No abstract

A wet detention pond in Norway has been monitored for 12 months. The pond receives runoff from a highway with a traffic load of 42,000 average daily traffic. Hydraulic conditions in terms of inflow, outflow, and pond water level were recorded every minute. Water quality was monitored by volume proportional inlet and outlet samples. During most of the year, excellent pollutant removal was achieved; however, during two snowmelt events the pollutant removal was poor or even negative. The two snowmelt events accounted for one third of the annual water load and for a substantial part of the annual pollutant discharge. The performance of the pond was analyzed using a dynamic model and pollutant removal was simulated by first-order kinetics. Good agreement between measurement and simulation could be achieved only when choosing different first-order rate constants for different parts of the year. However, no relation between the rate constants obtained and the time of year could be identified, and neither did the rate constants for different pollutants correlate. The study indicates that even detailed measurements of pollutant input and output allow only average performance to be simulated and are insufficient for simulating event-based variability in pond performance.