I don’t think many realise just how much energy capacity SA needs to build. It is stupefyingly large.

Thanks to the affable Eskom spokesperson Sikonathi Mantshantsha’s Twitter feed, I can tell you that on Friday night the grid’s available capacity was 25.3GW. Demand was 30GW so the country was on level 4 load-shedding.

There are lots of estimates about how much demand will be in the future. The Integrated Resource Plan, the official national picture, believes 78GW of capacity is needed by 2030. The IRP was last updated in 2019 and the assumptions on which it was based are already outdated. But let’s work with that figure.

While Eskom’s fleet has nameplate capacity of about 52GW, half of that is not working, and that energy availability factor is unlikely to increase given the deteriorating performance of the older coal power fleet.

There is much debate about what kind of energy availability factor is reasonable for Eskom, but it is safe to assume the utility’s coal fleet is not going to miraculously turn around its performance, even if the sabotage and poor management of stations are fixed. This is not an Eskom failing — the energy availability factor of coal plants everywhere falls as they age, and it deteriorates particularly fast after about 50 years of life, and most Eskom plants are older.

The Eskom fleet is gradually being put out of its misery. By 2035, the utility expects it will have retired 12 of its 15 coal plants, representing 33GW of its capacity.

So, if you make optimistic assumptions about Eskom’s ability to get Medupi and Kusile fully operating, and extend the life of the Koeberg nuclear plant, we are going to need 50GW to 60GW of new power generating capacity to be built over the next eight to 12 years. That is not very far away. And the truly astounding number is how much investment will be required to deliver that. A lot depends on which technologies you use and pricing trends.

Solar photovoltaic and onshore wind are now much cheaper than fossil fuel production, but you need storage to even out supply capacity. Based on current global capital costs for different types of technologies, we need to invest R1.8-trillion to R3-trillion to build that capacity, depending on technology spread. And that doesn’t even consider the investment required to expand the grid to handle the volumes.

It is a staggering amount. It is between a quarter and half of GDP. It is up to half of the entire pot of savings in all pension funds. It is three to six times the total capital in the banking system. Of course, not all of that must be invested at once — but even if we do it over 10 years, the amount invested each year will be 5% of GDP, about one-third of annual total capital investment in SA. It is more than it cost to build Medupi, every year.

To date, the standard financing model has been for developers (large international utility companies) and private equity firms to put up the equity and banks to leverage that up about five times. But bank balance sheet capacity is going to be tested if new investment is accelerated. We will need considerable offshore funding, and more direct funding by institutional investors such as pension funds and insurance companies, as well as listed instruments.

So, given the long lead times on electricity plants, how much are we busy building? An embarrassingly trivial amount. While the IPP office, a procurement mechanism set up between National Treasury and the department of energy, did a brilliant job of procuring renewable energy from 2010-16, it has since been stuttering.

An “urgent” risk mitigation round was launched in 2020 to rapidly acquire 2GW of new capacity at just about any price. It ended up in farce after Turkish floating gas turbine producer Karpowership won most of that capacity. A series of court cases have effectively ended that. Finally in June this year, Scatec Solar managed to achieve financial close on three projects with 150MW.

Round five of the renewable energy independent power producers’ procurement programme was held in 2021 and has also ended up in farce. Thanks to a dispute over local content levels with the department of trade, industry & competition, there were delays to closing the projects that were tendered. During the delays, the global energy investment world was turned upside down by gas prices and war. Only three of the 25 projects that were announced have managed to achieve financial close so far with 420MW.

Round six has so far called for 4.2GW of new supply and there will be another 1GW procured in the round too. We are waiting for the results after proposals totalling 9.6GW were submitted earlier this month.

One potentially important source of supply in the mix is embedded generation. This year, the National Electricity Regulator of SA has registered about 800MW of such projects. If a proper wheeling framework is established that allows companies to sell power to each other over the grid, there could be an explosion in the number of such projects.

Add that all together and we’ve managed to only deliver 1.4GW of new procurement in 2022.

• Theobald is chair of research-led consulting firm Intellidex. This article first appeared on Business Day.