The government has pitched offshore wind as a clean power source for Cape Town’s growing AI data centre industry — but the first project could not be commissioned before 2035, the laws needed to build it do not exist, and the power would cost roughly three times what onshore wind does today.
The Department of Electricity and Energy published a strategic framework for offshore wind on Thursday, identifying about 95GW of technical potential along South Africa’s 2,800km coastline. It singles out energy-hungry coastal users, data centres among them, as the customers that could justify the investment.
The numbers against municipal demand are striking. The framework identifies 6.4GW of floating offshore wind potential off Cape Town, against a peak municipal demand of roughly 2GW, and 17.4GW off Durban against eThekwini’s peak of about 1.7GW.
Cape Town, the report notes, “has already attracted a large data centre operator by guaranteeing access to clean energy, currently powered by the Northern Cape’s solar PV capacity.” With offshore generation added, the city could “develop as a clean energy data centre hub.”
A 2035 answer to a 2026 question
The difficulty is timing.
Under every scenario modelled, the first offshore wind project reaches commissioning in 2035 at the earliest, given development lead times. Data centre demand is arriving now. Cape Town approved two hyperscale facilities totalling around 174MW in July. McKinsey projects continental data centre demand could reach 2.2GW by 2030, roughly five times current levels.
Offshore wind cannot power that build-out. At best, it could power what comes after it.
That matters because power and water are now the centre of a live dispute. Cape Town advocacy group Housing Assembly and UK nonprofit Foxglove filed a legal challenge in August against two proposed Equinix data centres in the city, arguing they would compete with residents for scarce electricity and water, and are calling for a national moratorium on hyperscale development.
No legal route to the seabed
The framework is candid about the gaps.
South Africa has no mechanism for granting a developer exclusive rights to an area of ocean. “Site exclusivity regulations for OSW have not been enacted,” the report states, citing “a lack of legislation and regulation allowing OSW site leasing and exclusivity.”
The Marine Spatial Planning Act of 2018 does not explicitly address how offshore wind development areas should be designated. The national marine spatial plan is still being drafted. Offshore wind was left out of the National Coastal and Marine Spatial Biodiversity Plan entirely.
Responsibility has not been assigned either. The framework’s second recommendation asks the energy department and the Department of Forestry, Fisheries and the Environment to “determine and designate clear roles and responsibilities” for designating areas, granting exclusivity and leasing seabed. It notes regulatory changes in South Africa “have been known to take up to two years,” before public participation and parliamentary oversight.
Offshore wind also appears nowhere in national electricity planning. The draft Integrated Resource Plan 2023 “makes no specific reference to OSW,” and the framework’s first recommendation is to set explicit targets.
Energy Minister Kgosientsho Ramokgopa is careful about what the document is not. “This report is not a procurement plan, a policy determination, a capacity allocation or a commitment of public funds,” he writes in the foreword.
Deputy Minister Alexandra Abrahams, who launched it at the Africa Green Hydrogen Summit in Cape Town, was more bullish, saying investors “need transparency and a reasonable amount of predictability.”
The price premium
The cheapest fixed-foundation sites, off Saldanha Bay and Coega, are projected at $94 to $116 per megawatt-hour in 2035. Floating turbines — the overwhelming majority of the 95GW, because the continental shelf drops away sharply — start at $97 to $126.
Onshore wind procured in bid window 5 of the renewable energy independent power producer programme averaged $32 per megawatt-hour in 2021. The comparison is not exact — the offshore figures are modelled in 2023 dollars for 2035 delivery, while REIPPPP tariffs are inflation-indexed — but the gap is wide on any reading.
Costs fall materially only with volume. The framework models 5GW, 15GW and 40GW by 2050, supplying roughly 6%, 16% and 26% of national demand. Only the 40GW case — building 2.5GW to 3GW a year through the 2040s — brings floating wind down to $47 to $60 by 2050.
That is also the only scenario producing the headline benefit government has been citing: more than 300,000 full-time equivalent job-years and $6.7 billion in wages between 2030 and 2050. Job-years are not jobs, and the total accumulates over two decades.
The report concedes that offshore wind “may currently not be fully competitive with other renewable energy technologies in terms of price alone,” justifying it instead on industrialisation and employment.
Ports, grids and birds
Two further constraints sit behind cost. Only Saldanha Bay is close to ready, needing little beyond extra crane capacity. The medium and high scenarios require two to three ports each handling a gigawatt a year, implying quay extensions and dredging at Richards Bay or Durban, or the still-unbuilt Boegoebaai. Transmission upgrades are not costed at all; the framework defers that study to the National Transmission Company South Africa.
Marine protected areas, core habitat for the African penguin, Cape gannet and Cape cormorant, shipping lanes and a 10km coastal buffer are treated as hard exclusions. Fisheries, petroleum rights, subsea broadband cables and naval areas are flagged for further assessment — the cables being the same infrastructure that makes Cape Town attractive to data centre operators in the first place.
South Africa’s first proposed offshore project, the 800MW Gagasi wind farm off Richards Bay launched in 2022, has stalled. Its environmental authorisation application was withdrawn last year to redesign from floating to anchored turbines — “for technical and not environmental reasons,” according to Acer (Africa) Environmental Consultants MD Dieter Heinsohn.
The framework also assumes low social licence for anything visible from Cape Town, and calls for fair compensation for displaced fisheries.
Prepared by the Carbon Trust for the World Bank with the department, and funded by the Bank’s Energy Sector Management Assistance Programme, it is the ninth report in the Bank’s country-level offshore wind series.
None of this makes offshore wind implausible for South Africa. It means the 95GW describes a resource rather than a pipeline — and for the data centres being approved today, the power question will have to be answered some other way.v
