Stage 6: How did Eskom get it so wrong?

Just more than a day after Minister of Electricity Kgosientsho Ramokgopa was telling everyone who would listen that Eskom would handle the increased demand caused by a brutal cold front, the country was plunged back into Stage 4 load shedding at very short notice at 1pm on Monday.

According to News24, he said last Sunday: “Even if demand were to surge to the worst-case scenario, if these machines were performing at the rate at which we were expecting them to, we shouldn’t have a problem. Even if it [demand] gets to 34 000MW or 36 000MW, we shouldn’t have a problem.”

Turns out, this optimism was misplaced. The reality is that Eskom’s coal fleet is old, unreliable and unpredictable. According to data presented at that week-ago briefing, there are between 7 000MW and 8 000MW of units “at risk” at any given point in time.

(As Moneyweb cautioned last week, there is an awful lot of spin surrounding Eskom at present – expected, given there’s an election next year. Despite Ramokgopa’s assertions, Eskom’s energy availability factor actually declined from 58.29% in the last week of June to 57.88% in the first week of this month.)

Read: Beware government’s spin on ‘fixed’ Eskom

The story about how the country went from no load shedding to permanent Stage 4 to Stage 6 for the afternoons from Wednesday, to permanent Stage 6 from 5am on Friday to (an unexpected) 1am suspension on Sunday is a familiar one. Eskom continued to lose units across its fleet from Monday until this weekend.

On Monday afternoon, breakdowns totalled just 14 252MW, with planned maintenance at 4 118MW. This is pretty much the level at which Eskom had been able to sustain no load shedding during the day, with Stage 3 from 4pm to midnight. By Thursday afternoon, breakdowns had spiked to 17 679MW, with planned maintenance at 3 090MW.

Those 3 000MW of additional breakdowns across the week necessitated Stage 6 load shedding.

The writing was on the wall on Monday night already.

At that point, Eskom relied on 22 open-cycle gas turbines (OCGTs) and gas turbines as well as interruptible load supply of 447MW, a virtual power station of 42MW and a load curtailment of 1 458MW.

The data shows it threw all its emergency reserves and options at the problem on Monday, hoping it would somehow be able to bring back coal units for Tuesday’s peak. This did not happen. Instead, it burnt as much diesel as it was capable of (nearly 2 000MW) constantly throughout Monday and Tuesday. By Wednesday, it had exhausted about a third of that generation capacity. The OCGTs only contributed around 1 300MW throughout the day.

By Friday, it was all but out of diesel supply. This is why it confirmed Stage 6 through all of Friday and for the rest of the weekend.

	Eskom availability	Peak demand	Load shedding
Sunday 9 July	28 013MW	30 513MW	3 054MW
Monday 10 July	27 650MW	33 370MW	3 771MW
Tuesday 11 July	27 687MW	33 902MW	4 082MW
Wednesday 12 July	26 699MW	34 057MW	6 432MW
Thursday 13 July	Not published
Friday 14 July	27 086MW	31 707MW	6 416MW

But the supply side only tells half the story.

A study of the demand-side data shows that Eskom grossly underestimated just how high peak demand would be last week.

As recently as Monday, it forecast peak demand last week of 32 992MW. The week before that, it forecast a peak of just 32 141MW. In May, it expected demand for last week to be over 33 500MW.

Of course, slightly milder weather conditions meant that demand has been suppressed this winter.

But surely Eskom should’ve more accurately foreseen the impact of the cold snap? The minister himself spoke about it at length last Sunday!

Instead, Eskom thought demand would be 33 000MW. In reality, it was over 34 000MW. This is not simply about the extra 1 000MW required at peak. It factors into Eskom’s entire plan to balance supply and demand. The plan – on the coldest day of the year so far – was to have no load shedding on Monday and implement only Stage 3 load shedding from 4pm.

By lunchtime, it had been caught out – burning as much diesel as it could and running its pumped storage schemes at full capacity (>2 500MW). Wind was non-existent, and cloud cover meant the contribution from photovoltaic solar was lower than normal.

Had it instead implemented Stage 2 load shedding during the day and Stage 4 overnight, as it did this past Sunday (16 July), it wouldn’t have exhausted its emergency reserves on Monday already and battled to catch up all week. The end result, Stage 6, was entirely avoidable.

Read:
Grid storage is the missing link in SA’s electricity crisis
Stalled SA power projects for 1 850 MW risk losing grid access
How Eskom managed to tame load shedding