How to repair submarine cables: "The biggest effort is not in the mountains"

Jonas Franken from TU Darmstadt is investigating the resilience of critical information infrastructures on land and at sea. He explains the individual steps involved in repairing damaged submarine cables, as was recently the case in the Baltic Sea.

Submarine cables are damaged time and again, most recently at the end of January between Latvia and Sweden. How can these cables be repaired? We spoke to Jonas Franken, research assistant and doctoral student at the Chair of Science and Technology for Peace and Security (PEASEC) at TU Darmstadt, about this. His work there focuses on the resilience of critical information infrastructures on land and at sea. He previously sailed at sea with the navy and was involved in civilian sea rescue for many years.

MIT Technology Review: When damage to a submarine cable is discovered – what happens then?

Jonas Franken: This is a process that has been established for a very long time. Even on land, it is possible to estimate between which repeaters the damage will occur. Backscattering is used for this: damage to optical fibers scatters back some of the light. Such repeaters are installed every 60 to 80 kilometers.

What is the more common scenario: that a cable is completely severed or just kinked?

In terms of numbers, so-called shunt faults are somewhat more common: the electrical conductors are exposed, but the optical fibers are still undamaged. This can happen, for example, due to trailing cables. Incidentally, shunt faults do not necessarily mean that a cable fails immediately. Problems often only arise a long time later. In the case of anchor damage, however, it is usually the case that a cable breaks completely.

How exactly can the damage be localized?

If current is passed through the cable, it generates an electromagnetic field that can be measured by the sensors on a repair ship. If the signals change abruptly, then it is clear: Okay, this is where the damage is. If that's not enough, a sonar can be used, for example, to check whether there are fresh traces of abrasion on the seabed.

How are the damaged cables recovered?

Once the position of the damage is roughly known, a certain pattern is followed up and down, dragging a so-called grapnel across the seabed. This grapnel can be imagined as a mixture of grappling hook, sledge and plow. When the grapnel is pulled, you can tell that it has taken hold. Then you pull the cable on board.

Aren't there quite a few cables lying around at certain neuralgic points, for example at straits or harbor exits? How do I know I haven't grabbed the wrong one?

That can certainly happen in very narrow areas. But normally you look at the nautical charts and see what other infrastructure is nearby. And then you just sail a correspondingly narrower grid. Splicing takes the longest.

What happens once the cable is on board?

The end that you pull up is probably frayed or kinked. Parts of it have to be cut off. This can be several hundred meters. The remaining "good end" is sealed and hung on a buoy because the other end still has to be recovered. In the event of a shunt fault, the cable is cut completely and also shortened to the good ends.

Then a replacement cable is spliced on: all the glass fibers are connected together and the entire sheathing layers are applied again. The joint is then a little thicker than the original cable. Then move to the other end of the cable and splice it to the replacement cable as well. The piece of cable used is approximately two and a half times the length of the sea depth at the damaged area. It is laid on the seabed in a large loop to reduce the mechanical tension.

How long does a repair like this take?

About one to two days, not including the repair ship's journey. The biggest effort is not in recovering the cable itself, but in splicing it. For old cables with ten optical fibers, this is of course faster than for modern cables with 144 fibers. There is a special room on board for this, which is sealed so that no dust can get in. And you need a very well-trained team. Specialist personnel are the critical point.

"The classic danger zone is above 500 meters."

Up to what depth is a repair possible?

Theoretically, human interference is expected up to 1500 meters, and the longest trawls reach about that deep. But that is the exception rather than the rule. The classic danger zone is above 500 meters. There is a lot of human influence there, but also water currents that can cause abrasion. Most repair ships are designed for this depth.

How many of these ships are there – and how long does it take for them to arrive on site?

There are around 60 to 70 ships worldwide that are theoretically suitable for cable repairs. However, most of them are actually cable layers. There are around 25 to 30 special ships for repair and maintenance.

How quickly a repair ship is on site also depends on the maintenance contract. There is a kind of repair cooperative to which cable operators pay a certain annual fee for every kilometer of their submarine cables. However, if the ships in this cooperative are all busy, an operator may have to wait until another ship becomes available. A private maintenance contract is quicker, but also more expensive.

Geographically, the situation is worst in the South Pacific. Repairs can take two to three weeks there. However, repairs often take much longer because the approval process is so slow. There are actually countries that have an average processing time of six weeks.

"The really key point is redundancy."

Have the recent incidents in the Baltic Sea caused a rethink when it comes to repairs?

That depends on who you ask. The industry says, and from their perspective this is true: we are an infrastructure for peace, and in peacetime we work perfectly. There weren't very many cables that broke, and not many that were particularly relevant, with the possible exception of C-Lion between Germany and Finland. There are around 200 cable failures worldwide every year, some of which can be repaired at the same time.

On the other hand, political players and institutions such as the EU and NATO say: What happens in the event of a conflict? If we actually assume that many things will fail in a coordinated manner in the future, then the repair capacities will quickly be overwhelmed. They simply think in different scenarios than industry. In the First and Second World Wars, the international telegraph lines were one of the first infrastructures to be attacked. And that will probably happen again in an open conflict between major powers.

What are the most common causes of 200 damage to data cables?

Fishing and anchors account for around 70 percent, followed by seaquakes, landslides, dredging damage or wear and tear.

Given these figures, do you think all the excitement surrounding the latest incidents is exaggerated?

That is again a question of perspective. The industry's position is – understandable –: everything can be compensated for in the grids, for example via land connections. The political players are also right to ask themselves –: Which players could be behind this? Ultimately, we have to live with the fact that malicious actors exploit this uncertainty. I think a certain amount of caution and sensitivity is appropriate. But I found many political statements, especially from Germany, a little too alarmist.

How likely do you think it is that the recent cable damage in the Baltic Sea was caused accidentally – along the lines of: Oops, we lost our anchor and only noticed it after we had dragged it 150 kilometers behind us?

In fact, in most cases the anchor was dropped at night. The ships of the Russian shadow fleet in particular often have no warning devices. Something like that can happen. There has always been bad seamanship.

But when the same type of damage happens repeatedly over time to a similar class of ships – i.e. anchor damage to ships coming from or going to a Russian port, some of which belong to the Russian shadow fleet –, then this is very noticeable. This pattern is well above the usual level. I would say that there is already a relatively high probability that this is happening in some kind of coordinated way.

How can the cables be better protected?

I think the approach adopted for the Baltic Sea, especially at NATO level, is a bit one-sided. There are now F35 fighter jets flying around. Nobody has yet been able to explain to me what added value they offer compared to other maritime patrol aircraft for protecting submarine cables.

Submarine cable protection means much more: shore stations, repair ships, cable depots, software and so on. NATO can't guard these things somewhere in a media-effective way.

The really central point is redundancy. In the case of new cables, for example, the operators agree with operators of similar routes: If something fails on our end, we can route something through on yours and vice versa. This is how the industry has been doing it for ages and it ensures that the services usually run.

Then it's about the architecture itself. This is always a question of cost, so the industry naturally weighs up: Is it worth burying a cable? How heavily do we reinforce this cable? One thing is clear: you can't build a fence around submarine cables or lay them in concrete ducts for thousands of kilometers.

Part of prevention is also deterrence. Acoustic sensors in the cables can, for example, determine exactly how long an anchor has been dragged along. Together with other data from satellites, the AIS ship reporting system, cameras on bridges or radar stations on land, it is possible to pinpoint the culprits very precisely. In an open conflict, however, even the best sensor technology is useless - the cable is still broken.

Another factor is recovery. This also includes getting the permits faster.

Adaptation is about diversifying in the long term. If you look at a nautical chart of the Mediterranean, for example: Almost every cable lands in Marseille, or runs right across Egypt through the Suez Canal. This is of course incredibly cost-efficient. But people are now saying that the routes should also be diversified. And finally, you should also have a few strategies up your sleeve in case the internet does go down.

This article first appeared on t3n.de .

(kbe)