The promise of results from risk assessments is as seductive as the picture above. The results are as fuzzy and fantasy-based as the picture above. But doing risk assessments is as much fun as it would be to be a participant in the scene above. Let me explain.
We know that a deep deposit of tailings could take as many as five-hundred years to consolidate by self weight. Today I read a paper to be published at Paste 2014 that promises that if you add polymer to your tailings, the same deep deposit will take a mere thirty years to self-weight consolidate. Or if you insert wick drains into the tailings & place sand over the tailings, maybe you can reduce the five-hundred years to thirty.
Meanwhile you can quantify the risk of a breach in the perimeter dikes in thirty versus five-hundred years. Whatever way you do it, the risk is greater the longer the period of consolidation. Is the expense of polymer or wick drains low enough to ameliorate the longer-period risk? Is the risk that institutional controls will fail in five hundred years a justification for demanding action now to eliminate the risk of a perimeter dike breach four-hundred years hence?
The answer, I believe depends on your political outlook. We found on the UMTRA Project that Republicans overwhelmingly preferred rock covers to limit erosion of the closed tailings piles. Democrats on the other hand overwhelmingly preferred soil covers as being more “natural.” I bet small government folk would support one tailings consolidation approach and big government folk would prefer the alternative. Do a pole amongst your friends.
If you set as a design criterion that the tailings facility closure works should last for 1,000 years to the extent reasonable and at any rate for 200 years, you can estimate the probability of failure as a result of the big storm or the large earthquake, and sleep quietly at night. For what it is worth, we on UMTRA, simply designed for the probable maximum precipitation and the maximum credible earthquake.
Others, seeking to reduce closure costs, have formulated the design criteria for the closure works differently—and then used the criteria as a basis for risk assessments of dubious validity. One such formulation is: design for 200 years of good performance and tell us what the result is of the passage of 1,000 years. Applying this criterion as a risk period results in the need to design for small storms and tiny earthquakes. Even the story you need to tell about what can happen in 1,000 years is pusillanimous: again only slightly larger storms and slightly bigger earthquakes are the basis of the story. The rest is brushed away as a miniscule risk and therefore not worthy of attention.
Steve Vick has rejected this shorter period approach to justifying lesser closure designs. He takes the attitude that the closure works will be there forever–whatever that means. Thus the biggest storms, the biggest earthquakes, and countless cycles of vegetation growth & decay will occur. These and many other forces will, in the goodness of time, affect the closure works. Thus, he concludes, we as a society are obligated to provide closure works that will resist such forces.
This approach is enshrined in a document put out by consultants to the oil sands industry. A paper on their work is available in the proceedings of Tailings and Mine Waste 2013. I listened to the presentation when they said that closed oil sands tailings facilities should “respond to natural forces as does the natural topography in the vicinity of the site.” Noble, but achievable? Yes, if you add polymer or install wick drains. But at what cost?
Then consider the use of risk assessments as a way to compare alternative closure approaches. If you do the analysis on the basis of a mere 200 years and postulate continuation of institutional controls, there is little difference between alternatives no matter how much they differ. Conversely, if you consider infinite time (perpetuity) in your risk assessment, significant differences in risk of failure emerge. One alternative may be intrinsically geomorphically unstable and the other may be intrinsically geomorphically stable. An oil sand tailings facility of dewatered, non-flowable tailings is potentially geomorphically stable–maybe it will respond as the surrounding landscape to natural force. But tailings still undergoing consolidation, are susceptible to flowing out in the event of a breach of the perimeter dikes. The risks are different.
I know of no technical papers exploring these ideas. Yet I submit they are interesting & provocative enough to justify such a paper. And the implications of going one way or the other via risk assessments will be as profound as the sin or pleasure you see in the picture above and the humour in the picture below. Comment!