Durban's port is under real strain. Its cargo is not yet leaving the country.

By late October 2026, trade reporting suggested that shippers were turning from Durban to Walvis Bay, Dar es Salaam, Mombasa and air freight while the city's largest container terminal worked through a backlog that began on 30 June.

On the public record, no diversion of cargo has been measured at any level.

That is narrower than it sounds. It means the evidence for an exodus is forecasts, narration and price signals, not volumes, and it says nothing about what has not been published.

We asked a plain question of the claim: how exposed is Durban's gateway role to competing ports and routes, and which cargo could actually move? We traced carrier advisories, terminal notices, the weekly cargo-movement series, results filings and earlier disruptions back to their published sources. We then separated four kinds of move: between terminals inside Durban, between South African ports, to a foreign port, and from sea to air. Each segment was tested against five stages, from disruption through to durable loss.

The answer turns on geography. Maputo is the only foreign gateway that reaches Gauteng at all, and its published container capacity is 170k–270k TEU, against 2,650,419 TEU through Durban Container Terminal in 2024. No foreign port serves KwaZulu-Natal end to end. Gauteng and KwaZulu-Natal cargo is captive nationally in the near term. It is contestable between South African ports, at known penalties: Ngqura and Port Elizabeth add about 12 and 23 planned rail hours to Gauteng.

The strain is real

Rail evacuation from the affected terminal fell to about 233 boxes a day in mid-September, against 477 before the crisis. Carriers have priced against Durban. One Durban-wide congestion fee stands at USD450 per 20ft and USD900 per 40ft from 1 November, with no end date or trigger. The episode, at about 18 weeks, has already outlasted earlier Durban disruptions after which cargo came back.

Two things surprised us

Our view, for what it is worth

The risk to Durban is less a shipper's decision than a carrier's. The exposure we would watch is a network decision for 2027 taken while the terminal is still congested, because repeated episodes are the condition most often present before a comparable gateway lost share for good. That is a working hypothesis, not a finding. Durable loss can only be seen after the terminal recovers, so the test comes one to two quarters after normalisation, not now.

These are questions, not advice. The decision stays with the reader, and if the public record misses something you know, we would like to hear it.