EVALUATION OF SPATIAL AND TEMPORAL VARIABILITY OF STORM EVENTS: APPLICATION TO CSO FACILITY DESIGN

In the state of Washington, one method of evaluating combined sewer overflow reduction methods is via the utilization of models that simulate historical rainfall and subsequent runoff, sanitary flows, and collection system hydraulic dynamics where the simulated combined sewer o overflow frequencies and volumes are evaluated to assess regulatory compliance. The results of this work showed that the storm data (and the underlying storm dynamics) at local City of Spokane rain gages, which only have 6 years of data, differed statistically significantly from, and could be correlated with, precipitation measured at the regional Spokane International Airport rain gage, which has over 50 years of data. Based on statistical analyses of the differences between Airport and City rain gage rainfall volumes, regressions between these differences and distance from the Airport gage to the City rain gage provided a correlation coefficient of 0.64. This linear regression was then extended to the each combined sewer overflow basin to developed basin-specific design storms that are scaled from the Airport rain data. The benefit of developing customized rainfall data for each CSO basin based on regressions between the local gages and the City gage, can be demonstrated by comparing to two single, system-wide design storm scenarios: 1) use of the largest regression multiplier times the SIA data to ensure compliance for the worst-case basin and 2) use of the airport data directly. Using the first case, CSO reduction facilities were over-designed by approximately 70%, corresponding to increased costs of approximately 40 to 60 million, whereas, in the second case, CSO facilities would be under-designed by approximately 35%, corresponding to potential regulatory violations over one third of the time.