When conducting consequence assessments for dam safety, current industry practice is to select a single representative design event for each flood AEP of interest. These events are then run through the dambreak hydraulic model and consequence model, producing a single potential loss of life (PLL) estimate for each flood AEP. By picking a single representative design event for each flood AEP, the variability in design floods that are modelled in hydrological studies is lost when conducting consequence assessments. For large catchments or dams with extreme consequences it may be warranted to quantify the impact on PLL estimates associated with this variability in design floods.
This paper explores three anonymised case studies where multiple design storms were run through the consequence model for each flood AEP to better capture the impact of storm variability on PLL estimates. The spatial and temporal distributions of rainfall on the catchments were defined using spacetime patterns of rainfall from historic storms at each location. Craig et al., 2018 details the methodology on deriving local spacetime storms to use for design flood modelling. The consequence modelling is done in a dynamic simulation model, LifeSim (Woodrow and Needham, 2022), where the evacuation of individuals is modelled explicitly with respect to the flood behaviour.
The three case studies show that there is large variation in the PLL estimates dependent on which rainfall spacetime pattern was chosen for each design flood. For Case study 1, the impacts of two different spacetime patterns were assessed using LifeSim modelling. It was found that for the same flood AEP, there was a 95% difference in PLL due to the different spacetime patterns. For Case Study 2, four spacetime patterns were modelled and it was found that PLL varied by up to 125% across the four spacetime patterns. Case Study 3 modelled four spacetime patterns as well and showed a 40% difference in PLL.
The three case studies presented show the potential variation in PLL due to storm variability which can have a major impact on dam safety risk assessment. While the large variation in PLL seen in the three case studies are largely due to the specific hydrological catchment configuration and the population at risk due to dam failure, it shows that different design storms for the same flood AEP should at the minimum be considered when conducting consequence assessments.
$15.00
ANCOLD is an incorporated voluntary association of organisations and individual professionals with an interest in dams in Australia.