A novel safe-fail framework for the design of urban stormwater drainage infrastructures with minimal failure and flood severity

Abstract

The urban stormwater drainage structures (SDSs) are conventionally designed for their failure recurrences rather than addressing the consequences of the failures. Various methods that consider the effects of flooding in SDSs design largely rely on the overflow volume and do not consider actual inundation characteristics due to computational challenges. Hence, advanced methods are required to design SDSs while incorporating flood consequences. We introduce a novel method called the 'safe-fail framework' for designing SDSs, which determines the best design solutions that minimize flood consequences with computational efficiency. The safe-fail framework involves event-based simulations of urban flood inundation properties, iteratively to evaluate the flood consequences until the design return period satisfies the safe-fail criteria with acceptable failure consequences. We demonstrate the approach in an urban catchment in Chennai, India, by estimating the optimal design return periods of the SDSs to minimize the flood consequences. A structure-variable approach that reduces the complexities of multivariate problems to univariate to identify flood-causing extreme rainfall events was explored. We show that 12 conduits (among 30) in the catchment require design return periods ranging between 3 and 6 years (vis-a-vis initial 2 years) to achieve the safe-fail condition, evaluated through a newly developed Flood Severity Index. The results reiterate the inappropriateness of using a common design return period across the catchment due to the varying sensitivity of locations to flood flows. The safe-fail framework can potentially improve the SDSs design criteria while accounting for urban flood properties and deriving improved urban water resources management decisions.