This communication presents some ideas and topics concerning risk analysis and decision making in Rock Engineering. It comes from the author's contribution to Workshop W1 entitled "Uncertainty, reliability and risk". At large, all geotechnical related sciences and, in particular, Rock Mechanics are paying increasing attention to risk analysis and trying to base their decisions in probabilistic approaches rather than in deterministic evaluations. This trend can be associate to the number of recent and near future Conferences, Symposia and courses where this issues were or will be discussed. Some examples can be listed: - International Symposium on "Assessment and Prevention of Failure Phenomena in Rock Engineering" (Istanbul, Turkey 1993), - ICE Conference on "Risk and Reliability in Ground Engineering" (London, U.K. 1993), - 1ST North American Rock Mechanics Symposium comprises a session named "Uncertainty and Reliability in Rock Engineering" (Austin, USA 1994), - "Acceptable risks and practical decisions in Rock Engineering", a short course by Prof. E. Hoek (Santiago, Chile 1994), - 8TH Congress of the International Society of Rock Mechanics includes a topic designated "Methods and case studies of safety evaluation and prediction" (Tokyo, Japan 1995). Since this paper is a contribution to the Workshop, a special concern was taken to provide to future newcomers a relatively large number of bibliographic references. However, it is not intended to be a thorough list of all published works pertaining to risk and reliability in Rock Engineering. When looking for the 3 words that make up the theme of the workshop in a common english dictionary, some interesting and unexpected engineering related comments can be found: Uncertainty
insecure, precarious, doubtful, problematic
inability to predict or to place confidence in the unknown
not firm, stable or safe
likely to give away or fail
suggest great liability to failure or exposure to imminent danger Reliability
dependable, trustworthy, infallible
satisfactory performance can be expected
with complete confidence
worthy of being trusted and believed in Risk
venture, peril, jeopardy
the hazard or chance of injury or loss
a hazardous or dangerous chance of loss
the degree of probability of such loss.
At a first look it can be noted that the common everyday interpretation and use of these words is somehow different from their technical interpretation (Blockley 1992, Shu-Ho and Ming-O 1992). Uncertainty is simply lack or imprecision of knowledge. It begins to arise when, for a given purpose, part of the real world is represented, with inherent loss of complexity, by a model. Uncertainty increases when judging the parameters that are imputed in the model and in the methodology used to evaluate the performance of the model. Some examples often faced in geotechnical engineering are: - rock mass heterogeneities (weak zones, faults, etc.) have to be located using only an extremely limited amount of data - estimation of the in situ rock properties faces several problems, such as the inability to reproduce the correct field conditions.