Since the problems in the Japanese nuclear reactors have been publicized, the safety of nuclear power plants comes into question everywhere.
How are North American reactors in terms of earthquake safety?
Are nuclear reactors in Southern Ontario at risk of earthquake damage.
The Toronto Star has published its answer. No,
they say. The basis of their argument is that southwestern Ontario (particularly the Lake Ontario and Lake Huron basins) are not known for large earthquakes.
It is well known that the western margin of North America is seismically active. Less well known is the potential for seismic activity in eastern North America.
Here is the seismicity map for eastern North America from the
USGS. Sourced from
this document.
Right away we see four hot areas--in South Carolina; in Missouri; near the mouth of the Gulf of St. Lawrence; and along the Ottawa River. These areas represent the locations of sizable earthquakes in the past 200 years.
Four significant shocks (magnitude 7-8)
occurred in the New Madrid area between December 1811 and February 1812.
A
major earthquake occurred in South Carolina in 1886.
The Charlevoix seismic zone is the most active in eastern Canada.
Significant earthquakes have occurred along the Ottawa River from Montreal to Temsicaming.
The map is entirely dependent on the past record of earthquakes. Prior to 1811, it is unlikely that central Missouri was considered a seismic hazard area. The next major earthquake, wherever it occurs, will generate its own spot on the map.
To evaluate seismic risk, we need to evaluate structures. In eastern North America, a major plane of weakness intersects the earth's surface along the St. Lawrence River, and passes through Lake Ontario and Lake Erie. It passes southward, and follows the trace of the Mississippi River to the Gulf of Mexico. A failed rift arm extends through Lake Huron and Lake Superior (these structures go a long way towards explaining why these features are where they are).
The main structural feature passing through Lakes Ontario and Erie and up through the St. Lawrence is the same structure on which the near magnitude 8 events occurred in 1811-12. We therefore must recognize that there is potential for events of similar magnitude in southwestern Ontario. The reason why this is largely unrecognized is because such an event has not occurred in recorded history.
One of the charming things about earthquakes is that they demonstrate scale invariance. In particular, the size distribution of earthquakes in any given area shows a
1/f distribution. This relationship has long been observed in seismic studies--the famed
Gutenberg-Richter law is an example of an empirical law which has been shown to have predictive value.
If you have an area where historical records are short, and you have only observed small earthquakes, it is still possible to make use of the scale-invariant nature of earthquakes to predict the recurrence interval of earthquakes larger than noted in the historical record. Using this technique, Arsalan Mohajer calculated that the recurrence interval of a
magnitude 6.5 earthquake in western Lake Ontario was 1000 years. (see figure 2).
Figure 2 of Mohajer, 1993.
True, he doesn't actually state the numbers in the report, but the extrapolation of the data in the figure above gives us a recurrence interval for a magnitude 6.5 earthquake to be 1000 years.
Assuming that the power plants are to last 50 years, there is a 1 in 20 chance of an earthquake of this magnitude affecting them. The hazard assessment for earthquakes in Ontario is based more on the official data, which notes the lack of any such events in the past 250 years.
Reference
Mohajer, A. A., 1993. Seismicity and seismotectonics of the western Lake Ontario region.
Geographie physique et Quaternaire,
47:353-362.
