The settling dynamics of flocculating mud and sand mixtures: part 2—numerical modelling

Abstract

Estuarine and coastal sediment transport is characterised by the transport of both sand-sized particles (of diameter greater than 63 μm) and muddy fine-grained sediments (silt, diameter less than 63 μm; clay, diameter less than 2 μm). These fractions are traditionally considered as non-cohesive and cohesive, respectively, because of the negligible physico-chemical attraction that occurs between sand grains. However, the flocculation of sediment particles is not only caused by physico-chemical attraction. Cohesivity of sediment is also caused by biology, in particular the sticky extra-cellular polymeric substances secreted by diatoms, and the effect of biology in binding sediment particles can be much larger than that of physico-chemical attraction. As demonstrated by Manning (2008) and further expanded in part 1 of this paper (Manning et al., submitted), the greater binding effect of biology allows sand particles to flocculate with mud. In many estuaries, both the sand and fine sediment fractions are transported in significant quantities. Many of the more common sediment transport modelling suites now have the capability to combine mud and sand transport. However, in all of these modelling approaches, the modelling of mixed sediment transport has still essentially separated the modelling of sand and mud fractions assuming that these different fractions do not interact except at the bed. However, the use of in situ video techniques has greatly enhanced the accuracy and reliability of settling velocity measurements and has led to a re-appraisal of this widely held assumption. Measurements of settling velocity in mixed sands presented by Manning et al. (2009) have shown strong evidence for the flocculation of mixed sediments, whilst the greater understanding of the role of biology in flocculation has identified mechanisms by which this mud-sand flocculation can occur. In the first part of this paper (Manning et al., submitted), the development of an empirical flocculation model is described which represents the interaction between sand and mud particles in the flocculation process. Measurements of the settling velocity of varying mud-sand mixtures are described, and empirical algorithms governing the variation of settling velocity with turbulence, suspended sediment concentration and mud-sand content are derived. The second part of this paper continues the theme of examination of the effects of mud-sand interaction on flocculation. A 1DV mixed transport model is developed and used to reproduce the vertical transport of mixed sediment fractions. The 1DV model is used to reproduce the measured settling velocities in the laboratory experiments described in the part 1 paper and also to reproduce measurements of concentration of mixed sediments in the Outer Thames. In both modelling exercises, the model is run using the algorithms developed in part 1 and repeated using an assumption of no interaction between mud and sand in the flocculation process. The results of the modelling show a significant improvement in the ability of the 1DV to reproduce the observed sediment behaviour when the empirical equations are used. This represents further strong evidence of the interaction between sand and mud in the flocculation process.