Earthquakes at Koeberg?
2010-05-05 07:40
Nuclear power stations (NPSs) and earthquakes do not make a happy combination.
For my sins, I recently read the seismic risk assessment specialist report that forms part of the draft environmental impact report (EIR) for Eskom’s proposed Nuclear1 project. Eskom wants to build NPSs at Thyspunt near St Francis Bay in the Eastern Cape, Bantamsklip near Hermanus in the Western Cape and Duynefontein, the site of the existing Koeberg NPS outside Cape Town. The draft EIR is currently available for public comment and can be found here.
A red herring?
You might think that worries about earthquakes are unfounded in South Africa, but, as the report acknowledges, there are historic records from 1809 of a major earthquake with an estimated magnitude of 6.3 less than 25km from the Duynefontein site. The possibility of a recurrence of a similar seismic event is high and in the words of the report's authors, it would constitute "the most serious geo-scientific threat to a NPS". Not an unsubstantiated concern then...
What’s in a number?
I was a bit surprised to find that the report uses only a single quantitative parameter to evaluate the seismic risk at the three sites: the expected "peak ground acceleration" or PGA. The PGA values presented in the report are 0.16g for Thyspunt, 0.23g for Bantamsklip and 0.30g for Duynefontein. In the simplest terms, the higher the calculated PGA value for an area, the more severe the largest earthquake that can be expected there.
The authors of the report tell us that a PGA of 0.3 is "typically used in the seismic design of NPSs". I find the fact that the PGA value determined for Duynefontein is identical to this typical design specification cause for considerable consternation. I don’t know about you, but a bit more leeway between the expected PGA for the area and the specification to which a future NPS is built would be desirable.
But wait, it gets worse...
No errors
Numerical parameters such as the PGAs quoted in the report always have associated statistical errors. No such errors are provided. An earlier report commissioned for the ill-fated Pebble Bed Modular Reactor Demonstration Power Plant which was proposed for the Duynefontein site gives a slightly lower PGA of 0.27g. Significantly, it also provides us with an error: ±0.14g. If we make the reasonable assumption that the error associated with the PGA value in the current report is in the same ball park, the PGA for Duynefontein would be 0.30±0.14g, meaning that, within error, it could reach 0.44g - substantially higher than the typical specification for a new NPS!
Worried yet? Well, it gets worse...
Outdated methods
The authors of the report tell us - without a hint of irony - that the method used to calculate the PGAs doesn’t conform to the latest international standards and that the values may actually increase once additional, up-to-date studies have been conducted. The only thing we can take from the report seems to be that the seismic risk at Duynefontein is relatively higher than that at Bantamsklip which in turn is relatively higher than that at Thyspunt. Exactly what the PGAs are at each of the sites remains a mystery.
In summary, the single quantitative measure of the seismic risk at the proposed sites provided in the report is outdated, but could be significantly higher than the rating typically used in the design of NPSs, especially in the case of Duynefontein.
A nonsensical conclusion
Having ploughed through all this I expected the report to conclude that the seismic risk at the proposed sites remains undetermined and that until such time as it can be ascertained properly, no NPS should be built at either of them. The actual conclusion came as a jaw-dropper: "the seismic hazard does not preclude a standard export NPS at any of the proposed sites". What? Talk about snatching a predetermined answer from the jaws of all evidence to the contrary. Frankly, the whole affair smells like an exercise in rubber-stamping.
You can read my detailed comments about this report, which I’ll be submitting as part of the public participation process, here.
- Andreas manages Lobby Books, the independent book shop at Idasa's Cape Town Democracy Centre.
Send your comments to Andreas
Disclaimer: News24 encourages freedom of speech and the expression of diverse views. The views of columnists published on News24 are therefore their own and do not necessarily represent the views of News24.
- News24