Electricity in South Africa: Postcard from a Dystopian Future

REVIEW ESSAY
Truth to Power: My Three Years inside Eskom
by André de Ruyter
Penguin, 2023, 407 pages

For more than a century, anyone in South Africa lucky enough to have electricity has gotten it from Eskom. Born of a mash-up of the anglophone Electricity Supply Commission (escom) and the Afri­kaans Elektrisiteitsvoorsieningskommissie (evkom), the utility today is formally Eskom Holdings SOC. That appendage, SOC, is something readers of South Africa’s business pages may see with the regularity that one comes across LLC and LP in the private-sector-dominated United States; it denotes that Eskom is a state-owned corporation.

Eskom once stood as the preeminent example of a successful utility in the southern hemisphere. The Afrikaner government of South Africa was appalling in many ways, yet it allowed Eskom to be run in businesslike fashion, with the driving objective of producing plentiful electricity in support of the mining and industrial cash cows of South Africa. P. W. Botha, the country’s last strongman of apartheid, nevertheless demurred to a reproach by Eskom’s then CEO, who reportedly told him, “Presi­dent, you keep out of Eskom, and I’ll keep out of government.” After apartheid ended three decades ago, and Nelson Mandela’s African Na­tional Congress took the reins of power, Eskom continued to deliver, recognized in the early 2000s as the world’s best electric-power utility in a Financial Times awards ceremony.

Eskom has since fallen in grand fashion. On most days of the year in 2023, Eskom deployed its protocols for “load-shedding”—a term of industry jargon, though now grimly familiar to South Africans, used to describe a controlled loss of electricity service to customers resulting from an insufficient supply to meet demand. The job of a power-grid operator is usually to send signals to dispatch “up” or “down” power stations to meet variations in demand. During load-shedding, the operator turns its focus to managing demand, switching off parts of the grid that serve particular clusters of customers. These outages are then rotated around the wider grid until available supply once again can meet overall demand. If demand were not curtailed in this manner, the entire grid would collapse, as its system frequency of fifty hertz plunged. A so-called blackstart of the grid is a process that would take weeks or months, a prospect that would cause civil unrest in most places.

Load-shedding in South Africa has been going on for fifteen years, with escalating severity in the past few years. The one exception has been the months leading up to the country’s national elections in May, when electricity service—as if a miracle had occurred—suddenly re­sumed for over a month. More on that later. For most South Africans, normal life has meant familiarity with Eskom’s eight stages of load-shedding, each representing a level of demand on the grid that needs to be cut.¹ Indeed, one of the most popular digital apps in the country, whose name is a vulgar pun that blends “Eskom” with a homophone of Afrikaans slang for female anatomy, informs its users when and where load-shed has been announced, along with stage and implementation information.² There is not another example in the world of a previously reliable utility, serving a relatively modern economy, so comprehensively failing to provide its basic service. This failure announces itself seemingly everywhere.

The Powers That Be

Every country has a unique context around its development. So too South Africa, and then some. After apartheid, South Africa’s government launched an ambitious campaign to electrify the have-nots of its grossly unequal society. Building out the grid succeeded, but ensuring enough energy was on the grid to meet the new demand did not. Government ministers promised affordability in power supply, but this wasn’t possible when new capital expenditures at high interest rates were layered atop the aging, but cheap, infrastructure Eskom had depreciated over the decades. And this led to another problem.

It would already have been a struggle to supply more energy to the Rainbow Nation, but with age came the diminishing reliability of South Africa’s existing power plants. This was made much worse by corruption that took hold of one important sector: coal. South Africa has one of the most carbon-intensive power sectors in the world. With the exception of a single nuclear plant and a few hydroelectric assets, the country is more than 80 percent powered by coal, with most generating stations clustered in the country’s resource-rich Northeast.³ Coal has to be mined, transported, and burned to produce steam that spins turbines that produce electricity. There are many inputs and processes involved in the business of coal; the bureaucratic surface area to which corrupt actors may attach is vast. Also, the coal industry, in the places it does exist, is often the only game in town. It is also heavily unionized, and in South Africa, the liberation struggle—and the postindependence govern­ment for three decades, until just months ago—was constituted by a tripartite mash-up of the once exiled, outlawed, and imprisoned African National Congress, the South African Communist Party, and the country’s umbrella group of trade unions, cosatu. Ideologically, there was just as much interest in Eskom remaining in state ownership as there had been during the time of Afrikaner nationalism. Of course, there was also a very practical interest in this structure, since there were so many new mouths to feed among the country’s politically ascendant and nouveau riche, a class that was one and the same in the post-apartheid republic. Eskom and its suppliers today represent a conduit for these interests’ satisfaction.

A number of enterprising gambits have sprung up over the years to loot the utility. A single cargo of fuel is often delivered multiple times, a practice called “round-tripping,” with the driver passing by a (bribed) security guard, ostensibly unloading that fuel to a (bribed) procurement agent, then leaving the plant still fully loaded to drive around for a few hours before purporting to deliver a new cargo. Coal trains, the only affordable way to convey the fuel in the bulk required by power stations, have been sabotaged on behalf of these trucking cartels. High-quality coal has been siphoned off of stockpiles for sale on the lucrative export market, leaving Eskom’s own coal piles mixed in with heavy payloads of dirt and rocks. The quantity and quality issues alike take their toll on power plants’ operation.

Apart from its corrupted supply chain, the state-owned enterprise itself is hardly a meritocracy. “Power plant manager” may sound like a highly technical job title. It can also be a sinecure. André de Ruyter, Eskom’s CEO until early 2023, relates a dismal state of affairs in his extraordinary memoir Truth to Power: My Three Years Inside Eskom. Largely absentee plant managers, when they did show up, were dudded up in the pointed-toe dress loafers of a flaneur, instead of steel-toe footwear. The tone at the top trickled all the way down, and de Ruyter relates the basic failures of management and simple housekeeping that further imperiled Eskom’s ability to deliver power safely and reliably. The result of the twin threat of the corrupted supply chain and derelict workforce is that Eskom power plants’ equivalent availability factor (“EAF”)—a measure of a facility’s readiness to produce energy—sat at a shade above 50 percent last year. Similarly aged plants in the United States usually have EAFs in at least the 70 percent range, often higher.

Thus, while de jure national policy holds that the state-owned enter­prise Eskom should be devoted to making electricity widely available on affordable and reliable terms, the de facto policy of the country treated Eskom as the bounty for particular interests, closely adjacent to state actors, by dint of its state-owned characteristics. To be sure, this prob­lem goes beyond electricity in South Africa, where just as the public has become familiar with the once niche term “load-shedding,” so too has a bit of jargon from social science—“state capture”—become a commonplace of newspapers’ reportage. This year marks thirty years’ distance from apartheid, and at this point the country’s politics still often seems an exercise in rewarding the liberation struggle’s aging personalities—and their relations, friends, and financial patrons. The class of winners in the system is just as narrow, if not as racially limited, as during apartheid.

De Ruyter and others tried to put a halt to the sloth and corruption they found in various corners of Eskom and its hangers-on. When the formal channels of law enforcement proved lackadaisical or unwilling to pursue the graft, he sought funding from some of Eskom’s own customers—South Africa’s largest business interests—to fund a private investigation. He also took up the begging bowl in his visits abroad, helping to convince Western funders and the World Bank to channel $8.5 billion to South Africa on the condition that it build out gobs of wind and solar. For the West, this was a decarbonization maneuver, but in-country, the funding is poised to have the more significant benefit of expanding power supply and unchaining the sector from a nearly exclusive reliance on the corrupt coal supply chain. By investigating corruption and diversifying away from a politically favored industry, it is not mere supposition to say that de Ruyter got too close to powerful interests. His own political backers shied away from him. And then, after drinking a cappuccino before a meeting on December 12, 2022, at his office, André de Ruyter found himself losing his recall of basic words, struggling to breathe, his heart racing—he had been poisoned with cyanide. He lived, but resigned and left the country, spending the next year in another hemisphere, at Yale’s School of Management.

Coping with Load-Shedding

Eskom management and the state itself have been incapable of stopping the slide into unreliability, so individual citizens have taken power into their own hands. The well-worn adage about necessity’s maternal characteristics has rarely found a truer expression than during South Africa’s load-shedding crisis. In addition to the new information economy that conveys status updates to the population about the grid’s availability, a variety of consumer devices big and small have gained fast adoption in the country. The failure of the state-owned enterprise has led to a set of decentralized solutions for the “haves,” and workarounds for the “have-nots.”

Anyone who can afford one has what, in country, is called “an inverter”—which in South African parlance means any form of on-premises back-up generation. While in Cape Town this February, as load-shedding was occurring on a daily basis, I noticed hardly any trouble with electricity supply whatsoever. My friend’s rental home overlooking beautiful Camp’s Bay was fitted out with rooftop solar panels and two large batteries in the garage that seamlessly switched on when the grid-delivered power went off. The only one in the household affected was my friend’s toddler, who woke up crying to the blinking light of his baby monitor, which flashed on sensing a brief power interruption. Personally, I had noticed load-shedding only twice. Once, I was having drinks with, as luck would have it, an Eskom engineer and a solar developer. The lights went out at the hip bar on clubby Kloof Street where we were conferring; this establishment relied not on batteries, but a larger backup generator, so the latency between Eskom’s power-off and the back-up’s power-on was more like two minutes at the bar rather than two seconds at the residence. On the second occasion, I visited a charmingly rustic, organic vineyard in Stellenbosch which had had its credit card processor fried by a voltage excursion caused by a power interruption on their rural circuit; while booking my wine tasting in advance, I was told to bring cash. The inconveniences I faced were comedically small given the gravity of the situation. As a visiting tourist, I existed near the upper echelon of a society whose inequalities have been sharpened by the failure of this state public service. For me, it’s a beautiful country, safe enough, and inexpensive due to the relative levels of inflation that have battered the South African economy more than the American one. I’ll happily be back again, unless it all goes the way of Zimbabwe.

Most South Africans are not so fortunate. For those without backup power, ordinary life has to be timed around anticipated load-shedding; just check the app. Most load-shedding is announced in advance because demand is projected to exceed supply, and thus demand can be cut on relatively predictable schedules. But if the weather changes, or a “forced” outage at a power plant occurs, as opposed to a “planned” or “maintenance” outage, that will mean deeper cuts and ruining plans that had been formed around a more modest stage of load-shedding. In practice, load-shedding, in general, and the particular unpredictability of this tight margin of supply-and-demand balance have rolled back the clock on a basic middle-class existence that relies on refrigeration, a Wi-Fi router, and a television. Restaurants offer a “load-shed menu” to denote what they are capable of cooking during an outage. In order to preserve some lighting during the evening hours of load-shedding for their homework, students in modest circumstances rely on an invention I have never seen anywhere else: LEDs with tiny batteries attached to them, which sense resistance coming from the grid and switch on when grid-delivered electricity is unavailable.

Happily, there are attempts at orchestrating some of these decentralized solutions to benefit wider communities. Eskom itself and some of the municipal power companies that exist one rung below Eskom (and which depend on Eskom’s generation and transmission to supply their customers) have implemented “load-limiting” programs that use ad­vanced metering technology and Wi-Fi to ratchet the use of particular devices in customers’ homes, avoiding the bludgeon of load-shedding which plunges everything into the dark. Load-limiting can apply to what in South Africa are termed “geysers”—pronounced, hilariously, “geezers”: hot water heaters, a big source of electricity consumption for residential customers. Also, pool pumps: the affluent do their part by deferring their swimming pools’ refreshment in order to spare a little light for the vast underclass. The load-limiting program that Cape Town began this year will contain about thirty thousand residential meters and is expected to be good for sixty megawatts of demand reduction, enough to move off one stage of load-shedding, albeit just a small fraction of the capacity of a single utility-scale power plant. Cape Town is buying surplus generation from customer-sited solar and battery storage as well.

South Africa’s New Generation

Done well, a public utility levels up a citizenry. But in South Africa, the utility is a prominent symbol of the class divide that puts the lie to the ANC’s grand promises of a transformed and modern South African republic for all people. After a long period of invulnerability and complacency, it was load-shedding perhaps more than anything else that has finally spurred an electoral feedback loop that put the ANC on its heels.

Leading up to the country’s elections this May, many public con­veniences appeared as if by magic: refuse collections resumed, port-a‑potties appeared where public toilets had been out of service for years, and in the power sector, Eskom operated for nearly two months with­out resorting to load-shedding. The ANC and the opposition Democratic Alliance (DA) bickered about the cause of this miracle, the latter pointing to huge run-ups in diesel purchases and cuts to plant maintenance outages so that they could run go-for-broke style in supplying power. It was a strategy tied to an election campaign, rather than durable operations, the DA argued. In any case, a reprieve in load-shedding did occur, but not enough to erase the public’s memory. The ANC, for the first time since apartheid, lost its parliamentary majority and, at least temporarily humbled, accepted a coalition government with the Democratic Alliance and other minor parties. Neither side appears especially pleased about the arrangement, and it remains to be seen how it will work out. In the cabinet announced to run the country in June, there are some modest positive signs: The sprawling Ministry of Public Enterprises was terminated, and oversight for Eskom was concentrated in a new ministry—Electricity and Energy. That new post took additional powers from the Ministry of Energy Resources and Energy, led by the recalcitrant Gwede Mantashe, a union organizer and com­munist who opposed private investment in the sector and personified the conflicted relationship between coal interests and their major cus­tomer; it was right there in the ministry’s name.

Personalities and officialdom aside, the ANC acceded to a policy recommendation that was at least two decades old, at last deciding that the country’s dependency on a politically compromised and financially distressed gatekeeper, Eskom, needed to end. A measure to restructure Eskom, spinning off its electric transmission function, and creating a more competitive market for the purchase and sale of electricity, would place South Africa in the company of the majority of countries that have broken up their legacy power monopolies. Parliament passed legislation enabling these changes before the national elections, though the reform package as of this writing awaits presidential assent, and it could be returned to the reconstituted legislature for further revision. If not a heartfelt repudiation of the state ownership that was a central tenet of both the ANC and its Afrikaner predecessor’s philosophy of government, it is a practical one: only a paltry 150 megawatts of new generation (of a 25,000-megawatt demand system) had been added to the grid in the five years of Mantashe’s leadership through the old state-owned chan­nels. By contrast, the thousands of megawatts of capacity optimistically projected to come online in the next several years are overwhelmingly sourced by either direct, private purchases by customers from generators, or at-scaled Eskom purchases from third-party generators who bear the risk of cost overruns and failures to perform on their contractual obligations.⁴

There is still a huge mishmash of personalities and their associational attachments—ministries, regulatory agencies, state-owned industries—that could slow any item of the emergent reform agenda down, and the tenuousness of South Africa’s current political arrangement could doom big changes altogether. Many South Africans want to see immediate successes in an industry that often takes years to steer to meaningful reform and investment. The impatience for getting big, long-lead items accomplished tends to refocus governmental priorities on short-term band-aids. This reality may not bode well for a government that seems unstable. Meanwhile, it has never been entirely clear who specifically should be held to account for Eskom’s recent failures, but the public was collectively able to hold the ANC to account. Now that the party cannot be exclusively to blame for such failures due to its power-sharing arrangement, the internecine blame game may well be amplified. Whether South Africa’s power sector can revamp to reliably serve the country as a whole will be a strong indicator of whether the liberal reforms at hand can be workably located in a modern democracy and prominent member of the Global South.

A Postcard from a Dystopian Future

South Africa is a unique polity in many ways. It may be tempting to dismiss Eskom as a peculiar sidenote of South African history. That would be wrong. The challenges of the country’s power sector, and the power sector’s intersection with, regulation by, and ownership through the state, strike some eerily familiar chords with the American experience.

First, the debate between U.S. political factions has lately trended toward tribalism around certain technologies. Once generally united behind principles related to outputs of a well-managed grid, such as ubiquitous access, reliability, and affordability, America’s political parties have arrayed behind supposedly rival technologies that are inputs to the grid. Democrats talk up renewables and storage. Republicans have adopted nuclear as an avatar for their rebuttal on “the energy transition.” They also advocate for existing and new fossil capacity, both do­mestically and in the name of American influence through gas exports. South Africa presages this turn of American politics, where political movements and patronage networks are built around fuel supplies and power technologies, and these political networks then largely predetermine more essential conversations about how government should structure and regulate the industry in its service to customers.

Second, the stories of corruption in the South African power industry are quite colorful. But so, too, are they in the United States. Where utilities maintain a monopoly, and look to politicians as their primary customer, a common arc exists. Indeed, the biggest political corruption scandals of this century so far in the United States have featured regulated utilities. In Ohio, the utility FirstEnergy was the dark‑money financier of a $60 million political operation of the state’s House Speaker Larry Householder, who in return pushed through a package of subsidies to the utility’s nuclear and coal power plants. If anything, South African politicians were more discrete in their influence over Eskom. In Ohio, after bulldozing the opposition and passing the corruptly begotten legislation, known as House Bill 6, one utility lobbyist texted around an image of two utility executives, Householder, and a friendly regulator, their faces photoshopped atop Mount Rushmore, accompanied by the message “HB6 Fuck anybody who ain’t us.” Speaker Householder was sentenced to twenty years in prison last year, a lobbyist for the company died by suicide as the authorities closed in, the former chairman of the Public Utilities Commission of Ohio chose the same exit, and FirstEnergy’s former CEO and a government affairs executive will face trial.

At around the same time as the scandal in Ohio was occurring, much the same thing was going on in Illinois, albeit with greater subtlety—Chicagoland having a depth of experience in this stuff—but with much the same outcome: a gargantuan legislative package benefitting the incumbent utility, Commonwealth Edison, whose executives administered a patronage network to employ ne’er-do-well relatives of legisla­tors and the Democratic machine in no-show jobs. Four ComEd executives were indicted and convicted, a prelude to the trial on federal racketeering charges of former House Speaker Mike Madigan, which is set to begin this year. If the character names were Zulu or Xhosa, it could’ve been a story about Eskom. The closer the power industry is to government, the more it presents an opportunity for corruption no matter which hemisphere you live in.

Third, one may ask how utilities owned by the state, and in this case the national government, become captured by such local interests—even besides the overt corruption in question. Clearly, Eskom’s nation-building mission did not extend into a post-apartheid South Africa. Instead, it became captured by relatively local interests, where the coal sector had its roost. We may ask the same question about how the purposes of nationally owned utilities in the United States have trans­formed. Set up during the New Deal, and providing the energy to produce crucial materiel during the Second World War, the Bonneville Power Administration, the Tennessee Valley Authority, the Western Area Power Administration, the Bureau of Reclamation, the Army Corps of Engineers, and a couple more obscure agencies own or operate power generation and transmission infrastructure in parts of the United States. This is usually dominated by hydroelectricity on federal water­ways, but extends to certain coal, nuclear, and gas assets. These government-owned enterprises, which once built up America’s national might, now are largely devoted to relatively parochial interests, where the views of Wenachtee more than Washington dictate outcomes. Ironically, because of their local clienteles, these entities tend to be last in line when the regions in which they exist see movements to stitch together a more national grid. That is not necessarily a bad thing, but anyone who sees in America’s federally owned utilities a vindication of the national interest is very poorly informed. Of all federal agencies, theirs is the most attenuated dotted line to Washington, D.C. And so too Eskom, which in its power generation is concentrated in the coal industry based in Mpumalanga in the country’s Northeast. None of this, again, is surprising: state ownership of power utilities may have at its inception a broad national motivation, but the inputs into or customers of the sector tend to be geographically concentrated, and its long-term existence often is subordinated to these local interests.

And fourth, and finally, one should consider the hoped-for global decarbonization of the power sector. Remarkably, despite its supply inadequacies, South Africa plans to close certain coal-fired power sta­tions; they have fallen into such disrepair that, in view of the emissions they would produce if repaired and the enormous expense of doing so, it may be better to close them. This will put South Africa further still into the hole of supply inadequacy versus the demand on the Eskom system. U.S. grids are not so dire, and Americans not so inured to power outages, as to brook the outcome one does see and will probably continue to see in South Africa.

There is nevertheless a serious risk that America’s fossil plants are retiring so quickly that they cannot be replaced with the quantity and quality of new resources that will ensure similar levels of reliability as what could be expected under the status quo ante. After years of mostly reassuring messaging, interspersed with modest caveats about the tech­nical and economic feasibility of the ongoing transition toward a cleaner grid, this year has witnessed a notable change in the pitch of grid-reliability warnings from the people actually responsible for reliability. Take America’s largest grid, the Mid-Continent Independent System Operator, which for several years has issued a relatively soothing “Relia­bility Imperative” report. This year’s begins with a cover letter from MISO’s CEO John Bear that does not mince words, starkly warning in boldfaced font, “The transition that is underway to get to a decarbonized end state is posing material and adverse challenges to reliability.”⁵

On the demand side, America has not for many decades had a largely unelectrified population, like South Africa did after apartheid, whose connection to the grid gave rise to new sinks of demand. Yet on the horizon loom big new loads on America’s grids: electric vehicles, an electrification of heating, electrolytic hydrogen production, and most of all entirely new demand for energy from data processing for artificial intelligence and other applications. New demand introduced onto al­ready tight systems will stress the ability of the United States to produce adequate power for everyone and, should that become a problem, the issue of how supply inadequacy of grid-delivered power will be socialized broadly. Load-shedding is written into the protocols that govern American power systems, just as it is in South Africa’s. If load-shedding were to be implemented with any degree of regularity in the United States, it would naturally cascade into the same system of haves and have-nots as one sees in South Africa, in terms of those persons and industries financially able to secure backup supplies in defense against it.

Eskom’s recent history can teach us many things. It teaches us that a power sector too close to government, ironically, may not best serve the public interest. It suggests that state capacity reverts to localism. It illustrates what happens when the power sector fails—not in one heaping collapse, but gradually through diminished reliability, with a resulting social stratification between those able to privately do something about it and those who cannot. It is a warning about a politics of energy that debates attachments to particular technologies and fuels, rather than on a political and regulatory structure that orients toward outcomes of reliability, affordability, and basic competence. It is a postcard from a dystopian future we must avoid.

This article originally appeared in American Affairs Volume VIII, Number 3 (Fall 2024): 115–25.

Notes