South Africa’s electricity generation fell 5.2% year-on-year in September, while electricity distribution — labelled consumption in Stats SA’s summary — rose 1.5%. The latest release gives businesses a more complicated signal than either number alone: annual distribution improved, but distribution fell from August after seasonal adjustment.

Prepared and last checked: 6 November 2026. AI-assisted research and drafting; intended editorial reviewer: Tristan van Niekerk, review pending.

Key findings

  • Generation and distribution moved in opposite directions in the year-on-year comparison.
  • After seasonal adjustment, September distribution fell 0.9% from August, although the third quarter was 0.4% above the second.
  • Stats SA’s September accounting shows how generation, cross-border flows and power-station use combine to produce the distribution total. It does not establish why generation declined.
  • The figures measure electricity volumes. They do not establish a tariff reduction, a change in any particular business’s sales, or reliable supply at every address.

Evidence: three comparisons, three questions

Stats SA published its September electricity report on 5 November at 13:00. The reporting month is September, not November. Its results need to be read using the same reference periods and adjustments.

Electricity volume changes, September and Q3 2026
Comparison Generation (%) Distribution (%)
September 2026 versus September 2025, unadjusted −5.2 +1.5
September versus August 2026, seasonally adjusted +0.4 −0.9
Q3 versus Q2 2026, seasonally adjusted −0.4 +0.4

The annual comparison says distribution was higher than a year earlier. The monthly comparison says it lost ground after August. Both can be true, and the modest quarterly increase does not make September’s decline disappear.

How the September electricity account reconciles

Stats SA’s Table 7 provides the accounting bridge between electricity generated and electricity distributed. For all producers in September 2026, the published figures reconcile as follows:

September 2026 electricity accounting, all producers
Flow Treatment GWh
Generated Starting volume 17,826
Inflow into South Africa Add 594
Consumed in power stations and auxiliary systems Subtract 1,312
Outflow from South Africa Subtract 530
Distributed in South Africa Result 16,578

Arithmetic check: 17,826 + 594 − 1,312 − 530 = 16,578 GWh.

These are electricity-accounting flows, not proof of what caused September generation to fall. Stats SA reported that inflow was 78.4% higher and outflow 58.8% lower than a year earlier, while electricity used in power stations and auxiliary systems was 13.2% lower. Those movements help explain how distribution could rise while domestic generation fell; they do not establish the operational or commercial reasons behind any movement.

For an investor or business planner, that makes the figures a cross-check rather than a stand-alone signal to expand or cut capacity. A retailer’s sales, a factory’s orders and a mine’s production still need to be tested against their own evidence. National electricity distribution cannot identify which customers drove the change.

Why reliability is a separate question

In its 30 October operational update, Eskom reported a financial-year-to-date Energy Availability Factor of 68.15%, compared with 62.96% over the corresponding period a year earlier. That measure covers 1 April to 29 October. It is not the same period or measure as Stats SA’s September output.

A plant can be available without producing at its maximum throughout the month. Better availability and lower measured output therefore need not contradict each other. This does not prove why September generation declined; neither headline resolves that causal question.

SBL’s earlier report on Eskom’s diesel-free week examined operating performance and costs. The new Stats SA release adds a different test: how much electricity was generated and distributed.

Practical impact: decisions for households and businesses

  • Households: assess electricity costs using metered units, the applicable tariff and fixed charges. A national volume change alone gives no rand saving to put into a household budget.
  • Businesses with backup systems: use site-specific outage records, critical loads and backup costs when reviewing provision. This monthly release cannot establish the risk of a local interruption.
  • Businesses planning growth: compare the distribution signal with actual orders and production before changing staffing or equipment plans. The benefit is avoiding a costly decision based on one national headline.

Risks and limitations

The strongest counterargument to a cautious reading is the positive annual distribution figure and the quarterly increase: there is evidence of improvement on those comparisons. September’s monthly decline, however, prevents a claim of uninterrupted momentum.

These aggregate results do not identify winners and losers by business, establish profits, or quantify a household benefit. They also do not establish the risk of interruption at a particular site. Eskom’s operational figures are company-reported and cover a later, longer period.

What happens next

Stats SA’s advance schedule lists the October electricity release for 3 December 2026 at 13:00. The useful next test is whether the monthly distribution fall reverses and whether the quarterly improvement persists, allowing for revisions. A scheduled date is not a guarantee that publication will occur exactly then.

Sources and methodology

All table changes and GWh values are published Stats SA figures, not SBL forecasts. The reconciliation is a direct arithmetic check of Table 7. SBL analysis and practical implications are identified separately from source findings.

KEEP READING