You are currently viewing our desktop site, do you want to download our app instead?
Moneyweb Android App Moneyweb iOS App Moneyweb Mobile Web App

NEW SENS search and JSE share prices

More about the app

How renewables are helping Eskom keep the lights on

Without IPPs, we would be in a near-perpetual state of Stage 2 load shedding.
Eskom admits in its own integrated report that the role played by independent power producers (IPPs) is expanding. Picture: Waldo Swiegers/Bloomberg

A deft social media disinformation/misinformation campaign – deliberate or otherwise – during last month’s load shedding chaos sought to blame renewable independent power producers (IPPs) for the dysfunction. Former executives have also persistently blamed the high cost of electricity generated by these providers as the reason for Eskom’s financial mess (because of pass-through tariffs, this is most certainly not the case).

Quite how power from renewables would be the cause of more than a week’s worth of chaotic load shedding is anyone’s guess. In fact, data published by the Department of Public Enterprises and Eskom at last week’s update briefing shows just how important the contribution from solar and wind is becoming.

On April 2, a day with a reportedly elevated risk of load shedding, Eskom was able to provide 29 798 megawatts (MW) of supply during the evening peak. But demand peaked at just under 31 000 MW at around 6pm. The difference of more than 1 000 MW was met using wind and concentrated solar power (CSP) from IPPs.

Eskom used pumped water schemes and its hydro plants (total of 1 716 MW), open cycle gas turbines (546 MW) and demand-side management through its so-called ‘virtual power station’ (409 MW) to achieve its nearly 30 000 MW. Tellingly, it did not (have to) rely on either of the (very pricey) Avon or Dedisa gas peaking plants.

Wind comprised the bulk of the additional supply provided by the IPPs during the peak, with only a residual amount – easily less than 200 MW – from solar (CSP).

The picture through the day is startling.

Source: Department of Public Enterprises/Eskom presentation

Together, wind and solar – both photovoltaic (PV) and CSP – are already consistently supplying just under 2 000 MW of power. This equates to 7% of total supply, and during the ‘8 to 5’ workday is significantly more than the amount currently being supplied by imports from Cahora Bassa. In fact, the three together just about equate to (or exceed) the amount being supplied by Koeberg (1 845 MW).

The total capacity of IPPs, as at March 31, 2018, was 4 779 MW. Exclude the 1 005 MW from the Avon and Dedisa peaking plants (which Eskom seems to sometimes include and other times exclude from its ‘capacity’), and you’re left with 3 774 MW, primarily from solar and wind. Eskom points to an average load factor of 31.5% for fiscal 2018, but this is obviously distorted by the unavailability of solar overnight. During the day, the load factor is comfortably over 50%.

And, given the current dire performance from Eskom’s coal fleet, without renewable energy from the IPPs (at around 2 000 MW), we would be in a near-perpetual state of Stage 2 load shedding.

This is what no one at the utility and very few in government will admit.

Read: What Eskom is not telling us

Consider this: in fiscal 2018, Eskom purchased 9 584 gigawatt hours (GWh) of electricity from IPPs (renewables and gas). This is 24% more than the amount of power imported from Cahora Bassa in the same period! (Ironically the amount of electricity purchased from IPPs in FY 2018 was lower than in FY 2017, but this is not at all surprising given the previous leadership.)


FY 2018

FY 2017

FY 2016


9 584 GWh

11 529 GWh

9 033 GWh

Imports (Cahora Bassa)

7 731 GWh

7 418 GWh

9 703 GWh

Coal-fired stations

202 106 GWh

200 893 GWh

199 888 GWh

Nuclear (Koeberg)

14 193 GWh

15 026 GWh

12 237 GWh

Pumped storage schemes

4 479 GWh

3 294 GWh

2 919 GWh

Hydro (Eskom)

709 GWh

579 GWh

688 GWh

Wind (Eskom’s Sere Wind Farm)

331 GWh

345 GWh

311 GWh

Open cycle gas turbines (OCGT)

118 GWh

29 GWh

3 936 GWh

Eskom admits in its own integrated report that it operates “most of the base-load and peaking capacity, although the role played by independent power producers (IPPs) is expanding” (author’s emphasis).

By 2022/2023, Eskom expects supply from IPPs to increase to 19 883 GWh, more than double the amount in FY 2018. Although this number is conservative, even at this level it will come close to exceeding nuclear (Koeberg) and imports combined.

Coal supply has peaked. Regardless of the Medupi and Kusile power stations, the wheels of decommissioning have been set in motion – perhaps forcefully so, given the age of some Eskom plants – and dozens of units will be shut down over the coming years.

Energy from renewables (again, primarily solar and wind) will take up the slack. Already, there are signs that this transition – ‘just’ or otherwise – is going to be painful and disruptive. (The burst of disinformation/misinformation last month is only the proverbial tip of the iceberg.)

Government’s updated draft integrated resource plan (2018) puts coal at under 45% of total capacity by 2030, with solar (PV and CSP) at around 12%, wind at around 15% and gas/diesel at some 16%. At this point – should this plan be finalised and implemented in its current state – capacity from these sources will exceed coal in just over a decade from now.

Does Eskom – and government – fully grasp this future?

Hilton Tarrant works at YFM. He can still be contacted at

Please consider contributing as little as R20 in appreciation of our quality independent financial journalism.



Sort by:
  • Oldest first
  • Newest first
  • Top voted

You must be signed in to comment.


and dozens of units will be shut down over the coming years Which dozens? There is around 18 coal fired power stations in SA?

Each coal-fired plant has multiple units. Matla, for example, has six.

Thanks for very informative article.

Thanks for a great article

I think by dozens Hikton is referring to boiler sets and not the power stations.

Please publish an article on construction and permanent jobs created in IPP sector. Tobias Bischof-Niemz has done some seminal work on this.

Also there are positive and negative rumours around the local community funds. What actually happens in towns where the IPP is based. Case study?

If an inheritance can split up a family, imagine what happens if you dump a heap of cash on a community. Local Community Funds can be structured as a trust, a charity or a company, but the structure does not prevent the most unscrupulous members of that society from diverting the cash flow for their personal benefit. If you want to determine the level of sophistication of a community, ask them to distribute an annual income of R5 million among themselves in an equitable and fair manner.

When the IPP initially consults with the community, everybody is positive and cooperative. This peaceful and mutually beneficial agreement ends the moment the first tranche of payments arrive. Then, you see neighbours turn on each other like a pack of hyenas around a carcass of a skunk. Their struggle to get a part of the loot will eventually motivate them to burn down the IPP. You should “know your people”. Envy is the most explosive force in an unsophisticated community.

There is only one way to distribute the IPP community benefits in a fair and sustainable way. Use the local charitable organization that has been active in that community for 50 years. The members and management of this society should not receive any remuneration and the financial statements should be audited annually. The ACVV comes to mind.

Solar is BIG business in California. For instance, shopping malls offer free EV charging thus encouraging EV owners (and there is a LOT) to charge for free while shopping. We need to think on ways to create jobs and stop relying on outdated energy supply mechanisms. Another way to get away from the high tax on fuel in SA.

As usual it’s the unions and other vested interests in coal that are holding this country back. We could easily tap wind, solar and other renewables to supply most of our power but the fossil fuel industry will do everything it can to scupper this transition.

No ways. Germany invested over $580 billion in renewables over the past 10 years or so and early this year it was the 1st time that the renewables produced 70% of total power due to an unsually high wind at that period. The result is they are still running 40% to 60% regularly on coal depending on the output from the renewables. Should they have spent that on nuclear, all the coal stations would have been closed a few year ago already. South Africa will need an obscene amount of money to generate more than 50% on renewables

As an engineer – a picture speaks a thousand words. And if it is a graph, 10x more 🙂

What is obvious from the graph:
1) Morning and evening peaks are sizable and cannot be adequately covered by solar(PV).
2) Koeberg (i.e. nuclear) is saving our butts bigtime. When Koeberg needs to be re-fuelled, which takes some time to complete, we will be in trouble.
3) Pumped Water Actuals – Eskom is running this for base load and not just peak demand as originally intended.
4) Wind – I had no idea we had so much wind generation – this is a bit of an eye opener. In contrast to solar this does contribute to the daily peaks but yes it is dependent on the weather.

That’s why gas is such an important part of the IPP mix.
ps. During the load shedding chaos, pumped storage was being used for a LOT of base load power.
Agree on Koeberg.
Wind not too dependent. It’s always windy on the west coast and near PE. Performance, generally, seems very predictable.

Plenty of wind in KZN along the coast too…

Plenty of wind in Cape Town – mainly in the parliament…

“It’s always windy on the west coast and near PE.”
Actually Hilton, this is not true. There were 4 occasions in 2016 (the only year for which I could find public data) when all our solar and wind combined produced virtually no energy. What do you propose the country do at those times?

Do we have to have gas backup for the entire IPP fleet that only operates on the rare occasions that there is no wind/solar? What is the cost of a mamoth gas system just sitting there doing nothing 99% of the time waiting for a windless night? It is misleading to talk about wind or solar costs without including the cost of the gas backup system required.

Now all we need is a simplified no nonsense net metering program for SSEG’s, Eskom could easily add a couple of MW by just allowing people to bring their : PV, Wind, Hydro, or bicycle power to the grid.

PJJ , i totally agree with you. the problem is this government and it’s SOE’s do not honour their commitments .
Contributors would be tied up legally to be paid, exactly like these bums have done to a once proud construction industry

I recently noted that people do not trust the main stream media (MSM) and one of the main reasons is they lie or report selectively to promote an agenda.

take a close look as Eskom’s annual report.

According to Eskom they paid R19.3 billion for 9584GWh of ISP energy. This amounts to R2.01 per kWh (page 81 of the report). Their revenue per kWh was only R0.8505 per kWh (page 80). This means they lost R1.16 per kWh on ISP energy (for the taxpayer’s tab). This would by no means be the only reason they are financially in the dwang but this amounts to north of R11 billion that they lost buying renewable energy. Every year! This would slide directly on the bottom line loss. This is only going to get worse now that Kathu has come on stream. Kathu CSP should output 390000000kWh adding a cool R452 million per annum to the loss assuming that the difference between the price Eskom pays and the price they sell the energy for is constant. Maybe the R452 is plus minus a hundred million. Either way it is secret.

Q: Please explain what is a “pass though tariff” and how a loss of R11 billion annually is not a big deal in light of this?

Q : Why did this article not mention the cost per kWh of the ISP energy?

Let us look at Kathu CSP system. It occupies an area of 800Ha or 8km2. Scale this up to replace a large 3.6GW coal fired power station and we are looking at maybe 500 to 1000 km2. That is a lot of land to sterilise.

Q: How do other days compare with 2 April in terms of wind power? Was this day cherry picked? Wind power in other places is extremely fickle, why should South Africa be any different? Where can one get information in this regard?

From Australia:

The output is totally not reliable. The installed 1475 MW wind capacity means nothing in the scheme of things. It’s totally unreliable and depends on … the weather.

The day (02 April) was not “cherry-picked”. This is the data Eskom has made publicly available.

The author clearly states that April 02 was picked, and indicates the reason for this, which we all know well. Although it wasn’t a “cherry”, more like a prune.

By the way, it should be mentioned in comparing the costs from the different sources of power, that the cost to Eskom for power from Mozambique is R0.02c per kWh.

Great comment. Food for thought.

3 counters would be:

1. The electricity (megaflex) price is only this low because of the sunk capex of the older plants. New builds will see coal-powered unit prices easily top 2.00. The rate of increase may then be below the current 10 year trend of approx. 15%, but will still certainly be above CPI.

2. Eskom May sell for 80 c, but we certainly don’t pay 80c. The company I work for has 12 000 employees spread over numerous sawmilling sites, and our average blended cost of electricity last year was 1.95. 2.00 for renewables, and below-inflation increases, having green credentials and not being affected by load shedding looks very attractive.

3. We seriously looked at building a 36 MW biomass plant last year. Would need to sell at 1.42 per unit to make 12% IRR. Quite tempting. This would be a great price for us, but a terrible price for the utility, because they cannot pass these prices on to residential users (Political) and are too inefficient to survive on the reduced margin (economic). So we walked away.

I dont agree with this opinion piece. It was proven that one needs atleast 10 x the actual required capacity at different parts of the country for IPP’s to reliably supply baseload power(excluding large scale hydro of course).Secondly the pass through argument of the excess tariff from renewable doesn’t hold water since Municipalities and residents owe billions for their power consumption. So where did the IPP cost pass through to? I like the idea of renewable’s but as a personal installation at home. Nuclear and Hydro, and to a lesser extent CSP should be the main backbone of the network when coal is eventually phased out..

Where did the IPP costs pass through to? The taxpayer, who will continue to carry in increasing burden of especially bid windows one, two and three. We all know stats count as the biggest of all possible lies.

Nice seeing some stats!

A further renewables factor is the amount of embedded solar PV that reduces demand. I know businesses that generate half their own energy. My house is ⅔ self consumption.

If batteries and hybrid inverters get to slightly better cost, then between it and some load shifting in consumers, those morning and evening peaks can be handled. Businesses already pay over R3/kWh in peak periods, so batteries are even now close to competitive in avoided cost.

That shift will happen very quickly.

Would be interesting to add in the MW’s produced by private entities in terms of solar installs and generators etc. (eg. shopping malls) How much did they help keep the SA lights(read economy) on.

Limited by a 1MW limit. Let’s get rid of that cap and the monopoly

I did a bit of research on SA’s actual renewable energy production figures on a daily basis. I’ve always heard that “the wind is always blowing somewhere” many times but I was wondering how true this actually is. Fortunatey, the CSIR did a detailed analysis for 2016 which is very interesting.

In this report is shows that there are periods in 4 different months when there are hours that all renewables combined are supplying just 0.1% of total generation when they should be supplying about 5%. See these pages:
pg114 (15 May), pg115(17 June), pg119 (24 Oct), pg121 (21 Dec)

The implication of this is significant in terms of intermittency. 0.1% of total supply represents about 30MW. This is out of an installed renewable capacity of 3000MW. Assuming SA’s average generation needs of 30GW continuously through the day, this means that one would need 3000 GW of installed renewable capacity (700 Medupis!). Obviously that isn’t even close to being affordable.

Even if we aim to install 20GW of renewables, this means there will still be times when we get just 200MW from all of them combined, which means that we basically need an entire backup, dispatchable power system (gas) ready and waiting to operate for just a few hours in the year when renewables aren’t performing. That means we pay for two expensive parallel systems, but one (gas) is rarely ever used. Unfortunately pump storage can’t help us much because in those few hours it needs to supply a burst of energy to cover all the renewables which are out of action.

I can’t see an affordable way we can go big on renewables without spending huge amounts on duplicate, mostly redundant fossil fuel backup systems. I’d be interested in what energy experts have to say on the above.

Dave, I like your realistic take on wind power. It’s time to get the message out.

Great article, and thanks for the graph Hilton. Currently, on another website, a guy called Andrew Kenny is busy rubbishing the value of renewables using old data to support his ‘argument’. Google him, and it becomes clear he is an avowed nuclear man. We all have to beware of fake news about renewables. Prices are coming down fast, we have ample sunshine and wind, so let’s use it for all our benefit. No need to put ourselves in bondage to Russia with trillion-rand debts.

End of comments.





Follow us:

Search Articles: Advanced Search
Click a Company: