Dave Nicholls
Featured image: Stock

In South Africa, one of the key national issues at present is the formulation of the national Integrated Resource Plan (IRP).  Given the vital need for reliable and available electricity supply for the country to achieve its long-term industrial growth, this may be more important than the current debate over ‘state capture’ writes Dave Nicholls. 

In the debate, there are a series of ‘facts’ that are seen to be relevant to the formulation of a national IRP.  Some of these ‘facts’ are based upon either a lack of understanding of the real dynamics along with the national economics of the electricity supply industry or a deliberate distortion.

If one is to develop a theoretical model of the electricity system and then use that model to establish the best long-term solution, it would need to be benchmarked against real-world experience.

Current world experience clearly appears to show a very strong correlation between renewable energy penetration in the market and overall electricity tariffs. It also shows a very limited correlation between high penetration of renewables and low greenhouse gas emissions.

The only consistent international trend is that the combination of low carbon emissions and low tariff levels appear to occur when the dominant grid supply is large hydro electric installations and/or nuclear power. Given this empirical evidence one has to ask where the current IRP2018 assumptions have differed from, or ignored, the real world.

The nuclear ‘fact’

As a nuclear proponent (based on nearly 40 years in the electricity business) the first ‘fact’ I would dispute is that nuclear economics make no sense for SA’s economy. This is clearly a function of the input data used in the calculation. The studies done in the IRP to date assumed a real cost of capital (interest excl. inflation) of 8.2%.

This bears no relation to the real financing costs of a nuclear plant, which the recent Egyptian nuclear deal had financing of 3%-4% in US$. Given the US inflation over the last 10 years has averaged 1.9% this is a real cost of capital of about 2%. With this the capital element (the largest portion of the cost of nuclear power) would fall massively and make it very much least-cost in the long term.

This economic ‘fact’ is linked to the belief that Eskom’s balance sheet cannot alone undertake the capital energy requirements going forward, but this is completely irrelevant to the debate. As has been seen with the REIPPP funding all electricity investment has required 100% state guarantees (some R250bn so far) linked to 20 year index linked PPAs.

So all the financing risk ends up going back to the state, in some form or other. IPPs just mean that someone gets a very profitable position as the middleman between the state funds and the transmission system requirements. In fact, the easiest technology to obtain low cost funding for is nuclear as all the vendor countries offer full funding packages, such as the one the Russian’s gave the Egyptians noted above.

The renewable energy ‘fact’

Another key ‘fact’ is that renewable energy prices are decreasing. While this is true on a kWh basis at the point of connection to the national grid it ignores the time value of electricity as well as other technical costs. People want electricity when they need it, not just when the sun is shining. The peak demand on the system is the evening peak, between 5pm and 9pm, when solar PV is not functioning, so duplicate backup power is required. This is the logic used for making sure that (imported) natural gas is this back up. For renewable energy to compete with its own backup supplies its total cost of supply has to be less than the operating cost (i.e. fuel) of the backup. This occurs with imported gas but not with coal, hydro or nuclear.

If South Africa had fracking at the same prices as the US (below $3/GJ) then the model would create a total gas solution as the least cost and build no renewables. Of course, if you want a very low carbon footprint the model builds just nuclear and hydro. You must set the cost of capital and gas and the carbon constraints just right to get the solution you want!

Another system cost that is ignored is the cost of expanding the grid to connect the distributed renewable sources, such as the recent BRICS bank loan to Eskom of $180m to allow it to extend the grid to connect 670MW of intermittent renewable energy. This cost differential with conventional generation sources is explicitly ignored in the current IRP. Given the cost of $180m for 670MW this leads to a unaccounted cost of some $270/kWe, which, using the IRP cost of capital assumptions and the actual performance of solar and wind measured in South Africa, is a 20% addition to the long term IRP assumption on wind power costs and 30% for solar PV.

The scale of the system costs, over and above pure generation costs, is such that of the approximately R2/kWh that South African consumers pay for electricity only about 30c/kWh is actually for the pure energy (mainly coal) with the balance being for the other aspects (continuous availability, grid stability, transmission, distribution etc.).

The renewable energy sources, which only displace this energy cost, create many problems (costs) for the overall system management. It is notable that at a recent press conference the Eskom General Manager for grid operations stated that they took no credit in their plans for the availability of renewable energy to meet demand.

The job creation ‘fact’

The last ‘fact’ to mention is the job creation potential. Minister Radebe has stated in a parliamentary answer in July 2018 that the REIPPP programme would, during 20 years, create over 114,000 job-years. For a nuclear programme we have publicly available analyses, being the 2017 KPMG study into Koeberg’s actual economic impact showing some 35,733 direct, indirect and induced jobs being supported (over 20 years that is 714,660 job-years).

Also, the EIA study into the construction of a nuclear plant (Nuclear-1 EIA Economic Impact Assessment, Conningarth Economists/Imani Development (SA) (Pty) Ltd), showed that the construction of a new nuclear plant next to Koeberg would generate an average of 91,194 jobs for 7 years (that is 638,538 job-years). So I am not sure where the country would have better spent the R200+bn capital of the different alternatives.

I fear that many of the proponents of the current draft 2018 IRP have been more ideological than analytical in their concern over criticisms of the Draft 2018 IRP.

The author welcomes responses to this Op-Ed.

About Dave Nicholls

Dave Nicholls retired from Eskom in December 2018 as the Chief Nuclear Officer after 34 years with the company.  He served as the Chairman of the IAEA Technical Working Group on Advanced Light Water Reactors from 2010-2017 and is currently the Co-Chairman of the IAEA Technical Working Group on Nuclear Power Plant Operations. Prior to joining Eskom, he was an engineer officer in the UK Royal Navy, serving on nuclear submarines.