How Eskom stopped the blackouts: What South Africa's turnaround means for utilities that harvest depreciation

Eskom, the state-owned utility that powered South Africa through years of rolling blackouts, has reported its strongest system performance in nearly a decade: 441 consecutive days without load shedding as of July 2026, with a daily Energy Availability Factor (EAF) of 82.04% on 26 July, the highest since 2017.^[1] Unplanned outages fell below 5,000MW for the first time in eight years.^[5] Diesel spending collapsed to R807.41 million year-to-date, down 85.67% from R5.63 billion in the same period last year.^[8] The turnaround was not funded by a surcharge or a rate increase passed through to customers. It was built on a Generation Recovery Plan focused on returning existing coal plant to service through sustained, disciplined maintenance.

That sequence matters to every ratepayer in a jurisdiction where a utility is claiming it cannot afford grid upkeep without a hardening rider, resilience surcharge, or rate bump. Eskom's generation fleet was not broken because it lacked money; it was broken because maintenance spending had atrophied while the utility extracted dividends and let unplanned outages climb to 9,700MW in the worst years of the crisis. When political pressure forced a shift in accountability, the utility's own engineering teams were able to restore 5.9GW of capacity, not by building new plant but by returning deferred units to full operation.^[5] The lever was operational discipline and a credible threat of sanctions for non-performance, not a new revenue stream.

The mechanism Eskom deployed was simple in concept but alien to investor-owned utilities in the United States: a target for plant availability, a recovery plan with named milestones, and consequences for failure. The financial year-to-date EAF improved 7.28% year-over-year and 11.8% compared to three years earlier.^[5] More than 85% of the coal fleet now operates at EAF levels between 73% and 97%.^[5] The company did not request ratepayers to fund the same infrastructure twice (once through depreciation allowances in rates, once through a "hardening" rider after years of deferred maintenance). Instead, it reallocated its existing budget toward the plant that mattered most and enforced accountability.

In U.S. regulatory language, this is what a performance-based reliability incentive mechanism (PIM) looks like in practice. The utility earns its return on a demonstration of outcome, not on the volume of capital deployed. Hawaii adopted a version of this framework in 2020, and Britain's RIIO model (Revenue = Incentives + Innovation + Outputs) has run symmetric PIMs since 2015, placing revenue at risk against reliability targets and restoring the link between what a utility collects and whether customers' lights stay on. The absence of such a mechanism in most U.S. states means that a utility can let reliability degrade, then request a surcharge to fix it, then request a rate increase to fund the next cycle of deferred maintenance. Eskom's recovery proves the alternative works, and it cost shareholders and management credibility, not customers' bills.

The question for regulators and ratepayers now is straightforward: if Eskom, a state utility carrying years of budget pressure and political dysfunction, can restore 5.9GW of generation capacity through disciplined maintenance alone, why should an investor-owned utility be permitted to harvest depreciation allowances, distribute them as dividends, and then request a hardening surcharge when the deferred maintenance fails in a storm? The source of the funds to restore reliability is already in the rate base. The missing ingredient is the enforcement mechanism.