Why Undersea Cable Repairs Take Weeks—And What That Means for You

When an undersea cable breaks, the internet doesn’t stop. Traffic reroutes. But for the regions directly served by that cable, latency spikes, capacity drops, and some services slow or fail. A single cut can affect millions of users for weeks. Repairs take time—often three to five weeks, sometimes longer. Why? And what does that mean for how we think about the internet?

Undersea cables carry roughly 95 percent of intercontinental internet traffic. Satellites get the headlines, but they handle a tiny fraction of global data. The real backbone is fiber optic cables—thousands of miles of them, laid across ocean floors. They’re armored against pressure and abrasion, buried under the seabed where possible, and monitored for faults. They’re also fragile: a single fishing net drag, a ship anchor, or an earthquake can sever one. Fixing it requires a specialized ship, precise location data, and often a long voyage to a remote stretch of ocean.

Why repairs take so long

First, you have to find the fault. Cable operators monitor traffic and signal loss to pinpoint where a break occurred. That can narrow it down to a few kilometers, but the ocean is vast and visibility is zero. The repair ship deploys a grapnel—a weighted hook—and drags it along the cable route until it snags the severed end. That can take days. Weather matters: storms, high seas, and poor visibility delay operations. The ship might have to wait for a weather window before it can work safely.

Once the cable is located, the ship hauls it aboard. Crews inspect the damage, cut out the broken section, and splice in a new length. Splicing fiber optic cable is precise work: the glass fibers must align to within microns. It’s done in a controlled environment on deck, often under a tent to protect from wind and spray. A single splice can take hours.

There aren’t many repair ships

Globally, there are only about 60 cable repair ships. They’re owned or chartered by cable consortia and contractors. They’re expensive to operate and maintain. They’re also spread thin: a ship based in Europe might be the nearest asset for a break off West Africa. Transit time alone can add a week or more. If a ship is already on another repair job, the queue lengthens.

Some regions are better served than others. The North Atlantic has dense cable coverage and multiple repair ships nearby. The Pacific, the Indian Ocean, and routes along less developed coasts have fewer assets. A break in the Red Sea or off Mauritius can mean longer waits and longer outages for the countries that depend on that cable.

What this means for you

If you’re in a major hub—London, New York, Singapore, Tokyo—a single cable break rarely affects you directly. Traffic routes around it. You might notice slightly higher latency to some regions, but services keep working. The internet is designed for redundancy at that scale.

If you’re in a country or region served by only one or two cables, a break can be serious. East Africa, island nations, and some parts of South Asia have relied on single cables. When one goes down, capacity drops sharply. International calls, video calls, and cloud services can degrade or fail. Repairs take weeks. During that time, local providers ration bandwidth and prioritise critical traffic.

For businesses and remote workers, the takeaway is: geographic redundancy matters. If your operations depend on low-latency access to a specific region, a cable break can disrupt you even if you’re not physically there. Choosing cloud regions, backup links, and redundant paths isn’t paranoia—it’s recognising that the physical internet has single points of failure.

The future of cable resilience

New cables are being laid every year. Routes are diversifying. The more paths between regions, the more resilient the network. But cables will keep breaking—fishing, anchors, geology, and occasionally sabotage. Repair capacity isn’t scaling as fast as cable deployment. That mismatch will matter when multiple breaks occur in quick succession.

Understanding why repairs take weeks helps set expectations. The internet feels instantaneous and virtual. But it runs on physical infrastructure—cables, ships, and crews—that operate in the real world, with real constraints. When a cable breaks, the fix is slow because the ocean is big, the ships are few, and the work is hard. That’s not changing anytime soon.