How vulnerable are undersea cables?

In late 2024 and early 2025, several important data connections across the Baltic Sea were broken, including between Sweden and Lithuania. The cable breaks received a lot of attention and raised questions about the resiliency of undersea cables.

Marija Furdek Prekratic, associate professor at Chalmers and researcher in optical networks, says that it is difficult to overestimate how important submarine cables are for the internet.

“They provide strategically important connections between different continents and carry global communications traffic at very high speeds, high reliability, short response times and low cost. These networks handle an estimated 95 percent of intercontinental global Internet traffic and 99 percent of transoceanic digital communications,” Furdek Prekratic says. “As traffic continues to grow globally, the importance of submarine cables for global communications also increases. There is growing concern about the resilience of cable networks, and various stakeholders have begun to call for action.”

Hans Liwång, professor and researcher at the Swedish National Defense University, believes that the uproar over the cable breaks in the Baltic Sea has been exaggerated and may actually have done more harm than good.

“The suspected sabotage in the Baltic Sea is not the biggest threat to Swedish communication security or to our ability to access the internet. This method is too crude to be a major problem. We have good redundancy, and the repair capacity for communication cables is also good in the Baltic Sea, so the effect of this damage, both short-term and long-term, is very limited,” he says.

[[Read more: Google unveils $1B plan for subsea cables to Japan]

Liwång points out that cable breaks are something that has always occurred, and he does not believe that the operators of the cables have been naive and have not taken the risks seriously.

“Cables break all the time, usually as a result of accidents. We can’t stop that. The system solves it by having redundancy and repair capacity. A few possible sabotages don’t change that situation. The British’s first action in the First World War was to fish out and cut off German communication cables. This is not new and the risks are taken into account. By and large, the industry has made a reasonable risk assessment.”

Other parts of the Nordic region, such as Iceland and Svalbard, are more vulnerable: “They have lower redundancy, repair times are longer and the number of satellites is more limited, so there are fewer emergency options,” says Liwång.

Another region where cable breaks can have major consequences is the Red Sea. Seventeen cables pass through the 30-mile-wide Bab al-Mandab Strait between Djibouti and Yemen. In the spring of 2024, four Red Sea cables broke in a short period of time.

“The Red Sea cables provide a very important high-bandwidth, low-latency link between Europe and Asia. When cables were recently cut there, 25–70 percent of all traffic between the continents was affected,” says Furdek Prekratic.

Along the eastern side of the sea, fighting is ongoing between Houthi rebels and Yemeni government forces. The rebels have attacked ships linked to Israel. They have not claimed responsibility for the cable breaks, but that has not stopped speculation. Whether the cables were cut or damaged accidentally complicates the repair of the security situation in the region.

More generally, it can be difficult to base your decisions on maps of submarine cables, says Furdek Prekratic. There are areas with few such cables that still have very good connections to surrounding countries via cables on land, which are buried and better protected.

For Sweden, Hans Liwång does not believe that cable sabotage is what we should focus on: “The cables are vulnerable, but that doesn’t mean our internet access is vulnerable. There is no need to protect every cable, we need to ensure that the function is protected. That is by making sure we have good redundancy.”

The only way to effectively protect a cable against sabotage is to bury the entire cable, says Liwång, which is not economically justifiable. In the Baltic Sea, it is easier and more sensible to repair the cables when they break, and it is more important to lay more cables than to try to protect a few.

Burying all transoceanic cables is hardly feasible in practice either. In the spring of 2024, The Verge published a longer report from the Japanese cable ship KDDI Ocean Link, whose crew, among other things, talked about their experiences after the major earthquake off Fukushima in 2011. They carried out 11 out of 20 repairs after the quake and the deepest cable was at a depth of 6,200 meters.

Remote islands like Iceland have greater challenges, says Liwång, and may need to think more about other types of redundancy than laying many expensive cables. “An alternative for Iceland is to use satellites instead, but this will greatly affect capacity for areas near the poles. Satellite capacity is poor there, which makes the challenge extra great for Iceland.”

Designed to handle interruptions

Built-in redundancy means internet traffic can take many paths to get from point A to point B. If the shortest path suddenly becomes unusable, for example after a cable break, the routers through which the traffic passes can find an alternative path.

“Cable breaks are relatively common even under normal circumstances. In terrestrial networks, they can be caused by various factors, such as excavators working near the fiber installation and accidentally cutting it. In submarine cables, cuts can occur, for example due to irresponsible use of anchors, as we have seen in recent reports,” says Furdek Prekratic.

Network operators ensure that individual cable breaks do not lead to widespread disruptions, she notes: “Optical fiber networks rely on two main mechanisms to handle such events without causing a noticeable disruption to public transport. The first is called protection. The moment an optical connection is established over a physical path between two endpoints, resources are also allocated to another connection that takes a completely different path between the same endpoints. If a failure occurs on any link along the primary path, the transmission quickly switches to the secondary path. The second mechanism is called failover. Here, the secondary path is not reserved in advance, but is determined after the primary path has suffered a failure.”

Operators typically combine these two mechanisms to ensure access to a backup path via protection and provide the opportunity for additional flexibility via recovery, says Furdek Prekratic.

“All of this is implemented in the optical layer that transports aggregated traffic from the upper network layers, so that the protocols in the upper network layers do not even perceive the change in the underlying topology. This speeds up recovery without requiring updates in, for example, routing tables.”

New technology, old craftsmanship

The world’s first underseas cable, or submarine cable as it is also called, was laid between England and France in 1851. For many years, these cables were used only for telegraphy. The first transatlantic cable for telephony, for example, was put into service as late as 1956. It was then just over 30 years before the first fiber-optic cable crossed the Atlantic, TAT-8.

That cable had a total capacity of 280 megabits per second – a speed that is today surpassed by a normal home broadband connection.

Furdek Prekratic says that the fastest cable today, MAREA, has a total capacity of 200 terabits per second – 700,000 times faster than TAT-8. That may sound like a lot, but if all Swedish households tried to stream high-definition video from a server in the US at the same time, it would require a lot of such cables.

“Ambitious plans for groundbreaking projects were recently announced: 2Africa, which will connect three continents with a total of 46,000 kilometers of cable, and Project Waterworth from Meta, which will connect five continents with a large submarine cable network totaling 50,000 kilometers. It will be very exciting to see what capacity is achieved in these new networks.”

The craft of laying cables and building them hasn’t changed at all. Atlantic Cable Maintenance & Repair Agreement chairman Alasdair Wilkie told The Verge that the technology is essentially the same as it was 150 years ago.

Initially, gutta-percha (a filling material derived from certain trees) was used as insulation around the cable, but now polyethylene is used around steel wire and several other materials that protect the optical fibers.

The cable is coiled on large drums in the cable ships that lay it. Relays are placed at regular intervals to ensure that the signal does not weaken too much over the extremely long distances. A thinner cable is used in deep water and a significantly thicker one near land where the risk of damage to the cable is greater.

Sweden’s only cable installer

Today there is only one Swedish-flagged cable ship: C/S Pleijel, named after a former submarine cable expert at Televerket named Henning Pleijel. The ship was built in Denmark in 1972 but underwent a major rebuild in 2015, including becoming 13 meters longer to accommodate more cable. The magazine Sjömannen did a report from the ship in 2017, which you can read here.

This article originally appeared in PC för Alla