Predicting the existence of stormwater first flush from the time of concentration

Simulations were performed using a one-dimensional kinematic wave model to examine the relationships between the time of concentration (T(c)) and the first flush (FF) for impervious landuses. The model used chemical oxygen demand parameters from a highway runoff site in west Los Angeles for eight storm events. FF indices such as the mass first flush ratio (MFF(n)) and the partial event mean concentration to event mean concentration ratio (PEMC(t)/EMC) were calculated from simulations using various rainfall intensities, rainfall durations, bed slopes, and watershed lengths (L), and matched measured values with an RMS error of 25% for six events. Resulting values of MFF(20) and PEMC(60)/EMC were correlated to T(c) calculated using ASCE's kinematic wave formula, and showed a non-monotone relationship, with very short or long T(c) having low MFF(20) with a maximum for mid-range T(c). Contours of the MFF(20) and PEMC(60)/EMC were created to predict and quantify FF from known T(c), L, and rainfall duration (T).