TRISTEN TAYLOR: Eskom didn’t know just how dangerous Koeberg is

On July 19 the National Nuclear Regulator granted a 20-year lifespan extension for Koeberg’s unit 1. The decision also revealed that unit 2 is not in great shape, so much so that the regulator will only license it up to 2033.

Why? In its decision the regulator states that “the safety of the containment building is not fully demonstrated for unit 2 for the period of LTO [long-term operation]”. 

Eskom has consistently said in engineering reports, in its LTO application and in its statement to the public that both of Koeberg’s units will last another 20 years. Once the LTO licence was granted, we would be set to fearlessly harness the power of the atom for another two decades. 

But that’s not the case. The regulator’s reports on the LTO licence decision reveal a chain of events that led Eskom to conclude that unit 2’s containment structure was fine, when in reality we could have ended up joining the small but terrifying list of major nuclear accidents.

In the case of a full or partial meltdown there are a few layers of defence within a nuclear plant. The first is the pressure vessel, which encases the reactor’s core. As long as the extremely radioactive material remains inside the reactor pressure vessel and the emergency cooling system works, calamity can be more or less managed.  

However, if the reactor vessel doesn’t hold, problems compound within the nuclear plant, especially as the meltdown will produce considerable quantities of steam and hydrogen gas, the air pressure inside the containment building will rise. 

At this point, the containment building (what you see from the street) has to be able to withstand the increase in pressure. How much pressure the containment building can hold is an engineering question that revolves around the strength and integrity of the building’s concrete. If the pressure approaches the containment building’s threshold, the now extremely panicked nuclear operators might attempt to vent the radioactive gases into the atmosphere, though this is not a guaranteed “fix”. 

Radioactive gases were vented at both Fukushima and Three Mile Island to reduce the pressure arising from the respective full and partial meltdowns. The venting worked at Three Mile Island. At Fukushima, venting led to hydrogen explosions, compounding the catastrophe. The Japanese have already spent $82bn on cleanup and decommissioning costs, and will need another $82bn.

So, to be as sure as possible that Cape Town won’t be turned into a disaster zone, Koeberg’s containment building must be able to withstand the rise in internal air pressure coming from a meltdown. The structural integrity has to be there. Eskom has said this is the case; that the containment building would hold. 

As part of the licence process, the regulator reviewed Eskom’s engineering reports, including its assessment of the containment buildings’ structural integrity. In the regulator’s LTO safety evaluation report, it states that “data limitations, including inconsistencies and gaps, prompted Eskom to rely on engineering judgment and data analysis to reinterpret the data. These methods introduced discretionary criteria that are not part of surveillance procedures.” 

One of the data limitations, according to the regulator, was “the complete absence of [containment] dome data for unit 2”. To determine if unit 2 was good to go for 20 years, Eskom took data from unit 1’s containment structure and extrapolated it to unit 2. Take a moment here ... Eskom made an educated guess about one of the most important safety aspects of a nuclear power plant. 

Heaven wept iodine-131. 

In a much-delayed response to Business Day, taking two weeks to answer nine simple questions, the regulator went into spin mode, stating that its decision did “not imply that the containment dome would not keep/maintain integrity in a worst-case scenario”. 

Whatever should we make of this? It seems to contradict the regulator’s statements in its decision that “the structural safety margins for the dome have been eroded to the minimum” and that unit 2 was only validated “for a period of eight years, considering worst-case scenarios, assumptions, uncertainties introduced through data analysis and use of engineering judgment”. 

A bit on the subtleties of the English language here. Would is not might. For, if one cannot prove safety beyond eight years, then at the very least one cannot be certain of the containment’s integrity. 

The technical analysis of the plant’s structural integrity exists in 12 key documents. Eskom hasn’t responded to Business Day’s request for the documents. At first, the regulator agreed to provide some of them but then backtracked. Both organisations have reverted to secrecy mode. The public is deemed unfit to know what is going on in our ageing nuclear plant. 

There is a serious financial consequence to the regulator’s decision. Assuming that unit 2 will be granted an LTO licence in November 2025 and the licence will last until only 2033, the decommissioning of unit 2 would have to start then. In eight years from now, unit 2 will go from a source of income to an expense lasting for decades. 

How much will decommissioning cost? Business Day’s repeated queries to both the regulator and Eskom on this question have received no answer. Maybe they don’t know. The International Atomic Energy Agency estimates $500m-$1bn per generic reactor and a decommissioning timeline of 15- 20 years. 

The standard method of ensuring there are funds for decommissioning is to put a proportion of the income from the plant’s electricity sales into a dedicated ring-fenced account. Eskom’s 2020 decommissioning strategy states that “nuclear decommissioning expenditure is funded from the Eskom pool of funds”. 

In other words, plunder Eskom’s miserable balance sheet. 

There has, fortunately, been some improvement since 2020. As part of the LTO decision, Eskom must now ring-fence funds for decommissioning. The regulator stated in its LTO safety assessment that Eskom has “engagement with a banking institution for the creation of this new [ring-fenced] bank account”.

On August 22 the regulator told Business Day that Eskom has opened a bank account. Progress of sorts. But horrible engineering judgment and the inability to open a savings account for more than 40 years aren’t the sum total of Koeberg’s problems. A broken containment monitoring system, overdue repairs to concrete cracks and other issues of wonder and delight emerging from the LTO decision all await. Stay tuned. 

• Dr Taylor, a freelance journalist and photographer, is a research fellow in environmental ethics at Stellenbosch University.