ANTHONY BUTLER | Baseload power myth lingers as energy system shifts

Coal-era thinking persists despite growing need for storage, flexible grids and modern electricity planning

First published in Business Day

August 07 2026

The idea of baseload energy seems destined to live on in SA even as the age of baseload power draws to a close.

“Baseload” originally referred to the electricity needed as a floor even at the period of lowest demand. It was then used loosely to describe the supply that meets that floor: traditionally, big plants that run continuously at constant output because they’re cheap to run once built and expensive to ramp up or down. In SA, this meant coal, with a little bit of nuclear.

Baseload is fast becoming an outdated idea in the renewables-heavy grids that are proliferating around the world, where the focus has shifted to the challenge of “intermittency”: simply put, what to do when the sun doesn’t shine and the wind doesn’t blow.

This is less of a problem than it might seem. After all, wind tends to blow more at night and in winter, whereas solar peaks at midday and in summer. Adding hydroelectric, geothermal and biomass renewables further smooths supply, as do transmission lines that let power flow from where it’s windy or sunny to where it isn’t.

Smart charging of electric vehicles, industrial processes that run when power is abundant, smart thermostats, and time-of-use pricing are becoming widely used.

Meanwhile batteries for short-duration energy storage, and hydro storage thatpumps water uphill when power is abundant and releases it through turbines when it’s needed, are becoming massively cheaper, and long-duration storage options, including compressed air and thermal storage, are under advanced development.

The emerging logic is that renewable supply is somewhat variable, so what matters is having demand response options and storage. There is the added advantage that renewables, once built, don’t carry fuel-supply geopolitical risks (though they do have concentrated supply chain dependencies for rare earth minerals, batteries, and panels).

This probably leaves a residual need forsome “dispatchable” power, in the form of gas peaker plants, new generation nuclear generators, or hydro plants with reservoirs, whose output can be turned up or down in response to grid operators’ needs, as a backup for low probability extended renewable lulls.

“Baseload” lives on in SA’s energy debate in part because coal still accounts for 80% of the country’s power needs. Eskom relies on 15 coal-fired power stations, so coal-as-baseload is still the actual operating model of the grid today, and not yet a historical relic.

Eskom, however, has dramatically highlighted the impending challenge the country faces as a result of ongoing coal fleet retirement. The parastatal is scheduled to retire 8.4 GW of coal-fired capacity over the next three or four years, and this will create a first significant “baseload cliff” in the national system by 2030, one that demands non-coal alternatives or risks new supply shortages.

Ministers meanwhile continue to use the language of “baseload power” – which they treat as synonymous with reliability and affordability – to cover their failures over investment in storage, the transmission grid, and end-user distribution.

There is open conflict over scarce transmission capacity between renewable developers. Until grid capacity scales up substantially and becomes better aligned with optimal renewable generation locations, a flexible dispatchable system cannot be built.

Meanwhile, decaying local distribution infrastructure and municipal financial mismanagement are bringing a new crisis of unreliable electricity access quite unrelated to the generation mix. Ministers continue to imply that the origins of these failures lie in environmentalists’ championing of coal plant decommissioning.

The unique configuration of the pro-coal baseload lobby in SA – combining as it does major party funding donors, ANC-aligned trade unions, and a politically connected and bloated parastatal staff complement – means the baseload idea is set to exercise disproportionate influence, despite the impending disappearance of the baseload model itself.

Butler teaches public policy at the University of Cape Town