Recently, I read an article about the nuclear power deal.

In the article, you, Mmusi Maimane, leader of the Democratic Alliance (DA) write: “Jacob Zuma’s nuclear deal will be disastrous for South Africa. It will literally bankrupt us, diverting billions of rands from poverty reduction projects, while producing a nuclear white elephant in two decades’ time. But it will make a lot of money for Zuma, the Gupta family and African National Congress cronies in the short term, and they will be long gone by the time we feel the real pain.”

You continue: “For one thing, we simply will not need 9.6 GW of new nuclear energy in 2030. We currently have a generation capacity of 42 GW and are expecting an additional 11 GW from the Medupi, Kusile and Ingula projects in the early 2020s, which will offset the decommissioning of most of South Africa’s coal-fired power stations in the mid-2020s.”

I am quite sure that a great deal of corruption will be attendant upon part of the nuclear deal. This is how the present government works. Let us imagine that we do not care who bene- fits from the corruption. That is a financial issue. The question is more about whether we need 9.6 GW of additional power by 2030 and whether this should be nuclear power.

So, do we need 9.6 GW? We will need more than we have now, that is for sure. Some of the coal stations have to go. I am quite certain that we will not need more than 9.6 GW. It is probable that we will need less. It all depends on investment, which is dependent on perceptions of government, currently at historical low levels.

Let us now deal with the question of whether this should be nuclear power. Well, it cannot be coal – it is no good transporting coal to the coast to make electricity, some of which was used in coal transport. The new power station or power source has to be on the coast for the following reasons (none of which I am sure the DA will grasp).

The demand to the Western Cape is about 4 000 MW, and south of the Hydra substation it is about 6 000 MW. The transfer limit of the Eskom transmission system to the Western Cape, with no Koeberg generation, is 2 900 MW (currently), while the transfer limit to Hydra is about 4 000 MW.

Thus, with no Koeberg, we are short of 2 000 MW. This can be picked up by the Ankerlig and Gourikwa gas turbines, but this costs a lot. It is planned to link Port Elizabeth (PE) and Cape Town to the 765 kV interconnectors and build a 400 kV line PE/East London to the northern Transkei border. The transfer limits to the Western Cape and south of Hydra will fall away. The logical nuclear station location will then be at the northern Transkei border, since it can stabilise the KwaZulu-Natal voltage profile, can absorb kVAR against Gauteng major fault, can supply the Western Cape and the Eastern Cape for contingency with the loss of Hydra, can supply the Western Cape and the Eastern Cape for contingency loss of 765 kV interconnect, and can supply the Western Cape and the Eastern Cape for contingency loss of Koeberg.

Did you get that all? I bet not. The fact is that power stations are about reliability of supply location relative to load requirements. You can go on about wind turbines and solar photovoltaic (PV), gas turbines and biogas generation, but these are just not reliable enough.

The Council for Scientific and Industrial Reserach has stated that, to support 8 GW of baseload, you need 16 GW of wind power, 6 GW of solar PV and 8 GW of other generation. It strikes me that, if we just stick to the 9.6 GW of nuclear generation, we will be sorted. But is this affordable? Nuclear construction is the most rigorously controlled construction regime in the world. There will be backhanders but it is not likely that Luthuli House will be supplying reactor control rods any time soon.

Mr Maimane, get some engineers on board who know stuff instead of presenting dog mathematics about South Africa’s power requirements. Oh, and one thing: on the graph forming part of your presentation, the vertical axis should be ‘annual energy consumption’, not ‘annual energy demand’. Oops.