In this edited article first published in the Daily Friend on 19th January 2024, Ivo Vegter looks beyond the hype and concludes it is not very impressive.
A solar plant complex with battery storage has come online in the deep Karoo. It isn’t very impressive.
The site at Kenhardt comprises three vast solar photovoltaic fields with a total nominal capacity of 540 MW.
Its battery storage can nominally deliver up to 225 MW of power.
In practice, what this all adds up to is a 20-year contract with Eskom to supply 150 MW of dispatchable power consistently between 05:00 and 21:30 throughout the year. ‘Dispatchable’ means that it can be switched on or off, or ramped up or down, depending upon demand.
The project cost is in the region of $1 billion, which at the time of writing amounted to roughly R19 billion.
Project developer, Scatec describes this project as ‘innovative and large-scale’.
It certainly is ‘large-scale’ in terms of land use. The company reports that it ‘[spans] 879 hectares and [measures] 10 km north to south’. This sounds inconsistent since 879 hectares is 8.79 square kilometres, which would make the entire thing only 879 metres wide from east to west, but in terms of land use, this almost exactly matches the 883-hectare site on which the Medupi coal-fired power station was built.
So, let the comparisons begin.
Medupi, a 4 800MW six-unit monster, produces 5.4MW per hectare. Kenhardt achieves 0.17MW/ha, so per unit of land, Medupi is 32 times more efficient.
Medupi was originally budgeted to cost R80 billion, but since that was in 2007, and both inflation and project costs have since escalated, I’m going to use the most recent and the highest cost-to-completion figure I can find: R234-billion.
At that price, the cost-overrun disaster of Medupi comes in at R49-million per megawatt, while the Kenhardt plant costs R127-million per megawatt. The solar-battery-hybrid is, therefore, 2.5 times more expensive than a grossly overpriced coal-fired power station.
Taking into account Medupi’s energy availability factor of about 70% and Kenhardt’s inability to produce electricity 24 hours a day does not change this calculation by much, and not in Kenhardt’s favour, anyway.
A coal-fired power station does, of course, need coal, which balances the scales somewhat in favour of the solar plant, but 1.5 times the price of Medupi buys a hell of a lot of coal.
On the downside for the Kenhardt plant, Medupi will last 50 years. The typical lifespan of solar photovoltaic panels is 20 years.
Medupi is built on the coal fields of the Waterberg, where existing grid infrastructure is abundant. Kenhardt is a tiny town in the deep Karoo. The Scatec plant itself is about 91 km due south of Upington, which is the closest town with more than 50 000 people. It is almost 600 km from Cape Town, as the crow flies, and about 750km from Johannesburg.
The length of high-voltage connections to the grid, the rights-of-way, and the line losses, all favour power plants other than this solar installation.
To match a single large coal-fired power station, we would need 32 plants such as the Scatec plant at Kenhardt, most of which will be situated in remote locations far from available grid connections, to take advantage of the high levels of solar irradiation of hot, arid regions.
Of course, coal-fired power stations are heavily polluting, and they’re also controversial over their contribution to carbon dioxide emissions.
That doesn’t mean, however, that solar plus battery plants are an adequate substitute. With the short projected lifespan of the technologies involved, the permanently toxic waste of panels and batteries will also mount as the decades pass.
A few dozen 150 MW plants are not going to displace coal in South Africa.