Manual Sediment Transport Measurements

MANUAL SEDIMENT TRANSPORT MEASUREMENTS
LIST OF CONTENTS
1. INTRODUCTION, PROBLEMS AND APPROACHES Page

1.1 Introduction 1.1
1.2 Sedimentation and erosion problems in rivers, estuaries and coastal seas 1.2
1.2.1 Introduction 1.2
1.2.2 Sedimentation and erosion problems 1.2
1.2.3 Approach of sedimentation problems 1.12
2. MORPHOLOGICAL DEFINITIONS AND PROCESSES Page

2.1 General 2.1
2.2 Definitions 2.2
2.3 Fluid flow and sediment properties 2.4
2.3.2 Sediment classification 2.4
2.3.3 Fluid and sediment properties 2.6
2.4 Sediment transport processes 2.11
2.4.2 Sand transport 2.12
2.4.2.1 Sand transport in steady river flow 2.14
2.4.2.2 Sand transport in non-steady (tidal) flow 2.19
2.4.2.3 Sand transport in combined non-steady (tidal) flow and oscillatory flow (waves) 2.20
2.4.3 Mud transport 2.23
2.4.3.1 General characteristics, definitions and modelling approaches 2.23
2.4.3.2 Cohesion 2.24
2.4.3.3 Flocculation 2.25
2.4.3.4 Settling 2.26
2.4.3.5 Deposition 2.28
2.4.3.6 Saturation 2.29
2.4.3.7 Consolidation 2.30
2.4.3.8 Erosion 2.31
2.4.3.9 Transport of mud 2.32
3. MEASURING PRINCIPLES, STATISTICS AND ERRORS Page

3.1 Measuring principles for suspended load transport 3.1
3.1.1 Direct method 3.1
3.1.2 Indirect method 3.1
3.2 Measuring principles for bed load transport 3.2
3.2.1 Direct method 3.2
3.2.2 Indirect method 3.2
3.3 Measuring statistics 3.3
3.3.1 General aspects 3.3
3.3.2 Sampling site 3.3
3.3.3 Number of measurements for suspended load transport 3.4
3.3.3.1 General aspects 3.4
3.3.3.2 Number of points in a vertical 3.4
3.3.3.3 Number of verticals over bed-form length 3.7
3.3.3.4 Number of verticals in cross-section 3.7
3.3.3.5 Number of verticals per tide 3.10
3.3.4 Number of measurements for bed-load transport 3.12
3.3.4.2 Number of samples at each location and number of locations along bed form 3.12
3.3.4.3 Number of sampling locations over width of cross-section 3.13
3.3.5 Sampling frequency 3.14
3.3.6 Sampling methods 3.14
3.3.7 Sample preservation and in-situ sampling 3.14
3.3.8 Sampling flexibility 3.15
3.4 Measuring errors and required accuracy 3.16
LIST OF CONTENTS February 2006

Manual Sediment Transport Measurements Page: 1
4. COMPUTATION OF SEDIMENT TRANSPORT AND PRESENTATION OF RESULTS Page
4.1 Rivers 4.1
4.1.1 Total load transport per unit width 4.1
4.1.2 Bed-load transport per unit width 4.1
4.1.3 Suspended load transport per unit width 4.1
4.1.3.1 Partial method 4.1
4.1.3.2 Integral method 4.1
4.1.4 Total load transport per cross-section 4.6
4.1.5 Presentation of results 4.7
4.2 Estuaries 4.9
4.2.1 Tide-integrated total load transport 4.9
4.2.2 Presentation of results 4.10
5. MEASURING INSTRUMENTS FOR SEDIMENT TRANSPORT Page

5.1 General aspects 5.1
5.2 Instrument characteristics 5.1
5.3 Selection of sediment transport samplers 5.4
5.3.1 Guidelines for selection of sediment transport samplers 5.4
5.3.2 Sediment transport measurements in rivers 5.5
5.3.3 Sediment transport measurements in estuaries 5.8
5.3.4 Sediment transport measurements in coastal seas 5.10
5.4 Comparison of suspended load samplers 5.16
5.4.1 Comparison of USP-61, Delft Bottle and Pump-Filter sampler 5.16
5.4.2 Comparison of Pump-filter sampler and ASTM 5.19
5.4.3 Comparison of Pump-Filter sampler and Pump-Bottle sampler 5.20
5.4.4 Comparison of Pump-Sedimentation sampler and Pump-Filter sampler 5.21
5.4.5 Comparison of Pump-Sedimentation sampler and Bottle sampler 5.22
5.4.6 Comparison of OBS and Pump sampler 5.24
5.4.7 Comparison of ASTM and Pump sampler 5.28
5.4.8 Comparison of ASTM, OBS and Pump sampler 5.32
5.4.9 Comparison of ABS and Pump sampler 5.34
5.4.10 Overall conclusions with respect to OBS, ASTM and ABS instruments 5.38
5.5 Descripton of bed load samplers 5.39
5.5.1 Trap sampling 5.39
5.5.1.2 Bed-load transportmeter Arnhem (BTMA) 5.41
5.5.1.3 Helley Smith (HS) 5.45
5.5.1.4 Delft Nile sampler (DNS) 5.50
5.5.2 Bed form tracking 5.55
5.6 Description of suspended load samplers 5.57
5.6.1 Classification of samplers 5.57
5.6.2 Bottle and Trap samplers 5.58
5.6.2.2 Bottle sampler 5.60
5.6.2.3 Trap sampler 5.65
5.6.2.4 USP-61 point-integrating sampler 5.71
5.6.2.5 Delft Bottle sampler 5.74
5.6.2.6 USD-49 depth-integrating sampler 5.80
5.6.2.7 Collapsible-Bag depth-integrating sampler 5.84
5.6.3 Pump sampler 5.87
5.6.3.1 General aspects for sampling in unidirectional flow 5.87
5.6.3.2 General aspects for sampling in oscillatory flow 5.93
5.6.3.3 Pump-Filter sampler 5.95
5.6.3.4 Pump-Sedimentation sampler 5.101
5.6.3.5 Pump-Bottle sampler 5.107
5.6.4 Optical and Acoustical sampling methods 5.111
5.6.4.1 General principles 5.111
5.6.4.2 Optical backscatter point sensor (OBS) 5.114
5.6.4.3 Optical Laser diffraction point sensors (LISST) 5.125
5.6.4.4 Various other Optical point sensors 5.127

5.6.4.5 Acoustic point sensors (ASTM, UHCM, ADV) 5.132
5.6.4.6 Acoustic backscatter profiling sensors (ABS and ADCP) 5.141
5.6.5 Impact sensor 5.154
5.6.5.2 IOS impact sensor 5.154
5.6.6 Nuclear sensor 5.156
5.6.7 Conductivity sensor 5.157
6. MEASURING INSTRUMENTS FOR PARTICLE SIZE AND FALL VELOCITY Page

6.1.1 In-situ sampling 6.2
6.1.2 Formulae particle fall velocity 6.2
6.1.3 Definitions of sediment sizes 6.3
6.2 Instrument characteristics 6.8
6.3 Selection of instruments 6.9
6.4 Comparison of instruments 6.10
6.4.1 BAT and VAT for sand particles 6.10
6.4.2 BAT, PWT,Wet-sieving and Coulter-Counter for fine particles 6.12
6.4.3 PWT, BWT and BAT for fine particles 6.14
6.5 Description of instruments 6.16
6.5.1 Photographic instrument 6.16
6.5.2 Sieving instruments 6.17
6.5.2.2 Dry sieving 6.17
6.5.2.3 Wet sieving 6.22
6.5.2.4 Air-jet sieving 6.22
6.5.3 Sedimentation instruments 6.23
6.5.3.2 Accumulation tube 6.25
6.5.3.3 Bottom Withdrawal Tube (BWT) 6.32
6.5.3.4 Pipet-Withdrawal Tube (PWT) 6.41
6.5.4 Coulter Counter 6.48
6.5.5 Particle size and concentration by Laser Diffraction (LISST, COULTER, PARTEC) 6.49
6.5.6 In-situ photo and video camera 6.64
6.5.7 Particle size and velocity by Phase Doppler Anemometry (PDA) 6.70
6.5.8 Particle size by Laser Reflectance (PARTEC Laser) 6.73
7. MEASURING INSTRUMENTS FOR BED MATERIAL SAMPLING Page

7.1 General Aspects 7.1
7.2 Bed material samplers: grab, dredge and scoop samplers 7.2
7.3 Bed material samplers: core samplers 7.5
7.4 Particle size of bed materials 7.10
7.4.1 Based on metallic trace elements (MEDUSA) 7.10
7.4.2 Based on acoustic reflection (ROXANN) 7.12
7.5 Movement of bed material particles 7.13
7.5.1 Critical bed-shear stress for initiation of motion 7.13
7.5.2 Tracer studies 7.15
8. LABORATORY AND IN-SITU ANALYSIS OF SAMPLES Page

8.1 Sediment concentration 8.1
8.1.1 Evaporation method 8.1
8.1.2 Filtration method 8.1
8.1.3 Units 8.3
8.2 Bed material composition
8.4
8.2.1 General aspects
8.2.2 Detailed method 8.4
8.2.3 Simple method 8.4
8.3 Suspended sediment composition 8.7
8.3.1 General aspects 8.9
8.3.2 Sandy environment 8.9

8.3.3 Silty environment 8.10
8.3.4 Sandy-silty environment 8.11
8.4 Sediment density 8.12
8.4.1 Detailed method 8.12
8.4.2 Simple method
8.13
8.5 Chemical analysis
8.6 Laboratory equipment 8.14
8.16
9. IN-SITU MEASUREMENT OF WET BULK DENSITY Page
9.2 Mechanical core sampler 9.1
9.3 Acoustic sensor 9.3
9.4 Nuclear radiation sensor 9.4
10. INSTRUMENTS FOR BED LEVEL DETECTION Page

10.1 Introduction 10.1
10.2 Mechanical bed level detection in combination with DGPS 10.1
10.3 Acoustic bed level detection (Echo-sounding instruments) 10.2
10.4 Optical bed level detection 10.5
10.5 Conclusions 10.5
ANNEX A: MEASURING INSTRUMENTS FOR FLUID VELOCITY, PRESSURE Page

AND WAVE HEIGHT
A1 Introduction 1
A2 Velocity sensors 3
A2.1 Velocities and bed-shear stresses, instrument characteristics and accuracies 3
A2.2 Electro-Magnetic Velocitymeter (EMV) 6
A2.3 Acoustic Doppler Velocitymeter (ADV) 8
A2.4 Acoustic Doppler Current Profiler (ADCP, UVP) 10
A2.5 Phased Array Doppler Sonar (PADS) 12
A2.6 Coherent Doppler Velocity Profiler (CDVP) and Cross-Correlation Velocity Profiler 15
(CCVP)
A3 Comparison of measured velocities 16

A3.1 Electro-Magnetic Velocitymeter (EMV) and Laser Doppler Velocitymeter (LDV) 16
A3.2 Acoustic Doppler Velocitymeter (ASTM) and Electro-Magnetic Velocitymeter (EMV) 18
A3.3 Acoustic Doppler Velocitymeters (ADV) 22
A3.4 Ultra-sonic Velocity Profiler (UPV) and Particle Image Velocitymeter (PIV) 27
A4 Fluid pressure and wave height instruments 28

A4.1 General instrument characteristics, accuracies and selection criteria 28
A5 Comparison of measured wave heights

33
A5.1 Pressure sensor and capacity wire
33
A5.2 Pressure sensor and surface following wave gauge
34
A5.3 Pressure sensors
36
A5.4 Velocity sensor, fluid pressure sensor and capacity wires
38
A5.5 Pressure sensor and resistance wave staff
39
A5.6 Accelerometer and DGPS on wave rider bouy 41

1. INTRODUCTION, PROBLEMS AND APPROACHES
1.1 Introduction
In general the natural bathymetry (bottom configuration) of a hydraulic system is under the influence of a
large number of factors varying from geological processes to the complex interaction of fluid and sediment
particles. Most hydraulic systems can be considered to be in a state of dynamic equilibrium between
deposition and erosion. The general characteristics only change very slowly with time. Human interference
with the governing phenomena in such a delicate equilibrium will have morphological consequences. To
predict these consequences for a specific project, it is of essential importance to have detailed knowledge of
the local morphological variables such as the bed material size, the settling velocities of the suspended solids
and the transport rates. To obtain this information, an extensive field survey should be carried out.
An important phase prior to the actual field survey is the selection of the most appropriate instruments, which
usually is a rather difficult problem because a wide range of instruments has been developed from simple
mechanical samplers to sophisticated optical and acoustical samplers. The selection of instruments is largely
dependent on the variables to be measured, the available facilities (boat, winch) and the required accuracy.
Especially, the required accuracy should be considered carefully. For example, a reconnaissance study
requires the use of much less sophisticated instruments than a basic research study.
This manual provides information of all relevant aspects related to sediment transport measurements such as:
measuring principles and statistics,
type and accuracy of the instruments,
selection of the instruments,
analysis of the samples,
elaboration and presentation of the measuring results.
Any field worker knows that there is considerable difficulty and expense in sediment transport measurements
inherent to the required time and labour in sampling of processes that usually vary greatly in space and time
(see Wren et al., 2000). Traditional forms of sediment transport measurements where samples are taken in
the field and transferred to the laboratory for analysis may lead to inaccurate results (particle size). The
importance of measuring particle size using sophisticaed in-situ electronic instruments avoiding sample
collection and handling which may change the particle size distribution (disturbing solids and/or aggregates),
has been stressed by many field workers.
In this manual the attention is focused on those instruments which have been proven to be reliable and
successful in field conditions. Instruments that are in a developing stage are not considered. Typical
laboratory instruments are not considered.
This manual is composed by L.C. van Rijn, senior research and project engineer of the Delft Hydraulics
Laboratory.
CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Manual Sediment Transport Measurements Page:1.1
1.2 Sedimentation and erosion problems in rivers, estuaries and coastal seas
1.2.1 Introduction
This Section addresses sedimentation and erosion engineering problems in rivers, estuaries and coastal seas
and practical solutions of these problems based on the results of field measurements, laboratory scale models
and numerical models.
Often, the sedimentation problem is a critical element in the economic feasibility of a project, particularly
when each year relatively large quantities of sediment material have to be dredged and disposed at far-field
locations.
Although engineering projects are aimed at solving problems, it has long been known that these projects can
also contribute to creating problems at other nearby locations (side effects). Erosion often occurs in places
where sediment cannot be supplied by nature in sufficient quantities because it is trapped in another part of
the system. The trapping can be due to natural causes or due to man-made changes in the system. Dredging
of ship channels, construction of jetties, groynes and seawalls always results in the redistribution of sand
within the local system. Engineering works should be designed in a way such that side effects (sand trapping,
sand starvation, downdrift erosion) are minimum. Dramatic examples of side effects are presented by
Douglas et al. (2003), who state that about 1 billion m3 (109 m3) of sand are removed from the beaches of
America by engineering works during the past century. Nourishment and bypassing of sand are often
required to mitigate the inavoidable side-effects of engineering works. It is important to emphasize that
engineering works should be designed and constructed or built in harmony rather than in conflict with nature.
This building with nature approach requires a profound understanding of the sediment transport processes
in morphological systems (Van Rijn, 2005).
1.2.2 Sedimentation and erosion problems
Human interference in hydraulic systems often is necessary to maintain and extend economic activities
related to ports and associated navigation channels. In many situations engineering structures are required to
stabilize the shoreline, shoals and inlets, to reduce sedimentation, to prevent or reduce erosion, or to increase
the channel depth to allow larger vessels entering the harbour basin. Coastal protection against floods and
navigability are the most basic problems in many estuaries in the world.
Examples of engineering works in estuaries and coastal systems are shown in Figure 1.
Waves
Jetty
Navigation channel
Harbour
River flow
Shoal
Tidal flow
Training wall
Reservoir
Groynes
Offshore
breakwater
Waves
Figure 1
Examples of man-made structures in river, estuary and coastal systems

Sedimentation problems generally occur at locations where the sediment transporting capacity of the
hydraulic system is reduced due to the decrease of the steady (currents) and oscillatory (waves) flow
velocities and related turbulent motions. Examples are: the expansion of the flow depth and width due to
natural variations or artificial measures (dredging), the presence of vortex or eddy zones, flow separation
zones, dead water zones and lee zones of structures. Expansions of the navigational depth will reduce
velocities inducing shoaling. Similarly, the expansion of the width of turning and mooring basins inside a
harbour area will reduce velocities stimulating shoaling conditions. Piers or piling structures create eddies
resulting in increased shoaling.
Sedimentation problems are most often associated with human interference in the physical system such as
the construction of artificial structures or the dredging of sediment from the bed to increase the flow depth or
width. However, sedimentation (as well as erosion) also is a basic phenomenon of nature dealing with loose
sediments within the transporting cycle from source to sink locations. Natural sedimentation areas are known
as shoals, flats, banks, sheets, bars, etc. Human interference in these natural sedimentation areas will always
lead to relatively large maintenance cost and should therefore be avoided as much as possible.
Sedimentation problems in rivers, estuaries and coastal seas are herein classified, as follows:
Channel sedimentation Basin sedimentation Shoreline sedimentation

1) Navigation channels; 1) Harbour and port basins, 1) Updrift area of groynes and
2) Inlet channels; Docks breakwaters normal to shore;
3) Entrance channels of 2) Open settling basins, turning 2) Lee-side area of offshore
harbours, docks and water and mooring basins, mining breakwaters parallel to shore.
intakes; pits
4) Trenches for tunnels, 3) Water intake basins;
pipelines and cables; 4) Flood plains and reservoirs.
Navigation channels
Ports are of vital importance for the economy of coastal countries. The increasing draft of vessels requires
the dredging of deep-draft channels connecting the port to deep water. Generally, these channels suffer from
sedimentation requiring maintenance dredging to ensure safe passage of the ships under most conditions. The
costs related to capital and maintenance dredging often are critical in the economic functioning of ports.
Therefore, the channel should be designed in such a way that sedimentation is minimum.
When the flow passes a channel, the velocities decrease due to the increase of the water depths in the channel
and hence the transport capacity of bed load and suspended load decreases. As a result the bed-load particles
and a certain amount of the suspended sediment particles will be deposited in the channel (see Figure 2).
When waves are present, this process is considerably enhanced due to the sediment stirring action of the
wave motions in the near-bed region resulting in larger sediment concentrations, which are subsequently
transported by the flow.
Factors enhancing sedimentation, are:
deep and wide channel;
orientation almost normal to the flow;
strong flows and large waves passing the channel;
fine sediment (fine sand and mud);
alignment through shoaling areas.
Natural navigation channels in estuaries often suffer from the generation of bars/shoals at the transition of
flood-domiated and ebb-dominated channels.

PLAN VIEW
OF
CHANNEL
FLOW
UNIDIRECTIONAL FLOW
Bed level at time T
TIDAL FLOW
Figure 2
Channel sedimentation (plan view and cross-sections)
Inlet channels
Natural tidal inlet channels generally suffer from heavy sedimentation due to wave-induced longshore input
of sediments thereby reducing the navigability of the channel. Jetties (long dam-like structures) are
commonly built to eliminate the input of sediment by the longshore current, creating an inlet channel (see
Figure 3). The jetties should be so long that the sedimentation at the entrance of the channel due to
longshore bypassing of sediments is minimum. Sediment accumulation will generally take place on the
updrift side of the jetties and erosion on the downdrift side. Mechanical bypassing of sediment may be
required to reduce downdrift erosion.
Jetty design requires the following considerations to reduce sedimentation:
the length of the jetties should extend beyond the littoral transport zone;
the jetty spacing should be narrow, but not leading to excessively large channel velocities undermining
the jetties (wide jetties may lead to shoaling and meandering of the channel);
the jetties should be impermeable to prevent lateral passage of water and littoral sediments;

the jetties should be parallel, rather than curved or tapered (narrow entrance, channel widening); parallel
jetties tend to confine the ebb flow, raising ebb velocities and thereby flushing sediment out of the
channel into the sea;
three jetties creating two parallel channels, may be built to separate the inlet channel from the river
outflow;
a settling basin/trap may be dredged at the entrance of the channel as a bufffer for sedimentation to
ensure navigability of the channel.
Tidal flow
Tidal flow
Jetties
HARBOUR ENTRANCE
TIDAL INLET CHANNEL
Flow
Tidal flow
WATER INTAKE DOCK ENTRANCE
Figure 3
Sedimentation of inlets, entrances and intakes (plan view)
Entrance channels of harbours, docks and water intakes

The entrance to a harbour basin, dock or water intake basin generally suffers from sedimentation due to the
reduction of flow velocities and wave activity (see Figure 3). Often, a circulation cell is generated in the
entrance of the basin due to the geometry involved, which attracts sediments by exchange processes with the
main flow system outside the entrance. The breakwaters should be streamlined or training walls may be built
to reduce eddies and dead-water areas. Harbour entrances should never be built on the inside of bends, where
natural shoaling processes (point bars) generally occur.

Entrance design requires the following considerations to reduce sedimentation:
the entrance should not be located in shallow depth on the inside of a bend or close to other natural
shoaling areas (see Figure 4);
the entrance should be narrow and streamlined to reduce the generation of eddies, and circulation zones;
the breakwaters should be impermeable to prevent lateral passage of water and sediments;
a settling basin/trap may be dredged at the entrance of the channel as a buffer for sedimentation.
Point bar Pool
Bar
Point bar
(shallow area) Thalweg
Pool
(deep area) A
Cross-section A
Propagation of tidal wave

Ebb
Flood
Ebb
Flood Shoal
inflo
Flat
inflo
River inflow
Figure 4
Deposition patterns
Top: River flow; deposition (point bars) near inner bend; erosion (pools) near outer bend
Bottom: Tidal flow; deposition (flats) near inner channel bends and between channels (shoal)
Trenches for tunnels, pipelines and cables

Small scale trenches for pipelines and cables (Figure 5) are characterized by relatively small dimensions;
width of about 2 to 10 m and depth of 1 to 2 m, while the side slopes may be rather steep (1 to 3), depending
on soil conditions. These trenches generally have a triangular shape due to the construction method
(plowing). The depth of such a trench is often only a fraction of the water depth resulting in flow separation
and the generation of vortex cells inside the major part of the trench area. In mobile bed conditions (surf

zone near beach) the trench should have a considerable overwidth to serve as a buffer for sediment trapped in
the period between trenching and pipe/cable laying.
Tunnel trenches have dimensions similar to those of navigation channels and show the same sedimentation
patterns.
Spoil
Pipe
Slopes of 1 to 3
Tunnel
Slopes of 1 to 7
Figure 5
Trenches for pipelines, cables and tunnels
Harbours and ports

Sedimentation in harbours and ports is a problem that exists as long as harbours and ports exist and is related
to their basic function, providing shelter by creating quiescent conditions. In general, a harbour is a place that
provides shelter and mooring for ships, whereas a port is a place where cargoes are loaded and unloaded
from ships. Both types of places require quiescent conditions protected from wave penetration and strong
currents.
The design and construction of harbour basins is one of the oldest branches of engineering. One of the oldest
known artificial harbours was PHAROS, located on the open coast of Egypt at about 2000 B.C. It had two
parallel breakwaters, each about 2.5 km long which consisted of rubble-mound structures of very large
blocks of rock and this harbour undoubtly suffered from sedimentation.
There are various types of natural harbours:
well protected bays,
lee areas of islands or headlands,
lee areas of reefs,
lee areas in river mouths.
Artificial harbours can, in principle, be made at any place along the water front, but should preferably be
located in areas where the sedimentation is observed or estimated to be minimum.
Artificial harbour basins can be classified into:
open basins; basins or berthing places which are open to flow and/ or waves; these basins can be situated
in nearshore coastal areas and along river banks, see Figure 6; the basin generally is substantially deeper
than the surrounding area so that an access channel is required for vessels to enter the basin; mooring is
only possible in conditions with weak cross-currents and mild wave motions;
sheltered basins (enclosed); these types of enclosed basins are typified by a single entrance to reduce the
flow and wave motion as much as possible; three subclasses can be identified:
- coastal basin with a controlled entrance, generally consisting of twin-armed breakwaters or jetties
(Figure 6); the breakwaters may block the longshore sediment transport requiring sediment bypassing

methods; eddies and turbulence may produced in the entrance; a semi-enclosed basin is obtained when
only one breakwater or jetty is present on the updrift side;
- coastal dredged lagoon with an uncontrolled entrance (Figure 6), which may migrate along the shore
or may suffer from severe sedimentation due to wave-induced processes or may even be closed by wave-
induced sedimentation during storms; the entrance may suffer from migrating channels and bars;
relatively strong currents may be present due to tidal filling of the basin; entrance may also be controlled
by jetties
- interior basin (docks) along esturial or river channels (Figure 6) and/or lakes; eddies are generally
present in the entrance area, where sedimentation of silty and muddy materials usually is maximum.
Access channel
A Main flow direction
Deeper basin
River bank
Basin
B Main flow
Waves
Trestle
pier
Coast
Coastal flow
C Waves
Breakwater
Coast
Coastal flow
Coast
Lagoon
D
Waves
River
E
Coastal flow
Figure 6
Types of harbour basins
Open basins: A= basin along river bank; B= basin along trestle pier perpendicular to coast
Sheltered basins: C= basin protected by breakwaters, D= coastal lagoon; E= interior basin along river
The most important and difficult part of a sheltered harbour basin is the entrance. Many harbour entrances
have been found to be difficult for ships to navigate owing to currents, waves and morphology
(sedimentation).
In previous centuries many harbours suffered from heavy sedimentation threatening the economic
functioning of the harbour. Some harbours deteriorated completely due to heavy sedimentation in the

entrance area of the basin. For example, the harbour of Amsterdam (The Netherlands) located in the
southwestern part of an inner sea (former South Sea) was one of the most powerful harbours in the world in
the 16th and 17th centuries, but lost its dominating position due to problematic sedimentation processes which
could not be solved at that time. Owing to better dredges, it is now possible to keep almost each harbour
entrance at the required navigation level, although this may be a financial burden on the economic
performance of a harbour suffering from heavy sedimentation.
From observed sedimentation rates it can be concluded that harbour sedimentation in fresh water conditions
is much less (factor 5) than in salt and brackish water conditions. The generation of stratified flow with a
clear salt water wedge is a well known phenomenon in the tidal zone of major rivers ( tidal volume about
equal to fresh water volume over tidal period). The maximum silt and mud concentrations are generally
found in the area where the edge of the salt water front is moving up and down the river channel. This zone
where soft fluid mud layers are formed due to deposition at slack tides (especially neap tides) is known as the
turbidity maximum. Harbour basins should preferably be situated outside this zone to avoid that the
deposited fluid mud layers penetrate into nearby harbour basins.
Turning, mooring and settling basins

Expansions of width and depth are often required outside and inside harbours or docks for navigational
reasons. The flow in these basins is reduced considerably in proportion to the expansion dimensions resulting
in rapid siltation of fine material. This principle can also be used for the design of a settling basin in areas of
rapid shoaling to create a buffer (overdepth) for sediment from which dredging activities can be performed to
remove the sediment from the system.
Flood plains and reservoirs

The sedimentation of flood plains, irrigation channels, navigation channels, floodways, reservoirs is a basic
problem in many river systems. Erosion problems can occur due to construction of roads and bridges when
protective vegetation is removed and steeply sloping cuts are made and left unprotected. Bridge pier scour
may lead to the failure of a bridge crossing the river channel in extreme flood events. Stream and river
control works may have an adverse effect on increasing erosion and sedimentation especially if runoff
conditions are changed. These types of problems create large public cost in maintenance operations.
Sedimentation in flood plains is a severe problem, because it reduces the flood carrying capacity of the river
system resulting in more frequent overflows. Often natural levees (ridges) of coarse materials are formed
adjacent to the main channels while the fine materials are transported and deposited in areas behind these
levees. The natural levees (ridges) obstruct the surface drainage of flood waters back to the main channels
resulting in the swamping of flood-plain lands.
Reservoir sedimentation is caused by the inflow of water and sediment into the reservoir. When the river
flow enters a reservoir, its velocity and hence transport capacity are reduced and the sediment load is
deposited in the reservoir (Figure 7). The amount of sediment deposited depends on the types of sediment in
the river system, the shape of the reservoir, the detention storage time and the operating procedures. Often,
more than 90% of the incoming sediment load is trapped. The loss of storage of the reservoir may cause
severe problems as the primary function of the reservoir is to store water for flood control, water supply,
irrigation and power. To deal with these problems, it is often tried to generate density or turbidity currents
that carry fine sediments to the deepest parts of the reservoir.

Inflow
Siltation Dam
Figure 7
Siltation in a reservoir
WAVES
OFFSHORE BREAKWATER
WAVES Flow
sedimentation
erosion
SEDIMENTATION AND EROSION NEAR INLET
WAVES
Flow
erosion
SEDIMENTATION AND EROSION NEAR GROYNES
Figure 8
Shoreline accretion and erosion near coastal structures (plan view)

Shorelines
Shorelines may suffer from large-scale and small-scale sedimentation and erosion processes.
Structures such as groynes or breakwaters perpendicular or oblique to the shoreline (Figure 8) will lead to
sedimentation on the updrift side of the structure due to the blocking of the sediment transport (current-induced
and wave-induced transport of sediment). Similarly, sedimentation will occur in the lee of a shoreparallel
breakwater (Figure 8). Generally, erosion will take place on the downdrift side, because the sediment transport
will gradually restore itself eroding sediment particles from the downdrift side. Mechanical bypassing of
sediment may be necessary to deal with the downdrift erosion near structures.
Large scale shoreline sedimentation and erosion generally are caused by:
episodic flood or storm-induced processes;
obstruction of the sediment transport processes due to the presence of natural barriers or artificial barriers;
fluctuations in sediment supply and transport;
onshore/offshore-directed sediment transport to or away from the shoreline;
mining/dredging and dumping of sediment.
The available options to deal with typical erosion problems, are: (1) to accept retreat in areas where the
shorelines/banks are wide and high; (2) to maintain the shoreline at a fixed position (by hard and/or soft
structures or by dredging activities) and (3) to bring the shoreline forward by reclaiming land.
Dredging and dumping of sediments

Large-scale dredging of sand from the seabed for nourishment of beaches, for land reclamations and for
other commercial purposes is known as sand mining and has become increasingly popular due to limited
availability of sand on land. Dredging of sediments deposited in harbour basins and approach channels is
known as maintenance dredging and is a basic element of the economic performance of many ports. Port
authorities should be closely examining their management of maintenance dredging with a view to cost-
effectiveness. The aspects to be considered in such an evaluation are:
navigational requirements; required channel depth and width in relation to weather conditions, soil
conditions (nautical depth concept may be introduced; navigable bed may be defined as that level at
which the density is about 1200 kg/m3);
suitable dredging methods (mechanical, hydraulic or agitation dredging) and monitoring programs to
evaluate the effect of dredging;
disposal grounds; determination of the nearest disposal sites and environmental aspects involved
(contamination of the deposited materials, available dumping locations for contaminated materials, type
of dredging equipment for contaminated materials.
The three essential elements of dredging are:

excavation,
transport,
disposal (plumes).
These three elements are linked together in the dredging cycle and any change in one may effect the other
two and have a major effect on the cost of the dredging operation. Often, the most critical element is the
dumping of sediments at the disposal site due to environmental problems. In many cases the dredged
material has to be taken out to the sea. Dumping in rivers and estuaries is not allowed if the dredged material
is polluted. Dumping at sea also has become a big environmental issue, because the polluted sediments may
damage the organisms living in the ocean.
Efficient management of dredging works requires:

detailed and regular monitoring of the area considered;
sufficient knowledge of the sediment transport processes in the area considered;
sufficient knowledge of dredging and disposal methods;
sufficient knowledge of cost and price factors of various dredging methods.

1.2.3 Approach of sedimentation problems
Approach
The general approach to solve sedimentation and erosion problems consists of the following major elements:
1. Specification of the problem and definition of wider context (socio-economic, legal, political,
environmental, administrative aspects, etc.).
2. Formulation of general objectives and desired state of knowledge,
- required level of accuracy,
- available time and budget.
3. Determination of problem dimensions and analysis of physical system (current state of knowledge),
- relevant user functions,
- physical parameters of interest,
- space and time scales involved,
- state of the system (indicators),
- existing knowledge (literature, charts, interviews).
4. Formulation of hypotheses related to problem.
5. Generation of alternative solutions and cost estimates,
- selection and application of tools (existing databases, measurements/monitoring, models),
- application of specialist knowledge.
6. Selection of optimum solution.
The three most basic rules are:

1. try to understand the physical system based on available field data; perform new field measurements
if the existing field data set is not sufficient (do not reduce on the budget for field measurements);
2. try to estimate the morphological effects of engineering works based on simple methods (rules of
thumb, simplified models, analogy models, i.e. comparison with similar cases elsewhere);
3. use detailed models for fine-tuning and determination of uncertainties (sensitivity study trying to
find the most influencial parameters).
Analysis of the physical system

One of the most important activities within an engineering approach is a sound analysis of the physical
system considered, involving:
Geometry and scales of the system,
- plan shape of coast,
- dimensions of shoals and bed forms,
- depths of natural channels,
- sediment composition and distribution,
- time scales of natural sedimentation and erosion patterns (migration rates),
- dredging volumes.
Tides and currents,
- vertical tidal ranges (micro<1 m, meso1 to 4 m, macro> 4 m),
- peak current velocities of flood and ebb phases (incl. velocity profiles),
- duration and asymmetry of flood and ebb phases,
- penetration length into estuary,
- residual (tide-averaged) flow velocities,
- three-dimensional flow patterns (flow in bends, stratified flow),
- wind-driven currents.
Sediment transport,
- bed forms (type and dimensions),
- mud, silt and sand concentration profiles,
- suspended size composition (sand), percentages of mud and organic material,
- in-situ settling velocities (for mud),
Wave climate,
- dominant wind and wave direction,

- frequency and intensity of storms,
- types of coastal exposure (open, sheltered or exposed),
- presence of breaker bars.
River discharge,
- frequency and intensity of flood waves,
- water levels and flow strenghts,
- presence of upstream control structures (weirs, barrages and reservoirs).
Estuary phenomena
- stratification effects (salt wedge) near outlet,
- turbidity maximum (null zone),
- tidal flats and shoals (migration rates),
- flood and ebb channel crossings,
- tidal ebb delta and migrating mouth bars.
Tools
The tools available for solving problems are:
existing databases,
measurements and monitoring (field studies),
numerical and or physical modelling.
Field studies comprise:

Hydrodynamic measurements,
- water level recordings,
- current velocity at fixed positions,
- current velocity profiling (ADCP method) from moving vessel,
- discharge measurements across main channels,
- float trackings of curved streamlines,
- wave field close to shore.
Sediment transport measurements,
- types and composition of sediment (mud, silt, sand, gravel, mixtures),
- critical bed-shear stresses for erosion and deposition (mud),
- settling velocity (flocculation of mud),
- bed load transport in lowest 0.1 m of water column,
- sediment concentrations at various levels above bed,
- bulk density of bed material (consolidation of mud).
Ecological risk assessment measurements for determination of contaminant sources
(www.ead.anl.gov),
- chemical analysis of water-sediment samples (X-ray Fluorescence Spectrometry for metals,
UV Fluorescence Spectoscopy for Polycyclic Aromatics and Hydrocarbons; Immuno-assay
for PCBs, Pesticides and PAHs; determination of Dissolved Oxygen and PH-values),
- physical analysis of water-sediment samples (particle sizes; moisture content),
- biological analysis of water-sediment samples (Bio-assay methods for organic materials).
Morphology,
- bathymetry as function of time,
- bed form trackings,
- sedimentation and erosion volumes from bathymetry data.
Laboratory scale models consist of:

Fixed bed engineering or design models,
- tide levels and flow patterns,
- nearshore wave conditions,
- configuration of structures,
- strength of structures (breakwaters).

Movable bed engineering models,
- valuable for small-scale 3D phenomena (local scour and deposition patterns),
- scale effects due to incorrect representation of sediment mobility and bed forms,
- laboratory effects due to space limitations and simplified boundary conditions.
Process models,
- data for understanding of physics involved,
- data for calibration and validation of numerical models,
- systematic variation of parameters,
- immediate and repeatable results of experiments.
Mathematical models consist of:

1 dimensional models (1D),
- suitable for rivers and network system of ebb and flood channels in estuaries (non stratified),
- suitable for longshore coastal flows,
- cross-section-integrated equations,
- sediment transport capacity formulae,
- advantages: easy to apply, good results for tide levels and discharges, long term morphology,
- disadvantages:poor results for local currents; no information of lateral morphology,
2 dimensional-vertical models (2DV),
- suitable for modelling of streamtubes (information of streamtubes from 2DH model),
- vertical structure of velocity and sediment concentration is included,
- space and time lag effects are included,
- advantages: easy to apply, operational on PC, long term estimates,
- disadvantages: schematization into streamtubes required, geometry of each tube must be known,
2 dimensional-horizontal models (2DH),
- suitable for coastal seas and estuaries,
- depth-averaged mass and momentum equations in two horizontal directions,
- depth-averaged suspended sediment equations (including lag effects),
- flooding and drying procedures,
- curvi-linear grid for efficient computations,
- nesting of models for detailed computations,
- advantages: standard tool, operational on advanced PC, easy to combine with wave models,
- disadvantages: not for very irregular geometry, not for stratified flow, not for secondary flow,
3 dimensional models (3D),
- mass and momentum in three coordinates (hydrostatic pressure is generally assumed),
- curvi-linear grid for efficient computations,
- advantages: all effects included (stratified and non-stratified flow, secondary flow), many details,
- disadvantages: not easy to apply (not much experience), only for local problems, short duration.
Sedimentation predictions can only be done accurately, if there is sufficient understanding of the physical
processes based on field measurements. These types of measurements require experienced personnel to
handle the sophisticated electronic instruments under extreme conditions and are often expensive to cover the
long term natural variations of the hydraulic system considered. Fixed bed laboratory scale models can be
operated to determine the local flow and wave fields; movable bed models may be applied to get information
of local, small-scale morphological developments near structures. The results of these laboratory models
suffer, however, from scale errors and interpretation errors related to the schematized boundary conditions.
Mathematical models do not suffer from scale effects, but the interpretation errors due to simplified model
formulations and boundary conditions also limit the use of model results. Mathematical models are
relatively easy to use but are not particularly cheap to operate, as many runs performed by experts are
required to get a good feeling of the most important parameters and uncertainties involved.

References
Douglas, S. et al. 2003. The amount of sand removed from Americas beaches by engineering works,
Coastal Sediments, Florida, USA
Van Rijn, L.C., 2005. Principles of Sedimentation and Erosion Engineering in Rivers, Estuaries and
Coastal Seas, 600 p., incl. toolkit on CD-ROM. Aqua Publications
(WWW.AQUAPUBLICATIONS.NL), The Netherlands
Wren, D.G., Barkdoll, B.D., Kuhnle, R.A. and Derrow, R.W., 2000. Field techniques for suspended
sediment transport measurement. Journal of Hydraulic Engineering. Vol. 126, No. 2, p. 97-104

2. MORPHOLOGICAL DEFINITIONS AND PROCESSES
2.1 General
Sediment is fragmental material, primarily formed by the physical and chemical desintegration of rocks from
the earth's crust. Such particles range in size from large boulders to colloidal size fragments and vary in
shape from rounded to angular. They also vary in specific gravity and mineral composition, the predominant
material being quartz. Once the sediment particles are detached, they may either be transported by gravity,
wind or/and water.
When the transporting agent is water, it is called fluvial or marine sediment transport. The process of moving
and removing from their original source or resting place is called erosion. In a channel the water flow erodes
the available material in the banks and/or the stream bed until the flow is "loaded" with as much sediment
particles as the energy of the stream will allow it to carry.
Usually, three modes of particle motion are distinguished:

rolling and/or sliding particle motion,
saltating or hopping particle motion,
suspended particle motion.
When the value of the bed-shear velocity just exceeds the critical value for initiation of motion, the bed
material particles will be rolling and/or sliding in continuous contact with the bed. For increasing values of
the bed-shear velocity the particles will be moving along the bed by more or less regular jumps, which are
called saltations.
When the value of the bed-shear velocity begins to exceed the fall velocity of the particles, the sediment
particles can be lifted to a level at which the upward turbulent forces will be of comparable or higher order
than the submerged weight of the particles and as a result the particles may go into suspension.
Usually, the transport of particles by rolling, sliding and saltating is called bed-load transport, while the
suspended particles are transported as suspended load transport. The suspended load may also include the
fine silt particles brought into suspension from the catchment area rather than from the streambed material
(bed material load) and is called the wash load. A grain size of 50 mm is frequently used to make the
separation between bed material load and wash load. Sometimes a value of 63 mm is used (USA). Another
method of discrimination is given by Vlugter (1941) and Bagnold (1962). Based on energy considerations
they state that all particles with a fall velocity smaller than 1.6 u i (u = depth-averaged velocity, i = water
surface gradient) can be transported in unlimited quantities, the latter being a typical feature of wash load
transport. An important characteristic of wash load is that its concentration is approximately uniform for all
points of the cross-section of a river. This implies that only a single point measurement is sufficient to
determine the cross-section integrated wash-load transport by multiplying with discharge.
Bed load and suspended load may occur simultaneously, but the transition zone between both modes of
transport is not well-defined. Figure 1 can be used to get a rough estimate of the ratio of the suspended load
and total load (no wash load) as a function of hydraulic conditions expressed by the ratio of the particle fall
velocity and the bed-shear velocity (Van Rijn, 1984).
CHAPTER 2: MORPHOLOGICAL DEFINITIONS AND PROCESSES February 2006

Manual Sediment Transport Measurements Page: 2.1
Figure 1
Ratio of suspended load transport and total load transport as function of bed-shear velocity and particle fall
velocity

2.2 Definitions
The following classification and definitions in accordance with the ISO-standards (ISO 4363) are given:
moving as bed load
Bed material load

moving as suspended load
Total load transport
wash load moving as suspended load
Bed material: The material, the particle sizes of which are found in appreciable quantities in that
part of the bed that is affected by transport.
Bed material load: The part of the total sediment transport which consists of the bed material and which
rate of movement is governed by the transport capacity of the channel.
Suspended load: That part of the total sediment transport which is maintained in suspension by
turbulence in the flowing water for considerable periods of time without contact with
the stream bed. It moves with practically the same velocity as that of the flowing
water.
Bed load: The sediment in almost continuous contact with the bed, carried forward by rolling,
sliding or hopping.
Wash load: That part of the suspended load which is composed of particle sizes smaller than
those found in appreciable quantities in the bed material. It is in near-permanent
suspension and, therefore, is transported through the stream without deposition. The
discharge of the wash load through a reach depends only on the rate with which these
particles become available in the catchment area and not on the transport capacity of
the flow.
References
Bagnold, R.A., 1962. Autosuspension of Transported Sediment; Turbidity Currents. Proc. Royal Soc,
A, no.1, London, England
ISO-COMMITTEE, 1977. Methods for Measurement of Suspended Sediment. Ref.No. ISO 4363-1977,
The Netherlands
Van Rijn, L.C., 1984. Sediment Transport, Part II: Suspended Load Transport. Journal of Hydraulic
Engineering, Vol 110, No. 11
Vlugter, H., 1941. The Transport of Sediment by Running Water (in Dutch). De Ingenieur van Nederlands
Indi, No. 3. Morphological Definitions

2.3. Fluid flow and sediment properties
2.3.1 Introduction
This Section presents an overview of :

Sediment classification,
Fluid and sediment properties,
- bed-shear stress,
- fluid density and viscosity,
- sediment density,
- sediment shape, size and fall velocity,
- critical bed-shear stress,
Sediment transport processes,
- sand transport,
- sand transport in steady river flow,
- sand transport in non-steady (tidal) flow,
- sand transport in combined non-steady (tidal) flow and oscillatory flow (waves),
- mud transport.
Detailed information is presented by Van Rijn (1990, 1993, 1998, 2005) and Soulsby (1997).
2.3.2 Sediment classification
Sediment is fragmental material, primarily formed by the physical and chemical desintegration of rocks from the
earth's crust. Such particles range in size from large boulders to colloidal size fragments and vary in shape from
rounded to angular. They also vary in specific gravity and mineral composition, the predominant materials being
quartz mineral and clay minerals (kaolinite, illite, montmorillonite and chlorite). The latter have a sheet-like
structure, which can easily change (flocculation) under the influence of electrostatic forces (cohesive forces) in a
saline environment. Consequently, there is a fundamental difference in sedimentary behaviour between sand and
clay materials.
Sediments can be classified according to their genetic origin:

Lithogeneous sediments, which are detrital products of disintegration of pre-existing rocks,
Biogeneous sediments, which are remains of organisms mainly carbonate, opal and calcium phosphate,
Hydrogeneous sediments, which are precipates from seawater or from interstitial water.
Descriptive sediment classifications can also be used and are related to characteristics like color, texture, grain
size, organic content, etc. For example, a mixture of sand and clay is classified as a sandy clay when the
percentage of sand is between 25% and 50%. Similarly, clayey sands, gravelly sands, sandy gravels, clayey
gravels, and gravelly clays are distinguished.
Sediment particles larger than 62 mm and smaller than 2000 mm are usually referred to as sand particles.
Based on mineral and chemical composition, three types of sands can be distinguished:
silicate sands,
carbonate sands,
gypsum sands.
Silicate sands mainly consist of quartz and feldspar minerals, which are extremely insoluble in water.
Carbonate sands consist of calcite and aragonite, which are two different crystalline forms of calcium-carbonite
(CaC03), originating from shell and coral fragments (coral sands). The percentage of carbonate in a sample
usually is larger than about 80%. Carbonate sands are much more soluble in fresh water than silicate sands. In
sea water (especially in the tropics) which is already supersaturated with carbonate it is hardly soluble. Carbonate
sands usually exhibit some degree of cementation: weakly-cemented or well-cemented, which means that the
fragments cannot be manually broken.

Gypsym sands consist of crystal forms of gypsum (CaS04.2H2O), which is a moderately soluble mineral that can
survive only in arid regions.
The grain size scale of the American Geophysical Union for sediments with particle sizes smaller than 2 mm
consists of about 13 subclasses ranging from very coarse sand to very fine clay. Herein, five somewhat
broader subclasses are distinguished:
coarse sand (non-cohesive) 0.5 to 2 mm (500 to 2000 mm),
fine sand (non-cohesive) 0.062 to 0.5 mm (62 to 500 mm),
coarse silt (sometimes cohesive) 0.032 to 0.062 mm (32 to 62 mm),
fine silt (weakly cohesive) 0.08 to 0.32 mm ( 8 to 32 mm),
clay+very fine silt (very cohesive) <0.08 mm (< 8 mm),
The following class separation diameters are herein used: dgravel=2000 mm, dsand=62 mm, dsilt=32 mm, dcfs= 8
mm. Basically, the pure clay fraction is the fraction with sediments smaller than 2 mm (lutum). For practical
reasons (laboratory determination of the percentage<2 mm is extremely difficult), the cohesive fraction with
clay and very fine silt is herein defined to consist of particles with diameters smaller than 8 mm (clay-silt
dominated fraction). Bed samples consisting of mixtures of clay, silt and sand are herein classified as: mud,
sandy mud, silty mud or clayey (fat) mud, depending on the percentages of sand, silt, clay and organic
material (Table 1).
Type of sediment Percentage Percentage Percentage Percentage

of organic of Clay+Fine of Silt of Sand
material Silt (< 8 mm) (8 to 62 mm) (> 62 mm)
Sand (non-cohesive) 0% 0% 0% 100%
Muddy Sand (weakly-cohesive) 0-10% 0-5% 20-40% 60-80%
Sandy Mud (cohesive) 0-10% 5-10% 30-60% 60-30%
Mud (cohesive) 0-20% 10-20% 50-70% 0-10%
Silty Mud (cohesive) 0-20% 10-40% 60-80% 0%
Clayey (fat) Mud (cohesive) 0-20% 40-60% 40-60% 0%
Table 1 Types of sand-mud mixtures
The transport of bed material particles may be in the form of either bed-load or bed-load plus suspended load,
depending on the size of the bed material particles and the flow conditions. The suspended load may also include
some wash load (usually, clay-silt dominated fraction smaller than 8 mm).
In major rivers the clay-silt dominated fraction with particle sizes smaller than about 8 mm can hardly be
observed in bed material samples indicating that there is not much exchange of this fraction with the bed.
Therefore the presence of this very fine sediment fraction in the water is herein defined as the wash load
determined by upstream supply conditions. The very fine fraction can be transported in almost unlimited
quantities (autosuspension) depending on the supply rate of very fines to the river by soil erosion and surface
runoff. In the absence of reservoirs the majority of the bed material load and the wash load will be
transported to the mouth of the river (estuary) where it will be deposited at the mouth bed. Generally, a
distinct downstream fining pattern from sand to clay can be observed along the bed of the estuaries. At the
most distal locations in the estuary where there is a transition from fresh to saline water (turbidity maximum)
and where the tide is dominant, the bed usually consists of very fine cohesive sediments (mud). Under the
dominant regime of tidal motion superimposed by surface waves during windy conditions a continous cycle
of deposition, consolidation, fluidization, erosion, flocculation and deposition and so on of fine sediments is
established in strong interaction with the prevailing mud bed.
In the lower river reaches and in most tidal basins the sediments are generally deposited in distinct patterns of
sand (channels), silt and clay (flats). The deposits (flats) of fines in shallow water generally consist of thin layers
of clay, silt and fine sand; the clay and silt fractions being the dominant fractions. Mixing may take place by
biological processes. The presence of these different types of sediment (clay, silt and sand) in the system will
result in selective transport processes (particle sorting). This latter process is related to the selective
movement of different types of sediment particles near incipient motion at low bed-shear stresses and during

generalized transport at higher shear stresses. Sorting effects can only be represented by taking into account
the full size composition of the bed material, which may vary horizontally and vertically.
The sand, silt and clay mixture (on the flats) generally behaves as a mixture with cohesive properties when the
clay-silt dominated fraction (<8 mm) is larger than a critical value (about 10%). The distinction between non-
cohesive mixtures and cohesive mixtures can be related to a critical value (pcfs,cr) of the clay-silt dominated
fraction (<8 mm) in the mixture. Cohesive properties become effective when the clay-silt dominated fraction is
larger than about 10%. The critical mud fraction (particles<62 mm) will be a factor 2 to 3 larger (20% to 30%).
2.3.3 Fluid and sediment properties
Since the sediment particles are carried by the fluid flow with or without surface waves superimposed, it is of
essential importance to have accurate information of the fluid system. Basic dimensionless parameters of
sediment transport processes are:
particle size D*=d50[(s-1)g/n2]1/3;
particle mobility parameter (Shields parameter) q=tb/((rs-rw)gd50);
suspension parameter Z=ws/(ku*).
with: tb=bed-shear stress, u*=(tb/rw)0.5=bed-shear velocity, d50=median sediment particle diameter,
ws=sediment fall velocity (settling velocity), rs= sediment density, rw=fluid density, n=kinematic viscosity
coefficient, k=coefficient of Von Karmann, g=acceleration of gravity.
Bed-shear stress
One of the most basis parameters of the fluid flow system is the bed-shear stress or bed-shear velocity.
In steady, uniform river flow with depth h and energy gradient S, the bed-shear stress is defined as:
tb=rw g h S
Using the Chzy equation (V=C(hS)0.5; v=depth-averaged velocity, C=Chzy coefficient), the bed-shear
stress can also be defined as:
tb=rw g V2/C2
This latter definition can also be used to compute the bed-shear stress in estuaries.
In coastal environments the bed-shear stress is a combination of the current-related bed-shear stress and the
wave-related bed-shear stress (Van Rijn, 1993; Soulsby, 1997).
The fluid velocity distribution over the water depth in the most common hydraulic rough flow regime is
expressed as a logarithmic function, as follows:
v=(u*/k)ln(z/zo)
with: v=fluid velocity at height z above the bed, zo=0.033ks=zero velocity level, ks=equivalent bed roughness
height.
Kim et al. (2000) discuss various methods to determine the bed-shear stress from measured velocity profiles
using an Acoustic Doppler Velocitymeter (ADV-sensor) and Electromagnetic velocitymeters (EMV). The
measurements should be done in the constant stress layer (lowerst 10% of the flow depth). The ADV has a
relatively small sampling volume with size of about 5 mm. Determination of the optimum sensor height may
be problematic. Too close to the bottom, the data suffer from drastically increased noise because of increased
shear within the sampling volume and increased acoustic signal scatter associated from suspended sediments.
Best results were found at a height of about 15 cm above the bottom.
Bed-shear stresses can be determined from the measured velocity data using the following methods (Kim et
al., 2000):
Logarithmic profile method (LP): the vertical distribution of the near-bed velocities is assumed to be
logarithmic with the bed-shear stress (tb) and zero-velocity level (zo) as basic parameters which can be

derived by regression analysis from the velocity data (at least 5 sensors in the lowest 1 m of the water
column);
Covariance method (COV): the bed-shear stress (tb=-r<U/W/>) is assumed to be equal to the time-
averaged value (<..>) of the horizontal (U/) and vertical velocity (W/) fluctuations (turbulent
fluctuations);
Turbulent Kinetic Energy method (TKE): the bed-shear stress is related to the turbulent energy (tb=c1E
with E=0.5(<U/U/>+<V/V/>+<W/W/>), with c1=0.21), (also used: tb=c2<W/W/> with c2=0.9);
Velocity Spectrum method (VS); the bed-shear stress is derived from the spectral characteristics of the
instantaneous velocity data.
Kim et al. (2000) have evaluated these four methods. Their conclusions are:
The optimum sensor height of the ADV is problematic. Too close to the bottom, the velocity data suffer
from drastically increased Doppler noise because of increased acoustic signal scatter associated with
near-bed suspended sediments. Reasonable results were obtained using a sensor height of about 0.15 m
above the bed.
The LP method is sensitive, because it normally requires measurements at several heights up to a 1 meter
above the bed. The LP method generally gives the largest estimates of the bed shear stress.
The COV method is considered to give unbiased estimates of bottom stress as long as the sensor is
within the constant stress layer near the bed but sufficiently far form the bed to avoid noise problems.
The TKE method is considered to be the most consistent and exhibits the least variability; c1 is found to
be c1=0.21 and c2=0.9.
The VS method requires the sensor location to be somewhat further away from the bed to have an
adequate separation of production and dissipation scales. The bed-shear stresses are systematically
underestimated (10%).
Fluid density and viscosity

All real fluids have certain measurable characteristics or properties such as density, viscosity, compressibility,
capillarity, surface tension etc. Some properties are combinations of other properties. For example, kinematic
viscosity involves dynamic viscosity and density. Herein, the two most basic properties being the density and the
viscosity, are given.
The density of fresh water varies with temperature between 1000 and 995 kg/m3 for temperatures beween 4o and
20o Celsisus. The density of sea water (= 1025 kg/m) can be determined from the following expression:
rw=1000+1.455CL-0.0065(Te-4+0.4CL)2
in which: CL=Chlorinity (in ), Te=temperature (in C).

The Chlorinity follows from:
Sa=0.03+1.805CL
with: Sa=Salinity = total quantity of dissolved salt in grammes per kilogramme of sea water (in , weight ratio).
The kinematic viscosity coefficient (n) is defined as:
n=h/r
in which: n=kinematic viscosity coefficient (m/s), h=dynamic viscosity coefficient (Ns/m),r=fluid density
(kg/m).
The kinematic viscosity coefficient can be approximated by:
n==1.14-0.031(Te-15)+0.00068(Te-15)2]10-6
The dynamic viscosity coefficient is influenced by the sediment particles. For dilute suspensions (c<0.1) Einstein
(1906) has found:

hm=h(1+2.5c)
in which: hm=dynamic viscosity coefficient of fluid-sediment mixture, h=dynamic viscosity coefficient of

clear water, c=volumetric sediment concentration
Sediment density
The density of quartz and clay minerals is approximately equal to rs= 2650 kg/m. The density of carbonate
material may be somewhat smaller (2500 to 2650 kg/m). The specific gravity is defined as the ratio of the
sediment density and the fluid density, s = rs/rw= 2.65.
The conversion from values in mass (kg) to bulk volume (m3) requires information of the dry bulk density of
the deposits.
The dry sediment density is the dry sediment weight per unit volume (= concentration) and is equal to:
rdry=(1-p) rs
with: rdry= dry sediment density (kg/m3); rw = water density (kg/m3), rs = sediment density (kg/m3) and p=
porosity factor.
The wet density or volume weight of deposited material (assuming total saturation) is the weight of water and
sediment per unit volume (also known as the wet bulk density) and is equal to:
rwet=p rw + (1-p) rs
This can also be expressed as:
rwet= rw +[(rs-rw)/rs)] rdry
or as
rwet= rw rs /[rw +(rs-rw)WC]
The water content WC is defined as:
WC=(rwet-rdry)/ rwet
The dry sediment weight of mixtures (consisting of sand, silt and mud) can be estimated from empirical
relationships, based on analysis of in-situ samples. Generally, the dry bulk density is represented as :
rdry= a + b Psand
with:
a = coefficient of about 500 kg/m3,
b = coefficient of about 1000 kg/m3,
psand = fraction (0 to 1) of sand particles (d 62 mm) in bed material.
The porosity of sediment material is often related to the deposition history of the sediment bed. Loose packing
occurs when sediments settle from suspension in still water. Basically, four packing arrangements are possible
for spherical particles. The most unstable arrangement is the cubic arrangement with the sphere centres forming
a cube yielding a porosity of 48%. The Rhombohedral arrangement with the spheres in the hollows of each other
yields the most stable packing and the smallest porosity of 26%. Random packing of spheres yields porosity
ranges from 36% to 40%. Natural sediments with particles of various sizes have relatively small porosity values
because the smaller particles can occupy the large void spaces. A pourly sorted (many sizes) coarse sand has a

porosity of about 40%. A well sorted (almost uniform) fine sand has a porosity of about 45%. The porosity of
coral sand (mixture of coral and shell fragments) has been found in the range from 0.5 to 0.65.
Deposits consisting of clay, silt, sand and organic material are called mud deposits and can have a large porosity
factor (upto 80%).
Sediment shape, size and fall velocity

Most of the sand particles on the face of the Earth are more or less rounded because their edges and corners are
smoothed by abrasion as running water or wind moves the sand particles from their origin (source) to their final
resting place. Roundness is a function of abrasion induced by transport and it increases slowly with distance.
Thousands of kilometers of transport in a river are required to achieve even moderate rounding. Beaches where
sand moves in and out with each wave are ideal places for rounding of sand particles if they stay there for any
length of time.
Usually, sediments are refered to as gravel, sand, silt or clay. These terms refer to the size of the sediment
particle. The grain size scale (used in sieving methods) is based on powers of 2 mm, which yields a linear
logarithmic scale via the phi-parameter defined as f = - 2log d (with d in mm).
Various methods are available to determine the particle size. Cobbles can be measured directly with a ruler.
Gravel, sand and silt are analyzed by wet or dry sieving methods yielding sieve diameters. Clay materials are
analyzed hydraulically by using settling methods yielding the particle fall velocity from which the standard fall
diameter is computed. Modern electronic instruments are the Laser Diffraction and Reflectance techniques.
Thus, the size of a sediment particle is closely related to the analysis method.
Typical "diameters" are:

sieve diameter which is the diameter of a sphere equal to the length of the side of a square sieve opening
through which the given particle will just pass,
nominal diameter which is the diameter of a sphere that has the same volume as the particle,
standard fall diameter which is the diameter of a sphere that has a specific gravity of 2.65 and has the
same fall velocity as the particle in still, distilled water of 24C.
A natural sample of sediment particles contains particles of a range of sizes. The size distribution of such a
sample is the distribution of sediment material by percentages of weight, usually presented as a cumulative
frequency distribution.
The frequency distribution is characterized by:

median particle size d50 which is the size at which 50% by weight is finer,
mean particle size dm = (pi di)/100 with pi = percentage by weight of each grain size fraction di,
standard deviation sd = pi(di-dm)/100 or sd = 0.5(d50/d16 + d84/d50).
Often the phi-scale is used for size distribution representation f = -log(d) where d is the particle diameter in
millimeters.
Basically, the fall velocity is a behavioural property. The terminal fall velocity (ws) of a sphere is the fall velocity
when the fluid drag force on the particle is in equilibrium with the gravity force, yielding:
ws=[4 (s-1) g d/(3 Cd)]0.5
in which: ws=terminal fall velocity of a sphere in a still fluid, d=sphere diameter, s=specific gravity (= 2.65)
CD=drag coefficient, g=acceleration of gravity.
The drag coefficient CD is a function of the Reynolds number Re = wsd/n and shape factor. In the Stokes region
(Re<1) the drag coefficient is given by: CD = 24/Re, yielding:

ws=(s-1) g d2/(18n)
Outside the Stokes region there is no simple expression for the drag coefficient. The CD-value decreases rapidly
outside the Stokes region (Re<1) and becomes nearly constant for 10 < Re < 105 , yielding ws proportional to
0.5
d .
The effect of temperature on the fall velocity is taken into account by the kinematic viscosity coefficient n. The
largest effect occurs for the smallest sphere diameters.
The expressions valid for a sphere cannot be applied for a natural sediment particle because of the differences in
shape. The shape effect is largest for relatively large particles (> 300 mm) which deviate more from a sphere
than a small particle. Experiments show differences in fall velocity of the order to 30% for shape factors in the
range from 0.5 to 1. Various formulae are available to determine the terminal fall velocity of non-spherical
sediment particles (Van Rijn, 1993).
The fall velocity of coral sand may be considerably smaller than that of quartz sand. The differences are
mainly caused by differences in shape. Coral sand particles are more angular and have, therefore, a smaller
fall velocity. The density of coral sand may also be somewhat smaller (= 2500 to 2650 kg/m).
The fall velocity of a single particle is modified by the presence of other particles. A small cloud of particles in a
clear fluid will have a fall velocity which is larger than that of a single particle. Experiments with uniform
suspensions of sediment and fluid have shown that the fall velocity is strongly reduced with respect to that of a
single particle, when the sediment concentration is large. This effect, known as hindered settling, is largely
caused by the fluid return flow induced by the settling velocities. A state of fluidization may occur when the
vertical upward fluid flow is so strong that the upward drag force on the particles becomes equal to the
downward gravity forces resulting in no net vertical movement of the particles.
Critical bed-shear stress

Particle movement will occur when the instantaneous fluid force on a particle is just larger than the
instantaneous resisting force related to the submerged particle weight and the friction coefficient. The degree
of exposure of a grain with respect to the surrounding grains (hiding of smaller particles resting or moving
between the larger particles) obviously is an important parameter determining the forces at initiation of
motion. Cohesive forces are important when the bed consists of appreciable amounts of clay and silt
particles.
The driving forces are strongly related to the local near-bed velocities. In turbulent flow conditions the velocities
are fluctuating in space and time. This makes together with the randomness of both particle size, shape and
position that initiation of motion is not merely a deterministic phenomenon but a stochastic process as well.
The fluid forces acting on a sediment particle resting on a horizontal bed consist of skin friction forces and
pressure forces. The skin friction force acts on the surface of the particles by viscous shear. The pressure force
consisting of a drag and a lift force is generated by pressure differences along the surface of the particle. These
forces per unit bed surface area can be reformulated in a time-averaged bed-shear stress.
Initiation of motion in steady flow is defined to occur when the dimensionless bed-shear stress (q) is larger than
a threshold value (qcr). Thus: q>qcr, with q=tcr,o/((rs-rw)gd50); tcr,o= critical bed-shear stress of cohesionless
particles on a horizontal bed, rs=sediment density, rw=fluid density, s=rs/rw = relative density, d50=median
sediment diameter.
The qcr-factor depends on the hydraulic conditions near the bed, the particle shape and the particle position
relative to the other particles. The hydraulic conditions near the bed can be expressed by the Reynolds
number Re* = u*d/n. Thus: qcr = F(Re*). The viscous effects (n=kinematic viscosity coefficient) can also be
represented by a dimension particle size D*=d50((s-1)g/n2)1/3, (Van Rijn, 1993).
Many experiments have been performed to detemine the qcr-values as a function of Re* or D*. The experimental
results of Shields (1936) related to a flat, horizontal bed surface are most widely used to represent the critical

conditions for initiation of motion. The Shields-curve represents a critical stage at which only a minor part (say
1% to 10%) of the bed surface is moving (sliding, rolling and colliding) along the bed.
The effect of bed slope on the critical bed-shear stress can be taken into account by the Schoklitsch-factor for
longitudinal slopes and by the Leitner-factor for lateral slopes (Van Rijn, 1993).
The Schoklitsch-factor reads, as:
tb,cr,slope/tb,cr,o=cosb(1-tanb/tanf)=sin(f-b)/sinf
The Leitner-factor reads, as:
tb,cr,slope/tb,cr,o=cosg [1-(tang/tanf)2]0.5
with: tb,cr= critical bed-shear stress, tb,cr,o=critical bed-shear stress along horizontal bed, b=longitudinal slope
angle (positive value of b refers to downsloping bed, negative value of b refers to upsloping bed), g= lateral
slope angle, f=angle of repose.
Both effects are taken into account by using:
tb,cr,slope/tb,cr,o= [sin(f-b)/sinf][cosg [1-(tang/tanf)2]0.5]
The angle of (natural) repose is a behavioural property of sand particles. Grains piled up on each other have an
equilibrium slope which is called the angle of natural repose (fn). This parameter appears to be a function of size,
shape and porosity. The angle increases with decreasing roundness. Values from the literature are in the range of
fn1 = 30 to 40 for sand sizes from 0.001 to 0.01 m. Observations in nature on the avalanche lee slope of
desert dunes and river bed dunes also show values in the range of 30 to 40.
The angle of repose (f) also referred to as the angle of internal friction is a characteristic angle related to the
particle stability on a horizontal or sloping bed. The angle of repose (f) may differ from the angle of natural
repose (fn). Usually, the angle of repose is determined from initiation of motion experiments for horizontal
and sloping beds.
2.4 Sediment transport processes
2.4.1 Introduction
Morphological problems are strongly related to gradients of sediment transport processes as caused by either
natural phenomena or by human interference. Often, the sudden changes in morphological patterns can be
traced back to the construction of engineering works such as the deepening of a tidal channel resulting in an
increase of tidal penetration and hence increased sediment import or the construction of a training wall
creating lee areas with substantial sedimentation.
The consequences of changes in environmental conditions can be both dramatic and long-lasting, affecting
the entire morphological system (estuary and adjacent coast).
As a prerequisite to the understanding of the effects of engineering works, the basic flow and sediment
transport processes in rivers, estuaries and coastal seas should be known in sufficient detail.
In rivers, the fluctuations in water level are mainly produced by rainfall and land runoff, while gravity is the
primary driving force for the unidirectional transportation of fluid and sediment to the sea (or lake).
In the sea, the water levels are governed by tide-induced forces in combination with wave-, wind- and
barometric pressure-induced forces. The daily tidal cycle consists of flood and ebb phases with opposite flow
directions and a continuously changing flow strength. Both the flow direction and flow strength are

influenced by the rotation of the earth (Coriolis force) and large-scale weather systems (geostrophical
currents).
In the estuaries between the rivers and the sea, the flow of water and sediment is affected by the mixing of
fresh river water and saline seawater (density-induced forces), depending on the geometrical shape (both in
plan and in bathymetry) of the estuary. Components of tide-generating forces may be selectively amplified
while reflection of tidal energy can lead to distortion of the tidal curve so that the duration of the flood and
ebb phases are different and maximum flow occurs after the time of high and low water (tidal asymmetry).
The shallowing of the water depth also increases the tidal range and hence the tidal currents, although the
effect is offset by increased bed friction. In systems with low tidal range the fresh river water generally flows
out to the sea as a distinct layer over the top of the seawater (salt wedge), which penetrates inland along the
deepest parts of the estuary. This generates a seaward flow of salt water in the upper layers of the salt zone
which, in turn, causes a compensating landward flow of seawater near the estuary bed (vertical gravitational
circulation in low-energy systems where separation of river water and seawater occurs throughout the tidal
cycle). In high-energy situations the flow velocities and turbulence field (eddies) are sufficiently strong to
continuously mix the fresh and saline water over the depth.
The type of sediment found in any morphological system depends on hydrodynamical, sedimentological and
geological factors. Generally, fine sand is the dominant type of sediment in the lower river bed, the main
channel beds of the estuary and the bed of the coastal sea. Muddy and silty materials generally dominate
along the banks of the channels and tidal flats/shoals within the estuary and further offshore in coastal seas.
Onshore-offshore movement as well as longshore movement of sand under the influence of oblique waves
are the dominant phenomena on sandy beaches. The net direction of movement will depend on the
predominant wave climate.
The movement of sediment in the estuarine environment closely follows the water movement. Sediment is
moved into an estuary on the flood tide from coastal sources and is mixed there with sediments eroded from the
bed and banks of the estuary, as well as that entering from the river. Flocculation of individual clay particles
and organic materials is a basic process taking place in saline conditions. In a high-energy system the sediments
and flocs are mixed throughout the water column and carried to all parts of the estuary by tidal currents. On the
ebb tide, salt water mixed with sediments and flocs flows out of the estuary. At locations with lower ebb
velocities, not all the material settled during the previous flood tide will be eroded again by the ebb flow,
especially in the lower reaches of the estuary or along the banks. Thus, sediment is trapped in the estuary by
gravitational circulation effects. Fine sediments consisting of clay, silt and fine sand staying undisturbed on the
bed for a few weeks may rapidly consolidate, making resuspension more and more difficult. The trapping effect
is maximum in the zone with maximum turbidity (front of salt wedge zone). This zone is not fixed but moves
landwards as tides increase from Neaps to Springs and seawards as they decrease from Springs to Neaps. Large
river flows will also move this zone seawards. In conditions with exceptionally large river flows the total load
of fine sediments is completely transported into the coastal zone, where it is moved along the coastline as large
scale mudbanks (Amazon River).
In this section, the basic sediment transport processes consisting of sand and mud transport are briefly
discussed. Detailed information is given by Van Rijn (1993, 2005).
2.4.2 Sand transport
General characteristics, definitions and modelling approaches

Sand can be transported by gravity-, wind-, wave-, tide- and density-driven currents (current-related transport),
by the oscillatory water motion itself (wave-related transport) as caused by the deformation of short waves under
the influence of decreasing water depth (wave asymmetry) or by a combination of currents and short waves.
In rivers the gravity-induced flow generally is steady or quasi-steady generating bed load and suspended load
transport of particles in conditions with an alluvial river bed. A typical feature of sediment transport along an
alluvial bed is the generation of bed forms from small-scale ripples (order 0.1 m) up to large-scale dunes (order
100 m). The adjustment of large-scale bed forms such as dunes and sand waves may lead to non-steady effects

(hysteresis effects) as it takes time for these large-scale features to adjust to changed flow conditions (flood
waves).
In the lower reaches of the river (estuary or tidal river) the influence of the tidal motion may be become
noticeable introducing non-steady effects with varying current velocities and water levels on a diurnal or semi-
diurnal time scale. Furthermore, density-induced flow may be generated due to the interaction of fresh river
water and saline sea water (salt wedge).
In coastal waters the sediment transport processes are strongly affected by the high-frequency waves introducing
oscillatory motions acting on the particles. The high-frequency (short) waves generally act as sediment stirring
agents; the sediments are then transported by the mean current. Low-frequency waves (bound-long waves)
interacting with short waves may also contribute to the sediment transport process. Low-frequency waves such
as bound-long waves and surf beat may have a specific role in the shoreface and surf zones. Although the
associated net transport directions are partly unknown, these types of waves may be of importance for the
position of the inner breaker bars. The net direction of this type of transport depends largely on phase differences
between near-bed current velocities and sand concentrations.
Field experience over a long period of time in the coastal zone has led to the notion that storm waves cause
sediments to move offshore while fair-weather waves and swell return the sediments shorewards. During
conditions with low non-breaking waves, onshore-directed transport processes related to wave-asymmetry and
wave-induced streaming are dominant, usually resulting in accretion processes in the beach zone. During high-
energy conditions with breaking waves (storm cycles), the beach and dune zone of the coast are attacked
severely by the incoming waves, usually resulting in erosion processes.
Sand transport is herein defined as the transport of particles with sizes in the range of 0.05 to 2 mm as found
in the bed of rivers, estuaries and coastal waters. The two main modes of sand transport are bed-load
transport and suspended load transport. The bed-load transport is defined to consist of gliding, rolling and
saltating particles in close contact with the bed and is dominated by flow-induced drag forces and by gravity
forces acting on the particles. The suspended load transport is the irregular motion of the particles through
the water column induced by turbulence-induced drag forces on the particles. Detailed information is
presented by Van Rijn (1993).
The definition of bed-load transport is not universally agreed upon. Sheet flow transport at high bed-shear
stresses may be considered as a type of bed-load transport, but it may also be seen as suspended load
transport. Some regard bed-load transport as occurring in the region where concentrations are so high that
grain-grain interactions are important, and grains are not supported purely by fluid forces.
The suspended load transport can be determined by depth-integration of the product of sand concentration
and fluid velocity from the top of the bed-load layer to the water surface.
Herein, the net (averaged over the wave period) total sediment transport in coastal waters is defined as the
vectorial sum of net the bed load (qb) and net suspended load (qs) transport rates: qtot = qb + qs.
For practical reasons the suspended transport in coastal waters will be subdivided into current-related and
wave-related transport components. This division is necessary to study the effect of phase differences
between velocity and suspended sediment.
Thus, the suspended sand transport is represented as the vectorial sum of the current-related (qs,c in current
direction) and the wave-related (qs,w in wave direction) transport components, as follows:
qs = qs,c + qs,w = vc dz + <(V-v)(C-c)> dz
in which: qs,c = time-averaged current-related suspended sediment transport rate and qs,w = time-averaged
wave-related suspended sediment transport rate (oscillating component), v= time-averaged velocity, V=
instantaneous velocity, C= instantaneous concentration and c= time-averaged concentration, <> represents
averaging over time, and .. represents the integral from the top of the bed-load layer to the water surface.
The precise definition of the lower limit of integration is of essential importance for accurate determination
of the suspended transport rates. Furthermore, the velocity and concentration profiles must be known. The
velocity profile has a three-dimensional structure in the coastal zone, particularly in conditions with breaking
waves resulting in an undertow (see Figure 1).
The current-related suspended transport component (qs,c) is defined as the advective transport of sediment
particles by the time-averaged (mean) current velocities (longshore currents, rip currents, undertow currents);

this component therefore represents the transport of sediment carried by the steady flow. In the case of waves
superimposed on the current both the current velocities and the sediment concentrations will be affected by
the wave motion. It is known that the wave motion reduces the current velocities near the bed while, in
contrast, the near-bed concentrations are strongly increased due to the stirring action of the waves. These
effects are included in the current-related transport. The wave-related suspended sediment transport (qs,w) is
defined as the transport of sediment particles by the high-frequency and low-frequency oscillating fluid
components (cross-shore orbital motion). Low-frequency transport contributions are herein neglected, unless
otherwise specified. The wave-related transport components are commonly supposed to occur in the plane of
orbital motion (generally, a small sector around the main wave propagation direction). Often the current-
related and the wave-related transport components are studied separately to evaluate the relative magnitude
of both components, which is of significant importance for modelling purposes.
Finally, the suspended transport vector can be combined with the bed load transport vector to obtain the total
transport vector.
Figure 1
Instantaneous 3-dimensional velocity vector in coastal zone with breaking waves
Various modelling approaches are available to determine the sediment transport rate in practical situations:
using rules of thumb, often based on large-scale deposition volumes;
using black-box transport capacity formulae;
using process-based practical engineering models (e.g., models that represent the vertical structure of the
flow and sediment concentrations, either intra-wave or time-averaged, as a function of the basic wave,
current and sediment properties);
using process-based research models (e.g., models that represent the intra-wave variation of fluid
velocities and sediment concentrations in the boundary layer and over the water column above it).
2.4.2.1 Sand transport in steady river flow
Basic characteristics
The transport of bed material particles may be in the form of either bed-load or bed-load plus suspended load,
depending on the size of the bed material particles and the flow conditions. The suspended load may also contain
some wash load (usually, clay and silt particles smaller than 0.05 mm), which is generally defined as that portion
of the suspended load which is governed by the upstream supply rate and not by the composition and properties
of the bed material. The wash load is mainly determined by land surface erosion (rainfall, no vegetation) and not
by channel bed erosion. Although in natural conditions there is no sharp division between the bed-load transport
and the suspended load transport, it is necessary to define a layer with bed-load transport for mathematical
representation.

When the value of the bed-shear velocity just exceeds the critical value for initiation of motion, the particles will
be rolling and sliding or both, in continuous contact with the bed. For increasing values of the bed-shear velocity,
the particles will be moving along the bed by more or less regular jumps, which are called saltations. When the
value of the bed-shear velocity exceeds the fall velocity of the particles, the sediment particles can be lifted to a
level at which the upward turbulent forces will be comparable with or of higher order than the submerged weight
of the particles and as result the particles may go in suspension.
The sediment transport in a steady uniform current over an alluvial bed is assumed to be equal to the transport
capacity defined as the quantity of sediment that can be carried by the flow without net erosion or deposition,
given sufficient availability of bed material (no armour layer).
In general, a river flood wave is a relatively slow process with a time scale of a few days. Consequently, the
sediment transport process in river flow can be represented as a quasi-steady process. Therefore, the available
bed-load transport formulae and suspended load transport formulae can be applied for transport rate predictions.
An exception to this may be the sediment transport process in a flash flood wave (banjir) and in the tide-
influenced lower reach of the river.
Flume and field data show that the sand transport rate is most strongly related to the depth-averaged velocity.
The power of velocity is approximately 3 to 4. Analysis of field and laboratory data sets shows that a certain
value of the overall bed-shear stress can occur at two values of the mean velocity depending on the presence of
bed forms or a flat bed. This means that the sand transport rate is not a unique function of the bed-shear stress.
The bed material in natural conditions consists of non-uniform sediment particles. The effect of the non-
uniformity of the sediments will result in selective transport processes (grain sorting). Grain sorting is related
to the selective movement of sediment particles in a mixture near incipient motion at low bed-shear stresses
and during generalized transport at higher shear stresses. Sorting effects can only be represented by taking
the full size composition of the bed material, which may vary horizontally and vertically, into account.
Bed forms are relief features initiated by the fluid oscillations generated downstream of small local obstacles at
the bottom consisting of movable (alluvial) sediment materials. Many types of bed forms can be observed in
nature. The bed form regimes for steady flow over a sand bed can be classified into (see Figure 2):
lower transport regime with flat bed, ribbons and ridges, ripples, dunes and bars,
transitional regime with washed-out dunes and sand waves,
upper transport regime with flat mobile bed and sand waves (anti-dunes).
When the bed form crest is perpendicular (transverse) to the main flow direction, the bed forms are called
transverse bed forms, such as ripples, dunes and anti-dunes.
Ripples have a length scale much smaller than the water depth, whereas dunes have a length scale much larger
than the water depth. The crest lines of the bed forms may be straight, sinuous, catenary, linguoid or lunate.
Ripples and dunes travel downstream by erosion at the upstream face (stoss-side) and deposition at the
downstream face (lee-side). Antidunes travel upstream by lee-side erosion and stoss-side deposition. Bed forms
with their crest parallel to the flow are called longitudinal bed forms such as ribbons and ridges.
In the literature, various bed-form classification methods for sand beds are presented. The types of bed forms are
described in terms of basic parameters (Froude number, suspension parameter, particle mobility parameter;
dimensionless particle diameter).
A flat immobile bed may be observed just before the onset of particle motion, while a flat mobile bed will be
present just beyond the onset of motion. The bed surface before the onset of motion may also be covered
with relict bed forms generated during stages with larger velocities.
Small-scale ribbon and ridge type bed forms parallel to the main flow direction have been observed in
laboratory flumes and small natural channels, especially in case of fine sediments (d50 < 0.1 mm) and are
probably generated by secondary flow phenomena and near-bed turbulence effects (burst-sweep cycle) in the
lower and transitional flow regime. These bed forms are also known as parting lineations because of the

streamwise ridges and hollows with a vertical scale equal to about 10 grain diameters and these bed forms
are mostly found in fine sediments (say 0.05 to 0.25 mm).
Figure 2
Bed forms in steady flows (rivers)
When the velocities are somehwat larger (10%-20%) than the critical velocity for initiation of motion and the
median particle size is smaller than about 0.5 mm, small (mini) ripples are generated at the bed surface.
Ripples that are developed during this stage remain small with a ripple length much smaller than the water
depth.
The characteristics of mini ripples are commonly assumed to be related to the turbulence characteristics near
the bed (burst-sweep cycle). Current ripples have an asymmetric profile with a relatively steep downstream
face (lee-side) and a relatively gentle upstream face (stoss-side). As the velocities near the bed become
larger, the ripples become more irregular in shape, height and spacing yielding strongly three-dimensional
ripples. In that case the variance of the ripple length and height becomes rather large. These ripples are called
lunate ripples when the ripple front has a concave shape in the current direction (crest is moving slower than
wing tips) and are called linguoid ripples when the ripple front has a convex shape (crest is moving faster
than wing tips). The largest ripples may have a length up to the water depth and are commonly called mega-
ripples.
Another typical bed form type of the lower regime is the dune-type bed form. Dunes have an asymmetrical
(triangular) profile with a rather steep lee-side and a gentle stoss-side. A general feature of dune type bed
forms is lee-side flow separation resulting in strong eddy motions downstream of the dune crest. The length
of the dunes is strongly related to the water depth (h) with values in the range of 3 to 15 h. Extremely large
dunes with heights of the order of 7 m and lengths of the order of 500 m have been observed in the Rio

Parana River (Argentina) at water depths of about 25 m, velocities of about 2 m/s and bed material sizes of
about 0.3 mm.
The formation of dunes may be caused by large-scale fluid velocity oscillations generating regions at regular
intervals with decreased and increased bed-shear stresses, resulting in the local deposition and erosion of
sediment particles.
The largest bed forms in the lower regime are sand bars (such as alternate bars, side bars, point bars, braid
bars and transverse bars), which usually are generated in areas with relatively large transverse flow
components (bends, confluences, expansions). Alternate bars are features with their crests near alternate
banks of the river. Braid bars actually are alluvial "islands" which separate the anabranches of braided
streams. Numerous bars can be observed distributed over the cross-sections. These bars have a marked
streamwise elongation. Transverse bars are diagonal shoals of triangular-shaped plan along the bed. One side
may be attached to the channel bank. These type of bars generally are generated in steep slope channels with
a large width-depth ratio. The flow over transverse bars is sinuous (wavy) in plan. Side bars are bars
connected to river banks in a meandering channel. There is no flow over the bar. The planform is roughly
triangular. Special examples of side bars are point bars and scroll bars.
It is a well-known phenomenon that the bed forms generated at low velocities are washed out at high
velocities. It is not clear, however, whether the disappearance of the bed forms is accomplished by a decrease
of the bed form height, by an increase of the bed form length or both. Flume experiments with sediment
material of about 0.45 mm show that the transition from the lower to the upper regime is effectuated by an
increase of the bed form length and a simultaneous decrease of the bed form height. Ultimately, relatively
long and smooth sand waves with a roughness equal to the grain roughness are generated.
In the transition regime the sediment particles will be transported mainly in suspension. This will have a
strong effect on the bed form shape. The bed forms will become more symmetrical with relatively gentle lee-
side slopes. Flow separation will occur less frequently and the effective bed roughness will approach to that
of a plane bed. Large-scale bed forms with a relative height (/h) of 0.1 to 0.2 and a relative length (/h) of 5
to 15 were present in the Mississippi river at high velocities in the upper regime. The shape of the bed forms
was not reported. Probably, more or less symmetrical sand waves were generated because the effective bed
roughness was rather small at high velocities.
In the supercritical upper regime the bed form types will be plane bed and/or anti-dunes. The latter type of
bed forms are sand waves with a nearly symmetrical shape in phase with the water surface waves. The anti-
dunes do not exist as a continuous train of bed waves, but they gradually build up locally from a flat bed.
Anti-dunes move upstream due to strong lee-side erosion and stoss-side deposition. Anti-dunes are bed forms
with a length scale of about 10 times the water depth ( @ 10 h). When the flow velocity further increases,
finally a stage with chute and pools may be generated.
Bed load transport

Usually, the transport of particles by rolling, sliding and saltating is called the bed-load transport. For example,
Bagnold (1956, 1966) defines the bed-load transport as that in which the successive contacts of the particles
with the bed are strictly limited by the effect of gravity, while the suspended-load transport is defined as that in
which the excess weight of the particles is supported by random successions of upward impulses imported by
turbulent eddies. Einstein (1950), however, has a somewhat different approach. Einstein defines the bed-load
transport as the transport of sediment particles in a thin layer of 2 particle diameters thick just above the bed by
sliding, rolling and sometimes by making jumps with a longitudinal distance of a few particle diameters. The bed
layer is considered as a layer in which the mixing due to the turbulence is so small that it cannot influence the
sediment particles, and therefore suspension of particles is impossible in the bed-load layer. Further, Einstein
assumes that the average distance travelled by any bed-load particle (as a series of successive movements) is a
constant distance of 100 particle diameters, independent of the flow condition, the transport rate and the bed
composition. In the view of Einstein, the saltating particles belong to the suspension mode of transport, because
the jump lengths of saltating particles are considerably larger than a few grain diameters.

Suspended load transport
When the value of the bed-shear velocity exceeds the particle fall velocity, the particles can be lifted to a level at
which the upward turbulent forces will be comparable to or higher than the submerged particle weight resulting
in random particle trajectories due to turbulent velocity fluctuations. The particle velocity in longitudinal
direction is almost equal to the fluid velocity. Usually, the behaviour of the suspended sediment particles is
described in terms of the sediment concentration, which is the solid volume (m) per unit fluid volume (m) or
the solid mass (kg) per unit fluid volume (m).
Figure 3
Definition sketch of suspended sediment transport
Observations show that the suspended sediment concentrations decrease with distance up from the bed. The rate
of decrease depends on the ratio of the fall velocity and the bed-shear velocity (ws/u*).
The depth-integrated suspended-load transport (qs,c) is herein defined as the integration of the product of velocity
(u) and concentration (c) from the edge of the bed-load layer (z=a) to the water surface (z=h).
This definition requires the determination of the velocity profile, concentration profile and a known
concentration (ca) close to the bed (z=a), see Figure 3. These latter parameters are refered to as the reference
concentration and the reference level: ca at z=a.
Sometimes, the suspended load transport is given as a mean volumetric concentration defined as the ratio of the
volumetric suspended load transport (= sediment discharge) and the flow discharge: cmean=qs,c/q.
The mean concentration (cmean) is approximately equal to the depth-averaged concentration for fine sediments
(mud). The concentration can be expressed as a weight concentration (cg) in kg/m or as a volume concentration
(cv) in m/m. Sometimes the volume concentration is expressed as a volume percentage after multiplying
with 100%.
Some rivers carry very high concentrations of fine sediments (particles < 0.05 mm), usually refered to as the
wash load. Experience shows that the presence of fines enhances the suspended sand transport rate because the
fluid viscosity and density are increased by the fine sediments. As a result the fall velocity of the suspended sand
particles will be reduced with respect to that in clear water and hence the suspended sand transport capacity of
the flow will increase.
Generally, the sediment concentration distribution over the water depth is described by the diffusion approach,
which yields for steady, uniform flow: c ws + es dc/dz=0, with c= sand concentration, ws= particle fall velocity
and es= sediment diffusivity coefficient.

2.4.2.2 Sand transport in non-steady (tidal) flow
In non-steady flow the actual sediment transport rate may be smaller (underload) or larger (overload) than the
transport capacity resulting in net erosion or deposition assuming sufficient availability of bed material (no
armour layers).
Bed-load transport in non-steady flow can be modelled by a formula type of approach because the adjustment of
the transport of sediment particles close to the bed proceeds rapidly to the new hydraulic conditions.
Suspended load transport, however, does not have such a behaviour because it takes time (time lag effects) to
transport the particles upwards and downwards over the depth and therefore it is necessary to model the vertical
convection-diffusion process.
Tidal flow is characterized by a daily ebb and flood cycle with a time scale of 6 to 12 hours (semidiurnal or
diurnal tide) and by a neap-spring cycle with a time scale of about 14 days.
Sediment concentration measurements in tidal flow over a fine sand bed (0.05 to 0.3 mm) show a continuous
adjustment of the concentrations to the flow velocities with a lag period in the range of 0 to 60 minutes
Figure 4
Time lag of suspended sediment concentrations in tidal flow
The basic transport process in tidal flow is shown in Figure 4. Sediment particles go into suspension when the
current velocity exceeds a critical value. In accelerating flow there always is a net vertical upward transport of
sediment particles due to turbulence-related diffusive processes, which continues as long as the sediment
transport capacity exceeds the actual transport rate. The time lag period DT1 is the time period between the time
of maximum flow and the time at which the transport capacity is equal to the actual transport rate. After this
latter time there is a net downward sediment transport because settling dominates yielding smaller
concentrations and transport rates. In case of very fine sediments (silt) or a large depth, the settling process
can continue during the slack water period giving a large time lag (DT2) which is defined as the period
between the time of zero transport capacity and the start of a new erosion cycle. Figure 4 shows that the
suspended sediment transport during decelerating flow is always larger than during accelerating flow.
Time lag effects can be neglected for sediments larger than about 0.3 mm and hence a quasi-steady approach
based on the available sediment transport formulae can be applied.
Effect of salinity stratification

In a stratified estuary a high-density salt wedge exists in the near-bed region resulting in relatively high near-bed
densities and relatively low near-surface densities. Stratified flow will result in damping of turbulence because
turbulence energy is consumed in mixing of heavier fluid from a lower level to a higher level against the action
of gravity.

The usual method to account for the salinity-related stratification effect on the velocity and concentration
profiles is the reduction of the fluid mixing coefficient by introducing a damping factor related to the
Richardson-number (Ri).
Effect of neap-spring tidal cycle

The tidal range of a Neap-Spring tidal cycle varies as a function of time which is mainly caused by astronomical
effects. Apart from astronomical effects, there are also climatological effects (wind effects) which are
superimposed on the astronomical variations.
Sand transport computations in tidal conditions requires representation of the Neap-Spring cycle. This can be
simply done by multiplying the velocities of the mean tidal cycle by a (correction) factor to account for the
higher velocities and hence higher transport rates (power-law relationship of transport and velocity) during
springtide conditions. The correction factor varies roughly between 1.05 and 1.1.
Effect of mud
In most tidal basins the sediment bed consists of a mixture of sand and mud. The sand-mud mixture generally
behaves as a mixture with cohesive properties when the mud fraction (all sediments<0.05 mm) is dominant
(>0.3) and as a non-cohesive mixture when the sand fraction is dominant (>0.7). The distinction between non-
cohesive mixtures and cohesive mixtures can be related to a critical mud content (pmud,cr). Most important is the
value of the clay-fraction (sediments<0.005 mm) in the mixture. Cohesive properties become dominant when the
clay-fraction is larger than about 5% to 10%. Assuming a clay-mud ratio of 1/2 to 1/4 for natural mud beds, the
critical mud content will be about pmud,cr=0.2 to 0.4.
Laboratory and field observations have shown that the erosion or pick-up process of the sand particles is
slowed down by the presence of the mud particles. This behaviour can be quite well modelled by increasing
the critical bed-shear stress for initiation of motion of the sand particles.
2.4.2.3 Sand transport in combined non-steady (tidal) flow and oscillatory flow (waves)
Basic characteristics of coastal transport

Sand transport in a coastal environment generally occurs under the combined influence of a variety of
hydrodynamic processes such as winds, waves and currents. Figure 5 gives a schematic representation of
sand transport mechanisms along a cross-shore bed profile of a straight sandy coast.
Sand can be transported by wind-, wave-, tide- and density-driven currents (current-related or advective
transport), by the oscillatory water motion itself (wave-related or oscillating transport) as caused by the
deformation of short waves under the influence of decreasing water depth (wave asymmetry), or by a
combination of currents and short waves. The waves generally act as sediment stirring agents; the sediments
are transported by the mean current. Low-frequency waves interacting with short waves may also contribute
to the sediment transport process. Wind-blown sand represents another basic transport process in the beach-
dune zone.
In friction-dominated deeper water outside the breaker (surf) zone the transport process is generally
concentrated in a layer close to the sea bed; bed-load transport (bed form migration) and suspended transport
may be equally important. Bed load type transport dominates in areas where the mean currents are relatively
weak in comparison with the wave motion (small ratio of depth-averaged velocity and peak orbital velocity).
Suspension of sediments can be caused by ripple-related vortices. The suspended load transport becomes
increasingly important with increasing strength of the tide- and wind-driven mean current, due to the
turbulence-related mixing capacity of the mean current (shearing in boundary layer). By this mechanism the
sediments are mixed up from the bed-load layer to the upper layers of the flow.
In the surf zone of sandy beaches the transport is generally dominated by waves through wave breaking and
the associated wave-induced currents in the longshore and cross-shore directions. The longshore transport in
the surf zone is also known as the longshore drift. The breaking process together with the near-bed wave-
induced oscillatory water motion can bring relatively large quantities of sand into suspension (stirring) which
is then transported as suspended load by net (wave-cycle averaged) currents such as tide-, wind- and density

(salinity)-driven currents. The concentrations are generally maximum near the plunging point and decrease
sharply on both sides of this location.
The nature of the sea bed (plane or rippled bed) has a fundamental role in the transport of sediments by waves
and currents. The configuration of the sea bed controls the near-bed velocity profile, the shear stresses and the
turbulence and, thereby, the mixing and transport of the sediment particles. For example, the presence of ripples
reduces the near-bed velocities, but it enhances the bed-shear stresses, turbulence and the entrainment of
sediment particles resulting in larger overall suspension levels. Several types of bed forms can be identified,
depending on the type of wave-current motion and the bed material composition. Focussing on fine sand in the
range of 0.1 to 0.3 mm, there is a sequence starting with the generation of rolling grain ripples, to vortex ripples
and, finally, to upper plane bed with sheet flow for increasing bed-shear. Rolling grain ripples are low relief
ripples that are formed just beyond the stage of initiation of motion. These ripples are transformed into more
pronounced vortex ripples due to the generation of sediment-laden vortices formed in the lee of the ripple crests
under increasing wave motion. The vortex ripples are washed out under large storm waves (in shallow water)
resulting in plane bed sheet flow characterised by a thin layer of large sediment concentrations.
Figure 5
Sand transport mechanisms along cross-shore profile
Top: cross-shore distribution of wave heights
Middle: cross-shore distribution of longshore current
Bottom: sand transport processes along cross-shore profile
The sand transport processes induced by (non-breaking) waves in the lower rippled bed regime, are:
strong vortex motions over steep ripples yielding concentrated suspension clouds moving upward,
forward and backward in the water column; spatial sediment concentration variability is relatively large;
phase differences between the velocity peaks and the concentration peaks are prominent, which may
result in net offshore-directed sand transport under shoaling swell waves (see Fig. 6);
suspended sand transport processes generally are dominant.

The sand transport processes induced by (breaking) waves in the upper plane bed regime, are:
generation of a high-concentration layer (sheet flow layer) in direct contact with the bed (0-1 cm);
phase differences between the velocity peaks and the concentration peaks are generally small, but
become important for fine sediment (<0.2 mm) and strongly asymmetric wave motion, which may result
in net sand transport against the wave-current direction at certain levels in the oscillatory boundary
layer;
sediment concentrations above the sheet flow layer are generally relatively small (except in the presence
of strong wave-induced currents);
bed-load transport is generally dominant in non-breaking waves without external currents; suspended
transport is generally dominant in breaking wave conditions with external currents.
There is an essential difference between the sediment transport processes that occur above rippled and plane
sand beds. It is important therefore to define the boundary delineating these regimes. Steep ripples are
formed by relatively low waves in relatively deep water (i.e. typical offshore conditions). Such ripples tend
to be long crested (two-dimensional), and they have steepness (h/l) greater than about 0.15, where h and l
are the ripple height and length, respectively. The roughness (ks) of a rippled bed is equivalent to about 3-4
ripple heights (ks = 3h- 4h). Beneath steeper waves in shallower water (for example, at edge of the surf
zone), the ripples start to become washed out; their steepness decreases, causing ks to decrease, and they tend
to become shorter crested (transitional 2D-3D profiles). Finally, beneath very steep and breaking waves, the
ripples are washed away completely, the bed becomes plane (sheet flow regime), and the bed roughness ks
decreases still further (scaling on the sand grain diameter).
Figure 6
Transport processes in asymmetric wave motion over plane and rippled bed
Left: plane bed
Right: rippled bed
The net transport components induced by non-breaking waves along a straight coast, are:
net onshore-directed wave-related transport due to asymmetry of the near-bed orbital velocities with
relatively large onshore peak velocities under the wave crests and relatively small offshore peak
velocities under the wave troughs;
net onshore-directed current-related transport due to the generation of a quasi-steady onshore-directed
weak mean current in the wave boundary layer; net offshore-directed mean current may be generated in
strongly asymmetric wave motion over a steep slope; bed ripples may also have a directional effect on
the mean current;

net offshore-directed wave-related transport due to the generation of bound long waves associated with
variations of the radiation stresses under irregular wave groups (peak velocities and concentrations are
out of phase).
The net transport components induced by breaking waves along a straight coast, are:
net onshore-directed wave-related transport due to asymmetric wave motion;
net offshore-directed current-related transport due to the generation of a net return current (undertow) in
the near-bed layers balancing the onshore mass flux between the wave crest and trough;
wave-related transport rates due to the generation of low-frequency wave motion (bound long waves
and free waves may be generated, but the direction of net transport rate is highly uncertain);
offshore-directed current-related transport due to the generation of meso-scale circulation cells with
local offshore rip currents;
net longshore current-related transport due to cross-shore gradients in radiation stress.
2.4.3 Mud transport
2.4.3.1 General characteristics, definitions and modelling approaches
Sediment mixtures with a fraction of clay particles larger than about 10% have cohesive properties because
electro-statical forces comparable to or higher than the gravity forces are acting between the particles.
Consequently, the sediment particles do not behave as individual particles but tend to stick together forming
aggregates known as flocs whose size and settling velocity are much larger than those of the individual particles.
Most clay minerals have a layered (sheet-like) structure. The most important types of clay minerals are: kaolinite
(two-layer structure), montmorillonite (three-layer structure), illite (three-layer structure) and chlorite (four-
layer-structure). Herein, clay is defined as sediment with sizes smaller than 0.008 mm (<8 mm); silt as sediment
with sizes between 0.008 and 0.062 mm (8 to 62 mm) and sand as sediment with size between 0.062 and 2 mm
(62 to 2000 mm).
Mud is herein defined as a fluid-sediment mixture consisting of (salt) water, sands, silts, clays and organic
materials.
In a natural environment there is a continuous transport cycle of mud material which consists of: erosion,
settling, deposition, consolidation, erosion and so on. As a result of the complexity and the lack of fundamental
knowledge, the description of the various processes is largely empirical. Most information is based on laboratory
experiments. Pioneering work has been done by Krone (1962), Partheniades (1965) and Mehta-Partheniades
(1975). Laboratory results are, however, often not representative because the biogenic mechanisms and the
organic materials are missing, the water depths are small, the deposition and the consolidation history of the bed
is different from that in nature. Therefore, in-situ research should be carried out to identify the effects of organic
materials and sediment-inhabiting organisms such as bacteria and worms. There are no sediment deposits that
are not inhabited by biota and often the biological component is dominant.
As a result of the layered structure of mud suspensions with a high-concentration (fluid mud) layer near the bed
and a low-concentration (dilute mud suspension) layer near the surface, it is attractive to model each layer
separately.
The concentration in the upper part of the water column can be modelled as a depth-uniform concentration using
a two-dimensional horizontal advection-diffusion approach with critical bed-shear stresses for deposition and
erosion. Deposition being equal to the product of concentration and settling velocity acts as a sink, while erosion
prescribed as a vertical flux into the water column depending on the excess bed-shear stress is a source of
sediment from the bed.
The generation and transportation of fluid mud near the bed can be simulated by a combination of the mass
balance equation and the momentum equation (Odd and Cooper, 1989).
A detailed three-dimensional approach requires the representation of the high-concentration phenomena near
the bed such as hindered settling, damping of turbulence by large concentration gradients and consolidation
effects (Winterwerp, 1999).

Basic mud processes
The most important properties and processes are summarized herein (see Figure 1):
1. cohesion, viscosity and yield stress
2. flocculation,
3. settling,
4. deposition,
5. saturation,
6. consolidation,
7. erosion,
8. transport.
Water surface
Flocculation
Break-up
Erosion Resuspension
Sedimentation
Resuspension
Figure 7
Transport cycle of mud aggregates
2.4.3.2 Cohesion
If a cohesive soil sample with a low water content is submitted to shear stresses (t) under various normal
pressures (s) to the point of failure, the relationship between t and s can be expressed as (Law of Coulomb):
t=ty+s tanj, with ty = yield stress and j= angle of internal friction.
The yield stress is generally interpreted as the "cohesion" of the sample. Thus, a cohesive sediment sample is
able to withstand a finite shear stress for s= 0 (no deformation). The angle of internal friction represents the
mechanical resistance to deformation by friction and interlocking of the individual particles.
Plasticity is the property of cohesive material to undergo substantial permanent deformation without breaking.
Types of basic rheological behaviour of mud are shown in Figure 8. Dilute suspensions with concentrations
smaller than 10 kg/m show a Newtonian behaviour. Deviations from this latter behaviour tend to occur at
concentrations larger than 10 kg/m. High-concentration (>50 kg/m) suspensions of water, fine sand, silt, clay
and organic material usually have a pseudo-plastic or a Bingham plastic shearing behaviour, which means that
the relationship between shear-stress t and shear rate (du/dz) is non-linear (Figure 8). The slope of the curve
expresses the effective dynamic viscosity () of the material. The intercept with the shear stress axis is called the
yield stress and represents the initial particle interaction force which must be overcome to give a distortion of the
material. The yield stress of an ideal Bingham type mud is called the Bingham stress.
Viscosity and yield stress can be measured in a roto-viscometer. This instrument consists of two concentric
cylinders in which the sediment material is placed. One cylinder is rotated at a constant rate giving a constant
shearing rate (du/dz), while the force is measured on the other cylinder giving the shear stress.

Experimental data show that the yield stress is proportional to the sediment concentration. Many attempts have
been made to relate the yield stress to the chemical and physical properties of the fluid-sediment mixture.
General relationships, however, do not exist.
Shear stress (tau)

C D
A
Tau-y B
slope
Tau-b
Tau-y
Shear rate (du/dz)
Figure 8
Rheological behaviour
A. Pseudoplastic, B. Newtonian, C. Ideal Bingham, D. Bingham plastic
2.4.3.3 Flocculation
Most of the individual clay particles have a negative charge. The mutual forces experienced by two or more clay
particles in close proximity are the result of the relative strength of the attractive and repulsive forces. The
attractive forces, also known as Van der Waals (VA) forces, are due to the interaction of the electrical fields
formed by the dipoles of the individual molecules. The repulsive forces (VR) are due to ion clouds of similar
charge repelling each other. Positive ions present in the fluid form a cloud of ions around the negatively charged
clay particles (double layer theory). The result can be either attraction or repulsion depending on the relative
strength (depending on number of positive ions) and the distance between the particles.
In fresh water suspensions (few positive ions) the repulsive forces between the negatively charged particles
dominate and the particles will repel each other. In saline water the attractive forces dominate due to the
(abundant) presence of positive sodium-ions forming a cloud of positive ions (cations) around the negatively
charged clay particles resulting in the formation of flocs (aggregates), as shown in Figure 9.
Other binding forces are chemical forces (hydrogen bonds, cementation, coatings of organic materials).
Figure 9
Sizes of individual clay particles, flocs and floc groups

Flocculation requires particle collisions. The three collision mechanisms are:
the Brownian motions of the particles (< 0.005 mm; <5mm ) due to the random bombardment by the
thermally agigated water molecules; the number of collisions is linearly proportional to the concentration;
turbulent mixing due to the presence of velocity gradients in the fluid;
differential settling velocities because the larger particles have larger settling velocities and may therefore
"fall" on the smaller particles.
Other factors affecting flocculation are: size, concentration of particles, salinity, temperature and organic
material.
A small particle size in combination with a large concentration greatly intensifies the flocculation process
because these two factors yield a small relative distance between the particles.
Experimental research has shown that flocculation quickly reaches an equilibrium situation at a salinity of about
5 to 10 promille which is small compared to that (35 promille) of sea water.
A high temperature also enhances the flocculation process because the double layer repulsive energy decreases
in magnitude and this leads to a decreased repulsion.
Organic materials in and on the flocs significantly intensify the flocculation process, because of the binding
properties of the organic materials. The binding forces become larger due to the presence of organic material
(biogenic forces) and the flocs are becoming larger.
Break up of the flocs is caused by large shearing forces in the fluid when these forces are larger than the strength
of the flocs; the flocs are broken into smaller flocs or particles. Large shearing forces exist close to the bottom
where the velocity gradients are largest. Large shearing forces also exist in small-scale eddies everywhere in the
fluid. Under the influence of turbulent forces there is a continuous process of flocculation and break-up resulting
in a dynamic equilibrium of the flocs (size, density and strength).
In still water (no turbulence) the flocs may grow to larger sizes due to differential settling collisions. However, as
the flocs get larger they fall faster until the fluid shear on the flocs becomes greater than the floc strength
resulting in break-up.
Analysis of under-water photographs shows the presence of macroflocs with sizes in the range of 100 to 1000
mm, miniflocs with sizes in the range of 10 to 100 mm and single mineral particles smaller than about 10 mm.
When the flocs grow larger, the floc size increases but the density of the flocs (consisting of sediment, fluid,
organic materials) becomes smaller. Individual clay particles will have an excess density of about 1600 kg/m3.
Large flocs of about 1 mm may have a density in the range of 1 to 10 kg/m in excess of the fluid density,
because most of the floc consists of (pore) fluid.
2.4.3.4 Settling
An important parameter in sedimentation studies of cohesive materials is the settling velocity of the flocs.
Analysis of laboratory and field data has shown that the settling velocity of the flocs is strongly related to the
salinity and the sediment concentration (c). It may also be related to the type of measuring instrument.
The settling velocity increases with salinity for salinities up to 15 ppt (15 promille). A further increase of the
salinity does not yield a much larger settling velocity if the mud concentration is smaller than 1000 mg/l. When
the sediment concentration is larger than 1000 mg/l, the settling velocity increases with salinity over the full
range up to 30 ppt.
In saline suspensions with sediment concentrations up to about 1000 mg/l an increase of the settling velocity
with concentration has been observed as a result of the flocculation effect both in laboratory and in field
conditions.
When the sediment concentrations are larger than approximately 10,000 mg/l, the settling velocity decreases
with increasing concentrations due to the hindered settling effect (Figure 10). Hindered settling is the effect that
the settling velocity of the flocs is reduced due to an upward flow of fluid displaced by the flocs. At very large
concentrations the vertical fluid flow can be so strong that the upward fluid drag forces on the flocs become
equal to the downward gravity forces resulting in a temporary state of dynamic equilibrium with no net vertical
movement of the flocs. This state which occurs close to bed, generally is called fluid mud. In the laboratory the

hindered settling velocity can be quite accurately determined from consolidation tests by measuring the
subsidence of the sediment-fluid interface.
The settling velocity in the two ranges can be expressed as:
ws,m= k cm in flocculating suspensions (10-10,000 mg/l)
ws,m=ws(1-ac)b in hindered-settling suspensions (> 10,000 mg/l)
in which: ws,m =settling velocity of flocs in fluid-sediment mixture, ws= settling velocity of individual particles,
c= volume concentration, m =coefficient (= 1 to 2), k=coefficient, a=coefficient, b= coefficient (= 3 to 5).
Settling velocities based on in-situ settling tube measurements as a function of concentration in saline conditions
from all over the world are shown in Figure 10 (Severn, Avonmouth, Thames, Mersey in England; Western
Scheldt in The Netherlands; River Scheldt in Belgium; Brisbane in Australia; Chao Phya in Thailand, Demerara
in South America).
The settling velocities based on analysis of in-situ video camera recordings were found to be considerably larger
(factor 10) than those derived from settling tube measurements. It may be concluded that the use of an in-situ
settling tube may result in relatively small settling velocities because of floc destruction during the mechanical
sampling procedure. The video camera system may offer promising results, but it should be realized that the
larger flocs are better observed by the camera than the smaller flocs. The in-situ settling tube represents the total
distribution of particles, but the larger flocs may be destroyed during closing of the valves to trap the fluid-
sediment sample.
Various settling models were tested against settling velocity measurements. Based on these results it is apparent
that while some models incorporate more of the physical processes than others, they are not necessarily more
accurate for any given situation. The variation between measurements is of the same order as the variation
between measurement devices in which case use of a sophisticated settling model is unwarranted. Until the
quality and quantity of settling velocity data allows an informed choice to be made, it is suggested that the
simplest model is used. The settling velocity can also be derived from measured concentrations using the Rouse
concentration profile (during slack tide when deposition is maximum).
Figure 10
The influence of sediment concentration on the settling velocity

2.4.3.5 Deposition
Deposition is predominant when the bed-shear stress falls below a critical value for deposition (td).
Basically, two critical bed-shear stresses for deposition can be distinguished: lower (minimum) value td,full and
upper (maximum) value td,part. All sediment will settle out if the applied bed-shear stress is smaller than the
lower value: tb<td,full. No sediment will settle out if the applied bed-shear stress is larger than the upper value:
tb>td,part. Partial deposition of the larger flocs and particles will occur for: td,full<tb<td,part.
Deposition tests in flumes are typically conducted by initially suspending the sediment (initial concentration co)
at a very high flow velocity, then lowering the velocity (or bed-shear stress) and recording the decrease in
suspension concentration with time to the new equilibrium concentration (ceq). The new equilibrium
concentration (ceq) will be zero, if tb<td,full (all sediment will settle out; ceq=0). The new equilibrium concentration
will be equal to co, if tb>td,part (no deposition; ceq=co). In tidal conditions this process will continuously take
place during decelerating flow (decreasing velocities).
The time-dependent behaviour of the concentration can be described as:

c= ceq + (co-ceq) e-(aws/h)t with co= initial concentration, ceq= new equilibrium concentration, t= time, ws= settling
velocity, h= depth, a= coefficient.
The deposition rate is: D= h dc/dt= a (co-ceq) ws, with a<1 for tb<td,full and a=1 for tb>td,full
Fluid mud of sediment flocs is defined to be present for concentrations above 10 kg/m in which the flocs are
partially supported by the escaping fluid (known as the hindered settling effect) and partially supported by
interfloc contacts. In nature with large water depths, there will be a situation with concentrations increasing
towards the bed. The settling velocity reduces for increasing concentrations (hindered settling). Thus, the settling
velocity will be relatively small near the bed and relatively large at the upper layers resulting in the formation of
a near-bed fluid mud layer (especially in neap-tide conditions) with increasing sediment concentrations in the
range of 10 to 300 kg/m. The thickness of the fluid mud layer will increase as long as the deposition rate at the
upper side of the layer is larger than the consolidation rate at the bottom side. Values up to several metres have
been observed. The fluid mud layer can be transported horizontally as a turbidity current due to:
gravity forces on a sloping bottom,
pressure gradient forces due to horizontal differences in sediment concentrations and fluid densities,
shear forces at the interface generated by the overflowing water.
Turbulent mixing at the interface (lutocline) between the fluid mud layer and the overlying water generally is
small because of the stabilizing effect of the heavier sediment material which has to be moved against gravity.
The deposition process in the concentration range of 0.3 to 10 kg/m3 is dominated by flocculation effects. Two
groups of flocs exist. The first group is characterised by flocs with strong bonds which can resist the disruptive
near-bed shear stresses and will be able to reach the bottom and form strong bonds with deposited flocs. The
second group is characterised by flocs which do not have sufficiently strong bonds and will be broken down
before reaching the bed resulting in resuspension or are eroded very quickly after being deposited because the
bonds between the flocs and the bed are relatively weak. The deposited flocs develop bonds with the bed flocs.
The bed-shear stress to disrupt these bonds and to overcome the submerged floc weight has to be larger than that
at which the floc is deposited. This larger value is called the critical bed-shear stress for erosion (te).

Figure 11
Ratio of equilibrium concentration and initial concentration as a function of bed shear stress based on Mehta
and Partheniades (1975)
Based on the experimental results of Mehta and Partheniades (1975), the following three processes can be
distinguished (see also Figure 11).
1. Full deposition; All sediment particles and flocs are deposited when the bed-shear stress (tb) is smaller
than the bed-shear stress for full deposition.
2. Hindered or partial deposition; The group of relatively strong flocs is deposited, whereas the group of
relatively weak flocs with a shear strength smaller or equal to the applied bed-shear stress (tb) is broken up
and remains in suspension. The relative magnitude of both groups depends on the bed-shear stress.
3. No deposition; No sediment flocs are deposited when the bed-shear stress is larger than the maximum bed-
shear stress for deposition (td,part).
The flocculation effect appears to be of minor importance for concentrations smaller than 0.3 kg/m3. All
sediment particles will eventually be deposited yielding full deposition, if tb< td,full.
Based on the experimental results of Krone (1962) for low concentrations (c<0.3 kg/m), the following two
processes can be distinguished:
1. No deposition,
2. Full deposition.
Based on the experimental work of Mehta and Partheniades (1975), two critical bed-shear stresses for
deposition can be distinguished: td,full and td,part.
The minimum bed-shear stress for full deposition (td,full) is defined as the bed-shear stress below which full
deposition of the sediment flocs will occur. The maximum bed-shear stress for deposition (td,part) is defined as
the bed-shear stress above which no deposition of the sediment flocs will occur. Experimental values of td,full are
in the range of 0.05 to 0.15 N/m2; experimental values of td,part are in the range of 1.5 to 2 N/m2.
2.4.3.6 Saturation
Winterwerp (1999, 2001) discusses the existence of a saturation concentration for cohesive sediments in
depositional conditions. When the velocities are decreasing after peak flow in tidal conditions, the sediments
start to settle to form a layer of fluid mud, creating a two-layer fluid system. At the interface between the two
layers, the vertical turbulent mixing is damped strongly resulting in an overall decrease of the sediment-
carrying capacity in the water column. At a certain moment the turbulence field collapses and the sediment
concentrations are greatly reduced. The mud concentration just prior to this collapse is denoted by the term

saturation concentration. The conditions with concentration larger and smaller than the saturation
concentration are defined as supersaturation (overloading) and subsaturation (underloading). The concept of
the saturation concentration for cohesive sediment is based on empirical evidence that when the flux
Richardson number at a specific level in the water column exceeds a critical value (of about 0.2), the
turbulent field collapses above this level.
2.4.3.7 Consolidation
The increase or decrease of the fluid-bed interface is equal to the rate of deposition (from the suspension) or
erosion minus the rate of consolidation (of the bed). Consolidation is a process of floc compaction under the
influence of gravity forces with a simultaneous expulsion of pore water and a gain in strength of the bed
material.
Generally, three consolidation stages are distinguished:
initial stage (days) : The process consists of hindered settling and consolidation. The flocs in a
freshly deposited layer are grouped in an open structure with a large pore
volume. The weakest bonds are broken down first and the network gradually
collapses. The bed surface sinks with time t.
secondary stage (weeks) : The pore volume between the flocs is further reduced. Small thin vertical
pipes (drains) are formed allowing the pore water to escape. The bed surface
sinks with t0.5 or log(t).
final stage (years) : The pore volume inside the flocs is further reduced and the flocs are broken
down. The bed surface sinks with log(t).
The consolidation process is strongly affected by the:

initial thickness of the mud layer (ho),
initial concentration of the mud layer (co),
permeability (k) of the mud layer (sediment composition and size, content of organic material, salinity, water
temperature).
The consolidation of natural muds proceeds relatively fast in a thin layer and relatively slow in a thick layer.
Mud samples with a low initial density tend to have a somewhat lower final density than the samples with the
higher initial density.
The consolidation time scale of natural mud (with low percentage of sand) is much larger than that of pure
kaolinite. A kaolinite layer with a thickness of 0.1 m obtains its final density after about 7 days, whereas natural
muds show a time scale of about 1 month. This latter behaviour is probably related to the presence of organic
materials (with a dry density of about 1200 to 1500 kg/m) which cause biochemical reactions resulting in gas
production. Gas bubbles may escape during the initial consolidation phase when the material strength is low,
thereby retarding the consolidation process. During later stages (or deeper layers) with increasing material
strength, the gas bubbles may remain in the bed and affect the consolidation process, depending on the
permeability of the bed material.
Information of the vertical distribution of the sediment concentration (dry density) of consolidated mud layers
can be obtained from experimental data. The dry densities in the top layer of mud deposits are in the range of 50
to 250 kg/m depending on the sediment composition and consolidation time. The dry densities in the lower
layer are in the range of 400 to 1500 kg/m depending on the percentage of sand and thickness of the deposit.
Values of the wet and dry sediment densities in successive consolidation stages in estuarine conditions are given
in Table 2. Various consolidation stages are distinguished. Mud with a wet density of about 1200 kg/m is still
sailable for ships. The nautical depth in the harbour of Rotterdam is defined at a wet density of 1200 kg/m. The
transition from the fluid to the solid phase occurs at a wet density of about 1300 kg/m.

Consolidation stage Rheological Wet sediment density Dry sediment density
behaviour (kg/m) (kg/m)
freshly consolidated dilute fluid mud 1000-1050 0-100
(1 day)
weakly consolidated fluid mud (Bingham) 1050-1150 100-250

(1 week)
medium consolidated dense fluid mud 1150-1250 250-400

(1 month) (Bingham)
highly consolidated fluid-solid 1250-1350 400-550

(1 year)
stiff mud (10 years) solid 1350-1400 550-650
hard mud (100 years) solid > 1400 > 650
Table 2
Density ranges of consolidated mud
2.4.3.8 Erosion
Sediment particles, flocs or lumps of the bed surface (including fluid mud layers) will be eroded when the
applied current-induced or wave-induced bed-shear stress (tb) exceeds a critical value for erosion (te), which
depends on the bed material characteristics (mineral composition, organic materials, salinity, density etc.) and
bed structure. Experimental results show that the critical bed-shear stress for erosion is strongly dependent on the
deposition and consolidation history.
The critical bed-shear stress for erosion was (by many researchers) found to be larger than the critical bed-shear
stress for full deposition (te >td,full). Mehta and Partheniades (1975) observed partial deposition (and no
erosion) of mud material for (high) bed-shear stresses up to 1.5 N/m in steady flows, because the bed consisted
of strong dense flocs deposited during high-shear conditions in which the weak flocs could not be deposited.
Thus, te may be very large when the top layer of the bed consists of strong flocs deposited at relatively high
velocities (te >td,part).
In tidal conditions with decreasing velocities (near slack tide) the weak flocs will be deposited on top of the
strong flocs deposited earlier at higher velocities. At increasing velocities of the next tidal cycle, these weak flocs
can be eroded rather easily at low velocities.
Various types of erosion are distinguished in the literature: (i) particle or floc erosion (surface erosion) which is
the one by one removal of particles and/or aggregates and (ii) mass erosion which is the erosion of clusters or
lumps of aggregates due to failure within the bed. Many attempts have been made to relate the te values to basic
parameters as plasticity index, voids ratio, water content, yield stress and others. Generally-accepted
relationships, however, are not available. Determination of the te values must, therefore, be based on laboratory
tests using natural mud or on in-situ field tests.
Many researchers have observed that the critical bed-shear stress for erosion is related to the sediment
concentration (dry density). The annular flume experiments of Parchure and Mehta (1985) clearly demonstrate
that the te value increases with depth in the bed layer, because the dry sediment density increases with depth. For
a given constant current-related bed-shear stress an equilibrium concentration is established after some time. A
further increase of the current velocity and hence the bed-shear stress will result (after some time) in a new but
larger equilibrium concentration, see Figure 12.

This sequence can be interpreted as erosion of the top layer down to a level at which the bed strength (te) is equal
to the applied bed-shear stress. Then, the erosion is stopped. This interpretation is confirmed by the results of a
special experiment in which the fluid-sediment suspension above the bed was slowly replaced by clear water
showing no further erosion (replaced fluid remained clear). Thus, the establishment of an equilibrium
concentration is caused by an erosion stop (tb= te) and not by saturation of the suspension.
Figure 12
Mud concentration as a function of time after Parchure and Mehta (1985)
The erosion rates for a bed of constant density (te = constant) is, generally, expressed as (Partheniades,
1965):
E=dm/dt=M [(tb-te)/te]
The M-coefficient (in mass per unit area and time) is a material "constant" depending on mineral
composition, organic material, salinity etc. Reported values are in the range of M = 0.00001-0.0005 kg/sm2
for soft natural muds.
2.4.3.9 Transport of mud
Generally, the load of fine sediments in a river is termed the wash load and consists of the finest-grained
portion (<8 mm) of the total sediment load. The wash load shows an almost uniform concentration profile
(small vertical gradients). The wash load typically is controlled by the upstream supply rate and not by the
local river bed characteristics. Generally, the wash load bears no relationship to the discharge of the river.
Concentrations tend to be large when there is a ready source of fine sediments in the drainage basin (soil
erosion). Extremely large concentrations (hyper-concentrations up to 50% sediment by weight in the fluid)
are a common occurrence during flash floods in semi-arid environments. Generally, this may occur in small
rivers. The Yellow river is a dramatic exception with hyper-concentrations in the range of 500 to 1000 kg/m
due to presence of extensive loess deposits in the upper river basin. About 10% of this material consists of
clay (< 0.05 mm). The remaining material consists of silt and fine sand (0.025 to 0.05 mm). Hyper-
concentrations may enhance the total bed material transport because the suspended sand particles have a reduced
settling velocity and remain longer in suspension before they are deposited again on the bed.

Mud can also be transported as a high concentration layer above sloping beds under the action of gravity.
This phenomenon is generally known as a turbidity current. Turbidity currents consisting of a high fraction
of fine silt and clay materials have been observed in reservoirs built in heavy silt-laden rivers in China and
the USA.
An indication for the generation of a turbidity current at the upstream end of a reservoir is the plunging of the
muddy river flow into the deeper reservoir. The generation of a turbidity current into a reservoir usually is a
short-period phenomenon (days), because the incoming flood discharge in arid regions increases and decreases
abruptly. During a flood the silt and clay concentration (due to land surface erosion) increases enormously.
Field observations in tidal flow show the presence of a cyclic process of erosion, transport, settling,
deposition and consolidation.
Generally, a three-layer system can be distinguished in vertical direction (see Figure 13), as follows:
Consolidated mud layer at the bottom with concentrations larger than about 300 kg/m. The flocs and
particles are supported by the internal floc framework.
The mud interface is detectable by echosounding instruments (30 kHz).
Fluid mud suspension layer with concentrations in the range of 10 to 300 kg/m. The layer thickness is
of the order of 0.1 to 1 m in normal conditions and up to 5 m in extreme conditions, as present at the
Amazon shelf. Marked interfaces (lutoclines) can be observed from echosounder recordings or from
nuclear density recordings. The flocs and particles in the fluid mud layer are supported by fluid drag
forces exerted by the escaping fluid (hindered settling effect). The fluid mud layer can be subdivided into
a turbulent upper layer (mixed fluid mud; 10 to 100 kg/m3) and a laminar (viscous) lower layer (100 to
300 kg/m3), depending on conditions. The turbulent upper fluid mud layer will try to (i) mix mud from the
laminar lower layer into the upper turbulent layer and (ii) mix fluid from the upper dilute suspension layer
into the fluid mud layer if the upper fluid mud layer is more turbulent than the dilute suspension layer (in
accelerating spring tide flow). This results in a decreasing concentration in the upper fluid mud layer and a
rise of the upper interface between the dilute layer and the fluid mud layer. If the dilute suspension layer is
more turbulent than the fluid mud layer (in neap tide flow), the fluid mud layer will be relatively thin
(much more stratified) and mud will be mixed up from the fluid mud layer into the dilute suspension layer
by vortices from that layer. Stratified fluid mud near the bed is enhanced by salinity stratification during
neap tide (damping of turbulence) at the frontal zone between river water and sea water.
Dilute mud suspension (Figure 13) with concentrations in the range of 0 to 10 kg/m, which are
detectable by optical methods and mechanical sampling. Mud will be mixed into dilute suspension layer
from the fluid mud layer if the the upper dilute layer is more turbulent than the fluid mud layer.
Flocculation is dominant in the dilute suspension layer. The flocs and particles are supported by
turbulence-induced fluid forces and transported by tide-driven and wind-driven currents.
Figure 13
Vertical distribution of mud concentrations

Figure 14
Variation of mud concentration in tidal flow
Vertical layers of different densities are influenced by gravity processes which oppose the mixing processes.
The stability of the system can be characterized by the gradient Richardson number (Ri). For Ri larger than
unity (based on experimental data) a stable system will be present (interfacial instabilities will die out).
The mud transport variation over a tidal cyle is shown in Figure 14. The Spring-Neap cycle may have a
marked influence on the vertical structure of the mud suspension:
Springtide: nearly uniform (well-mixed) concentration distribution at maximum flood and ebb velocities;
formation of fluid mud interfaces (lutoclines) near the bottom during slack water; lutocline is
moving downward; re-entrainment of mud during the next tide;
Neaptide : tidal velocities are generally too small to cause erosion at the near-bed fluid mud layer; the
fluid mud layer may survive several tidal cycles as a stationary layer.

Figure 15
Flow of water and silt/mud near the salt wedge in a stratified estuary (turbidity maximum)
Figure 16
Characteristic velocity profiles during ebb and flood tide in case of a horizontal landward salinity (density)
gradient
Mathematically, the transport of cohesive sediments in a well-mixed estuary (no vertical salinity
stratification) can be described by the convection-diffusion equation. Since the vertical concentration
distribution above the fluid mud layer is nearly uniform, the convection-diffusion equation can be averaged
over the depth. The bed layer should be represented as a number of sublayers, each with its own thickness,
density, shear strength and consolidation stage. When deposition occurs, the thickness of the bed layer
increases. When erosion occurs, the bed layer thickness decreases. Consolidation causes a continuous
decrease of the bed-layer thickness.

In (partially) stratified estuaries the maximum silt and mud concentrations (turbidity maximum) are usually
found in the area where the salt wedge is migrating during the tidal cycle. Figure 15 schematically shows the
ebb-tide flow of water and mud. Heavy sedimentation will occur in the salt wedge area resulting in the
formation of soft mud layers (fluid mud) on the channel bottom.
During the ebb tide, the river flow erodes the landward end of the mud layer. Near the toe of the wedge the
fresh river water is carried upward from the bottom and flows seaward over the heavy saline water. Intensive
mixing will occur at the interface between the fresh and saline water. In the salt wedge the bottom current is
landward. This saline water meets the fresh water near the toe of the wedge where it is carried upwards.
The mud carried by the river enters the area of the salt wedge together with the mud eroded from the
landward end of the mud layer. This mud is mixed (near the interface) with mud already suspended in the
saline water resulting in additional flocculation and increased settling velocities. Settling will occur seaward
of the sharp salt wedge interface where the mixing is reduced. The mud particles fall towards the bottom
where they are transported landward again by relatively strong bottom currents of saline water (see Fig. 16).
Summarizing, there are relatively strong landward bottom-currents with high mud concentrations during
flood tide and relatively weak seaward bottom-currents with low mud concentrations during ebb tide
resulting in a net transport of mud in landward direction. Deepening of the main channels by dredging may
lead to increased deposition at more inland locations as a result of increased tidal penetration and increased
tidal currents giving rise to higher mud concentrations.
In most tidal basins the sediment bed consists of a mixture of sand and mud. As the mud content (all clay and
silt particles<0.05 mm; 50 mm) increases, clay and silt particles will fill the spaces between the sand particles.
The sand-mud mixture generally behaves as a mixture with cohesive properties when the mud fraction (all
sediments<0.05 mm) is dominant (>20% to 30%) and as a non-cohesive mixture when the sand fraction is
dominant (>80% to 70%). A non-cohesive bed has a granular structure without cohesive forces between the
individual particles. A cohesive bed forms a coherent mass due to electrochemical interactions between the
particles and the bed exhibits a certain shear strength before deformation. Three basic types of mud beds can be
distinguished: liquid, plastic or solid, depending on the water content and clay content (particles<0.005 mm; <5
mm). The minimum clay content at which plastic (cohesive) behaviour occurs, is about 5% to 10%.
Thus, a sand-mud mixture will form a coherent cohesive mass, if there is a sufficient quantity of clay material
present in the mixture. The minimum amount of clay is about 5% to 10% to generate cohesive properties of the
mixture. Hence, the (abrupt) transition from non-cohesive to cohesive behaviour of a mixed sand/mud bed
strongly depends on the critical clay content (of about 5 to 10%). The cohesive strength of a natural bed
increases with increasing clay content and decreasing water content.
Sediment samples consisting of mixtures of clay, silt and sand can be roughly described as: mud, sandy mud,
silty mud or clayey (fat) mud, depending on the percentages of sand, silt, clay and organic material (Table 1).
Summarizing, the distinction between non-cohesive mixtures and cohesive mixtures can be related to a critical
mud (=clay+silt) content (pmud,cr). Most important is the value of the clay-fraction (sediments<0.008 mm; <8 mm)
in the mixture. Cohesive properties become dominant when the clay-fraction is larger than about 0.1 (10%).
Assuming a clay-mud ratio of 1/2 to 1/4 for natural mud beds, the critical mud content will be about pmud,cr=0.2
to 0.4 (mud is defined as clay+silt fraction).
If the mud content is below the critical value (pmud<pmud,cr), the sand-mud mixture is herein assumed to be
homogeneous with depth and to have non-cohesive properties. Furthermore, the erosion of the sand particles
is the dominant erosion mechanism. The mud particles will be washed out together with the sand particles.
Field observations have shown a significant increase of the mud concentrations near the bed during and after
periods with large waves. Flume experiments with waves over a consolidated mud bed have shown that large
waves can easily fluidize the top layer of the mud bed and generate a thin fluid mud layer with
concentrations larger than 100 kg/m. Fluidization of the top mud layer is initiated by wave-induced pressure
variations at the bed, which lead to an increase of the water pressure in the pores and hence to a reduction of
the internal soil shear strength. Various types of sediment mixtures in saline water were tested by Van Rijn
and Louisse (1987). In four experiments the bed consisted of pure (100%) kaolinite with different densities.

The dry density of the top layer (thickness of about 0.01 m) was less than about 200 kg/m3 in these tests. In
two other experiments the bed consisted of a mixture of fine sand (= 0.1 mm) and kaolinite. Each experiment
consisted of executing a sequence of tests with increasing wave heights. At each constant wave height, an
equilibrium concentration was established after a few hours. Equilibrium concentrations ranging from 2.5
kg/m near the bed to 0.5 kg/m near the surface were established after 200 minutes in conditions with a
relative wave height of H/h=0.2 (h=0.25 m). After this period, the wave generator was stopped and the
suspension was allowed to settle and consolidate for 43 hours resulting in a clear fluid with locally small
scour holes (depth = 0.02 m) on the bed. At a relative wave height of about 0.3 a fluid mud layer (= 0.025 m)
with concentrations of 100 kg/m, was generated. The concentrations above the fluid mud layer ranged from
20 kg/m near the bed to 4 kg/m near the surface. Dye injections showed that the fluid mud layer was slowly
(= 0.02 m/s) moving in the wave direction by wave induced velocities (drift). The generation of a fluid mud
layer has a stabilizing effect on the erosion process because velocity gradients and hence bed-shear stresses
remain relatively small.
Experiments with mixtures of fine sand (= 0.1 mm) and kaolinite showed that a consolidated bed of 75%
kaolinite and 25% fine sand had a similar erosional behaviour as a pure (100%) kaolinite bed. A bed consisting
of 25% kaolinite and 75% fine sand had a completely different behaviour. Fluidization of the top layer did not
occur. The kaolinite concentrations were not larger than about 0.3 kg/m, because only the top layer of the bed
was washed out. The sand concentrations were also quite small (factor 30 smaller than in case of 100% sand
bed) due to a strong suppression of sand ripples (height = 0.002 m).
Winterwerp et al. (1993) tested natural mud (70% clay and silt, 20% sand, 10% organics) in a wave flume
and found a clear influence of the dry density in the top layer of the bed on the fluidization process. It was
shown that a mud bed with a top layer having a dry density less than about 200 kg/m3can be easily fluidized
by wave action, resulting in a thin near-bed layer with high concentrations (200 to 400 kg/m3).

References
Bagnold, R.A., 1956. The Flow of Cohesionless Grains in Fluids. Proc. Royal Soc. Philos.Trans., London,
Vol. 249.
Bagnold, R.A., 1966. An Approach to the Sediment Transport Problem from General Physics. Geological
Survey Prof. Paper 422-I, Washington.
Einstein, A., 1906. Eine neue Bestimmung der Mlekl dimensionen. Annalen der Physiek, Leipzig (4),19.
Einstein, H.A., 1950. The Bed-Load Function for Sediment Transportation in Open Channel Flow.
Technical Bulletin No. 1026, U.S. Dep. of Agriculture, Washington, D.C.
Kim, S.C., Friedrichs, C.T., Maa, J.P.Y. and Wright, L.D., 2000. Estimating bottom stress in tidal
boundary layer from Acoustic Doppler Velocitymeter data. Journal of Hydraulic Engineering, ASCE,
Vol. 126, No. 6, p. 399-406
Krone, R.B., 1962. Flume Studies on the Transport of Sediment in Estuarine Shoaling Processes. Hydr.
Eng. Laboratory, Univ. of Berkeley, California, USA.
Mehta, A.J. and Partheniades, E., 1975. An Investigation of the Depositional Properties of Flocculated
Fine Sediment. Journal of Hydraulic Research, Vol. 13, No. 4, p. 361-381.
Mehta, A.J. and Lott, J.W., 1987. Sorting of fine sediment during deposition, p. 348-362. Coastal
Sediments, New Orleans, USA
Odd, N.V.M. and Cooper, A.J., 1989. A two-dimensional model for the movement of fluid mud in a high-
energy turbid estuary. Journal of Coastal Research, SI 5 , p. 185-193
Parchure, T.M. and Mehta, J.A., 1985. Erosion of Soft Cohesive Sediment Deposits. Journal of Hydraulic
Engineering, ASCE, Vol. 111, No. 10.
Partheniades, E., 1965. Erosion and Deposition of Cohesive Soils. Journal of the Hydraulic Division,
ASCE, Vol. 91, No. HY1.
Shields, A., 1936. Anwendung der hnlichkeitsmechanik und der Turbulenz Forschung auf die Geschiebe
bewegung. Mitteilungen der Preuss. Versuchsamst. fr Wasserbau und Schiffbau, Heft 276, Berlin,
Deutschland
Soulsby, R.L., 1997. Dynamics of marine sands. Thomas Telford
Van Rijn, L.C., 1990. Principles of fluid flow and surface waves in rivers, estuaries and coastal seas. Aqua
Publications, The Netherlands (WWW.AQUAPUBLICATIONS.NL)
Van Rijn, L.C., 1993. Principles of sediment transport in rivers, estuaries and coastal seas. Aqua
Publications, The Netherlands (WWW.AQUAPUBLICATIONS.NL)
Van Rijn, L.C., 1998. Principles of coastal morphology. Aqua Publications, The Netherlands
(WWW.AQUAPUBLICATIONS.NL)
Van Rijn, L.C., 2005. Principles of sedimentation and erosion engineering in rivers, estuaries and coastal
seas. Aqua Publications, The Netherlands (WWW.AQUAPUBLICATIONS.NL)
Van Rijn, L.C. and Louisse, C.J., 1987. The Effect of Waves on Cohesive Bed Surfaces. Proc. Coastal and
Port Eng. Conf., Beijing, China.
Winterwerp, H., 1999. On the dynamics of high-concentrated suspensions. Doctoral Thesis, Department of
Civil Engineering, Delft University of Technology, Delft, The Netherlands
Winterwerp, J.C., 2001. Stratification effects by cohesive and non cohesive sediment. Journal of
Geophysical Research, Vol. 106, No C10, p. 22,559-22,574
Winterwerp, J.C. et al., 1993. Experiments with Surface Waves on a Natural Bed. Report 40, Delft
Hydraulics, Delft, The Netherlands.

3. MEASURING PRINCIPLES, STATISTICS AND ERRORS
3.1 Measuring principles for suspended load transport
Samplers for suspended sediment transport were developed in the past according to two different principles:
the direct and the indirect measuring of the sediment transport. To explain this, the following expression for
the local suspended sediment transport is given:
Ss,z=<UsCs>=<(us+us/)(cs+cs/)> = <uscs> + <us/cs/>
with:
Ss,z = time-averaged suspended sediment transport at height z above the bed,
Us = instantaneous sediment particle velocity at height z,
Cs = instantaneous sediment concentration at height z,
us = time-averaged sediment particle velocity at height z,
cs = time-averaged sediment concentration at height z,
us/ = fluctuation of sediment particle velocity at height z,
cs/ = fluctuation of sediment concentration at height z.
3.1.1 Direct method
The direct method is based on the direct measurement of the time-averaged sediment transport (u c ) in a
certain point (point-integrating) or over a certain depth range (depth-integrating). This latter procedure
implies vertical movement at a uniform speed of the sampler over a certain depth range. Using a mechanical
sampler based on the collection of a water-sediment sample, this procedure is only applicable in shallow
streams. Examples of samplers based on the direct measuring principle are the Delft Bottle and the Acoustic
samplers (see Chapter 5).
3.1.2 Indirect method
The indirect method is based on the simultaneous but separate measurement of the time-averaged fluid
velocity and the time-averaged sediment concentration, which are multiplied to obtain the time-averaged
sediment transport. This method implies two assumptions which introduce errors: (1) the u/c/term is
assumed to be zero and (2) the fluid and sediment particle velocity are assumed to be equal. Some
information on the inaccuracy of the indirect method is given by Anderson (1975). Based on his results, the
sediment transport according to the direct method is about 10% smaller than that according to the indirect
method. Mulder et al. (1985) report a value of 5%, while Soulsby et al. (1985) report values of about 1%.
The time-averaged concentration can be measured in a single point (point-integrating) or over a
certain depth range (depth-integrating). Examples of samplers based on the indirect measuring principle
are the simple bottle and trap samplers, the USD-49, the USP-61, the pump-samplers and the optical
samplers (See Chapter 5).
References
Anderson, A.C., 1975. Comparison of Direct and Indirect Method of Measuring Suspended Sediment
Transport. St.Anthony Falls Hydr. Lab., Univ. of Minnesota, USA
Mulder, H.P.J., Van der Kolk, A.C. and Kohsiek, L.H.M., 1985. Measurements of Suspended
Sand Concentration and Velocity in the Eastern Scheldt Estuary, The Netherlands Euromech 192,
Munich, West-Germany
Soulsby, R.L., Salkfield, A.P., Haine, R.A. and Wainwright, B., 1985. Observations of the Turbulent
Fluxes of Suspended Sand near the Sea-bed Euromech 192, Munich, Germany
CHAPTER 3: MEASURING PRINCIPLES, STATISTICS AND ERRORS February 2006

3.2 Measuring principles for bed load transport
3.2.1 Direct method
The most widely used method for the sampling of bed load is the direct method by means of mechanical
trap-type samplers. Many versions of the trap-type sampler have been used with varying amount of success.
The problems of the trap-type sampler are the lowering and raising of the sampler to and from the streambed
and the efficiency of the sampler in collecting the particles. Trap-type samplers are described in Chapter 5.
3.2.2 Indirect method
The most widely-known indirect measuring methods are:

bed-form migration studies (tracking),
sediment deposition and erosion studies,
tracer studies.
Bed form tracking

Bed form tracking implies periodic depth measurements along longitudinal profiles (previously fixed). By
comparison of sequential profiles the migration velocity of the bed forms can be determined (see Par. 5.5.2).
Sediment deposition and erosion studies

Deposition and erosion studies imply periodic soundings of local bed levels (natural or dredged channel,
scour near structure, harbour basin). By comparison of sequential profiles the bed-load transport rate can be
obtained and related to the prevailing hydraulic conditions. An accurate (but costly) method to determine the
bed-load transport rate in a river is the dredging of a trench across the river. This method yields a reliable
time- and space-averaged value.
Tracer studies
For sediment transport measurements tracers can be used. Grains that are marked in such a way that their
transport characteristics are not changed, are added to the flow in small quantities and their displacements are
determined. These latter values can be converted to transport rates. Several types of tracers are used:
Fluorescent (luminofores): Marked grains can be detected after sampling under U.V. light. Different
types can be used simultaneously.
Radioactive: Natural sand is provided with a coating with radioactive material. Large quantities are
necessary to remain above the bavk-ground level of radioactivity.
Activation: Particles are marked and radiated after sampling.
Several methods for the interpretation are used:
Constant injection method.
A constant amount of tracer material (rate t) is distributed over the profile and injected during a long
time- interval. At a downstream cross section samples are collected and concentration as a function of
time is determined. After some time the concentration becomes constant = co. Then, the rate of transport
can be computed from the relation: s=t/co.
Point injection method.
At a certain time an amount of tracer material is injected. At several downstream locations concentration
is determined as a function of distance. From the diplacement of the centre of gravity of the
concentration-distance curves the average transport velocity can be computed. Multiplication with the
effective depth of transportation gives the rate of transport. The effective depth is of the order of half the
height of the bed forms and can also be determined by sampling in the bed.
Problems are the length of the measuring interval, the fact that the external conditions have to be constant
and the large number of observations. Some of these restrictions can be diminished by applying "dispersion
methods". The data are compared with a theoretical dispersion model.
A review of existing measuring techniques is given by Jansen et al.(1979).
References
Jansen, P. et al, 1979. Principles of River Engineering. Pitman, London

3.3 Measuring statistics
3.3.1 General aspects
A critical aspect of any morphological study is the field survey during which the samples to be analysed are
collected. It is important to be aware of the fact that the quality of the total study can only be as good as the
quality of the information gained through sampling. Thus, any errors incurred during sampling will manifest
themselves by limiting the accuracy of the study.
The objective of a field survey is to obtain samples from the project area with the purpose of characterizing
the area sampled. The sample size should be small enough to be conveniently handled and transported and
yet sufficient to meet the requirements of accuracy. The quality of the sampling process and analysis is
dependent upon:
selecting representative sampling sites in the project area,
collecting sufficient samples at each sampling site,
using appropriate sampling methods,
protecting the samples during the storage period (sample preservation),
flexibility of the sampling programme.
References
Environmental Protection Agency, Corps, of Eng., 1981. Procedures for Handling and Chemical Analysis
of Sediment and Water Samples. Env. Lab., U.S. Army Eng. Wat. Exp. Station Vicksburg, Mississippi, USA
3.3.2 Sampling site
The selected sites should be well-distributed over the project area and be representative for the (mean annual)
prevailing hydraulic and morphologic conditions. Some general requirements are:
located in a straight reach, i.e. over five channel widths upstream from the measuring location,
located in a stable cross-section (no erosion or deposition), uniform bed topography,
regular spatial velocity distributions (no converging or diverging streamlines, no vortices, backward
flow or dead water zones),
located normal to the main flow direction, uniform wave characteristics,
sufficiently deep with respect to the dimensions of the sampling equipment,
accessible and clear of natural and/or artificial obstacles (trees, bridges, piers),
well-defined geometrical dimensions (local depth, width, position).
In raising the question of representativeness, it is possible to define two populations: one population is the
actual population at the sampling point; the second population is the population of the collected samples.
Ideally, both populations should be the same. However, it is necessary to be aware of the fact that differences
may exist between these two populations because of bias in the sampling programme. Factors
that can contribute to the bias are over- and undersampling and equipment limitation.
Several criteria have been established to define the representative nature of a sampling programme:
well-defined project area,
well-distributed sampling sites (cross-sections).
The purpose of collecting representative samples is to define the range of characteristics that may be found in
the project area. The easiest method is to focus on the location of maximum characteristic values. However,
the results from these locations are not representative for the whole range in the project area. Additional
sampling should be conducted in areas of low hydraulic activity or energy because these areas favor the
settling of smaller-sized sediments (inside bend of channels, backwater areas, side channels and areas of
heavy shoaling or deposition). It may also be useful to select a reference station or control station at each
sampling site.
References
ISO-COMMITTEE, 1979. Velocity-Area Method. Ref. No. ISO 748-1979, The Netherlands

3.3.3 Number of measurements for suspended load transport
3.3.3.1 General aspects
The total load consists of bed material load and wash load. The bed-material load can be subdivided in bed
load and suspended load transport.
The wash load consists of sediment particles (fines < 62 mm) with particle sizes smaller than those found in
appreciable quantities in the bed material. The fine particles usually are uniformly distributed over the entire
cross-section. The sediment discharge can simply be obtained by multiplication of the flow discharge and the
concentration. Since the concentration is approximately constant over the corss-section, the number of
samples can be limited to a few samples. Usually, the samples are taken in reference points which have been
found to be representative for the entire cross-section. This should be checked regularly (dry and wet season;
low and high discharge).
The suspended sediment discharge (particles > 62 mm) usually is measured by taking a number of samples
over the depth and over the width of the river. The cross-section is divided into a number of subsections. The
sediment discharge passing through each subsection is determined by taking (point or depth-integrated)
samples along one vertical within each subsection. The accuracy of the suspended sediment discharge
depends on:
the number of points over the depth,
the number of verticals over the bed-form length in each subsection,
the number of verticals over the width (cross-section),
the number of verticals over time (flood period, ebb period).
Errors related to measurements can be classified as systematic errors and random errors.
A typical systematic error is the sediment discharge in the unmeasured zones below the lowest sampling
point (near the bed) and above the highest sampling point (near the water surface). Systematic errors
accumulate with increasing number of measurements. Random errors can be eliminated (averaged out) by
taking more measurements.
Factors influencing the accuracy are:
measuring method,
instruments available (calibrations),
natural fluctuations of concentration, velocity, transport,
calculation method (extrapolation unmeasured zones).
3.3.3.2 Number of points in a vertical
The depth-integrated suspended load transport per unit width (S) can be approximated by (see Section 4.1.3):
S=Sk(uici Dzi) + e (1)
with:
u = time-averaged velocity at height z (point i) above the bed,
c = time-averaged concentration at height z (point i) above the bed,
Dz = vertical increment (=zi-zi-1),
e = interpolation error,
k = number of points over the depth.
The interpolation error is related to the representation of a smooth curve by a limited number of linear
segments (e 0 for k ). The magnitude of this error depends on the number of points over the depth and
the distribution of these points over the depth. Since, the sediment concentrations are largest near the bed, the
interpolation error can be reduced by taking small increments (many points) near the bed.

Assuming that the variables u, c and Dz are independent stochastic variables with a normal Gaussian
distribution, the relative error of the suspended sediment transport can be estimated by:
(ss/S)2= (1/k)[(su/u)2 + (sc/c)2 + (sDz /Dz)2] + (erel)2 (2)
with s= standard deviation.
Relative standard deviations

Taking a time-averaging period of 2 to 3 minutes, the relative standard deviation of the local flow velocities
will be about su/u@0.1.
The relative standard deviation of the sediment concentration will be relatively large, especially close to bed
(see Par 5.2, Figure 2), say sc/c@0.3.
The relative error of the measuring elevation (height above bed) may also be quite large; mainly due to
vertical boat movements during the measuring period. Close to the bed, the relative error may be as large as
sDz /Dz @0.3.
Relative interpolation error

The relative interpolation error (erel = e/S) can be determined from flume and field measuring results.
The flume experiments of Barton and Lin (1955) have been used. During Run 20 (depth h = 0.13 m, mean
velocity u = 0.91 m/s, sediment size d50 = 180 mm) the flow velocities and sediment concentrations were
measured in 14 points distributed over the depth (between z = 0.01h and 0.9h). Based on these values, the
depth-integrated suspended sediment transport rate was computed. The sediment concentrations in the
unmeasured zone (near the bed) were represented by theoretical curves fitted through the lowest three
measuring points. The computed transport rate (based on 14 points) was adopted as the true transport rate.
Thus, the relative interpolation error is assumed to be zero for 14 points (erel = 0 for k = 14). These
computations were repeated for k = 11, 9, 7, 5 and 4 points, using two different schemes (A and B) for the
distribution of the points over the depth, as follows:
scheme A: z = 0.01h, 0.05h, 0.1h, ..., 0.9h

scheme B: z = 0.05h, 0.1h, 0.25h, ..., 0.9h
The transport rates were computed for each case (k = 11, 9, 7, 5 and 4 points) and compared to the true
transport rate (k = 14 points) to determine the relative error (erel), see Table 1.
Number of points Relative interpolation error Relative interpolation error

Scheme A Scheme B
k erel erel
14 0% 0%
11 0.5% -8.3%
9 1.1% -7.9%
7 1.2% -8.4%
5 4.0% -9.0%
4 11% -10%
Table 1
Relative interpolation error of depth-integrated transport rate (+=overestimation, - =underestimation)
A similar analysis for field conditions (waterdepth = 16 m) was performed by Rijkswaterstaat (1979, 1980).
The measuring elevations were in the range of z = 0.03h to z = 0.9h. The number of points was varied from k
= 10 to k = 2, using various distribution schemes. The true transport rate was assumed to be obtained for k =
10. In all, 89 cases (profiles) were considered. The results are presented in Table 2.

Number of points Mean relative interpolation error
k erel
10 0%
9 4.1%
8 5.2%
7 6.5%
6 6.3%
5 14.6%
4 15.1%
3 16.1%
2 76.1%
Table 2
Relative interpolation error of depth-integrated transport rate
Relative error of depth-integrated transport rate

The relative error of the depth-integrated transport rate can be determined from Eq. (2) with
su/u = 0.1, sc/c = 0.3, sDz /Dz = 0.3 and erel = 0.10-0.15 for k = 10 to 4. The results are presented in Table 3.
Number of points Relative error of depth-integrated

transport rate
k ss/S
10 17%
9 18%
8 19%
7 21%
6 22%
5 24%
4 27%
Table 3
Relative error of depth-integrated transport rate
These results are also shown in Figure 1. Based on these results the maximum relative error of the depth-
integrated suspended transport rate will be about 20% for 6 measuring points (k = 6). The following
distribution of the measuring points over the depth is proposed: z = 0.01h, 0.05h, 0.1h, 0.25h, 0.5h, 0.9h
above the bed.
Figure 1
Number of points in vertical

3.3.3.3 Number of verticals over bed-form length
When large-scale bed forms such as sand dunes (with length of about 7 h) are present, the error of the depth-
integrated transport rate will also depend on the location of the measuring station with respect to dune crest.
To reduce the error related to this effect, measurements should be performed in at least five verticals
distributed equally over the length of the bed form. Quantitative information of this error related to suspen-
ded sediment transport is not available.
3.3.3.4 Number of verticals in cross-section
The cross-section integrated suspended sediment transport rate T can be approximated by:
T=Sm(Si Dbi) + d (1)
with: Si = depth-integrated suspended transport rate in subsection i, Dbi= width of subsection i, m= number
of subsections in lateral direction, d= interpolation error.
The interpolation error depends on the number of verticals and also on the distribution of the verticals over
the cross-section. In this analysis equal distances between the verticals have been assumed. Other methods
are equal areas or equal discharges per subsection. This latter method is less practical because the lateral
velocity distribution must be known (a priori). When only a few verticals are selected in a wide irregular
cross-section, it is better to select representative locations (in the deepest channels where the velocities are
maximum). Assuming that the variables S and Db are independent stochastic variables with a normal
Gaussian distribution, the relative error of the suspended sediment transport (T) can be estimated by:
(sT/T)2= (1/m)[(sS/S)2 + (sDb /Db)2] + (drel)2 (2)
Relative standard deviation

According to the results presented in Par. 3.3.3.2, the relative error of the depth-integrated transport rate is
about sS/S=0.2 for a scheme with 6 points over the water depth.
The relative error of the width of each lateral subsection is estimated to be about sDb /Db= 0.05.
Interpolation error
An estimate of the relative interpolation error (drel = d/T) can be obtained by using a transport formula to
compute the suspended sediment transport rate at various verticals in a wide irregular cross-section
(consisting of 2 main channels with a depth of h= 4 m separated by a shallow part with a depth of 2 m; total
surface width of 500 m; d50 = 400 mm; u = 1 m/s). The total number of verticals is varied from m = 4 to 20
(equal distance between the verticals). The true transport is assumed to be obtained for m = 20. The relative
interpolation errors are given in Table 1.
Number of verticals Relative interpolation error of cross-section
integrated transport rate
m drel
20 0%
18 -1%
16 -2%
14 +5%
12 -5%
10 +11%
8 -10%
6 +17%
4 -20%
Table 1
Relative interpolation error of cross-section integrated transport rate
(+=overestimation, - =underestimation)

Relative error of cross-section-integrated transport rate
The relative error of the cross-section-integrated suspended sediment transport rate can be determined from
Eq. (2) with sS/S=0.2, sDb /Db=0 and drel =0.2 to 0.5 for m=4 to m=20. The results are presented in Table 2
and Figure 1A.
Number of verticals Relative error of cross-section integrated

transport rate
m sT/T
20 5%
18 7%
16 8%
14 8%
12 9%
10 12%
8 13%
6 17%
4 23%
Table 2
Relative error of cross-section integrated transport rate
Bonacci (1981) used a numerical simulation model to determine the relative error os/S in relation to the
number of verticals. The results are presented in Fig. 1A.
A method given by the U.S. Geological Survey (1988) is shown in Fig. 1B. The variability of the sand
concentration at different sampling verticals is assumed to be related to the variability of V2/D with V =
depth-mean velocity (ft/S) and D = water depth (ft).
In the example of Fig. 1B, the acceptable relative standard error is 15%, the bed-material is 100% sand, the
ratio V2/D = 2 resulting in a required number of verticals of seven. Table 2 and Fig. 1A also yield a number
of seven verticals for this example.
The total number of verticals to be selected should be such that the total required measuring period is
relatively small with respect to the flood period of the river discharge (discharge should be approx. constant).
In flash flood conditions this may give problems and hence a compromise must be made between accuracy
and measuring time available.
Kleinhans and Ten Brinke (2001) studied the accuracy of cross-schannel sampled sediment transport in
large sand-gravel-bed rivers during high discharges and found that the cross-channel integrated suspended
and bed load transport can be measured with an uncertainty of about 10% to 20%. The uncertainty mainly
depends on the balance between the number of subsections and measurement positions, and the rate of
change of hydrodynamic conditions. Sampling of 30 bed load samples and two suspended transport verticals
per subsection in five subsections is considered to be the minimum effort in a single channel. The presence
of bed forms is an important contribution to the final uncertainty. The temporal and spatial variation due to
moving bed forms are sufficiently included when at least 30 samples of bed load and two verticals of
suspended load are taken distributed over the length of the bed form (dune). Variations in hydrodynamic
conditions during a non-steady flood wave event cause systematic differences in transport between
subsections belonging to one cross-section, thus fouling the pure spatial variation. This can be corrected by
calculating a discharge-transport relationship and using this relationship to recalculate all transport such that
they all refer to the same discharge. The maximum acceptable correction of transport with this method is
estimated to be 25%. Therefore, the time to complete a cross-section must be kept as low as possible to
minimize the change in hydrodynamic conditions. Since, a minimum of samples is necessary for covering
the sptial variation in transport, a compromise has to be found between the number of subsections and the
rate of change of the discharge.

References
Barton, J.R. and Lin, P.N., 1955. A Study of the Sediment Transport in Alluvial Channels
Report No. 55JRBZ, Civ. Eng. Dep. of Colorado College, Fort Collins, USA
Bonacci, 0., 1981. Accuracy of Suspended Sediment Measurements in Natural Stream Flows
Journal of Hydraulic Research, Vol. 19, No. 3
Kleinhans, M.G. and Ten Brinke, W.B.M., 2001. Accuracy of cross-channel sampled sediment transport
in large sand-gravel-bed rivers. Journal of Hydraulic Engineering, Vol.127, No. 4, p. 258-269.
Rijkswaterstaat, 1979. Influence of Number of Points over the Depth on Computed Suspended Load
Transport (in Dutch). Report 44.001.02, District South West, The Netherlands
Rijkswaterstaat, 1980. Study on Accuracy of Suspended Load Measurements in a Tidal River
Report 72.002.01, District South West, The Netherlands
U.S. Geological Survey, 1988. Field Methods for Measurement of Fluvial Sediment
Report 86-531, Reston, Virginia, USA
Figure 1
Number of verticals in cross-section

3.3.3.5 Number of verticals per tide
The suspended sediment transport per unit width integrated over half a tidal period (flood tide or ebb tide)
can be computed as:
M=Sn(Si Dti) + g (1)
in which:
M = tide-integrated suspended sediment transport per unit width,
Si = depth-integrated suspended sediment transport per unit width at time i,
Dti = duration of time period i,
n = number of time intervals over half a tidal period (ebb or flood),
g = interpolation error.
The interpolation error depends on the number of time steps over the flood (or ebb) period and on the
distribution over the period. In this analysis equal time intervals are assumed.
Assuming that the variables S and Dt are independent stochastic variables with a normal Gaussian
distribution, the relative error of the tide-integrated suspended load per unit width can be estimated by:
(sM/M)2= (1/n)[(sS/S)2 + (sDt /Dt)2] + (grel)2 (2)
Relative standard deviations

According to the results presented in Par. 3.3.3.2, the relative error of the depth-integrated suspended
transport rate is about sS/S =0.2 for a scheme of 6 points.
The relative error of the time intervals is estimated to be about sDt /Dt= 0.05.
Interpolation error
An estimate of the relative interpolation error (grel= g/M) can be obtained by assuming a sinusoidal tidal
velocity curve v = vmax sin(wt) and a sediment transport formula s=a(v)4.
The tide-integrated transport was obtained by analytical integration and by numerical integration (trapezoidal
rule), giving a relative error:
grel= 8p2/(9n2) (3)
with; n=number of intervals.

The results are presented in Table 1. A time interval of 30 min yields a 6% underestimation of the transport.
Time interval Number of intervals during 6 Relative interpolation error

hour tidal period
Dt (min) n grel
10 37 -0.7%
20 19 -2.5%
30 12 -6.0%
45 8 -13%
60 6 -24%
Table 1
Relative interpolation error of tide-integrated transport rate (+=overestimation, - =underestimation)

Relative error of the tide-integrated transport rate
The relative error of the tide-integrated (flood or ebb) suspended transport can be determined from Eq. (2)
with sS/S =0.2, sDt /Dt = 0.05 and grel= 0.01 to 0.25 for n = 37 to n = 6. The results are presented in Table 2.
Time interval Number of intervals during 6 Relative error of tide-integrated

hour tidal period transport rate
Dt (min) n sM/M
10 37 3%
20 19 6%
30 12 9%
45 8 15%
60 6 25%
Table 2
Relative error of tide-integrated transport rate

3.3.4 Number of measurements for bed-load transport
The number of measurements of mechanical trap type bed-load samplers is presented in this Section.
Typical sampling problems related to the variability of the physical processes involved are (see Carey, 1985;
Delft Hydraulics, 1991) :
sampling duration of individual measurements (see Par. 5.5.1),
number of samples at each location,
number of sampling locations along the bed form length,
number of locations over the width of the cross-section.
3.3.4.2 Number of samples at each location and number of locations along bed form
The variability of the bed-load transport rate over the length of a bed form is so large (factor 10 to 100) that
the mean transport rate can only be determined accurately by taking a large number of samples. Different
locations equally distributed along the bed-form length should be selected and many samples should be taken
at each location. The mean bed-form length should be known a priori (echo soundings). Quite often the
sampling location is fixed because the sampler is operated from a bridge or a non-movable boat etc. In that
case accurate determination of the mean transport rate requires sequential sampling over a period long
enough for a (migrating) bed form to pass the sampling location.
Measurements in the Waal river in the Netherlands (Delft Hydraulics, 1991, 1992; Gaweesh and Van Rijn,
1994) have been used to determine the number of samples which should be taken along a bed form to obtain
a minimum error of the mean transport rate. Table 1 shows the variation coefficient in relation to the number
of samples. The variation coefficient is defined as:
VC =s/(mN0.5)
in which:
s = standard deviation of measured transport rate (over N samples),
m = mean value of measured transport rates (over N samples),
N = number of samples.
Results of Kleinhans and Ten Brinke (2001) are also shown in Table 1.
Number of samples Variation coefficient of mean Variation coefficient of mean

transport rate transport rate
N
VC VC
Gaweesh and Van Rijn (1994) Kleinhans and Ten Brinke (2001)
10 0.30 0.8
20 0.25 0.4
30 0.20 0.3
40 0.15 0.2
50 0.10 0.15
Table 1
Variation coefficient of mean transport rate
It is recommended to select at least 5 locations distributed equally along a bed form (dune) and to take 10
samples at each location yielding a total number of 50 samples. All measurements should be averaged to
obtain the mean transport rate. The mean transport based on these 50 samples will be within 10% of the true
mean transport rate (see Table 1). At each location a velocity profile should be measured to determine the
depth-averaged velocity (umean) , the bed-shear velocity (u*) and the bed roughness height (ks).

3.3.4.3 Number of sampling locations over width of cross-section
The number of sampling locations (equal distance) in the cross-section is primarily related to the
interpolation error, as discussed in Par. 3.3.3.4.
Accepting a relative interpolation error of 15%, at least seven locations in the cross-section are required. At
each location approximately 50 samples should be taken to obtain a relative error of 10% in the local
transport rate (see Table 1, Par. 3.3.4.2). Thus, about 7 x 50 = 350 samples are required to obtain an overall
relative error of (152 + 102)0.5 = 20% of the cross-section integrated bed-load transport rate.
The number of samples (including different boat positions, etc.) that can be collected per day is about 50
resulting in a measuring effort of 7 days to complete a cross-section. A significant reduction of the amount of
samples required can be obtained at the expense of less accurate results. For example, taking 3 locations
along a bed form with 5 samples per location and 7 locations in the cross-section will result in 3 x 5 x 7 =
105 samples which can be collected in 2 days. The relative inaccuracy of the local transport rate (based on 3
x 5 = 15 samples) will be about 30%, see Table 1, Par. 3.3.4.2. The overall relative error of the cross-section
integrated transport rate will be about (302 + 152)0.5 = 35%
References
Carey, P., 1985. Variability in Measured Bed-Load Transport Rate.Water Resources Bulletin, Vol. 21, No.
1, Paper no.84156
Delft Hydraulics, 1991. Delft Nile Sampler, Bed-Load Transport Measurements in the River Waal
Report Q1300, Part 1, Delft, The Netherlands
Delft Hydraulics, 1992. Bed-Load Transport Measurements in the Waal River near Woudrichem and
Druten. Report Q1300, Part 2 and 3, Delft, The Netherlands
Gaweesh, M.T.K. and Van Rijn, L.C., 1994. Bed load sampling in sand-bed rivers. Journal of Hydraulic
Engineering, Vol.120, No. 12, p. 1364-1384.
Kleinhans, M.G. and Ten Brinke, W.B.M., 2001. Accuracy of cross-channel sampled sediment transport
in large sand-gravel-bed rivers. Journal of Hydraulic Engineering, Vol.127, No. 4, p. 258-269.

3.3.5 Sampling frequency
The frequency of sampling depends on the frequency of the characteristic hydraulic conditions, the available
resources and the size of the project. Seasonal fluctuations are important. A sampling frequency of twice a
year will probably be sufficient for normal projects.
General rules are:
Rivers
during low run-off periods,
during and after high run-off periods,
Estuaries
daily variations,
during neap, mean and spring tides,
during summer and winter period,
during dry and wet period,
Coasts
during calm periods,
during and after storm periods.
3.3.6 Sampling methods
Usually, the sampling equipment is selected based on reliability, efficiency and accuracy criteria.
Information of these characteristics is summarized in Chapter 5 and 6. It must be stressed that the selection
of the sampling equipment may affect the apparent sample variability and thereby increase the number of
samples that is necessary or reduce the accuracy of the results. For example, the use of a grap sampler in an
area with a stratified bed material layer may introduce variability into the results that is a function of grab
penetration rather than being a property of the sample.
3.3.7 Sample preservation and in-situ sampling
The importance of sample preservation between the time of collection and the time of analysis must be
emphasized. Any change in sample composition can invalidate conclusions. Ideally, all samples should be
analysed directly after collection. However, this is usually not practical because of the number of samples
collected, the equipment and steps involved in the sample analysis. Therefore, some preservation method
may be necessary to retard biological action, hydrolysis and/or oxidation of chemical constituents. The
methods which can be used are pH-control, chemical addition, sample isolation and temperature control
(refrigeration and/or freezing).
Selection of a preservation method should be based on the purpose of the study and the characteristics to be
measured.
Sediment samples should always be sealed in airtight containers to preserve the anaerobic state of the sample
and maintain the solid-liquid phase equilibrium.
Suspended sediment particles in estuaries and coastal seas generally consist of solid and aggregated (flocs)
materials with densities as low as 1050 kg/m3. Particle surfaces may be coated with absorbed humuc
molecules.
In-situ measurements of sediment particles and flocs in these conditions are essential as natural flocs are
disrupted easily by physical manipulation such as sampling by bottles or pumps. True particle size
distributions of natural suspended sediments can only be achieved by in-situ systems. Most optical particle
size methods are potentially non-disruptive.

Eisma et al. (1991) have used an in-situ photocamera to measure size distributions of suspended sediment in
various West-European estuaries. In addition, they have also determined the size distributions of sediment
samples collected in bottles using the traditional pipette analysis method (sedimentation method) and the
Coulter Counter method. The bottle samples were either quickly brought back to the laboratory on land or
were analysed several hours later when the survey ship was at anchor with the engines off (as the pipette
method is sensitive to mechanical vibrations and temperature-induced circulations in the sample). The
analysis results reveal no relation between the in-situ size distributions (based on photocamera method) and
the size distributions from the bottle samples (pipette or Coulter Counter method). The maximum size of the
in-situ sediments was about 800 mm and the maximum size of the bottle sediments was a bout 125 mm. Both
the pipette and Coulter Counter analyis methods were performed on suspended samples that were sampled
and brought to the laboratory. During sample analysis in the laboratory the original flocs were disrupted so
that actually the size of the individual solids and/or floc fragments were measured. Both methods gave
similar but erroneous results. The results also depended on the way the samples were treated and stored
before analysis.
Phillips and Walling (1995) using a field-portable Laser-reflectance particle size analyser (PARTEC
200/300) have also shown that in-situ determination of particle size distributions of fluvial sediments is of
essential importance, either by making direct in-situ measurements in the water column or by taking bottle
samples and measuring the particle sizes directly after sampling. On-site measurements (immediately after
sample collection) of bottle samples were broadly similar to direct in-situ measured size distributions.
Analysis results of water-sediment samples collected in bottles and returned to the laboratory showed
significant differences in particle size distributions due to floccutation of sediments in the bottle samples,
even if the sediments in the bottle were artificially resuspended. In general the longer a sample was
allowed to settle the greater the increase in volume mean size upon resuspension. The bonding of flocs
formed during the settling period appears to become stronger with time.
References
Eisma, D. et al., 1991. Suspended matter particle size in some west-european estuaries; part I: particle size-
distribution. Netherlands Journal of Sea Research, Vol. 28, No. 3, p. 193-214
Phillips, J.M. and Walling, D.E., 1995. An assessment of the effects of sample collection, storage and
resuspension on the representativeness of measurements of the effective particle size distribution of
fluvial sediment. Water resources, Vol. 29, N0. 11, p. 2498-2508
3.3.8 Sampling flexibility
Any field survey program should be sufficiently flexible to allow changes based on field observations. The
presence of the responsible project engineer during the initial stage of the survey is of essential importance,
particularly when the variability of the main hydraulic parameters is unknown.
It may even be necessary to conduct a preliminary field survey (reconnaissance survey) to better define the
final sampling methods and programmes.

3.4 Measuring errors and required accuracy
Inevitably, errors are made in sediment transport measurements due to instrumental errors and sampling
errors. The errors can be systematic as well as random due to the stochastic nature of the variables to be
measured. An example of a systematic instrumental error is the inefficient trapping of fine sediment particles
(<100 mm) by the Delft Bottle.
A good example of a systematic sampling error is the error introduced by computing the time-averaged
suspended sediment transport (S) as the product of the time-averaged velocity (u) and concentration (c)
resulting in S=uc, while it should be computed as the time-averaged product of the instantaneous velocity (u)
and concentration (c), resulting in S = <u c>, (see also Par. 3.1).
To get a better understanding of the consequences of the errors of the measured sediment transport in relation
to the errors in the other hydraulic variables, the total load formula of Engelund-Hansen (1967) is
considered. This formula reads as:
S=a u2b(hi)1.5/(D2g0.5d50)
with:
S = total load transport (m3/s),
u = cross-section averaged velocity (m/s),
b = width (m),
h = water depth (m),
i = water surface slope (-),
D =(rs-rw)/rw= relative density (-),
g = acceleration of gravity (m/s2),
d50 = median particle size (m),
a =coefficient (=0.05).
Generally, the coefficient proposed by Engelund-Hansen (a=0.05) is not the optimum value for a specific
river. Improvement can be obtained by performing special calibration measurements (field survey).
Assuming that the basic variables are stochastically independent, the relative error of the a-coefficient can be
expressed as:
(sa/a)2=(sS/S)2 + 4(su/u)2 + (9/4)(sh/h)2 + (9/4)(si/i)2 + (sb/b)2 + (sd50/d50)2
with: s=standard deviation of each variable.
Typical relative errors of the velocity, water depth, width and particle size are: su/u=0.1, sh/h=0.05,
sb/b=0.05, sd50/d50=0.05. The relative error of the water surface slope may be rather large, especially in
isolated field conditions, say si/i =0.15 to 0.25.
Using these values, the relative errors of the a-coefficient have been computed. The results are shown in
Figure 1. Accepting a relative error of 40% in S, the relative error in a is about 60% which is not much
larger than the minimum value of about 45%. Thus, it is not very efficient to use a very sophisticated
measuring instrument with an instrumental error less than 10%, even when the number of samples is so large
that the random error introduced by the stochastic nature of the sediment transport process is small. Usually
however the number of samples to determine the sediment transport in the entire cross-section is so small,
that the stochastic sampling error becomes dominant. This is another reason for applying relative simple
measuring instruments. A disadvantage of the simple instruments, especially the trap-type instruments, is the
relatively expensive laboratory analysis of the samples. The more sophisticated optical and acoustic
instruments do not have this disadvantage and are, therefore, attractive although their use may not be
necessary for reasons of accuracy.

Not only the instrumental and sampling errors are of importance, but also the required accuracy of
morphological predictions. Ideally, one might first wish to determine what level of accuracy would be
acceptable in morphological predictions for a specific project and to select a sampling method which would
meet this requirement. Thus, the question of required accuracy cannot be answered in general, but
it should be studied for each project. It largely depends on the nature of the project and the applied prediction
methods, mathematical model or physical scale model. Interesting papers on this subject have been presented
by De Vries (1982, 1983).
References
Engelund, F. and Hansen, E., 1967. A Monograph on Sediment Transport in Alluvial Streams.
Teknische Forlag, Copenhagen, Denmark
Vries, M. de, 1982. A Sensitivity Analysis Applied to Morphological Computations.
Third Congress APD-IAHR, Bandung, Indonesia
Vries, M. de, 1983. On Morphological Forecasts for Rivers. Second Int. Symp. River Sedimentation,
Nanjing, China
Figure 1
Results of computed standard errors

4 COMPUTATION OF SEDIMENT TRANSPORT AND PRESENTATION OF RESULTS
4.1 Rivers
4.1.1 Total load transport per unit width
The total load transport can be determined as:
S= Sb + Ss (1)
with:
S = total load transport per unit width (kg/s/m),
Sb = bed load transport per unit width (kg/s/m),
Ss = suspended load transport per unit width (kg/s/m).
4.1.2 Bed load transport per unit width
The bed-load transport per unit width (kg/s/m) can be computed as given in Chapter 5.
4.1.3 Suspended load transport per unit width
When the suspended sediment samples are collected as point-integrated samples, there are two methods to
compute the depth-integrated suspended load transport (see Figure 1). First, there is the so-called partial
method which gives the suspended load transport between the bed and the highest sampling point using a
linear interpolation between adjacent (measured) values. Second, there is the so-called integral method,
which gives the total suspended load transport between the bed and the water surface by fitting a theoretical
distribution to the measured flow velocity and concentration profiles. Applying this latter method, the
suspended load in the unsampled zone is taken into account. The transport rate of the suspended silt (<50
mm) and suspended sand particles (>50 mm) should be computed separately. If necessary, more fractions can
be used.
4.1.3.1 Partial method
The suspend load transport per unit width (kg/s/sm) can be computed as:
Ss=[0.5u1c1z1 + Sn-1 0.5(uici + ui+1ci+1)(zi+1 - zi)] (2)
or
Ss=[0.5q1z1 + Sn-1 0.5(qi + qi+1)(zi+1 - zi)] (3)
with:
n = number of measuring points in the vertical,
ui = flow velocity in point i above the bed (m/s),
ci = concentration in point i above the bed (kg/m3),
zi = height above the bed of point i (m)
qi = suspended sediment transport (direct method) in point i above the bed (kg/sm2).
The first term in equations (2) and (3) is an estimate for the transport between the lowest sampling point and
the bed. The depth-averaged flow velocity (m/s) is:
umean=(1/zn)[0.5(u1z1) + Sn-1 0.5(ui + ui+1)(zi+1 - zi)] (4)
with: zn= height above bed of highest measuring point.
CHAPTER 4: COMPUTATION OF SEDIMENT TRANSPORT February 2006

4.1.3.2 Integral method
The velocities and concentrations in the unmeasured zone can be estimated, as follows.
Velocity
The velocities between the bed and the first measuring point (z1) can be represented by (see Figure 2A) :
v=v1(z/z1)0.25 for 0 < z < z1 (5)
in which:
v1 = fluid velocity in first measuring point above the bed,
z1 = height above bed of first measuring point.
The velocities between the last measuring point (zL) and the water surface can be taken equal to the velocity
(vL) in the last measuring point, see Figure 2A. Thus:
v = vL for zL < z < h (6)
Concentration
The sediment concentrations between the last measuring point and the water surface can be represented by a
linear function giving a zero concentration at the surface, as follows (see Figure 2A).
c = [(h-z)/(h-zL)]cL for zL < z < h (7)
in which:
cL = concentration in last measuring point,
zL = height above bed of last measuring point.
Since the exact distribution of the sediment concentrations in the near-bed zone is not known and considering
the relative importance of the concentrations in this zone, three different extrapolation methods are proposed
to represent the concentration profile between the bed and the first measuring point. Method 1 is supposed to
give a lower limit, while method 2 is supposed to give an upper limit.
Method 1
The sediment concentrations between the bed and the first measuring point (z =z1) are assumed to be equal to
the concentration (c1) in the first measuring point (see Figure 2A). Thus:
c = c1 for 0 < z < z1 (8)
Method 2
The sediment concentrations between the bed and the first measuring point are computed by (see Figure 2B):
c = A YB (9)
in which:
Y = (h-z)/z = dimensionless vertical coordinate,
z = vertical coordinate above bed,
h = water depth,
A,B = coefficients.

The A and B coefficients can be determined by a regression method applying the measured concentrations of
the first three measuring points above the bed, as follows:
select B=0.1,
compute A=S3(YkBck)/S3(YkB YkB), (10)
compute T=S3(AYkB-ck),
select B=0.2 (B is varied over range 0.1 to 5),
repeat procedure.
Finally, the A and B coefficients corresponding to a minimum T-value are selected as the "best" coefficients.
Applying Equation (9), the sediment concentrations are computed in 50 (equidistant) points between the bed
(defined at z = 2D50) and the first measuring point (z = z1). The maximum concentration is assumed to be
1600 kg/m3.
Method 3
The sediment concentrations between the bed and the first measuring point are represented by (Figure 2C):
c = eAZ + B (11)
in which:
z = height above bed
A, B = coefficients.
The A and B coefficients are determined by a linear regression method applying the measured concentrations
in the first three measuring points above the bed, as follows:
A= [3S3(zk ln ck) - S3(zk) S3(ln ck)] / [3S3(zkzk) (S3(zk))2] (12)
B= [S3(zk zk) S3(ln ck) - S3(zk) S3(zk ln ck)] / [3S3(zkzk) (S3(zk))2] (13)
Applying Equation (11), the sediment concentrations are computed in 50 (equidistant) points between the
bed (defined at z = 2D50) and the first measuring point (z = z1). The maximum concentration is assumed to be
1600 kg/m3.
Suspended load transport

Numerical computation of the depth-integrated suspended sediment transport (Ss) requires the specification
of velocities and concentrations at equal elevations above the bed (at equal z-values). When the z-values of
the velocities and concentrations are not corresponding, linear interpolation should be applied to obtain the
required data. The depth-integrated suspended sediment transport (Ss) is computed as:
Ss=SN [0.5(vici + vi-1ci-1)(zi - zi-1)] (14)
in which:
vi = fluid velocity at height z above the bed (m/s),
ci = sediment concentration at height z above the bed (kg/m3),
N = total number of points (including extrapolated and interpolated values).
If three different methods are applied to represent the sediment concentrations in the unmeasured zone near
the bed, three different values of the suspended sediment transport are obtained. These values can be
averaged to obtain a reliable estimate of the depth-integrated suspended load transport rate.
References
Van Rijn, L.C., 1991. Data base, Sand Concentration Profiles and Sand Transport for Currents
and/or Waves. Report H1148, Delft Hydraulics, Delft, The Netherlands

Figure 1
Suspended load transport per unit width

Figure 2
Suspended load transport per unit width

4.1.4 Total load transport per cross-section
The total load transport (kg/s) in a cross-section can be computed as (see Figure 1):
S=Snk(Sb,k + Ss,k)bk (15)
in which:
Sb,k = bed-load transport per unit width in section k (kg/sm),
Ss,k = suspended load transport per unit width (kg/sm),
bk = width of section k (m),
nk = number of sections in lateral direction.
The flow discharge (m3/s) in the cross-section can be computed as:
Q=Snk(umean,k hk bk) (16)
in which:
umean,k = depth-averaged flow veloccity (m/s) in section k according to partial or integral method,
hk = mean depth in section k (m).
The cross-section averaged flow velocity (m/s) is:
ucross=Q/ Snk( hk bk) (17)
The discharge-weighted concentration (mg/l) is:
cmean=(S/Q)10-3 (18)
Figure 1
Total load transport per cross-section

4.1.5 Presentation of results
Figures 1 and 2 show examples of data presentation.
References
GUY, H.P. and NORMAN, V.W., 1970. Field Methods for Measurement of Fluvial Sediment
Techniques of Water Resources Investigations of the U.S. Geological Survey, Chapter C 2. U.S.
Government Printing Office, Washington, USA
Figure 1
The flow discharge, sediment discharge and suspended sediment concentration as a function of time (daily,
monthly)

Figure 2
The sediment concentration and water surface level as a function of time (hourly)

4.2 Estuaries
4.2.1 Tide-integrated total load transport
In case of tidal flow conditions the sediment transport is a function of time. This implies the simultaneous
measurement of velocity and concentration profiles as a function of time. The time period needed to
complete the measurement of each velocity and concentration profile should be as short as possible (< 30
min).
The total amount of water (V) and sediment (M) passing a cross-section are (see also Figure 1):
Vflood = nfS[0.5Q1(t1-to) + 0.5(Qi+1 + Qi)(ti+1 - ti) + 0.5Qn(tM-tn)] (19)
Vebb = neS[0.5Q1(t1-tM) + 0.5(Qi+1 + Qi)(ti+1 - ti) + 0.5Qn(tL-tn)] (20)
VR =Vflood - Vebb (21)
in which:
VR = resulting water volume (m3),
Qi = discharge of water (m3/s) passing a cross-section at time t (according to Equation 16, Par. 4.1.4).
Mflood = nfS[0.5S1(t1-to) + 0.5(Si+1 + Si)(ti+1 - ti) + 0.5Sn(tM-tn)] (22)
Mebb = neS[0.5S1(t1-tM) + 0.5(Si+1 + Si)(ti+1 - ti) + 0.5Sn(tL-tn)] (23)
MR =Mflood - Mebb (24)
in which:
MR = resulting sediment mass (kg),
Si = total load transport (kg/s) passing a cross-section at time t (according to Equation 15, Par. 4.1.4),
nf = number of measurements during flood,
ne = number of measurements during ebb.
References
Van Rijn, L.C. van, 1979. SUDATA; Computer Program for Sediment Transport Measurements (in
Dutch). Delft Hydraulics Laboratory, Report R1267-1, The Netherlands
Figure 1
Tide-integrated total load

4.2.2 Presentation of results
Figure 1
The suspended sand concentration, flow velocity, and transport as a function of height above the bed (partial
and integral method, see Paragraph 4.1.3)
Figure 2
The suspended sand concentration and flow velocity as a function of height above the bed and time (tidal
cycle)
Figure 3
The depth-averaged flow velocity and the depth-integrated suspended sand transport as a function of time
(tidal cycle) according to partial and integral method
Figure 4
The local depth, depth-averaged flow velocity, concentration and transport as a function of time
Figure 5
The sediment concentration as a function of height above the bed and time
Figure 6
The sediment concentration as a function of height above the bed, longitudinal distance and time.
Figure 7
The sediment concentration as a function of longitudinal distance, height above the bed at various times
Figure 8
A plan view of suspended silt concentration in an estuary
References
Delft Hydraulics, 1980. Report R1267-IV (in Dutch)

Delft Hydraulics, 1982. Report R1586(in Dutch)
Delft Hydraulics, 1981. Report R1543
NEDECO, 1982. Access Channel Study, Beira Port. The Hague, The Netherlands
Parker, W.R. and Kirby, R., 1981. The Behaviour of Cohesive Sediment in the Inner Bristol
Channel and Severn Estuary in Relation to Construction of the Severn Barrage Report No. 117,
Institute of Oceanographic Sciences, Taunton, England

Figure 1
Sand concentration, flow velocity and transport

Figure 2
Sand concentrations and flow velocities as a function of height and time

Figure 3
Depth-averaged flow velocity and depth-integrated sand transport as a function of time
Figure 4
Depth-integrated transport rate as a function of time

Figure 5
Sediment concentration as a function of depth

Figure 6
Sediment concentration as a function of space

Figure 6.1
Salinity profiles from Khor Abdullah to Zubair Harbour

Figure 7
Sediment concentration as a function of space

Figure 8
Plan view of distribution of suspended silt concentration

5.1 General aspects
In this chapter various instruments for measuring the sediment transport rate are described. Usually the
sediment transport is represented as the summation of the bed load and suspended load transport. Figure 1 of
Par. 2.1 can be used to get an estimate of the relative importance of both modes of transport.
To measure the bed load transport, two measuring methods are available. There are the simple mechanical
trap-type samplers, that collect the sediment particles transported close to the bed. Another possibility is the
recording of the bed profile as a function of time (bed-formtracking).
To measure the suspended load transport, a wide range of instruments is available from simple mechanical
samplers to sophisticated optical and acoustical samplers. Most samplers are used as point-
integrating samplers which means the measurement of the relevant parameters in a specific point
above the bed as a function of time. Some instruments are used as depth-integrating samplers, which means
continuous sampling over the water depth by lowering and raising the instrument at a constant transit rate.
All instruments are described in terms of their measuring principle, ractical operation, inaccuracy and
technical specifications.
To get a better understanding of the accuracy of the various instruments, special attention is given to
comparative measurements (see paragraph 5.4). The instruments are described in Par. 5.5.
5.2 Instrument characteristics

In this section the most important characteristics of the point-integrating suspended load samplers are
summarized. The results are presented in Figure 1. Hereafter, some of the characteristics are discussed.
Sampling period
The sampling period is the time period during which a sample is collected. For the bottle-type and trap-type
samplers the sampling period is equal to the filling period of the bottle or trap. The sampling period of the
Delft Bottle is restricted by the size of the sediment catch which should be small compared with the size of
the sedimentation chamber of the instrument. The sampling period of the pump-filter sampler is restricted by
the filter characteristics. A small 50 mm-filter may be blocked rather easily, especially in a silty environment.
The sampling period of the optical and acoustical samplers is free. For time-averaging additional equipment
should be used allowing a digital reading.
Minimum cycle period

The minimum cycle period is the minimum time period between two successive measurements at adjacent
points in a vertical. In case of a free sampling period a minimum sampling period of about 5 minutes is used
to obtain the minimum cycle period, as given in Figure 1. The cycle period can be used to evaluate the time
period needed to cover a full concentration profile measurement. In case of tidal flow conditions this latter
period should be small in relation to the tidal period.
Overall accuracy
The overall accuracy is an estimate of the overall error of a single measurement due to (systematic and
random) measuring errors and stochastic (fluctuation) errors related to the physical process. The latter errors
are introduced by the stochastic fluctuations of the physical parameters to be measured. Figure 2 presents
some information of sediment concentration fluctuations, as observed by Eyster and Mahmood in some
Pakistan irrigation channels. They used a pint-bottle sampler yielding a time-averaged value over a five-
second period. Based on these results, each single bottle-measurement may have an error of about 100%.
This error can be reduced by using a larger sampling period or by collecting more samples. The overall
accuracy of the samplers with a free sampling period, as presented in Figure 1, is based on a relatively long
sampling period (say 5 minutes). The accuracy of the optical and acoustical samplers is largely dependent on
the number and accuracy of the calibration samples. More information of the inaccuracies involved can be
obtained from comparative measurements (see Paragraph 5.4).
References
Eyster, C.L. and Mahmood, K., 1976. Variation of Suspended Sediment in Sand Bed Channels
River Sedimentation, Vol. II, Fort Collins, USA
Suspended Suspended load Measured Measuring Response Sampling Minimum Overall accuracy
load Silt Sand Parameter Range Time Period Cycle Silt Sand
samplers (<50 mm) (>50 period
mm)
(mg/l) (s) (min) (min)
Mechanical Instruments
Bottle yes yes concentration >1 - 1 5 100% 100%
Trap yes yes concentration >1 instantaneous instan. 5 100% 100%
USP-61 yes yes concentration >1 - 1 5 100% 100%
transport
Delft Bottle no yes transport >10 - 5-30 10 - 50%
Pump-Filter no yes concentration >10 - 5-15 10 - 20%
Pump- no yes concentration >50 - 5-15 15 - 20%
Sedimentation
Pump Bottle yes yes concentration >1 - 1-5 5 20% 20%
Electronic Instruments
Optical-OBS yes no concentration 10-100000 <1 free 5 50% -
Optical- yes yes concentration 10-500 <1 free 5 30% 30%
LISST particle size
Pump- OBS yes no concentration 10-100000 - free 5 50% -
Acoustic- no yes concentration 10-10000 <1 free 5 - 50%
point ASTM transport
profiling ABS
Figure 1
Instrument characteristics for Sediment Concentration and Transport
Figure 2
Sand concentration as a function of time
Instrument characteristics (after Eyster and Mahmood, 1976)
5.3 Selection of sediment transport samplers
5.3.1 Guidelines for selection of sediment transport samplers
The selection of the most appropriate sampling technique should be based on a number of criteria, as
follows:
1. type of process/parameters to be measured:
- bed material;
- bed load;
- suspended load;
- particle size of bed material and suspended sediments;
- particle size of flocculated suspended sediments;
- fall velocity of sediments;
2. type of sampling environment:
- rivers, estuaries, coastal seas;
- sediments involved: mud, silt, sand, gravel, mixtures; flocculated materials;
- depth range and velocity range involved;
3. type of sampling:
- on-line measurements (short term) using hand-held instruments in shallow streams or instruments
attached to a survey ship (through a cable-winch system) or structure (bridge, platform) in the flow,
- stand-alone measurements (long-term) using instruments attached to frames, tripods or other structures
in the flow,
- point measurements using sensors that can measure in one point only (multiple points to determine
vertical distribution);
- profiling (or depth-integrated) measurements using instruments traversing (part of) the flow depth or
instrument signals (acoustics or optics) traversing the flow depth or part of the flow depth;
4. type of project and required accuracy:
- reconnaissance study to get a rough estimate of parameters/proccesses involved;
- process study to understand the details of the physical system (requiring turbulent fluxes);
- studies focussing on high spatial and time resolution (dredging and dumping plumes);
- input for mathematical modelling (boundary conditions) for design purposes;
- data for verification of models;
5. available instruments and available budget.
To select the most appropriate sampling instrument, quantitative information of the physical parameters to be
measured should be available prior to the actual field survey. It is important to have information of the
various transport modes at the sampling site such as the (relative) value of the wash and bed material load,
the bed load and the suspended load transport rates, the sediment concentrations (low or high) and the
particle size ranges of the suspended sediments (clay, silt, sand; flocculated sediments) involved. To get this
information, existing data sets should be analyzed or a reconnaissance study should be carried out.
Most measurements in rivers and estuaries are based on on-line data sampling using instruments attached to a
survey ship (through cable-winch system). Stand-alone measurements using a package of electronic
instruments attached to a frame, tripod or other structure placed on the bed or in the flow are conducted when
survey ships cannot be used due to the presence of surface waves (coastal envirionments) or when long-term
measurements (weeks to months) are required. The instruments usually are operated in burst mode, which
means that measurements of limited duration are taken at regular intervals in time (15 min per hour) to
reduce on data storage.
Examples of instrumented stand-alone measuring facilities are:
HSM-tripod of University of Utrecht (Grasmeijer et al., 2005);
AMF-tripod of Rijkswaterstaat (RIZA, 2001);
STABLE tripod of Proudmann Oceanographic Laboratory POL (Williams et al., 2003).
The availability of these advanced facilities with electronic equipment greatly improves data quality and data
density over the traditional mechanical equipment, and long-term costs will be reduced, as there is no longer
need for intensive sample processing and analysis in the laboratory.
Another important criterion is the purpose of the study and the accuracy required. For example, the predicted
deposition volumes of a planned harbour approach channel should be rather accurate when the maintenance
dredging costs of the channel are critical with respect to the economic feasibility of the project. In that case
the most accurate instruments and the most sophisticated facilities should be used. Although these type of
measurements are expensive, the overall costs of field surveys are only a fraction of the total project costs.
When research (process) studies are performed, it is often required to measure the time-series of the
fluctuating parameters so that the turbulent fluxes can be determined. For engineering studies it is usually
sufficient to measure the time-averaged velocities and sediment concentrations. Simple instruments such as
the bottle and trap samplers can then be used, although the analysis costs of the many samples involved are
relatively high.
5.3.2 Sediment transport measurements in rivers
The available instruments (and accuracies involved) for measuring:

bed load transport,
suspended sediment concentrations and transport rates,
particle sizes and fall velocities,
are presented in Tables 1,2 and 3 below; order of preference is based on the overall sampling accuracy.
Simple mechanical instruments such as the bottle-type, the trap-type and the pump-type samplers are still
very attractive because of their robustness and easy handling, particularly when used at isolated field sites.
The accuracy of the measured parameters involved can increased by increasing the number of samples
collected. Analysis costs of all samples involved may be critical with respect to the available budget. Optical
and acoustic instruments are attractive when large numbers of data have to be collected. As calibration is
involved, the accuracy strongly depends on the quality/reliability of the calibration curves. Hence, many
calibration samples are required using a pump sampler with the nozzle as close as possible to the
optical/acoustic sensor.
A major technological advance for measuring suspended load transport is the in-situ Laser diffraction
instrument (LISST). This instrument can measure the particle size distribution and sediment concentration
simultaneously. An attractive solution is the LISST-ST, which includes a streamlined body to improve
sampling accuracy and is equipped with a pressure sensor and current meter for measuring the sampling
height above the bed and the ambient velocity. The velocity data are used to control pump sampling
(isokinetic sampling) across the internal Laser arrangement. The velocity and concentration data are used to
compute fluxes for up to 32 particle size classes at points, verticals, or in the entire stream cross-section. All
data are transmitted via a cable to the survey vessel (on-line measurement). Limitations (related to light
penetration) are the maximum concentration ranges of about 150 mg/l for fines (mud/silt) and 500 mg/l for
sand particles. Hence, the instrument cannot be used in high-concentration conditions (close to bed; upper
flow regime).
Selection guidelines for sampling instruments in rivers focussing on the typical mechanical US-instruments
(see Figure 1) are given by the U.S. Geological Survey (Davis, 2005). This report and many other reports of
various typical US-samplers can be downloaded from the FISP-site (http://fisp.wes.army.mil). These US-
samplers are manufactured by Rickly Hydrological Company (www.rickly.com).
The manuals: Field methods for measurement of fluvial sediment (Edwards and Glysson, 1999) and Fluvial
sediment concepts (Guy) can be downloaded from USGS-site (www.usgs.gov).
Bed load transport
Type of sediment Type of method Type of sampling Accuracy

Silt Bed form tracking On-line factor 2
Sand/Gravel Bed-form tracking On-line 50% (many tracks)
BTMA trap sampler On-line factor 2 (many samples required)
Helley-Smith trap On-line factor 2 (many samples required)
sampler
Delft Nile trap sampler On-line factor 2 (many samples required)
Table 1
Characteristics of bed-load sampling methods in rivers
Suspended sediment concentrations and transport
Type of sediment Type of Method Type of sampling Accuracy

Mud/Silt Pump-Bottle On-line 20% (bottle filled by pumping during 15
minutes on board of survey ship)
USP-61 Bottle On-line 20%-30% (many samples required); not
(hand-held in depth<1m) close to bed
Bottle/Trap On-line 20%-30% (many samples required);
(hand-held in depth<1m) not close to bed
Laser (LISST) On-line 20%-30% (concentration< 150 mg/l)
Stand-alone
Optical (OBS) On-line 30%-50% (depending on number of
Stand-alone calibration samples)
Pump-Optical On-line 30%-50% (depending on number of
(OBS) calibration samples; bottles filled on
board of survey ship; sensor in bottle)
Sand Pump-Bottle On-line 20% (bottle filled by pumping during 15
Pump-Filter On-line 20% (undersampling of very fine
particles <filter size)
close to bed
not close to bed
Acoustic On-line 20%-30% (depending on number of
(ASTM, ABS) Stand-alone calibration samples)
Stand-alone
Delft Bottle On-line 50% (many samples required;
undersampling of finer particles); not
close to bed
Table 2
Characteristics of suspended load sampling methods in rivers
In-situ particle size and fall velocity

Mud/Silt Mechanical in-situ On-line 50% (intrusive for flocs; probably
settling tubes destroyed during sampling procedure)
(settling velocity) Bottom withdrawal tubes
Side withdrawal tubes
In-situ videocamera On-line 50% (undersampling of very fine
(size and settling sediments; lower limit of about 20 mm)
velocity; mainly VIS, INSSEV
flocs)
In-situ Laser- On-line 20%-30% (lower limit of about 5 mm and
diffraction Stand-alone upper limit of about 500 mm)
(particle size incl. LISST-100, LISST-25,LISST-SL
flocs)
In-sity Laser- On-line 20%-30% (lower limit of about 5 mm and
diffraction + Stand-alone upper limit of about 500 mm )
Settling-tube LISST-ST
(size and settling
velocity, incl. flocs)
reflectance Stand-alone upper limit of about 500 mm )
(particle size; not PARTEC
for flocs)
Sand In-situ Laser- On-line 20%-30% (lower limit of about 5 mm and
(particle size) LISST-100, LISST-25, LISST-SL
(particle size) PARTEC
Table 3
Characteristics of sampling methods for particle size and fall velocity in rivers
5.3.3 Sediment transport measurements in estuaries

bed load transport,
are presented in Tables 1, 2 and 3 below; order of preference is based on the overall sampling accuracy.
Simple mechanical instruments such as the bottle-type and the trap-type samplers are not attractive because
of the very short sampling times involved. Accuracy cannot be improved by increasing number of samples
due to time-variation of sediment concentrations within the tidal cycle.
Point-samples should be taken over the entire water column in strong tidal flows as the sediments will be
mixed over the water column by turbulent eddies. Data sampling can be confined to the bottom region in
weak tidal flows. Flocculation often is a dominant process in muddy estuaries. The LISST-ST which is an
in-situ Laser diffraction instrument in combination with a settling tube offers a powerful solution to measure
particle sizes, concentrations and densities of the individual particles as well as the flocculated aggregates
(low concentrations <150 mg/l) in on-line mode. This instrument (LISST-ST) is not yet suitable for long-
term, stand-alone measurements due to insufficient robustness, relatively low concentration range (<150
mg/l) and biological fouling problems.
Bed load transport

Silt Bed-form tracking On-line factor 2 to 3 due to limited number
of tracks
Sand/Gravel Bed-form tracking On-line factor 2 to 3 due to limited number
of tracks
Delft Nile trap sampler On-line factor 2 to 3 (limited number of
samples)
Table 1 Characteristics of bed-load sampling methods in estuaries

Mud/Silt Pump-Bottle On-line 20%-30% (bottle filled by pumping
during 15 minutes on board of survey
ship)
Stand-alone
Optical (OBS) On-line 50% (depending on number of calibration
Stand-alone samples)
Bottles/Traps On-line factor 2 (limited number of samples ); not
close to bed
Stand-alone
Bottles/Traps On-line factor 2 to 3 (limited number of samples);
not close to bed
Table 2 Characteristics of suspended load sampling methods in estuaries

flocs)
flocs)
In-sity Laser- On-line 20%-30% (lower limit of about 5 mm and
diffraction + Stand-alone upper limit of about 500 mm )
(size and settling
for flocs)
Table 3
Characteristics of sampling methods for particle size and fall velocity in estuaries
5.3.4 Sediment transport measurements in coastal seas

bed load transport,
Field instruments to measure bed-load transport in oscillatory flow are hardly available. Katori (1983) using
a stand-alone mechanical trap-type sampler, performed experiments to measure the cross-shore sediment
transport rate at and near the bed on a sandy beach at the Pacific Ocean coast of Japan (shallow surf zone).
Instruments available for measuring suspended sediment concentrations and transport in coastal
environments are: mechanical traps (streamer traps in shallow surf zone <1 m), pump samplers, optical
samplers and acoustic samplers.
Mechanical trap samplers have been used extensively in Japan, in Poland (Antsyferov et al, 1990), in the
USA (Kraus, 1987) and in Russia (Antsyferov et al, 1990). These types of instruments, which can only be
used in the shallow surf zone (depths <1 m) by operators wading in the water, are simple and reliable.
Simultaneous measurements in many points and stations can be performed. Kraus (1987) used portable
traps, consisting of long rectangular bags of polyester sieve cloth mounted on a vertical rack to measure
longshore suspended load transport rates. An operator standing downcurrent attends the trap during the
sampling interval (5 to 10 min), The use of the trap is restricted to surf zones with wave heights less than
about 0.5 m in water depths upto l m. Antsyferov et al (1990) used traps mounted on tripods placed in and
outside the surf zone. The traps were replaced by divers. Typical sampling periods are ranging from 10 min
to 1 day. The sediment mass collected by each trap gives an indication of the local time-averaged
concentration. Determination of the trapping coefficient is problematic. Therefore, relative concentrations
rather than absolute concentrations are obtained. Kana (1979) used an instantaneous trap sampler in the surf
zone. Many samples at the same location are required to eliminate the random fluctuations.
Pump samplers were used by many researchers to measure time-averaged sediment concentrations. These
types of samplers can only be used from a pier or platform. The intake nozzles should be directed
downwards.
Optical and acoustic probes are available to measure instantaneous sediment concentrations from a pier or
platform or from a stand-alone tripod. Data transmission can take place by telemetry or on-line to a
computer or datalogger. Calibration of the probes is required. Optical probes cannot be used in conditions
with both sand and silt particles in suspension. The optical instruments are relatively sensitive to fine mud
particles. Hence, the mud background concentration must be small (<50 mg/1). Otherwise, the sand
concentrations cannot be measured accurately.
Acoustic probes cannot be used in plunging breaking wave conditions due to the presence of air bubbles.
Nuclear probes which have been used in Russia and in China, cannot be used in low-energy conditions
where the concentrations are relatively small. The threshold concentration is of the order of 500 mg/1.
Suspended sediment transport measurements in conditions with combined current and wave conditions
cannot be performed from moored or sailing survey ships. Two options are possible (see Figure 1):
1. on-line sampling from piers connected to shore, platforms resting on seabed or sledges/trailers towed by
vehicles (only in shallow surf zone);
2. stand-alone sampling from frames/tripods/poles on/in the seabed or from drift bouys (profiling mode
from surface to bed) using a package of sophisticated electronic sensors (electromagnetic and acoustic
flowmeters, optical and acoustic backscattering sediment concentration meters); see Figures 3 and 4.
A typical instrument package attached to a stand-alone tripod in coastal environments is, as follows:
electromagnetic point velocitymeters (EMV), acoustic Doppler point velocitymeters (ADV) and acoustic
Doppler Current Profilers (ADCP);
optical backscatter point sensors (OBS) for silt/mud; acoustic backscatter point sensors (ASTM) and
acoustic backscatter profiler (ABS) for sand;
in-situ Laser diffraction sensors (LISST) for particle size;
acoustic bed profiler (line and sector scanning) for bed and ripple pattern detection.
A problem of all stand-alone facilities is the wave and current-induced scour near the legs of the structures
involved and the scour underneath and around the instrument sensors. Black and Rosenberg (1991) have
found by camera observations that a minimum distance of about 0.05 m should be kept between the probe
(sensor) and the bed in oscillatory flow conditions to prevent scour of bed material. The free span of the
instrument supports and legs should always be as large as possible to allow measurements between the legs
with minimum disturbance. A dramatic example of a poorly designed research pier is the Duck Pier (North
Carolina, USA) with a longshore pile spacing is 4.6 m and cross-shore spacing is 12.2 m, which has caused
considerable scour in the vicinity of the pier. An example of a slender pier with cross-shore pile spacings of
36 m is the Sally Wharf near Bordeaux, France (Figure 2).
Stand-alone tripods (Figures 2, 3, 4) are placed on the seabed from a vessel during calm conditions. They are
vulnerable in high wave conditions (overturning) or in areas with intensive fishing activities. Another
problem is the absence of control of the measuring elevations of the instruments in conditions with rapidly
changing bed levels (surf zone). The instruments can be damaged easily when placed close to the bed.
Williams et al. (2003) concluded that the turbulence levels underneath their tripod was slightly increased.
The bed-shear stresses were also slightly over-estimated (15%). Evidence of scour in the immediate vicinity
of the legs was detected, but was sufficiently localised not to effect the ripples beneath the main instrument
package on the tripod.
Piles/poles driven or washed into the bed offer a simple solution for hydrodynamic measurements of waves
and winds, but cannot be used for measurements of velocities and sediment concentrations due to the
influence of vortices and scour generated by the pile/pole construction.
Towed sledges or trailers with a mast above the water surface have been used for cross-shore bed level
soundings and wave height measurements in the surf zone during storms. However, they cannot be used for
measurements of velocities and sediment concentrations due to the disturbing influence of vortices generated
by the structure itself.
Reviews of field instrumentation for the coastal environment have been given by Terwindt et al. (1992),
Basinski (1989) and White (1998).
Bed load transport

Sand Bed-form tracking Sand-alone factor 2 (in deep water with non-breaking
of ripples using waves)
acoustic profilers factor 2 to 3 in surf zone
(side scan sonar)
Table 1
Characteristics of bed-load sampling methods in coastal seas

Mud/Silt Optical (OBS) On-line factor 2 to 3 (no in-situ calibration
Stand-alone samples available)
Sand Mechanical trap- On-line factor 2 to 3 (only usable in shallow surf
type sampler zone <1 m)
(streamer traps)
Pump sampler On-line 50% (only usable for longshore transport
from pier or platform)
Acoustic On-line factor 2 to 3 (no in-situ calibration
(ASTM, ABS) Stand-alone samples available)
Table 2
Characteristics of suspended load sampling methods in coastal seas

Mud/Silt In-situ videocamera On-line 50% (undersampling of very fine
velocity; mainly VIS, INSSEV (outside surf zone)
flocs)
(particle size incl. LISST-100, LISST-25
flocs) (outside surf zone)
(particle size; not PARTEC (outside surf zone)
for flocs)
(particle size) LISST-100, LISST-25 (outside surf zone)
(particle size) PARTEC (outside surf zone)
Table 3
Characteristics of sampling methods for particle size and fall velocity in coastal seas
References
Antsyferov, S.M., Belberov, Z.K. and Massel, S., 1990. Dynamical Processes in Coastal Regions
Publishing House, Bulgarian Academy of Sciences.
Basinski, T., 1989. Field studies on sand movement in the coastal zone. Polska Akademia Nank, Instytut
Budownictwa Wodnego, Gdansk, Poland
Black, K.P. and Rosenberg, M.A., 1991. Hydrodynamics and Sediment Dynamics in Wave-Driven
Environments Technical Report No. 13, Victorian Institute of Marine Science, Victoria, Australia.
Davis, B.E., 2005. A guide to the proper selection and use of federally approved sediment and water-quality
samplers. Report QQFederal Interagency Sedimentation Project (FISP), Waterways Experiment
Station, Vickburg, USA (http: //fisp.wes.army.mil)
Edwards, T.K. and Glysson, G.D., 1999. Field methods for measurements of fluvial sediment. Techniques
of water resources investigations of the US Geological Survey, Book 3, Applications of Hydraulics,
Chapter C2. USGS, Box 25286, Denver CO 80225, USA
Grasmeijer, B.T. et al., 2005. Description of measurements and database of field and laboratory data,
Paper M. In: Sandpit edited by Van Rijn et al., 2005. ISBN 90-800356-7-X (www.aquapublications.nl)
Guy, H.P., Fluvial sediment concepts. Techniques of water resources investigations of the US Geological
Survey, Book 3, Applications of Hydraulics, Chapter C1. USGS, Box 25286, Denver CO 80225, USA
Hemsley, J.M., McGehee, D.D. and Kucharski, W.M., 1991. Nearshore oceanographic measurements:
Hints on how to make them. Journal of Coastal Research, Vol. 7, No. 2
Kana, T.W., 1979. Suspended Sediment in Breaking Waves. Techn, Report No. 18-CRD, Coastal Res. Div.,
Dep. of Geology, Univ. of South Carolina, Colombia, USA
Katori, S., 1983. Measurement of Sediment Transport by Streamer Trap (In Japanese). Report No. 17, TR-
82-1, Nearshore Environment Research Center, Japan
Kraus, N.C., 1987. Application of Portable Traps for Obtaining Point Measurements of Sediment
Transport Rates in the Surf Zone. Journal of Coastal Research, Vol. 3, No. 2
Rijkswaterstaat/RIZA, 2001. Autonomous Measuring Facility for measurement of water and sediment
movements (in Dutch). Report RIZA 2001.073X, Lelystad, The Netherlands
Terwindt, J.H.J, Greenwood, B.J. and Short, A., 1992. Users report on instruments for coastal research.
Dep. of Physical Geography, University of Utrecht, Utrecht, The Netherlands
White, T.E., 1998. Status of measurement techniques for coastal sediment transport. Coastal Engineering,
35, p. 17-45
Williams, J.J. et al., 2003. Interactions between a benthic tripod and waves on a sandy bed. Continental
Shelf Research, Vol. 23, p. 355-375
Figure 1
Selection diagram for typical US-suspended samplers (Davis, 2005)
Figure 2
Instrument facilities for coastal environments
Figure 3
AMF tripod of Rijkswaterstaat/RIZA (2001), The Netherlands
Figure 4
HSM tripod of University of Utrecht (Grasmeijer et al., 2005), The Netherlands
5.4 Comparison of suspended load samplers
5.4.1 Comparison of USP-61, Delft Bottle and Pump-Filter sampler
During May 1979 a field investigation using the USP-61, the Delft Bottle and the Pump-Filter Sampler was
carried out in the Danube River near Ilok, Yugoslavia (Dijkman, 1982). At this location the measuring
section is straight, while the cross-section is rather wide and regular. The mean width was about 560 m and
the mean depth was about 8 m. The water temperature varied from 14-18C. The instruments are described
in Paragraph 5.6.
The three instruments were used at the same level above the bed within a few metres from each other. The
(vertical) positioning was done by means of an echo-sounder present on each instrument. Figure 1 shows the
results of a relative comparison of the instruments.
Figures 2 and 3 show typical results of the sediment transport per grain-size fraction measured by each
instrument. The results can be summarized, as follows:
USP-61
The "measured" sediment transport shows rather large fluctuations, even for the silt particles (< 70 mm). The
fluctuations are caused by the relatively small sampling time of about 30 seconds. For individual samples the
accuracy may be as large as 50%. To obtain a reliable average value in a statistical sense, at least 10
samples must be collected. In tidal flow conditions, where the concentrations vary as a function of time, the
USP-61 is not practical.
Delft Bottle
The measured values for each grain-size fraction were corrected using standard calibration curves, as given
in Paragraph 5.6.2.5. Although the sampling time was rather long (600-900 s), the fluctuations in the
measured transport rates are still rather large. Probably, these are caused by sediment losses during the
hoisting of the instrument.
Compared with the USP-61, the results of the Delft Bottle are systematically smaller, particularly for
the 70-150 mm grain-size fraction. For all measurements the average differences between the Delft Bottle
and the USP-61 Bottle varies from about 10%-100%, depending on the grain-size fraction and the nozzle
type (Figure 1).
Pump-Filter Sampler
The water-sediment mixture was pumped through 50 mm filters. Laboratory analysis showed that almost all
particles smaller than 70 mm were lost. Figures 2 and 3 show relatively small fluctuations, which means that
a sampling period of about 300 s is sufficiently large to obtain a reliable average value. The mean value of
the sand transport (> 70 um) measured by means of the Pump-Filter Sampler agrees rather well with the
USP-61 Bottle. For all measurements the average difference between the USP-61 and the Pump Filter
Sampler varies from 5%-15% (Figure 1).
References
Dijkman, J.P.M. and Milisic, V., 1982. Investigations on Suspended Sediment Samplers. Delft Hydraulics
Laboratory - Jaroslav Cerni Institute, Report S410, The Netherlands
Figure 1
Comparison of suspended sediment transport measured with Delft Bottle, USP-61 and Pump-Filter
___ ___ ___ Delft Bottle (small nozzle), measuring time=600-900 s

- - - - - - - - - USP-61, measuring time= 30 s
__________ Pump-Filter sampler, measuring time= 300 s
Location: Danube River near Ilok, Yugoslavia (16 May 1979)

Sampling height=0.75 m above bed; mean depth= 8m, mean velocity=0.9 m/s
Figure 2
___ . ___ . _ Delft Bottle (big nozzle), measuring time= 600-900 s
___ ___ ___ Delft Bottle (small nozzle), measuring time= 600-900 s

Figure 3
5.4.2 Comparison of Pump-Filter sampler and ASTM
The measurements were carried out in the Eastern Scheldt, a wide tidal estuary in the south-west part of the
Netherlands. The local water depth was about 10 m, the flow velocities ranged from 0.4 - 1.0 m/s. The local
bed material consisted of sand with D50 = 200 mm. The intake nozzle of the pump-filter sampler was installed
next to the sensor of the acoustic sensor ASTM (AZTM in Dutch). The instruments are described in
Paragraph 5.6. The flow velocities and sediment concentrations were measured at a level of 1 m above the
bed during flood tide.
The results, shown in Figure 1, show reasonably good agreement. The maximum deviation is about 30% for
small concentrations (< 50 mg/1).
References
Van Rijn, L.C., 1979. Pump-Filter Sampler. Delft Hydraulics Laboratory, Report S404 I, The Netherlands
Figure 1
Comparison of sand concentrations measured with Pump-Filter Sampler and acoustic AZTM Sampler
5.4.3 Comparison of Pump-Filter sampler and Pump-Bottle sampler
Sediment concentration measurements were carried out in the Oude Maas, a side branch of the Nieuwe
Waterweg (New Waterway) near Rotterdam, the Netherlands. The local flow depth was about 10 m, while
the tidal flow velocities varied from 0-1 m/s. The bed material consisted of sand particles with D = 200 mm.
The pump-filter sampler was operated with filters of 50 mm. The sampling period in each point was 5
minutes. Simultaneously a pump-bottle sampler was used to determine the sediment concentrations. This
method consisted of the filling of 1 liter-bottles by a deck-mounted pump. The intake nozzle of the pump-
bottle system was installed next to the nozzle of the pump-filter system. The bottles were returned to the
laboratory for analysis. Figure 1 shows the sand concentrations (> 50 mm) measured with both systems. As
can be observed, the sand concentrations measured with the pump-filter sampler are systematically smaller
than those measured with the pump-bottle method, which is, probably, caused by the loss of sediment
particles through the filter material. Another cause for deviations may be the relatively short filling
(sampling) period of the bottles, being about 1-2 minutes, compared with the sampling period of 5 minutes
for the pump-filter sampler. On the average, the deviations are about 20-30%.
References
Figure 1
Comparison of sand concentrations measured with Pump-Filter Sampler and Pump Bottle Sampler
5.4.4 Comparison of Pump-Sedimentation sampler and Pump-Filter sampler
Concentration measurements were carried out in the Nieuwe Waterweg (New Waterway) near Rotterdam,
the Netherlands. The local flow depth varied from 15-20 m; the (tidal) flow velocities were upto 1.5 m/s. The
bed material consisted of fine sand with D50 = 280 mm.
The intake nozzles of the pump sedimentation sampler and the pump filter sampler were installed next to
each other on a sampling carrier. Both instruments are described in Paragraph 5.6.
The sampling period for the pump sedimentation system was about 3 minutes, while a settling period of 5
minutes was used. The sampling period for the pump filter system was only 1-2 minutes due to rapid filter
blocking (silt particles). Figure 1 shows the sand concentrations (particles > 50 mm) at 1.0 m above the bed
measured with both systems. The overall agreement is quite satisfactory. On the average, the deviations are
about 20-30%. It can also be observed that the scatter in the concentrations measured with the pump filter
system are somewhat larger than those measured with the pump sedimentation system which is probably
caused by the relatively small sampling period of the pump filter system.
References
Van Rijn, L.C., 1980. Methods for in-situ separation of water and sediment. Delft Hydraulics, Report S404
II, The Netherlands
Figure 1
Comparison of sand concentrations measured with Pump-Sedimentation Sampler and Pump Filter Sampler
5.4.5 Comparison of Pump-Sedimentation sampler and Bottle sampler
Sediment concentration measurements were carried out in the Westerschelde (Western Scheldt), a tidal
estuary in the south-west part of the Netherlands. The local flow depth was about 15 m; the tidal flow
velocities were upto 1.5 m/s. The bed material consisted of sand particles (D50= 250 mm). The Bottle sampler
consisted of the filling of bottles.
One-liter bottles were used, which were attached under an angle of 35 with the vertical direction (see Par.
5.6.2.2) to a frame that was lowered to the channel bed. The exact angle of the bottle with the flow direction
under water was unknown (not measured). The bottles were opened by pulling a rope from the survey vessel.
The intake nozzle of the pump sedimentation sampler was installed close to the opening of the bottle. After
each measurement the frame was hoisted out of the water to replace the bottle.
The sampling period of the pump sedimentation system was about 1.5-2 minutes. The-filling (sampling)
period of the bottle is unknown but is assumed to vary from 0.5-1.5 minutes.
Figure 1A shows the sand concentrations (particles > 50 mm) measured with both systems. The results show
large deviations upto 150%. On the average, the sand concentrations measured with the pump sedimentation
sampler are 30-40% smaller. For concentrations smaller than 50 mg/1 the deviations can be explained by
inefficiency of the pump sedimentation method, but for high sand concentrations the loss of small particles
for the pump sedimentation method is rather small (Par. 5.6.3.4). Further, it can be observed that the sand
concentrations measured with the bottle method show a relatively large scatter.
Based on these results and those of the flume tests (Par. 5.6.2.2) it may be questioned if the (simple) bottle
method is a reliable method for measuring sand concentrations.
Figure 1B shows the silt concentrations (particles < 50 mm) measured with both methods. These results show
rather good agreement. On the average the silt particles measured with the pump method are about 10%
smaller.
References
Van Rijn, L.C., 1983. Sediment transport in the Western Scheldt (in Dutch). Delft Hydraulics Laboratory,
Report R1586, The Netherlands
Figure 1
5.4.6 Comparison of OBS and Pump sampler
Instruments
Grasmeijer (in Walstra et al, 1998) did measurements over a sand bed in the Large Oscillating Water
Tunnel (LOWT) of Delft Hydraulics using the optical sensor OBS and a pump sampling system. The LOWT
is described in Section 2.2. The optical OBS sensors were calibrated using the sand bed material in the
LOWT (d50 = 0.12-0.13 mm and 0.19-0.21 mm; d50 varied slightly based on samples before and after the
tests). Three OBS-sensors have been attached to a vertical rod on a footplate (see photographs above); an
EMF velocity probe is also connected to the rod. The footplate is resting on the bed and can move
downwards (due to the movable instrument arrangement) when the bed surface is eroding. Using this
arrangement, an approximately constant distance can be maintained between the bed and the measurement
elevations of both EMF and OBS-sensors during a short measurement period of say 15 to 30 minutes The
EMF-sensor is placed at 0.05 m above the upper side of the footplate. The OBS-sensors are placed at
distances of 0.03, 0.05 and 0.1 m above the upper side of the footplate. The effective measurement elevations
of the OBS-sensors with respect to the surrounding bed surface are approximately 0.025, 0.045 and 0.095 m,
as the instrument will sink down into the bed over about 0.005 m (due to local erosion).
The pump sampler consisted of a vertical array of 10 intake tubes of 3 mm internal diameter connected to the
pumps by plastic hoses. The lowest intake tubes were placed at about 0.01 m above the bed, with the intake
openings placed in a direction transverse to the plane of orbital motion. The intake velocity was about 1 m/s,
satisfying sampling requirements. The 10 liter samples were collected in calibrated buckets. The pump
sampler was operating for 15 minutes giving an average concentration over the measuring time. In case of
fast bed level variations due to large transport rates the operation time of the pump sampler was reduced to 8
minutes giving 5 liter samples. The suspended sand samples at each level above the bed were analyzed to
determine the size composition of the sand.
It was found from the suspended sand samples that in case of relatively coarse bed material (d50 = 0.19-0.21
mm), the median diameter of the suspended sand at z = 0.025 m and z = 0.095 m above the bed is
approximately 20%, respectively 30% smaller than the median grain size of the bed material. In case of
relatively fine bed material (d50 = 0.12-0.13 mm) the vertical sorting was less pronounced. For nearly all tests
with fine sand and at all three heights above the bed the median diameter of the suspended sand was
approximately 10% smaller than the median grain size of the bed material.
To take the effect of a varying grain size with height above the bed into account, the OBS concentrations
were determined from the calibration curve using the particle size of the suspended sand at the same level as
the OBS.
Experimental programme
Two different test series were performed using two different sediment sizes with a median diameter of 0.19-
0.21 mm and 0.12-0.13 mm respectively. Regular asymmetric wave motion (second order Stokes) and
irregular wave motion were generated. Orbital velocities were measured by means of a Laser Doppler
velocity meter at about 0.1 m and 0.2 m above the bed. Most tests were repeated twice to determine the
variation related to small differences in bed arrangement (refilling of sand in the tunnel).
The test procedure was as follows:
positioning of OBS transport meter on bed; foot plate flush with bed surface;
positioning of vertical array of intake tubes with lowest intake tube at 0.01 m above bed;
generation of oscillatory flow during 10 to 20 min.;
establishment of equilibrium bed conditions during about 10 min;
sampling of water-sediment mixture through each intake tube after establishment of equilibrium
conditions;
separation of water and sand; determination of wet and dry sand volumes of samples.
Results
Figures 1 and 2 show examples of time-averaged concentration profiles measured with OBS sensors and the
pump sampler in the Large Oscillating Water Tunnel. It can be observed that the concentrations measured
with the OBS and the concentrations measured with the pump sampler are of the same order of magnitude.
On average, the OBS sensors gave values that were 15% larger (two largest deviations are 250% and -70%)
in case of coarse sand (0.19-0.21 mm) and 30% larger (two largest deviations are 150% and -50%) in case of
fine sand (0.12-0.13 mm) than the values determined with the pump sampler. The largest differences were
found for test B2 and H2 in which the concentration gradient is relatively large (fine sand, small orbital
velocity and weak current compared to other tests).
Values from the OBS sensors showed favorable comparison to values from the pump sampler in case of
relatively large sand concentrations (larger than about 1 kg/m3).
References
Walstra, D.J.R., Van Rijn, L.C., Aarninkhof, S.G.J., 1998. Sand transport at the middle and lower
shoreface of the Dutch coast. Report Z2378, Delft Hydraulics, the Netherlands.
Concentrations
test B00-1, B00-2 and B00-3
0.30
B00-1 pump sampler

0.25
B00-1 OBS
height above bed (m) B00-2 pump sampler
0.20 B00-2 OBS
B00-3 pump sampler
0.15 B00-3 OBS
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Concentrations
test B2-2, B2-3, B2-4 and B2-5
0.30
B2-2 pump sampler
0.25 B2-2 OBS
B2-3 pump sampler
height above bed (m)
0.20 B2-3 OBS

B2-4 pump sampler
0.15 B2-4 OBS

B2-5 pump sampler
B2-5 OBS
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Figure 1
OBS concentrations and pump concentrations in Large Oscillating Water Tunnel.
B00: reg. asym., U1/3,on = 0.52 m/s, U1/3,off = 0.33 m/s, d50 = 0.19-0.21 mm
B2: irreg. asym, U1/3,on = 0.84 m/s, U1/3,off = 0.48 m/s, d50 = 0.19-0.21 mm
Concentrations
test H5-1, H5-2 and H5-3
0.30
H5-1 pump sampler
0.25 H5-1 OBS
H5-2 pump sampler

H5-2 OBS
0.20
H5-3 pump sampler
H5-3 OBS
0.15
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Concentrations
0.30
H8-1 pump sampler
0.25 H8-1 OBS
H8-2 pump sampler
H8-2 OBS
0.20
H8-3 pump sampler
H8-3 OBS
0.15
0.10
0.05
0.00
0.01 0.1 1 10
concentration (g/l)
Figure 2
H5: reg. symm, Umax = 1.30 m/s, d50 = 0.12-13 mm
H8: reg. symm,Umax = 0.67 m/s, d50 = 0.12-0.13 mm
5.4.7 Comparison of ASTM and Pump sampler
Experimental setup
Special experiments (Chung and Grasmeijer, 1999) have been performed in a large-scale flume of Delft
Hydraulics, in which a horizontal sand bed was placed over a length of about 40 m (Figure 4.11). The flume
has a total length of 233 meters, a depth of 7 meters and a width of 5 meters. A piston activated wave board
on one side of the flume generates the waves.
A sand bed ( 0.5 meter high) was placed in the Delta flume from position x = 100 meters to x = 140 meters.
During the first test series (July 1997) this sand bed had a d50 of 0.33 mm, and during the second test series
(August 1997) the d50 was 0.16 mm. The waterdepth was 4.55 m in all experiments.
The acoustic sand transportmeter ASTM was mounted in a tripod, which was placed on the sand bed at
location x = 125 m (Figure 1). The ASTM was used to measure the instantaneous fluid velocities and sand
concentrations at five points above the bed simultaneously. The ASTM was attached to an in vertical
direction movable arm to position the sensors at a known level above the bed. In most tests the measurement
levels were: z= 0.075, 0.125, 0.225, 0.475 and 1.075 m above the bed. A pump sampler (PS) was used to
determine the time-averaged concentrations and to get suspended sand samples at the same levels. Five
intake tubes of the pump sampler were attached at the ASTM sensors (horizontally within 0.2 m; vertically
within 0.01 m from the measurement volume of ASTM) and five other intake tubes were attached to a
supporting rod outside the tripod on the upwave side of it to study the effect of the tripod and ASTM
arrangement on the sand concentrations. These latter intake nozzles of the pump sampling unit were within 2
m of the ASTM sensors (at the same levels above the bed). During each test the instruments were operated
for about 15 minutes to sample over a representative wave record.
Results
Sand concentration profiles based on the pump sampler near the ASTM sensors inside the tripod (these pump
samples were used to calibrate the ASTM) and the pump sampler outside the tripod are presented in Figures
2 and 3 for the tests over a sand bed of 0.16 mm. The concentrations measured with the pump sampling
system were averaged over 7 to 9 tests. Only results for conditions with irregular waves are shown. Standard
deviations in concentration are represented by x-error bars and possible deviations in bed level (due to ripple
migration) are represented by y-error bars. The ripple height was found to be approximately 0.07 m.
All measured concentrations are within a factor 2 of each other without the presence of systematic
differences, which is a remarkably good result. Differences in concentrations can be caused by:
scour and extra turbulence induced by the supporting structure and footplate of the pump sampling system
outside the tripod,
extra turbulence induced by ASTM sensors and supports of tripod,
small local variations in bed forms and bed level inside and outside the tripod.
Turbulence related to the presence of the tripod and the ASTM sensors may produce larger concentrations
near the bed (more sediment is being stirred up). Also, larger concentrations at higher elevations above the
bed may be expected because of an increased turbulent mixing (sediment is distributed more easily by
vortices induced by the presence of the sensors).
Figures 4 and 5 show sand concentration profiles based on the pump sampler near the ASTM sensors inside
the tripod and the pump sampler attached to the flume wall for the tests over a sand bed of 0.33 mm. The
concentrations near the ASTM sensors are systematically larger (factor 2) than the concentrations measured
near the flume wall. Higher up in the water column these effects are supposed to be related to some extra
turbulence induced by the instrumental arrangement. In the near-bed region the observed differences in
concentrations are most likely be caused by differences in bed levels and differences in bed form dimensions
between the location of the pump sampler near the flume wall and the tripod location in the centre of the
flume. From visual observations of the sand bed after draining of the flume it became clear that the presence
of the pump sampling system near the flume wall had caused a scour hole at that location. At locations more
landward and more seaward of the pump sampler near the wall (approximately 1 m from the pump sampler),
an increase in bed level was found (local bump). In the centre of the flume near the ASTM sensors inside the
tripod location, scour effects were much less pronounced. Moreover, well-developed vortex ripples were
found near the ASTM sensors in the centre of the flume while no ripples were found near the pump sampler
attached to the flume wall.
The comparison of the sand concentrations measured by the pump sampler near the ASTM sensors inside the
tripod and measured by the other pump sampler outside the tripod shows that the sand concentration profiles
inside and outside the tripod generally are within a factor 2 of each other. The deviations between the ASTM
and pump sampling concentrations are most probably caused by variations in morphodynamic (bed)
conditions at the measurement locations (presence or absense of scour holes and/or ripples) rather than by
instrumental errors and instrumental arrangement. The ASTM concentrations higher up in the water column
may be somewhat too large due to extra turbulence produced by the tripod (effect of instrumental
arrangement).
Overall, the effect of the instrumental arrangement (ASTM sensors and supports within the tripod) on the
concentrations measured by the ASTM is assumed to be sufficiently small.
References
Chung, D.H. and Grasmeijer, B.T., 1999. Analysis of sand transport under regular and irregular waves in
large-scale wave flume. Report R99-05, Department of Physical Geography, University of Utrecht.
Wave generator
Beach slope of stones
4.5 m
Sand bed
0.5 m
123 m
100 m 121 m 125 m 140 m
Figure 1
Sketch of experimental setup
1.2
Pump samples near ASTM
sensor
1.0
Pump samples from separate
system
0.8
0.6 Irregular waves:

Hs = 1.0 m
h=5m
0.4 Tp = 5 s
D50 = 0.16 mm
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 2
Pump concentrations near ASTM sensors inside tripod and outside tripod (separate system) in case of
irregular waves Hs = 1.0 m and fine sand bed (0.16 mm); average values and standard deviations derived
from 7 tests
1.2
1.0
Irregular waves: Pump samples near ASTM
Hs = 1.25 m sensor
0.8 h=5m Pump samples from
Tp = 5 s separate system
D50 = 0.16 mm
0.6
0.4
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 3
Pump concentrations near ASTM sensors inside tripod and outside tripod (separate system) for irregular
waves Hs = 1.25 m and fine sand bed (0.16 mm); results from 9 tests
1.2
Pump samples near ASTM sensors

1.0
Pump samples near flume wall
0.8
0.6
0.4
Irregular waves:
Hs = 1.0 m
0.2 h=5m
Tp = 5 s
D50 = 0.33 mm
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 4
Pump concentrations near ASTM sensors inside tripod and near flume wall in case of irregular waves Hs =
1.0 m and coarse sand bed (0.33 mm); results from 4 tests
1.2
1.0 Pump samples near ASTM sensors
height above bed (m) Pump samples near flume wall

0.8
0.6
Irregular waves:
Hs = 1.25 m
0.4
h=5m
Tp = 5 s
D50 = 0.33 mm
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 5
5.4.8 Comparison of ASTM, OBS and Pump sampler
Experimental setup
The experiments have been carried out in the Grosser Wellenkanal (GWK) in Hannover, Germany. It has a
length of about 300 m, a depth of 7 m and a width of 5 m. The piston-type wave generator can produce
regular or irregular waves with periods between 1 and 15 s and heights up to 2.5 m. The wave reflections
from the beach are compensated directly at the wave paddle. For the present experiments irregular waves
were generated with a significant wave height of H1/3 = 1.25 m and wave spectrum peak period of Tp = 6.0 s.
In addition, two tests with regular waves (H = 1.6 m, T = 6.5 s) were done. The water depth was kept
constant at a value of 4.2 m in front of the wave generator. The water depth at the test section was 3.5 m. A
sand bed with a length of approximately 50 m was constructed in the flume (Figure 1).
Measurements have been done using an instrumented tripod designed and built by the Laboratory for
Physical Geography of the University of Utrecht, the Netherlands. The tripod was placed in the center of the
flume at x 113 m (Figure 1). Instruments mounted on the tripod included a five-fold acoustic sediment
transport meter (ASTM), three optical backscatterance sensors (OBS), six pump sampling intake tubes, three
electromagnetic velocity meters (EMF), a pressure sensor and a ripple profiler.
The acoustic sand transportmeter ASTM is connected to an arm, which can be moved in vertical direction by
an electromotor. A sensor to detect the bed surface was attached to the lower end of the arm. Using the
movable arm, the ASTM sensors can be placed at pre-selected elevations above the bed. The OBS and EMF
sensors are also attached to the movable arm. Three Optical Backscatter Sensors (OBS) were used during the
present experiments. Two sensors (serial number 455 and 458) were installed on a supporting rod at about
0.03 m above the sand bed. One OBS sensor (serial number 408) was placed at about z = 0.10 m (near the
lowest ASTM sensor for comparison). The lowest EMF sensor (velocity) was positioned at about 0.05 m
above the bed. The pump sampler consisted of a pumping system with 6 intake tubes. Five intake tubes were
attached to the five ASTM sensors at about 0.10 m, 0.15 m, 0.25 m, 0.50 m and 1.0 m above the sand bed.
One intake tube was mounted near two OBS sensors at about 0.03 m above the bed. The intake tubes of 3
mm internal diameter were connected by plastic hoses to the pumps. The intake openings were placed in a
direction transverse to the plane of orbital motion. The intake velocity was about 1.3 m/s, satisfying sampling
requirements. The 8-liter samples were collected in calibrated buckets. The pump sampler was operating for
15 minutes giving an average concentration over the measurement period.
Sand was used with grain size characteristics: d10 = 0.14 mm, d50 = 0.23 mm, d90 = 0.34 mm. The sand bed
was ended by asphalt constructions in which sandtraps were built. The sandtraps were intended for
determining the sediment transport rates based on volumetric changes. Previous experiments showed that
this approach was unsuccessful.
Results
Figure 2 shows a comparison between time-averaged concentrations measured with the pump sampler and
those measured with OBS sensors. Relatively large OBS concentrations were observed on each day during
the first tests, which was caused by resuspension of fine muddy and silty materials that had settled onto the
bed during the night. These relatively large OBS values were removed from the data set. Most OBS
concentrations are smaller than the pump concentrations. The maximum discrepancy is a factor 2.
Figure 3 shows a comparison between time-averaged concentrations measured with the ASTM sensors
(based on standard calibration curve; concentrationASTM = 0.257*OutputASTM) and the pump sampler. The
ASTM values have been interpolated to compare concentrations from the ASTM and the pump sampler at
the same height above the bed. For concentrations smaller than 1 kg/m3 the ASTM concentrations and the
pump concentrations agree very well (ASTM on an average 5% smaller). For concentrations larger than 1
kg/m3 the ASTM concentrations are slightly smaller than the pump sampler concentrations (ASTM on an
average 22% smaller).
wave paddle
3.5 m tripod sand trap

4.2 m asphalt 1:6
asphalt sand 1:10
sand bed
63.1 80.6 82.1 86.8 128.2 133.0 134.5 149.0
Figure 1
Sketch of Grosser WellenKanal
3
OBS (g/l)
B
0
0 1 2 3 4 5
pump sampler (g/l)
Figure 2
Comparison of time-averaged OBS concentrations and pump concentrations
4
conc. ASTM (kg/m )
3
0
0 1 2 3 4 5
3
conc. pump sampler (kg/m )
Figure 3
Comparison of time-averaged ASTM concentrations and pump concentrations
5.4.9 Comparison of ABS and Pump sampler
Experimental setup
Thorne et al. (2002) tested the acoustic backscatter sensor ABS system in the large-scale wave tank of Delft
Hydraulics (see also Williams et al., 2000). The ABS was mounted in a tripod of Proudman Oceanographic
Laboratories, which was placed on the sand bed (0.33 mm sand; rippled bed). A pump sampling system was
used to determine the sand concentrations at 10 points above the bed surface. The intake nozzles were
installed at a horizontal separation distance of 0.3 m from the vertical ABS beam. Figure 1 shows a
comparison of time-averaged pump and ABS concentrations (over about 15 min) at various levels based on
results of tests with regular and irregular waves (no current). The sand concentrations were processed from
the ABS signal between 0.01 and 1 m with a vertical resolution of 0.01 m. The sand concentration at one
height and the suspended sand sizes over the depth were assumed to be known from the pump sampling
results to calibrate the ABS concentrations. Hence, the ABS concentrations were not independently
determined. The deviations between the results of both instruments are up to 30%.
Figure 2 (Top and Bottom) shows a comparison of pump concentrations Pc and ABS concentrations Ac for a
site in a British estuary (0.16 mm sand bed; bed covered with sand waves; tidal current; no waves). The
horizontal separation distance between the intake nozzles (at 4 heights) and the ABS beam was 0.5 m. The
ABS concentrations were determined independently of the pump concentrations. The deviations between the
results of both instruments are up to 20%.
Rose and Thorne (2000) tested the multi-frequency (1, 2.5 and 5MHz) ABS-system mounted in a tripod in
conditions with tidal flow velocities between 0.6 and 1 m/s and water depths between 2 and 3 m (River Taw
Estuary, UK). The bed was composed of fine sand with d50=0.17 mm and d90= 0.2 mm. Ripples were present
with approximate wave heights of 0.025 m and wave lenghts of 0.2 m. Figure 3 shows four ABS-
concentration profiles and pumped concentrations at 4 heights (= 0.1, 0.2, 0.4 and 0.8 m) above the bed. The
agreement between the concentrations of both instruments is quite good with maximum deviations of about
15% near the bed and 30% higher up in the water column. The magnitude of the acoustically derived
suspended sediment estimates were slightly adjusted (factor 0.8 to 1.1) to optimize the agreement with the
direct pumped sampling measurements. The 5 MHz data was not utilised owing to inversion difficulties at
periods of high concentrations when sediment attenuation of the MHz signal resulted in problematic
concentrations.
References
Thorne, P.D., Williams, J.J. and Davies, A.C., 2002. Suspended sediments under waves measured in a
large-scale flume facility. Journal of Geophysical Research, Vol. 107, No. C8, 3178, p. 4.1-4.16
Rose, C. and Thorne, P., 2000. Measurements of suspended transport parameters in a tidal estuary.
Nearshore and Coastal Oceanography/Continental Shelf Research.
Williams, J. et al., 2000. Observed and predicted vertical suspended sediment concentration profiles and
bed forms in oscillatory-only flow. Journal of Coastal Research
Title:
deltaflume.eps
Creator:
MATLAB, The Mathworks, Inc.
Preview:
This EPS picture was not saved
with a preview included in it.
Comment:
This EPS picture will print to a
PostScript printer, but not to
other types of printers.
Figure 1
Comparison of pump Cp and ABS Ca concentrations in large scale wave tank of Delft Hydraulics
Title:
estuary.eps
Creator:
MATLAB, The Mathworks, Inc.
Preview:
This EPS picture was not saved
with a preview included in it.
Comment:
This EPS picture will print to a
PostScript printer, but not to
other types of printers.
Figure 2
Top: Comparison of suspended sand size; As=particle size derived from ABS signal; Ps= particle
size derived from pump samples
Bottom: Comparison of suspended sand size; Ac=concentration derived from ABS signal; Pc=
concentration derived from pump samples
Figure 3
Sand concentration profiles of ABS system (circles) and pumped concentrations (crosses); River Taw
Estuary (UK)
5.4.10 Overall conclusions with respect to OBS, ASTM and ABS instruments
The following conclusions are given:

the OBS sensors are highly sensitive to particle size;
the OBS concentrations should be corrected for the smaller size of the suspended sand relative to that of
the sand bed material used during calibration; this requires samples of suspended sand during field
measurements;
the effective concentration range for OBS sensors is 1 to 100 kg/m3; concentrations measured within this
range have an inaccuracy of about 50%;
the OBS sensors often show a reasonably steady offset concentration, which is related to the background
concentration of relatively fine sediments (silt and mud), which should be subtracted from the original
time series data; if the background concentration to be subtracted from the record is of the same order of
magnitude as the sand concentration, the OBS concentrations will be rather inaccurate;
OBS sand concentrations below 1 kg/m3 are not accurate (inaccuracy larger than factor 2) and should not
be used;
Acoustic concentration sensors (ASTM) for point measurements are not very sensitive to particle size; the
effective measurement range is 0.1 to 10 kg/m3; concentrations measured by sensors mounted in a tripod
have an inaccuracy of about 50%;
Acoustic concentration sensors (ABS) for profile measurements are sensitive to particle size; the effective
measurement range is 0.1 to 20 kg/m3; three different acoustic frequencies are used to simultaneously
determine the size and the concentration of the suspended sediment involved; the sand concentration
profiles can be determined with an inaccuracy of about 30% if the suspended sand size is known and
some pump concentrations are available for calibration; the optimum conditions for the ABS system are:
rather uniform fine sand (0.1 to 0.3 mm) in non-breaking wave conditions;
the mass concentration at range r from the ABS transducer is estimated from a complex function,
depending on the voltage V measured at range r, the sediment density, the speed of sound in water and the
attenuation of sound by water and sediment of radius a; this function is not yet properly defined resulting
in relatively large inaccuracies (factor 2 to 3) of the measured concentration, if calibration samples of
concentration and sediment size are not available;
the ABS system is rather sensitive for the presence of air bubbles in the water column (surf zone
conditions with breaking waves) resulting in an increase of the concentration.
The basic principle of mechanical trap-type bed-load samplers is the interception of the sediment particles
which are in transport close to the bed over a small incremental width of the channel bed. Most of the
particles close to the bed are transported as bed load but the sampler will inherently collect a small part of the
suspended load (related to vertical size of intake mouth).
The bed-load transport measured by a mechanical sampler is dependent on its efficiency (instrumental
errors), on its location with respect to the bed form geometry (spatial variability) and on the near-bed
turbulence structure (temporal variability).
The efficiency of the bed-load sampler depends on the hydraulic coefficient, the percentage of width of the
sampler nozzle in contact with the bed during sampling and on sampling disturbances generated at the
beginning and the end of the sampling period.
The hydraulic coefficient, defined as the ratio of the inflow velocity and the ambient flow velocity, depends
on the geometry and construction of the sampler nozzle, on the position of the nozzle at the bed, on the
percentage of filling of the bag with sand particles (volume of catch) and on the percentage of blocking of
the bag material by fine sand, silt or clay particles and organic materials, depending on the mesh size of the
bag (Beschta, 1981) . Laboratory experiments have shown (Emmett, 1980; Delft Hydraulics, 1991) that the
sampler (bag or basket) can be filled to about 40% of its capacity without reduction of the hydraulic
efficiency and without loss of sediment particles during raising of the instrument. Thus, the problems of bed-
load transport measurements are related to the instrumental design and to the physical processes involved.
Typical instrumental problems of a (bag-type) bed-load sampler are:

the initial effect; sand particles of the bed may be stirred up and trapped when the instrument is placed on
the bed (oversampling),
the gap effect; a gap between the bed and the sampler mouth may be present initially or generated at a
later stage under the mouth of the sampler due to migrating ripples or erosion processes (undersampling),
the blocking effect; blocking of the bag material by sand, silt, clay particles and organic materials will
reduce the hydraulic coefficient and thus the sampling efficiency (undersampling),
the scooping effect; the instrument may drift downstream from the survey boat during lowering to the
bed and it may be pulled forward (scoop) over the bed when it is raised again so that it acts as a grab
sampler (oversampling).
Typical sampling problems related to the variability of the physical processes of bed-load transport are (see
Carey, 1985; Delft Hydraulics, 1991) :
the number of measuring locations along a bed form (dune),
the number of measurements at each location,
the sampling duration of each measurement,
the number of locations in the cross-section.
Here, the sampling duration will be discussed. The number of measurements is discussed in Par. 3.3.4.
Three effects related to the sampling period are considered:

the volume of the nylon bag (maximum sampling period),
the blocking of the nylon material by sediment particles,
the presence or generation of a gap under the sampler mouth.
The size of the bag imposes a maximum sampling period which depends on the transport rate and hence on
the current velocity.
Taking a maximum filling percentage of 50%, the maximum sand catch of a 2 liter bag is about 2000 grams.
The bed-load transport formula of Meyer-Peter-Muller has been used to compute the maximum sampling
period for bed material of 500 mm and velocities in the range of 0.75 to 2 m/s
The maximum transport rate is assumed to be equal to three times the mean rate (qb,max = 3 qb,mean).
The results are given in Table 1.
Depth-averaged velocity (m/s) Maximum sampling period (sec)
0.75 450
1.0 120
1.25 50
1.5 30
1.75 15
2.0 10
Table 1
Maximum sampling period
Blocking of standard nylon bags by fine sediments was studied by Beschta (1981) using the Helley-Smith
bed-load sampler in flume and field conditions. The 200 mm-bag was found to be rapidly clogged by fine
sediment particles reducing the sampling efficiency to 50% after 30 seconds and to only 10% after 300
seconds. Thus, the sampling period must be rather short (< 30 seconds) whenever a standard bag is used. A
special bag consisting of nylon material with a mesh size of 250 mm and a patch (0.10 x 0.15 m2) of 500 mm
at the upper side of the bag has been used by Van Rijn and Gaweesh (1992). The application of the special
bag yields good results as regards the hydraulic coefficient.
Camera observations (Delft Hydraulics, 1991, 1992) did show the presence of an initial gap or the
generation of a gap at a later stage (after 2 or 3 minutes) during some of the sampling periods. Given the
stochastic nature of the generation of gaps (migrating ripples), its effect on the average transport rate will be
the least when many samples of short duration are taken (gap effect is averaged out). Furthermore, the
application of a short sampling period reduces the time available for the generation of gaps when initially no
gap is present.
Based on the above-given considerations, it is advised to use a maximum sampling period of 3 minutes for
depth-averaged velocities upto 0.8 m/s
At higher velocities (higer transport rates) the maximum sampling period must be reduced (between 0.5 and
3 minutes) or a larger bag should be used. The largest sand catch should not be larger than 50% of the bag
volume.
References
Beschta, R.L., 1981. Increased Bag Size improves Helley-Smith Bed Load Sampler for Use in
Streams with High Sand and Organic Matter Transport. Symposium Erosion and Sediment Transport
Measurement, Florence, Italy
Carey, P., 1985. Variability in Measured Bed-load Transport Rates. Water Resources Bulletin, Vol. 21, No.
1, Paper No. 84156.
Delft Hydraulics, 1991. Delft Nile Sampler, Bed Load Transport Measurements in the River Waal.
Report Q1300 part 1, Delft, The Netherlands
Delft Hydraulics, 1992. Bed Load Transport Measurements in the Waal River near Woudrichem and
Druten. Report Q1300 part 2 and 3, Delft, The Netherlands
Emmett, W.W., 1980. A Field Calibration of the Sediment Trapping Characteristics of the
Helley-Smith Bed Load Sampler. Geological Survey Professional Paper 1139, Washington, USA
Vries, M. de, 1973. On Measuring Discharge and Sediment Transport in River Flow. Delft Hydraulics
Laboratory, Publication No. 106, The Netherlands
Vries, M. de, 1982. Lecture Notes on Measuring Sediment Transport in Rivers. Delft University of
Technology, Delft, The Netherlands
5.5.1.2 Bed load transportmeter Arnhem (BTMA)
Principle
The instrument is based on the collection of sediment particles by means of a basket type sampler. The
basket, consisting of fine wire mesh and mounted in a frame, is pressed (by means of a spring) on the
channel bed after lowering of the frame (Figure 1). The form of the basket causes a pressure reduction
behind the instrument so that the water and sediment particles enter the mouth of the basket with the same
velocity as that of the ambient flow, provided that the sediment content, already in the basket, is relatively
small. The sampler can collect particles coarser than 0.3 mm (mesh size basket) but finer than 50 mm
(opening height).
Practical operation
1. lower bed-load sampler to the bed
2. use a preset sampling time (usually 2 minutes)
3. raise bed load sampler (gently)
4. wash sediment catch from the basket into a calibrated funnel
5. read (immersed) volume of sediment catch
6. wash sediment catch into a sample container (for laboratory analysis)
7. change sampling location (if necessary) and repeat sampling procedure.
8. note data on measuring sheet (see Figure 2)
Remarks
1. Before starting a series of measurements, an echo sounding of the longitudinal bed section at the
sampling location must be made to determine the bed form length. If the bed form length (dunes) is large
compared with the boat length, the sampling location must be changed regularly in longitudinal direction
to assure random sampling (Figure 1) and at least 20 samples should be collected.
2. The sampling time may be larger than 2 minutes but never larger than the time needed to fill the
sample basket for about 40% of the capacity.
3. It is sufficiently accurate to determine the immersed volume of the sediment catch on board of the
vessel by using a calibrated funnel. Single samples should be returned to the laboratory to determine the
porosity factor of the samples.
Laboratory analysis
1. determine dry weight of each sample (if necessary)
2. accumulate samples of each measuring location and determine size distribution of the sediment
particles by sieving or settling tests.
Results and accuracy
The bed-load transport (in kg/sm) can be determined as:
Sb=a rs(1-p)Vs/(bT) or Sb= Gs/(bT)
in which:
a = calibration factor (= 2 for BTMA)
p = porosity factor (= 0.4)
rs = density of sediment particles (= 2650 kg/m3)
Vs = immersed volume of sediment catch (m3)
Gs = dry mass of sediment catch (kg)
b = width of intake opening (= 0.085 m for BTMA)
T = sampling period (s)
CHAPTER 5: INSTRUMENTS SEDIMENT TRANSPORT February 2006

The accuracy of the measured bed-load transport is strongly dependent on the accuracy of the calibration
factor a, the number of measurements and the sampling procedure (skilled personel). Figure 1 presents
calibration curves for the BTMA showing considerable variations. In practice a calibration factor equal to 2
is used (drawn line).
Assuming ideal sampling, at least 20 samples must be collected at each location to obtain a bed-load
transport rate with a standard deviation (error) of about 20% (De Vries, 1973). In practice the sampling error
will be considerably larger (say 100%) particularly due to the sampling procedure. More information of the
calibration of bed load samplers is given by Hubbell et al. 1985.
Technical specifications
intake opening: width of 0.085 m; height of 0.050 m

dimensions: frame 0.40 x 0.8 x 1.85 m; basket 0.1x0.15x0.5 m
mesh size of basket: 300 mm (0.3 mm)
weight of sampler: 32 kg
cycle period: 5 min (minimum period between two measurements)
Advantages
1. simple and reliable instrument
Disadvantages
1. large sampling effort (time consuming)
2. winch facility needed
3. unreliable calibration factor
References
Delft Hydraulics, 1958. Calibration of BTMA (in Dutch) Report M6Q1-I, The Netherlands
Delft Hydraulics, 1966. Development of Bed Load Samplers (in Dutch) Report M601-II, The Netherlands
Delft Hydraulics, 1969. Calibration of Bed Load Samplers (in Dutch) Report M601-III, The Netherlands
De Vries, M., 1973. On Measuring Discharge and Sediment Transport in River Flow. Delft Hydraulics
Hubbell.D.W., Stevens, H.H., Skinner, J.V. and Beverage, J.P., 1985. New Approach to Calibrating Bed
Load Samplers. Journal of Hydraulic Engineering, Vol.111, No.4

Figure 1
B.T.M.A.

Figure 2
Measuring sheet bed load sampler

5.5.1.3 Helley Smith (HS)
Principle
The Helley-Smith bed-load sampler is a modified version of the BTMA-sampler (Helley and Smith, 1971).
The Helley-Smith sampler consists of a nozzle, sample bag and frame (Figure 1).
The sampler has a square entrance nozzle (0.076 x 0.076 m) and a sample bag constructed of 250 pm- mesh
polyester. Several different versions of the sampler have been used for various field conditions. Larger
nozzles are generally used to sample larger sediment sizes and heavier samplers become necessary as deeper
and faster rivers are sampled. An important advantage of the Helley-Smith sampler is the extensive
calibration (based on about 10,000 samples) and its simple operation.
Practical operation (see Paragraph 5.5.1.2)
Laboratory analysis (see Paragraph 5.5.1.2)

Sb=a rs(1-p)Vs/(bT) or Sb= Gs/(bT)
in which:
a = calibration factor (= 0.5 for particles of 0.25 to 0.5 mm)
(= 1.0 for particles of 0.5 to 16 mm)
(= 1.5 for particles of 16 to 32 mm)
rs = density of sediment particles (= 2650 kg/m3)
b = width of intake opening (= 0.0762 m for HS)
Figures 2 and 3 present calibration curves relating the sediment catches and the actual transport rates
for various size fractions. The actual transport rate has been assumed to be represented by
simultaneous measurements with a conveyer belt system just downstream of the sampling position of
the Helley-Smith sampler (Emmett, 1980).
The conveyer-belt trap consists of a concrete slot (width = 0.4 m, depth = 0.6 m) in the channel bed,
orthogonal to the flow direction. Along the bottom of the concrete slot passes an endless belt of rubber
(width = 0.3 m). Sediment falling into the open slot drops on the moving belt, and is carried laterally to
the riverbank where it is scraped off the belt, weighted and returned to the river flow (depth = 1.2 m, width
= 15 m, discharge = 20 m3/s).
Based on Figures 2 and 3, Emmet (1980) concluded that the Helley-Smith Sampler has an efficiency of
100% (calibration factor, a = 1) for particle sizes in the range of 0.5 mm to 16 mm. For particles in
the range of 0.25 to 0.5 mm the efficiency is found to be about 175%.
(resulting in a calibration factor a= 0.5) which is assumed to be caused by the collection of suspended
sediment particles.
For particles in the range of 16 to 32 mm the efficiency is found to be about 70% (resulting in a
calibration factor a = 1.5) which is assumed to be caused by the paucity of large particles moving as
bed load in the flow.
Summarizing:
a= 0.5, for particles in the range of 0.25 to 0.5 mm,
a= 1.0, for particles in the range of 0.5 to 16 mm,
a= 1.5, for particles in the range of 16 to 32 mm.
More information of the calibration of bed load samplers is given by Hubbell et al., 1985.
The original Helley Smith sampler was tested by Delft Hydraulics (1996, 1997). The sampler was tested in a
laboratory flume (known bed load transport) with water depths in the range of 0.5 to 0.6 m and sediment of
D50=0.46 and 0.75 mm. About 25 samplings were required to obtain a variation coefficient lower than 0.1.
The results are given in the following Table. The HS sampler was found to suffer from the problem of
oversampling with a factor of 2 to 3, especially at low velocities.
Bed load D50=0.46 mm D50=0.75 mm

transport D10=0.33 mm D10=0.20 mm
measurements D90=0.65 mm D90=9 mm
v=0.5 m/s v=0.75 m/s v=1 m/s v=0.8 m/s v=0.9 m/s
1) Bed load 4 gr/s/m 70 gr/s/m 210 gr/s/m 70 gr/s/m 95 gr/s/m
transport in
Flume
transport
measured by
Helley Smith
Ratio of 3.0 1.6 1.7 2.2 2.5
2) and 1)
dimensions: frame 0.18x0.32x1 m
mesh size of basket: 250 mm (0.25 mm)
Advantages
2. easy to handle
Disadvantages
References
Delft Hydraulics, 1996. Test measurements of Helley Smith and modified Helly Smith (in Dutch). Report Q
2141, Delft, The Netherlands
Delft Hydraulics, 1997. Calibration and comparison of Helley Smith. Report Q 2345,Delft, The Netherlands
Emmett, W.W., 1980. A Field Calibration of the Sediment Trapping Characteristics of the Helley-Smith
Bed Load Sampler. Geological Survey Professional Paper 1139, Washington, USA
Helley, E.J. and Smith, W., 1971. Development and Calibration of a Pressure-Difference Bed Load
Sampler. U.S. Geological Survey Open File Report, Washington, USA
Hubbell, D.W., Stevens, H.H., Skinner, J.V. and Beverage, J.P., 1985. New Approach to Calibrating Bed
Figure 1
Helley Smith
Figure 2
Calibration curves Helley Smith
Figure 3
5.5.1.4 Delft Nile sampler (DNS)
Principle
The instrument consists of a bed-load sampler and a suspended sampler attached to a supporting frame (see
Figure 1). The sampler has a weight of about 60 kg. The suspended load sampler consists of 7 intake nozzles
(inner diameter = 0.003 m) which are connected to plastic hoses and operated by pumps.
The bed-load sampler consists of a nozzle (entrance width = 0.096 m, entrance height = 0.055 m, length =
0.085 m, rear width = 0.105 m, rear height = 0.06 m) connected to a bag. The bag consists of nylon material
with a mesh size of 150 or 250 mm, depending on the size of bed material. At the upper side of the bag a
patch (0.10 x 0.15 m2) consisting of 500 mm material is present to reduce the blocking effect by fine particles
as much as possible (see Figure 1). The bottom side of the sampler nozzle has a forward-tilting slope of 1 to
10. The bed-load sampler is connected to a swing arm, which can move in the vertical direction (upward and
downward) over a distance of about 0.15 m with respect to the supporting frame. In its hanging position
during lowering of the instrument, the bed-load sampler is in its most upward position. When the frame of
the instrument is placed on the bed, the sampler nozzle makes contact with the bed by relaxing the tension in
the cable. When the instrument is raised, first the nozzle of the bed-load sampler is raised over 0.15 m and
then the frame is lifted from the bed. The swing-arm construction was designed to reduce the gap effect and
the scooping effect as much as possible. The instrument was developed for bed-load measurements in the
Nile river.
The hydraulic coefficient of the bed-load sampler, defined as the ratio of the inflow velocity and the ambient
flow velocity, was determined by measurements in a flume (see Table 1). A small Ott-propeller meter (dia-
meter = 0.02 m) was placed in the entrance of the nozzle. The velocity in the nozzle was compared with the
flow velocity measured simultaneously at the same height in a section two metres upstream of the sampler.
Ten successive measurements of 30 seconds each were carried out. Three types of bags were used, as
follows:
1. 250 mm-nylon bag and 500 mm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag (see Fig. 1),
2. 150 mm-nylon bag and 500 mm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag,
3. impermeable plastic bag and 500 mm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag.
This latter bag is supposed to simulate a nylon bag which is almost fully blocked by fine silt and clay
material as present in most natural conditions (Beschta, 1981). Four filling percentages of the bag were
tested: 0%, 25%, 50% and 75%.
The hydraulic coefficients of the 250 mm -bag and the 150 mm -bag, both with a patch of 500 mm at the upper
side, are about unity for filling percentages in the range of 0% to 50%. A filling percentage of 75% reduces
the hydraulic coefficient to about 0.75. The hydraulic coefficient of an impermeable plastic bag with a patch
of 500 mm at the upper side, simulating a blocked nylon bag, is about 0.8 for a filling percentage in the range
of 0% to 50%. A filling percentage of 75% reduces the hydraulic coefficient to about 0.70. Based on these
results, a maximum filling percentage of 50% is advised to be used. Blocking of the nylon material by fine
sediments will result in a hydraulic coefficient of about 0.8.
Finally, it is noted that the importance of a hydraulic coefficient equal to unity should not be
overemphasized, because actually the sampling efficiency is the most important parameter. A hydraulic
coefficient of unity does not necessarily give a sampling efficiency of unity, because other factors are
involved.
Velocity Mesh size Mesh size Hydraulic coefficient
in mouth patch
Filling Filling Filling Filling

percentage percentage percentage percentage
(m/s) (mm) (mm) 0% 25% 50% 75%
250 500 1.04 1.01 0.98 0.91
0.5 150 500 1.07 1.04 0.98 0.94
0 (impermeable) 500 0.82 0.82 0.82 0.82
250 500 1.05 1.02 1.01 0.75

0.8 150 500 1.08 1.03 0.99 0.75
0 (impermeable) 500 0.84 0.80 0.80 0.70
Table 1
Hydraulic coefficients (average values)
The sampling efficiency is defined as the ratio of the bed-load transport measured by the sampler at a certain
location during a certain period and the true bed-load transport at the same location during the same period
(if the sampler had not been there). The sampling efficiency will express sampling errors related to the initial
and scooping effect, the gap effect and the loss of sediment particles through the patch of the bag.

Sb= a(Gs Go)/(bT)
in which:
a = calibration factor (= 1 to 1.5),
Gs = dry mass of sediment catch (kg),
Go = dry mass of sediment catch related to initial and scooping effect determined by zero-sampling
(sampler is lowered to bed and immediately raised),
b = width of intake opening (= 0.096 m for DNS),
T = sampling period (s).
The Go-values should be determined by taking 10 zero-samplings at each location. Practical experience sofar
shows a value of about 0.03 kg. The sampling efficiency (a-factor) was determined by means of tests in a
flume at the Hydraulics and Sediment Research Institute in Deltabarrage, Egypt. Different velocity and sand
size ranges were considered. The a-factor was found to be in the range of 1 to 1.5.
dimensions: frame 0.5x0.6x1.1 m
mesh size of basket: 150 and 250 mm
Advantages
2. easy to handle
Disadvantages
References
Van Rijn, L.C. and Gaweesh, M., 1992. A New Total Load Sampler. Journal of Hydraulic Engineering,
Vol. 118, No. 12
Delft Hydraulics, 1991. Delft Nile Sampler; Bed Load Measurements in the River Waal. Report Q1300, part
Delft Hydraulics, 1992. Bed Load Measurements in the River Waal near Woudrichem and near Druten
Report Q1300, part 2 and 3, Delft, The Netherlands
Figure 1
Delft Nile Sampler
Figure 2
Delft Nile Sampler
5.5.2 Bed form tracking
Principle
The basic principle is the computation of the bed-load transport from bed-form profiles measured at
successive time intervals under similar flow conditions (Figure 1).
Assuming steady flow conditions and undisturbed bed-form migration, the bed-load transport rate can be
computed from the bed form dimensions (Engel and Lau, 1980, 1981, De Boer, 1996).
Sb=as rs(1-p) f a D
in which:
as = coefficient (0.5 to 0.6),
p = porosity factor (= 0.4),
rs = density of sediment particles (= 2650 kg/m3),
f = shape factor (2V/(D l)),
a = average migration velocity (m/s),
V = volume of bed form per unit width,
D = average bed form height (m),
l = bed form length.
To apply this equation, the migration velocity and the bed form height must be determined from the bed
profiles. The bed-load transport rate can also be computed directly from the (successive) profile data using
all data instead of selecting the characteristic parameters such as the average migration velocity and the bed-
form height (see Havinga, 1982). To collect the bed profile data along a prefixed course, an accurate three-
dimensional measuring system must be available consisting of a two-dimensional horizontal positioning
system and a one-dimensional vertical sounding system.
In (isolated) field conditions, where an accurate positioning system is too complicated, a much simplier
method can be used. By means of an analoque echo sounder two or more successive bed-profile registrations
can be made in a longitudinal section between two well-defined cross-sections (bank marker). Using a simple
hand method, the average migration velocity and bed-form height can be determined quite easily, as
shown in Figure 1A.
Engel and Wiebe report an overall inaccuracy of about 40 to 50% for flume conditions. Figure 1B shows
measured and computed transport rates for flume conditions (Simons et al, 1965). For field conditions the
inaccuracy may be as large as 100%.
De Boer (1996) has developed a dune-track computer program to estimate the bed load transport from
successive bed profile measurements. The as-factor was found to be 0.64 for the Dutch IJssel river. The
length of each profile should at least be between 1 and 3 km. The local bed load transport may vary between
50% and 200% of the average value for one profile.
References
De Boer, A.G., 1996. The applicability of the dune-track method (in Dutch). Department of Physical
Geography, University of Utrecht, Utrecht, The Netherlands
Engel, P. and Wiebe, K., 1979. A Hydrographic Method for Bed-Load Measurement. Proc. Fourth
Nat. Hydro-Techn. Conf. River Basin Man., Vol. I, page 98-113. Vancouver, Canada
Engel, P. and Lau, Y.L., 1980. Computation of Bed Load Using Bathymetric Data. Journal of the
Hydraulics Division, ASCE, HY 3
Engel, P. and Lau, Y.L., 1981. Bed Load Discharge Coefficient. Journal of the Hydraulics Division,
ASCE, HY 11
Havinga, H., 1982. Bed Load Determination by Dune Tracking. Dir. Water Management and Water
Motion, District South East, Rijkswaterstaat, The Netherlands
Simons, D.B., Richardson, E.V. and Nordin, C.F., 1965. Bed Load Equation for Ripples and Dunes. U.S.
Geol. Survey Prof. Paper 462 H, Washington, USA
Figure 1
Bed load transport according to bed form tracking
5.6.1 Classification of samplers
The suspended load samplers described in the following paragraphs can be classified according to their
measuring principles (see Par. 3.1), as follows:
Suspended load samplers Point-integrating Depth-integrating
Direct method Delft-Bottle sampler

Acoustic samplers
Indirect method Trap samplers USD-49

Bottle sampler Collapsible-Bag sampler
USP-61
Pump samplers
Optical samplers
Impact samplers
The trap and bottle-type samplers can only be used in (quasi) steady flow conditions. The other samplers can
be used in unidirectional and oscillatory flow (waves).
The basic principle of all mechanical bottle and trap samplers is the collection of a water-sediment sample to
determine the local sediment concentration, transport and/or particle size by physical laboratory analysis.
Optimal sampling of a water-sediment volume by means of a mechanical instrument requires an intake
velocity equal to the local flow velocity (iso-kinetic sampling) or a hydraulic coefficient, defined as the ratio
of the intake velocity and local flow velocity, equal to unity (hc=1).
Differences between the intake velocity and local flow velocity result in sampling errors. Sampling at a
lower velocity than that of the ambient flow would result in a higher sediment concentration than present in
the flow due to diverging flow lines, which cannot be followed by the sediment particles, being of higher
density than the water particles (Figure 1A). Conversely, sampling at a higher velocity than that of the
ambient flow would result in a lower sediment concentration (Figure 1B). The magnitude of error due to
incorrect intake velocities in suspensions of various concentrations has been determined experimentally by
Nelson and Benedict (1950), whose results are shown in Figure 2. The results show a decreasing effect with
a decreasing particle size. It is also clear that a hydraulic coefficient smaller than 1 results in a relatively
large error. Further Nelson and Benedict found no significant change of the sampling error in relation to the
flow velocity (range 0.9 to 1.5 m/s), the size of the intake nozzle (range of 4 to 7 mm), the angle between
flow direction and nozzle axis (range 0 to 20), for a hydraulic coefficient in the range of 0.5 to 2 in all tests.
Crickmore and Aked (1975) report that no systematic trends can be distinguished for a hydraulic
coefficient in the range 0.5 to 4 and particle sizes from 60 pm to 250 mm. The maximum experimental error
found was about 10%.
Laboratory tests using an intake nozzle normal to the flow direction show errors of about 15% for particle
sizes of 150 pm and 450 mm, while for 60 mm - sediment no errors were found (Nelson and Benedict, 1950).
Hulsbergen (1981) using an intake nozzle normal to the flow direction reports an error of about -15%
(actual concentration larger than measured concentration) for 60 mm-sand, of about -25% for 110mm and 270
mm-sand and of about -33% for 500 mm-sand. In all tests the hydraulic coefficient was in the range 1 to 4.
Crickmore and Aked (1975) found a maximum error of 20% for a misalignment of 180 using a hydraulic
coefficient larger than 1.
Bottle and Trap samplers are usable in steady flow conditions (rivers), but not in time-dependent flow
conditions (estuaria and coasts).
References
Crickmore, M.J. and Aked, R.F., 1975. Pump Sampler for Measuring Sand Transport in Tidal Waters
Conference on Instrumentating Oceanography, I.E.R.E., Conference Proceeding No. 2, Bangor,
England
Hulsbergen, C.H., 1981. Determination of Sand Concentration by Pump Sampling normal to the flow.
Delft Hydraulics, M1267, Delft, The Netherlands
Nelson, M.E.and Benedict, P.C., 1950. Measurement and Analysis of Suspended Sediment Loads in
Streams. A.S.C.E.-Proceedings, Volume 76, U.S.A.
Van Rijn, L.C. van, 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Research Report S404, The
Netherlands
Figure 1
Influence of intake velocity on sediment paths
5.6.2.2 Bottle sampler
Principle
The method is based on the filling of a bottle to determine the silt and/or sand concentration at a specific
point in the flow. Usually, the bottle is placed vertically in a container (Figure 1) and lowered to the
sampling point, where the bottle is opened (mechanical or electrical). A cork ball should be present to close
the bottle after filling. The exact filling time is unknown but may vary from about 20 to 400 sec, depending
on the bottle orientation, flow velocity and sampling height as shown in Figure 2A (Delft Hydraulics,
1980).
Rapid profile measurement can be achieved by using a rack with 5 bottles (or more), which are opened at
prefixed depths.
Practical operation
1. lower container to sampling position (echo-sounder attached to bottle)
2. remove stop (mechanical or electrical)
3. wait till bottle is filled (3~
5 minutes)
4. raise container (slowly) and replace bottle.
5. note data on measuring sheet (Figure 3)
Laboratory analysis
The bottles should be stored away from direct sunlight. Analysis should be completed before the growth of
organisms.
The standard procedure for determining the sediment concentration is as follows (see also Paragraph 8.1.2,
Fig. 1):
1. measure water volume of sample
2. allow sample to settle for 24 hours and decant sediment-free water
3. wash sample over a 50 mm - sieve suspended in a glass cylinder under
light brushing to separate sand (use distilled water to remove salt)
4. wash sand sample over a pre-weighed non-hygroscopic nylon filter (0.5 mm)
5. filter silt sample through a pre-weighed 0.5 mm-filter (use non- hygroscopic nylon or glass-fiber filter
material)
6. use distilled water to wash free from salt
7. dry and weigh sand and silt samples
8. store sand sample for size analysis.
In the case of a large amount of samples the laboratory analysis can be reduced considerably by using an
optical method to determine the silt concentration (after separation of sand particles using the 50 mm - sieve).
About 10% of the samples are filtered to obtain a calibration curve. However, this method may introduce
additional errors due to scatter of the calibration curve.
Another alternative method is the use of ultra-centrifuge tubes to separate the silt particles.

The silt and sand concentrations are determined as:
c=Gs/V
in which:
Gs = dry mass of sediment (mg),
V = volume of water sample (l).
M
easurements in a laboratory flume ( Delft Hydraulics, 1980) have shown that the efficiency of a bottle in
collecting the sand particles (> 50 mm) is strongly dependent on the orientation of the bottle opening to the
main flow direction. Figure 2B shows the bottle efficiency as a function of the angle (a) with the flow
direction. The optimal angle is about 35. For a <35 the measured concentration is too small, while for a >
35 the measured concentration is too large.
Using a bottle in a vertical position, the inaccuracy of the measured sand concentration may be as large as
50%. In field conditions the inaccuracy of the measured sand concentration may even be larger when the
filling time is small compared with the characteristic time scale of the fluctuating concentrations.
The inaccuracy of the measured silt concentrations may be about 10%, which is indicated by field
measurements (see Paragraph 5.4.5).
dimensions: bottle 0.5 to 2 litre
weight: 25 to 50 kg (incl. container)
measuring range: > 1 mg/l
Advantages
1. simple and reliable instrument for silt and sand particles
2. easy to handle (no electricity)
3. usable in rough weather conditions
4. applicable in wave conditions
Disadvantages
1. many samples required; large sampling effort (time consuming)
2. short sampling period
3. small sediment catch (size analysis)
4. not for sampling close to bed
5. inaccurate sampling of sand particles
References
Delft Hydraulics, 1980. Investigation Vlissingen Bottle (in Dutch). M1710. Delft, The Netherlands
Hayes, F.Ch., 1978. Guidance for Hydrographic and Hydrometric Surveys. Delft Hydraulics Laboratory,
Publication No. 200, The Netherlands
Figure 1
Bottle sampler
Figure 2
Efficiency Bottle sampler
Figure 3
Measuring sheet bottle/trap sampler
5.6.2.3 Trap sampler
A. Instantaneous trap sampler
Principle
The instantaneous trap sampler consists of a horizontal cylinder equipped with end valves which can be
closed suddenly (by a messenger system) to trap a sample instantaneously, as shown in Figure 1. The water
is allowed to flow through the horizontal cylinder while the sampler is lowered to the desired point.
Practical operation
1. Lower the trap to the measuring position (echo sounder attached to trap)
2. Close valves by releasing the messenger
3. Raise the trap sampler
4. Transfer water-sediment sample to a bottle (see Fig. 1)
5. Note data on measuring sheet (see Paragraph 5.6.2.2.)
Laboratory analysis
See Paragraphs 5.6.2.2. and 8.1

c=Gs/V
in which:
As the trap sampler yields an instantaneously measured concentration, many samples are necessary to obtain
a statistically reliable average value.
dimensions: 0.6x0.1x0.1 m
Advantages
Disadvantages
B. Time-integrating trap sampler
The trap sampler used in lakes and oceans consists of a cylinder-shaped or funnel-shaped box with a closed
bottom and placed vertically in the water column, see Fig. 2. These traps are used to determine:
the deposition rate at a certain elevation,
the sediment concentration at a certain elevation.
In still water the trapping efficiency is unity. In flowing water the trapping efficiency depends on many
factors: turbulence generated inside trap, trap geometry, concentration and fall velocity of particles.
Cylinders were found to give the most accurate results. The trapping efficiency was found to be 0.8 to 1.2 for
mud suspensions in current velocities of 0.05 to 0.1 m/s (Bloesch and Burns, 1980; Gardner, 1980) . The
ratio of the height and diameter of the trap should be in the range of 3 to 5. Funnel-shaped traps were found
to give an underestimation of the deposition rate in flowing water. Traps with a small mouth and a wide body
(bottle) were found to give a large overestimation of the deposition rate.
Antsyferov et al (1990) used tripods and poles equipped with traps to measure the sediment concentrations
in coastal conditions (surfzone). The traps consisted of cylinder-shaped boxes (height = 0.1 m, diameter =
0.075 m) with six openings of 7.5 mm at the upper part of the boxes, see Fig. 2. The traps were set into
operation (by divers) by aligning corresponding openings in the trap body and in the cover.
The concentration of size fraction i at height z is given as:
ci(z)= Mi/(ke F DT Ut)
in which:
Mi = sediment mass of fraction i in trap,
F = area of intake openings projected normal to wave direction,
DT = sampling period,
Ut = time-averaged value of the absolute horizontal orbital velocity at height z,
ke = trapping coefficient.
The ke-value was obtained from field and laboratory calibrations. Concentrations were measured by means of
a pump sampler close to the sand traps. Velocity measurements were also performed close to the trap (0.5
to 0.8 m/s). Antsyferov et al (1990) found:
ke = 0.26 for field conditions,
ke = 0.21 for laboratory conditions.
The ke-values are only valid for steady wave conditions (during period DT). In storm conditions with growth,
stabilization and decay of the waves, the ke-value is different. It was found that 70% of the sediment mass
was trapped during the growth and decay phase of the waves (inside breaker zone).
Katori (1983) and Kraus (1987) used portable streamer traps to measure transport rates in the surf zone, see
Fig. 3. The traps consist of long rectangular bags of polyester sieve cloth material (100 mm) vertically
mounted on a stainless steel rack (Kraus, 1987). An operator standing downcurrent attends the trap during a
sampling interval of 10 min. The use of these traps is restricted to shallow water (1 m) with wave heights less
than about 0.5 m.
Katori (1983) used a similar trap to measure cross-shore transport at the bed. The trap was mounted on a
rubber mat resting on the bed (to prevent scour), see Fig. 3.
Advantages of streamer traps are:

absolute measurement of transport rate,
short sampling period of 10 min,
vertical distribution can be measured,
simultaneous deployment at many locations,
simple, robust and cheap.
Disadvantages are:
disturbances of flow fields,
scour around the traps,
many operators involved,
analysis of many samples,
restricted to shallow water.
References
Antsyferov, S.M., Belberov, Z.K. and Massel, S., 1990. Dynamical Processes in Coastal Regions.
Publishing House, Bulgarian Academy of Sciences
Bloesch, J. and Burns, N.M., 1980. A Critical Review of Sedimentation Trap Technique. Schweiz. Z.
Hydrol. 42, No. 1, p. 15-55
Gardner, W.D., 1980. Sediment Trap Dynamics and Calibration. Journal of Marine Research, Vol. 38, No.
1, p. 16-39
Gardner, W.D., 1980. Field Assessment of Sediment Traps. Journal of Marine Research, Vol. 38, No. 1, p.
41-52
Katori, S., 1983. Measurement of Sediment Transport by Streamer Trap (in Japanese). Report No. 17, TR-
Kraus, N.C., 1987. Application of Portable Traps for Obtaining Point Measurements of Sediment Transport
Rates in the Surf Zones. Journal of Coastal Research, Vol. 3, No. 2
Figure 1
Trap sampler
Figure 2
Trap sampler
Figure 3
Trap sampler
5.6.2.4 USP-61 point-integrating sampler
Principle
The sampler consists of a streamlined bronze casting (= 50 kg), which encloses a small bottle (= 500 ml), as
shown in Figure 1A. The sampler head is hinged to provide access to the bottle. The intake nozzle, which can
be opened or closed by means of an electrically operated valve, points directly into the approaching flow.
To eliminate a sudden inrush after opening of the intake nozzle, the air pressure in the bottle is balanced with
the hydrostatic pressure prior to opening of the valve. This is accomplished by means of an air bell in a body
cavity connecting the bottle and the surrounding stream. After opening of the valve, the air in the bottle can
escape through a special air-exhaust tube pointing downstream on the side of the sampler head. As a result
the hydraulic coefficient is approximately unity during sampling. The filling time varies from 10 to 30
seconds, as shown in Figure 1B. To avoid a circulation flow, the bottle should only be filled for about 75%.
The USP-63 is a heavier version (= 100 kg) of the USP-61.
The sampler is manufactured by Rickly Hydrological Company (www.rickly.com).
Practical operation
1. lower instrument to sampling position (echo sounder attached to sampler)
2. open (electrical) valve
3. wait till bottle is filled for about 75% (Figure 1B)
4. close (electrical) valve
5. hoist instrument (slowly) and replace bottle.
6. note data om measuring sheet (see Paragraph 5.6.2.2)
Laboratory analysis

c=Gs/V
in which:
The silt and sand transport (in kg/s/m2) can be determined as:
S=Gs/(F T)
in which:
F = area of nozzle (m2)
The sampling efficiency of the USP-61 is strongly dependent on the ratio of the intake velocity and the local
flow velocity (hydraulic coefficient).
Extensive laboratory measurements, summarized by Dijkman (1978), have shown that the hydraulic
coefficient varies from about 0.8 to 1.3 depending on the water temperature, sample height above the bed and
the nozzle orientation (maximum deviation with flow direction of 20). For this range a maximum sampling
error in the concentration of about 10% may be expected in case of a steady concentration (see Figure 2, Par
5.6.2.2). In the case of field conditions with fluctuating concentrations the inaccuracy of individual
samples may be as large as 50%. To obtain a reliable average value in a statistical sense, a large number of
samples (say 10) should be collected at each sampling point.
nozzle diameter: 0.005 m
weight: 50 kg (USP-61) and 100 kg (USP-63)
energy: 24 volt (DC)
Advantages
3. direct determination of transport
Disadvantages
4. not for sampling close to bed,
5. fragile intake nozzle
References
Dijkman, J., 1978. Some Characteristics of the USP-61 and Delft Bottle. Delft University of Technology,
Dep. of Civ. Eng., Int. Report No. 5- 78, The Netherlands
Laboratory and Jaroslav Cerni Institute, Report S410, The Netherlands
FEDERAL INTER-AGENCY SEDIMENTATION PROJECT. Instructions for USP-61, Suspended
Sediment Sampler Minnesota 5541K, U.S.A.
Figure 1
USP 61 sampler
5.6.2.5 Delft Bottle sampler
Principle
The Delft Bottle (Figures 1 and 2) is based on the flow-through principle, which means that the water
entering the intake nozzle leaves the bottle at the backside. As a result of a strong reduction of the flow
velocity due to the bottle geometry, the sand particles larger than about 100 ym settle inside the bottle. Using
this instrument, the local average sand transport is measured directly.
Practical operation
The Delft Bottle (DB) can be used as follows:
1. suspended at a wire using a straight nozzle,
2. suspended in a frame resting on the bottom using a bended or straight nozzle for measurements close
to the bed.
Depending on the flow conditions, a small nozzle (velocities < 1 m/s) with an internal diameter of 15.5mm
or a big nozzle with an internal diameter of 22 mm can be used.
The operational procedure is as follows:

1. lower DB into the flow,
2. let the air in the bottle escape through the nozzle and/or openings in the backside,
3. lower DB to the sampling position (note lowering time), use echo sounder to determine exact position,
4. start sampling period (note sampling time),
5. raise DB slowly (note raising time),
6. remove water from DB (gently),
7. transfer sediment catch from DB to a calibrated (funnel) glass and read volume,
8. transfer sediment catch to bottle for laboratory analysis (if necessary),
9. note data on measuring sheet.
Laboratory analysis
As the sediment sample only consists of sand particles (fine silt particles are washed through the DB) the
laboratory analysis is rather simple:
1. decant surplus water from sampler container,
2. determine dry mass of sand particles (if necessary),
3. determine size distribution of sand particles by sieving or settling tests,
4. determine calibration factor (see Figure 3).
Remarks
As the overall efficiency of the Delft Bottle is rather low, it is sufficiently accurate to determine only the
(immersed) volume of the sand catch on board of the vessel. A small number of the samples can be returned
to the laboratory to determine the porosity factor of the sediment sample and the particle size distribution.

The local average sediment transport (in kg/m2/s) is determined as:
S= a (1-p) rs Vs/(F T) or S=a Gs/(F T)
in which:
a = calibration factor according to Figure 3,
p = porosity factor,
rs = density of sediment (2650 kg/m3),
Vs = volume of sediment sample, including pores (m3),
F = area of nozzle (m2),
Sampling errors are introduced by:
1. incorrect intake velocity compared with local flow velocity; the hydraulic coefficient (ratio
of intake velocity and local flow velocity) varies from 1 to 1.5 (Dijkman, 1978, 1981),
2. inefficiency of the sampler to collect relatively fine sediment material (particles finer than 100 mm),
3. additional sampling during raising and lowering of the instrument,
4. sediment losses during removal of the sand catch from the DB.
Usually, only the first two errors are corrected using a calibration factor a according to Figure 3, which is
based on extensive laboratory measurements (Dijkman, 1981). The a-factor varies from 0.7 to 2.5
depending on the nozzle type, particle size and local flow velocity. The sampling error due to the collection
of sediment particles during lowering and raising of the instrument can be reduced by using a relatively large
sampling period (15 minutes). Otherwise, an additional calibration factor is necessary (Dijkman, 1981). The
minimum sampling time is about 5 minutes to obtain a statistically reliable result. An additional advantage of
a long sampling period is the collection of a large sediment catch enabling an accurate determination of
particle size (by sieving).
Field measurements show sampling errors up to 50% for individual samples, even after the application
of the calibration factor (Paragraph 5.4.1). Considering these large errors, the Delft Bottle can only be used
to obtain a rough estimate of the local sand transport. Therefore, it is sufficiently accurate to determine only
the volumetric quantity of the sand sample (in-situ). In that case the laboratory analysis is rather limited,
which is an advantage of the Delft Bottle method. The Delft Bottle should not be used in tidal flow
conditions with relatively small sediment concentrations because of the long sampling period which is
required to obtain a measurable sediment catch.
nozzle diameter: 0.0155 or 0.022 m (straight or bended)
dimensions: bottle 1.15x0.2x0.2 m; frame 1x1x0.85 m; funnel 0.5x0.2x0.2 m
weight: bottle 20 kg; frame 66 kg; funnel 3 kg
cycle period: 15-25 min (minimum period between two measurements)
Advantages
1. simple and reliable instrument for sand particles (>100 mm)
2. long sampling period
3. large sand catch (size analysis)
5. no electricity
Disadvantages
1. not for particles < 100 mm
2. application of calibration factors
4. not for tidal conditions (large cycle time)
References
Dep. of Civ.Eng., Int. Report No. 5-78, The Netherlands
Dijkman, J., 1981. Investigation of Characteristic Parameters of Delft Bottle. Delft Hydraulics Laboratory,
Report S362, The Netherlands
Dijkman, J. and Milisic, V., 1982. Investigations on Suspended Sediment Samplers. Delft Hydraulics
Figure 1
Delft Bottle
Figure 2
Delft Bottle
Figure 3
Calibration factors of Delft Bottle sampler
Figure 4
Measuring sheet of Delft Bottle sampler
5.6.2.6 USD-49 depth-integrating sampler
Principle
The USD-49 is a depth integrating sampler. The sampler is lowered at a uniform rate from the water surface
to the streambed, instantly reversed, and then raised again to the water surface. The sampler continues to take
its sample throughout the time of submergence. At least one sample should be taken at each vertical selected
in the cross-section of the stream. A clean bottle is used for each sample. The USD-49 sampler has a cast
bronze streamlined body in which a round or square pint-bottle sample container is enclosed. The head of the
sampler is hinged to permit access to the sample container (see Figure 1). The head of the sampler is drilled
and tapped to receive the -inch, 3/16-inch or 1/8-inch intake nozzle which points into the current for
collecting the sample. The transit rate depends on the mean velocity in the vertical, the water depth and the
nozzle diameter, as shown in Figure 2. The USD-49 is suitable for depth integration of streams less than
about 5 m in which the velocities do not exceed 2 m/s.
The sampler is manufactured by Rickly Hydrological company (www.rickly.com).
Practical operations
1. determine the depth to be sampled and divide twice the depth (round-trip integration) by the sampling
time to obtain the required rate for lowering and raising,
2. lower and raise the sampler at the selected transit rate (use stopwatch for time control),
3. replace bottle or container on board of vessel,
4. note the lowering and raising times,
5. observe sample volumes and transit rates carefully, so the transit rate can be adjusted (if necessary).

The depth-averaged concentration can be determined as
cmean=Gs/V
in which:
The depth-integrated suspended sediment transport (in kg/m/s) can be determined as:
Ss= Gs h//(F T) or as Ss=cmean umean h= (Gs/V) umean h
in which:
h/ = depth of sampled zone (m),
umean = depth-averaged velocity (m/s).
The sampler cannot sample down to the stream bed surface. When the sampler touches the bed, the distance
between the sample nozzle and the bed is about 0.1 m (see Figure 1). Thus, the depth of the sampled zone is
about equal to the water depth minus 0.1 m (h/ = h - 0.1 in m). Another problem is the short sampling period
at each specific point in the vertical. As a result concentration fluctuations are not averaged out and repeat
samples are necessary. Information of the vertical concentration distribution cannot be obtained.
nozzle diameter: 0.006, 0.0045, 0.003 m (1/4, 3/16 or 1/8 inch)
weight: 28 kg
Advantages
Disadvantages
1. for depth smaller than 5 m
2. small sediment catch
3. many samples required
5. fragile intake nozzles
References
INTER-AGENCY COMMITTEE on Water Resources, 1963. Determination of Fluvial Sediment

Discharge Report no. 14, St. Anthony Falls Hydr. Lab., Minneapolis, USA
Figure 1
Depth-integrating sampler USD-49
Figure 2
Allowable ratio of uniform transit rate RT to mean velocity Vm for depth-integration using litre milk bottle
5.6.2.7 Collapsible-Bag depth-integrating sampler
Principle
The Collapsible-Bag sampler is based on the principle that the static pressure acting on the outside surface of
the flexible bag (devoid of air) creates at the nozzle exit a pressure equal to the hydrostatic pressure at the
nozzle entrance. Using this method, samples can be collected throughout any depth.
The sampler consists of a wide-mouth, perforated, rigid plastic container enclosed in a cage-like metal frame.
The head of the frame supports a plastic intake nozzle (6 or 13 mm) and swings open to permit the plastic
container to be removed. When the head is closed, the end of the nozzle extends slightly into the mouth of
the container. Perforations in the container allows the air in the container to escape during submergence. For
sampling, a collapsed flexible plastic bag is placed inside the rigid container. The neck of the flexible bag is
stretched over the neck of the rigid container and this unit is placed into the sampler. When the sampler is
lowered into the flow, water enters the perforations in the rigid container, surrounds the collapsed bag and
equalizes the hydrostatic pressure at the nozzle entrance. The velocity head forces water through the intake
nozzle into the collapsed bag, which unfolds to conform to the rigid container. This sampling action
eliminates any possibility for flow to rush in due to unequal pressures and insures that water-sediment
mixture is collected at stream velocity regardless of velocity distribution and depth, provided that the sample
container does not overfill and the vertical transit rate is smaller than 0.4 of the depth-averaged velocity.
Nordin et al. have used a large-volume (6 1) bag sampler in combination with an Ott-type current meter in
water depths up to 30 m. The rotation of the propeller was transmitted into a rate meter. This latter provided
a direct read out of velocity. By lowering and raising the current meter over the depth, the instantaneous
velocity profile can be recorded.
The constant vertical transit rate of the bag sampler should be fast enough so that the sample container is
not completely filled, but not greater than 0.4 v (v= depth-averaged velocity). Nordin et al. tried to keep the
transit rate smaller than 0.2 v . The sample volumes that will be collected for a transit rate equal to 0.2 v are
given in Figure 1 A. The mean velocity v can be estimated from the surface velocity. The time required to
collect one liter of sample for various mean velocities and nozzle diameters is given in Figure 1B. If the
cross-channel distribution of sediment discharge is needed, each sample is processed separately. If the depth
and velocity are fairly uniform across the section, it often is possible to use one nozzle diameter and transit
rate at all verticals and the individual samples can be composited into a single discharge-weighted sample.
For most purposes 10 to 15 depth-integrated samples equally spaced across the section are adequate. If the
maximum and minimum surface velocities are observed when the cross-section is sounded initially, that
information can be used with the data plotted in Figure 1B to select the transit rate. Usually, it is accurate
enough to select a transit rate smaller than 0.2 v using the velocity v of the deepest en fastest vertical.
Assuming that v does not vary more than a factor 2, the transit rate will not be greater than 0.4 v in the
vertical with the lowest velocity.
Practical operation
See Paragraph 5.6.2.6.
Laboratory analysis
See Paragraph 5.6.2.2 and 8.1

See Paragraph 5.6.2.6
nozzle diameter: 0.006 or 0.013 m (1/4, 1/2 inch)
weight: 45 kg
Advantages
3. also for deep flows
Disadvantages
1. handling of large sample containers
3. no information of vertical distribution
References
Christiansen, H., 1985. Suspended Sediment Measurements in Elbe Estuary at Hamburg using a Cux
Sampler. Euromech 192, Neubiberg, Germany
Nordin, C.F., Cranston, C.C., Abel, M.B., 1983. New Technology for Measuring Water and Suspended
Sediment Discharge of Large Rivers.
Stevens, H.H., Lutz, G.A. and Hubbel, D.W., 1980. Collapsible-Bag Suspended Sediment Sampler.
Journal of Hydraulic Division, ASCE, Vol. 106, No. Hy4
Figure 1
Collapsible Bag depth-integrating sampler
5.6.3.1 General aspects for sampling in unidirectional flow
To obtain a reliable average sediment concentration, the sampling or measuring period should be rather large
(about 300 seconds). Furthermore, the collection of a large sediment sample for size-determination by
sieving or settling tests requires the sampling of a relatively large water volume (about 25 to 50 litres).
Both requirements can be satisfied by collecting water samples by means of a pump in combination with an
in-situ separation of water and sediment particles.
Usually a pump sampler consists of a submergible carrier (with intake nozzle, current meter and echo-
sounder), a deck-mounted pump and a flexible hose connecting the intake nozzle and the pump. The hose
diameter should be as small as possible to reduce the stream drag on the hose. Using a hose diameter (bore)
in the range of 0.003 to 0.016 m, the pump discharge will be in the range of 1 to 30 litres per minute. In case
a deck-mounted pump is used the maximum suction lift will be about 7 m. Assuming a static lift (= height of
pump above water level) of about 2 m, the suction lift available for operation of the pump will be about 5 m
resulting in a maximum hose length of about 50 m (Van Rijn, 1979). In extreme deep waters an underwater
pump must be used. Operation of a pump sampler is limited to flow conditions with velocities smaller than 2
m/s because of excessive stream drag on the pumphose and carrier.
Type of pump
Peristaltic pumps or propeller type pumps can be used.
Peristaltic pumps (24 volt/220 volt) (see www.globalw.com) have proven to be very efficient for pump
sampling in river and coastal conditions. The discharge is relatively small (1 1/min) yielding a relatively
small water-sediment sampling which can be handled easily. The hose diameter is extremely small (0.006
m), which reduces the fluid dragforces on the hose. The pump direction can be easily changed to remove
small objects (shell fragments, organic materials, etc.) blocking the intake nozzle. The intake velocity in
relation to the static suction lift (= height of pump above water surface) and the hose length is given in Fig. 3.
Propeller type pumps (common garden pumps; www.metabo.cm/eb/com/en/produkte/gardenpumps) produce
a relatively large discharge (10 1/min) resulting in the handling of a large water-sediment volume. The hose
diameter is in the range of 0.01 to 0.016 m.
Intake velocity and accuracy

Ideally, the intake velocity of the water-sediment sample should be equal to the local flow velocity (see
Paragraph 5.6.2.1). In conditions with varying velocities this would mean a continuous adjustment of the
intake velocity. For practical reasons it is preferable to operate the pump system as much as possible with a
fixed discharge and hence a fixed intake velocity. This can be done by using a fixed intake velocity for each
class of flow velocities (see following table).
Using these values, the hydraulic coefficient will be in the range of 0.8 to 2.0 during pump sampling,
resulting in a maximum error in the concentration of about 20%, which is quite acceptable for concentration
measurements (see Paragraph 5.6.2.1). For reasons of effective sampling, the flow velocity in the hose must
be larger than 1.0 m/s (Van Rijn, 1979).
Local flow Intake Pump discharge

velocity velocity (l/min)
Intake nozzle Intake nozzle Intake nozzle

(m/s) (m/s) diameter=0.003 m diameter=0.01 m diameter=0.016 m
0.5 - 1.0 1.0 0.45 5.0 12
1.0 - 1.5 1.25 0.55 6.25 15
1.5 - 2.0 1.75 0.80 8.75 21
2.0 - 2.5 2.25 1.0 11.25 27

Determination of hose dimensions and carrier type
The maximum hose length is about 50 m using a deck-mounted pump, which enables sampling in channels
with flow depths upto 25 m. In shallow water the hose dimensions and also the carrier dimensions can be
reduced substantially.
Figure 1 can be used to select the required hose dimensions (bore diameter and length) for a propeller type
pump and carrier type, given specific flow conditions. Figure 2 shows an example of an under-water carrier,
as used in tidal waters in the Netherlands.
The carrier should satisfy the following specifications:

1. streamlined body,
2. small cross-section area,
3. low centre of gravity,
4. suspension cable attached to swivel,
5. large tail fin,
6. supporting pins (at bottom side) and hand grips for transportation,
7. adjustable ballast weight (lead).
Determination of sampling position

To determine the actual position of the intake nozzle above the bed, an echo-sounder attached to the carrier
should be used. Usually it is desirable to have a sampling position close to the bed. This can be achieved by
placing the carrier on the bed. In that case the position of the intake nozzle is equal to the distance between
the nozzle and the underside of the carrier (see Figure 1).
When measurements at a few centimetres above the bed are necessary, special equipment should be used
which allows the vertical adjustment of the intake nozzle over a certain range (remote controlled) or
additional intake nozzles should be attached to the carrier.
Handling of water volume after sampling

When a propeller type pump is used, a large water-sediment volume is obtained. The handling of a large
water-sediment volume requires the in-situ separation of water and sediment particles. A practical solution
can be obtained by using the filtration method or the sedimentation method. In case the collection of a large
quantity of sediment particles for size analysis is not of importance, it is advisable to take a relatively small,
but representative sample of the total water sample to determine the sediment concentration. This latter
method is discussed as the pump-bottle method (Paragraph 5.6.3.5) .
When a peristaltic pump is used, a relatively small amount of water is obtained (2 liters in 5 minutes) . This
small sample can be stored in a bottle and returned to the laboratory for analysis.
Automatic pump samplers

Site location, flow conditions, frequency of collection and operational costs sometimes make collection of
sediment data by manual methods impractical. For these reasons automatic pumping type samplers have
been developed (FTS: see www.ftsinc.com); (www.ismatec.com).
Such a sampler consists of:

1. intake nozzle and tubing system,
2. pump to draw water-sediment samples from the flow (flushing of tubing system after sampling is
necessary),
3. sample container unit to hold sample bottles in position for filling,
4. sample distribution system to divert a pumped sample to the correct bottle,
5. activation system that starts and stops the sampling cycle.
Propeller garden pump

weight: 5 kg
discharge: 0 to 30 l/min
energy: 220 volt
hose length and diameter: 10 to 50 m; 0.01 and 0.016 m
Peristaltic pump
type: Ismatec (www.ismatec.com)
weight: 5 kg
discharge: 0.1 to 0.5 l/min
energy: 220 volt or 24 volt
hose length and diameter: 5 to 50 m; 0.004 m
References
Conference on Instrumentating Oceanography, I.E.R.E., Proc. No. Bangor, England
Rijn, L.C. van, 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Report S404, The Netherlands
Figure 1
Under-water carrier
Figure 2
Under-water carrier for pump samplers
Figure 3
Hose length and intake velocity of peristaltic pump
5.6.3.2 General aspects for sampling in oscillatory flow
Pump sampling is an attractive method for concentration measurements in coastal conditions because a
relatively long sampling period can be used which is of essential importance to obtain a reliable time-
averaged value. The sampling period should be rather long (15 min) in irregular wave conditions (at least
100 waves).
A problem of sampling in conditions with irregular waves is that the magnitude and direction of the fluid
velocity is changing continuously. This complicates the principle of isokinetic sampling in the flow
direction. A workable alternative may be the method of normal (or transverse) sampling, which means that
the intake nozzle of the sampler is situated normal to the plane of fluid velocity.
Bosman et al (1987) studied the sampling error related to the orientation of the intake nozzle, because they
were interested in pump sampling under wave conditions. They found that a transverse pumping direction
yields good results. The intake nozzle is directed downward or normal to the plane of orbital motion. Figure
1 shows the ratio c/co as a function of the ratio u/uo and the nozzle orientation for 170 mm-sediment (c =
measured concentration, co = original concentration, u = intake velocity, uo = local ambient velocity). For
transverse orientation (90) and ratio of u/uo> 2, the c/co-ratio is about a=0.7 to 0.8 which means a
systematic error of 20% to 30% in the measured concentration. Similar results were obtained for 220 mm,
280 mm, 360 mm and 450 mm-sediment. This systematic error can be eliminated by multiplying the measured
concentrations with a factor l/a.
Peristaltic pumps have proven to be very efficient in coastal conditions.
References
Antsyferov, S.M., Basinski, T. and Pykhov, N.V., 1983. Measurements of Coastal Suspended Sediment
Concentrations. Coastal Engineering, 7, Elsevier Publishers, Amsterdam, The Netherlands
Bosman, J.J., Velden, E.T.J.M., van der and Hulsbergen, C.H., 1987. Sediment Concentration
Measurement by Transverse Suction. Coastal Engineering, Vol. 11, p. 353-370
Figure 1
Directional sensitivity of 3 mm nozzle for 170 mm sediment
5.6.3.3 Pump-Filter sampler
Principle
The water-sediment sample is pumped through a filter which separates all particles larger than the mesh size
of the applied filter material. The method is shown schematically in Figure 1A. To separate the sand fraction,
nylon filter material with a mesh size of 50 mm can be used. The water volume is recorded by means of a
(simple) volume meter. After taking a sample, the filter system is opened and the filter material with the sand
catch is removed and returned to the laboratory for drying, weighing and size analysis. During removal of the
filter, the pumping is continued using a bypass system. The filtration method cannot be used in a silty
environment with silt concentrations larger than about 50 mg/1 because of rapid filter blocking by the fine
silt particles.
Practical operation
1. lower intake nozzle to measuring position (use echo-sounder),
2. open filter house and install 50 ym- filter,
3. close filter house (note initial reading of volume meter),
4. adjust intake velocity depending on measured flow velocity (discharge meter and valve A, see Figure
1A),
5. wait one minute to flush the pump hose (through bypass system),
6. pump through filter by using valve B (if filter is blocked, indicated by discharge meter, stop
sampling),
7. use sampling period of 5 minutes,
8. pump through bypass system by using valve B (note final reading of volume meter),
9. open filter house and remove filter with sediment catch (wash inside of filter house free from
sediment particles before removal of filter),
10. install new filter,
11. note data on measuring sheet (Figure 3).
Remarks
1. the filter system may never be used without filter material to prevent the volume meter from
damage,
2. cavitation (noise!) due to low water pressure can be eliminated by closing valve A somewhat till the
cavitation noise has disappeared.
Laboratory analysis
Simple
1. wash filter and sediment sample with fresh water to remove silt and salt particles,
2. dry and weigh filter with sand particles,
3. remove sand particles from filter material by brushing,
4. weigh filter material and determine dry mass of sand particles,
5. determine size distribution of sand particles by sieving (or settling tests).
Detailed
1. wash sediment sample off the filter material into a beaker,
2. decant through a 50 mm- sieve to remove silt particles,
3. wash sand sample with distilled water to remove silt and salt articles,
4. wash sand sample in a small beaker,
5. dry and weigh sand sample,
6. determine size distribution of sand sample.
The sand concentrations are determined as:
c=Gs/V
in which:
To determine the overall accuracy of the pump-filter method, some concentration measurements were carried
out in a laboratory flume using a siphon sampler for comparison. The siphon sampler consisted of a short
hose connected to an intake nozzle in the flume. The intake nozzles of both methods were installed next to
each other at a fixed height above the flume bottom. In lateral direction an uniform concentration was
assumed. To simulate field conditions as much as possible, the pump-filter system was operated with a static
lift of 2.0 m and a hose diameter (bore) of 0.016 m and a hose length of 50 m. Filter material with a mesh
size of 50 mm is used for separation of the sand fraction. The hydraulic coefficient of the pump sampler was
varied in the range 0.7 to 1. Figure 1B shows the sampling error in the concentration for two types of sand.
The concentration range is 50 to 1000 mg/1. Each point represents an average value, while also the highest
and lowest value are indicated. The maximum overall error for all measurements is about 10% (Van Rijn,
1979).
Field measurements (see Paragraph 5.4) indicate a sampling error in the concentration of about 20%
weight: 50 to 100 kg
Advantages
1. simple and reliable instrument for sand particles
2. simple laboratory analysis
3. large sample size (for size analysis)
4. usable in wave conditions
Disadvantages
1. not for silty materials
2. electricity and pump required
3. winch facility required
5. fragile mechanical parts
References
Van Rijn, L.C., 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Report S4CW, The Netherlands
Figure 1
Pump Filter sampler
Figure 2A
Pump Filter sampler; filter with sediment catch
Figure 2B
Pump Filter sampler; filtration unit
Figure 2C
Pump Filter sampler
Figure 3
Measuring Sheet Pump Filter sampler
5.6.3.4 Pump-Sedimentation sampler
Principle
The method is based on the filling of a large calibrated container (= 50 liters), in which the sand particles can
settle (bottle 1), as shown schematically in Figure 1A. Using a settling height of about 0.75 m, the sand
particles larger than 50 a 60 mm can be separated in about 5 minutes. A high separation efficiency can be
obtained by using a conical container and a vibrator to avoid settlement of the sand particles on the inside of
the container (Van Rijn, 1980). To determine the silt concentration (particles smaller than 50 mm), a small
water sample (bottle 2) can be tapped during emptying of the container.
Practical operation
2. install bottle 1 and 2 (Fig. 1A ),
3. bring bottle 1 under opening of container (slide valve C to position that bottle 1 can be filled),
4. adjust intake velocity depending on measured flow velocity (discharge meter and valve A),
5. wait one minute to flush the pump hose (through bypass-system),
6. fill container (valve B) in about 5 minutes,
7. take a settling period of 3 minutes without using vibrator,
8. start vibrator to remove sand from inside of container (vibration period of about 2 minutes),
9. note water volume (gauge glass),
10. stop vibrator,
11. empty container (slide valve C to position that filling of bottle 1 is blocked),
12. take a water-silt sample (bottle 2), use a filling time equal to emptying time of container (tap D),
13. remove bottle 1 and 2,
Remarks
1. a calibrated measuring glass can be used for the volumetric determination of the sand sample (bottle 1),
2. the inside of the container must be cleaned regularly.
Laboratory analysis
Bottle 1
1. wash sand sample over 50 mm-sieve, use distilled water to remove silt and salt particles,
2. wash sand sample into a small beaker,
4. determine size-distribution of sand sample (sieving or settling test).
Bottle 2
See Paragraphs 5.6.2.2 and 8.1.
csilt=Gsilt 2/Vbottle 2
csand=Gsand 1/Vcontainer + Gsand 2/Vbottle 2
in which:
Gsilt 2 = dry mass of silt sample from bottle 2 (mg)
Vbottle 2 = volume of water sample in bottle 2 (1)
Gsand 1 = dry mass of sand sample in bottle 1 (mg)
Vcontainer = volume of water sample in container (1)
Laboratory measurements were carried out to determine the sampling efficiency. A siphon sampler was used
to determine the original concentrations near the intake nozzle. The maximum sampling error was found to
be about 15% for sand concentrations larger than about 50 mg/1 (see Figure 1B). Probably, a minor part of
the sand particles settles on the inside of the container and is washed off during emptying of the container
(Van Rijn, 1980).
Field measurements in a tidal estuary (Western Scheldt, The Netherlands) have shown the presence of a
small amount of sand particles in bottle 2. Figure 3 presents the ratio of the sand concentration from bottle 2
and 1 as a function of the sand concentration from bottle 1. For sand concentrations larger than about 50
mg/1, the sand concentration from bottle 2 is less than 20% of the sand concentration from bottle 1 and may,
therefore, be neglected resulting in a considerable reduction of the laboratory analysis. Size analysis of
sediment samples collected in field conditions shows that nearly all (sand) particles larger than MO pm do
settle in bottle 1 during the sedimentation period of about 5 minutes (Van Rijn, 1980). Field measurements
have also shown that the silt concentration (bottle 2) can be determined with an inaccuracy of about 10%
compared with a simple bottle sampler (Paragraph 5.4.5).
measuring range: > 50 mg/l for sand and > 10 mg/l for silt
energy: 24 volt (vibrator)
Advantages
2. in-situ separation of sand and silt
3. relatively large sampling period (5 minutes)
Disadvantages
1. winch facility
3. relatively large cycle period
4. many bottles for laboratory analysis
References
Van Rijn, L.C., 1980. Methods for in-situ Separation of Water and Sediment. Delft Hydraulics Laboratory,
Report SH04 II, The Netherlands
Figure 1
Pump Sedimentation sampler
Figure 2
Sedimentation container
Figure 3
Efficiency of Pump-Sedimentation sampler
Figure 4
Measuring Sheet Pump Sedimentation sampler
5.6.3.5 Pump-Bottle sampler
Principle
This simple method is based on the continuous pumping (propeller type pump) of a water-sediment mixture.
On board of the survey vessel a small part of the pump discharge is used to fill a 1 liter-bottle or 2 liter-bottle
in 3 to 5 minutes by using a small siphon tube (Fig. 1A). Using this method, a relatively long sampling
period and hence a (statistically) reliable concentration measurement can be obtained.
When a peristaltic pump is used (discharge = 0.5-1 1/min), the bottle can be filled directly.
An optical sensor can be used to determine the silt concentration in the bottle after settling of the sand
particles.
Practical operation
1. lower intake nozzle to sampling position (use echo sounder),
2. adjust intake velocity (discharge meter and valve A),
3. wait one minute to flush the pump hose,
4. fill bottle by opening tap B (2 liters in 4 minutes),
5. remove bottle,
Laboratory analysis

The silt and sand concentration can be determined as:
csilt=Gsilt/V and csand=Gsand/V
in which:
To determine the sampling efficiency of the pump-bottle method, laboratory measurements were carried out
in a flume using sand with D10 = 150 mm, D50= 220 mm and D90= 330 mm. A siphon sampler was used to
determine the actual concentration (co) at the position of the intake nozzle. The sand concentration in the
flume was varied from 30 to 1700 mg/1. The intake velocity of the pump-bottle system was equal to the local
flow velocity in the flume (iso-kinetic sampling). The tap discharge Qs was varied from 0.2 to 2 liters per
minute. Figure 1B presents the average error in the concentration as a function of the discharge Qs showing
an average error smaller than 20% for a discharge Qs in the range 0.2 to 1 1/min.
For each concentration the largest and smallest deviation are also indicated.
Figure 1B shows a trend from a positive error for a small discharge to a negative error for a large discharge,
which can be explained by means of the hydraulic coefficient of the bottle filling process. For Qs = 2 l/min
the ratio of the tap velocity and pump velocity (= hydraulic coefficient) is 1.5 resulting in a negative
sampling error (see also Paragraph 5.6.2.1). For Qs smaller than 1 1/min, the hydraulic coefficient is smaller
than 1 resulting in a positive sampling error. Optimal sampling requires a discharge of about 0.5 1/min (2
liter bottle in 1 minutes). Grain-size analysis of the sediment particles collected in the bottle showed a D50 =
200 mm which is about 10% smaller as the original sediment (D50 = 220 mm).
dimensions: length of 0.25 m; hose diameter 0.016 m; siphon diameter of 0.004 m
weight: 1 kg
Advantages
2. relatively large sampling period (3 to 5 minutes)
3. small cycle period (5 minutes)
Disadvantages
3. small sediment samples
Figure 1
Pump Bottle sampler
Figure 2
Measuring Sheet Pump Bottle sampler
5.6.4.1 General principles
Optical and acoustical sampling methods enable the continuous and contactless measurement of sediment
concentrations, which is an important advantage compared to the mechanical sampling methods. Although
based on different physical phenomena, optical and acoustical sampling methods are very similar in a
macroscopic sense. For both methods the measuring principles can be classified in (see Figure 1):
transmission,
scattering,
transmission-scattering.
Transmission
The source and detector are placed in an opposite direction of each other at a distance 1. The
sediment particles in the measuring volume reduce the beam intensity resulting in a reduced detector
signal. The relationship between the detector signal (It) and the sediment concentration (c) is:
It=k1 e-k2 c
in which:
k1 = calibration constant depending on instrument characteristics, fluid properties and travel distance (l),
k2 = calibration constant depending on particle properties (size, shape), wave length and travel distance (l).
Scattering
The source and detector are placed at an angle (f) relative to each other (see Figure 1B). The detector
receives a part of the radiation scattered by the sediment particles in the measuring volume. The relationship
between detector signal (Is) and sediment concentration (c) is:
Is=k3 c e-k2 c
in which:
k3 = calibration constant depending on instrument characteristics, fluid and particle properties (size, shape),
wave length and travel distance (l).
An important disadvantage of the scattering method is the strong non-linearity of the relation between the
detector signal and sediment concentration for large concentrations.
Transmission-scattering
This method is based on the combination of transmission and scattering, as shown in Figure 1C. If the travel
distance for transmission and scattering is equal, a linear relationship for the ratio of both signals is obtained
I=Is/It=k4 c
in which:
k4 = calibration constant depending on instrument characteristics and particle properties.
Important advantages are the absolute linearity between the output signal (I) and the sediment concentration,
the independence of water colour and the reduced influence of fouling.
Calibration
For all measuring principles an in-situ calibration for determining the constants is necessary, if possible
under representative flow conditions covering the whole range of flow velocities and measuring positions
(close to bed and water-surface). Regular calibration is required because the constants may change in time
due to variations in temperature, salinity and pollution.
In practice, the optical and acoustical sampling methods can only be used in combination with a mechanical
sampling method to collect water-sediment samples for calibration. Usually, about 10% of the measurements
should be used for calibration.
The inaccuracy of field measurements may sometimes be rather large because of calibration problems
(Kirby et al, 1981), particularly for optical samplers. The main problem is the lack of synchronity between
the optical and mechanical sample collection. To minimize synchronity errors, the optical samplers should be
calibrated bij measuring the silt concentration on board of the ship using a pre-collected water-silt sample.
Measuring range
For an optimal sampling resolution the wave length and particle size must be of the same order of magnitude.
Therefore the optical method is most suitable for silt particles (> 50 mm). Laboratory experiments using the
optical sampler, have shown that the addition of sand particles with a concentration equal to the silt
concentration increased the output signal with about 10% (Der Kinderen, 1981). The upper concentration
limit for optical samplers is about 25000 mg/1 (Kirby et al., 1981).
The acoustic method is most suitable for sand particles (>50 um). The upper concentration limit is about
10000 mg/1.
Advantages
An important advantage of optical and acoustical samplers is the continuous measurement of the suspended
sediment concentration. In combination with a chart recorder for data collection a relatively long period (one
month) can be sampled continuously and automatically. When there is very little variation of the silt
concentrations in lateral direction of the cross-section, measurements at one point can be considered as
representative for the whole cross-section. In that case the sensor can be fixed to a bridge pier or river side
installation. The measuring location must be easily accessible for regular cleaning of the sensor and changing
of batteries and chart records. Energy consumption and recorder maintenance can be minimized by using a
switch system activating the sensor and recorder only for short periods (5 min) at preset intervals (1 hour) as
reported by Brabben (1981). Another advantage of the continuous signal is the possibility of determining
continuous concentration profiles by raising the optical or acoustical sensor from the bed to the watersurface
(rapid profile method, Kirby et al 1981). Using this latter method a complete concentration profile can be
determined in one minute. To check the representativeness of these profiles, occasionally the concentration
profile should also be determined by means of a number of point-integrated measurements. The horizontal
variability can be determined by towing the sensor at a (monitored) depth below the water surface.
Finally, it is remarked that both sampling methods can also be used to measure the instantaneous sediment
concentration under wave conditions, provided the respons period is small enough.
References
Brabben, T.E., 1981. Use of Turbidity Monitor to assess Sediment Yields in East Java. Proc.Symp. Erosion
and Sediment Transport Measurements, Florence, Italy
Der Kinderen, W.J.G.J., 1980. Silt Concentration Meters; Evaluation (in Dutch).
Delft Hydraulics Laboratory, Report S453 I, The Netherlands
Der Kinderen, W.J.G.J., 1982. Silt Concentration Meters (in Dutch). Delft Hydraulics Laboratory, Report
M1799 I, Delft, The Netherlands
Kirby, R. and Parker, W.R., 1981. The Behaviour of Cohesive Sediment in the inner Bristol
Channel and Severn Estuary. Institute Oceanographic Sciences, Report No. 117, Taunton, England
Figure 1
General principles optical and acoustic sampling
5.6.4.2 Optical backscatter point sensor (OBS)
Principle
The OBS is an optical sensor for measuring turbidity and suspended solids concentrations by detecting
infrared light scattered from suspended matter (see Figures 1A and 1B). The response of the OBS sensors
strongly depends on the size, composition and shape of the suspended particles (see Figure 2). Battisto et al.
(1999) show that the OBS response to clay of 2 mm is 50 times greater than to sand of 100 mm of the same
concentration. Hence, each sensor has to be calibrated using sediment from the site of interest (see Figures 1
to 5). The measurement range for sand particles (in water free of silt and mud) is about 1 to 100 kg/m3. The
sampling frequency generally is 2 Hz.
The OBS sensors consist of a high intensity infrared emitting diode (IRED), a detector (four photodiodes),
and a linear, solid state temperature transducer (Downing et al., 1981). The (Optical Back Scatter) sensor
measures infrared radiation scattered by particles in the water at angles ranging from 140 to 165. Infrared
radiation from the sensor is strongly attenuated in clear water (more than 98% after traveling just 0.2 m),
(D&A instruments, 1989). Therefore, even bright sunlight does not interfere with measurements made in
shallow water.
The diameter of the sensor is about 0.02 m (see Figure 1); the length is about 0.05 m (see Photographs 1, 2
and 3 below). The IRED produces a beam with half power points at 50 in the axial plane of the sensor and
30 in the radial plane. The detector integrates IR-light scattered between 140 and 160. Visible light
incident on the sensor is absorbed by a filter. Sensor components are potted in glass-filled polycarbonate
with optical-grade epoxy.
The sensor gain of the OBS has to be adjusted in order to match the highest output voltage expected from the
OBS during the measurements with the input span of the data logger. Undesirable results will be obtained if
the gain is not correctly adjusted. When the gain is too high, data will be lost because the sensor output is
limited by the supply voltage and will saturate before peaks in sediment concentration are detected. If the
gain is too low, the full resolution of the data logger will not be utilized.
The performance of the OBS-sensor is claimed to be superior to most other in-situ turbidity sensors, because
of: small size and sample volume, linear response and wide dynamic range, insensitivity to bubbles and
phytoplankton, ambient light rejection and low temperature coefficient and low cost.
The OBS sensors are about the same size (or larger) as the length of gradients in the sand concentration
being measured. This may cause hydrodynamic noise in the output signal because the turbulent flow around
the sensor redistributes the particles in the water and increases the variation of sediment concentration above
natural levels. Furthermore, the volume sampled by the OBS sensors depends on how far the IR beam
penetrates into the water. This decreases as sediment concentration increases and so the sample volume is
constantly varying with concentration which may also cause random noise in the output signal. From limited
tests performed by the manufacturer it appeared unlikely that the random noise would exceed 30% of the
mean signal in situations with high concentrations of coarse sediment. The manufacturer recommends post-
processing the data with a low-pass filter to reduce the random noise in the output signal.
Other noise in the output signal may be caused by electronic noise or environmental conditions. According to
specifications, the electronic noise is insignificant for most applications. Some causes for environmental
noise are: biofouling, excess in suspended sediment resulting from scour around instrument structures and
cables moving in front of the OBS sensor with the currents.
Experiments have shown that the sensor gain varies with particle size. Ranging from mud (< 10 mm) to sand
(> 200 mm) the gain decreases approximately by a factor 10.
Hatcher et al. (2000) have used OBS sensors measuring at wavelengths of 442, 470, 510, 589, 620 and 671
nm with source beams originating from colour LEDs (six channel OBS; multi-spectral OBS) which can be
used to measure concentrations of sediment mixtures (multiple grain sizes). This makes it possible to
measure spectral responses of suspended particle concentrations across the optical range of wave lengths.
Using the differential response of the backscatter coefficient of the suspended constituents at six wave
lengths, an accurate estimation of concentration of mixtures can be obtained. This method is based on the
simultaneous solution of linear equations that relate output of optical backscatter sensors to concentrations of
various constituents of suspended sediments (see Green and Boon, 1993). The basic requirements are: 1)
linear sensor response to concentration of a particular sediment size, 2) different sensor response to different
sediment sizes and 3) grain shielding and multiple scattering should be negligible.
Calibration results from Utrecht University
A detailed description of the calibration of OBS sensors is given by Van de Meene (1994). The OBS sensors
were calibrated in a calibration tank of the Physical Geographic Laboratory at Utrecht University. Water is
circulated in a closed circuit by a strong slurry pump. The sediment is added from above in a large perspex
cylinder. The circulating water-sediment mixture is jetted into the cylinder, where the flow expands and
decelerates. A flow straightener is present to make the flow as smooth as possible. The water sediment
mixture flows undisturbed along the sensors with a velocity of approximately 0.25 m/s, which is large
enough to suppress inhomogeneities due to settling and small enough to prevent inhomogeneities due to
turbulence. Two OBS sensors can be calibrated simultaneously. A suction tube is present near the sensors to
draw concentration samples. The calibrations were carried out using cinput (=mass of sand in system divided
by volume of water) as the actual concentration. According to Van de Meene (1994) the sediment
distribution across the horizontal plane in the measurement region appeared reasonably homogeneous.
Variations were of the order of 5 to 10% of the mean concentration.
Figure 3 shows examples of the calibration curves for the OBS sensors used for the experiments carried out
in the large wave flume (GWK) in Hannover, Germany (grain size characteristics are d10 = 0.14 mm, d50 =
0.23 mm, d90 = 0.34 mm).
Figures 4 to 6 show calibration results using the bed material from tests in the wave tunnel (LOWT) of Delft
Hydraulics (two types of sand: d50 = 0.12-0.13 mm and 0.19-0.21 mm; d50 varied slightly based on samples
before and after the tests). The different response of the OBS sensors to the two different grain sizes is
reflected by the different slopes of the calibration curves.
Figure 5 shows this influence of the grain size on the calibration factor (slope of calibration curve). It can be
observed that the calibration coefficient is 2 to 3 times smaller when the grain size decreases with 30%.
Figure 6 shows the OBS concentrations measured in the calibration tank compared to the sand concentrations
from a pump sampler. It can be seen that the OBS concentrations show favourable comparison to pump
concentrations larger than 1 kg/m3. OBS values significantly deviate from pump concentrations smaller than
1 kg/m3. A systematic overestimation of the measured values can be observed for concentrations below 1
kg/m3.
The OBS sensors often show a reasonably steady offset concentration, which is related to the background
concentration of relatively fine sediments (silt and mud). It is common practice to subtract this offset value
from the original time series data. The offset can be defined as the minimum value of the data record (burst)
or as the 1% to 5% lowest value of the signal. For example, Battisto et al (1999) found that the most
appropriate cut-off voltage at the Duck site (USA) was 1% to 5% of the signal values.
Figure 7 shows time series values of two OBS sensors and one acoustic backscatter point sensor (ASTM) for
an experiment (M2) carried out in the large scale wave tank of Delft Hydraulics (Chung and Grasmeijer,
1999). The time-averaged ASTM-concentrations were about 1.3 kg/m3 at 0.115 m above the bed and 0.6
kg/m3 at 0.215 m above the bed. The OBS signal shows a background voltage of about 50 mV, which is
equivalent to a concentration of about 0.5 to 1 kg/m3. Hence, the background concentration to be subtracted
from the record is of the same order of magnitude as the sand concentration, which makes the application of
the OBS sensors rather dubious in the sand concentration range below 1 kg/m3. The acoustical ASTM sensor
does not show a background cocncentration due to fine sediments. This instrument is not sensitive for fine
sediments (<0.05 mm; smaller than the sand range).
Calibration results from Duck site, USA

Battisto et al. (1999) have made a comparison between OBS and pump sampler concentrations measured in
the surf zone at the Duck site (USA) during October 1997. For this study, OBS sensors were calibrated
separately using sand and mud collected at the Duck site. OBS voltage gain associated with mud was found
to be an order of magnitude larger than that for sand. Based on this calibration, Battisto et al. show that the
concentration of particles smaller than 63mm pumped at the Duck site during October 1997 correspond to the
lowest 1% to 5% of the output voltage recorded by the OBS sensors (background turbidity). The intake tubes
of the pump sampler were positioned approximately 0.1 to 0.2 m above the bed.
Calibrated OBS response above this background turbidity level was consistent with pumped sand
concentration as long as corrections were made for 1) varying size of suspended sand, 2) the precise time of
pump sampling, 3) apparent noise in the OBS records. Corrections for the smaller size of the suspended sand
relative to that used during calibration resulted in a decrease of the OBS sand concentration by about 50%.
Accounting for signal noise resulted in a decrease of the OBS sand concentration by about 0.05 to 0.2 kg/m3.
Despite these corrections the OBS concentrations are considerably larger (factor 2 to 5) than the pump
concentrations for sand concentrations smaller than 1 kg/m3. Hence, OBS data are unreliable for c<1 kg/m3.
OBS sensors are supplied by D&A instruments (www.d-a-instruments.com) and by Seapoint-

instruments (www.seapoint.com).
Technical specifications D&A sensor

dimensions sensor 0.018 x 0.05 m
housing 0.06 x 0.23 m
weight 1.3 kg
power 8-35 V/70 mA
measuring range mud 5 - 5000 mg/1
sand 100 - 100 000 mg/1
response period 10 Hz
temperature drift 0.05% per C
Technical specifications Seapoint sensor

dimensions sensor 0.025 x 0.12 m
power 7-20 VDC, 3.5mA, 6 mA pk
measuring range mud 5 - 5000 mg/1
sand 100 - 100 000 mg/1
response period 10 Hz
temperature drift < 0.05% per C; 0 to 65 oC
output 0-5 VDC
RMS noise < 1 mV
light source wavelength 880 nm
sensing distance < 5 cm
linearity <2% deviation 0-750 FTU
material ABS plastic, epoxy
Practical operation (on-line point measurements)
1. lower or raise sensor to sampling position (use echo sounder)
2. select sampling period (3 to 5 min)
3. read output signal (time-averaged)
4. collect water sample simultaneously for calibration (if necessary)
Advantages
1. small size and small sample volume
2. linear response; high-frequency response
3. insensitive to air bubbles and ambient light
4. large measuring range
5. long-term, stand-alone deployments (field-proven reliability)
6. low cost
Disadvantages
1. strongly dependent on particle size; regular calibration required
2. not usable in conditions with combined clay, silt and sand particles
3. not accurate for sand concentrations below 1000 mg/l (=1 kg/m3)
References
Battisto, G.M., Friedrichs, C.T., Miller, H.C. and Resio, D.T., 1999. Response of OBS to mixed grain size
suspensions during Sandy Duck97. Coastal Sediment Conference 99, ASCE, New York. pp. 297-312.
Connor, C.S. and De Visser, A.M., 1992. A laboratory investigation of particle size effects of an optical
backscatterance sensor. Marine Geology, Vol. 108, p. 151-159
D and A Instruments, 1989. Optical Backscatterance Turbidity Monitor. Instruction Manual
Tech. Note 3, 2428, 39th Street, N.W., Washington, D.C., 20007, USA
Downing, J.P., Sternberg, R.W. and Lister, C.R.B., 1981. New Instrument for the Investigation of
Sediment Suspension Processes in the Shallow Marine Environment. Marine Geology, 42, p. 19-34
Green, M.O. and Boon, J.D., 1993. The measurement of constituent concentrations in nonhomogeneous
sediment suspensions using optical backscatter sensors. Marine Geology, Vol. 110, p. 73-81
Hatcher, A., Hill, P., Grant, J. and Macpherson, P., 2000. Spectral optical backscatter of sand in
suspension: effects of particle size, composition and colour. Marine Geology, Vol. 168, p. 115-128
Van de Meene, J.W.H., 1994. The shoreface connected ridges along the central Dutch coast, The
Netherlands, Doctoral Thesis, Utrecht University, Department of Physical Geography, The
Netherlands.
Photographs 1, 2, 3, and 4
OBS sampler in wave tunnel of Delft Hydraulics
Figure 1A
OBS Sensor (D&A instruments)
Figure 1B
OBS Sensor (Seapoint instruments)
Figure 2
OBS Calibration curves for sediments between 37 and 121 mm (Connor and De Visser, 1992)
2500
OBS 358
2000 OBS 355
OBS 408
ouptut OBS (mV)
Linear (OBS 358)

1500
Linear (OBS 355)
Linear (OBS 408)
1000
500
0
0 5 10 15 20 25 30 35
concentration (kg/m 3)
10000
1000 OBS 358

ouptut OBS (mV)
OBS 355
OBS 408
100
10
1
0.01 0.1 1 10 100
Figure 3
Calibration of OBS sensors used during GWK Hannover experiments
80
OBS 350
70 OBS 355 D50 = 0.12 mm
OBS 408
60
Linear (OBS 350)
concentration (kg/m 3) Linear (OBS 355)
50
Linear (OBS 408)
40
30
20
10
0
0 500 1000 1500 2000 2500
output (mV)
80
70 D50 = 0.19 mm
60
50
40 OBS 350
OBS 355
30 OBS 408
Linear (OBS 350)
20 Linear (OBS 355)
Linear (OBS 408)
10
0
0 500 1000 1500 2000 2500
output (mV)
Figure 4
Calibration of OBS for LOWT experiments
0 .0 70
O BS 35 0
0 .0 60
O BS 35 5
O BS 40 8
0 .0 50
e xtrapo lation
c alibratio n fac to r (-)

0 .0 40
0 .0 30
0 .0 20
0 .0 10
0 .0 00
0.05 0.07 0 .0 9 0.1 1 0.13 0.15 0 .1 7 0.1 9 0.21 0.23
gra in s ize D 50 (m m )
Figure 5
Influence of sediment grain size on slope of OBS calibration curve, Oscillating Water Tunnel Experiments;
calibration factor is slope of calibration curves (in kg/m3 per millivolt)
100
predicted concentration (g/l)
OBS 350
10
0.1
0.1 1 10 100
measured concentration (g/l)
100
OBS 355
10
0.1
0.1 1 10 100
100
10
1
OBS 408
0.1
0.1 1 10 100
Figure 6
OBS concentration based on calibration curves as a function of pump concentration in the calibration tank;
two calibration curves with different sediment sizes (d50 = 0.12-0.13 and 0.19-0.21 mm) were used.
120
100
uncalibrated ASTM at z = 0.125 m
output ASTM (mV)

80
60
40
20
0
10:46 10:48 10:50 10:52 10:54 10:56
time (hh:mm)
500
uncalibrated OBS at z = 0.115 m
400
output OBS (mV)
300
200
100
0
10:46 10:48 10:50 10:52 10:54 10:56
time (hh:mm)
500
uncalibrated OBS at z = 0.215 m
400
output OBS (mV)
300
200
100
0
10:46 10:48 10:50 10:52 10:54 10:56
time (hh:mm)
Figure 7
Comparison of uncalibrated ASTM (or USTM) and OBS signals.
Delta flume M2, Hm0 = 1.5 m, Tp = 5.0 s, h = 4.55 m, D50 = 0.16 mm.
Figure 8
Measuring Sheet for Optical sampler
5.6.4.3 Optical Laser diffraction point sensors (LISST)
Principle
Various Laser diffraction instruments are commercially available to measure the particle size and
concentration of suspended sediments. The LISST instruments (Laser In-Situ Scattering and
Transmissometery) are manufactured by Sequoia Inc, USA (www.sequoiasci.com), (Agrawal and
Pottsmith, 2000, 2002). Detailed information is given in Chapter 6.
LISST-100: This instrument is the most widely used Laser diffraction instrument, which delivers the size
distribution by inversion of the 32-angle scattering measurements.
LISST-ST; This instrument has been designed to obtain the settling velocity distribution of sediments of
different sizes. In this case, a sample of water is trapped and particles are allowed to settle in a 30 cm tall
settling column at the end of the instrument-housing. Movable doors are present on both ends of the tube,
which are programmed to open at regular intervals. Using a motorised propeller, a water sample is drawn
into the tube through 8 openings of 20 mm diameter. Throughout, the size distribution is monitored near
the bottom of the settling tube. After sampling, a few seconds are allowed for turbulence to break down
before the doors are closed and the sample is allowed to settle for several hours. During settlement of 12
and 24 hours runs, respectively 72 and 83 Laser scans are made in logarithmically scheduled time intervals.
Over time, the size distribution shows zero concentration in sizes that have settled out. The time for settling
is used to estimate settling velocity. From knowledge of the size versus settling velocity, mass density can
be estimated. This instrument obtains the settling velocity and particle density for 8 size classes in the 5 to
500 micron range. The assumption that all particles settle independently in a complete stagnant fluid is
often violated. As a result, the calculated particle density ditribution often becomes unrealistically wide to
compensate for effects such as convection and particle interaction.
LISST-25A and 25X; This instrument is a simpler, less expensive version of the LISST-100. Replacing
the multi-ring detector of the LISST-100, a special shape for a focal plane detector was invented. This
shape (comet-detector) is the result of solving the mathematical problem: does there exist a detector shape
that would measure light scattering in a manner that it holds calibration for all sizes? Indeed, the LISST-25
holds calibration for spheres over a 200 to 1 size range, where earlier sensors would vary in calibration by a
factor of 200! The LISST-25 instrument is a superior sensor to the LISST-100 when only concentration
measurement is required. The LISST-100 obtains sediment concentration by first inverting the 32 multi-
angle scattering data to construct the size distribution and then summing the concentrations in the 32 size
classes. When small numbers of particles are present, as can happen with coarse particles, the inversion can
miss them due to noise. In contrast, since the comet-detector directly estimates concentration from the
weighted sum of angular scattering, it misses nothing. A second attribute of the LISST-25 is that this
device obtains particle area concentration from the optical transmission. The ratio of the volume
concentration and area concentration is called the Sauter Mean Diameter (SMD), first introduced in the
aerodynamics-droplet combustion literature. The two types of LISST-25 refer to an analog output only
version and a second version that is fully recording and presents a coarse fraction concentration in addition
to the total suspended load. The LISST-25X instrument has new comet shapes built in to separate between
wash load finer than 63 micron and the sand load larger than 63 micron. The two new comet shapes
deliver the total concentration and SMD in the entire size range and concentration and SMD in the coarse
sand range. The comet shapes assume nothing regarding the underlying size distribution of sediments. The
only requirement is spherical shape for particles. Inaccuracies of perhaps as much as 100% may occur if the
particle composition changes from mineral to biogenic.
LISST-SL: This instrument is a streamlined body that draws a sediment-laden stream into it for Laser
measurements. It incorporates a Laser, optics, multi-ring detector identical to the LISST-100 and electronics
for signal amplification and data scheduling and transmission. A pump is also built-in to ensure isokinetic
withdrawal rates. The pump is controlled by a microprocessor, which is fed information about the river
velocity by a propeller type current meter to ensure isokinetic velocity sampling. The propeller is mounted
above the body itself and a sensor is employed to count the number of its rotations in a short period of
time. This device includes pressure transducers to record the depth of sampling. The LISST-SL has been
designed to provide real-time data on sediment concentrations and particle-size distributions. The velocity
and concentration data are used to compute fluxes (on-line) for up to 32 particle size classes at points,
verticals or in the entire stream cross-section (Gray, 2004).
The LISST-SL offers a very powerfull instrument for on-line measuring of particle size, concentration per
size class (32 classes) and hence transport using a separate sensor for velocity measurements in rivers and
estuaries. A severe limitation is the relatively small concentration range (up to 500 mg/l) due to
insufficient light penetration of the optical sensor in conditions with concentrations larger than 500 mg/l.
see Section 6.5.5 and website: www.sequoiasci.com
Advantages
1. simultaneous measurement of particle size, concentration and fall velocity
2. simulataneous measurement of mud, silt and sand particles
3. instrument can measure in stand-alone mode
Disadvantages
1. limited concentration range for silty and muddy sediments (up to 500 mg/l)
2. not close to bed (relatively large sampler size)
3. fragile instrument in coastal conditions with surface waves
4. calibration required for non-spherical particles
References
Agrawal, Y.C. and Pottsmith, H.C., 2000. Instruments for particle size and settling velocity observations
in sediment transport. Marine Geology, Vol. 168, p. 89-114
Agrawal, Y.C. and Pottsmith, H.C. 2002. Laser Diffraction Method: two new sediment sensors. Sequoia
Inc., USA (www.sequoiasci.com)
Gray, J.R., 2004. The LISST-SL streamlined isokinetic suspended-sediment profiler. Proc. 19th Int. Symp. on
River Sedimentation, Yichang, China.
5.6.4.4 Various other Optical point sensors
Various types of optical samplers were and are commercially available. Herein, the following types of
optical instruments are discussed:
Eur Control Mex 2,
Partech Twin-Gap,
Metrawatt GTU 702,
Monitek 230/134.
Submersible Optical sensor Eur Control Mex 2

The instrument is based on the transmission of light waves using two light paths with different lengths, as
shown in Figure 1A. The (submersible) sensor consists of two light sources (A,B) and two photodiodes
(C,D). The light sources are used in turn for a period of one second. Firstly, only source A is off. Diode C
receives light over a distance x1 , while diode D receives light over a distance x2. The diode currents are
converted by a logarithmic amplifier and stored. Secondly, the process is repeated using source B. The result
of both phases are added.
The influence of (uniform) pollution is negligible small. An important disadvantage of this instrument is
the sensitivity to ambient light due to unequal exposure of the photodiodes.
light paths: 0.122 and 0.027 m
dimensions: sensor 0.3x0.1x0.05 m
weight: 1 to 10 kg
measuring range: 100-10000 mg/l
response period: 5s
References
Der Kinderen, W.J.G.J., 1980. Silt Concentration Meters; Evaluation (in Dutch). Delft Hydraulics
Laboratory, Research Report S453 I.
Der Kinderen, W.J.G.J., 1981. Silt Concentration Meters; Experimental Comparison (in Dutch). Delft
Hydraulics Laboratory, Research Report S453 II.
Der Kinderen, W.J.G.J., 1982. Silt Concentration Meters (in Dutch). Delft Hydraulics Laboratory,
Report M1799 I, Delft, The Netherlands.
Jansen, R.H.J., 1980. Methods for Measuring Velocity and Sediment Concentration in the
Breaker zone (in Dutch). Delft Hydraulics Laboratory, Report R971, The Netherlands.
Figure 1
Eur Control Mex 2
Figure 2
Eur Control Mex 2
Submersible Optical sensor Partech Twin Gap
The instrument is based on light-transmission, using two light paths with different length as shown
schematically in Figure 3A. The (submersible) sensor contains a light source with on both sides a light
sensitive resistor being part of a wheatstone bridge. The influence of ambient light is relatively small due to a
small aperture angle of the detector. The influence of uniform pollution is negligible small.
light paths: 0.0127 and 0.0064 m
weight: 1 to 10 kg
response period: <1 s
Figure 3/4
Partech Twin Gap
Pump-Optical sensor Metrawatt GTU 702
This instrument is based on the continuous pumping of a water-sediment mixture through the measuring
volume of the optical sensor (with light source and detector) on board of the survey vessel (Figure 5A). The
optical determination of the silt concentrations is based on the transmission-scattering principle using infra-
red light, as shown schematically in Figure 5B. Therefore, the influence of the electronic characteristics of
the instrument and also pollution is negligible small.
An important advantage of the pump-optical method is the relative simple calibration facility.
light paths: 0.032 m

weight: 1 to 10 kg
energy: 220 volt
response period: 10 s
Figure 5
Metrawatt
Pump-Optical sensor Monitek 230/134
This instrument is based on the pump-optical sampling method. The fluid sediment mixture is pumped
through the instrument (above water). The optical system is based on the transmission-scattering principle, as
shown schematically in Figure 6B. The light source is a simple electronic lamp that produces a narrow light
beam (2-3 mm) using a system of lenses. Two photodiodes are used to detect the directly transmitted light
and the forward scattered light. In an electronic circuit the ratio of both detector signals is determined
resulting in a linear output signal. An advantage of this instrument is the relatively large measuring range.
Another important advantage of the pump-optical method is the relative simple calibration facility.
light paths: 0.04 m
weight: 10 to 20 kg
energy: 220 volt
response period: 1s
Figure 6
Monitek 230/134
5.6.4.5 Acoustic point sensors (ASTM, UHCM, ADV)
Principle
Delft Hydraulics (1994) has developed an instrument (ASTM or USTM; Acoustic or Ultrasonic Sand
Transport Meter; in Dutch: Acoustische Zand Transport Meter) for measuring the velocity and sand
concentration in a point. The USTM or ASTM (Figures 1 and 2) is an acoustic instrument for measuring the
flow velocity in 1 or 2 horizontal dimensions and the sand concentration.
The Acoustic Sand Transport Monitor (ASTM) is based on the transmission and scattering of ultrasound
waves by the suspended sand particles in the measuring volume, as shown schematically in Figure 1A. Using
the amplitude and frequency shift of the scattered signal, the concentration and velocity and hence the
transport of the sand particles can be determined simultaneously and continuously. The ASTM consists of a
sensor with a pre-amplifier unit mounted on a submersible carrier and a separate converter with panel
instruments and switches (Figure 2). The velocity measurement if mounted on a carrier is one-dimensional
and related to the carrier orientation, which is measured by means of a magnetic compass. The vertical
position is measured by a pressure gauge (height beneath water surface) and an echosounder (height above
bed) mounted on the carrier.
A transmitting frequency of 4.5 Mhz has been chosen to minimize the particle size dependency and to make
the instrument insensitive to silt particles (< 50 mm). The influence of temperature and salinity variations is
also negligible.
Delft Hydraulics has designed a five-fold two-dimensional ASTM consisting of five identical sensors
(Figures 1B and 1C) for use as stand-alone instrument mounted in a tripod. Each sensor consists of one
transmitter and two receivers in a horizontal arrangement (measurement volume is about 0.2 m from
transmitter). The transducer heads are connected by a cable of 5 m to the electronics container (diameter of
0.27 m and length of 0.6 m). The transmitter produces a 4.4 MHz signal, which is scattered by the sediment
in suspension in front of the transmitter. This signal is subsequently sampled with a frequency of 2 Hz. The
backscattered signals are analysed to obtain the signal intensity and the frequency shift (Doppler effect). The
velocity of the sand particles can be derived (without calibration) from the frequency shift. The signal
intensity is a measure of the sand concentration and also depends on local sediment characteristics such as
the texture, the angularity and the density of the sediment (calibration curve). The measuring range for the
sand concentration of the ASTM (linear response) is about 0.1 to 10 kg/m3. According to the manufacturer
(Delft Hydraulics) the maximum error amounts to about 3% of the measured velocity value and 30% of the
measured concentration value. The five-fold ASTM was successfully used during long-term field
deployments in the North Sea to measure the sand transport process under combined current and wave
conditions (Grasmeijer et al., 2005).
The UHCM-instrument (only concentration) is a small-sized instrument (Figure 1B right) which has been
developed for the high concentration range of 1 to 100 kg/m3 near the bed (see Figure 1A). This instrument is
based on the measurement of the attenuation of ultra-sound by the sediment particles. The transducer heads
are close together at a distance of about 10 to 20 mm (depending on application; user-specified). Figures 4
and 5 show results of the acoustic sensor UHCM in a wave tunnel experiment (Van der Werf, 2006).
Rijkswaterstaat (2005) has made an attempt to use the backscattered signal from the point-measuring
Acoustic Doppler Velocitymeter (ADV-ocean) from SONTEK-instruments. This instrument was tested in a
laboratory flume and in field conditions. The laboratory tests showed that there is a positive correlation
between sediment concentration and backscattered signal. The field tests in the Western Scheldt estuary
(near Hansweert) consisted of concentration-measurements using the point-measuring ADV and another
point-measuring acoustic instrument (ASTM from Rijkswaterstaat). This latter instrument mounted in a
streamlined body suspended at a cable from the survey boat measures velocity and concentration
simultaneously in one point above the bed. The ADV was operated with a time-averaging interval of 0.5 s.
The sampling frequency was in the range of 10 to 20 Hz. The measured concentrations were in the range of
0.1 to 4 gr/l. Analysis of the measured concentrations using both instruments shows that the ADV cannot
measure concentrations larger than about 0.5 gr/l (500 mg/l) due to saturation effects. The concentrations
show a weak non-linear behaviour in the lower range between 0.1 and 0.5 gr/l. The velocities of the ADV
and ASTM show close agreement. The overall conclusion is that the ADV is potentially usable as sand
concentration and transportmeter, but its working range is insufficient at present (maximum concentration
of about 0.5 gr/l) to cover the near-bed region of the water column where the concentrations are largest and
most of the sand transport takes place.
Fugate and Friedrichs (2002) have successfully used the ADV backscattered signal (Sontek ADV) to
determine the concentrations in the range of 20 to 100 mg/l at a site in Chesapeake bay (USA). The ADV
proved to be a versatile instrument for characterizing the suspended sediment dynamics. Acoustic backscatter
from the ADV was relatively insensitive to grain size differences, thus producing good estimates of mass
concentrations. In addition, the ability of the ADV to measure Reynolds bed stresses (<u/w/>) and vertical
Reynolds fluxes (<w/c/>) of sediment allows estimation of the critical shear velocities and the indirect
analytical estimation of settling velocity of the sediments involved.
Calibration curve
The ASTM has been calibrated using pump sampling concentrations obtained during experiments in the
Deltaflume (Chung and Grasmeijer, 1999). The five intake openings of the pump sampling equipment
were positionned close to the acoustic sensors.
The time-averaged sand concentrations measured by the five intake tubes of the pump sampler near the
acoustic sensors have been used to determine the calibration curve of the ASTM. The time-averaged (over
about 15 min) sand concentrations are between 0 and 3.7 kg/m3 for the coarse sediment (0.33 mm sand) and
between 0 and 2.7 kg/m3 for the finer sediment (0.16 mm sand). The results are shown in Figure 3. It can be
observed that for concentrations larger than 0.05 g/l the output of the ASTM varies linearly with the
sediment concentration. For smaller concentrations (<0.05 g/l) the ASTM output is larger than may be
expected for a linear relationship. This is in agreement with the measuring range of the ASTM according to
technical specifications (0.1< c <10 kg/m3).
The calibration curve can be represented by: ConcentrationASTM = 0.257*OutputASTM. The effects of particle
size in sand size range of 0.16 to 0.33 mm (d50= 0.16 mm and d50= 0.33 mm) is negligibly small.
Practical operation for on-line measurements

1. lower (or raise) sensor to sampling position (use echo sounder)
2. select sampling period (5 minutes)
3. read output signals
Remarks
1. use (electronic) time integrator
2. clean sensor regularly.
Results and accuracy of ASTM

The relationship between the output signal (I) and the sand concentration (c) is as follows:
I=kc
in which:
k = calibration constant depending on properties of the transducers and sand particles (size, shape).
The inaccuracy of the measured velocity is about 2% (of the full scale). The inaccuracy of the measured sand
concentration follows from the inaccuracy (scatter) of the calibration curve. Figure 1A shows an example for
field conditions (Eastern Scheldt, The Netherlands), where a pump filter sampler has been used for
calibration. The overall accuracy is about 20%.
Technical specifications (ASTM)
sound paths: 0.15 m; frequency 4.5 MHz
weight: 10 to 20 kg
energy: 220 volt, 50-60 Hz, 40 watt
output signal: 0-10 volt (analog)
outputs: 0 to 10 V for concentration (linear); -10 V to +10 V for velocity
RS-232C port of data transfer to PC, baud rate 9600
measuring range: 0.03-3 m/s; 10-10000 mg/l(particles 50 to 500 mm)
response period: 0.1 s
cable: 5m
conditions: fully immersible to 20 m; temperature range of 5 to 30 degrees
Advantages
1. rapid, simultaneous measurement of velocity, concentration and transport
2. linear response between 10 and 10,000 mg/l (concentration range)
3. small response period
4. weakly sensitive to sand particle size; not sensitive to silt
5. insensitive to temperature, salinity variations and sensor pollution
7. long term, stand-alone deployments (field proven reliability)
Disadvantages
1. calibration required for concentration
2. fragile electronic equipment
3. large dimensions and weight (carrier)
4. intrusive transducer head arrangement; large dimensions of electronics container
5. not suitable for silt
References
Delft Hydraulics, 1980. Methods for measuring velocity and sediment concentrations in breaker
zone (in Dutch). Report R971. Delft, The Netherlands
Delft Hydraulics, 1994. Manual ASTM (in Dutch). Report B329, Delft. The Netherlands
Fugate, D.C. and Friedrichs, C.T., 2002. Determining concentration and fall velocity of estuarine particle
populations using ADV, OBS and LISST. Continental Shelf Research, Vol. 22, p. 1867-1886
Grasmeijer, B.T. et al. 2005. Suspended sand concentrations and transports in tidal flow with and without
waves (Paper U). In: Sand transport and morphology of offshore sand mining pits edited by Van Rijn
et al., ISBN90-800356-7-x. Aqua Publications, The Netherlands (www.aquapublications.nl)
Jansen, R.H.J., 1978. The in-situ Measurement of Sediment Transport by Means of Ultrasound
Scattering. Delft Hydraulics Laboratory, Publication No. 203, Delft, The Netherlands
Jansen, R.H.J., 1979. An Ultrasonic Doppler Scatterometer for Measuring Suspended Sand
Transport. Ultrasonics Internal Conference, Graz, Austria
Jansen, R.H.J., 1981. Combined Scattering and Attenuation of Ultrasound. IAHR-Workshop on Particle
Motion and Sediment Transport, Rapperswil, Switzerland
Rijkswaterstaat, 2005. Acoustic Doppler Velocitymeter as sand transportmeter (in Dutch). Document
RIKZ/KW/2005.110W. RIKZ, The Haque, The Netherlands
Schaafsma, A.S. and Der Kinderen, W., 1985. Ultrasonic Instruments for the Continuous Measurement of
Suspended Sand Transport. IAHR-symposium Measuring Techniques, Delft, The Netherlands
Van der Werf, J.J., 2006. Sand transport over rippled beds in oscillatory flow. Doctoral Thesis,
Department of Civil Engineering, University of Twente, The Netherlands
Figure 1A
Acoustic ASTM
Figure 1B
Left: ASTM (or USTM); five-fold and two-dimensional
Right: UHCM transducer heads
Figure 1C
The five-fold two-dimensional ASTM (or USTM); side view (left) and topview (right)
Figure 2
Acoustic AZTM
10

1
0.1
0.01
0.001
0.001 0.01 0.1 1 10
output USTM (V2)
Figure 3
Time-averaged output of ASTM sensors plotted against the sediment concentration measured with the pump
sampling system (Chung and Grasmeijer, 1999)
Figure 4
Time-averaged and ripple-averaged sand concentration in wave tunnel experiment of UHCM compared with
results of OPCON and Pumps samples (TSS)
Figure 5
Time-series of sand concentrations above ripple crest in wave tunnel experiment
_________ Acoustic sensor (UHCM); ------- Optical sensor (OPCON)
Figure 6
Measuring Sheet acoustic sampler AZTM
5.6.4.6 Acoustic backscatter profiling sensors (ABS and ADCP)
Principle
Acoustic backscatter (ABS) measurement is a non-intrusive technique for the monitoring of suspended
sediment particles in the water column and changing sea bed characteristics. An acoustic backscatter
instrumentation package comprises acoustic sensors, data acquisition, storage and control electronics, and
data extraction and reduction software. An overview of the ABS-technique is given by Smerdon, Rees
and Vincent (www.aquatecgroup.com). Hereafter, a summary of this is given.
The basic principle of the acoustic backscatter approach is as follows. A short pulse (10 Ps) of acoustic
energy is emitted by a sonar transducer (1 to 5 MHz). As the sound pulse spreads away from the
transducer it insonifies any suspended material in the water column. This scatters the sound energy,
reflecting some of it back towards the sonar transducer, which also acts as a sound receptor. With
knowledge of the speed of sound in water, the scattering strength of the suspended material and the sound
propagation characteristics, a relationship may be developed between the intensity of the received echoes
and the characteristics of the suspended material. With typical acoustic ranges in excess of 1 metre, the
acoustic head remains outside the area of study and therefore makes the instrument non-intrusive. The
magnitude of the backscattered signal can be related to the sediment concentration, particle size and the
time delay between transmission and reception. The acoustic backscatter intensity from a uniform field of
particles of constant concentration is assumed to be an inverse function of the distance from the source
with corrections for attenuation due to water and particles. Calibration in uniform suspensions is required
to find this relationship. The theoretical background of the acoustics is described in detail by Thorne and
Hanes (2002). Early work was done by Hay (1983).
The sensor comprises acoustic transducers that emit pulses of sound, which are incident on the sea bed.
They receive sound reflected by the sea bed and suspended sediment in the intervening water mass. The
instrument records the amplitude of reflected sound at gated intervals, thus building a reflected sound profile.
With low angles of incidence, the technique may be used to monitor the formation and progress of sea bed
ripples. Perpendicular incidence angles will yield information on sediment suspension between the sensor
head and the sea bed, and on the erosion or accretion of the bed level. The vertical resolution is limited by
the length of the acoustic pulse and by the speed at which the signal is digitized and recorded. A vertical
resolution of about 1 cm is feasible. Temporal resolution depends on the pulse repetition rate and on the
number of pulses which must be averaged to produce statistically meaningfull backscatter profiles. Vincent
et al. (1991) used a pulse repetition rate of 10 Hz and four profiles were averaged before storing the data on
disc. On average, a profile was recorded every 0.58 s; 1250 average profiles were recorded during each burst
(12 min).
Libicki et al. (1989) identified two difficulties in estimating the suspended load from the backscatter
signals. First, their instrument did not measure in situ the attenuation of sound caused by the suspended
load, which was introducing errors when sediment concentrations were high. Second, their instrument was
unable to distinguish between changes in particle size distribution and sediment concentration, although
in their experiments they showed that the assumption of a time-invariant particle size distribution did not
introduce substantial errors. They felt that to decouple particle size distribution from sediment
concentration would require multiple frequency devices with impractically high upper frequency limits.
The acoustic method is most appropriate for particle size distributions on the order of tens to hundreds of
microns (say 10 to 500 microns).
Libicki et al. (1989) presented a detailed analysis of the sources of interference in the ABS signal. Besides
interference from non-sedimentary targets including fish, plankton and other marine biota, they analysed
the effect of bubbles, whose target strength can be highly significant with respect to that of sediment. They
concluded that the expected lifetime of 3MHz resonant bubbles was sufficiently short to be insignificant
for their system at their deployment depths. However, they concluded that there may be significant
interference from superresonant bubbles entrained by wave activity at depths of less than 3 m and that
deployment in or very near the surf zone would not yield accurate sediment concentration measurements.
They also identified various noise sources and suggested means of mitigating them. Thermal noise is
generated in the electronics of the receive circuitry, while ambient noise arises from the marine
environment. Neither is coherent and both may be minimised by reducing the input signal bandwidth
through effective filtering. It may be quantified by periodically turning off the transmitter and recording
sound profiles without any echo return. This is normally done at the beginning of an ABS data gathering
cycle, when the first group of data recorded in exactly this manner.
Experimental and theoretical work by Thorne, Holdaway et al. (1995) has addressed the problem of
sound attenuation by suspended sediment by measuring the signal strength of the bottom echo. Hay and
Sheng (1992) carried out an analysis of three-frequency acoustic backscatter signals in which they
developed a procedure to extract not only suspended sediment concentration, but also particle size
distribution with encouraging results. By simultaneously measuring both parameters, it is now possible,
in principle, to estimate directly the vertical mass flux by settling.
The uncertainties associated with measurements from earlier single frequency instruments have now been
overcome. A detailed theoretical understanding of the acoustic response of suspended sediment, backed
up with laboratory and field studies, has established the multi-frequency acoustic backscatter instrument as a
useful tool for the study of near-bed sediment processes.
Usually, a multi-frequency acoustic instrument (ABS) is used to determine the sand concentrations in the
near-bed region (lowest 1 meter of the water column).
Attempts have been made to use the backscattered signal of the (ship-mounted or bottom-mounted) Acoustic
Doppler Current Profiler (ADCP) for determining the suspended sediment concentrations (Hoitink and
Hoekstra, 2005). Similarly, the ADV was used to determine the suspended sediment concentrations (Visser,
1997).
Instrument description (ABS)

To minimise effects on water flow in the vicinity of the measurement area, the small acoustic head is
mounted remotely from the main instrument housing and attached to it by an umbilical cable. Most of the
study of near-bed sediment processes is carried out in shallow coastal waters, to which a light plastic
housing of PVC or similar material is well suited. The ABS-instrument supplied by Aquatecgroup
(www.aquatecgroup.com) is based on three frequency ranges. The three-frequency acoustic head
contains three separate transducers, each tuned to a different operating frequency. Each transducer
operates both as a transmitter and as a receiver. The transducers are each mounted on separate acoustic
backings but placed closely together to minimise the type of inaccuracies ascribed to high spatial
separation by Hay and Sheng (1992). Figure 1 shows the acoustic head. The choice of frequencies and
the transducer characteristics are essential in determining how the system will operate. The following
guidelines may be used.
a) Higher frequencies are used to resolve smaller particles. Thorne, Waters et al. (1995) have obtained
through empirical methods, backed up with theoretical analysis, a simplified form function for a
suspension of irregularly shaped scatterers, which can be applied to suspended sediment. To determine
particle size, the choice of frequencies should be such that the particle sizes of interest span the regions of
variable response.
b) Higher frequencies are attenuated more. The range of interest for sediment studies is normally limited
to less than 2 metres. From Coates (1990), the two-way attenuation due to sound absorption by water
(frequencies in excess of 1MHz) is found to depend on temperature and frequency. The percentages of
attenuation of sound power due to water absorption for different frequencies and temperatures are given in
the following table.
.
Temp/Frequency 1 MHz 2 MHz 3 MHz 5 MHz 10 MHz
5C 7.9% 28.1% 52.4% 87.3% 99.97%
15C 5.4% 19.9% 39.3% 75.0% 99.61%
25C 3.7% 14.1% 29.0% 61.4% 97.79%
c) The dimensions and shape of the transducer affect the horizontal spatial resolution of the system and the
characteristics of the acoustic response. The volume of water that is insonified can be roughly divided into
two regions. In the far field of the transducer, the sound source may be treated as if it were a point source.
For a circular transducer, the insonified volume may be approximated as a cone with its tip at the
transducer. In the near field of the transducer, the relationship between range and intensity becomes
complex. The standard suspended sediment monitoring configuration insonifies a narrow vertical column,
with each of the three frequencies incident on a compact spatial volume. However, the characteristics of the
transducer may be adjusted to achieve specific, non-conical beam patterns. For example, by using
transducers that are long and thin, a beam pattern is created that has a narrow beam in one axis and a wide
beam in the orthogonal axis. Such a pattern can be used to examine changes in bedform shape by aiming
the transducer at a low incidence angle to the seabed. A narrow strip of seabed along the transducer axis is
insonified. Using a principle similar to sidescan sonar, the changing characteristics of bedform ripples may
be observed.
The sounder transmit and receive electronics is controlled by a dedicated microcontroller. This controls the
generation of the three transmitter and three receiver local oscillator frequencies, which are monitored for
stability and corrected if necessary. Each channel has a tuned amplifier and transformer to match the
transducers to the drive electronics for maximum efficiency. The transmitter generally operates at
approximately 1W, with a pulse duration corresponding to a pulse length of just under 10 mm. In the
transducer head, there are receiver preamplifiers for each channel, which are multiplexed into a balanced
Schottky-diode mixer. The mixed down frequency is amplified with a linear time-varied gain amplifier.
This is used to compensate for the decrease in received signal intensity due to geometrical spreading
losses and also helps to overcome attenuation due to absorption. The received signal is detected and filtered
to eliminate aliasing when sampled by the controller.
The data logger and controller system is a two-board assembly with a hard disk. The first board is a general
purpose combined computer and data logger card, which is based on the IBM PC/XT architecture. The
second board is a specially designed interface board. The PC/XT architecture is chosen as being the most
practical approach to interface to a hard disk, since the system has a DOS compatible operating system. The
second board is designed to interface to the sounder control signals and output and to provide a trigger and
data transfer interface to the user. Opto-isolated trigger inputs allow the use of external events to trigger the
logging cycle and the type of logging. A second serial port is interfaced to the datalogger card. This is a
high speed port capable of sustained transmissions at 115200 baud, that is used to set up the logger initially
and then to recover data from it after a deployment. The serial interface standard chosen is RS422. This is
a four-wire interface that can operate over several hundred metres and therefore overcomes the difficulty
of attempting to communicate with the logger on a wet ship deck. A timer is also interfaced to the XT bus.
This is used to time the data acquisition and sounder triggering cycle. The board also includes power supply
and control circuitry, so that the system can operate efficiently from a variety of battery types and voltages,
and can switch into and out of low power modes according to internal or external triggers. Data buffers are
fitted to allow the hard disk to be safely turned off while other parts of the system remain powered up. A
long period watchdog timer is included, which can provide a system power up and reset, typically every
4096 seconds in the absence of normal logging activity, if required. This safeguards against
microprocessor crashes or a failure to receive external triggers. Finally, two Flash EPROMs are used in
place of that on the data logger card to hold not just the BIOS and DOS, but also the main data logger
application program. The hard disk now fitted is a 520MB AT-style IDE hard disk, of the type typically
found in notebook computers. The system originally operated with XT hard disks, available up to 80 MB.
However, these have now become obsolete, and a new disk BIOS has been developed to control the AT
disk from the XT electronics. This type of disk is extremely rugged and when not performing disk
accesses, can withstand shocks in excess of 100g, thus protecting it from the level of shock that may be
experienced during instrument deployment or recovery. The data logger is supplied with a software package
that allows the user to set up the various logging parameters for a deployment. These include parameters
governing the triggering of data acquisition, the duration of logging, the acquisition rate, the type of
averaging to be performed, how many frequencies to use and whether only selected bins of data should be
processed.
Instrument calibration
The early history of ABS use was characterised by concerns about interpretation of backscatter data. In
particular, Libicki et al. (1989) were concerned both about the increased likelihood of error caused by
signal attenuation by the suspended load itself, and the inability, without using many frequencies, to
distinguish particle size and to separate variation in particle size from variation in suspended load. In a
series of theoretical studies of backscatter characteristics from suspended glass spheres (Thorne and
Hardcastle, 1991; Thorne and Campbell, 1992; Thorne et al. 1992 and Thorne, Manley et al., 1993),
the general form function for a suspension of regular shaped scatterers was derived and tested against
laboratory experiments. The use of glass spheres was chosen because the main constituent of both glass and
marine sediments is quartz. The usefulness of these experiments was further enhanced by recent work on
irregularly shaped scatterers (Thorne et al., 1995), which show the general principles applied to regularly
shaped scatterers with moderate size distribution is similar to a first order approximation to that of
irregularly shaped scatterers, such as would be found in marine sediment.
When sediment concentrations are high, the technique for calculating the concentration profile becomes
inaccurate. This is because the attenuation term (a) includes a contribution from suspended load-related
attenuation. The iterative approach required to solve for load introduces errors that can cause divergence
from the solution. Recently, Thorne, Holdaway et al. (1995) examined how the attenuation due to load
may be estimated independently, thereby constraining the load solution. The principle behind their method
is to monitor the signal strength of the seabed echo during calm periods, when there is little or no sediment
in suspension. This may be used to calculate the attenuation due to water alone, or simply to act as a
reference signal strength. By monitoring the strength of the seabed return during periods of high suspended
load, the attenuation due to suspended load may be derived.
A method of particle size determination using multi-frequency acoustic backscatter was described and tested
by Hay and Sheng (1992). The principle is to use the frequency-dependent variation in backscatter form
function for particles where ka<2 with k=2Sf/c=wave number, f=frequency, c=speed of sound, a=mean
equivalent radius. For particles that lie on this sloping region for at least one frequency and which return
sufficient backscatter to be detected, a ratio between two or more frequency responses may be calculated.
There must be sufficiently high concentration levels to calculate with reasonable accuracy. In the
analyses described, if the measured standard error of estimates of particle size exceeded a certain level, the
estimates were discounted, even though this may have resulted in some valid fluctuations in concentration
being rejected. Using frequencies of 1MHz, 2.25MHz and 5MHz, they were able to estimate particle sizes in
the range 50 Pm to 170 Pm to between 10% and 20%. They also note that once the relative sensitivities of
the three frequencies have been established, the calibration of the system is site independent.
The expression for the suspended load profile relates suspended load to acoustic backscatter pressure. In a
practical system it is necessary to calibrate the system response to known suspensions to take account of
variations in the transducer sensitivity, amplifier gain, TVG response and perhaps transducer radiation
characteristics. The latter was noted by Downing et al. (1995) to be significant in the correction for near-
field results, in which the measured radiating dimensions of their test transducers varied from the physical
dimensions by up to 15%. The standard method of calibration is either to use a suspended sediment jet, as
described by Hay and Sheng (1992) or a sediment tank in which a homogeneous sediment suspension is
circulated. The former method was used to calibrate the relative sensitivities of the three transceiver systems
to jets of known particle size distribution. The latter system is commonly used to calibrate the overall
system response to a uniform distribution, which may then be extrapolated using the previously described
equations for suspended load profiles. Where possible, suspended or seabed sediment samples are taken
from the deployment site to back up general calibration data and to analyse particle size distributions for a
given site.
Field and laboratory deployments

Early ABS-experiments were mainly conducted with single frequency systems. Green and Vincent
(1991) described experiments to evaluate a model of concentration profile in a combined wave and current
flow. They analysed the predictions of sediment reference concentration and vertical rate of decay of
concentration. Soulsby et al. (1991) described an experiment to monitor suspended sediment over sand
waves. The ABS-data were transferred by cable to the beach. Acoustic measurements were supported by
pump samplers at key elevations. Van Hardenberg et al. (1991) took measurements using a 3MHz ABS-
system. They reported on a method to remove the kinematic effects of wave motion from their results to
leave only the effects from physical forcing of suspension. Vincent and Green (1991) observed
substantial differences in suspension patterns for two instances of similar wave and current conditions,
leading them to suggest that the time scale of bedform processes may be much longer than the 8 minute
samples they took. Vincent et al. (1991) postulated that seabed roughness had a significant effect on the
level of resuspension. Their work also highlighted many of the early shortcomings of the single frequency
ABS-approach.
Green et al. (1995) present measurements of wave heights, near-bed currents, bed shear stresses and
suspended sediment concentrations and fluxes during a severe storm using multi-frequency ABS
systems. The acoustic backscatter data revealed wave resuspension of bed sediment, modulation of
sediment concentration by wave groups and advection of clouds by currents. Heyse et al. (1995) and
Vincent et al. (1998) present results of a study of sand bank mobility on the Middelkerke Bank in the North
Sea off the Belgian coast. This includes a brief description of the correlation between OBS and ABS
sensors in monitoring the transport of sand by resuspension. A new instrumentation platform for the
monitoring of seabed boundary layer processes, ALICE, was introduced by Green (1996). In addition to
the ABS- system, the platform includes four electromagnetic current meters, an acoustic Doppler
velocimeter for measuring near-bed three-dimensional turbulence, a precision pressure sensor for measuring
waves and tide state, five optical backscatter sensors, a time lapse video camera and an automated water
sampler. The instruments are controlled by a master data logger.
A typical example of ABS-concentration profiles as a function of time within a burst is shown in Figure 2
(River Tees, UK).
Osborne et al. (1994) have used a single-frequency profiling ABS (2.8 MHz) in combination with a point-
measuring optical backscater instrument (OBS) in the nearshore zone at Stanhope Lane Beach, Prince
Edward Island, Canada (water depth of about 1.5 m, sand bed with d50 of about 0.23 mm). The ABS was
calibrated in a laboratory recirculation tank using sand taken from the field deployment site. As the ABS is
sensitive to any change in the acoustic impedance and so will respond to bubbles and organic material in
suspension, the measurements were taken well outside the breaker zone (bubble contamination was small;
organic material content was also small). Figure 3 shows typical results (with relatively large scatter) of ABS
and OBS concentrations close to the bed averaged over the wave half-cycle.
Vincent and Green (1999) described a field arrangement on the Continental Shelf (Pacific East Coast of
New Zealand) with three transducers (F1= 1, F2= 2 and F4= 4 MHz) and a pulse-repetition rate of 80 Hz; each
profile recorded consisted of an average of 16 pulses (5 Hz). The vertical resolution is 1 cm. The
concentration range is about 0.1 to 20 kg/m3. The system was calibrated in a laboratory recirculating
suspension tank using sand from the deployment site (0.33 mm sand).
The mass concentration at range r from the acoustic transducer is estimated from a function, depending on
the voltage V measured at range r, the sediment density, the speed of sound in water and the attenuation of
sound by water and sediment of radius a. The attenuation is a complex form function of ka, which describes
the efficiency with which sediment of radius (a) backscatters sound of acoustic wavenumber (k). Three
different acoustic frequencies are used to simultaneously determine the size and concentration of the
suspended sediment involved.
The strongest acoustic echoes are used to identify the position of the sand bed. Close to the bed, the bed echo
dominates the backscattered signal. The concentration at 1cm above the bed is defined at the height at which
the first uncontaminated echo occurs, which is identified from a break-in-slope in the concentration profile
close to the maximum backscattered signal in the burst-averaged profile. The uncertainty in height is about
r0.5 cm.
The concentration profiles measured by the three transducers should be identical, if the calibration conditions
are perfect, which means that the suspended sand has the same size distribution at all heights in the water
column at the field site and in the laboratory tank. Vincent and Green (1999) show examples of
concentration profiles based on the three frequencies, which have relatively large differences (factor 3, see
Figure 4) in concentrations. The concentration profiles were calculated using the results of the calibration
tank (based on 0.33 mm sand from the bed at the field site). The concentration profiles differ systematically
with F1-concentration < F2-concentration < F4-concentration. When the backscatter data are re-processed
using F1 simultaneously to obtain concentration and size, the sand concentrations are between those of F2 and
F4-concentration and the suspended sand size varies between 0.25 and 0.15 mm. It is assumed that the
suspended sand has a Gaussian distribution at every height and that the width of the distribution is constant.
When the F1 and F4-frequency pair is used, the suspended sand sizes become smaller and the concentrations
become larger; the latter show a discontinuity due to the shape of the form function yielding ambiguous
results for F1-F4 pair. This latter combination of frequencies is very sensitive to small errors in backscatter
intensity. These analysis results suggest that the sizes of the suspended sand at the field site differ
significantly from that of the bed material used in the calibration procedure. The concentrations derived from
the F1-F2 pair were found to be the most reliable. Vincent and Green concluded that the applied form
function is not quite right and should be reconsidered.
Another problem is the elimination of the effects of air bubbles in the water column, if the ABS-system
(highly sensitive to air bubbles) is used in the surf zone with breaking waves (Huck et al., 1999). This can be
done by analysis of the time-averaged concentration profiles, which should show a decreasing concentration
with increasing height above the bed. If large amounts of bubbles are present, the concentration profiles
derived from the ABS will show an increase of concentration at higher levels. These data records should be
excluded from the analysis.
The optimum conditions for the ABS-system are: rather uniform fine sand (0.1 to 0.3 mm) in non-breaking
wave conditions.
Thorne et al. (2002) have used an ABS in the large-scale Deltaflume of Delft Hydraulics to measure the
near-bed sand concentration profiles under regular and irregular waves over a sand bed with d50 of 0.33 mm
(see Figures 5 and 6). A pump sampler was operated very close to the sampling profile of the ABS at
nominally 0.05, 0.07, 0.10, 0.13, 0.18, 0.25, 0.65, 1.05 and 1.55 m above the bed. Ten liter samples were
collected and dry weighed. Size analysis (by sieving) was performed to a subset of samples to determine the
sediment size profile in the near-bed region. The ABS was operated at three frequencies of 1, 2 and 4 MHz,
collecting backscatter profiles at each frequency at a sampling rate of 128 Hz, with a spatial resolution of
0.01 m, over an operating range of 1.28 m. The transducers were mounted at nominally 1.24 m above the
sediment bed. To evaluate the acoustic concentration profiles an explicit inversion scheme was applied to the
data (Thorne and Hardcastle, 1997). This approach uses a pumped sample concentration measurement at
one height above the bed as a calibration concentration. Using this explicit approach allows the concentration
to be computed as:
c=R2/[Ro2/Mo Ir]
with: Ir= ror (4]sR2)dr, R=[V\r/ks]e2rD, V=backscattered voltage, r=range from transducer, \=departure from
spherical spreading within transducer field, ks and ]s=acoustic backscattering and attenuation parameters,
D=attenuation due to water absorption, Mo=calibration concentration at range ro, Ro=value of R at r=0.
Figure 5 shows time-averaged concentration profiles (over 17 min) for typical cases of regular and irregular
wave conditions. Over the 17 min period there was generally some ripple migration below the ABS-
transducers and therefore the measured concentration profiles were considered to be approximately quasi-
ripple-averaged profiles. The results from the three ABS-frequencies are comparable with one another.
However, there are differences, due in part to the calibration of the voltage transfer function for the system
and also to our present limited knowledge of the variability of the backscattering and attenuation
characteristics of different sediments. The concentration averaged over the three ABS-frequencies shows
generally good agreement with the pumped sample data, though there is some divergence at greater heights
above the bed. A comparison between all the contemporaneously measured pumped samples and acoustic
concentrations is shown in Figure 6.
Grasmeijer et al. (2005) have successfully used a point-measuring (in 5 points above the bed) acoustic sand
transportmeter (ASTM; 4.5 MHz) and an acoustic profiling instrument (ABS; three frequencies). Both
instruments were operated from two stand-alone tripods in the Dutch sector of the North Sea. The local water
depth was about 13 m. The seabed consisted of fine sand with d50 of about 0.22 mm. One major storm with
significant wave heights up to 4 m did occur during the deployment period. Peak tidal velocities were of the
order of 0.6 to 0.8 m/s. On average, the suspended transport rates derived from the ABS-data were somewhat
smaller (factor 2) than those of the ASTM. The ABS tends to underestimate the concentrations of particles
smaller than 0.2 mm due to the frequency used (2 MHz).
Visser (1997) has studied the backscattered signals from ADCP data (RD type DR-BB, frequency of 1200
KHz, beam angle of 20 degrees, mounted at 1.25 m beneath the water surface) collected from a moored ship
in the Western Scheldt estuary, The Netherlands. For calibration water-sediment samples were taken at 5
heights in the water column every 2 hours. The local water depth varied between 12 and 16 m. The
suspended sediment samples were used to determine the various sediment size classes and sediment
concentrations per size class. The inversion model of Thorne et al. (1995) was used to translate the acoustic
signals to sediment concentration profiles. It was found that the acoustic concentrations are very sensitive to
the Dcoefficient which represents the acoustic attenuation due to water phase. Figure 7 shows Dvalues in
the range of 0.01 to 0.1 to obtain reasonable agreement between the acoustic concentrations and the water-
sample concentrations (calibration samples in the range of 10 to 300 mg/l). Increasing the Dvalue from
about 0.042 to 0.08 results in an increase of the concentrations by a factor of 2 to 5. The Dvalues were
found to vary from case to case (without any systematic dependency).
Hoitink and Hoekstra (2005) have successfully used a standard 1.2 MHz ADCP-instrument in combination
with an in-situ calibrated optical backscatter instrument (OBS) to measure the mud concentration profiles
(<20 mg/l) at a field site in Indonesia. The response of the ADCP showed a linear behaviour. The
interpretation of the ADCP signal was complicated by the presence of flocculated materials and by the
prevalence of anomalous scatters due to phytoplankton in the water column.
DRL software Ltd has tried to determine the suspended sediment concentrations from bottom-mounted
ADCP signals in USA coastal waters (near Cape Fear at Bald Head Island; August 2001). For comparison
data from other instruments were used (OBS, LISST, water samples). At each site a single calibration was
applied to the full data set. Many in-situ calibration surveys are required to interpret long-term monitoring
data to deal with the environmental variability of particle sizes, concentrations and other suspended
materials.
Merckelbach and Ridderinkhof (2005) have successfully used an acoustic backscatter ADCP instrument
and an optical backscatter sensor OBS to measure the suspended sediment concentration in the mouth of a
large-scale tidal inlet (Marsdiep, The Netherlands). The OBS sensor was carefully calibrated on the site
using in-situ water samples. The acoustic backscatter signal was found to consist of a Rayleigh
backscattering component proportional to SSC (suspended sediment concentration) and a backscatter
component generated by turbulence-induced variability of SSC. Comparing data from the OBS and ADCP, it
was found that the deviations between both time series were, on average, within 10 mg/l. Occasionally, the
deviations were as large as 20 mg/l.
Technical specifications of ABS (Aquascat) of Aquatecgroup (www.aquatecgroup.com)

frequencies: up to 4 frequencies, up to 5 MHz
transducers: typically 10 mm diameter ceramic cells
vertical range: 128 cm standaard
transmission rate: 1W transmitted pulse
data averaging: cell ensembles averaged over time by powers of 2 up to 64 before storage
range cells: 10 mm standaard
logging period: from 1 to 60 minutes
power requirement: 10V to 28V dc; typically 2W when logging, 4W when writing to disc
Advantages
1. non-intrusive profiler covering the lowest 1 meter of the water column where most of the suspended sand
transport takes place; bed detection included;
2. time-resolution of concentrations within wave cycle can be accurately determined;
3. particle sizes can be determined using multiple frequencies (calibration required);
4. usable in coastal environments (stand-alone tripod deployments);
5. almost no effects of biological fouling.
Disadvantages
1. in-situ calibration is required for accurate results; in-situ calibration is problematic in coastal
environments;
2. not suitable for upper concentration range (>10 gr/l);
3. measurement of velocity requires multiple sensors (cross-correlation);
4. sensitive to air bubbles and organic materials in the water column; not suitable for breaking wave
conditions (surf zone).
References
Coates, R.F.W., 1990. Underwater Acoustic Systems. Macmillan, Basingstoke.

Downing, A., Thorne, P.D., and Vincent, C.E., 1995. Backscattering from a suspension in the near
field of a piston transducer. Journal of the Acoustical Society of America, 97 (3), p. 1614-1620.
DRL Software Ltd, 2001. Monitoring of experiment disposal mound at Cape Fear: sediview
calibration of ADCPs and comparison with other measurement techniques. DRL Software,
Godalming, Surrey, UK (www.drl.com)
Grasmeijer, B.T., Dolphin, T., Vincent, C. and Kleinhans, M.G., 2005. Suspended sand
concentrations and transports in tidal flow with and without waves. Paper U in Sandpit book
ISBN90-800356-7-x, edited by Van Rijn et al. Aqua Publications, The Netherlands
(www.aquapublications.nl)
Green, M.O., 1996. Introducing ALICE , Water & Atmosphere (NIWA), 4 (2), p. 8-10.
Green, M.O. and Vincent, C.E., 1991. Field measurements of time-averaged suspended-sediment
profiles in a combined wave and current flow. In: Soulsby, R. and Bettess, E. (Editors), Sand
Transport in Rivers, Estuaries and the Sea. A.A. Balkeema, Rotterdam, p. 25-30.
Green, M.O., Vincent, C.E., McCave, I.N., Dickson, R.R., Rees, J.M., and Pearson, N.D., 1995. Storm
sediment transport: observations from the British North Sea shelf. Continental Shelf Research, 15,
p. 889-912.
Hay, A.E., 1983. On the remote acoustic detection of suspended sediment at long wavelengths. Journal
of Geophysical Research, 88 (C12), p. 7525-7542.
Hay, A.E. and Sheng, J., 1992. Vertical profiles of suspended sand concentration and size from
multifrequency acoustic backscatter. Journal of Geophysical Research, 97 (C10), p. 15661-15677.
Heyse, I., Chamley, H., De Batist, M., De Moor, G., De Schaepmeester, G., De Winne, E., Houthuys,
R., Lankneus, J., Marsset, T., Pichot, G., Pollentier, A., Porter, C., Stolk, A., Terwindt, J.,
Tessier, B., van Cauwenberghe, C., Van Wesenbeeck, V., Vincent, C., 1995. Sediment transport
and bedform mobility in a sandy shelf environment. In: Marine Science and Technologies. Second
MAST Days and Euromar Market. Project Reports. Weydert, M., Lipiatou, E., Goi, R.,
Fragakis, C., Bohle-Carbonell, M., Barthel, K.-G. (Eds.). Commission of European Communities,
Brussels, p. 1393-1408.
Hoitink, A.J.F. and Hoekstra, P., 2005. Observations of suspended sediment from ADCP and OBS
measured in a mud-dominated environment. Coastal Engineering, Vol. 52, p. 103-118
Huck, M.P. et al., 1999. Vertical and horizontal coherence length scales of suspended sediments. Coastal
Sediments, p. 225-240
Libicki, C., Bedford, K.W., and Lynch, J.F., 1989. The interpretation and evaluation of a 3-Mhz
acoustic backscatter device for measuring benthic boundary layer sediment dynamics. Journal of
the Acoustical Society of America, 85 (4), p. 1501-1511.
Merckelbach, L.M. and Ridderinkhof, H., 2005. Estimating suspended sediment concentration from
ADCP backscatterance at a site with strong tidal currents. Submitted to Ocean Dynamics,
(merckel@nioz.nl)
Osborne, P.D., Vincent, C.E. and Greenwood, B., 1994. Measurement of suspended sand
concentrations in the nearshore. Continental Shelf Research, Vol. 14, No. 23, p. 159-174
Smerdon, A.M., 1996. AQ59:C - ABS System User Manual. Aquatec Electronics Limited, Hartley
Wintney.
Smerdon, A.M., Rees, J.M. and Vincent, C.E., An acoustic backscatter instrument to measure near-bed
sediment processes. www.aquatecgroup.com
Soulsby, R.L., Atkins, R., Waters, C.B., and Oliver, N, 1991. Field measurements of suspended
sediment over sandwaves. In: Soulsby, R. and Bettess, E. (Editors), Sand Transport in Rivers,
Estuaries and the Sea. A.A. Balkeema, Rotterdam, p. 155-162.
Thorne, P.D. and Hardcastle, P.J., 1991. Application of acoustic backscattering to measuring
suspended sediment processes. In: Soulsby, R. and Bettess, E. (Editors), Sand Transport in Rivers,
Estuaries and the Sea. A. A. Balkeema, Rotterdam, p. 111-116.
Thorne, P.D. and Campbell, S.C., 1992. Backscattering by a suspension of spheres. Journal of the
Acoustical Society of America, 92 (2), Pt 1, p. 978-986.
Thorne, P.D., Vincent, C.E., Hardcastle, P.J., Rehman, S., and Pearson, N., 1991. Measuring
suspended sediment concentrations using acoustic backscatter devices. Marine Geology, 98, p. 7-16.
Thorne, P.D., Hayhurst, L., and Humphrey, V.F., 1992. Scattering by non-metallic spheres.
Ultrasonics, 30 (1), p. 15-20.
Thorne, P.D., Hardcastle, P.J. and Soulsby, R.L., 1993. Analysis of acoustic measurements of
suspended sediments. Journal of Geophysical Research, 88 (C1), p. 899-910.
Thorne, P.D., Manley, C., and Brimelow, J., 1993. Measurements of the form function and total
scattering cross section for a suspension of spheres. Journal of the Acoustical Society of America,
93 (1), p. 243-248.
Thorne, P.D., Waters, K.R. and Brudner, T.J., 1995. Acoustic measurements of scattering by objects
of irregular shape. Journal of the Acoustical Society of America, 97 (1), p. 242-251.
Thorne, P.D., Holdaway, G.P. and Hardcastle, P.J., 1995. Constraining acoustic backscatter estimates
of suspended sediment concentration profiles using the bed echo. Journal of the Acoustical Society
of America, 98 (4), p. 2280-2288.
Thorne, P.D. and Hardcastle, P.J., 1997. Acoustic measurements of suspended sediments in turbulent
currents and comparison with in-situ sampling. Journal of Acoustic Society Am., Vol. 101, p.
2603-2614
Thorne, P.D., Williams, J.J. and Davies, A.G., 2002. Suspended sediments under waves measured in a
large-scale flume facility. Journal of Physical Research, Vol. 107, No. C8, p. 4.1-4.16
Thorne, P.D. and Hanes, D.M., 2002. A review of acoustic measurement of small-scale sediment
processes. Continental Shelf Research, Vol. 22, p. 603-632
Van Hardenberg, B., Hay, A.E., Sheng, Y. and Bowen, A.J., 1991. Field measurements of the vertical
structure of suspended sediment. Coastal Sediments '91. A.S.C.E., New York, p. 300-312.
Vincent, C.E. and Green, M.O., 1991. Patterns of suspended sand. In: Soulsby, R. and Bettess, E.
(Editors), Sand Transport in Rivers, Estuaries and the Sea. A.A. Balkema, Rotterdam, p. 117-124.
Vincent, C.E., Hanes, D.M., and Bowen, A.J., 1991. Acoustic measurements of suspended sand on the
shoreface and the control of concentration by bed roughness. Marine Geology, 96, p. 1-18.
Vincent, C.M. et al., 1998. Sand suspension and transport on the Middelkerke Bank by storms and tidal
currents. Marine Geology
Vincent, C.M. and Green, M.O., 1999. The control of resuspension over megaripples on the continental
shelf. Coastal Sediments, p. 269-280
Visser, R., 1997. Relationship between the reflection of sound in water and suspended sediment
concentration (in Dutch). Kamminga BV, Zoetermeer, The Netherlands
Figure 1
Acoustic head of ABS instrument
Figure 2
Typical example of ABS profiles (height above bed on vertical axis) as function of time (on horizontal axis)
with high concentrations just above the bed.
Figure 3
ABS concentrations versus OBS concentrations; wave-averaged values of half wave cycles; Osborne et al.
(1994)
Figure 4
Left: Time-averaged sand concentration profiles calculated for F1, F2 , F4, F1-F2 pair simultaneously
and F1 -F4 pair simultaneously based on a mean sand size of 0.33 mm; Pacific East Coast New
Zealand
Right: Time-averaged values of suspended sand size derived from F1 -F2 signals and from F1 -F4
signals (Vincent and Green, 1999)
Figure 5
Sand concentration profiles in regular waves (left) and irregular waves (right), Thorne et al. (2002)
Figure 6
Regression plot of mean acoustic concentration and pumped sample concentration (calibration curve);
Thorne et al. (2002)
Figure 7
Sand concentration profiles from ADCP backscattered signal (lines) and water samples (solid squares) in
Western Scheldt (The Netherlands); concentration (in gr/l) on vertical axis, height above bed on horizontal
axis); Visser (1997)
5.6.5 Impact sensor
Impact probes are based on the momentum-transfer principle. The high density of sediment particles gives
them excess momentum over the surrounding water so that they tend to strike a transducer placed in the
stream rather than follow the path of the water particles. This effect discriminates between sand and silt
particles. Silt particles do not possess sufficient excess momentum to impact. The sand concentration can be
determined from the impact rate and the independently measured water velocity. The height of the voltage
pulse is proportional to the momentum of the sand grains so that a crude grain size distribution can be
obtained by pulse height discrimination. The impact transfers some of the sand particle momentum and
energy to an impact sensor which converts it to an electric signal. The output is a count of frequency of
impacts.
References
Salkfield, A.P., Le Good, G.P. and Soulsby, R.L., 1981. Impact Sensor for Measuring Suspended Sand
Concentration. Conf. on Electronics for Ocean Technology, Birmingham, England
5.6.5.2 IOS-impact sensor
Principle
The IOS-impact probe consists of a 1 x 10 mm stainless steel strip coupled to a slab of piezo-electric ceramic
(see Figure 1A). The ceramic slab is totally enclosed, protected and screened from electrical pick-up. The
body of the probe is formed from a stainless steel/glassepoxy laminate/stainless steel sandwich, covered in a
2-part epoxy coating.
The sand concentration can be derived from the impact rate and the (independently) measured velocity.
Thus, a separate instrument for measuring the fluid velocity is necessary.
Practical operation
1. lower (or raise) sensor to sampling position (use echo sounder)
2. select sampling period (u3e time-integrator)
3. read output signals
5. note data on measuring sheet

Probe calibration is necessary to obtain a relationship between count rate and sand concentration. A typical
field calibration (using pump sampler) is given in Figure 1B. The impacts per litre were calculated by
assuming that the geometric area of the probe was the effective active area and from simultaneously
measured velocities.
weight: 1 kg
response period: 0.2 s
cycle period: 5 min
Advantages
1. small sensor
2. small response period
Disadvantages
1. calibration required
2. not suitable for silt
Figure 1
IOS impact sensor
5.6.6 Nuclear sensor
Principle
Nuclear samplers for suspended sediment concentrations have been used in Russia, Hungary, Poland and
China. The principle is based on the absorption of radio-active energy by the sediment particles. In China the
nuclear sampler consists of a PU238 X-ray source. Other sources are Cs137, Am241 and Cd109. The radio-
activity is measured by (radiation) counters. Calibration is required. The concentration range is 0.3 to 1000
kg/m3 with an inaccuracy of 20% for low concentrations and 5% for high concentrations.
References
Basinski, T., 1989. Field Studies on Sand Movement in the Coastal Zone. Polska Akademia Nauk, Instytut
Budownictwa Wodnego, Gdansku, Poland
Liu Yu-Ren, 1987. Summary on the Development of Nuclear Suspended Sediment Flux Gauge HDD-l
Institute of Hydraulic Research, Yellow River Conservancy Commission, China
5.6.7 Conductivity sensor
Principle
Delft Hydraulics has developed a small-scale conductivity sensor (CCM) for measuring sand concentrations
in the high concentration regime (100 to 2000 kg/m3). The sensor (size of 0.01 m) measures the conductivity
of the fluid sediment mixture near the sensor points. The sensor has been used to measure sand
concentrations in the sheet flow layer close to the bed.
5. MEASURING INSTRUMENTS FOR SEDIMENT TRANSPORT
5.1 General aspects

In this chapter various instruments for measuring the sediment transport rate are described. Usually the
sediment transport is represented as the summation of the bed load and suspended load transport. Figure 1 of
Par. 2.1 can be used to get an estimate of the relative importance of both modes of transport.
To measure the bed load transport, two measuring methods are available. There are the simple mechanical
trap-type samplers, that collect the sediment particles transported close to the bed. Another possibility is the
recording of the bed profile as a function of time (bed-formtracking).
To measure the suspended load transport, a wide range of instruments is available from simple mechanical
samplers to sophisticated optical and acoustical samplers. Most samplers are used as point-
integrating samplers which means the measurement of the relevant parameters in a specific point
above the bed as a function of time. Some instruments are used as depth-integrating samplers, which means
continuous sampling over the water depth by lowering and raising the instrument at a constant transit rate.
All instruments are described in terms of their measuring principle, ractical operation, inaccuracy and
technical specifications.
To get a better understanding of the accuracy of the various instruments, special attention is given to
comparative measurements (see paragraph 5.4). The instruments are described in Par. 5.5.
5.2 Instrument characteristics

In this section the most important characteristics of the point-integrating suspended load samplers are
summarized. The results are presented in Figure 1. Hereafter, some of the characteristics are discussed.
Sampling period
The sampling period is the time period during which a sample is collected. For the bottle-type and trap-type
samplers the sampling period is equal to the filling period of the bottle or trap. The sampling period of the
Delft Bottle is restricted by the size of the sediment catch which should be small compared with the size of
the sedimentation chamber of the instrument. The sampling period of the pump-filter sampler is restricted by
the filter characteristics. A small 50 Pm-filter may be blocked rather easily, especially in a silty environment.
The sampling period of the optical and acoustical samplers is free. For time-averaging additional equipment
should be used allowing a digital reading.
Minimum cycle period

The minimum cycle period is the minimum time period between two successive measurements at adjacent
points in a vertical. In case of a free sampling period a minimum sampling period of about 5 minutes is used
to obtain the minimum cycle period, as given in Figure 1. The cycle period can be used to evaluate the time
period needed to cover a full concentration profile measurement. In case of tidal flow conditions this latter
period should be small in relation to the tidal period.
Overall accuracy
The overall accuracy is an estimate of the overall error of a single measurement due to (systematic and
random) measuring errors and stochastic (fluctuation) errors related to the physical process. The latter errors
are introduced by the stochastic fluctuations of the physical parameters to be measured. Figure 2 presents
some information of sediment concentration fluctuations, as observed by Eyster and Mahmood in some
Pakistan irrigation channels. They used a pint-bottle sampler yielding a time-averaged value over a five-
second period. Based on these results, each single bottle-measurement may have an error of about 100%.
This error can be reduced by using a larger sampling period or by collecting more samples. The overall
accuracy of the samplers with a free sampling period, as presented in Figure 1, is based on a relatively long
sampling period (say 5 minutes). The accuracy of the optical and acoustical samplers is largely dependent on
the number and accuracy of the calibration samples. More information of the inaccuracies involved can be
obtained from comparative measurements (see Paragraph 5.4).
References
Eyster, C.L. and Mahmood, K., 1976. Variation of Suspended Sediment in Sand Bed Channels
River Sedimentation, Vol. II, Fort Collins, USA

Suspended Suspended load Measured Measuring Response Sampling Minimum Overall accuracy
load Silt Sand Parameter Range Time Period Cycle Silt Sand
samplers (<50 Pm) (>50 period
Pm)
(mg/l) (s) (min) (min)
Mechanical Instruments
Bottle yes yes concentration >1 - 1 5 100% 100%
Trap yes yes concentration >1 instantaneous instan. 5 100% 100%
USP-61 yes yes concentration >1 - 1 5 100% 100%
transport
Delft Bottle no yes transport >10 - 5-30 10 - 50%
Pump-Filter no yes concentration >10 - 5-15 10 - 20%
Pump- no yes concentration >50 - 5-15 15 - 20%
Sedimentation
Pump Bottle yes yes concentration >1 - 1-5 5 20% 20%
Electronic Instruments
Optical-OBS yes no concentration 10-100000 <1 free 5 50% -
Optical- yes yes concentration 10-500 <1 free 5 30% 30%
LISST particle size
Pump- OBS yes no concentration 10-100000 - free 5 50% -
point ASTM transport
profiling ABS
Figure 1
Instrument characteristics for Sediment Concentration and Transport

Figure 2
Sand concentration as a function of time
Instrument characteristics (after Eyster and Mahmood, 1976)

5.3 Selection of sediment transport samplers
5.3.1 Guidelines for selection of sediment transport samplers
The selection of the most appropriate sampling technique should be based on a number of criteria, as
follows:
1. type of process/parameters to be measured:
- bed material;
- bed load;
- suspended load;
- particle size of bed material and suspended sediments;
- particle size of flocculated suspended sediments;
- fall velocity of sediments;
2. type of sampling environment:
- rivers, estuaries, coastal seas;
- sediments involved: mud, silt, sand, gravel, mixtures; flocculated materials;
- depth range and velocity range involved;
3. type of sampling:
- on-line measurements (short term) using hand-held instruments in shallow streams or instruments
attached to a survey ship (through a cable-winch system) or structure (bridge, platform) in the flow,
- stand-alone measurements (long-term) using instruments attached to frames, tripods or other structures
in the flow,
- point measurements using sensors that can measure in one point only (multiple points to determine
vertical distribution);
- profiling (or depth-integrated) measurements using instruments traversing (part of) the flow depth or
instrument signals (acoustics or optics) traversing the flow depth or part of the flow depth;
4. type of project and required accuracy:
- reconnaissance study to get a rough estimate of parameters/proccesses involved;
- process study to understand the details of the physical system (requiring turbulent fluxes);
- studies focussing on high spatial and time resolution (dredging and dumping plumes);
- input for mathematical modelling (boundary conditions) for design purposes;
- data for verification of models;
5. available instruments and available budget.
To select the most appropriate sampling instrument, quantitative information of the physical parameters to be
measured should be available prior to the actual field survey. It is important to have information of the
various transport modes at the sampling site such as the (relative) value of the wash and bed material load,
the bed load and the suspended load transport rates, the sediment concentrations (low or high) and the
particle size ranges of the suspended sediments (clay, silt, sand; flocculated sediments) involved. To get this
information, existing data sets should be analyzed or a reconnaissance study should be carried out.
Most measurements in rivers and estuaries are based on on-line data sampling using instruments attached to
a survey ship (through cable-winch system). Stand-alone measurements using a package of electronic
instruments attached to a frame, tripod or other structure placed on the bed or in the flow are conducted when
survey ships cannot be used due to the presence of surface waves (coastal envirionments) or when long-term
measurements (weeks to months) are required. The instruments usually are operated in burst mode, which
means that measurements of limited duration are taken at regular intervals in time (15 min per hour) to
reduce on data storage.
Examples of instrumented stand-alone measuring facilities are:
x HSM-tripod of University of Utrecht (Grasmeijer et al., 2005);
x AMF-tripod of Rijkswaterstaat (RIZA, 2001);
x STABLE tripod of Proudmann Oceanographic Laboratory POL (Williams et al., 2003).
The availability of these advanced facilities with electronic equipment greatly improves data quality and data
density over the traditional mechanical equipment, and long-term costs will be reduced, as there is no longer
need for intensive sample processing and analysis in the laboratory.

Another important criterion is the purpose of the study and the accuracy required. For example, the predicted
deposition volumes of a planned harbour approach channel should be rather accurate when the maintenance
dredging costs of the channel are critical with respect to the economic feasibility of the project. In that case
the most accurate instruments and the most sophisticated facilities should be used. Although these type of
measurements are expensive, the overall costs of field surveys are only a fraction of the total project costs.
When research (process) studies are performed, it is often required to measure the time-series of the
fluctuating parameters so that the turbulent fluxes can be determined. For engineering studies it is usually
sufficient to measure the time-averaged velocities and sediment concentrations. Simple instruments such as
the bottle and trap samplers can then be used, although the analysis costs of the many samples involved are
relatively high.
5.3.2 Sediment transport measurements in rivers

x bed load transport,
x suspended sediment concentrations and transport rates,
x particle sizes and fall velocities,
are presented in Tables 1,2 and 3 below; order of preference is based on the overall sampling accuracy.
Simple mechanical instruments such as the bottle-type, the trap-type and the pump-type samplers are still
very attractive because of their robustness and easy handling, particularly when used at isolated field sites.
The accuracy of the measured parameters involved can increased by increasing the number of samples
collected. Analysis costs of all samples involved may be critical with respect to the available budget. Optical
and acoustic instruments are attractive when large numbers of data have to be collected. As calibration is
involved, the accuracy strongly depends on the quality/reliability of the calibration curves. Hence, many
calibration samples are required using a pump sampler with the nozzle as close as possible to the
optical/acoustic sensor.
A major technological advance for measuring suspended load transport is the in-situ Laser diffraction
instrument (LISST). This instrument can measure the particle size distribution and sediment concentration
simultaneously. An attractive solution is the LISST-ST, which includes a streamlined body to improve
sampling accuracy and is equipped with a pressure sensor and current meter for measuring the sampling
height above the bed and the ambient velocity. The velocity data are used to control pump sampling
(isokinetic sampling) across the internal Laser arrangement. The velocity and concentration data are used to
compute fluxes for up to 32 particle size classes at points, verticals, or in the entire stream cross-section. All
data are transmitted via a cable to the survey vessel (on-line measurement). Limitations (related to light
penetration) are the maximum concentration ranges of about 150 mg/l for fines (mud/silt) and 500 mg/l for
sand particles. Hence, the instrument cannot be used in high-concentration conditions (close to bed; upper
flow regime).
Selection guidelines for sampling instruments in rivers focussing on the typical mechanical US-instruments
(see Figure 1) are given by the U.S. Geological Survey (Davis, 2005). This report and many other reports of
various typical US-samplers can be downloaded from the FISP-site (http://fisp.wes.army.mil). These US-
samplers are manufactured by Rickly Hydrological Company (www.rickly.com).
The manuals: Field methods for measurement of fluvial sediment (Edwards and Glysson, 1999) and Fluvial
sediment concepts (Guy) can be downloaded from USGS-site (www.usgs.gov).

Bed load transport

Silt Bed form tracking On-line factor 2
Sand/Gravel Bed-form tracking On-line 50% (many tracks)
BTMA trap sampler On-line factor 2 (many samples required)
Helley-Smith trap On-line factor 2 (many samples required)
sampler
Delft Nile trap sampler On-line factor 2 (many samples required)
Table 1
Characteristics of bed-load sampling methods in rivers

Mud/Silt Pump-Bottle On-line 20% (bottle filled by pumping during 15
(hand-held in depth<1m) close to bed
(hand-held in depth<1m) not close to bed
Stand-alone
Optical (OBS) On-line 30%-50% (depending on number of
Stand-alone calibration samples)
close to bed
not close to bed
Stand-alone
Delft Bottle On-line 50% (many samples required;
undersampling of finer particles); not
close to bed
Table 2
Characteristics of suspended load sampling methods in rivers


(size and settling sediments; lower limit of about 20 Pm)
flocs)
In-situ Laser- On-line 20%-30% (lower limit of about 5 Pm and
diffraction Stand-alone upper limit of about 500 Pm)
flocs)
In-sity Laser- On-line 20%-30% (lower limit of about 5 Pm and
diffraction + Stand-alone upper limit of about 500 Pm )
(size and settling
reflectance Stand-alone upper limit of about 500 Pm )
for flocs)
Sand In-situ Laser- On-line 20%-30% (lower limit of about 5 Pm and
Table 3
Characteristics of sampling methods for particle size and fall velocity in rivers

5.3.3 Sediment transport measurements in estuaries

Simple mechanical instruments such as the bottle-type and the trap-type samplers are not attractive because
of the very short sampling times involved. Accuracy cannot be improved by increasing number of samples
due to time-variation of sediment concentrations within the tidal cycle.
Point-samples should be taken over the entire water column in strong tidal flows as the sediments will be
mixed over the water column by turbulent eddies. Data sampling can be confined to the bottom region in
weak tidal flows. Flocculation often is a dominant process in muddy estuaries. The LISST-ST which is an
in-situ Laser diffraction instrument in combination with a settling tube offers a powerful solution to measure
particle sizes, concentrations and densities of the individual particles as well as the flocculated aggregates
(low concentrations <150 mg/l) in on-line mode. This instrument (LISST-ST) is not yet suitable for long-
term, stand-alone measurements due to insufficient robustness, relatively low concentration range (<150
mg/l) and biological fouling problems.
Bed load transport

Silt Bed-form tracking On-line factor 2 to 3 due to limited number
of tracks
Sand/Gravel Bed-form tracking On-line factor 2 to 3 due to limited number
of tracks
Delft Nile trap sampler On-line factor 2 to 3 (limited number of
samples)
Table 1 Characteristics of bed-load sampling methods in estuaries

Mud/Silt Pump-Bottle On-line 20%-30% (bottle filled by pumping
during 15 minutes on board of survey
ship)
Stand-alone
Optical (OBS) On-line 50% (depending on number of calibration
Stand-alone samples)
Bottles/Traps On-line factor 2 (limited number of samples ); not
close to bed
Stand-alone
Bottles/Traps On-line factor 2 to 3 (limited number of samples);
not close to bed
Table 2 Characteristics of suspended load sampling methods in estuaries


flocs)
flocs)
In-sity Laser- On-line 20%-30% (lower limit of about 5 Pm and
diffraction + Stand-alone upper limit of about 500 Pm )
(size and settling
for flocs)
Table 3
Characteristics of sampling methods for particle size and fall velocity in estuaries

5.3.4 Sediment transport measurements in coastal seas

Field instruments to measure bed-load transport in oscillatory flow are hardly available. Katori (1983) using
a stand-alone mechanical trap-type sampler, performed experiments to measure the cross-shore sediment
transport rate at and near the bed on a sandy beach at the Pacific Ocean coast of Japan (shallow surf zone).
Instruments available for measuring suspended sediment concentrations and transport in coastal
environments are: mechanical traps (streamer traps in shallow surf zone <1 m), pump samplers, optical
samplers and acoustic samplers.
Mechanical trap samplers have been used extensively in Japan, in Poland (Antsyferov et al, 1990), in the
USA (Kraus, 1987) and in Russia (Antsyferov et al, 1990). These types of instruments, which can only be
used in the shallow surf zone (depths <1 m) by operators wading in the water, are simple and reliable.
Simultaneous measurements in many points and stations can be performed. Kraus (1987) used portable
traps, consisting of long rectangular bags of polyester sieve cloth mounted on a vertical rack to measure
longshore suspended load transport rates. An operator standing downcurrent attends the trap during the
sampling interval (5 to 10 min), The use of the trap is restricted to surf zones with wave heights less than
about 0.5 m in water depths upto l m. Antsyferov et al (1990) used traps mounted on tripods placed in and
outside the surf zone. The traps were replaced by divers. Typical sampling periods are ranging from 10 min
to 1 day. The sediment mass collected by each trap gives an indication of the local time-averaged
concentration. Determination of the trapping coefficient is problematic. Therefore, relative concentrations
rather than absolute concentrations are obtained. Kana (1979) used an instantaneous trap sampler in the surf
zone. Many samples at the same location are required to eliminate the random fluctuations.
Pump samplers were used by many researchers to measure time-averaged sediment concentrations. These
types of samplers can only be used from a pier or platform. The intake nozzles should be directed
downwards.
Optical and acoustic probes are available to measure instantaneous sediment concentrations from a pier or
platform or from a stand-alone tripod. Data transmission can take place by telemetry or on-line to a
computer or datalogger. Calibration of the probes is required. Optical probes cannot be used in conditions
with both sand and silt particles in suspension. The optical instruments are relatively sensitive to fine mud
particles. Hence, the mud background concentration must be small (<50 mg/1). Otherwise, the sand
concentrations cannot be measured accurately.
Acoustic probes cannot be used in plunging breaking wave conditions due to the presence of air bubbles.
Nuclear probes which have been used in Russia and in China, cannot be used in low-energy conditions
where the concentrations are relatively small. The threshold concentration is of the order of 500 mg/1.
Suspended sediment transport measurements in conditions with combined current and wave conditions
cannot be performed from moored or sailing survey ships. Two options are possible (see Figure 1):
1. on-line sampling from piers connected to shore, platforms resting on seabed or sledges/trailers towed by
vehicles (only in shallow surf zone);
2. stand-alone sampling from frames/tripods/poles on/in the seabed or from drift bouys (profiling mode
from surface to bed) using a package of sophisticated electronic sensors (electromagnetic and acoustic
flowmeters, optical and acoustic backscattering sediment concentration meters); see Figures 3 and 4.
A typical instrument package attached to a stand-alone tripod in coastal environments is, as follows:
x electromagnetic point velocitymeters (EMV), acoustic Doppler point velocitymeters (ADV) and acoustic
Doppler Current Profilers (ADCP);

x optical backscatter point sensors (OBS) for silt/mud; acoustic backscatter point sensors (ASTM) and
acoustic backscatter profiler (ABS) for sand;
x in-situ Laser diffraction sensors (LISST) for particle size;
x acoustic bed profiler (line and sector scanning) for bed and ripple pattern detection.
A problem of all stand-alone facilities is the wave and current-induced scour near the legs of the structures
involved and the scour underneath and around the instrument sensors. Black and Rosenberg (1991) have
found by camera observations that a minimum distance of about 0.05 m should be kept between the probe
(sensor) and the bed in oscillatory flow conditions to prevent scour of bed material. The free span of the
instrument supports and legs should always be as large as possible to allow measurements between the legs
with minimum disturbance. A dramatic example of a poorly designed research pier is the Duck Pier (North
Carolina, USA) with a longshore pile spacing is 4.6 m and cross-shore spacing is 12.2 m, which has caused
considerable scour in the vicinity of the pier. An example of a slender pier with cross-shore pile spacings of
36 m is the Sally Wharf near Bordeaux, France (Figure 2).
Stand-alone tripods (Figures 2, 3, 4) are placed on the seabed from a vessel during calm conditions. They are
vulnerable in high wave conditions (overturning) or in areas with intensive fishing activities. Another
problem is the absence of control of the measuring elevations of the instruments in conditions with rapidly
changing bed levels (surf zone). The instruments can be damaged easily when placed close to the bed.
Williams et al. (2003) concluded that the turbulence levels underneath their tripod was slightly increased.
The bed-shear stresses were also slightly over-estimated (15%). Evidence of scour in the immediate vicinity
of the legs was detected, but was sufficiently localised not to effect the ripples beneath the main instrument
package on the tripod.
Piles/poles driven or washed into the bed offer a simple solution for hydrodynamic measurements of waves
and winds, but cannot be used for measurements of velocities and sediment concentrations due to the
influence of vortices and scour generated by the pile/pole construction.
Towed sledges or trailers with a mast above the water surface have been used for cross-shore bed level
soundings and wave height measurements in the surf zone during storms. However, they cannot be used for
measurements of velocities and sediment concentrations due to the disturbing influence of vortices generated
by the structure itself.
Reviews of field instrumentation for the coastal environment have been given by Terwindt et al. (1992),
Basinski (1989) and White (1998).

Bed load transport

Sand Bed-form tracking Sand-alone factor 2 (in deep water with non-breaking
of ripples using waves)
acoustic profilers factor 2 to 3 in surf zone
(side scan sonar)
Table 1
Characteristics of bed-load sampling methods in coastal seas

Mud/Silt Optical (OBS) On-line factor 2 to 3 (no in-situ calibration
Stand-alone samples available)
Sand Mechanical trap- On-line factor 2 to 3 (only usable in shallow surf
type sampler zone <1 m)
(streamer traps)
Pump sampler On-line 50% (only usable for longshore transport
from pier or platform)
Acoustic On-line factor 2 to 3 (no in-situ calibration
(ASTM, ABS) Stand-alone samples available)
Table 2
Characteristics of suspended load sampling methods in coastal seas

Mud/Silt In-situ videocamera On-line 50% (undersampling of very fine
velocity; mainly VIS, INSSEV (outside surf zone)
flocs)
(particle size incl. LISST-100, LISST-25
flocs) (outside surf zone)
(particle size; not PARTEC (outside surf zone)
for flocs)
(particle size) LISST-100, LISST-25 (outside surf zone)
(particle size) PARTEC (outside surf zone)
Table 3
Characteristics of sampling methods for particle size and fall velocity in coastal seas

References
Antsyferov, S.M., Belberov, Z.K. and Massel, S., 1990. Dynamical Processes in Coastal Regions
Publishing House, Bulgarian Academy of Sciences.
Basinski, T., 1989. Field studies on sand movement in the coastal zone. Polska Akademia Nank, Instytut
Budownictwa Wodnego, Gdansk, Poland
Black, K.P. and Rosenberg, M.A., 1991. Hydrodynamics and Sediment Dynamics in Wave-Driven
Environments Technical Report No. 13, Victorian Institute of Marine Science, Victoria, Australia.
Davis, B.E., 2005. A guide to the proper selection and use of federally approved sediment and water-quality
samplers. Report QQFederal Interagency Sedimentation Project (FISP), Waterways Experiment
Station, Vickburg, USA (http: //fisp.wes.army.mil)
Edwards, T.K. and Glysson, G.D., 1999. Field methods for measurements of fluvial sediment. Techniques
of water resources investigations of the US Geological Survey, Book 3, Applications of Hydraulics,
Chapter C2. USGS, Box 25286, Denver CO 80225, USA
Grasmeijer, B.T. et al., 2005. Description of measurements and database of field and laboratory data,
Paper M. In: Sandpit edited by Van Rijn et al., 2005. ISBN 90-800356-7-X (www.aquapublications.nl)
Guy, H.P., Fluvial sediment concepts. Techniques of water resources investigations of the US Geological
Survey, Book 3, Applications of Hydraulics, Chapter C1. USGS, Box 25286, Denver CO 80225, USA
Hemsley, J.M., McGehee, D.D. and Kucharski, W.M., 1991. Nearshore oceanographic measurements:
Hints on how to make them. Journal of Coastal Research, Vol. 7, No. 2
Kana, T.W., 1979. Suspended Sediment in Breaking Waves. Techn, Report No. 18-CRD, Coastal Res. Div.,
Dep. of Geology, Univ. of South Carolina, Colombia, USA
Katori, S., 1983. Measurement of Sediment Transport by Streamer Trap (In Japanese). Report No. 17, TR-
Kraus, N.C., 1987. Application of Portable Traps for Obtaining Point Measurements of Sediment
Transport Rates in the Surf Zone. Journal of Coastal Research, Vol. 3, No. 2
Rijkswaterstaat/RIZA, 2001. Autonomous Measuring Facility for measurement of water and sediment
movements (in Dutch). Report RIZA 2001.073X, Lelystad, The Netherlands
Terwindt, J.H.J, Greenwood, B.J. and Short, A., 1992. Users report on instruments for coastal research.
Dep. of Physical Geography, University of Utrecht, Utrecht, The Netherlands
White, T.E., 1998. Status of measurement techniques for coastal sediment transport. Coastal Engineering,
35, p. 17-45
Williams, J.J. et al., 2003. Interactions between a benthic tripod and waves on a sandy bed. Continental
Shelf Research, Vol. 23, p. 355-375

Figure 1
Selection diagram for typical US-suspended samplers (Davis, 2005)
Figure 2
Instrument facilities for coastal environments

Figure 3
AMF tripod of Rijkswaterstaat/RIZA (2001), The Netherlands
Figure 4
HSM tripod of University of Utrecht (Grasmeijer et al., 2005), The Netherlands

5.4 Comparison of suspended load samplers
5.4.1 Comparison of USP-61, Delft Bottle and Pump-Filter sampler
During May 1979 a field investigation using the USP-61, the Delft Bottle and the Pump-Filter Sampler was
carried out in the Danube River near Ilok, Yugoslavia (Dijkman, 1982). At this location the measuring
section is straight, while the cross-section is rather wide and regular. The mean width was about 560 m and
the mean depth was about 8 m. The water temperature varied from 14-18C. The instruments are described
in Paragraph 5.6.
The three instruments were used at the same level above the bed within a few metres from each other. The
(vertical) positioning was done by means of an echo-sounder present on each instrument. Figure 1 shows the
results of a relative comparison of the instruments.
Figures 2 and 3 show typical results of the sediment transport per grain-size fraction measured by each
instrument. The results can be summarized, as follows:
USP-61
The "measured" sediment transport shows rather large fluctuations, even for the silt particles (< 70 Pm). The
fluctuations are caused by the relatively small sampling time of about 30 seconds. For individual samples the
accuracy may be as large as r50%. To obtain a reliable average value in a statistical sense, at least 10
samples must be collected. In tidal flow conditions, where the concentrations vary as a function of time, the
USP-61 is not practical.
Delft Bottle
The measured values for each grain-size fraction were corrected using standard calibration curves, as given
in Paragraph 5.6.2.5. Although the sampling time was rather long (600-900 s), the fluctuations in the
measured transport rates are still rather large. Probably, these are caused by sediment losses during the
hoisting of the instrument.
Compared with the USP-61, the results of the Delft Bottle are systematically smaller, particularly for
the 70-150 Pm grain-size fraction. For all measurements the average differences between the Delft Bottle
and the USP-61 Bottle varies from about 10%-100%, depending on the grain-size fraction and the nozzle
type (Figure 1).
Pump-Filter Sampler
The water-sediment mixture was pumped through 50 Pm filters. Laboratory analysis showed that almost all
particles smaller than 70 Pm were lost. Figures 2 and 3 show relatively small fluctuations, which means that
a sampling period of about 300 s is sufficiently large to obtain a reliable average value. The mean value of
the sand transport (> 70 um) measured by means of the Pump-Filter Sampler agrees rather well with the
USP-61 Bottle. For all measurements the average difference between the USP-61 and the Pump Filter
Sampler varies from 5%-15% (Figure 1).
References
Laboratory - Jaroslav Cerni Institute, Report S410, The Netherlands

Figure 1
___ ___ ___ Delft Bottle (small nozzle), measuring time=600-900 s


Figure 2

___ . ___ . _ Delft Bottle (big nozzle), measuring time= 600-900 s
___ ___ ___ Delft Bottle (small nozzle), measuring time= 600-900 s

Figure 3

5.4.2 Comparison of Pump-Filter sampler and ASTM
The measurements were carried out in the Eastern Scheldt, a wide tidal estuary in the south-west part of the
Netherlands. The local water depth was about 10 m, the flow velocities ranged from 0.4 - 1.0 m/s. The local
bed material consisted of sand with D50 = 200 Pm. The intake nozzle of the pump-filter sampler was installed
next to the sensor of the acoustic sensor ASTM (AZTM in Dutch). The instruments are described in
Paragraph 5.6. The flow velocities and sediment concentrations were measured at a level of 1 m above the
bed during flood tide.
The results, shown in Figure 1, show reasonably good agreement. The maximum deviation is about 30% for
small concentrations (< 50 mg/1).
References
Figure 1
Comparison of sand concentrations measured with Pump-Filter Sampler and acoustic AZTM Sampler

5.4.3 Comparison of Pump-Filter sampler and Pump-Bottle sampler
Sediment concentration measurements were carried out in the Oude Maas, a side branch of the Nieuwe
Waterweg (New Waterway) near Rotterdam, the Netherlands. The local flow depth was about 10 m, while
the tidal flow velocities varied from 0-1 m/s. The bed material consisted of sand particles with D = 200 Pm.
The pump-filter sampler was operated with filters of 50 Pm. The sampling period in each point was 5
minutes. Simultaneously a pump-bottle sampler was used to determine the sediment concentrations. This
method consisted of the filling of 1 liter-bottles by a deck-mounted pump. The intake nozzle of the pump-
bottle system was installed next to the nozzle of the pump-filter system. The bottles were returned to the
laboratory for analysis. Figure 1 shows the sand concentrations (> 50 Pm) measured with both systems. As
can be observed, the sand concentrations measured with the pump-filter sampler are systematically smaller
than those measured with the pump-bottle method, which is, probably, caused by the loss of sediment
particles through the filter material. Another cause for deviations may be the relatively short filling
(sampling) period of the bottles, being about 1-2 minutes, compared with the sampling period of 5 minutes
for the pump-filter sampler. On the average, the deviations are about 20-30%.
References
Figure 1
Comparison of sand concentrations measured with Pump-Filter Sampler and Pump Bottle Sampler

5.4.4 Comparison of Pump-Sedimentation sampler and Pump-Filter sampler
Concentration measurements were carried out in the Nieuwe Waterweg (New Waterway) near Rotterdam,
the Netherlands. The local flow depth varied from 15-20 m; the (tidal) flow velocities were upto 1.5 m/s. The
bed material consisted of fine sand with D50 = 280 Pm.
The intake nozzles of the pump sedimentation sampler and the pump filter sampler were installed next to
each other on a sampling carrier. Both instruments are described in Paragraph 5.6.
The sampling period for the pump sedimentation system was about 3 minutes, while a settling period of 5
minutes was used. The sampling period for the pump filter system was only 1-2 minutes due to rapid filter
blocking (silt particles). Figure 1 shows the sand concentrations (particles > 50 Pm) at 1.0 m above the bed
measured with both systems. The overall agreement is quite satisfactory. On the average, the deviations are
about 20-30%. It can also be observed that the scatter in the concentrations measured with the pump filter
system are somewhat larger than those measured with the pump sedimentation system which is probably
caused by the relatively small sampling period of the pump filter system.
References
Van Rijn, L.C., 1980. Methods for in-situ separation of water and sediment. Delft Hydraulics, Report S404
II, The Netherlands
Figure 1

5.4.5 Comparison of Pump-Sedimentation sampler and Bottle sampler
Sediment concentration measurements were carried out in the Westerschelde (Western Scheldt), a tidal
estuary in the south-west part of the Netherlands. The local flow depth was about 15 m; the tidal flow
velocities were upto 1.5 m/s. The bed material consisted of sand particles (D50= 250 Pm). The Bottle sampler
consisted of the filling of bottles.
One-liter bottles were used, which were attached under an angle of 35 with the vertical direction (see Par.
5.6.2.2) to a frame that was lowered to the channel bed. The exact angle of the bottle with the flow direction
under water was unknown (not measured). The bottles were opened by pulling a rope from the survey vessel.
The intake nozzle of the pump sedimentation sampler was installed close to the opening of the bottle. After
each measurement the frame was hoisted out of the water to replace the bottle.
The sampling period of the pump sedimentation system was about 1.5-2 minutes. The-filling (sampling)
period of the bottle is unknown but is assumed to vary from 0.5-1.5 minutes.
Figure 1A shows the sand concentrations (particles > 50 Pm) measured with both systems. The results show
large deviations upto 150%. On the average, the sand concentrations measured with the pump sedimentation
sampler are 30-40% smaller. For concentrations smaller than 50 mg/1 the deviations can be explained by
inefficiency of the pump sedimentation method, but for high sand concentrations the loss of small particles
for the pump sedimentation method is rather small (Par. 5.6.3.4). Further, it can be observed that the sand
concentrations measured with the bottle method show a relatively large scatter.
Based on these results and those of the flume tests (Par. 5.6.2.2) it may be questioned if the (simple) bottle
method is a reliable method for measuring sand concentrations.
Figure 1B shows the silt concentrations (particles < 50 Pm) measured with both methods. These results show
rather good agreement. On the average the silt particles measured with the pump method are about 10%
smaller.
References
Van Rijn, L.C., 1983. Sediment transport in the Western Scheldt (in Dutch). Delft Hydraulics Laboratory,
Report R1586, The Netherlands

Figure 1

5.4.6 Comparison of OBS and Pump sampler
Instruments
Grasmeijer (in Walstra et al, 1998) did measurements over a sand bed in the Large Oscillating Water
Tunnel (LOWT) of Delft Hydraulics using the optical sensor OBS and a pump sampling system. The LOWT
is described in Section 2.2. The optical OBS sensors were calibrated using the sand bed material in the
LOWT (d50 = 0.12-0.13 mm and 0.19-0.21 mm; d50 varied slightly based on samples before and after the
tests). Three OBS-sensors have been attached to a vertical rod on a footplate (see photographs above); an
EMF velocity probe is also connected to the rod. The footplate is resting on the bed and can move
downwards (due to the movable instrument arrangement) when the bed surface is eroding. Using this
arrangement, an approximately constant distance can be maintained between the bed and the measurement
elevations of both EMF and OBS-sensors during a short measurement period of say 15 to 30 minutes The
EMF-sensor is placed at 0.05 m above the upper side of the footplate. The OBS-sensors are placed at
distances of 0.03, 0.05 and 0.1 m above the upper side of the footplate. The effective measurement elevations
of the OBS-sensors with respect to the surrounding bed surface are approximately 0.025, 0.045 and 0.095 m,
as the instrument will sink down into the bed over about 0.005 m (due to local erosion).
The pump sampler consisted of a vertical array of 10 intake tubes of 3 mm internal diameter connected to the
pumps by plastic hoses. The lowest intake tubes were placed at about 0.01 m above the bed, with the intake
openings placed in a direction transverse to the plane of orbital motion. The intake velocity was about 1 m/s,
satisfying sampling requirements. The 10 liter samples were collected in calibrated buckets. The pump
sampler was operating for 15 minutes giving an average concentration over the measuring time. In case of
fast bed level variations due to large transport rates the operation time of the pump sampler was reduced to 8
minutes giving 5 liter samples. The suspended sand samples at each level above the bed were analyzed to
determine the size composition of the sand.
It was found from the suspended sand samples that in case of relatively coarse bed material (d50 = 0.19-0.21
mm), the median diameter of the suspended sand at z = 0.025 m and z = 0.095 m above the bed is
approximately 20%, respectively 30% smaller than the median grain size of the bed material. In case of
relatively fine bed material (d50 = 0.12-0.13 mm) the vertical sorting was less pronounced. For nearly all tests
with fine sand and at all three heights above the bed the median diameter of the suspended sand was
approximately 10% smaller than the median grain size of the bed material.
To take the effect of a varying grain size with height above the bed into account, the OBS concentrations
were determined from the calibration curve using the particle size of the suspended sand at the same level as
the OBS.
Experimental programme
Two different test series were performed using two different sediment sizes with a median diameter of 0.19-
0.21 mm and 0.12-0.13 mm respectively. Regular asymmetric wave motion (second order Stokes) and
irregular wave motion were generated. Orbital velocities were measured by means of a Laser Doppler
velocity meter at about 0.1 m and 0.2 m above the bed. Most tests were repeated twice to determine the
variation related to small differences in bed arrangement (refilling of sand in the tunnel).
The test procedure was as follows:
x positioning of OBS transport meter on bed; foot plate flush with bed surface;
x positioning of vertical array of intake tubes with lowest intake tube at 0.01 m above bed;
x generation of oscillatory flow during 10 to 20 min.;
x establishment of equilibrium bed conditions during about 10 min;
x sampling of water-sediment mixture through each intake tube after establishment of equilibrium
conditions;
x separation of water and sand; determination of wet and dry sand volumes of samples.
Results
Figures 1 and 2 show examples of time-averaged concentration profiles measured with OBS sensors and the
pump sampler in the Large Oscillating Water Tunnel. It can be observed that the concentrations measured
with the OBS and the concentrations measured with the pump sampler are of the same order of magnitude.
On average, the OBS sensors gave values that were 15% larger (two largest deviations are 250% and -70%)

in case of coarse sand (0.19-0.21 mm) and 30% larger (two largest deviations are 150% and -50%) in case of
fine sand (0.12-0.13 mm) than the values determined with the pump sampler. The largest differences were
found for test B2 and H2 in which the concentration gradient is relatively large (fine sand, small orbital
velocity and weak current compared to other tests).
Values from the OBS sensors showed favorable comparison to values from the pump sampler in case of
relatively large sand concentrations (larger than about 1 kg/m3).
References
Walstra, D.J.R., Van Rijn, L.C., Aarninkhof, S.G.J., 1998. Sand transport at the middle and lower
shoreface of the Dutch coast. Report Z2378, Delft Hydraulics, the Netherlands.

Concentrations
test B00-1, B00-2 and B00-3
0.30
B00-1 pump sampler

0.25
B00-1 OBS
height above bed (m) B00-2 pump sampler
0.20 B00-2 OBS
B00-3 pump sampler
0.15 B00-3 OBS
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Concentrations
test B2-2, B2-3, B2-4 and B2-5
0.30
B2-2 pump sampler
0.25 B2-2 OBS
B2-3 pump sampler
0.20 B2-3 OBS

B2-4 pump sampler
0.15 B2-4 OBS

B2-5 pump sampler
B2-5 OBS
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Figure 1
B00: reg. asym., U1/3,on = 0.52 m/s, U1/3,off = 0.33 m/s, d50 = 0.19-0.21 mm
B2: irreg. asym, U1/3,on = 0.84 m/s, U1/3,off = 0.48 m/s, d50 = 0.19-0.21 mm

Concentrations
0.30
H5-1 pump sampler
0.25 H5-1 OBS

H5-2 pump sampler

H5-2 OBS
0.20
H5-3 pump sampler
H5-3 OBS
0.15
0.10
0.05
0.00
0.01 0.1 1 10 100
concentration (g/l)
Concentrations
0.30
H8-1 pump sampler
0.25 H8-1 OBS
H8-2 pump sampler
H8-2 OBS
0.20
H8-3 pump sampler
H8-3 OBS
0.15
0.10
0.05
0.00
0.01 0.1 1 10
concentration (g/l)
Figure 2
H5: reg. symm, Umax = 1.30 m/s, d50 = 0.12-13 mm
H8: reg. symm,Umax = 0.67 m/s, d50 = 0.12-0.13 mm

5.4.7 Comparison of ASTM and Pump sampler
Experimental setup
Special experiments (Chung and Grasmeijer, 1999) have been performed in a large-scale flume of Delft
Hydraulics, in which a horizontal sand bed was placed over a length of about 40 m (Figure 4.11). The flume
has a total length of 233 meters, a depth of 7 meters and a width of 5 meters. A piston activated wave board
on one side of the flume generates the waves.
A sand bed (r 0.5 meter high) was placed in the Delta flume from position x = 100 meters to x = 140 meters.
During the first test series (July 1997) this sand bed had a d50 of 0.33 mm, and during the second test series
(August 1997) the d50 was 0.16 mm. The waterdepth was 4.55 m in all experiments.
The acoustic sand transportmeter ASTM was mounted in a tripod, which was placed on the sand bed at
location x = 125 m (Figure 1). The ASTM was used to measure the instantaneous fluid velocities and sand
concentrations at five points above the bed simultaneously. The ASTM was attached to an in vertical
direction movable arm to position the sensors at a known level above the bed. In most tests the measurement
levels were: z= 0.075, 0.125, 0.225, 0.475 and 1.075 m above the bed. A pump sampler (PS) was used to
determine the time-averaged concentrations and to get suspended sand samples at the same levels. Five
intake tubes of the pump sampler were attached at the ASTM sensors (horizontally within 0.2 m; vertically
within 0.01 m from the measurement volume of ASTM) and five other intake tubes were attached to a
supporting rod outside the tripod on the upwave side of it to study the effect of the tripod and ASTM
arrangement on the sand concentrations. These latter intake nozzles of the pump sampling unit were within 2
m of the ASTM sensors (at the same levels above the bed). During each test the instruments were operated
for about 15 minutes to sample over a representative wave record.
Results
Sand concentration profiles based on the pump sampler near the ASTM sensors inside the tripod (these pump
samples were used to calibrate the ASTM) and the pump sampler outside the tripod are presented in Figures
2 and 3 for the tests over a sand bed of 0.16 mm. The concentrations measured with the pump sampling
system were averaged over 7 to 9 tests. Only results for conditions with irregular waves are shown. Standard
deviations in concentration are represented by x-error bars and possible deviations in bed level (due to ripple
migration) are represented by y-error bars. The ripple height was found to be approximately 0.07 m.
All measured concentrations are within a factor 2 of each other without the presence of systematic
differences, which is a remarkably good result. Differences in concentrations can be caused by:
x scour and extra turbulence induced by the supporting structure and footplate of the pump sampling system
outside the tripod,
x extra turbulence induced by ASTM sensors and supports of tripod,
x small local variations in bed forms and bed level inside and outside the tripod.
Turbulence related to the presence of the tripod and the ASTM sensors may produce larger concentrations
near the bed (more sediment is being stirred up). Also, larger concentrations at higher elevations above the
bed may be expected because of an increased turbulent mixing (sediment is distributed more easily by
vortices induced by the presence of the sensors).
Figures 4 and 5 show sand concentration profiles based on the pump sampler near the ASTM sensors inside
the tripod and the pump sampler attached to the flume wall for the tests over a sand bed of 0.33 mm. The
concentrations near the ASTM sensors are systematically larger (factor 2) than the concentrations measured
near the flume wall. Higher up in the water column these effects are supposed to be related to some extra
turbulence induced by the instrumental arrangement. In the near-bed region the observed differences in
concentrations are most likely be caused by differences in bed levels and differences in bed form dimensions
between the location of the pump sampler near the flume wall and the tripod location in the centre of the
flume. From visual observations of the sand bed after draining of the flume it became clear that the presence
of the pump sampling system near the flume wall had caused a scour hole at that location. At locations more
landward and more seaward of the pump sampler near the wall (approximately 1 m from the pump sampler),
an increase in bed level was found (local bump). In the centre of the flume near the ASTM sensors inside the
tripod location, scour effects were much less pronounced. Moreover, well-developed vortex ripples were

found near the ASTM sensors in the centre of the flume while no ripples were found near the pump sampler
attached to the flume wall.
The comparison of the sand concentrations measured by the pump sampler near the ASTM sensors inside the
tripod and measured by the other pump sampler outside the tripod shows that the sand concentration profiles
inside and outside the tripod generally are within a factor 2 of each other. The deviations between the ASTM
and pump sampling concentrations are most probably caused by variations in morphodynamic (bed)
conditions at the measurement locations (presence or absense of scour holes and/or ripples) rather than by
instrumental errors and instrumental arrangement. The ASTM concentrations higher up in the water column
may be somewhat too large due to extra turbulence produced by the tripod (effect of instrumental
arrangement).
Overall, the effect of the instrumental arrangement (ASTM sensors and supports within the tripod) on the
concentrations measured by the ASTM is assumed to be sufficiently small.
References
Wave generator
Beach slope of stones
4.5 m
Sand bed
0.5 m
123 m
100 m 121 m 125 m 140 m
Figure 1
Sketch of experimental setup
1.2
Pump samples near ASTM
sensor
1.0
Pump samples from separate
system
0.8
0.6 Irregular waves:

Hs = 1.0 m
h=5m
0.4 Tp = 5 s
D50 = 0.16 mm
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 2
Pump concentrations near ASTM sensors inside tripod and outside tripod (separate system) in case of
irregular waves Hs = 1.0 m and fine sand bed (0.16 mm); average values and standard deviations derived
from 7 tests

1.2
1.0
Irregular waves: Pump samples near ASTM
height above bed (m) sensor
Hs = 1.25 m
0.8 h=5m Pump samples from
Tp = 5 s separate system
D50 = 0.16 mm
0.6
0.4
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 3
Pump concentrations near ASTM sensors inside tripod and outside tripod (separate system) for irregular
waves Hs = 1.25 m and fine sand bed (0.16 mm); results from 9 tests
1.2
Pump samples near ASTM sensors

1.0
Pump samples near flume wall
0.8
0.6
0.4
Irregular waves:
Hs = 1.0 m
0.2 h=5m
Tp = 5 s
D50 = 0.33 mm
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 4

1.2
1.0 Pump samples near ASTM sensors
height above bed (m) Pump samples near flume wall

0.8
0.6
Irregular waves:
Hs = 1.25 m
0.4
h=5m
Tp = 5 s
D50 = 0.33 mm
0.2
0.0
0.001 0.01 0.1 1 10
concentration (g/l)
Figure 5

5.4.8 Comparison of ASTM, OBS and Pump sampler
Experimental setup
The experiments have been carried out in the Grosser Wellenkanal (GWK) in Hannover, Germany. It has a
length of about 300 m, a depth of 7 m and a width of 5 m. The piston-type wave generator can produce
regular or irregular waves with periods between 1 and 15 s and heights up to 2.5 m. The wave reflections
from the beach are compensated directly at the wave paddle. For the present experiments irregular waves
were generated with a significant wave height of H1/3 = 1.25 m and wave spectrum peak period of Tp = 6.0 s.
In addition, two tests with regular waves (H = 1.6 m, T = 6.5 s) were done. The water depth was kept
constant at a value of 4.2 m in front of the wave generator. The water depth at the test section was 3.5 m. A
sand bed with a length of approximately 50 m was constructed in the flume (Figure 1).
Measurements have been done using an instrumented tripod designed and built by the Laboratory for
Physical Geography of the University of Utrecht, the Netherlands. The tripod was placed in the center of the
flume at x | 113 m (Figure 1). Instruments mounted on the tripod included a five-fold acoustic sediment
transport meter (ASTM), three optical backscatterance sensors (OBS), six pump sampling intake tubes, three
electromagnetic velocity meters (EMF), a pressure sensor and a ripple profiler.
The acoustic sand transportmeter ASTM is connected to an arm, which can be moved in vertical direction by
an electromotor. A sensor to detect the bed surface was attached to the lower end of the arm. Using the
movable arm, the ASTM sensors can be placed at pre-selected elevations above the bed. The OBS and EMF
sensors are also attached to the movable arm. Three Optical Backscatter Sensors (OBS) were used during the
present experiments. Two sensors (serial number 455 and 458) were installed on a supporting rod at about
0.03 m above the sand bed. One OBS sensor (serial number 408) was placed at about z = 0.10 m (near the
lowest ASTM sensor for comparison). The lowest EMF sensor (velocity) was positioned at about 0.05 m
above the bed. The pump sampler consisted of a pumping system with 6 intake tubes. Five intake tubes were
attached to the five ASTM sensors at about 0.10 m, 0.15 m, 0.25 m, 0.50 m and 1.0 m above the sand bed.
One intake tube was mounted near two OBS sensors at about 0.03 m above the bed. The intake tubes of 3
mm internal diameter were connected by plastic hoses to the pumps. The intake openings were placed in a
direction transverse to the plane of orbital motion. The intake velocity was about 1.3 m/s, satisfying sampling
requirements. The 8-liter samples were collected in calibrated buckets. The pump sampler was operating for
15 minutes giving an average concentration over the measurement period.
Sand was used with grain size characteristics: d10 = 0.14 mm, d50 = 0.23 mm, d90 = 0.34 mm. The sand bed
was ended by asphalt constructions in which sandtraps were built. The sandtraps were intended for
determining the sediment transport rates based on volumetric changes. Previous experiments showed that
this approach was unsuccessful.
Results
Figure 2 shows a comparison between time-averaged concentrations measured with the pump sampler and
those measured with OBS sensors. Relatively large OBS concentrations were observed on each day during
the first tests, which was caused by resuspension of fine muddy and silty materials that had settled onto the
bed during the night. These relatively large OBS values were removed from the data set. Most OBS
concentrations are smaller than the pump concentrations. The maximum discrepancy is a factor 2.
Figure 3 shows a comparison between time-averaged concentrations measured with the ASTM sensors
(based on standard calibration curve; concentrationASTM = 0.257*OutputASTM) and the pump sampler. The
ASTM values have been interpolated to compare concentrations from the ASTM and the pump sampler at
the same height above the bed. For concentrations smaller than 1 kg/m3 the ASTM concentrations and the
pump concentrations agree very well (ASTM on an average 5% smaller). For concentrations larger than 1
kg/m3 the ASTM concentrations are slightly smaller than the pump sampler concentrations (ASTM on an
average 22% smaller).

wave paddle
3.5 m tripod sand trap

4.2 m asphalt 1:6
asphalt sand 1:10
sand bed
63.1 80.6 82.1 86.8 128.2 133.0 134.5 149.0
Figure 1
Sketch of Grosser WellenKanal
3
OBS (g/l)
B
0
0 1 2 3 4 5
pump sampler (g/l)
Figure 2
Comparison of time-averaged OBS concentrations and pump concentrations
4
conc. ASTM (kg/m )
3
0
0 1 2 3 4 5
3
conc. pump sampler (kg/m )
Figure 3
Comparison of time-averaged ASTM concentrations and pump concentrations

5.4.9 Comparison of ABS and Pump sampler
Experimental setup
Thorne et al. (2002) tested the acoustic backscatter sensor ABS system in the large-scale wave tank of Delft
Hydraulics (see also Williams et al., 2000). The ABS was mounted in a tripod of Proudman Oceanographic
Laboratories, which was placed on the sand bed (0.33 mm sand; rippled bed). A pump sampling system was
used to determine the sand concentrations at 10 points above the bed surface. The intake nozzles were
installed at a horizontal separation distance of 0.3 m from the vertical ABS beam. Figure 1 shows a
comparison of time-averaged pump and ABS concentrations (over about 15 min) at various levels based on
results of tests with regular and irregular waves (no current). The sand concentrations were processed from
the ABS signal between 0.01 and 1 m with a vertical resolution of 0.01 m. The sand concentration at one
height and the suspended sand sizes over the depth were assumed to be known from the pump sampling
results to calibrate the ABS concentrations. Hence, the ABS concentrations were not independently
determined. The deviations between the results of both instruments are up to 30%.
Figure 2 (Top and Bottom) shows a comparison of pump concentrations Pc and ABS concentrations Ac for a
site in a British estuary (0.16 mm sand bed; bed covered with sand waves; tidal current; no waves). The
horizontal separation distance between the intake nozzles (at 4 heights) and the ABS beam was 0.5 m. The
ABS concentrations were determined independently of the pump concentrations. The deviations between the
results of both instruments are up to 20%.
Rose and Thorne (2000) tested the multi-frequency (1, 2.5 and 5MHz) ABS-system mounted in a tripod in
conditions with tidal flow velocities between 0.6 and 1 m/s and water depths between 2 and 3 m (River Taw
Estuary, UK). The bed was composed of fine sand with d50=0.17 mm and d90= 0.2 mm. Ripples were present
with approximate wave heights of 0.025 m and wave lenghts of 0.2 m. Figure 3 shows four ABS-
concentration profiles and pumped concentrations at 4 heights (= 0.1, 0.2, 0.4 and 0.8 m) above the bed. The
agreement between the concentrations of both instruments is quite good with maximum deviations of about
15% near the bed and 30% higher up in the water column. The magnitude of the acoustically derived
suspended sediment estimates were slightly adjusted (factor 0.8 to 1.1) to optimize the agreement with the
direct pumped sampling measurements. The 5 MHz data was not utilised owing to inversion difficulties at
periods of high concentrations when sediment attenuation of the MHz signal resulted in problematic
concentrations.
References
Thorne, P.D., Williams, J.J. and Davies, A.C., 2002. Suspended sediments under waves measured in a
large-scale flume facility. Journal of Geophysical Research, Vol. 107, No. C8, 3178, p. 4.1-4.16
Rose, C. and Thorne, P., 2000. Measurements of suspended transport parameters in a tidal estuary.
Nearshore and Coastal Oceanography/Continental Shelf Research.
Williams, J. et al., 2000. Observed and predicted vertical suspended sediment concentration profiles and
bed forms in oscillatory-only flow. Journal of Coastal Research

10
6
C (kgm3)
5
a
0
0 1 2 3 4 5 6 7 8 9 10
C (kgm3)
p
Figure 1
Comparison of pump Cp and ABS Ca concentrations in large scale wave tank of Delft Hydraulics

1
10
a
As / Ps
0
10
1
10
0 50 100 150
time (min)
b
0
10
Ac (kgm3)
2
10
4
10
4 3 2 1 0 1
10 10 10 10 10 10
3
Pc (kgm )
Figure 2
Top: Comparison of suspended sand size; As=particle size derived from ABS signal; Ps= particle
size derived from pump samples
Bottom: Comparison of suspended sand size; Ac=concentration derived from ABS signal; Pc=
concentration derived from pump samples

Figure 3
Sand concentration profiles of ABS system (circles) and pumped concentrations (crosses); River Taw
Estuary (UK)

5.4.10 Overall conclusions with respect to OBS, ASTM and ABS instruments
The following conclusions are given:

x the OBS sensors are highly sensitive to particle size;
x the OBS concentrations should be corrected for the smaller size of the suspended sand relative to that of
the sand bed material used during calibration; this requires samples of suspended sand during field
measurements;
x the effective concentration range for OBS sensors is 1 to 100 kg/m3; concentrations measured within this
range have an inaccuracy of about 50%;
x the OBS sensors often show a reasonably steady offset concentration, which is related to the background
concentration of relatively fine sediments (silt and mud), which should be subtracted from the original
time series data; if the background concentration to be subtracted from the record is of the same order of
magnitude as the sand concentration, the OBS concentrations will be rather inaccurate;
x OBS sand concentrations below 1 kg/m3 are not accurate (inaccuracy larger than factor 2) and should not
be used;
x Acoustic concentration sensors (ASTM) for point measurements are not very sensitive to particle size; the
effective measurement range is 0.1 to 10 kg/m3; concentrations measured by sensors mounted in a tripod
have an inaccuracy of about 50%;
x Acoustic concentration sensors (ABS) for profile measurements are sensitive to particle size; the effective
measurement range is 0.1 to 20 kg/m3; three different acoustic frequencies are used to simultaneously
determine the size and the concentration of the suspended sediment involved; the sand concentration
profiles can be determined with an inaccuracy of about 30% if the suspended sand size is known and
some pump concentrations are available for calibration; the optimum conditions for the ABS system are:
rather uniform fine sand (0.1 to 0.3 mm) in non-breaking wave conditions;
x the mass concentration at range r from the ABS transducer is estimated from a complex function,
depending on the voltage V measured at range r, the sediment density, the speed of sound in water and the
attenuation of sound by water and sediment of radius a; this function is not yet properly defined resulting
in relatively large inaccuracies (factor 2 to 3) of the measured concentration, if calibration samples of
concentration and sediment size are not available;
x the ABS system is rather sensitive for the presence of air bubbles in the water column (surf zone
conditions with breaking waves) resulting in an increase of the concentration.

5.5 Description of bed load samplers
5.5.1 Trap sampling
The basic principle of mechanical trap-type bed-load samplers is the interception of the sediment particles
which are in transport close to the bed over a small incremental width of the channel bed. Most of the
particles close to the bed are transported as bed load but the sampler will inherently collect a small part of the
suspended load (related to vertical size of intake mouth).
The bed-load transport measured by a mechanical sampler is dependent on its efficiency (instrumental
errors), on its location with respect to the bed form geometry (spatial variability) and on the near-bed
turbulence structure (temporal variability).
The efficiency of the bed-load sampler depends on the hydraulic coefficient, the percentage of width of the
sampler nozzle in contact with the bed during sampling and on sampling disturbances generated at the
beginning and the end of the sampling period.
The hydraulic coefficient, defined as the ratio of the inflow velocity and the ambient flow velocity, depends
on the geometry and construction of the sampler nozzle, on the position of the nozzle at the bed, on the
percentage of filling of the bag with sand particles (volume of catch) and on the percentage of blocking of
the bag material by fine sand, silt or clay particles and organic materials, depending on the mesh size of the
bag (Beschta, 1981) . Laboratory experiments have shown (Emmett, 1980; Delft Hydraulics, 1991) that the
sampler (bag or basket) can be filled to about 40% of its capacity without reduction of the hydraulic
efficiency and without loss of sediment particles during raising of the instrument. Thus, the problems of bed-
load transport measurements are related to the instrumental design and to the physical processes involved.
Typical instrumental problems of a (bag-type) bed-load sampler are:

x the initial effect; sand particles of the bed may be stirred up and trapped when the instrument is placed on
the bed (oversampling),
x the gap effect; a gap between the bed and the sampler mouth may be present initially or generated at a
later stage under the mouth of the sampler due to migrating ripples or erosion processes (undersampling),
x the blocking effect; blocking of the bag material by sand, silt, clay particles and organic materials will
reduce the hydraulic coefficient and thus the sampling efficiency (undersampling),
x the scooping effect; the instrument may drift downstream from the survey boat during lowering to the
bed and it may be pulled forward (scoop) over the bed when it is raised again so that it acts as a grab
sampler (oversampling).
Typical sampling problems related to the variability of the physical processes of bed-load transport are (see
Carey, 1985; Delft Hydraulics, 1991) :
x the number of measuring locations along a bed form (dune),
x the number of measurements at each location,
x the sampling duration of each measurement,
x the number of locations in the cross-section.
Here, the sampling duration will be discussed. The number of measurements is discussed in Par. 3.3.4.
Three effects related to the sampling period are considered:

x the volume of the nylon bag (maximum sampling period),
x the blocking of the nylon material by sediment particles,
x the presence or generation of a gap under the sampler mouth.
The size of the bag imposes a maximum sampling period which depends on the transport rate and hence on
the current velocity.

Taking a maximum filling percentage of 50%, the maximum sand catch of a 2 liter bag is about 2000 grams.
The bed-load transport formula of Meyer-Peter-Muller has been used to compute the maximum sampling
period for bed material of 500 Pm and velocities in the range of 0.75 to 2 m/s
The maximum transport rate is assumed to be equal to three times the mean rate (qb,max = 3 qb,mean).
The results are given in Table 1.
Depth-averaged velocity (m/s) Maximum sampling period (sec)

0.75 450
1.0 120
1.25 50
1.5 30
1.75 15
2.0 10
Table 1
Maximum sampling period
Blocking of standard nylon bags by fine sediments was studied by Beschta (1981) using the Helley-Smith
bed-load sampler in flume and field conditions. The 200 Pm-bag was found to be rapidly clogged by fine
sediment particles reducing the sampling efficiency to 50% after 30 seconds and to only 10% after 300
seconds. Thus, the sampling period must be rather short (< 30 seconds) whenever a standard bag is used. A
special bag consisting of nylon material with a mesh size of 250 Pm and a patch (0.10 x 0.15 m2) of 500 Pm
at the upper side of the bag has been used by Van Rijn and Gaweesh (1992). The application of the special
bag yields good results as regards the hydraulic coefficient.
Camera observations (Delft Hydraulics, 1991, 1992) did show the presence of an initial gap or the
generation of a gap at a later stage (after 2 or 3 minutes) during some of the sampling periods. Given the
stochastic nature of the generation of gaps (migrating ripples), its effect on the average transport rate will be
the least when many samples of short duration are taken (gap effect is averaged out). Furthermore, the
application of a short sampling period reduces the time available for the generation of gaps when initially no
gap is present.
Based on the above-given considerations, it is advised to use a maximum sampling period of 3 minutes for
depth-averaged velocities upto 0.8 m/s
At higher velocities (higer transport rates) the maximum sampling period must be reduced (between 0.5 and
3 minutes) or a larger bag should be used. The largest sand catch should not be larger than 50% of the bag
volume.
References
Beschta, R.L., 1981. Increased Bag Size improves Helley-Smith Bed Load Sampler for Use in
Streams with High Sand and Organic Matter Transport. Symposium Erosion and Sediment Transport
Measurement, Florence, Italy
Carey, P., 1985. Variability in Measured Bed-load Transport Rates. Water Resources Bulletin, Vol. 21, No.
1, Paper No. 84156.
Delft Hydraulics, 1991. Delft Nile Sampler, Bed Load Transport Measurements in the River Waal.
Report Q1300 part 1, Delft, The Netherlands
Delft Hydraulics, 1992. Bed Load Transport Measurements in the Waal River near Woudrichem and
Druten. Report Q1300 part 2 and 3, Delft, The Netherlands
Emmett, W.W., 1980. A Field Calibration of the Sediment Trapping Characteristics of the
Helley-Smith Bed Load Sampler. Geological Survey Professional Paper 1139, Washington, USA
Vries, M. de, 1973. On Measuring Discharge and Sediment Transport in River Flow. Delft Hydraulics
Vries, M. de, 1982. Lecture Notes on Measuring Sediment Transport in Rivers. Delft University of
Technology, Delft, The Netherlands

5.5.1.2 Bed load transportmeter Arnhem (BTMA)
Principle
The instrument is based on the collection of sediment particles by means of a basket type sampler. The
basket, consisting of fine wire mesh and mounted in a frame, is pressed (by means of a spring) on the
channel bed after lowering of the frame (Figure 1). The form of the basket causes a pressure reduction
behind the instrument so that the water and sediment particles enter the mouth of the basket with the same
velocity as that of the ambient flow, provided that the sediment content, already in the basket, is relatively
small. The sampler can collect particles coarser than 0.3 mm (mesh size basket) but finer than 50 mm
(opening height).
Practical operation
1. lower bed-load sampler to the bed
2. use a preset sampling time (usually 2 minutes)
3. raise bed load sampler (gently)
4. wash sediment catch from the basket into a calibrated funnel
5. read (immersed) volume of sediment catch
6. wash sediment catch into a sample container (for laboratory analysis)
7. change sampling location (if necessary) and repeat sampling procedure.
Remarks
1. Before starting a series of measurements, an echo sounding of the longitudinal bed section at the
sampling location must be made to determine the bed form length. If the bed form length (dunes) is large
compared with the boat length, the sampling location must be changed regularly in longitudinal direction
to assure random sampling (Figure 1) and at least 20 samples should be collected.
2. The sampling time may be larger than 2 minutes but never larger than the time needed to fill the
sample basket for about 40% of the capacity.
3. It is sufficiently accurate to determine the immersed volume of the sediment catch on board of the
vessel by using a calibrated funnel. Single samples should be returned to the laboratory to determine the
porosity factor of the samples.
Laboratory analysis
1. determine dry weight of each sample (if necessary)
2. accumulate samples of each measuring location and determine size distribution of the sediment
particles by sieving or settling tests.
Sb=D Us(1-p)Vs/(bT) or Sb= Gs/(bT)
in which:
D = calibration factor (= 2 for BTMA)
Us = density of sediment particles (= 2650 kg/m3)
b = width of intake opening (= 0.085 m for BTMA)

The accuracy of the measured bed-load transport is strongly dependent on the accuracy of the calibration
factor D, the number of measurements and the sampling procedure (skilled personel). Figure 1 presents
calibration curves for the BTMA showing considerable variations. In practice a calibration factor equal to 2
is used (drawn line).
Assuming ideal sampling, at least 20 samples must be collected at each location to obtain a bed-load
transport rate with a standard deviation (error) of about 20% (De Vries, 1973). In practice the sampling error
will be considerably larger (say 100%) particularly due to the sampling procedure. More information of the
calibration of bed load samplers is given by Hubbell et al. 1985.

dimensions: frame 0.40 x 0.8 x 1.85 m; basket 0.1x0.15x0.5 m
mesh size of basket: 300 Pm (0.3 mm)
Advantages
Disadvantages
2. winch facility needed
References
Delft Hydraulics, 1958. Calibration of BTMA (in Dutch) Report M6Q1-I, The Netherlands
Delft Hydraulics, 1966. Development of Bed Load Samplers (in Dutch) Report M601-II, The Netherlands
Delft Hydraulics, 1969. Calibration of Bed Load Samplers (in Dutch) Report M601-III, The Netherlands
De Vries, M., 1973. On Measuring Discharge and Sediment Transport in River Flow. Delft Hydraulics
Hubbell.D.W., Stevens, H.H., Skinner, J.V. and Beverage, J.P., 1985. New Approach to Calibrating Bed

Figure 1
B.T.M.A.

Figure 2
Measuring sheet bed load sampler

5.5.1.3 Helley Smith (HS)
Principle
The Helley-Smith bed-load sampler is a modified version of the BTMA-sampler (Helley and Smith, 1971).
The Helley-Smith sampler consists of a nozzle, sample bag and frame (Figure 1).
The sampler has a square entrance nozzle (0.076 x 0.076 m) and a sample bag constructed of 250 pm- mesh
polyester. Several different versions of the sampler have been used for various field conditions. Larger
nozzles are generally used to sample larger sediment sizes and heavier samplers become necessary as deeper
and faster rivers are sampled. An important advantage of the Helley-Smith sampler is the extensive
calibration (based on about 10,000 samples) and its simple operation.

Sb=D Us(1-p)Vs/(bT) or Sb= Gs/(bT)
in which:
D = calibration factor (= 0.5 for particles of 0.25 to 0.5 mm)
(= 1.0 for particles of 0.5 to 16 mm)
(= 1.5 for particles of 16 to 32 mm)
Us = density of sediment particles (= 2650 kg/m3)
b = width of intake opening (= 0.0762 m for HS)
Figures 2 and 3 present calibration curves relating the sediment catches and the actual transport rates
for various size fractions. The actual transport rate has been assumed to be represented by
simultaneous measurements with a conveyer belt system just downstream of the sampling position of
the Helley-Smith sampler (Emmett, 1980).
The conveyer-belt trap consists of a concrete slot (width = 0.4 m, depth = 0.6 m) in the channel bed,
orthogonal to the flow direction. Along the bottom of the concrete slot passes an endless belt of rubber
(width = 0.3 m). Sediment falling into the open slot drops on the moving belt, and is carried laterally to
the riverbank where it is scraped off the belt, weighted and returned to the river flow (depth = 1.2 m, width
= 15 m, discharge = 20 m3/s).
Based on Figures 2 and 3, Emmet (1980) concluded that the Helley-Smith Sampler has an efficiency of
100% (calibration factor, D = 1) for particle sizes in the range of 0.5 mm to 16 mm. For particles in
the range of 0.25 to 0.5 mm the efficiency is found to be about 175%.
(resulting in a calibration factor D= 0.5) which is assumed to be caused by the collection of suspended
sediment particles.
For particles in the range of 16 to 32 mm the efficiency is found to be about 70% (resulting in a
calibration factor D = 1.5) which is assumed to be caused by the paucity of large particles moving as
bed load in the flow.
Summarizing:
D= 0.5, for particles in the range of 0.25 to 0.5 mm,
D= 1.0, for particles in the range of 0.5 to 16 mm,
D= 1.5, for particles in the range of 16 to 32 mm.
More information of the calibration of bed load samplers is given by Hubbell et al., 1985.

The original Helley Smith sampler was tested by Delft Hydraulics (1996, 1997). The sampler was tested in a
laboratory flume (known bed load transport) with water depths in the range of 0.5 to 0.6 m and sediment of
D50=0.46 and 0.75 mm. About 25 samplings were required to obtain a variation coefficient lower than 0.1.
The results are given in the following Table. The HS sampler was found to suffer from the problem of
oversampling with a factor of 2 to 3, especially at low velocities.
Bed load D50=0.46 mm D50=0.75 mm

transport D10=0.33 mm D10=0.20 mm
measurements D90=0.65 mm D90=9 mm
v=0.5 m/s v=0.75 m/s v=1 m/s v=0.8 m/s v=0.9 m/s
transport in
Flume
transport
measured by
Helley Smith
Ratio of 3.0 1.6 1.7 2.2 2.5
2) and 1)
dimensions: frame 0.18x0.32x1 m
mesh size of basket: 250 Pm (0.25 mm)
Advantages
2. easy to handle
Disadvantages
References
Delft Hydraulics, 1996. Test measurements of Helley Smith and modified Helly Smith (in Dutch). Report Q
Delft Hydraulics, 1997. Calibration and comparison of Helley Smith. Report Q 2345,Delft, The Netherlands
Emmett, W.W., 1980. A Field Calibration of the Sediment Trapping Characteristics of the Helley-Smith
Bed Load Sampler. Geological Survey Professional Paper 1139, Washington, USA
Helley, E.J. and Smith, W., 1971. Development and Calibration of a Pressure-Difference Bed Load
Sampler. U.S. Geological Survey Open File Report, Washington, USA
Hubbell, D.W., Stevens, H.H., Skinner, J.V. and Beverage, J.P., 1985. New Approach to Calibrating Bed

Figure 1
Helley Smith

Figure 2

Figure 3

5.5.1.4 Delft Nile sampler (DNS)
Principle
The instrument consists of a bed-load sampler and a suspended sampler attached to a supporting frame (see
Figure 1). The sampler has a weight of about 60 kg. The suspended load sampler consists of 7 intake nozzles
(inner diameter = 0.003 m) which are connected to plastic hoses and operated by pumps.
The bed-load sampler consists of a nozzle (entrance width = 0.096 m, entrance height = 0.055 m, length =
0.085 m, rear width = 0.105 m, rear height = 0.06 m) connected to a bag. The bag consists of nylon material
with a mesh size of 150 or 250 Pm, depending on the size of bed material. At the upper side of the bag a
patch (0.10 x 0.15 m2) consisting of 500 Pm material is present to reduce the blocking effect by fine particles
as much as possible (see Figure 1). The bottom side of the sampler nozzle has a forward-tilting slope of 1 to
10. The bed-load sampler is connected to a swing arm, which can move in the vertical direction (upward and
downward) over a distance of about 0.15 m with respect to the supporting frame. In its hanging position
during lowering of the instrument, the bed-load sampler is in its most upward position. When the frame of
the instrument is placed on the bed, the sampler nozzle makes contact with the bed by relaxing the tension in
the cable. When the instrument is raised, first the nozzle of the bed-load sampler is raised over 0.15 m and
then the frame is lifted from the bed. The swing-arm construction was designed to reduce the gap effect and
the scooping effect as much as possible. The instrument was developed for bed-load measurements in the
Nile river.
The hydraulic coefficient of the bed-load sampler, defined as the ratio of the inflow velocity and the ambient
flow velocity, was determined by measurements in a flume (see Table 1). A small Ott-propeller meter (dia-
meter = 0.02 m) was placed in the entrance of the nozzle. The velocity in the nozzle was compared with the
flow velocity measured simultaneously at the same height in a section two metres upstream of the sampler.
Ten successive measurements of 30 seconds each were carried out. Three types of bags were used, as
follows:
1. 250 Pm-nylon bag and 500 Pm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag (see Fig. 1),
2. 150 Pm-nylon bag and 500 Pm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag,
3. impermeable plastic bag and 500 Pm-nylon patch (0.1 x 0.15 m2) at the upper side of the bag.
This latter bag is supposed to simulate a nylon bag which is almost fully blocked by fine silt and clay
material as present in most natural conditions (Beschta, 1981). Four filling percentages of the bag were
tested: 0%, 25%, 50% and 75%.
The hydraulic coefficients of the 250 Pm -bag and the 150 Pm -bag, both with a patch of 500 Pm at the upper
side, are about unity for filling percentages in the range of 0% to 50%. A filling percentage of 75% reduces
the hydraulic coefficient to about 0.75. The hydraulic coefficient of an impermeable plastic bag with a patch
of 500 Pm at the upper side, simulating a blocked nylon bag, is about 0.8 for a filling percentage in the range
of 0% to 50%. A filling percentage of 75% reduces the hydraulic coefficient to about 0.70. Based on these
results, a maximum filling percentage of 50% is advised to be used. Blocking of the nylon material by fine
sediments will result in a hydraulic coefficient of about 0.8.
Finally, it is noted that the importance of a hydraulic coefficient equal to unity should not be
overemphasized, because actually the sampling efficiency is the most important parameter. A hydraulic
coefficient of unity does not necessarily give a sampling efficiency of unity, because other factors are
involved.

Velocity Mesh size Mesh size Hydraulic coefficient
in mouth patch
Filling Filling Filling Filling

percentage percentage percentage percentage
(m/s) (Pm) (Pm) 0% 25% 50% 75%
250 500 1.04 1.01 0.98 0.91
0.5 150 500 1.07 1.04 0.98 0.94
0 (impermeable) 500 0.82 0.82 0.82 0.82
250 500 1.05 1.02 1.01 0.75

0.8 150 500 1.08 1.03 0.99 0.75
0 (impermeable) 500 0.84 0.80 0.80 0.70
Table 1
Hydraulic coefficients (average values)
The sampling efficiency is defined as the ratio of the bed-load transport measured by the sampler at a certain
location during a certain period and the true bed-load transport at the same location during the same period
(if the sampler had not been there). The sampling efficiency will express sampling errors related to the initial
and scooping effect, the gap effect and the loss of sediment particles through the patch of the bag.

Sb= D(Gs Go)/(bT)
in which:
D = calibration factor (= 1 to 1.5),
Gs = dry mass of sediment catch (kg),
Go = dry mass of sediment catch related to initial and scooping effect determined by zero-sampling
(sampler is lowered to bed and immediately raised),
b = width of intake opening (= 0.096 m for DNS),
The Go-values should be determined by taking 10 zero-samplings at each location. Practical experience sofar
shows a value of about 0.03 kg. The sampling efficiency (D-factor) was determined by means of tests in a
flume at the Hydraulics and Sediment Research Institute in Deltabarrage, Egypt. Different velocity and sand
size ranges were considered. The D-factor was found to be in the range of 1 to 1.5.
dimensions: frame 0.5x0.6x1.1 m
mesh size of basket: 150 and 250 Pm

Advantages
2. easy to handle
Disadvantages
References
Van Rijn, L.C. and Gaweesh, M., 1992. A New Total Load Sampler. Journal of Hydraulic Engineering,
Vol. 118, No. 12
Delft Hydraulics, 1991. Delft Nile Sampler; Bed Load Measurements in the River Waal. Report Q1300, part
Delft Hydraulics, 1992. Bed Load Measurements in the River Waal near Woudrichem and near Druten
Report Q1300, part 2 and 3, Delft, The Netherlands

Figure 1
Delft Nile Sampler

Figure 2
Delft Nile Sampler

5.5.2 Bed form tracking
Principle
The basic principle is the computation of the bed-load transport from bed-form profiles measured at
successive time intervals under similar flow conditions (Figure 1).
Assuming steady flow conditions and undisturbed bed-form migration, the bed-load transport rate can be
computed from the bed form dimensions (Engel and Lau, 1980, 1981, De Boer, 1996).
Sb=Ds Us(1-p) f a '
in which:
Ds = coefficient (0.5 to 0.6),
p = porosity factor (= 0.4),
Us = density of sediment particles (= 2650 kg/m3),
f = shape factor (2V/(' O)),
a = average migration velocity (m/s),
V = volume of bed form per unit width,
' = average bed form height (m),
O = bed form length.
To apply this equation, the migration velocity and the bed form height must be determined from the bed
profiles. The bed-load transport rate can also be computed directly from the (successive) profile data using
all data instead of selecting the characteristic parameters such as the average migration velocity and the bed-
form height (see Havinga, 1982). To collect the bed profile data along a prefixed course, an accurate three-
dimensional measuring system must be available consisting of a two-dimensional horizontal positioning
system and a one-dimensional vertical sounding system.
In (isolated) field conditions, where an accurate positioning system is too complicated, a much simplier
method can be used. By means of an analoque echo sounder two or more successive bed-profile registrations
can be made in a longitudinal section between two well-defined cross-sections (bank marker). Using a
simple hand method, the average migration velocity and bed-form height can be determined quite
easily, as shown in Figure 1A.
Engel and Wiebe report an overall inaccuracy of about 40 to 50% for flume conditions. Figure 1B shows
measured and computed transport rates for flume conditions (Simons et al, 1965). For field conditions the
inaccuracy may be as large as 100%.
De Boer (1996) has developed a dune-track computer program to estimate the bed load transport from
successive bed profile measurements. The Ds-factor was found to be 0.64 for the Dutch IJssel river. The
length of each profile should at least be between 1 and 3 km. The local bed load transport may vary between
50% and 200% of the average value for one profile.
References
De Boer, A.G., 1996. The applicability of the dune-track method (in Dutch). Department of Physical
Geography, University of Utrecht, Utrecht, The Netherlands
Engel, P. and Wiebe, K., 1979. A Hydrographic Method for Bed-Load Measurement. Proc. Fourth
Nat. Hydro-Techn. Conf. River Basin Man., Vol. I, page 98-113. Vancouver, Canada
Engel, P. and Lau, Y.L., 1980. Computation of Bed Load Using Bathymetric Data. Journal of the
Hydraulics Division, ASCE, HY 3
Engel, P. and Lau, Y.L., 1981. Bed Load Discharge Coefficient. Journal of the Hydraulics Division,
ASCE, HY 11
Havinga, H., 1982. Bed Load Determination by Dune Tracking. Dir. Water Management and Water
Motion, District South East, Rijkswaterstaat, The Netherlands
Simons, D.B., Richardson, E.V. and Nordin, C.F., 1965. Bed Load Equation for Ripples and Dunes. U.S.
Geol. Survey Prof. Paper 462 H, Washington, USA

Figure 1
Bed load transport according to bed form tracking

5.6 Description of suspended load samplers
5.6.1 Classification of samplers
The suspended load samplers described in the following paragraphs can be classified according to their
measuring principles (see Par. 3.1), as follows:
Suspended load samplers Point-integrating Depth-integrating
Direct method Delft-Bottle sampler

Acoustic samplers
Indirect method Trap samplers USD-49

Bottle sampler Collapsible-Bag sampler
USP-61
Pump samplers
Optical samplers
Impact samplers
The trap and bottle-type samplers can only be used in (quasi) steady flow conditions. The other samplers can
be used in unidirectional and oscillatory flow (waves).

5.6.2 Bottle and Trap samplers
The basic principle of all mechanical bottle and trap samplers is the collection of a water-sediment sample to
determine the local sediment concentration, transport and/or particle size by physical laboratory analysis.
Optimal sampling of a water-sediment volume by means of a mechanical instrument requires an intake
velocity equal to the local flow velocity (iso-kinetic sampling) or a hydraulic coefficient, defined as the ratio
of the intake velocity and local flow velocity, equal to unity (hc=1).
Differences between the intake velocity and local flow velocity result in sampling errors. Sampling at a
lower velocity than that of the ambient flow would result in a higher sediment concentration than present in
the flow due to diverging flow lines, which cannot be followed by the sediment particles, being of higher
density than the water particles (Figure 1A). Conversely, sampling at a higher velocity than that of the
ambient flow would result in a lower sediment concentration (Figure 1B). The magnitude of error due to
incorrect intake velocities in suspensions of various concentrations has been determined experimentally by
Nelson and Benedict (1950), whose results are shown in Figure 2. The results show a decreasing effect with
a decreasing particle size. It is also clear that a hydraulic coefficient smaller than 1 results in a relatively
large error. Further Nelson and Benedict found no significant change of the sampling error in relation to the
flow velocity (range 0.9 to 1.5 m/s), the size of the intake nozzle (range of 4 to 7 mm), the angle between
flow direction and nozzle axis (range 0 to 20), for a hydraulic coefficient in the range of 0.5 to 2 in all tests.
Crickmore and Aked (1975) report that no systematic trends can be distinguished for a hydraulic
coefficient in the range 0.5 to 4 and particle sizes from 60 pm to 250 Pm. The maximum experimental error
found was about 10%.
Laboratory tests using an intake nozzle normal to the flow direction show errors of about 15% for particle
sizes of 150 pm and 450 Pm, while for 60 Pm - sediment no errors were found (Nelson and Benedict, 1950).
Hulsbergen (1981) using an intake nozzle normal to the flow direction reports an error of about -15%
(actual concentration larger than measured concentration) for 60 Pm-sand, of about -25% for 110Pm and 270
Pm-sand and of about -33% for 500 Pm-sand. In all tests the hydraulic coefficient was in the range 1 to 4.
Crickmore and Aked (1975) found a maximum error of 20% for a misalignment of 180 using a hydraulic
coefficient larger than 1.
Bottle and Trap samplers are usable in steady flow conditions (rivers), but not in time-dependent flow
conditions (estuaria and coasts).
References
Conference on Instrumentating Oceanography, I.E.R.E., Conference Proceeding No. 2, Bangor,
England
Hulsbergen, C.H., 1981. Determination of Sand Concentration by Pump Sampling normal to the flow.
Delft Hydraulics, M1267, Delft, The Netherlands
Nelson, M.E.and Benedict, P.C., 1950. Measurement and Analysis of Suspended Sediment Loads in
Streams. A.S.C.E.-Proceedings, Volume 76, U.S.A.
Van Rijn, L.C. van, 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Research Report S404, The
Netherlands

Figure 1
Influence of intake velocity on sediment paths

5.6.2.2 Bottle sampler
Principle
The method is based on the filling of a bottle to determine the silt and/or sand concentration at a specific
point in the flow. Usually, the bottle is placed vertically in a container (Figure 1) and lowered to the
sampling point, where the bottle is opened (mechanical or electrical). A cork ball should be present to close
the bottle after filling. The exact filling time is unknown but may vary from about 20 to 400 sec, depending
on the bottle orientation, flow velocity and sampling height as shown in Figure 2A (Delft Hydraulics,
1980).
Rapid profile measurement can be achieved by using a rack with 5 bottles (or more), which are opened at
prefixed depths.
Practical operation
1. lower container to sampling position (echo-sounder attached to bottle)
2. remove stop (mechanical or electrical)
3. wait till bottle is filled (3~5 minutes)
4. raise container (slowly) and replace bottle.
Laboratory analysis
The bottles should be stored away from direct sunlight. Analysis should be completed before the growth of
organisms.
The standard procedure for determining the sediment concentration is as follows (see also Paragraph 8.1.2,
Fig. 1):
1. measure water volume of sample
2. allow sample to settle for 24 hours and decant sediment-free water
3. wash sample over a 50 Pm - sieve suspended in a glass cylinder under
light brushing to separate sand (use distilled water to remove salt)
4. wash sand sample over a pre-weighed non-hygroscopic nylon filter (0.5 Pm)
5. filter silt sample through a pre-weighed 0.5 Pm-filter (use non- hygroscopic nylon or glass-fiber filter
material)
6. use distilled water to wash free from salt
7. dry and weigh sand and silt samples
8. store sand sample for size analysis.
In the case of a large amount of samples the laboratory analysis can be reduced considerably by using an
optical method to determine the silt concentration (after separation of sand particles using the 50 Pm - sieve).
About 10% of the samples are filtered to obtain a calibration curve. However, this method may introduce
additional errors due to scatter of the calibration curve.
Another alternative method is the use of ultra-centrifuge tubes to separate the silt particles.

c=Gs/V
in which:

Measurements in a laboratory flume (Delft Hydraulics, 1980) have shown that the efficiency of a bottle in
collecting the sand particles (> 50 Pm) is strongly dependent on the orientation of the bottle opening to the
main flow direction. Figure 2B shows the bottle efficiency as a function of the angle (D) with the flow
direction. The optimal angle is about 35. For D < 35 the measured concentration is too small, while for D >
35 the measured concentration is too large.
Using a bottle in a vertical position, the inaccuracy of the measured sand concentration may be as large as
50%. In field conditions the inaccuracy of the measured sand concentration may even be larger when the
filling time is small compared with the characteristic time scale of the fluctuating concentrations.
The inaccuracy of the measured silt concentrations may be about 10%, which is indicated by field
measurements (see Paragraph 5.4.5).
dimensions: bottle 0.5 to 2 litre
Advantages
Disadvantages
References
Delft Hydraulics, 1980. Investigation Vlissingen Bottle (in Dutch). M1710. Delft, The Netherlands
Hayes, F.Ch., 1978. Guidance for Hydrographic and Hydrometric Surveys. Delft Hydraulics Laboratory,
Publication No. 200, The Netherlands

Figure 1
Bottle sampler

Figure 2
Efficiency Bottle sampler

Figure 3
Measuring sheet bottle/trap sampler

5.6.2.3 Trap sampler
A. Instantaneous trap sampler
Principle
The instantaneous trap sampler consists of a horizontal cylinder equipped with end valves which can be
closed suddenly (by a messenger system) to trap a sample instantaneously, as shown in Figure 1. The water
is allowed to flow through the horizontal cylinder while the sampler is lowered to the desired point.
Practical operation
1. Lower the trap to the measuring position (echo sounder attached to trap)
2. Close valves by releasing the messenger
3. Raise the trap sampler
4. Transfer water-sediment sample to a bottle (see Fig. 1)
5. Note data on measuring sheet (see Paragraph 5.6.2.2.)
Laboratory analysis

c=Gs/V
in which:
As the trap sampler yields an instantaneously measured concentration, many samples are necessary to obtain
a statistically reliable average value.
Advantages
Disadvantages

B. Time-integrating trap sampler
The trap sampler used in lakes and oceans consists of a cylinder-shaped or funnel-shaped box with a closed
bottom and placed vertically in the water column, see Fig. 2. These traps are used to determine:
x the deposition rate at a certain elevation,
x the sediment concentration at a certain elevation.
In still water the trapping efficiency is unity. In flowing water the trapping efficiency depends on many
factors: turbulence generated inside trap, trap geometry, concentration and fall velocity of particles.
Cylinders were found to give the most accurate results. The trapping efficiency was found to be 0.8 to 1.2 for
mud suspensions in current velocities of 0.05 to 0.1 m/s (Bloesch and Burns, 1980; Gardner, 1980) . The
ratio of the height and diameter of the trap should be in the range of 3 to 5. Funnel-shaped traps were found
to give an underestimation of the deposition rate in flowing water. Traps with a small mouth and a wide body
(bottle) were found to give a large overestimation of the deposition rate.
Antsyferov et al (1990) used tripods and poles equipped with traps to measure the sediment concentrations
in coastal conditions (surfzone). The traps consisted of cylinder-shaped boxes (height = 0.1 m, diameter =
0.075 m) with six openings of 7.5 mm at the upper part of the boxes, see Fig. 2. The traps were set into
operation (by divers) by aligning corresponding openings in the trap body and in the cover.
The concentration of size fraction i at height z is given as:
ci(z)= Mi/(ke F 'T Ut)
in which:
Mi = sediment mass of fraction i in trap,
F = area of intake openings projected normal to wave direction,
'T = sampling period,
Ut = time-averaged value of the absolute horizontal orbital velocity at height z,
ke = trapping coefficient.
The ke-value was obtained from field and laboratory calibrations. Concentrations were measured by means of
a pump sampler close to the sand traps. Velocity measurements were also performed close to the trap (0.5
to 0.8 m/s). Antsyferov et al (1990) found:
ke = 0.26 for field conditions,
ke = 0.21 for laboratory conditions.
The ke-values are only valid for steady wave conditions (during period 'T). In storm conditions with growth,
stabilization and decay of the waves, the ke-value is different. It was found that 70% of the sediment mass
was trapped during the growth and decay phase of the waves (inside breaker zone).
Katori (1983) and Kraus (1987) used portable streamer traps to measure transport rates in the surf zone, see
Fig. 3. The traps consist of long rectangular bags of polyester sieve cloth material (100 Pm) vertically
mounted on a stainless steel rack (Kraus, 1987). An operator standing downcurrent attends the trap during a
sampling interval of 10 min. The use of these traps is restricted to shallow water (1 m) with wave heights less
than about 0.5 m.
Katori (1983) used a similar trap to measure cross-shore transport at the bed. The trap was mounted on a
rubber mat resting on the bed (to prevent scour), see Fig. 3.
Advantages of streamer traps are:

x absolute measurement of transport rate,
x short sampling period of 10 min,
x vertical distribution can be measured,
x simultaneous deployment at many locations,
x simple, robust and cheap.

Disadvantages are:
x disturbances of flow fields,
x scour around the traps,
x many operators involved,
x analysis of many samples,
x restricted to shallow water.
References
Antsyferov, S.M., Belberov, Z.K. and Massel, S., 1990. Dynamical Processes in Coastal Regions.
Publishing House, Bulgarian Academy of Sciences
Bloesch, J. and Burns, N.M., 1980. A Critical Review of Sedimentation Trap Technique. Schweiz. Z.
Hydrol. 42, No. 1, p. 15-55
Gardner, W.D., 1980. Sediment Trap Dynamics and Calibration. Journal of Marine Research, Vol. 38, No.
1, p. 16-39
Gardner, W.D., 1980. Field Assessment of Sediment Traps. Journal of Marine Research, Vol. 38, No. 1, p.
41-52
Katori, S., 1983. Measurement of Sediment Transport by Streamer Trap (in Japanese). Report No. 17, TR-
Kraus, N.C., 1987. Application of Portable Traps for Obtaining Point Measurements of Sediment Transport
Rates in the Surf Zones. Journal of Coastal Research, Vol. 3, No. 2

Figure 1
Trap sampler

Figure 2
Trap sampler

Figure 3
Trap sampler

5.6.2.4 USP-61 point-integrating sampler
Principle
The sampler consists of a streamlined bronze casting (= 50 kg), which encloses a small bottle (= 500 ml), as
shown in Figure 1A. The sampler head is hinged to provide access to the bottle. The intake nozzle, which
can be opened or closed by means of an electrically operated valve, points directly into the approaching flow.
To eliminate a sudden inrush after opening of the intake nozzle, the air pressure in the bottle is balanced with
the hydrostatic pressure prior to opening of the valve. This is accomplished by means of an air bell in a body
cavity connecting the bottle and the surrounding stream. After opening of the valve, the air in the bottle can
escape through a special air-exhaust tube pointing downstream on the side of the sampler head. As a result
the hydraulic coefficient is approximately unity during sampling. The filling time varies from 10 to 30
seconds, as shown in Figure 1B. To avoid a circulation flow, the bottle should only be filled for about 75%.
The USP-63 is a heavier version (= 100 kg) of the USP-61.
The sampler is manufactured by Rickly Hydrological Company (www.rickly.com).
Practical operation
1. lower instrument to sampling position (echo sounder attached to sampler)
2. open (electrical) valve
3. wait till bottle is filled for about 75% (Figure 1B)
4. close (electrical) valve
5. hoist instrument (slowly) and replace bottle.
6. note data om measuring sheet (see Paragraph 5.6.2.2)
Laboratory analysis

c=Gs/V
in which:
The silt and sand transport (in kg/s/m2) can be determined as:
S=Gs/(F T)
in which:
F = area of nozzle (m2)
The sampling efficiency of the USP-61 is strongly dependent on the ratio of the intake velocity and the local
flow velocity (hydraulic coefficient).
Extensive laboratory measurements, summarized by Dijkman (1978), have shown that the hydraulic
coefficient varies from about 0.8 to 1.3 depending on the water temperature, sample height above the bed
and the nozzle orientation (maximum deviation with flow direction of 20). For this range a maximum
sampling error in the concentration of about 10% may be expected in case of a steady concentration (see
Figure 2, Par 5.6.2.2). In the case of field conditions with fluctuating concentrations the inaccuracy of
individual samples may be as large as 50%. To obtain a reliable average value in a statistical sense, a large
number of samples (say 10) should be collected at each sampling point.

nozzle diameter: 0.005 m
weight: 50 kg (USP-61) and 100 kg (USP-63)
energy: 24 volt (DC)
Advantages
Disadvantages
4. not for sampling close to bed,
5. fragile intake nozzle
References
Dep. of Civ. Eng., Int. Report No. 5- 78, The Netherlands
FEDERAL INTER-AGENCY SEDIMENTATION PROJECT. Instructions for USP-61, Suspended
Sediment Sampler Minnesota 5541K, U.S.A.

Figure 1
USP 61 sampler

5.6.2.5 Delft Bottle sampler
Principle
The Delft Bottle (Figures 1 and 2) is based on the flow-through principle, which means that the water
entering the intake nozzle leaves the bottle at the backside. As a result of a strong reduction of the flow
velocity due to the bottle geometry, the sand particles larger than about 100 ym settle inside the bottle. Using
this instrument, the local average sand transport is measured directly.
Practical operation
The Delft Bottle (DB) can be used as follows:
1. suspended at a wire using a straight nozzle,
2. suspended in a frame resting on the bottom using a bended or straight nozzle for measurements close
to the bed.
Depending on the flow conditions, a small nozzle (velocities < 1 m/s) with an internal diameter of 15.5mm
or a big nozzle with an internal diameter of 22 mm can be used.
The operational procedure is as follows:

1. lower DB into the flow,
2. let the air in the bottle escape through the nozzle and/or openings in the backside,
3. lower DB to the sampling position (note lowering time), use echo sounder to determine exact position,
4. start sampling period (note sampling time),
5. raise DB slowly (note raising time),
6. remove water from DB (gently),
7. transfer sediment catch from DB to a calibrated (funnel) glass and read volume,
8. transfer sediment catch to bottle for laboratory analysis (if necessary),
9. note data on measuring sheet.
Laboratory analysis
As the sediment sample only consists of sand particles (fine silt particles are washed through the DB) the
laboratory analysis is rather simple:
1. decant surplus water from sampler container,
2. determine dry mass of sand particles (if necessary),
3. determine size distribution of sand particles by sieving or settling tests,
4. determine calibration factor (see Figure 3).
Remarks
As the overall efficiency of the Delft Bottle is rather low, it is sufficiently accurate to determine only the
(immersed) volume of the sand catch on board of the vessel. A small number of the samples can be returned
to the laboratory to determine the porosity factor of the sediment sample and the particle size distribution.

The local average sediment transport (in kg/m2/s) is determined as:
S= D (1-p) Us Vs/(F T) or S=D Gs/(F T)
in which:
D = calibration factor according to Figure 3,
p = porosity factor,
Us = density of sediment (2650 kg/m3),

Sampling errors are introduced by:
1. incorrect intake velocity compared with local flow velocity; the hydraulic coefficient (ratio
of intake velocity and local flow velocity) varies from 1 to 1.5 (Dijkman, 1978, 1981),
2. inefficiency of the sampler to collect relatively fine sediment material (particles finer than 100 Pm),
3. additional sampling during raising and lowering of the instrument,
4. sediment losses during removal of the sand catch from the DB.
Usually, only the first two errors are corrected using a calibration factor a according to Figure 3, which is
based on extensive laboratory measurements (Dijkman, 1981). The D-factor varies from 0.7 to 2.5
depending on the nozzle type, particle size and local flow velocity. The sampling error due to the collection
of sediment particles during lowering and raising of the instrument can be reduced by using a relatively large
sampling period (15 minutes). Otherwise, an additional calibration factor is necessary (Dijkman, 1981). The
minimum sampling time is about 5 minutes to obtain a statistically reliable result. An additional advantage of
a long sampling period is the collection of a large sediment catch enabling an accurate determination of
particle size (by sieving).
Field measurements show sampling errors up to 50% for individual samples, even after the application
of the calibration factor (Paragraph 5.4.1). Considering these large errors, the Delft Bottle can only be used
to obtain a rough estimate of the local sand transport. Therefore, it is sufficiently accurate to determine only
the volumetric quantity of the sand sample (in-situ). In that case the laboratory analysis is rather limited,
which is an advantage of the Delft Bottle method. The Delft Bottle should not be used in tidal flow
conditions with relatively small sediment concentrations because of the long sampling period which is
required to obtain a measurable sediment catch.
nozzle diameter: 0.0155 or 0.022 m (straight or bended)
dimensions: bottle 1.15x0.2x0.2 m; frame 1x1x0.85 m; funnel 0.5x0.2x0.2 m
weight: bottle 20 kg; frame 66 kg; funnel 3 kg
Advantages
1. simple and reliable instrument for sand particles (>100 Pm)
2. long sampling period
3. large sand catch (size analysis)
5. no electricity
Disadvantages
1. not for particles < 100 Pm
2. application of calibration factors
4. not for tidal conditions (large cycle time)
References
Dep. of Civ.Eng., Int. Report No. 5-78, The Netherlands
Dijkman, J., 1981. Investigation of Characteristic Parameters of Delft Bottle. Delft Hydraulics Laboratory,
Report S362, The Netherlands
Dijkman, J. and Milisic, V., 1982. Investigations on Suspended Sediment Samplers. Delft Hydraulics

Figure 1
Delft Bottle

Figure 2
Delft Bottle

Figure 3
Calibration factors of Delft Bottle sampler

Figure 4
Measuring sheet of Delft Bottle sampler

5.6.2.6 USD-49 depth-integrating sampler
Principle
The USD-49 is a depth integrating sampler. The sampler is lowered at a uniform rate from the water surface
to the streambed, instantly reversed, and then raised again to the water surface. The sampler continues to take
its sample throughout the time of submergence. At least one sample should be taken at each vertical selected
in the cross-section of the stream. A clean bottle is used for each sample. The USD-49 sampler has a cast
bronze streamlined body in which a round or square pint-bottle sample container is enclosed. The head of the
sampler is hinged to permit access to the sample container (see Figure 1). The head of the sampler is drilled
and tapped to receive the -inch, 3/16-inch or 1/8-inch intake nozzle which points into the current for
collecting the sample. The transit rate depends on the mean velocity in the vertical, the water depth and the
nozzle diameter, as shown in Figure 2. The USD-49 is suitable for depth integration of streams less than
about 5 m in which the velocities do not exceed 2 m/s.
The sampler is manufactured by Rickly Hydrological company (www.rickly.com).
Practical operations
1. determine the depth to be sampled and divide twice the depth (round-trip integration) by the sampling
time to obtain the required rate for lowering and raising,
2. lower and raise the sampler at the selected transit rate (use stopwatch for time control),
3. replace bottle or container on board of vessel,
4. note the lowering and raising times,
5. observe sample volumes and transit rates carefully, so the transit rate can be adjusted (if necessary).

The depth-averaged concentration can be determined as
cmean=Gs/V
in which:
The depth-integrated suspended sediment transport (in kg/m/s) can be determined as:
Ss= Gs h//(F T) or as Ss=cmean umean h= (Gs/V) umean h
in which:
h/ = depth of sampled zone (m),
umean = depth-averaged velocity (m/s).
The sampler cannot sample down to the stream bed surface. When the sampler touches the bed, the distance
between the sample nozzle and the bed is about 0.1 m (see Figure 1). Thus, the depth of the sampled zone is
about equal to the water depth minus 0.1 m (h/ = h - 0.1 in m). Another problem is the short sampling period
at each specific point in the vertical. As a result concentration fluctuations are not averaged out and repeat
samples are necessary. Information of the vertical concentration distribution cannot be obtained.

nozzle diameter: 0.006, 0.0045, 0.003 m (1/4, 3/16 or 1/8 inch)
weight: 28 kg
Advantages
Disadvantages
1. for depth smaller than 5 m
2. small sediment catch
3. many samples required
5. fragile intake nozzles
References
INTER-AGENCY COMMITTEE on Water Resources, 1963. Determination of Fluvial Sediment

Discharge Report no. 14, St. Anthony Falls Hydr. Lab., Minneapolis, USA

Figure 1

Figure 2
Allowable ratio of uniform transit rate RT to mean velocity Vm for depth-integration using litre milk bottle

5.6.2.7 Collapsible-Bag depth-integrating sampler
Principle
The Collapsible-Bag sampler is based on the principle that the static pressure acting on the outside surface of
the flexible bag (devoid of air) creates at the nozzle exit a pressure equal to the hydrostatic pressure at the
nozzle entrance. Using this method, samples can be collected throughout any depth.
The sampler consists of a wide-mouth, perforated, rigid plastic container enclosed in a cage-like metal frame.
The head of the frame supports a plastic intake nozzle (6 or 13 Pm) and swings open to permit the plastic
container to be removed. When the head is closed, the end of the nozzle extends slightly into the mouth of
the container. Perforations in the container allows the air in the container to escape during submergence. For
sampling, a collapsed flexible plastic bag is placed inside the rigid container. The neck of the flexible bag is
stretched over the neck of the rigid container and this unit is placed into the sampler. When the sampler is
lowered into the flow, water enters the perforations in the rigid container, surrounds the collapsed bag and
equalizes the hydrostatic pressure at the nozzle entrance. The velocity head forces water through the intake
nozzle into the collapsed bag, which unfolds to conform to the rigid container. This sampling action
eliminates any possibility for flow to rush in due to unequal pressures and insures that water-sediment
mixture is collected at stream velocity regardless of velocity distribution and depth, provided that the sample
container does not overfill and the vertical transit rate is smaller than 0.4 of the depth-averaged velocity.
Nordin et al. have used a large-volume (6 1) bag sampler in combination with an Ott-type current meter in
water depths up to 30 m. The rotation of the propeller was transmitted into a rate meter. This latter provided
a direct read out of velocity. By lowering and raising the current meter over the depth, the instantaneous
velocity profile can be recorded.
The constant vertical transit rate of the bag sampler should be fast enough so that the sample container is
not completely filled, but not greater than 0.4 v (v= depth-averaged velocity). Nordin et al. tried to keep the
transit rate smaller than 0.2 v . The sample volumes that will be collected for a transit rate equal to 0.2 v are
given in Figure 1 A. The mean velocity v can be estimated from the surface velocity. The time required to
collect one liter of sample for various mean velocities and nozzle diameters is given in Figure 1B. If the
cross-channel distribution of sediment discharge is needed, each sample is processed separately. If the depth
and velocity are fairly uniform across the section, it often is possible to use one nozzle diameter and transit
rate at all verticals and the individual samples can be composited into a single discharge-weighted sample.
For most purposes 10 to 15 depth-integrated samples equally spaced across the section are adequate. If the
maximum and minimum surface velocities are observed when the cross-section is sounded initially, that
information can be used with the data plotted in Figure 1B to select the transit rate. Usually, it is accurate
enough to select a transit rate smaller than 0.2 v using the velocity v of the deepest en fastest vertical.
Assuming that v does not vary more than a factor 2, the transit rate will not be greater than 0.4 v in the
vertical with the lowest velocity.
Practical operation
See Paragraph 5.6.2.6.
Laboratory analysis
See Paragraph 5.6.2.2 and 8.1

See Paragraph 5.6.2.6
nozzle diameter: 0.006 or 0.013 m (1/4, 1/2 inch)
weight: 45 kg

A
...d
1
2 vasa rn
ilm tp
ca lg
e sn
a d eilfn
lab ile sftru m e torfe silta
n dsa n
dpa
rticles
3
D
1 isa d svte
o afonirtdra to
see frsmp tw
o n
sslva o a n sn p
2
3
R
C
iS
N
.e
h
o frrh
ite
nro
s
tvS
d en
d
fn
in
a
,Jso
le
cn
mC
.u sp
siHe
g
lra
Fn
m
a lo
p
,r,..tH
E
C
airn
rfL,ua
g
1
e
o
9
o
n
fc8
tydm 5
o
e
o
m
.cn
e
rS
h
,G
pcltso
iu
1
C
a
p
9
lcb
.D2e
oe
,ivnN
stte
id
disd
a
riin
Au
lio
b
S
b
rtsio
u
e d
r,u
,H
n
m
g
,b
M
.Se
n
tl,e
G
B M
rm
1 a
s
9
8u
n
y
3
.r
e
Nmewentse
T inE
lp
b
lseE
ylse
ft-u
ra
rM
ya
tsH
amb
u
rg
usrid
ng
a
Cu
xsa
S e d m en
tlo DiH su
c h a
rz,u ge o
fli.cA L a rn g e R
inrAveCs
E
,V D
o.lW
1,6
019
N 8
o0.H yc4
C h
ln
oo
a g
ib o
Bage
a
S u
re
pin
g
W
d
ea
te
S ia
n
d
tS
m
eu p
e
n
d
le
mr.
M
a
n
u
a
lS
e
d
im
e
n
tT
ra
n
sp
o
rtM
e
a
su
re
m
e
n
ts P
a
g
e
:5
.8
5

Figure 1
Collapsible Bag depth-integrating sampler

5.6.3 Pump sampler
5.6.3.1 General aspects for sampling in unidirectional flow
To obtain a reliable average sediment concentration, the sampling or measuring period should be rather large
(about 300 seconds). Furthermore, the collection of a large sediment sample for size-determination by
sieving or settling tests requires the sampling of a relatively large water volume (about 25 to 50 litres).
Both requirements can be satisfied by collecting water samples by means of a pump in combination with an
in-situ separation of water and sediment particles.
Usually a pump sampler consists of a submergible carrier (with intake nozzle, current meter and echo-
sounder), a deck-mounted pump and a flexible hose connecting the intake nozzle and the pump. The hose
diameter should be as small as possible to reduce the stream drag on the hose. Using a hose diameter (bore)
in the range of 0.003 to 0.016 m, the pump discharge will be in the range of 1 to 30 litres per minute. In case
a deck-mounted pump is used the maximum suction lift will be about 7 m. Assuming a static lift (= height of
pump above water level) of about 2 m, the suction lift available for operation of the pump will be about 5 m
resulting in a maximum hose length of about 50 m (Van Rijn, 1979). In extreme deep waters an underwater
pump must be used. Operation of a pump sampler is limited to flow conditions with velocities smaller than 2
m/s because of excessive stream drag on the pumphose and carrier.
Type of pump
Peristaltic pumps or propeller type pumps can be used.
Peristaltic pumps (24 volt/220 volt) (see www.globalw.com) have proven to be very efficient for pump
sampling in river and coastal conditions. The discharge is relatively small (1 1/min) yielding a relatively
small water-sediment sampling which can be handled easily. The hose diameter is extremely small (0.006
m), which reduces the fluid dragforces on the hose. The pump direction can be easily changed to remove
small objects (shell fragments, organic materials, etc.) blocking the intake nozzle. The intake velocity in
relation to the static suction lift (= height of pump above water surface) and the hose length is given in Fig. 3.
Propeller type pumps (common garden pumps; www.metabo.cm/eb/com/en/produkte/gardenpumps) produce
a relatively large discharge (10 1/min) resulting in the handling of a large water-sediment volume. The hose
diameter is in the range of 0.01 to 0.016 m.
Intake velocity and accuracy

Ideally, the intake velocity of the water-sediment sample should be equal to the local flow velocity (see
Paragraph 5.6.2.1). In conditions with varying velocities this would mean a continuous adjustment of the
intake velocity. For practical reasons it is preferable to operate the pump system as much as possible with a
fixed discharge and hence a fixed intake velocity. This can be done by using a fixed intake velocity for each
class of flow velocities (see following table).
Using these values, the hydraulic coefficient will be in the range of 0.8 to 2.0 during pump sampling,
resulting in a maximum error in the concentration of about 20%, which is quite acceptable for concentration
measurements (see Paragraph 5.6.2.1). For reasons of effective sampling, the flow velocity in the hose must
be larger than 1.0 m/s (Van Rijn, 1979).
Local flow Intake Pump discharge

velocity velocity (l/min)
Intake nozzle Intake nozzle Intake nozzle

(m/s) (m/s) diameter=0.003 m diameter=0.01 m diameter=0.016 m
0.5 - 1.0 1.0 0.45 5.0 12
1.0 - 1.5 1.25 0.55 6.25 15
1.5 - 2.0 1.75 0.80 8.75 21
2.0 - 2.5 2.25 1.0 11.25 27

Determination of hose dimensions and carrier type
The maximum hose length is about 50 m using a deck-mounted pump, which enables sampling in channels
with flow depths upto 25 m. In shallow water the hose dimensions and also the carrier dimensions can be
reduced substantially.
Figure 1 can be used to select the required hose dimensions (bore diameter and length) for a propeller type
pump and carrier type, given specific flow conditions. Figure 2 shows an example of an under-water carrier,
as used in tidal waters in the Netherlands.
The carrier should satisfy the following specifications:

1. streamlined body,
2. small cross-section area,
3. low centre of gravity,
4. suspension cable attached to swivel,
5. large tail fin,
6. supporting pins (at bottom side) and hand grips for transportation,
7. adjustable ballast weight (lead).
Determination of sampling position

To determine the actual position of the intake nozzle above the bed, an echo-sounder attached to the carrier
should be used. Usually it is desirable to have a sampling position close to the bed. This can be achieved by
placing the carrier on the bed. In that case the position of the intake nozzle is equal to the distance between
the nozzle and the underside of the carrier (see Figure 1).
When measurements at a few centimetres above the bed are necessary, special equipment should be used
which allows the vertical adjustment of the intake nozzle over a certain range (remote controlled) or
additional intake nozzles should be attached to the carrier.
Handling of water volume after sampling

When a propeller type pump is used, a large water-sediment volume is obtained. The handling of a large
water-sediment volume requires the in-situ separation of water and sediment particles. A practical solution
can be obtained by using the filtration method or the sedimentation method. In case the collection of a large
quantity of sediment particles for size analysis is not of importance, it is advisable to take a relatively small,
but representative sample of the total water sample to determine the sediment concentration. This latter
method is discussed as the pump-bottle method (Paragraph 5.6.3.5) .
When a peristaltic pump is used, a relatively small amount of water is obtained (2 liters in 5 minutes) . This
small sample can be stored in a bottle and returned to the laboratory for analysis.
Automatic pump samplers

Site location, flow conditions, frequency of collection and operational costs sometimes make collection of
sediment data by manual methods impractical. For these reasons automatic pumping type samplers have
been developed (FTS: see www.ftsinc.com); (www.ismatec.com).
Such a sampler consists of:

1. intake nozzle and tubing system,
2. pump to draw water-sediment samples from the flow (flushing of tubing system after sampling is
necessary),
3. sample container unit to hold sample bottles in position for filling,
4. sample distribution system to divert a pumped sample to the correct bottle,
5. activation system that starts and stops the sampling cycle.

Propeller garden pump

weight: 5 kg
discharge: 0 to 30 l/min
energy: 220 volt
hose length and diameter: 10 to 50 m; 0.01 and 0.016 m
Peristaltic pump
type: Ismatec (www.ismatec.com)
weight: 5 kg
discharge: 0.1 to 0.5 l/min
hose length and diameter: 5 to 50 m; 0.004 m
References
Conference on Instrumentating Oceanography, I.E.R.E., Proc. No. Bangor, England
Rijn, L.C. van, 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Report S404, The Netherlands

Figure 1
Under-water carrier

Figure 2
Under-water carrier for pump samplers

Figure 3
Hose length and intake velocity of peristaltic pump

5.6.3.2 General aspects for sampling in oscillatory flow
Pump sampling is an attractive method for concentration measurements in coastal conditions because a
relatively long sampling period can be used which is of essential importance to obtain a reliable time-
averaged value. The sampling period should be rather long (15 min) in irregular wave conditions (at least
100 waves).
A problem of sampling in conditions with irregular waves is that the magnitude and direction of the fluid
velocity is changing continuously. This complicates the principle of isokinetic sampling in the flow
direction. A workable alternative may be the method of normal (or transverse) sampling, which means that
the intake nozzle of the sampler is situated normal to the plane of fluid velocity.
Bosman et al (1987) studied the sampling error related to the orientation of the intake nozzle, because they
were interested in pump sampling under wave conditions. They found that a transverse pumping direction
yields good results. The intake nozzle is directed downward or normal to the plane of orbital motion. Figure
1 shows the ratio c/co as a function of the ratio u/uo and the nozzle orientation for 170 Pm-sediment (c =
measured concentration, co = original concentration, u = intake velocity, uo = local ambient velocity). For
transverse orientation (90) and ratio of u/uo> 2, the c/co-ratio is about D=0.7 to 0.8 which means a
systematic error of 20% to 30% in the measured concentration. Similar results were obtained for 220 Pm,
280 Pm, 360 Pm and 450 Pm-sediment. This systematic error can be eliminated by multiplying the measured
concentrations with a factor l/D.
Peristaltic pumps have proven to be very efficient in coastal conditions.
References
Antsyferov, S.M., Basinski, T. and Pykhov, N.V., 1983. Measurements of Coastal Suspended Sediment
Concentrations. Coastal Engineering, 7, Elsevier Publishers, Amsterdam, The Netherlands
Bosman, J.J., Velden, E.T.J.M., van der and Hulsbergen, C.H., 1987. Sediment Concentration
Measurement by Transverse Suction. Coastal Engineering, Vol. 11, p. 353-370

Figure 1
Directional sensitivity of 3 mm nozzle for 170 Pm sediment

5.6.3.3 Pump-Filter sampler
Principle
The water-sediment sample is pumped through a filter which separates all particles larger than the mesh size
of the applied filter material. The method is shown schematically in Figure 1A. To separate the sand fraction,
nylon filter material with a mesh size of 50 Pm can be used. The water volume is recorded by means of a
(simple) volume meter. After taking a sample, the filter system is opened and the filter material with the sand
catch is removed and returned to the laboratory for drying, weighing and size analysis. During removal of the
filter, the pumping is continued using a bypass system. The filtration method cannot be used in a silty
environment with silt concentrations larger than about 50 mg/1 because of rapid filter blocking by the fine
silt particles.
Practical operation
2. open filter house and install 50 ym- filter,
3. close filter house (note initial reading of volume meter),
4. adjust intake velocity depending on measured flow velocity (discharge meter and valve A, see Figure
1A),
5. wait one minute to flush the pump hose (through bypass system),
6. pump through filter by using valve B (if filter is blocked, indicated by discharge meter, stop
sampling),
7. use sampling period of 5 minutes,
8. pump through bypass system by using valve B (note final reading of volume meter),
9. open filter house and remove filter with sediment catch (wash inside of filter house free from
sediment particles before removal of filter),
10. install new filter,
Remarks
1. the filter system may never be used without filter material to prevent the volume meter from
damage,
2. cavitation (noise!) due to low water pressure can be eliminated by closing valve A somewhat till the
cavitation noise has disappeared.
Laboratory analysis
Simple
1. wash filter and sediment sample with fresh water to remove silt and salt particles,
2. dry and weigh filter with sand particles,
3. remove sand particles from filter material by brushing,
4. weigh filter material and determine dry mass of sand particles,
5. determine size distribution of sand particles by sieving (or settling tests).
Detailed
1. wash sediment sample off the filter material into a beaker,
2. decant through a 50 Pm- sieve to remove silt particles,
3. wash sand sample with distilled water to remove silt and salt articles,
4. wash sand sample in a small beaker,
6. determine size distribution of sand sample.

The sand concentrations are determined as:
c=Gs/V
in which:
To determine the overall accuracy of the pump-filter method, some concentration measurements were carried
out in a laboratory flume using a siphon sampler for comparison. The siphon sampler consisted of a short
hose connected to an intake nozzle in the flume. The intake nozzles of both methods were installed next to
each other at a fixed height above the flume bottom. In lateral direction an uniform concentration was
assumed. To simulate field conditions as much as possible, the pump-filter system was operated with a static
lift of 2.0 m and a hose diameter (bore) of 0.016 m and a hose length of 50 m. Filter material with a mesh
size of 50 Pm is used for separation of the sand fraction. The hydraulic coefficient of the pump sampler was
varied in the range 0.7 to 1. Figure 1B shows the sampling error in the concentration for two types of sand.
The concentration range is 50 to 1000 mg/1. Each point represents an average value, while also the highest
and lowest value are indicated. The maximum overall error for all measurements is about 10% (Van Rijn,
1979).
Field measurements (see Paragraph 5.4) indicate a sampling error in the concentration of about 20%
Advantages
1. simple and reliable instrument for sand particles
2. simple laboratory analysis
Disadvantages
1. not for silty materials
3. winch facility required
References
Van Rijn, L.C., 1979. Pump Filter Sampler. Delft Hydraulics Laboratory, Report S4CW, The Netherlands

Figure 1
Pump Filter sampler

Figure 2A
Pump Filter sampler; filter with sediment catch
Figure 2B
Pump Filter sampler; filtration unit

Figure 2C
Pump Filter sampler

Figure 3
Measuring Sheet Pump Filter sampler

5.6.3.4 Pump-Sedimentation sampler
Principle
The method is based on the filling of a large calibrated container (= 50 liters), in which the sand particles can
settle (bottle 1), as shown schematically in Figure 1A. Using a settling height of about 0.75 m, the sand
particles larger than 50 a 60 Pm can be separated in about 5 minutes. A high separation efficiency can be
obtained by using a conical container and a vibrator to avoid settlement of the sand particles on the inside of
the container (Van Rijn, 1980). To determine the silt concentration (particles smaller than 50 Pm), a small
water sample (bottle 2) can be tapped during emptying of the container.
Practical operation
2. install bottle 1 and 2 (Fig. 1A ),
3. bring bottle 1 under opening of container (slide valve C to position that bottle 1 can be filled),
4. adjust intake velocity depending on measured flow velocity (discharge meter and valve A),
5. wait one minute to flush the pump hose (through bypass-system),
6. fill container (valve B) in about 5 minutes,
7. take a settling period of 3 minutes without using vibrator,
8. start vibrator to remove sand from inside of container (vibration period of about 2 minutes),
9. note water volume (gauge glass),
10. stop vibrator,
11. empty container (slide valve C to position that filling of bottle 1 is blocked),
12. take a water-silt sample (bottle 2), use a filling time equal to emptying time of container (tap D),
13. remove bottle 1 and 2,
Remarks
1. a calibrated measuring glass can be used for the volumetric determination of the sand sample (bottle 1),
2. the inside of the container must be cleaned regularly.
Laboratory analysis
Bottle 1
1. wash sand sample over 50 Pm-sieve, use distilled water to remove silt and salt particles,
2. wash sand sample into a small beaker,
4. determine size-distribution of sand sample (sieving or settling test).
Bottle 2
csilt=Gsilt 2/Vbottle 2
csand=Gsand 1/Vcontainer + Gsand 2/Vbottle 2
in which:
Gsilt 2 = dry mass of silt sample from bottle 2 (mg)
Vbottle 2 = volume of water sample in bottle 2 (1)
Vcontainer = volume of water sample in container (1)

Laboratory measurements were carried out to determine the sampling efficiency. A siphon sampler was used
to determine the original concentrations near the intake nozzle. The maximum sampling error was found to
be about 15% for sand concentrations larger than about 50 mg/1 (see Figure 1B). Probably, a minor part of
the sand particles settles on the inside of the container and is washed off during emptying of the container
(Van Rijn, 1980).
Field measurements in a tidal estuary (Western Scheldt, The Netherlands) have shown the presence of a
small amount of sand particles in bottle 2. Figure 3 presents the ratio of the sand concentration from bottle 2
and 1 as a function of the sand concentration from bottle 1. For sand concentrations larger than about 50
mg/1, the sand concentration from bottle 2 is less than 20% of the sand concentration from bottle 1 and may,
therefore, be neglected resulting in a considerable reduction of the laboratory analysis. Size analysis of
sediment samples collected in field conditions shows that nearly all (sand) particles larger than MO pm do
settle in bottle 1 during the sedimentation period of about 5 minutes (Van Rijn, 1980). Field measurements
have also shown that the silt concentration (bottle 2) can be determined with an inaccuracy of about 10%
compared with a simple bottle sampler (Paragraph 5.4.5).
measuring range: > 50 mg/l for sand and > 10 mg/l for silt
energy: 24 volt (vibrator)
Advantages
2. in-situ separation of sand and silt
3. relatively large sampling period (5 minutes)
Disadvantages
1. winch facility
3. relatively large cycle period
References
Van Rijn, L.C., 1980. Methods for in-situ Separation of Water and Sediment. Delft Hydraulics Laboratory,
Report SH04 II, The Netherlands

Figure 1
Pump Sedimentation sampler

Figure 2
Sedimentation container

Figure 3
Efficiency of Pump-Sedimentation sampler

Figure 4
Measuring Sheet Pump Sedimentation sampler


Manual Sediment Transport Measurements

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Manual Sediment Transport Measurements

Hochgeladen von

Copyright:

Verfügbare Formate

MANUAL SEDIMENT TRANSPORT MEASUREMENTS

1. INTRODUCTION, PROBLEMS AND APPROACHES Page

2. MORPHOLOGICAL DEFINITIONS AND PROCESSES Page

3. MEASURING PRINCIPLES, STATISTICS AND ERRORS Page

LIST OF CONTENTS February 2006

5. MEASURING INSTRUMENTS FOR SEDIMENT TRANSPORT Page

LIST OF CONTENTS February 2006

6. MEASURING INSTRUMENTS FOR PARTICLE SIZE AND FALL VELOCITY Page

7. MEASURING INSTRUMENTS FOR BED MATERIAL SAMPLING Page

8. LABORATORY AND IN-SITU ANALYSIS OF SAMPLES Page

LIST OF CONTENTS February 2006

10. INSTRUMENTS FOR BED LEVEL DETECTION Page

ANNEX A: MEASURING INSTRUMENTS FOR FLUID VELOCITY, PRESSURE Page

A3 Comparison of measured velocities 16

A4 Fluid pressure and wave height instruments 28

A5 Comparison of measured wave heights

LIST OF CONTENTS February 2006

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

1.2.2 Sedimentation and erosion problems

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Channel sedimentation Basin sedimentation Shoreline sedimentation

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Bed level at time T

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

TIDAL INLET CHANNEL

WATER INTAKE DOCK ENTRANCE

Entrance channels of harbours, docks and water intakes

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Point bar Pool

Propagation of tidal wave

Trenches for tunnels, pipelines and cables

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Harbours and ports

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Turning, mooring and settling basins

Flood plains and reservoirs

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

SEDIMENTATION AND EROSION NEAR INLET

SEDIMENTATION AND EROSION NEAR GROYNES

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Dredging and dumping of sediments

The three essential elements of dredging are:

Efficient management of dredging works requires:

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

The three most basic rules are:

Analysis of the physical system

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Field studies comprise:

Laboratory scale models consist of:

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Mathematical models consist of:

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

CHAPTER 1: INTRODUCTION, PROBLEMS AND APPROACHES February 2006

Usually, three modes of particle motion are distinguished:

CHAPTER 2: MORPHOLOGICAL DEFINITIONS AND PROCESSES February 2006

CHAPTER 2: MORPHOLOGICAL DEFINITIONS AND PROCESSES February 2006

moving as bed load

Bed material load

wash load moving as suspended load

CHAPTER 2: MORPHOLOGICAL DEFINITIONS AND PROCESSES February 2006

This Section presents an overview of :

2.3.2 Sediment classification

Sediments can be classified according to their genetic origin: