On the need to go beyond bucket modelling to evaluate NBS risk of failure under urban constraints

Abstract

The need to combine simple analytical solutions to control the full spectrum of stochastic events, not just extreme rainfall events related to flood risk, in order to restore the natural water cycle has led to the adoption of probability analysis (PA) as a powerful design tool for urban drainage systems. Applications of PA to sustainable urban drainage and nature-based solution (NBS) design are numerous but often rely on a bucket modelling (BM) approach to assess the efficiency of retention and detention structures.Managing urban growth and climate change requires careful attention to 1) the shape of NBSs to fit into an existing urban grid and 2) the intensity of rainfall, which is expected to increase significantly. The BM approach is insensitive to the shape of technical solutions and the variability of rainfall intensity, so the PA was revised, coupled with a hydrodynamic modelling (HM) approach and compared with previous BM-based approaches.In the low-intensity rainfall regime, HM and BM predict completely different infiltration system (IS) overflow probabilities. In the high-intensity precipitation regime, both provide a bell-shaped probability density function of overflow volumes, although BM tends to underestimate IS overflow volumes with low shape factor and overestimate IS overflow volumes with high shape factor.