Summary

  • ReuNION has a published €93.1 million funding plan and a 16-pair design, but the public record does not yet show how a non-consortium operator would obtain capacity across the landing station and terrestrial handoff.
  • The 12/2/1/1 pair allocation is a planned consortium benefit, not proof of ownership or exclusion; the operating access rule is the missing evidence.

Réunion’s next cable has moved beyond an announcement. A regional project page sets out the money, the participants’ planned fibre allocation and a timetable; the European Commission’s Funding & Tenders Portal lists ReuNION as an ongoing project with a €20 million EU contribution. Orange says it will be the landing party, responsible for permits, civil works, a secure station and connection to the island’s terrestrial infrastructure. Orange Marine is to lay the cable, while Alcatel Submarine Networks is named for design and manufacture.

Those roles matter because “access” is not delivered by a fibre pair alone. The published architecture assigns 12 of 16 pairs to La Réunion Connectée, two to Orange, one to Telco OI and one to Réunicable. The project promoter describes each member as benefiting from dedicated pairs. Orange, meanwhile, says the landing link will make the cable accessible to all operators and users. Both statements can be true: the allocation can describe consortium capacity rights while third parties buy a service. But the public material does not specify how that service works.

The €93.1 million plan is unusually legible: €30 million in ERDF funding, €20 million from the Connecting Europe Facility, €6.5 million from the French State and €36.6 million from the operator consortium. The Commission portal separately lists €20 million of EU contribution, including participant allocations. A published plan and an active grant record establish a financing framework; they do not establish that every contribution has been paid, that construction milestones are complete or that wholesale access conditions are settled.

The physical handoff is the commercial control point. A cable may reach the shore while an operator still needs station space, optical cross-connects, power, monitoring, maintenance windows and terrestrial backhaul to reach a usable network. Orange’s landing-party duties explicitly include the station and interconnection. The public claim of universal accessibility therefore needs an operational counterpart: who receives requests, which technical interface is offered, what the price basis is, how capacity is assigned, and what happens when a fault crosses the submarine and terrestrial segments.

Réunion’s regulatory history makes these questions concrete, but not predetermined. In a 2014 market analysis, ARCEP described Orange as owning the landing stations for SAFE and LION 1 and said alternative operators had to buy terrestrial-extension or optical-deport services. In 2004, the regulator required wholesale transport and IP-transit offers over SAT3/WASC/SAFE after operators and the Regional Council raised disputes. These are precedents on older cables, not evidence about ReuNION’s ownership, present obligations or future prices.

They show why the landing-to-backhaul boundary deserves explicit treatment before a new system is operational.

The risk case also needs careful attribution. La Réunion Connectée says SAFE is scheduled to stop in 2027 and cites the 2023 combination of a SAT3 outage and LION2 maintenance as a practical warning. The Commission project page uses a relative end-of-life horizon, while Orange says some regional systems are nearing the end of operation. Those statements establish the project’s stated rationale, not an independently verified SAFE retirement notice.

Nor does a route to South Africa by itself prove end-to-end diversity: landing sites, terrestrial paths, power, upstream transit and repair arrangements determine whether a second cable survives the same failure.

The next public evidence should be more specific than another resilience claim. A final route map, station and asset responsibilities, access reference offer, capacity-allocation method, fault and service-level rules, and grant-linked milestones would let operators and funders test whether the physical project can become a usable, contestable service. Until then, ReuNION’s capital plan is more visible than its access regime.

Sources: Orange project announcement; La Réunion Connectée project, funding and schedule; European Commission project 101241199; Réunion Regional Council deliberation record; ARCEP 2014 market analysis; ARCEP 2004 ruling summary.