keepcalculator

06 Numbers and Powers

How Undersea Fibre Carries Internet

Card No. 06·62418 · Updated · Standard formula, shown below

This guide explains the physical path your internet data takes across oceans: how undersea fibre-optic cables are built, laid, and operated to carry global traffic. You’ll learn the key components, the step-by-step process from shore to shore, and what to do when things go wrong. No special equipment is needed—just curiosity about how the internet physically works.

What You’ll Need

  • A basic understanding of light and electricity (not required, but helpful)
  • Access to a map of submarine cables (e.g., TeleGeography’s Submarine Cable Map) to follow along
  • Patience—this is a complex process, but we’ll break it down

Step-by-Step

  1. Understand the core: light pulses in glass. Undersea cables contain hair-thin strands of glass called optical fibres. Data is sent as pulses of laser light, typically at wavelengths around 1550 nanometres (infrared). Each pulse represents a binary 1, and the absence of a pulse is a 0. The light travels through the glass by total internal reflection—the fibre’s core has a higher refractive index than the cladding around it, so light bounces along the core without escaping.
  2. Know the cable’s physical layers. A modern submarine cable is not just bare fibre. From the inside out: the core has multiple fibre pairs (often 8 to 16 pairs), each pair carrying separate signals. Around the fibres is a buffer tube, then a strength layer of steel wires or Kevlar to handle the tension during laying and recovery. Then comes a copper or aluminium tube that carries electrical power for repeaters. Finally, an outer jacket of polyethylene (plastic) protects against water and pressure. In shallow water (less than about 1,000 metres), additional steel armour layers are added to protect against ship anchors and fishing trawlers.
  3. Plan the route. Cable systems are not laid in a straight line. Route planners must avoid underwater mountains, volcanic activity, and fishing grounds. They also consider political boundaries and landing points. The route is surveyed using sonar and satellite data to map the seafloor. A typical transatlantic cable might be 6,000 km long, but the actual fibre path is longer because of slack and detours.
  4. Build the shore station and landing. The cable starts at a cable landing station on land. This building houses the terminal equipment: lasers, receivers, and multiplexers that convert electrical data signals from the internet backbone into light signals. The cable is buried in a trench from the station to the low-water mark, then laid along the seafloor. In shallow water, it’s usually buried under the seabed using a water jet or plough to protect it.
  5. Lay the cable from a ship. A specialised cable-laying ship carries the cable in large tanks, coiled in a figure-eight to avoid kinks. The ship moves slowly (about 5-10 knots) and pays out the cable over the stern or bow. The cable is tensioned carefully to avoid slack that could snag, but not too tight to cause stress. In deep water, the cable is simply laid on the seafloor; in shallower areas, it’s buried. The ship uses dynamic positioning to stay on course.
  6. Install repeaters (amplifiers) every 50-100 km. Light attenuates (weakens) as it travels through glass. To keep the signal strong, repeaters are spliced into the cable at intervals. A repeater is an optical amplifier that boosts the light signal without converting it to electrical. It uses a pump laser and a short length of erbium-doped fibre to amplify the signal. These repeaters are powered by a high-voltage direct current (typically 3,000-15,000 volts) sent along the copper conductor from the shore stations. The voltage is dangerous, so only trained personnel handle it.
  7. Splice and test. When the ship reaches the other shore, the cable is brought to the landing station. The fibres are spliced together using a fusion splicer that melts the glass ends. Each splice is tested with an optical time-domain reflectometer (OTDR) to ensure low loss. The system is then connected to the terminal equipment on both ends.
  8. Turn on the light and transmit data. Once the cable is connected, the shore stations send test signals. After verification, the system is handed over to the network operators. Data now flows as light pulses. To carry more data, modern systems use dense wavelength division multiplexing (DWDM): multiple lasers at different colours (wavelengths) are sent down the same fibre simultaneously. Each wavelength can carry 100 Gbps or more, and a single fibre pair can carry hundreds of wavelengths. Thus, a cable with 8 fibre pairs can have a total capacity of several terabits per second.
  9. Monitor and maintain. After installation, the cable is monitored 24/7 from network operations centres. They track signal power, bit error rates, and electrical current. If a fault occurs, they use OTDR to locate the break within a few metres. A repair ship is dispatched to the location, grapples the cable, brings it to the surface, and splices in a new section. Repairs can take days to weeks, depending on weather and depth.

Common Mistakes

  • Thinking the cable is a single fibre. In reality, it contains many fibre pairs, each capable of carrying many wavelengths.
  • Assuming data travels as electrical signals. It’s light, not electricity, that travels through the glass. Electricity is only used to power repeaters.
  • Believing the cable is laid straight. It has slack and is laid in a path that avoids hazards; it’s not a straight line.
  • Underestimating the power. The voltage used to power repeaters is lethal. Never open a cable or touch the copper conductor.
  • Ignoring the difference between shallow and deep water. Shallow water requires burial and armour; deep water doesn’t.

Tips

  • Use a cable map to visualise. Sites like TeleGeography’s map show real cable routes and landing points.
  • Remember the speed of light in glass. Light travels about 200,000 km/s in glass, so a transatlantic cable has a latency of roughly 30 ms one-way.
  • Think of repeaters as amplifiers, not regenerators. They boost the light, but they don’t decode the data.
  • Note that newer cables use coherent optics to encode data in phase and amplitude, not just on/off pulses.
  • Understand that capacity is not unlimited. Each wavelength has a maximum bit rate, and the number of wavelengths is limited by the amplifier bandwidth.

Troubleshooting

If your internet is slow or down, it’s rarely due to a submarine cable fault—most issues are local. But if you suspect an undersea cable issue:

  • Check cable status. Many operators publish status pages or use cable maps that show outages. If a major cable is cut, you’ll see reports in the news.
  • Run a traceroute. Use a tool like traceroute to see where packets stop. If they stop at a landing station, the issue might be on the cable.
  • Contact your ISP. They have monitoring tools and can tell you if the issue is on their network or upstream.
  • If you’re a technician, never attempt to repair a cable without proper training. The voltage is lethal, and the splicing requires precision.
  • For home users, remember that most outages are due to Wi-Fi, router, or local ISP issues. Reboot your router before assuming a cable fault.

Frequently Asked Questions

Q: How deep are undersea cables?
A: In deep ocean, they lie on the seafloor, which can be 8,000 metres deep. In shallow water, they are typically buried 0.5 to 3 metres under the seabed.

Q: How many undersea cables are there?
A: As of 2025, there are over 500 active submarine cable systems worldwide, with more being planned.

Q: Can sharks bite through cables?
A: Shark bites have occurred, but they are rare and usually not the cause of major outages. Most damage comes from fishing trawlers and anchors.

Q: How long does it take to repair a broken cable?
A: It can take from a few days to several weeks, depending on depth, weather, and the availability of a repair ship.

Q: Do undersea cables carry all internet traffic?
A: Most intercontinental traffic goes through cables, but a small percentage (about 1%) uses satellites, primarily for remote areas or as backup.

Q: Why don’t they use satellites instead?
A: Cables offer much higher capacity and lower latency. Satellites have higher latency (geostationary: 600 ms) and limited bandwidth.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.

More cards in this drawer

The whole Numbers and Powers drawer →