By Hügo Krüger
The South African Department of Energy had a final session of public participation today to explain the latest revision of the draft Integrated Resources Plan (IRP). I have some reservations that I will explain below, but I still want to congratulate the modellers for listening to those of us who sent our criticism. They did take some of it to heart.
I put a small thread together on twitter regarding what we can anticipate: https://x.com/hkrugertjie/status/1861466983637606783
The government finally published realistic cost comparisons for various scenarios and it admits that offshore wind and concentrated solar is more expensive than nuclear. By implication, if the cost to firm the electricity (i.e. make it reliable) is taken into account then renewable regress to the mean of large thermal generation.
That detail aside, the IRP modellers finally put a long term window going all the way to 2050. This is sensible, because the previous practice of using two 10 year windows effectively locked out LNG and large nuclear power plants (that do take 10 years to procure, after which you can stagger them).
The “reference” or “least cost” case contains wind and low levels of natural gas, presumably, by looking at the above graphs, this is onshore wind. It remains to be seen if our grid (that is congested) can accommodate it.
With all the variations in the coal fleet, it isn’t surprising to me that a computer model would select for high levels of variable wind and not solar generation. The reason for preferring the former and not the latter, is because load shedding was most pronounced during the evening peak. That is still the weakness in the system. What the modellers lacked was understanding why their computer might select for it, and they don’t seem to add much comments.
However, being the “least cost pathway” isn’t necessarily to say that the government “should” strictly follow this pathway. Notably, if the reference cast doesn’t include natural gas commitments then the demand for wind and solar can be even higher, but do notice how it also increases the demand for diesel (OCGT).
Where it gets interesting is the other pathways that the government proposes. The one that I like, for obvious reasons, is the nuclear pathway. One large nuclear reactor is a start, but I see no comment on the ability to stagger nuclear plants. Notice how the modellers selected for technologies that provide stabilized generation? The issue is not “more generation”, it is more “firm generation”.
South Africa does not need more generation, because our electricity demand is actually declining. They have missed our incredible pumped storage potential - despite being a water dry country. SA has about 5GW of untapped potential, that would ideally integrate renewable more easily and also smooth out the dying coal fleet.
Then what the IRP just gets wrong, flat wrong, is to not use maintaining coal generation, i.e. the brownfield generation as its reference case. The generated reference case assumes a smooth roll out of LNG. But in a country without LNG infrastructure, I find that implausible that the timeline is realistic. As an engineering rule of thumb, it is always easy to maintain the existing infrastructure, because properly executed infrastructure has long lead times.
What is termed “delayed shutdown” should be the reference case, because it is the most easily achievable. In practice we call it the “brownfield’s scenario”.
The IRP has another positive aspect to it, which is to point out that even if we maintain the coal fleet, that SA would still be decarbonizing. Rooftop solar in particular is expanding at incredible rates and therefore even if we keep burning the same amount of coal, we would be reducing our relative footprint. This implies that ANY carbon tax or carbon policy is pointless.
Nobody, not even coal advocates believe that we should build more coal power stations willy-nilly. Even maintaining coal has a positive impact on water generation. In fact all pathways have a reduced water demand, this is critical for the 25th driest country in the world.
Many of you would recall the debate about the levelised cost of generation, which is why you cannot compare the price per MWh of nuclear with solar for example. Well the IRP modellers figured this out and finally, and perhaps the most importantly, the various pathways have been compared using the total system cost. The graph below shows that the cost between the various systems do not differ widely.
By implication, the type of power generation that SA will pursue in the coming decade will include:
- more renewable, onshore wind and solar in particular
- more LNG
- and perhaps more nuclear (but it remains a democratic decision)
- most importantly, a commitment to keep part of the coal fleet going longer than their initially planned retirement age
I give the modellers 65% for their achievement (I am a strict teacher), they did not get a distinction because they failed to see that the reference pathway should be to fix the coal fleet. They passed, because they did consider it. But even if they got that right, they would not get 80+% (an equivalent of an A+ in South Africa), because they didn’t add more nuclear energy to the long term mix.
But for the sake of ending the debate, with all my reservations, I am willing to accept this IRP, on the condition that we start making preparations for a new nuclear reactor, that is now in the mix, tomorrow.
Hügo Krüger is a YouTube podcaster, writer and civil nuclear engineer who has worked on a variety of energy-related infrastructure projects ranging from Nuclear Power, LNG and Renewable Technologies.
BUSINESS REPORT