Volume 28, Issue 19, 15 November 2008, Pages 2668–2677

Modeling flocculation processes of fine-grained particles using a size-resolved method: Comparison with published laboratory experiments

  • School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, NY, USA

Abstract

The transport of fine-grained particles in estuarine and coastal waters is influenced by flocculation processes (aggregation and floc breakup). As a consequence, the particle size varies with time in the water column, and can be orders of magnitude larger than those of primary particles. In this study the variations in floc size is simulated using a size-resolved method, which approximates the real size distribution of particles by a range of size bins and solves a mass balance equation for each bin. To predict the size distribution both aggregation and breakup processes are included. The conventional rectilinear aggregation kernel is used which considers both turbulent shear and differential settling. The breakup kernel accounts for the fractal dimension of the flocs. A flocculation simulation is compared to the settling column lab experiments of Winterwerp [1998. A simple model for turbulence induced flocculation of cohesive sediment, Journal of Hydraulic Research, 36, 309–326], and a one-dimensional sediment transport model is verified with the observed variations in floc size and concentration over tidal cycles in a laboratory flume experiment of Bale et al. [2002. Direct observation of the formation and break-up of aggregates in an annular flume using laser reflectance particle sizing. In: Winterwerp, J.C., Kranenburg, C. (Eds.), Fine Sediment Dynamics in the Marine Environment. Elsevier, pp. 189–201]. The numerical simulations compare qualitatively and quantitatively well with the laboratory measurements, and the analysis of the two simulation results indicates that the median floc size can be correlated to the sediment concentration and Kolmogorov microscale. Sensitivity studies are conducted to explore the role of settling velocity and erosion rate. The results are not sensitive towards the formulation of settling velocity, but the parameterization of erosion flux is important. The studies show that for predicting the sediment deposition flux it is crucial to include flocculation processes.

Keywords

  • Flocculation;
  • Fine-grained sediment;
  • Numerical simulation

Figures and tables from this article:

Full-size image (41 K)
Fig. 1. 

Variations of median floc size (D50) with time (upper panel), and floc size distributions at steady state (lower panel) for three test cases of flocculation simulation.

Full-size image (30 K)
Fig. 2. 

Temporal evolution of floc size distributions for test case T69.

Full-size image (18 K)
Fig. 3. 

Regression of median floc size (D50) with C0/G0.5; the line indicates a linear regression fit (γ2=0.99).

Full-size image (31 K)
Fig. 4. 

Variations of median floc size (μm) with depth and time (hour) from 1-D simulation.

Full-size image (41 K)
Fig. 5. 

Variations in floc size (solid line) over tidal cycles (upper panel) and variations in SSC (solid line) (lower panel) at 0.1 mab for 1-D simulation. The dash line is the frictional velocity, and the triangles are laboratory data adapted from Bale et al. (2002).

Full-size image (14 K)
Fig. 6. 

The temporal changes of floc mass density distribution at 0.1 mab by (a) breakup processes (solid line) and aggregation processes (dash line), (b) the sum of changes by aggregation and breakup processes, (c) changes by diffusion (solid line) and sedimentation (dash line), (d) the sum of changes by the above four processes, and (e) by erosion (solid line) and deposition (dash line) at hours 4.5 (left) and 5.5 (right).

Full-size image (34 K)
Fig. 7. 

Regression of median floc size (D50) with C0/G0.5; the line indicates a linear regression fit (γ2=0.52) (upper panel). Regression of median floc size (D50) with C0/G; the line indicates a linear regression fit (γ2=0.90) (lower panel).

Full-size image (26 K)
Fig. 8. 

Variations of SSC (solid line) and frictional velocity (dash line) of 1-D simulation without flocculation; SSC (dash-dotted line) with flocculation is included for comparison.

Full-size image (27 K)
Fig. 9. 

The settling velocity of Sternberg et al. (1999) (solid line) and Winterwerp (2002) (dash line).

Full-size image (25 K)
Fig. 10. 

Variations of medium floc size (solid line) and frictional velocity (dash line) over tidal cycles for 1-D simulation with the erosion flux of Warner et al. (2005); the median floc size of the base case (dash-dotted line) is included for comparison.

Table 1.

Parameters and results in the three test cases (adopted from Winterwerp, 1998)

C0 is the total mass concentration, G is the shear rate, and D50 is median particle size at equilibrium.

View Within Article
Table 2.

The median floc size (μm) under different fractal dimension (nf) simulations

View Within Article
Table 3.

The median floc size (μm) under different particle stickiness (α) simulations

View Within Article
Corresponding author contact information
Corresponding author. Tel.: +1 631 889 0232.
1

Now at the Department of Atmospheric Sciences, University of Illinois at Urbana-Champaign, Urbana, Illinois.