ABSTRACT
BC Hydro is a committed user of quantitative risk assessment (QRA) techniques in its dam safety program. By considering the safety of dams from a failure driven perspective in conjunction with the traditional analysis driven approach, all aspects of safety are addressed and quantified in a consistent manner both for individual dams and between dams. BC Hydro has adopted qualitative risk assessment guidelines for review of reservoir slope stability which provide a comparison of the allowable annual probabilities of exceedance proposed for potential landslides to those adopted by Hydro for floods and earthquakes. However this paper presents a strong caution concerning full adoption of quantitative techniques for large singular landslides. BC Hydro’s experience is that while QRA can be very useful for situations involving rockfall hazard areas or stream valleys prone to debris torrents, one must be very careful in using this technique for large singular slopes or slides. In the initial category, there is opportunity for having or being able to obtain historical or statistical data. In the latter category there is almost no opportunity.
The emphasis usually must be on lessening geological uncertainty and evaluating geological complexity. Ongoing monitoring will point out where early judgments made with limited information are incorrect and where new problems can arise. The authors suggest that the evaluation of large singular slopes/landslides will continue to rely heavily on traditional, more deterministic approaches complemented by judgment, qualitative risk analysis and long term monitoring. While probabilistic analysis can be used to justify the selection and credibility behind each geological model proposed for a slope, it cannot supplant the important roles of empiricism and judgment in choosing the correct geological model.
