Professional opinion concerning the usefulness of risk analysis in dam safety practice is almost as divided now as it was in the early 1980s

It should not come as a surprise that risk analysis in dam safety practice is controversial because it involves the mathematics of probability. These are the words of Dr Des Hartford, specialist engineer in hydroelectric risk management at BC Hydro in Canada.

In his paper ‘Risk analysis in geotechnical and earthquake engineering state-of-the-art and practice for embankment dams’, presented at the Fourth International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Hartford argues the case for probabilities based on scientifically valid analysis as opposed to subjective engineering judgements of probability.

Despite the widespread publication of papers on quantitative risk analysis for dams over the past 20 years, risk analysis is still not widely used in dam safety practice. Recent experience even suggests that although it generated a great deal of enthusiasm in the 1990s, professional opinion concerning the usefulness of risk analysis in dam safety practice is almost as divided now as it was in the early 1980s.


Hartford traces the evolution of quantitative risk analysis in dam safety practice. However, before even discussing the applications of risk analysis in this area, the author believes that the engineering profession should first develop an understanding of:

  • What risk is and, in particular, what dam risk is.

  • How risks posed by dams can be analysed.

  • What are the experience and qualification requirements for participants in the risk analysis process?

  • What are the strengths and weaknesses of approaches?

  • What do the outputs of a risk analysis for a dam mean?

  • How can methods and results of risk analyses be validated?

  • How can the results of risk analysis be used in dam safety decision-making?

  • What are the roles and responsibilities of all participants in the overall process?

In its most complete sense, risk analysis characterises the uncertainty that is inherent in the answer to the question: how safe is the dam? Risk analysis itself is not a decision-making process but an integral part of a risk assessment. Risk analysis and risk assessment add new dimensions to dam safety decision-making as they involve explicit characterisation of the uncertainty that pervades all aspects of dam safety decision-making, as well as the complex concept of societal risk.

Hartford says it is important to note that, with one or two notable exceptions, risk assessment aimed at determining if a dam is safe enough is generally not being carried out. And despite the proliferation of published papers on the subject in recent years, in general there is no formal regulatory acceptance of proposed risk analysis and risk assessment methods by dam safety regulators.

In his paper Hartford discusses the role of risk analysis in dam safety decision-making and management from a broad perspective. He reflects on the experiences of a regulated owner (BC Hydro) which has formally embraced the concepts of risk management in dam safety as a matter of corporate policy. He also draws on his ten years of experience of experimenting with the various proposed approaches, subjecting them to critical review, rejecting some and improving others and, ultimately, subjecting the most satisfactory approach to a simulated test of (risk) regulatory acceptability.

Historical perspective

Hartford gives an historical perspective of risk analysis, focusing on the principal stages in the development of this branch of engineering as applied to dams. He goes on to describe the latest attempts to quantify risk associated with earth dams for two failure modes: seismically induced liquefaction and a proposed procedure for seismically induced non-liquefaction deformation failure. Focusing on the Hugh Keenleyside dam, Hartford gives an abridged account of the quantitative seismic risk analysis for the dam.

In the conclusions of his paper Hartford restates his concern about the questions surrounding quality and reliability of contemporary literature on risk assessments for dams (IWP&DC September 2000, pp18-23). The author warns that there are risks associated with this if dam owners only rely on their interpretations of the literature and/or those who write on the subject. There is a clear need to improve the quality of literature to raise the level of the debate concerning risk analysis of dams, as well as improve professional practice.

‘If these vitally important issues are not addressed,’ Hartford warns, ‘scepticism and lack of confidence in risk analysis in dam safety practice will become even more entrenched.’