Load shedding in 2023 worse than the last eight years combined
The problem? Eskom’s old, dirty, unreliable coal fleet.
The severity of load shedding this year, where Friday (8 September) marks the 40th day of stage 6 in 2023, means that simply counting the number of days of power cuts is no longer helpful (there’s only been a single day since January – 21 March – where Eskom was able to meet all demand for 24 hours.).
Data from utility app Loadshedding Notifier shows that when measuring intensity, in other words the gigawatt hour (GWh) impact, there have been 19 794 GWh of power cuts this year (to Wednesday).
This is more than the 19 103 GWh impact in 2015, 2018, 2019, 2020, 2021 and 2022 combined! (There was no load shedding in 2016 or 2017.)
This measure is useful as an hour of stage 1 is not the same as an hour of stage 4, and is not the same as an hour of stage 6. By definition, the last of these is six times more severe than the first – it requires 6 000MW of demand to be ‘removed’.
Loadshedding intensity measured in gigawatt hours (GWh)
We’ve had more days of stage 6 this year than in any other year, plus days of stage 5 in 2022 and 2023 combined. It’s that bad.
Gone is the bullish bluster from Electricity Minister Kgosientsho Ramokgopa. A week ago, he was telling anyone who would listen that “We are very bullish about the prospects going into the future.”
A month or so back, he confidently (and absurdly) predicted that the generation output of Eskom’s coal fleet would improve during summer (versus winter). This is factually inaccurate (and rather laughable).
Output has improved during every single winter where cooler Lowveld temperatures help greatly. That’s just how its power stations work.
This week, at a hastily convened media briefing, he admitted that there would be “intense” load shedding over the coming months as Eskom ramps up maintenance.
Maintenance and breakdowns
“We are going to stick to planned maintenance, we’re going to stick to philosophy maintenance. We do accept that, in the short-term, it’s going to result in the possibility of intensified load shedding,” said Ramokgopa.
Wasn’t this what former Eskom CEO Andre De Ruyter and COO Jan Oberholzer had been preaching for years?
Its latest weekly system status report shows that units out for maintenance last week increased to 12.82% of its total capacity (about 6 000MW). That’s the highest level since April, and double what has been offline during winter.
Included in this is the 900MW from Koeberg, where the steam generator replacement on one unit continues to run months and months over schedule.
Breakdowns last week were at about 15 000MW. This week, those spiked to 16 000MW. Combined with issues at its pumped storage schemes, this saw the move to permanent stage 6 load shedding (from a ‘peak’ of stage 4 last week).
Eskom doesn’t even have the headroom anymore to alternate between stage 6 in the peak and overnight and stage 4 during the day.
It has been burning as much diesel as possible this week. Between 06:00 on Saturday and 23:59 on Tuesday, it burned diesel at its open cycle gas turbines for 88 out of the 89 hours (sometimes as much as 1 000MW overnight).
The problem, as ever, is Eskom’s old, dirty, unreliable coal fleet.
The last time these power stations together managed to generate more than 19 000MW was last Thursday. From noon on Saturday until Tuesday at 23:59, as plants were removed for scheduled maintenance and units broke down, it has managed to generate more than 18 000MW from its coal fleet for three out of 84 hours.
The cascading set of coal units breaking down this week forced Eskom’s hand. It’s using a combination of diesel (including calling on two independent power producers, the Avon and Dedisa peaking power plants) and its pumped storage schemes to keep the lights on, as well as from shifting to an unprecedented stage 7.
On Tuesday, it was helped by a strong contribution from wind generation (fairly consistent 1 500MW to 2 000MW).
The good news? (Is there any?)
Temporary repairs at Kusile are running ahead of schedule. One of the units could return to service next month – two months sooner than anticipated.
This will add 720MW to the grid immediately, with the other 1 400MW likely to be back before the end of the year. To expect a constant contribution of 2 160MW from those three units is unreasonable. There will be breakdowns and trips.
Kusile hasn’t exactly been a top performer – even before the flue duct (chimney) collapse.
We can probably bank on an average of 1 800MW from those three units. There will be good weeks and bad ones. This should be enough to stave off a repeat of last year’s stage 6 nightmare this December.
Still, there are many unknowns… including exactly when that one Koeberg unit is going to come back online.