Change the world


This article appeared in the Business Day of 25 November 2016.

Koeberg nuclear power just outside Cape Town. Picture: SUNDAY TIMES

It would be near impossible to construct a nuclear power station safely at Thyspunt, near Jeffrey’s Bay, because of deep, hidden canyons in the bedrock covered by sand and soft rock, a geological study has found.

The study was corroborated by another from a PhD candidate at Nelson Mandela Metropolitan University that detects evidence of a large paleo-seismic event along an Eastern Cape fault line 10,000 years ago. It looks set to blow a hole in Eskom’s environmental impact assessment study for the area that found the area technically safe and environmentally friendly for the construction of a nuclear power station.

Thyspunt, 90km from Port Elizabeth, is one of Eskom’s top two preferred sites for a nuclear power station. It comprises a relatively flat area of hard rock close to sea level, covered by wetland and fynbos.

Eskom applied for a licence for the site with the National Nuclear Regulator in March.

The study was the work of a geology master’s student at the university. Professor Maarten de Wit, director of the Africa Earth Observatory Network, has written a report summarising the findings of the two in which he warns that the risks at Thyspunt have not been properly examined and that the evidence is overwhelming that the site is unsuitable.

The first study reveals what he describes as "a simple but to date unexpected message from the subsurface: cut into the hard bedrock on which the nuclear station must be built, are paleo-valleys and canyons hidden below the … dune sands and soft sedimentary rock".

The study discovered four such previously unknown paleo-canyons that extend inland well below the present sea level, he says. At Thyspunt, the canyon cuts into more than 1,000m² of bedrock to a depth of 16m below sea level.

"When I saw her [the master’s student’s] work I was shocked. Eskom looked at all this data and they missed the most important thing," he said.

It implied the plant would need to be built on, or adjacent to, an area that is well below sea level, De Wit said.

"The hard rock is almost 20m below sea level. Just one big sweep of sea could dislodge that sand and rock," he says.

Rising sea levels, surging storms due to climate change, and far-field tectonic-induced tsunamis pose dangers in the presence of the canyons.

The new work implies that the hazard for the southern to southeastern Cape should now be upgraded, to include local 7.0-magnitude events

Eskom responded to the report on Thursday by saying it was confident its seismic hazard assessment was adequate. The methodology used had taken account of various alternative models and interpretations, including the "extreme scenario" in De Wit’s report, it said.

"Eskom will also apply conservative decision-making in the design and operation of a future Thyspunt nuclear installation. It is confident that the data used and knowledge gained through the Thyspunt [probabilistic seismic hazard analysis] project is accessible to the broader scientific community, and that it has contributed to our understanding of the local and regional geology and the hazard posed by seismic events,"

The paper quotes studies that show the sea level has been rising along Nelson Mandela Bay at about 2mm a year over the past 36 years and may rise between 1m and 2m by the end of this century as the power station nears the end of its life. Nelson Mandela Bay is also particularly exposed to storm surges.

"From the above, one may conclude that the chance of seawater penetration into buried canyons and valleys beneath the Thyspunt dunes is more likely to increase into the future rather than stagnate or decrease," the report reads.

 The PhD study explores the fault system that runs about 600km along the Cape mountains to Coega, and several fault systems nearby. The area has witnessed historic earthquakes such as the 6.3-magnitude Ceres-Tulbagh quake in 1969.

De Wit says new "paleo-seismicity work" outlined in the thesis confirms the fault "can be best considered ‘active’"; and a major earthquake and surface rupture occurred about 10,000 years ago along the Kango Fault segment, close to De Rust.

A conservative estimate of the magnitude of this paleo-earthquake is between 6.9 and 7.2, the study reads.

"The new work implies that the hazard for the southern to southeastern Cape should now be upgraded, to include local 7.0-magnitude events.

"If in future, for example such a similar earthquake occurs along the extended St Croix Fault offshore, it is likely to generate a large submarine slump, and possible significant local tsunami that would affect the coastal region, including Thyspunt," the report reads.

De Wit criticises Eskom for making little of its own technical work available on an open-source data bank, which would allow other scientists to scrutinise it. This, he says, is neither transparent nor international best practice.

The risks are compounded by indications from Eskom that it does not intend building a state-of-the-art nuclear plant that would include all the post-Fukushima safety features standard in GenerationIII reactors.


Contact information
Prof Maarten De Wit
Tel: 27 41 504 2277