Wet coal aside, Eskom’s immediate outlook for its available generation capacity is much improved from a month ago. At that time, its three-month outlook published in its weekly system status reports indicated that until mid-April, it would “definitely” be between 1GW and 2GW short to meet its reserve margin and “possibly demand” for six of the 13 weeks. This picture has steadily improved from early January until now.

In the most recent update, published at the end of last week, it sees risk in just one week (specifically next week) where it is likely to be between 1GW and 2GW short to meet demand (including its operating reserve). For seven of the next 11 weeks, it may be less than 1GW short to “meet reserves”.

There are an awful lot of assumptions here, chiefly that 12GW of capacity will be offline due to breakdowns (added to this is 2.2GW required for its operating reserve to get to the 14.2GW “planned risk level”).

If there is no sudden spike in breakdowns, the picture looks significantly more positive than it did at the start of the year.

Even if there is a forecasted shortfall, Eskom is typically able to meet this peak demand with its peaking power plants — either open-cycle gas turbines (OCGTs) and pumped storage schemes, or OCGTs operated by independent power producers (IPPs). The utility is usually also able to ride out breakdowns of the odd unit (generally an impact of 500MW to 800MW) by using this peaking power as baseload power.

However, it’s when a number of units trip or break down simultaneously that it is forced to implement load shedding, as these outages typically total 2GW (or higher).

Improving generation

The generation outlook for situations when there is this spike in breakdowns does not improve until the end of March.

For context, Eskom said on Tuesday that 14.4GW of capacity is “unavailable due to breakdowns and delays”. This necessitated stage-2 load shedding overnight as it needed to replenish its emergency generation reserves (chiefly its pumped storage schemes which it is using for baseload power, given the high level of breakdowns).

From its near-term outlook, the overall improvement has not come at the expense of a reduced forecast in demand, nor at the expense of the level of planned maintenance. This suggests it could only be as a result of steadily improving generating capacity.

Eskom has also made a further material change to its assumptions from April: that unplanned outages would be 11GW, not the historical 12GW level that has been used for a number of years.

Kendal unit 5, which can produce 640MW and has been on a long-term forced outage, is the only possible explanation for this as it returns to service in April. It is unclear whether Eskom is continuing to plan using a risk level of 14.2GW in outages (or a likely risk scenario of 16.2GW), but this is not likely as it requires only 2.2GW as operating reserve.

In its system status report from last week, Eskom also for the first time published a forecast for the next year. It typically only shares a three-month view.

Its outlook shows that it expects a small shortfall to meet reserves in most weeks for the remainder of the year. When (or if) breakdowns do spike, load shedding is very likely (red) until April when that Kendal unit returns to service (Koeberg unit 1 returns to service in May).

Beyond that, the outlook does not look encouraging. Eskom’s ability to juggle breakdowns across an ageing fleet will determine just how much load shedding we endure until December.