Combined sewer overflows (CSOs) cause environmental problems and health risks, but poor guidance exists on the use of rainfall data for sizing optimal CSO control solutions. This study first reviews available types of rainfall information as input for CSO modelling and, secondly, assesses the impacts of three rainfall data selection methods (continuous simulation, historical rainstorms selected based on rainfall depth or maximum intensity and IDF-derived storms) on the estimation of CSO volume thresholds to control in order to reach specific seasonal CSO frequency targets. The methodology involves hydrological/hydraulic modelling of an urban catchment in the Province of Québec (Canada). Continuous simulation provides the most accurate volume estimations and shows high sensitivity to the number of simulated *Revised Manuscript with no changes marked Click here to view linked References 2 years. Alternatively, when historical events extracted from rainfall data separated by a minimum inter-event time (MIT) criterion are selected based on their total rainstorm depth, the CSO volumes are underestimated significantly; whereas an analysis based on rainstorm maximum intensities over durations similar to the time of concentration provides more conservative volumes. Finally, synthetic storms constructed from multiple points of an IDF curve tend to underestimate slightly the CSO volumes, but provide acceptable results compared to single point derived storms. It was found that the overflow structures local characteristics had a marginal influence on results obtained from continuous simulation compared to event-based simulation. The use of design rainfall events should thus be restricted to preliminary assessment of CSO volume thresholds, and the final volume estimation for solution sizing should be reviewed under continuous simulation. The innovative contribution lies in the improvement of modelling procedures for solutions design to achieve a maximum CSO frequency, such as specified by many regulating agencies.

An innovative optimization‐simulation framework is applied to a case study of the Province of Quebec, Canada, to optimize the spatial distribution of green infrastructure (GI), the capacity and location of gray infrastructure, and the parameters specific to real‐time control (RTC) operating rules of a sewer system for reducing combined sewer overflows (CSOs) frequency and volume. GI, gray infrastructure, and RTC are applied either individually or in integration through eight optimization scenarios which are simulated over a nine‐year period of historical rainfall data. Among all scenarios, spatial optimization of GI with RTC leads to maximal CSO volume reduction (98%) and is the most cost‐effective option analyzed (70$/m3 of seasonal average CSO reduction compared to 140$/m3 for the scenario involving gray infrastructure alone). However, it requires a high GI implementation level and the CSO frequency under this scenario is sensitive to varying GI design parameters. The findings suggest that the best alternative for CSO control is the integration of the optimization of green and gray infrastructures with RTC as it still provides high CSO volume reduction (95%) and remains a cost‐effective solution (90$/m3 of CSO reduction), while providing robustness under cost and design uncertainties.

No abstract

No abstract