At the conference in Vail earlier this week, somebody at dinner said this: “I am sick of hearing people talk about other people’s tailings impoundment failures. Why can’t anyone stand up and say ‘this is why the dam I designed failed’?” [I hope I get the punctuation right.]
I designed a slimes dam once. I stood a while. Then it failed. Here is what I wrote about it in some obscure publication. I cannot now say it any better.
Construction of the new Bafokeng impoundment was underway, and we were sent to Kimberly to consult on a new slimes dam for the Big Hole. For as long as anybody could recall, the slimes had been discharge from a single point and allowed to flow wherever it wished. And this it did; it had flowed miles over the gently sloping ground. Because it showed no intention of stopping, understandably, many years before, some enterprising slimes-man had pushed up a long one- to two-meter high dike along the perimeter to catch the stuff.
But even that was now in danger of being overtopped. The mine had decided they needed a new slimes dam. The obvious place was on top of the old one; preferably near the discharge point where in theory the slimes should have been coarser. But this is a diamond tailings—it is not coarse in any part of the gradation curve. The sun had, however, beaten down for years and years and dried out the tailings, forming a reasonable foundation.
We designed something with drains and pool control and all the bright new ideas we were flush with from the Transvaal. But the miners would have none of these new-found ideas from the liberal north. No Drains—We Will Control The Water was their mantra. It took only one lunch at the Kimberly Club to persuade us to acquiesce.
The dam was built and operated for a few years. But the inevitable occurred: the miners lost control of the pool which came too close to the edge and it all failed. No real environmental harm. The flowing-failing materials ran out of the breach and simply continued on down and over the old tailings that had been deposited for decades in an uncontrolled way. The replacement dam included drains.
This case history along with the story of the Bafokeng dams reinforced the need to provide drains, control seepage within the tailings outer slopes and limit the quantity of water on the top of the impoundments. None of these lessons learnt fully entered the system and some years later, after we had both left South Africa, the slimes dam at Merriespruit failed for much the same reasons as had caused the failure of the dams we described here.