Electricity Generation in South Africa

There is so much said about this in South Africa and so many vested interests that it can be difficult to make sense of the subject. Complicating this is the fact that generating projects often have long lead times. Some of the factors clouding the issue are:

  • Jacob Zuma appears totally committed to the nuclear option with contracts for Russian supply possibly already in place, as announced first by Russia and then by SA in a word for word translation before there was some hasty backtracking.
  • Not unusually the Guptas seem to working in tandem on this after buying uranium assets that few others deemed valuable.
  • Brian Molefe at Eskom seems to have bought into the Zupta model in punting the nuclear option and already buying coal from the Guptas on favourable terms. He is also reluctant to have too much of the Eskom generating monopoly taken away (distribution is more of a natural monopoly). This counts against IPP (independent power Producers) whether for renewable or fossil fuels.
  • Those pushing renewable quote low costs per kWh but an apples with apples comparison with other sources is not possible unless the power can be generated 24/7 (eg biomass) or there is a cheap way of storage (currently molten salt in the concentrated solar plants).
  • The Department of Energy has not finalised an integrated development plan for many years over which time there have been technological developments and a reduced demand on Eskom, greatly assisting in preventing power outages for the past year.

Looking at the various options:

  • Coal has its back to the wall in view of carbon emission pledges. A coal power station project has a very long leadtime and Eskom has gone from being a world leader in project managing these to an also ran. Thermal efficiency is as low as 33%. A combined heat and power station such as that located in AvedøreDenmark, can achieve 49% conversion into electricity and 94% for combined electricity and industrial and domestic heating. Because coal is a natural constituent of the earth it contains numerous trace elements. I recall being surprised in my studies in the early 60s that the release of radiation from a coal fired power station exceeds that under normal operating conditions from a nuclear plant, and I now see it quoted at 100 times the release from an equivalent nuclear plant. A 1,000MW coal fired generator releases about 5.2 metric tonnes of uranium per year. Despite the environmental aspect the Department of Energy recently conducted the first auction for non-renewable energy in the form of 2,500MW of coal generation. Commentator Chris Yellend estimates the cost of power from Medupi and Kusile to be in the R1-1.20/kWh region.
  • Oil does not seem to have been considered as an option. Although plants can be constructed in far less time than for coal the environmental issues remain. Eskom of course relied heavily on diesel powered open cycle gas (sic) turbines for peak demand power, but as I recall at around R4/kWh.
  • Gas has earned a name for itself as being less polluting than coal (about half the CO2 and far less sulphur, soot etc), almost completely flexible (relatively easy to switch on and off) and more efficient, especially the combined cycle rather than open cycle turbines. However there have been concerns raised recently. One ‘advantage’ that gas has is that it is a normally transparent pollutant. Thus when 100,000 tons spewed into the atmosphere in Aliso Canyon, California a few months ago there was little for the media to show even though thousands of homes had to be evacuated. The largest constituent of natural gas is methane which has 25 times the global warming effect of CO2 but lasts only about one fortieth the time in the atmosphere. More worrying though is the leakage which appears to occur along the supply chain, put at 2.5% by the US Environmental Protection Agency from relatively mundane problems such as leaking valves. Again it is a case of what the eye don’t see….. Whether fracking takes off in SA could be a factor.
  • Renewables recently have concentrated on solar and wind. Except for the concentrated solar power plants with molten salt storage their output is variable and inconsistent. Countries which have gone into this big time often have high electricity costs. Those which prioritise offtake from renewable sources before baseload thermal and nuclear plants put the latter in a difficult position as they cannot cut back on their production. Thus there was a windy sunny day in Germany when renewables supplied all the midday requirements but an opposite cloudy calm occurrence in Australia when renewable output was negligible and the purchase price went through the roof. Obviously constant wind conditions are the ideal and I understand that Africa’s largest wind farm being established in Kenya by one of the recipients of this email is about as close to ideal as possible. Where industry and domestic users go off grid with solar or wind the main generator and distributor are left with a smaller base to carry all their maintenance and other non-fuel costs. Hydroelectric power stations are still being built but appear to be mainly in developing countries. Storage of power from renewables is still an immense problem and almost exclusively by way of pumped storage schemes such as we have in the Drakensberg. These are expensive projects and necessarily net users of electricity such that on a graph of output capacity from all sources they are shown as negative contributions. Lithium-ion batteries are still apparently something like 10 times the cost necessary for them to become a viable storage medium. Meanwhile renewable energy costs continue to fall. At the first auction in 2011 the average price was R2.37/kWh but by 2015 was down to R0.77. This is an obvious threat to Eskom such that chairman Ben Ngubane has told Tina Joemat-Petterson that the utility was not willing to enter into any further purchase power agreements.
  • This is the most contentious of the options by virtue of the cost, project size, long leadtime and corruption potential among others. The advances in renewables and storage and the possibility of more consumers going off grid have given rise to questions on the need for the suggested 9,600MW. Business Day reports that the overnight capital cost (the cost in today’s terms, ie excluding inflation and interest costs) included in government modelling for the 2016 draft plan is $6,000/MW. If I plug this figure into a very simple model assuming 9,600MW, a gradually increasing capex pa to a 2028 finalisation date at an FX rate of R14.00:$1,00, with interest accumulating at the R186 bond yield of 8.7% and inflation of 6%pa I arrive at the following figures in trillion rand:
    • Overnight cost 8
    • Inflated capex excluding interest 1.3
    • Interest 0.3
    • Total actual cost 1.67

This represents quite a leap in the dark given the lack of any up to date published plan and the uncertainty implicit in demand and other projections so far forward. Meanwhile other countries are tending to move away from nuclear. Only recently Theresa May delayed the £18bn Hinkley Point nuclear station. As I understand this, because nuclear persistently fails to show a return on such plants David Cameron’s government was prepared to guarantee generous offtake prices for 35 years.

Since writing the above I see that Chris Yelland has made a calculation based on an overnight cost of $5,776/kW, a shorter buildtime and lower WACC (weighted average cost of capital) to come to R1.30/kWh at current prices.

(Business Day, Wikipedia, The Daily Maverick, The Daily Insider, The Economist, Moneyweb)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s