Local determination of constituent's decay coefficients for river water quality modelling considering the catchment scale

Modelling water quality of polluted heavy loaded water courses as those crossing highly urbanized catchment areas are a complex task that involves several processes based on coefficients like unit loads, decay rates and self-depuration effects among others. The fate of pollutants as organic matter and nutrients are usually done through transport equations added by sourcing and sinking terms that implies the use of decay factors. The so-called k’s coefficients are present in the literature and were derived from typical water samples that don’t represent the local conditions founded in the urban rivers, usually affected by the catchment scale. This article presents an approach method used for local determination of the k’s coefficient for Biological Oxygen Demand (BOD-k1), Atmospheric Reaeration (AR-k2) and Sediment Oxygen Demand (SOD-k4) and compares results to typical adopted values bye modelists. The approach is based on local waters and sediment laboratory tests adjusted to consider specific driving forces as water temperature, constituent concentration and flow turbulence. Considering the catchment area (286 km²) and the river reach (25 km long) 4 sampling stations were defined to collect depth integrated water samples and the bed material. The k1 coefficient is the most sensible one due to the influence of the biological components and the relation between labile and refractory fractions, that varies along the reach with the contribution from the sub catchments’ land use and sanitary infrastructure. Considering Fujimoto’s and Thomas’ equations, different values of local dependent k1 were found. For k2, the concept of river shear stress velocity was applied to correlate the oxygen mass transferred to the water in the JAR test run for the observed range of velocity gradients in the natural flow. Results lead to a more realistic air entrainment rate due to hydraulic, superficial tension and presence of oils and greases. The SOD-k4 were determined after bottom sediment samples collected in the same defined stations for 3 different concentrations in clean water. The continuous oxygen demand for each sample was taken hourly in the first day and daily in the next 5, and then converted to the Toro’s active sediment layer demand considering local porosity and specific weight. Results showed considerably lower than the typical values referenced in the literature for all k’s, denoting the influence of the above-mentioned characteristics and the level of uncertainties that could affect modelling results when non-local parameters are employed.