A method for adjusting design storm peakedness to reduce bias in hydraulic simulations

Abstract

In the UK, decision makers use hydraulic model outputs to inform funding, connection consent, adoption of new drainage networks and planning application decisions. Current practice requires the application of design storms to calculate sewer catchment performance metrics such as flood volume, discharge rate and flood count. With flooding incidents occurring more frequently than their designs specify, hydraulic modelling outputs required by practice are questionable. The main focus of this paper is the peakedness factor (ratio of maximum to average rainfall intensity) of design storms, adjudging that this is a key contributor to model bias. Hydraulic models of two UK sewer catchments were simulated under historical storms, design storms and design storms with modified peakedness to test bias in modelling outputs and the effectiveness of peakedness modification in reducing bias. Sustainable drainage systems (Suds) were implemented at catchment scale and the betterment achieved in the modelling outputs was tested. The proposed design storm modification reduced the bias that occurs when driving hydraulic models using design storms in comparison with historical storms. It is concluded that Suds benefits are underestimated when using design rainfall because the synthetic rainfall shape prevents infiltration. Thus, Suds interventions cannot accurately be evaluated by design storms, modified or otherwise.