Theme D

Risk Analysis – assessment of reliability for concrete dams

Formulators: Fredrik Johansson (KTH/SWECO), Marie Westberg Wilde (KTH/ÅF) and Luis Altarejos García (Universidad Politécnica de Cartagena)

Reliability-based methodology has yet found only limited application to dam safety assessment, even though it is well known that the same numerical value of a factor of safety may imply very different safety margins in the actual design. A larger factor of safety does not necessarily mean a smaller level of risk due to the presence of large uncertainties in the design environment.

In the last 15-20 years the attention and application of reliability-based methodology for concrete dams is increasing. One example is the 2011 ICOLD Benchmark where one of the themes was to estimate the probability of sliding failure of a concrete dam. The differences among the results presented by different teams were significant. The reason was that the engineers had to make decisions and choose between different options at some points of the process of finding the solution and that even a simple problem may exhibit relatively high degrees of freedom.

In 2016 a reliability-based methodology for design and assessment of concrete dams founded on rock was published in Sweden. The methodology can be seen as a first attempt to put together in a consistent way some of the rules, regulations and explanations necessary for design and assessment of concrete dams from a probabilistic point of view.

In the 14th ICOLD Benchmark Workshop participants are asked to define limit states, estimate probability of failure, sensitivity values and system reliability of a concrete gravity dam using the “Probabilistic model code for concrete dams”. The goal is that the outcome will further improve probabilistic assessment of concrete dams and, in the future provide a framework for unified reliability calculations and thus increase the interest and confidence in risk analysis.

 

The instructions and data can be downloaded from the following link