2024 – How Do We Assign Population at Risk? A Critical Assessment of Threshold Levels

Nicholas Thomas-Kinsella, Chris Nielsen, Rob Fowden

ANCOLD Guidelines on the Consequence Categories for Dams (2012) (the ANCOLD guideline) describe methods to establish population at risk (PAR) from credible dam failure events. Specific hydraulic thresholds for flood depth, velocity and hazard are applied. Regulatory authorities adopt the same hydraulic thresholds as ANCOLD, or modifications to them, to determine whether to enforce regulatory oversight of dam safety standards and to apply conditions on their management.

The approach described in the ANCOLD guideline is sensitive to selection of the natural (no failure) flood event, which strongly influences the number of incremental, or dambreak, PAR. To account for this, some regulatory authorities nominate an additional hydraulic exclusion of a threshold depth difference between the natural flood event and the same natural flood event combined with a dambreak failure. This threshold depth difference appears to have a clear, documented justification, however the value assigned to it is somewhat subjective and can strongly influence PAR calculations.

In this paper, sensitivity testing of a range of hydraulic threshold parameter assignments was conducted using recent consequence analyses for a range of dams of various types, sizes and geographical locations. The results highlight sensitivity to the assignment of the threshold depth difference and flood hazard thresholds applied in the ANCOLD guideline approach and some regulatory authority guidelines. An alternative approach for determining PAR is proposed that considers the hazard categorisations widely used in flood risk investigations in Australia, based on research and documented in Australian Rainfall and Runoff. The approach is more closely and intuitively aligned to techniques used to establish incremental impacts using flood modelling tools and there is a clearer physical basis for assigned hydraulic threshold parameters. Sensitivity testing suggests that a shift to a hazard category approach would not, for most dams, impact PAR numbers to the extent that would change consequence categories.