>What bollocks. I think the actual question to ask is how it's possible to create the conditions for an very large (the size of the mine)and extremely low density (the concentration of natural ore) nuclear reactor.
No bollocks involved - those laws depend on the fundamental constants. Scientists have speculated for decades about the possibility that these may have been slightly different in the distant past - and thus the laws of physics would not be exactly the same.
This is quite controversial, mavericky science because it's very hard to test - but it's actually become less so in the past 20 years or so because some evidence from astronomy (in particular the cosmic background radiation) is suggesting that they may have been slightly different in the very early days of the universe.
Oklo offers a chance to look more recently (on a universal scale) but still a long time ago - 2 billion years, about half the lifetime of the planet.
If there had been subtle and slight changes over the years - then 2 billion years ago should be enough to detect some - much smaller even than what cosmic radiation data has hinted at, but on the same line (that said there are other theories that could explain the radiation data - the question is unanswered at the moment since none of them have any other supporting evidence yet either).
Now there's no proof the fundamental constants have changed at all since the big bang, but there's no proof they haven't. For most physics it's perfectly adequate to assume they have always been constant, but if they weren't and we could determine that, it would change a lot of our understanding of physics - particularly the physics of the early universe.
By factoring in those different values we could possibly explain a lot of the other things which currently remain open questions.
So while it's unlikely - it's nevertheless and most decidedly NOT bollocks. It's maverick science for sure - but it's still science and still done according to the scientific method. If it yields results those results will be greatly valuable.
Just because there's a 99.999% chance your theory is a dead end, doesn't mean it's not proper science to damn well test it and make sure.