The recent earthquakes near Cape Town have brought renewed focus on the risks of running a nuclear power station 27 kilometres from the city centre. Over a weekend in September 2020, a 6.2 magnitude quake occurred 1 600km off the coast of Port Elizabeth, as well as a 2.5 and then a 2.7 magnitude in Durbanville, about 20km from the Koeberg nuclear power station.
This was followed by a 3.6 quake in November about 60km offshore from Koeberg. The plant was built near what used to be called the Milnerton fault line, but recent research has uncovered further faults which comprise what is now referred to as the Table Bay fault zone.
Koeberg was built during the late 1970s with a designed operating life of 40 years. It
started generating electricity in 1984, and so is scheduled to be shut down in July 2024. But Eskom has plans to extend this by another 20 years, pending approval by the National Nuclear Regulator. In addition the minister for energy issued a section 34 determination in August to authorise a second nuclear plant to be built at the Koeberg site. This has been sent to the National Energy Regulator for approval.
Koeberg was the first nuclear plant built with earthquake protection in the foundations in the form of a concrete raft resting on neoprene rubber pads. Later designs included horizontal fail safe buffers, but this was not done for Koeberg. That means that if these rubber pads don’t perform as expected, or if there is an unexpectedly powerful earthquake, it could result in the foundation cracking.
These pads have now been in a marine environment for 44 years, and it is not clear what condition they are in. Extending the life of Koeberg would mean these rubber pads are expected to last for more than 68 years. This is particularly concerning in the light of the history of inadequate maintenance at Eskom plants, and specifically at Koeberg. A damning 2006 report by the National Energy Regulator found, among other problems, negligence, inappropriate risk assessment and a “complete failure to notify supervisors of abnormalities”.
To demonstrate that the Koeberg emergency plan can be carried out the regulator does an exercise every two years. This is intended to show how, should it be necessary, tens of thousands of people could be provided with potassium iodate tablets, evacuated using the MyCiti buses and drivers, decontaminated, and taken to temporary shelters such as the Bellville Velodrome. The plan calls for some major roads being converted to one way, and for parents to evacuate without collecting their children from school, because transport will be provided to evacuate them. The plan does not cover where all these people would then be housed, nor where their children would be able to go to school. Other potential effects such as the loss of the Western Cape tourism, wine and fruit exporting industries have never been quantified.
In the past, members of the press and representatives of civil society organisations have been present to observe these emergency exercises. In an unprecedented move, Eskom and the regulator decided that no observers were allowed at the exercise on 8 October this year. Despite claiming to be able to carry out an actual evacuation of tens of thousands of
people safely during a pandemic, they say that a handful of observers would have been at risk due to Covid-19.
Also being kept out of the public eye is a study by Eskom into the seismic risks associated with the Koeberg site. The regulator acknowledged this had been submitted to it by Eskom in 2017, but rejected a request for the study. If it came to reassuring conclusions, it seems likely that Eskom and the regulator would have made it public by now; it is not reassuring that both bodies choose to keep it secret. Adding to the uncertainty, a tender was issued by the Council for Geosciences (with a closing date of 6 November) for a new study into the seismic hazards at the Koeberg site.
If experts were so confident the risk was small enough to be ignored, there would be no need to spend money on a new expert study.
But could South Africa afford to do without Koeberg? Eskom’s status report for the week up to 4 October showed a reserve margin of between 2% and 18%. A well run utility should have a minimum reserve margin of about 15%, and hopefully Eskom will make progress towards that under new leadership.
Koeberg generates only 2% to 5% of our generating capacity and this is important in view of the threat of loadshedding. But, with a 10% or 15% reserve margin, it wouldn’t be significant, and hopefully soon it won’t be.
Although large earthquakes are much less frequent around Cape Town than in some other areas of the world, history has shown they do happen. It is more a question of when, rather than if, and the recent cluster of seismic events may be an indication of more to come.
Those in the nuclear industry will say how Koeberg is built to withstand an earthquake. If you had asked the Fukushima nuclear plant manager In February 2011 whether the plant was built to be safe from earthquakes, what would he have said? Or if you had asked the Chernobyl plant manager if that plant was safe from human error?
Theoretical safety is one thing but the reality is the world had three major nuclear reactor meltdowns in the period 1979 to 2011. Each of these was from something unexpected, so it is no reassurance to hear that Koeberg is expected to withstand an earthquake of an expected magnitude.
With ageing infrastructure, a seismic hazard report that is being kept secret, Eskom’s poor record of plant maintenance and a questionable emergency plan, every year Koeberg is kept running is a roll of the dice.
The effect of doing without Koeberg will be low, and the potential consequences of seismic damage to a nuclear plant so close to Cape Town would be catastrophic. A risk and benefits analysis of extending the life of the old plant at Koeberg by 20 years shows it is simply not worth it.