Summary

  • Geoff Huston’s APNIC-hosted essay cites a 2025 model that puts the 2040 cost of replacing 15 repair vessels and adding five at about US$3 billion; it is a scenario, not a current appropriation or purchase order.
  • The ITU/IAB report’s YZ case study says 15-year maintenance contracts supported vessel financing, while five-year terms weakened lender confidence. The report does not identify a funded fleet plan or a party committed to paying the modeled total.

The number is a scenario

Geoff Huston’s 14 October essay on the APNIC Blog cites a TeleGeography and Infra-Analytics model, finalized in 2025, of what cable maintenance could require by 2040. It assumes cable kilometres increase about 48% and roughly two-thirds of maintenance vessels reach the end of service. The model estimates about US$3 billion for 15 replacement ships and five additional vessels to maintain current service levels as the network expands.

The estimate combines renewal and growth. It is not an announced US$3 billion funding request, an approved budget or evidence that 20 ships have been ordered. The APNIC page is a signed essay and says the views belong to its author; it is not an APNIC procurement decision.

A fleet count does not name a buyer

The July 2026 report of the International Telecommunication Union’s International Advisory Body on Submarine Cable Resilience provides a separate inventory. It describes about 70 vessels with technical capability to repair cable, but says many primarily lay cable or are no longer actively maintained. Only around 20–25 are formally dedicated under active maintenance agreements or long-term charters. The counts use different definitions and should not be read as a contradiction.

For the investment case, the distinction is consequential: a technically capable ship is not necessarily backed by a revenue contract that supports its financing and upkeep. The report estimates a purpose-built repair vessel at US$100–150 million and says ageing ships lose 15–20% of annual availability to dry-dock work. Replacement therefore competes with the recurring cost of keeping the existing fleet serviceable.

The report’s YZ case study links contract length to capital access. It says 15-year maintenance agreements supported financing and newbuilds; when terms shortened to five years, financial institutions became less willing to fund vessels and upgrades. That is evidence from one arrangement, not a rule that every vessel requires a 15-year contract. It does show why a fleet forecast cannot be evaluated only as a ship count: lenders also need to see who has promised to buy maintenance capacity, for how long and on what terms.

Recommendations are not committed capital

The ITU/IAB report recommends longer maintenance contracts, jointly funded or multi-contracted ships at critical hubs, coverage-gap mapping and public–private maintenance hubs in underserved regions. Those proposals do not identify buyers, financing terms, procurement schedules or appropriated funds. The forecast remains a modeled requirement until owners and customers attach it to contracts and funded projects.

The central unanswered question is how the projected bill would be divided. A replacement vessel preserves existing capacity; an additional ship expands it. Those choices distribute costs differently among cable owners, maintenance-zone members and public partners. The report supports discussing those mechanisms, but does not show that any has been agreed.

Sources