Only three of Eskom’s coal plants are consistently producing over 2 000MW

So far this year, only three of Eskom’s coal power stations are consistently generating more than 2 000 megawatts (MW) of power, on average.

It has a total of 14 coal power stations, of which only three have total capacity below 2 000MW. Of the remaining 11, seven have generating capacity of more than 3 500MW*. These are big power plants.

This was revealed by executives from the utility in a presentation to the Mineral Resources and Energy Portfolio Committee in parliament on Friday.

Read: These four charts show the Eskom crisis is just beginning

For sure, the sample period for this is relatively short – just three weeks (to 22 January) – but the performance of Eskom’s generating fleet remains horrendously below par.

In the first three weeks of the year, the energy availability factor (EAF), a measure of available generating capacity, for its coal units was 48%.

Overall, EAF was 53.3% for the bulk of January. This is laughably below its stated 65% target for 2023/24.

What affects EAF

EAF is impacted by planned maintenance and breakdowns. Eskom has kept the planned outage factor elevated at 13.3% of the fleet for this period, a touch higher than the 12.2% achieved in the comparable period in 2022.

Read: It won’t get better for the next 6 to 12 months – Eskom spokesperson

Breakdowns include full load losses, outage slips, boiler tube leaks, trips, and partial load losses. It is the last of these that has been the major contributor to unplanned losses in recent years.

Until October, it had contributed to losses of nearly 13% of total capacity. This was followed by full load losses (nearly 10%). Of that 10%, more than half were defined as “major events”, meaning large units (±600MW) being offline for nearly a month. At that point, total unplanned losses equalled 32.2% of installed capacity.

Top three

It is no surprise that the three that are producing more than 2 000MW consistently are among the best-performing stations across Eskom’s fleet.

Lethabo, whose manager Thomas Conradie is currently acting group head of generation, produced (on average) 2 759MW for the first three weeks of the year (from total capacity of 3 558MW). This equates to an EAF of 77.5%.

Matimba, which has also been a ‘dependable’ station historically, achieved a very similar EAF figure of 74.3%, with 2 742MW being produced out of 3 690MW, on average.

Medupi, where performance has steadily improved over the past year, achieved average output of 2 075MW (from 3 600MW), meaning an EAF of 57.7%. Unit 4 remains offline since it was blown to pieces just more than a month after the ‘official’ completion of the entire power station in 2021. This unit will return by August next year.

Read: Eskom reports blast at Medupi Power Station unit [Aug 2021]

 ** Peaking plants include pumped storage schemes and open cycle gas turbines.

One can see from this detailed breakdown (sorted by EAF) just how shockingly poor the performance is at some large power stations.

Duvha’s performance of 21% is diabolical, as is Tutuka’s at 33%. Duvha was a ±3 600MW six-pack power plant until Unit 4 exploded in 2011.

Eskom took the insurance payout and decided to not return that unit to service. Duvha Unit 2 remains out of service after a fire broke out in June. The unit had been offline since January 2022 for a general overhaul. It is assumed this unit will return in this financial year (that is, before April).

Kusile’s issues will be detailed in a Moneyweb article later this week.

Selective focus

The six highlighted power stations above are Eskom’s top six stations of focus.

Maintenance is being prioritised at these plants, as any improvements will yield the greatest possible returns (crudely, in megawatts).

In the next six months, Eskom sees nearly 2 000MW returning to service from four of these power stations: 600MW from Tutuka, 475MW from Duvha (likely Unit 2), 422MW from Matla and 365MW from Kendal.

Remember that Eskom’s maintenance budget is finite and there is the constraint of time. There is very little point in investing billions of rands, many months, and the required management oversight/effort on the three smallest stations in the coal fleet (Grootvlei, Hendrina and Camden). Together, these three stations are about the size of one of Eskom’s 3 500/3 600MW plants. (Plus Grootvlei and Camden are performing ok.)

Also, crucially, the recovery of generation capacity is a constantly moving target. Big units will come back online in any rolling six-month window. Plus, big units will need to be taken offline or, worse, accidents will force them out of service in that same six-month period.

The trick is to ensure that quality maintenance is done on units to ensure they produce a reliable supply of power (and to make sure you’re bringing back more capacity than you’re taking offline).

Not even Koeberg can plug the gap.

The nuclear power station’s EAF won’t be much higher than 60% for the year – that’s if the steam generator replacement at Unit 1 goes according to plan (and it returns in June). Unit 2 will be taken offline for the same work later this year.

It is outrageous that Eskom’s peaking units have a combined EAF of nearly 90%. By definition, pumped storage schemes and OCGTs (open cycle gas turbines) ought to only be used for around a third of the day (for the morning and evening peaks).

The pumped storage schemes are being used as baseload power throughout the day (hence the higher stages of load shedding overnight as these require more than 3 000MW to pump water back to their top dams).

And, of course, we know just how many billions Eskom has spent burning diesel …

* Technically, Matla’s capacity is ‘only’ 3 450MW (with nameplate capacity of 3 600MW), but let’s not split hairs over 50MW.