Over the weekend South Africa’s energy minister, Dr Kgosientsho Ramokgopa, announced an ambitious new plan to more than triple the country’s energy production infrastructure within fifteen years. This was revealed in the latest variant of the government’s Integrated Resource Plan (IRP).

Since 2010, IRPs have been the government’s blueprint for electricity supply strategy.

The centrepiece of this latest plan is the addition of over 100 000MW in new electricity generation capacity by 2039, at a projected cost exceeding R2 trillion, all while moving towards net-zero carbon emissions by 2050. This generation target is highly ambitious, given that current total installed capacity is roughly 50 000MW. Installed capacity refers to the maximum amount of electricity that all power plants combined could produce if they ran at full power, without stopping, under ideal conditions.

In reality, actual output is always less, because maintenance, breakdowns, and weather (for wind and solar) mean power plants rarely operate at maximum output.

According to the government, coal currently accounts for around 60% of the roughly 50 000MW installed capacity. Wind and solar together contribute about 30%, with the balance coming from nuclear (just under 5%) and other sources, including diesel turbines.

By 2039, the plan envisages coal’s share dropping to 30% of a projected 160 000MW installed capacity, with wind and solar at 40%, gas at 10%, nuclear at 5%, and the remainder from other sources and storage technologies.

Misleading

However, the plan to an extent, and certainly in how it has been reported, is misleading in two key respects.

First, it creates the impression that South Africa will triple its generation capacity from around 50 000MW to 160 000MW. In reality, installed capacity differs greatly from actual electricity produced. Wind and solar, for instance, only deliver about 20% of their installed capacity and hence you don’t get that many MWs fed into the economy from every MW of installed production.

Second, the suggestion that electricity available to support industrial expansion will lift sharply is also questionable. For energy to support industrial expansion, it must be connected to the grid and be dispatchable, which means available when and where needed.

Currently, there is almost no space on the west-east electricity grid, meaning a new grid, approximating the size of the existing one, would have to be built to carry new solar and wind energy. This requires financing and construction. That energy would also need to be stored to make it available to industrial users on demand, but storage infrastructure is extremely expensive.

The plan includes 6 300MW of new storage capacity which is significant, but also still requires financing and construction. The dispatchability issue may be partly addressed by using wind at night and solar by day as long as the wind rarely stops and the sun shines. Gas infrastructure, which is forecast to supply about 10% of capacity, still needs to be built, as does any new nuclear capacity.

Ambitious and complex

In summary, the plan is extremely ambitious and complex, even under ideal conditions without delays. It is also extremely costly. At R2 trillion, the bill, assuming no overruns, amounts to 25% of GDP and would have to be borrowed, but this is not included in current debt or deficit projections (the government’s Medupi and Kusile power station build costs overran by 200% to 300%).

Frans Cronje told The Common Sense he is wary of the plan’s complexity, cost overruns, and unresolved questions around storage, grid capacity, and dispatchability: “What I don’t understand is why such an immensely complicated and complex approach has again been chosen to address what is actually quite a simple problem. South Africa has coal, over 20 000MW of defunct coal stations, and surplus grid capacity to link those stations to industry. If it refit those stations and ran them at 70% capacity South Africa could quickly put its economic growth rate on track to reach 4%. All of this can be done cheaply and for very little money – I would really recommend doing that first before setting off on a trajectory as ambitious as the new IRP.”

Cronje’s initial estimates suggest achieving the government’s plan by its due date is not feasible: “Even if it is partly achieved, and taking into account the capacity factors of solar and wind, total peak daily production will rise only modestly, but not in excess of what coal refitment would achieve.”

As one of the first analysts to predict South Africa’s decade of loadshedding, and the first to call its end, Cronje has tracked government IRPs for 15 years: “In each case, there was vast complexity far beyond the ability and resources of the state to execute. The first IRP, in 2010, predicted that by today the state would supply between 400 and 500 terawatt hours (TWh) of electricity annually (a TWh is equal to 1 000 000 megawatt-hours). At the time, only 250TWh was being supplied. Had the IRP’s targets been met, and the 400–500TWh achieved, South Africa’s economy might have grown by 4% to 5% annually. The plan failed because it was too complex, unfunded, essentially ignored coal and grid constraints, and focused on too many peripheral and environmental issues. As a result, only 220TWh was consumed last year and growth has hovered at around 1% for a decade. Fifteen years later, even after the African National Congress lost an election, which I maintain would not have happened if sensible energy policies had been pursued, the government appears to have learned nothing.”