Historical Storm Events as an Alternative to Design Storms for SCMs

Throughout the stormwater industry, center-peaking design storms historically are the foremost method for managing peak flow. As the availably of historical rainfall data begins to grow, alternative and possibly more appropriate design approaches are becoming of interest. With 1-minute rainfall data readily available, it is possible to design Stormwater Control Measures (SCMs) using historical rainfall events. The benefit of using historical storm events to determine peak flow is the greater understanding of how the system will perform during a real event distribution as opposed to the overly conservative design storm S-curve. Design storm peak intensities are chosen based on maximum the gage has recorded, a peak intensity often associated with convective events of small depths, which may not be appropriate to represent in a frontal 2-year/24-hour storm of long duration. This research evaluates the use of historical rainfall events of similar depth to the design storm instead of traditional design storm distributions for SCM design. Airport rain gages in four cities in PA (Philadelphia, Erie, Harrisburg, and Pittsburgh) between 2005 through 2020 are explored to find historical rainfall events with approximately equivalent rainfall depth (between NOAA 14 median depth and upper 90% confidence interval). In Pennsylvania, the 2-, 5-, 10-, 50-, and 100-year/24-hour storm events post-construction peak flows often must be managed back to the preconstruction peak flows. Duration of historical rainfall events are limited to 24 hours or under. A rain garden SCM and detention basin in series are modeled in EPA SWMM to compare the differences between the two methods (i.e. design storms vs historical storms).