Experimental study of pollutant accumulation on an urban road surface J. Vaze * , Francis H.S. Chiew Department of Civil and Environmental Engineering, Cooperative Research Centre for Catchment Hydrology, University of Melbourne, Victoria 3010, Australia Received 29 September 2000; received in revised form 6 March 2002; accepted 30 April 2002 Abstract Anunderstandingofpollutantcharacteristicsonimpervioussurfacesisessentialtoestimatepollutantwashoffcharacteristicsand to design methods to minimise the impacts of pollutants on the environment. This paper presents data on surface pollutant characteristics on an urban road surface in Melbourne, Australia, from samples collected over a 36 day period. The data indicate thatbuildupoverthedrydaysoccursrelativelyquicklyafterarainevent,butslowsdownafterseveraldaysasredistributionoccurs. Thesurfacepollutantalsobecomesfineroverthedrydaysasitisdisintegrated.Thewashoffofsurfacepollutantisdependentonthe rainfallandrunoffcharacteristics,buttheresultshereshowthatcommonstormsonlyremoveasmallproportionofthetotalsurface pollutantload.Thedataalsoshowthatstreetsweepingmayhaveanadverseimpactonpollutantwashoffbecausethestreetsweeper releases the finer material but only removes some of them, making the fine sediment available for washoff by the next storm. The dataalsoshowthatmostofthenutrientsareattachedtothefinersediments,andtoeffectivelyreducenutrientloadsinparticulates, treatment facilities must be able to remove the finer particles (down to 50 lmforTPanddownto10 lmforTN),andnotjustthe total sediment or suspended solid load. Ó 2002 Elsevier Science Ltd. All rights reserved. Keywords: Pollutant; Road surface; Nutrient; Particle size; Street sweeping 1. Introduction Stormwater pollution is a major problem in urban areas. The loadings and concentrations of suspended solids, nutrients and other contaminants are much higher in urban stormwater runoff than in runoff from unimpaired and rural areas (Sartor & Boyd, 1972). Es- timates of urban stormwater pollutant loads are re- quired to assess the impact of stormwater pollution on drainage waterways and receiving waters and to design methods for minimising the impacts. An inaccurate representation of non-point source pollution can lead to the design of undersized and ineffective measures, or oversized measures with excessive capital costs and maintenancerequirements.Itisalsoimportanttohavea clear understanding of the amount of pollutants asso- ciated with different particle size ranges so that the treatment facilities can be effectively designed to target the most polluted sediment sizes. Stormwater quality modelling started in early 1970s and many models have been proposed. The models typically view stormwater pollution as a two-stage pro- cess,pollutantbuildupandpollutantwashoff.Buildupis theaccumulationofpollutantsonthecatchmentsurface during dry periods and washoff is the removal of the pollutants by rainfall and runoff. The pollutant avail- ability on the catchment surface (buildup) is typically estimated as a linear, exponential, power or Michaelis– Menton function of the number of dry days since the previous storm event. Pollutant washoff is com- monlymodelledasanexponentialdecayoftheavailable surface pollutant load, with rainfall intensity, rainfall volume, runoff rate or runoff volume used as the ex- planatory variable(s) to describe the decay. There have been numerous studies on pollutant washoff (Duncan, 1995a; Duncan, 1995b). Although the pollutant availability on the catchment surface is an important variable used in most existing washoff models, there have been few studies (Barrett, Malina, Charbeneau, & Ward, 1995; Deletic, Maksimovic, Loughreit, & Butler, 1998) on pollutant buildup since * Fax: +61-2-6299-2743. E-mail address: jvaze@dlwc.nsw.gov.au (J. Vaze). 1462-0758/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved. PII:S1462-0758(02)00027-4 Urban Water 4 (2002) 379–389 www.elsevier.com/locate/urbwat