Modelling of flocculation and transport of cohesive sediment from an on-stream stormwater detention pond
Introduction
Stormwater ponds are used extensively to enhance stormwater quality by removal of suspended solids and attached pollutants by settling. Thus, for effective design of these facilities, it is required to understand fine sediment transport processes in such ponds. Typical interactive transport processes comprise sediment transport by advection and turbulent diffusion, with concomitant particle aggregation (flocculation) and disaggregation (break-up), and the resulting sediment deposition or scouring [1]. Sediment flocculation strongly affects the density of the sediment aggregates (flocs) and their settling characteristics. The sediment reaching the bottom of the pond is subjected to elevated levels of bed shear stress that prevail at the sediment-water interface and may prevent sediment deposition. The bed shear stress could also erode previously deposited fine sediment and entrain more sediment into the water column. A quantitative understanding of all these processes is necessary for predicting the net deposition of sediment in the pond.
The methods currently used to evaluate suspended solids settling in stormwater ponds are not satisfactory, and are typically based on two approaches: (a) the ideal settling tank concept, and (b) computational fluid dynamics (CFD) models. The ideal settling tank concept was recommended for calculation of dynamic settling in ponds [2], but its underlying assumptions require further scrutiny. In this concept, the pond is approximated by a rectangular settling basin with a uniformly distributed, steady inflow along the upstream edge, uniform distribution of suspended particles in the lateral direction, and discrete particle settling without flocculation. Obviously, the applicability of these assumptions to actual ponds with irregular shapes, unsteady flow, non-uniform flow distribution in the pond, and the presence of flocculation is questionable [1]. CFD models provide a better representation of flow distribution in ponds [3], but their performance in simulation of low velocity fields has been questioned and their applicability to the settling/scouring of cohesive sediments has not been verified by field data.
To advance the understanding of stormwater sediment settling, Krishnappan and Marsalek [4] measured deposition characteristics of fine stormwater pond sediment in a rotating circular flume and concluded that the pond sediment underwent flocculation, while settling under different turbulent shear flows. Such findings emphasise the need for predictive methods that would consider the flocculation process explicitly in calculating the settling behaviour of fine sediment particles. As the first step toward developing such methods, the settling and flocculation model (S&FM) developed by Krishnappan [5] for still water was extended to the class of flows observed in the laboratory rotating flume, and then applied to the deposition experiments conducted with pond sediment [4]. The model features, calibration parameters and simulation results are presented in the following sections.
Section snippets
Salient features of the settling and flocculation model (S&FM)
In the S&FM [5], the motion of sediment particles is considered in two stages, namely, a settling stage and a flocculation stage. These stages were assumed to occur alternately during each time step of modelling. The settling stage was analysed using a one-dimensional unsteady advection–diffusion equationwhere Ck is the volumetric concentration of sediment of the kth size fraction and wk is the fall velocity of that fraction, D is the turbulent diffusion coefficient, t
Model application to rotating flume experiments
The deposition characteristics of fine sediments from an on-stream stormwater detention pond in Kingston, Ont., Canada were studied in a rotating circular flume at the National Water Research Institute at Burlington, Ont., Canada. The flume consists of a circular channel, which is 5.0 m in mean diameter, 0.30 m in width and 0.30 m in depth. The channel rests on a rotating platform, which is 7.0 m in diameter. A counter-rotating top cover that fits inside the flume with close tolerance (1.5 mm gap on
Size classes
The measured size distribution at the end of the high-speed operation was used to describe the size classes and the initial distribution of the particle sizes. As an example, the size classes and the volumetric concentration of particles in each of the size classes are given in Fig. 3 for the highest shear stress test (Test No. 1). The size classes shown in this figure were chosen according to Eq. (9).
Settling velocity, wk
The settling velocity of the sediment flocs is calculated using the Stokes’ equation and the
Coagulation factor, β
This parameter attains various values for specific types of sediment and is determined by calibration.
Collision frequency functions
The collision frequency function due to Brownian motion was considered to be unimportant compared to other mechanisms. The flow parameter that controls the frequency functions is the dissipation rate of kinetic energy of turbulence (ε) and it was evaluated from the results of the PHOENICSTM model.
Floc disruption due to turbulence
The growth-limiting effect of turbulence was modelled after Tambo and Watanabe [16]. According to
Results and discussion
Comparison of model predictions of concentration of the suspended sediment in the water column with those measured in the flume is shown in Fig. 4. The test with the highest shear stress, i.e. Test No. 1, was used as a calibration test and the calibration coefficients, a,b and β were determined by matching the predicted concentration vs. time curve and the size distribution profiles with the measured data. The numerical values of these coefficients were found to be a=0.02,b=1.45 and β=0.075.
Conclusions
A novel approach to modelling settling of fine sediment in stormwater ponds was proposed. It was based on Krishnappan's flocculated settling model for still water (1990), which was extended to dynamic conditions, and applied to flocculated settling of stormwater pond sediment in a laboratory rotating flume. Some model input parameters, such as floc properties, had to be obtained by calibration, and inputs describing flow field properties were obtained from a k−ε turbulence model. The model
Acknowledgements
The authors wish to acknowledge the contribution of Mr. Robert Stephens of the National Water Research Institute in carrying out the Laboratory experiments in the rotating circular flume for the Kingston Pond sediment. The authors also thank the anonymous reviewers for their constructive comments that helped to improve the quality of the manuscript.
References (18)
- et al.
Transport characteristics of fine sediments from an on-stream storm water detention pond
Urban Water
(2002) - et al.
Physical aspect of flocculation process—I. Fundamental treatise
Water Res
(1979) - et al.
Use of a field based water elutriation system for monitoring the in-situ particle size characteristics of fluvial suspended sediment
Water Res
(1993) - Lau YL, Krishnappan BG. Measurement of size distribution of settling flocs. NWRI Publication No. 97-223, National Water...
- Driscoll ED. Long term performance of water quality ponds. In: Roesner LA, Urbonas B, Editors. Proceedings of the...
FEM-modelling of open stormwater detention ponds
Nord Hydrol
(1997)Modelling of settling and flocculation of fine sediments in still water
Can J Civ Eng
(1990)- et al.
Numerical solution of convective transport problems
Am Inst Chem Eng J
(1963) - et al.
Numerical simulation of a sedimentation basin. 1. Model development
Environ Sci Technol
(1984)
Cited by (55)
Microplastics in urban catchments: Review of sources, pathways, and entry into stormwater
2023, Science of the Total EnvironmentCitation Excerpt :However, data from a settling basin with a design surface loading of 10 m/h showed that a finer fraction (<50 μm) was preferentially transported by runoff into the receiving waters (Klöckner et al., 2020). This apparent disconnect in particle size data may have two explanations: the fine material (<50 μm) may be deposited when inflow is low or zero, which is typical for interevent periods, or naturally induced flocculated settling may lead to the settling of small primary particles and consolidation of cohesive sediments in the basin, as described for stormwater settling in ponds by Krishnappan and Marsalek (2002). Such consolidation would also protect the bottom sediments against scouring caused by shear stresses based strictly on particle sizes.
Deposition and erosion behaviour of cohesive sediments in the upper River Taw observatory, southwest UK: Implications for management and modelling
2021, Journal of HydrologyCitation Excerpt :This approach can be used to determine the effects of flow field on erosion and deposition of cohesive sediments collected from different geographical locations and aquatic systems (Graham et al., 1992; Widdows et al., 1998; Pope et al., 2006), to improve knowledge of flocculation (Manning et al., 2007), to evaluate the effects of gravel on fine sediment dynamics (Glasbergen et al., 2014) and to study the roles of consolidation and bio-stabilization for sediment erodibility (Stone et al., 2011). Critically, data derived from flume experiments have enabled the development of models such as RIVFLOC (Krishnappan, 1991, 1997) that can be used to simulate fine sediment transport at multiple scales in storm water management ponds (Krishnappan and Marsalek, 2002), harbours (Skafel and Krishnappan, 1999) and large river systems in Canada (Milburn and Krishnappan, 2003; Droppo and Krishnappan, 2016). Even where in-channel sediment transport and deposition are represented explicitly within catchment modelling frameworks, the need for empirical calibration of relevant parameters remains important to ensure acceptable model performance on a catchment-specific basis.
Modeling behaviors of permeable non-spherical micro-plastic aggregates by aggregation/sedimentation in turbulent freshwater flow
2021, Journal of Hazardous MaterialsDeveloping a model to determine the settling velocity of ballasted flocs
2020, Journal of Environmental Chemical Engineering