Uncertainty analysis of the rational method in stormwater management - A case study in Gråbo, Lerum municipality, using multi-criteria analysis

Abstract

Stormwater management is an important aspect for municipalities in Sweden and it is of high importance to have a sufficient stormwater system to ensure sustainable cities in regard to economic, social and environmental aspects. The rational method is the most widely used approach for estimating runoff rates in urban catchments. The method incorporates four parameters: rainfall intensity, catchment area, runoff coefficient and a climate factor. Depending on the estimations that are made for the inputs, the calculated, dimensioning flow will have an uncertainty attached to it. No evaluations of the extent of the uncertainty could be found prior to this work. Therefore, the aim with this study was to identify the potential uncertainties in the input parameters of the rational method and to evaluate what effect these uncertainties might have on stormwater management. Segerstaden is a residential area in Gråbo, Lerum municipality, Sweden, that has problems with unwanted water accumulation. This study investigated suitable stormwater measures for Segerstaden and evaluated the measures with a multicriteria analysis. The results from an uncertainty analysis of the rational method were integrated in a multi-criteria analysis to evaluate what effects parameter un certainties in the rational method might have when choosing stormwater measures. The uncertainty analysis was made by means of Monte Carlo simulation in Excel using the ad-in software @Risk. The study showed that all parameters in the rational method bring some uncertainty to the dimensioning flow as they are based on estimations. The magnitude of the uncertainty showed to be significant in this study, with the 5th percentile flow equal to 118 l/s and 95th percentile flow equal to 756 l/s. The rainfall intensity was shown to have the largest impact on the overall uncertainties in the rational method followed by the catchment area. However, all parameters contribute to the overall uncertainty and the magnitude of it will vary depending on how rough the estimations for the input parameters are. Since the stormwater measures are designed based on a dimensioning flow, the uncertainties in the rational method will have a direct impact on the design of the measures and hence the stormwater management.