South Africa’s energy transition has a paradox at its centre that rewards precise reading: the closure schedules are real, and so is the money still being made. For anyone selling into coal, the question is not ideology but sequencing, which assets run hard for another decade, which are managed for cash, and which quietly disappear. The evidence gives clearer answers than the headlines do.
What does the closure schedule actually say?
Read the revealed preferences, not the plans. On paper: around four South African coal mine closures by 2025, eleven between 2026 and 2030, twelve more by 2035, from a base of roughly 108 operating mines. In practice, the one hard datapoint is that a single Eskom station, Komati, has been decommissioned since 2022, while the board pushed Camden, Grootvlei and Hendrina out to 2030 at an accepted extra cost of R85 to 90 billion, and the environment ministry licensed five stations to run at relaxed emission limits until March 2030. The asymmetry is the lesson: closure dates have slipped every time they met load-shedding or job politics. Extension decisions, once made, have held. Weight your planning accordingly.
Why did Komati matter so much?
Because it converted the just transition from a policy into a measurable result, and the result was poor. The showcase closure delivered job losses, falling household income and repurposing programmes that arrived late and small, documented enough that the Presidential Climate Commission itself flags the delays. The political consequence is mechanical: every subsequent closure decision now carries Komati as evidence for the opposition, which is a large part of why the 2030 extensions faced so little resistance. For sector reading, treat Komati as the reason the schedule’s error bars all point one way, toward later.
What are the miners doing with the cash-rich years?
Executing three distinct playbooks, each visible in the capital flows. Seriti, the largest producer at around 62 million tonnes a year, is becoming its own power utility: the 155 MW Ummbila Emoyeni wind build, online 2026, is stage one of a planned 900 MW cluster whose economics rest on displacing Eskom tariffs at its own operations, a hedge that pays regardless of coal policy. Exxaro bought growth outside coal entirely, putting about 640 million dollars into control of the Tshipi Borwa and Mokala manganese mines. Thungela runs the return-cash playbook, paying out while two mines supplying Sasol’s Secunda complex reach their planned end in 2026. What no one is doing is building major new coal capacity, which tells you the industry’s own terminal-value assumption even as current earnings stay strong.
How solid is the export leg?
Solid demand, constrained delivery, quantified precisely by the gap between two numbers: Richards Bay Coal Terminal’s 91 million tonne design capacity and its actual throughput, 52.1 million tonnes in 2024, recovering from 47.2 million the year before as Transnet improved. Around 55 million tonnes was budgeted for 2025. Every tonne of that roughly 35 to 40 million tonne shortfall is revenue the mines cannot access for want of rail, which is why producers co-fund locomotives and why private rail access is the single reform with the largest cash value to the sector. Asian buyers continue to absorb South African thermal coal as Europe fades. The market is not the constraint. Watch Transnet’s weekly rail run-rate as the sector’s true leading indicator.
What this means for suppliers
Price coal as a decade of well-funded maintenance demand with a defined tail. The buying centres: plant availability and efficiency work on the surviving fleet, rail and loadout debottlenecking anywhere on the RBCT corridor, water treatment as compliance tightens, and the miners’ own renewable builds, wind and solar packages procured by coal companies. Two growth niches with long runways: mine rehabilitation and closure engineering as the schedule advances, and the repurposing work at retired stations. The ground-level operational map is in our Mpumalanga guide. AMIQ tracks 2,200+ African mining projects with verified owner and engineer contacts, the coalfields included. Join at AMIQ, and see coal mining in South Africa for the operating picture.
Frequently asked questions
Where is coal mined in South Africa?
Overwhelmingly in Mpumalanga, source of about four fifths of national output, with the remainder mainly from Limpopo’s Waterberg and KwaZulu-Natal’s anthracite fields.
When will South Africa stop using coal?
Later than scheduled, on the evidence. One station has closed since 2022, three others were extended to 2030 at an accepted R85 to 90 billion cost, and mine closures are phased into the 2040s. Plan on coal dominating the grid for at least another decade.
Are new coal mines being built in South Africa?
No major ones. Producers are extending existing operations, spending on rail access and efficiency, and deploying surplus cash into renewables and other minerals, a capital pattern that prices coal as a strong but finite cash generator.
