Not that it was put that far below ground for safety reasons, it was mostly to save on the amount of land required and to shield the detectors from pesky background radiation (not the other way around).
AFAIK, these areas are restricted while the LHC is operating, so there would be nobody near it (and I believe trying to enter these areas triggers an interlock which stops the whole system before you can get near it).
IIRC from my short time intersecting with the finer points of collider discussion, they locate them in such areas where generally if such things happen, it's a non-issue; but take great pains to ensure that such events don't happen, as it would require potential excavation and removal of a lot of hardware to replace the irradiated sections to restore a contamination free loop. It's also, if I recall correctly, part of why circular paths were chosen instead of things like ovals or capsules for the path shape among other factors I think like optimization for cheapest construction.
Any loss of beam containment would continue on a tangent ensuring one and only one segment of the collider would potentially have to be replaced; other shapes would end up with magnets that were relatively more catastrophic to have fail due to the excursion geometry having more opportunities to escape along a path near parallel to the track, allowing the irradiation of additional accelerator segments. Which would be a hell of a lot more expensive to remediate.
I sometimes laugh, because in the business world, people like to talk about what is 'reasonably forseeable'. I'm a bit of a stick in the mud at such times because I can generally say that exactly no one in the room is actually interested in ptobing the actual boundaries of the reasonably forseeable, as that would require doing too much math. I'm regularly right in that regard.