I'd assume that "geologically stable" was one of the requirements when they first made a list of potential locations.
Fortunately, all of the heavy-water CANDU reactors currently in commercial service in Ontario are fueled by unenriched uranium, so the worst possible outcome shouldn't result in much more contamination being released into the environment than we would see with a natural uranium deposit of comparable size nearby. Which isn't nothing, but the result would be a smallish statistical increase in cancers, not people dropping dead from Acute Radiation Syndrome. Many industrial sites do more damage and are less scrutinized, but we get all weird about radiation the moment the word "nuclear" comes up, and have a hard time putting the risks in perspective.