Representation of Particulate Matter COD in Rainfall Runoff from Paved Urban Watersheds

For a half century, total suspended solids (TSS) has been the most commonly utilized particulate matter (PM) gravimetric index for wastewater. While TSS has been extended to urban runoff, runoff phenomena are unique. Runoff is unsteady and transports heterodisperse inorganic granulometry, giving rise to the PM index, suspended sediment concentration (SSC). With respect to PM-associated chemical oxygen demand (COD_p) in runoff, it is hypothesized that, while the TSS method can represent effluent COD_p, the SSC method is required to represent influent COD_p. COD_p and PM indices (TSS and SSC) for runoff events with mass balances and manual sampling are analyzed to investigate this hypothesis. This study examined a series of rainfall-runoff events captured from an instrumented fully paved urban catchment subject to traffic loadings in Baton Rouge, LA. Results indicate TSS generated substantial event-based mass balance errors for COD_p and Δm _p (mg/g) across a hydrodynamic separator (HS) as compared to SSC. TSS underestimates sediment-bound COD (>75 µm), a significant portion (maximum of 63% and median of 50%) of influent load. Negative bias by the TSS method for influent COD_p load increases as the heterodisperse particle size distribution becomes coarser. Above a PM of 250 mg/L, underestimation of COD_p by the TSS method is statistically significant. Utilizing the SSC method, COD_p reduction by a HS upstream of a batch clarifier (BC) indicates that a HS does not provide COD_p reduction, compared to a BC with 60 min of residence time. Representative PM and COD_P assessment suggests frequent BMP and drainage system maintenance to ensure proper operation and reduce pollutant elution.