Abstract

In this study, a simple laboratory settling experiment is performed to examine the hindered settling process of high-concentration sediment suspension. In the suspension, the height of the clear water–turbid water interface undergoes an initial, rapidly decreasing phase followed by a slowly decreasing phase with increasing settling time. The influences of initial settling height, primary sediment concentration and the size distribution of the sediment sample on the hindered settling process of the suspension are investigated. A large initial settling height of the suspension leads to a slow settling velocity of the suspension during the rapidly decreasing phase. The larger the primary particle concentration of the suspension, the more gently the vertical position of the interface decreases during the rapidly decreasing phase. Increasing the primary particle concentration also causes the slowly decreasing phase to appear later. Finally, a fine-grained sediment suspension results in a gentle decrease in the vertical position of the clear water–turbid water interface and a small settling velocity of the interface during the rapidly decreasing phase.

INTRODUCTION

The settling of sediment particles is an important process that has been investigated extensively in many natural sources (such as rivers, reservoirs, lakes, estuarine and coastal waters) and some industrial systems (such as sedimentation tanks and wastewater treatment plants) (Cheng 2009; Vahedi & Gorczyca 2011; Maggi 2013). In estuarine and coastal waters, the settling of sediment particles plays an essential role in determining the vertical flux of sediment transport. Sediment transport in turn is of great importance for determining some geophysical processes, such as morpho-dynamic change and ecosystem function variation related to the water quality, and the maintenance and management of navigation channels, ports and harbours (Winterwerp 2002; Maggi 2013; Priya et al. 2015).

The factors influencing the settling velocity of sediment particles include sediment particle properties (such as size, shape and structure), sediment concentration, the surrounding fluid properties (such as salinity and viscosity) and the turbulence condition of the surrounding fluid (Cuthbertson et al. 2008). There have been many studies to investigate the impact of increasing suspension concentration on the settling of particles (Cheng 1997; Winterwerp 2002; Camenen & Bang 2011): for the case that the particle settles within a still and homogeneous fluid, the terminal settling velocity could be determined by application of Stokes' law and modified expressions; while with increasing suspension concentration, the particle no longer settles independently due to the interaction of other suspending particles, resulting in a settling velocity lower than that for individual, isolated particles, which has been termed ‘hindered settling’. The simple Richardson and Zaki equation has been adopted widely to estimate the hindered settling velocity, , of a sediment particle in the suspension (e.g. Dankers & Winterwerp 2007; Cuthbertson et al. 2008; Pal & Ghoshal 2013): , where is the settling velocity of a sediment particle in clear fluid, c is the volumetric concentration of the suspension, and n is the exponent of reduction of the settling velocity. The exponent n depends on the particle Reynolds number and some empirical expressions have been proposed to describe the relationship between them (Pal & Ghoshal 2013).

There have also been some studies investigating the influence of salinity on the water–sediment suspension characteristics. Migniot (1968) showed salinity had a clear effect on the settling velocity of cohesive material when salinity varies from 0 to 13‰. Van Leussen (1999) introduced the concept of ‘flocculation ability’ and identified that variations in flocculation ability are at the root of the large differences in settling properties of suspended sediment in estuaries. The experiment of Thill et al. (2001) showed that Rhone river particulate matter has a poor average reactivity regarding salt-induced flocculation. The experimental observations of Mikeš et al. (2004) indicated salinity induces flocculation throughout a threshold action. Furthermore, Portela et al. (2013) carried out a laboratory settling experiment to examine the effect of salinity on the settling velocity of fine sediment samples collected in a harbour basin in the Tagus estuary, and concluded that settling velocity increased by a factor of 6.5 between freshwater and marine conditions.

By contrast, for a settling experiment, the influences of initial settling height and the size distribution of sediment particles on the hindered settling characteristics of a high-concentration suspension are not yet completely understood. This study aims to investigate the possible impacts of initial settling height, primary sediment concentration and size distribution of sediment particles on the hindered settling of the suspension by carrying out a set of controlled laboratory settling experiments, and provide a reference for further understanding of the hindered settling characteristics of high-concentration sediment suspension. The structure of this work is simply arranged as follows. The second section briefly introduces the experimental materials and methods, and the experimental results and simple discussions are presented in the third section. Finally the fourth section contains some concluding remarks.

MATERIALS AND METHODS

Sediment material collected from Huayuankou station on the Yellow River of China was adopted in the settling experiments of the high-concentration sediment suspension (referred to as sediment sample #1 in this study). Its particle size distribution was determined using a laser particle size analyser (Horiba LA-920; produced by Horiba Corporation, Tokyo, Japan). The median size of the sediment sample was 22.75 μm, and Table 1 presents its size range and mineral composition measured by an X-ray diffractometer (D/Max-RC; produced by Shimadzu Corporation, Tokyo, Japan). Before the experiment, the sediment sample was dried at a temperature of 90–100 °C in an oven for almost 30 minutes and then was cooled to contains room temperature.

Table 1

Size distribution and mineral composition of the three sediment samples in the study

Sample number Size distribution
 
Mineral composition
 
Median size (micron) Size range (micron) Mixed-layer clay Montmorillonite Illite Kaolinite Quartz Calcite Feldspar 
Sediment sample #1 22.75 1–65 20% 30% 10% – 15% 5% 20% 
Sediment sample #2 13.98 1–63 30% 15% 20% – 10% 15% 10% 
Sediment sample #3 8.12 1–21 30% – 20% 15% 10% 15% 10% 
Sample number Size distribution
 
Mineral composition
 
Median size (micron) Size range (micron) Mixed-layer clay Montmorillonite Illite Kaolinite Quartz Calcite Feldspar 
Sediment sample #1 22.75 1–65 20% 30% 10% – 15% 5% 20% 
Sediment sample #2 13.98 1–63 30% 15% 20% – 10% 15% 10% 
Sediment sample #3 8.12 1–21 30% – 20% 15% 10% 15% 10% 

As many previous studies have shown, when the primary particle concentration exceeds a critical value, the water–sediment suspension forms a space-filling network structure (referred to as a gel in the work of Winterwerp (2002)) and experiences a measurable increase in strength (Winterwerp 2002). In this set of experiments, the settling experiments involving the high-concentration water–sediment suspensions were carried out in a standard 1,000 ml volumetric measuring cylinder. A graduated scale was arranged along the outside edge of the settling column (with an accuracy of 1 mm), where the zero point of the scale corresponded to the 1,000 ml scale line of the volumetric cylinder and increased as it ran down the length of the cylinder. The inner diameter of the measuring cylinder was 55.73 mm.

A pre-arranged high-concentration water–sediment suspension was injected into the volumetric cylinder, and after adequate stirring by using the agitator, the experiment was initiated. Due to the high primary sediment concentration, the water–sediment suspension in the settling column had an obvious clear water–turbid water interface due to the formation of a space-filling network structure. Above the interface, the suspension was clear, and it contained fewer sediment flocs. Below this interface, the water suspension was turbid, and it contained many suspended sediment flocs. During the experiment, the clear water–turbid water interface gradually decreased with time, leading to a gradually increasing volume of clear water suspension and a gradually decreasing volume of turbid suspension. Although some microscopic mechanisms regarding the formation of the space-filling network structure remain unclear (Winterwerp 2002; Zhu 2014), some studies have considered the appearance of a clear water–turbid water interface to be indicative of a space-filling network structure (Winterwerp 2002; Guo et al. 2015). In this study, the clear water–turbid water interface was simply considered to indicate the space-filling network structure in the high-concentration water–sediment suspension, and the impacts of the initial settling heights, different primary sediment concentrations and different size distributions of sediment samples on the flocculation and hindered settling characteristics of the network structure in the settling column were investigated. During the experiment, the temporal variation in the vertical position of the clear water–turbid water interface was recorded. The settling velocity of the interface could be readily calculated as the change in the vertical position of the interface divided by the recorded settling time.

To study the influence of initial settling height on the flocculation and sedimentation behaviours of the high-concentration sediment suspension, six settling heights, H, were chosen in this experiment: H = 15 cm, 20 cm, 25 cm, 30 cm, 35 cm and 41 cm, under three primary sediment concentrations of 200 kg/m3, 250 kg/m3 and 300 kg/m3 respectively. Simply fixing the initial setting height of the water–sediment suspension at 25 cm, this study adopted five different kinds of primary sediment concentrations, s, as follows: s = 200 kg/m3, 225 kg/m3, 250 kg/m3, 275 kg/m3 and 300 kg/m3 respectively, in order to investigate the impact of different primary sediment concentrations on the flocculation and settling characteristics of the water–sediment suspension.

Furthermore, another two sediment samples were also used in this settling experiment to study the influence of different size distributions of sediment sample on the sedimentation behaviours of the water–sediment suspension by simply fixing the initial settling height at 41 cm and the primary sediment concentrations at 200 kg/m3 and 250 kg/m3 respectively. The first sediment sample was also collected from Huayuankou station on the Yellow River of China (referred to as sediment sample #2 in this study), but had a small median size of 13.98 μm (its size distribution and mineral composition measured by an X-ray diffractometer are also shown in Table 1). The rest sediment sample was as adopted in the work of Liang (2004) (referred to as sediment sample #3 in this study): it was also collected from Huayuankou station on the Yellow River, but it had a much finer median size of 8.12 μm (its size distribution and mineral composition measured by an X-ray diffractometer are also shown in Table 1).

RESULTS AND SIMPLE DISCUSSION

Impact of initial settling height

Figure 1(a)–1(c) show the time evolution of the vertical position of the clear water–turbid water interface of the suspension for six initial settling heights, H = 15, 20, 25, 30, 35 and 41 cm at three primary sediment concentrations, s = 200 kg/m3, 250 kg/m3 and 300 kg/m3 respectively. Here, z is the vertical positon of the clear water–turbid water interface, and t is the settling time. For any given case, it is clear that the height of the clear water–turbid water interface in the suspension initially experienced a rapid decrease (rapidly decreasing phase) followed by a slow decrease (slowly decreasing phase) with increasing settling time.

Figure 1

Time evolution of vertical position of clear water–turbid water interface for six initial settling heights: 15, 20, 25, 30, 35 and 41 cm at three primary sediment concentrations: (a) 200 kg/m3, (b) 250 kg/m3 and (c) 300 kg/m3 respectively; (d) average settling velocity of the rapidly decreasing phase of the interface with respect to the six initial settling heights at three primary sediment concentrations.

Figure 1

Time evolution of vertical position of clear water–turbid water interface for six initial settling heights: 15, 20, 25, 30, 35 and 41 cm at three primary sediment concentrations: (a) 200 kg/m3, (b) 250 kg/m3 and (c) 300 kg/m3 respectively; (d) average settling velocity of the rapidly decreasing phase of the interface with respect to the six initial settling heights at three primary sediment concentrations.

A feasible interpretation of the appearance of both the rapidly decreasing and slowly decreasing phases is presented below.

When the settling experiment was initiated, the particle concentration was uniform across the total settling column, and some primary sediment particles began to collide and aggregate into flocs of varying sizes due to Brownian motion. It can be predicted that these flocs have larger sizes, and they have a settling velocity larger than the primary particles (Winterwerp 1998). During the settling process, flocs with a large settling velocity overtake smaller particles with a slower settling velocity, and as a result, collisions between these particles can lead to flocculation due to differential settling. In a high-concentration sediment suspension, more primary particles can aggregate into more and/or larger flocs due to Brownian motion and differential sedimentation. These flocs further cohere with each other and join into a space-filling network structure. In this network structure, there is weak strength between flocs. The network structure is porous, with free water filling in all of the pores between flocs in the structure. When the network structure settles, this free water is easy to remove from the structure. This leads to a rapid height reduction of the clear water–turbid water interface during the initial stage of the settling experiment. With a further increase in the settling time, most of the free water in the pores between flocs has been discharged from the network structure. At this point, the structure gets denser and starts to squeeze the water inside the floc. Because of this, the water now has a cohesive force with the solid core of the floc and therefore becomes more difficult to discharge. Therefore, the height of the clear water–turbid water interface begins to decrease more slowly. This slowly decreasing phase can last a long time; finally, the settling velocity approaches zero, and the sedimentation of the clear water–turbid water interface reaches a steady state.

The settling velocity, U, of the clear water–turbid water interface can be calculated by U = (z2z1)/(t2t1), where z1 and z2 are the vertical positions of the interface at time points t1 and t2, respectively. Figure 1(d) plots the estimated average settling velocity, u, of the rapidly decreasing phase of the interface with respect to the six initial settling heights of the suspension at the three primary sediment concentrations respectively. It is obvious that an increase in the initial settling height leads to a slower settling velocity. The reason may be simply because a large initial settling height implies a large friction effect on the sidewall of the volumetric cylinder during the settling of the suspension, leading to slow sedimentation behaviour of the water–sediment suspension. This result may imply that the effect of the sidewall might not be simply negligible in some settling column experiments.

Impact of different sediment concentrations

Fixing the initial settling height of the suspension at H = 25 cm, Figure 2(a) shows the time evolution of the vertical position of the clear water–turbid water interface at the five primary particle concentrations: 200 kg/m3, 225 kg/m3, 250 kg/m3, 275 kg/m3 and 300 kg/m3 respectively. It shows that the larger the primary particle concentration of the water–sediment suspension, the more gently the vertical position of the clear water–turbid water interface decreases during the rapidly decreasing phase. The critical time points of transition from the fast decreasing phase to the slowly decreasing phase are different depending on the different primary particle concentrations, as shown by Figure 2(b). Increasing the primary particle concentration causes the slowly decreasing phase to appear later. A simple explanation is presented as follows: when the primary particle concentration in the water–sediment suspension is high, more and more primary particles aggregate into flocs, which promotes the formation of the space-filling network structure in the suspension. As more flocs fill in the network structure, the network structure becomes denser. The free water inside the network structure will become more difficult to squeeze, and the complete discharge of all of the free water takes longer. This leads to a gentler decrease of the clear water–turbid water interface and a later appearance of the subsequent slowly decreasing phase. A deep mathematical model for describing the relationship between the critical time and sediment concentration seems to be worthy of further investigation in future study. Figure 2(c) shows the time evolution of the settling velocity of the clear water–turbid water interface at these five primary particle concentrations. It can be observed that a suspension with a low primary particle concentration generally experiences a larger settling velocity during the rapid settling phase of the clear water–turbid water interface, despite some scattering and fluctuation. This agrees with the qualitative understanding based on the developed equation for hindered settling velocity: , that an increase in particle concentration leads to a slow settling velocity of the suspension (e.g. Dankers & Winterwerp 2007; Cuthbertson et al. 2008; Pal & Ghoshal 2013). Some previous experiments have shown this behaviour (e.g. Cheng 1997; Winterwerp 2002), and a deep discussion regarding the hindered settling velocity can be found in the work of Pal & Ghoshal (2013).

Figure 2

(a) Time evolution of vertical position of clear water–turbid water interface at the five primary particle concentrations: 200 kg/m3, 225 kg/m3, 250 kg/m3, 275 kg/m3 and 300 kg/m3 respectively; (b) critical time points of transition from the fast decreasing phase to the slowly decreasing phase at these five primary particle concentrations; (c) time evolution of the settling velocity of the clear water–turbid water interface at these five primary particle concentrations respectively.

Figure 2

(a) Time evolution of vertical position of clear water–turbid water interface at the five primary particle concentrations: 200 kg/m3, 225 kg/m3, 250 kg/m3, 275 kg/m3 and 300 kg/m3 respectively; (b) critical time points of transition from the fast decreasing phase to the slowly decreasing phase at these five primary particle concentrations; (c) time evolution of the settling velocity of the clear water–turbid water interface at these five primary particle concentrations respectively.

Impact of different particle sizes

Simply fixing the initial settling height at H = 41 cm, Figure 3(a) and 3(c) show the time evolution of the vertical position of the clear water–turbid water interface for three kinds of sediment samples of different size distribution with the median size D50 = 22.75 μm, 13.98 μm and 8.12 μm, at two primary sediment concentrations, s = 200 and 250 kg/m3 respectively. Time evolution of the settling velocity of the interface is plotted in Figure 3(b) and 3(d) respectively. It is obvious that the larger the primary particle size of the water–sediment suspension, the more rapidly the vertical position of the clear water–turbid water interface decreases and the larger the settling velocity of the interface during the rapidly decreasing phase, even though there is a small difference between the case of D50 = 13.98 μm and that of D50 = 8.12 μm.

Figure 3

Time evolution of vertical position of clear water–turbid water interface for three kinds of sediment samples of different size distribution with the median size D50 = 22.75 μm, 13.98 μm and 8.12 μm, at two primary sediment concentrations, (a) s = 200 kg/m3 and (c) s = 250 kg/m3 respectively. Time evolution of the settling velocity of the interface at two primary sediment concentrations, (b) s = 200 kg/m3 and (d) s = 250 kg/m3 respectively.

Figure 3

Time evolution of vertical position of clear water–turbid water interface for three kinds of sediment samples of different size distribution with the median size D50 = 22.75 μm, 13.98 μm and 8.12 μm, at two primary sediment concentrations, (a) s = 200 kg/m3 and (c) s = 250 kg/m3 respectively. Time evolution of the settling velocity of the interface at two primary sediment concentrations, (b) s = 200 kg/m3 and (d) s = 250 kg/m3 respectively.

A simple explanation for this phenomenon is presented as follows. For a fine-grained sediment suspension, the specific surface area of the primary sediment particle (equal to the surface area divided by the volume, and equal to six divided by the volume for a spherical particle) is large, and as a result it is prone to flocculation, since the flocculation capacity of cohesive sediment particles is proportional to their specific surface area as revealed by several studies (Zhang et al. 1989; Kretzschmar et al. 1997; Sakhawoth et al. 2017). With an increasing rate of flocculation occurring between primary particles, more and more flocs are formed in the suspension and the network structure in the suspension becomes increasingly compact. As a result, the free water in the pores between flocs inside the network structure will become more difficult to squeeze, and the complete discharge of all of the free water takes longer.

CONCLUDING REMARKS

In this study, the laboratory settling experiment of a high concentration sediment suspension was carried out. The influences of the initial settling height, the primary sediment concentration and the size distribution of sediment samples on the flocculation and hindered settling characteristics of the sediment suspension were investigated. It was obvious that the height of the clear water–turbid water interface in the water–sediment suspension initially experienced a rapid decrease (rapidly decreasing phase) followed by a slow decrease (slowly decreasing phase) with increasing settling time. The former was attributed to the easy discharge of free water between adjacent flocs from the network structure, whereas the latter is because of the greater difficulty in releasing the remaining water inside the floc as discharge continues. Two qualitative conclusions could be obtained as follows:

  • (1)

    A large initial settling height of the suspension leads to a slow settling velocity of the water–sediment suspension during the rapidly decreasing phase due to a large friction effect on the sidewall of the volumetric cylinder during the settling process of the suspension. This result may imply that the effect of the sidewall might not be simply negligible in some settling column experiments.

  • (2)

    The larger the primary particle size of the water–sediment suspension, the more rapidly the vertical position of the clear water–turbid water interface decreases and the larger the settling velocity of the interface during the rapidly decreasing phase. The reason is that a fine-grained sediment suspension experiences a strong flocculation effect, more and more flocs are formed in the suspension and the network structure in the suspension becomes increasingly compact. As a result, the free water in the pores between flocs inside the network structure will become more difficult to squeeze, and the complete discharge of all of the free water takes longer.

ACKNOWLEDGEMENTS

This work is jointly supported by the National Natural Science Foundation of China (51509004).

REFERENCES

REFERENCES
Cheng
N. S.
1997
Effect of concentration on settling velocity of sediment particles
.
Journal of Hydraulic Engineering
123
,
728
731
.
Cuthbertson
A. J. S.
,
Dong
P.
,
King
S.
&
Davies
P. A.
2008
Hindered settling velocity of cohesive/non-cohesive sediment mixtures
.
Coastal Engineering
55
,
1197
1208
.
Dankers
P. J. T.
&
Winterwerp
J. C.
2007
Hindered settling of mud flocs: theory and validation
.
Continental Shelf Research
27
,
1893
1907
.
Guo
S. J.
,
Zhang
F. H.
,
Song
X. G.
&
Wang
B. T.
2015
Deposited sediment settlement and consolidation mechanisms
.
Water Science and Engineering
8
,
335
344
.
Kretzschmar
R.
,
Sticher
H.
&
Hesterberg
D.
1997
Effects of adsorbed humic acid on surface charge and flocculation of kaolinite
.
Soil Science Society of America Journal
61
,
101
108
.
Liang
C.
2004
The Study on the Gel-Like Network of Hyperconcentrated Cohesive Sediment Suspensions
.
Master's thesis
,
Tsinghua University
,
Beijing
,
China
.
Mikeš
D.
,
Verney
R.
,
Lafite
R.
&
Belorgey
M.
2004
Controlling factors in estuarine flocculation processes: experimental results with material from the Seine Estuary
,
Northwestern France. Journal of Coast Research
41
,
82
89
.
Pal
D.
&
Ghoshal
K.
2013
Hindered settling with an apparent particle diameter concept
.
Advances in Water Resources
60
,
178
187
.
Portela
L. I.
,
Ramos
S.
&
Teixeira
A. T.
2013
Effect of salinity on the settling velocity of fine sediments of a harbour basin
.
Journal of Coastal Research
(Special Issue)
65
(
2
),
1188
1193
.
Priya
K. L.
,
Jegathambal
P.
&
James
E. J.
2015
On the factors affecting the settling velocity of fine suspended sediments in a shallow estuary
.
Journal of Oceanography
71
,
163
175
.
Thill
A.
,
Moustier
S.
,
Garnier
J. M.
,
Estournel
C.
,
Naudin
J.-J.
&
Botero
J.-V.
2001
Evolution of particle size and concentration in the Rhône river mixing zone: influence of salt flocculation
.
Continental Shelf Research
21
,
2127
2140
.
Winterwerp
J. C.
1998
A simple model for turbulence induced flocculation of cohesive sediment
.
Journal of Hydraulic Research
36
,
309
326
.
Winterwerp
J. C.
2002
On the flocculation and settling velocity of estuarine mud
.
Continental Shelf Research
22
,
1339
1360
.
Zhang
R. J.
,
Xie
J. H.
,
Wang
M. F.
&
Huang
J. T.
1989
River Sediment Dynamics
.
China Water Power Press
,
Beijing
,
China
.