Yesterday evening we had a fine dinner and then to The Railroad Club to listen to live music and dance some. I got drunk, so left the car in the office parking and took a cab home. Today I rode my bike down to the SeaBus and hence to the office to collect my car. For years I have safely navigated the gravel path beneath the bridge. Today, for no good reason, I kept to the left as I came around the blind corner. As luck would have it there was another cyclist coming the other way. I hit the brakes, slammed into him, and as the front wheel brakes took before the back wheel brakes, went over the handle bars. No great harm except torn skin on the right knee and some pains in otherwise places.
The lesson learnt is that even if we have done it before, we can err, and an accident can occur. Calculate the statistics that on this very day, at that exact time, I would go to the left and there would be another cyclist. Never seen another cyclist coming that way in years. I had gotten to believing the path was mine—although it is a public place through a lonely park besides the river.
This strange statistic set me thinking about the presentation made just before me and the one I made at the conference Tailings and Mine Waste 2013. At this link is the blog posting by Franco and Cesar Oboni of Riskope. Here is the link to the full paper. In essence they say: Mine tailings dams have failed at a lesser rate than nuclear power plants. At least the rate of failure of tailings facilities is less than the rate of failure of nuclear power generating facilities. The fact that there are so many more tailings facilities failures than nuclear power plant failures is the reason it seems that the failure rate is greater—its just that there are more tailings facilities and hence for a give failure rate, more failures. We are not instinctively able to distinguish numbers and rates (probabilities).
Here is the link to the presentation I made immediately following Franco Oboni. I tell of the three 2012 tailings facility failures I picked up on the web. On the Friday preceeding the conference there was a failure of a tailings pond in Alberta that poured coal tailings into the Athabasca River. A commenter on my presentation reminded me of a failure in Quebec earlier in 2013. This is what I found on the web about that 2013 tailings failure:
Aurizon Mines Ltd. (ARZ), the gold producer that agreed in March to be acquired by Hecla Mining Co. (HL), fell the most in more than four years after authorities reported a dam broke at the Casa Berardi mine in western Quebec, spilling thousands of liters of contaminated water. Aurizon dropped 0.7 percent to C$4.23 in Toronto. The shares earlier plunged as much as 12 percent, the most intraday since December 2008, before they were halted. About 150,000 cubic meters (40 million gallons) of liquid and 15,000 cubic meters (530,000 cubic feet) of solid material were spilled yesterday, Quebec’s Ministry of Environment said in a statement posted on its website. While a breach of an internal tailings dam spilled liquids and “suspended solids,” most of the material was contained within the tailings-pond area, Aurizon said in a statement today. “A residual film of material has affected a limited surface area of outside the external dyke,” the company said. Environmental authorities are on site and there won’t be any impact on operations, it said. The spill is in an isolated region a few hundred kilometers south of James Bay, said Stephanie Lemieux, a ministry spokeswoman. The ministry was informed of the spill late yesterday and an investigation began today to determine the limits of the damage. There is a creek about 1 kilometer (0.6 mile) from the site of the leak that’s being assessed, she said.
I know of three near-misses of tailings near-failures in 2013. Can’t write about them, for they are confidential. We have yet a month and a half left of this year. Let us hope there are no more tailings failures this year. But as Franco pointed out to me after our presentations: “Three in 2012, and now two real and three near in 2013, is about the historical average. Still a lower rate per dam than per nuclear power plant.”
This is strangely comforting, but strangely discomforting. I tells me that just as I crashed and tumbled off my bike today, any tailings facility could fail tomorrow. As a result of getting caught up in the statistics, getting overconfident, or simple trained-operator inattention–going to the left instead of going right.
What is comforting is that the highly irregular tailings industry–an industry far less regulated and overseen than the nuclear industry, has so low a rate of failure. Are we smart, lucky, or just dealing with a more robust entity? I leave you to decide. Read the Oboni paper. Read my paper. Consider the stats. And let us keep working hard to keep tailings facilities safe.