The Effect of Groynes On Rivers PDF

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/228716610
The Effect of Groynes on Rivers
Article
CITATION READS
1 561
1 author:
Mohamed Yossef
Deltares
21 PUBLICATIONS 124 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Sediment management in rivers and reservoirs View project
Improving the navigability of the Danube River - hydrodynamic and morphological modelling View project
All content following this page was uploaded by Mohamed Yossef on 23 January 2014.
The user has requested enhancement of the downloaded file.

The Effect of Groynes
Literature review
on Rivers
Mohamed F. M. Yossef
Dioc "Water"
The Effect of Groynes on Rivers
Literature review
Mohamed F. M. Yossef
Delft University of Technology

Faculty of Civil Engineering and Geosciences
Section of Hydraulic Engineering
e-mail: [email protected]
Delft Cluster project no. 03.03.04
23 August, 2002
Delft University of Technology

Faculty of Civil Engineering and Geosciences
Section of Hydraulic Engineering
23 August, 2002
Foreword
The following report is a literature review study carried out to acquire the background
knowledge, and the state of the art concerning the effect of groynes on rivers, as part of the
author’s Ph.D. study. This study is carried out within the framework of DIOC Water, theme
1.3 (Intermediate-scale morphological developments in rivers due to human interventions)
and Delft Cluster, theme 3 (Coasts and Rivers).
Abstract
Groynes are structures constructed at an angle to the flow in order to deflect the flowing
water away from critical zones. They are made of stone, gravel, rock, earth, or piles,
beginning at the riverbank with a root and ending at the regulation line with a head. They
serve to maintain a desirable channel for the purpose of flood control, improved navigation
and erosion control. In the River Rhine, which is considered the backbone of North-western
European waterways network, the primary objective of groynes is to provide a fairway of
sufficient depth and width. For example the River Waal, the most important branch of the
Rhine River in the Netherlands is regulated by around 500 groyne.
Within the framework of the research project “Space for the Rhine Branches” several
measures have been devised to achieve a decrease of the water levels at peak discharges,
one of those measures, is lowering of the existing groynes. The rationale behind this
proposal is that; due to large-scale erosion of the low-water bed through the past decades,
the groynes are now higher than necessary for keeping the main channel at depth. Lowering
the groynes along certain reaches of the river would result in a reduction of the effective
roughness during high water conditions thus, increasing the river’s flood conveyance
capacity.
If the groynes are lowered, however, the balance of hydrodynamic forces acting on the
groyne-fields will change, and there will be a large-scale morphological impact. To identify
this impact, a thorough understanding to the effect of groynes on the morphology of the
river is necessary. The sediment exchange between the groyne-fields and the main channel
needs to be more comprehensible.
The purpose of this report is to acquire the background knowledge required to study the
effect of groynes on a river. The characteristics of the existing groyne-fields along the Waal
River are presented. The hydrodynamic and morphological impact of groynes on a river
is described. Moreover, because navigation plays an important role in the interaction
between the groyne fields and the main channel, the navigation induced water motion and
its effect on the flow in groyne-fields is described. Finally, a review of some prediction
attempts to the interaction between the groyne-fields and the main channel is presented.
Table of Contents
1. Introduction
1.1. Background-----------------------------------------------------------------------------------------------1
1.2. Objective --------------------------------------------------------------------------------------------------2
1.3. Scale of the problem ------------------------------------------------------------------------------------2
1.4. Outlines of this report ----------------------------------------------------------------------------------3
2. Characteristics of Groynes
2.1. General ----------------------------------------------------------------------------------------------------4
2.2. Transverse structures – groynes ----------------------------------------------------------------------4
2.2.1. Types of groynes..........................................................................................................4
2.2.2. Design Considerations for Groynes .............................................................................5
2.3. Groynes in the Netherlands----------------------------------------------------------------------------7
2.3.1. Historical background ..................................................................................................7
2.3.2. Characteristics of the groyne-fields along the Waal River ..........................................9
3. The Effect of Groynes on a River

3.1. Flow near groynes------------------------------------------------------------------------------------- 12
3.1.1. Flow near a single groyne ..........................................................................................12
3.1.2. Flow pattern in groyne-fields.....................................................................................14
3.1.3. Submerged groynes....................................................................................................17
3.2. Morphological Effect of Groynes on a River ---------------------------------------------------- 18
3.2.1. General .......................................................................................................................18
3.2.2. Bed Degradation Caused By Long Constriction........................................................18
3.2.3. Bed Degradation Caused By Series of Groynes ........................................................20
3.2.4. Local Scour Near Groynes.........................................................................................21
4. Hydraulic Disturbances Caused by Navigation

4.1. Navigation induced water motion------------------------------------------------------------------ 25
4.1.1. Water movement around a ship .................................................................................25
4.1.2. Return current and water level depression .................................................................26
4.1.3. Ship waves .................................................................................................................28
4.2. Effect of navigation on the flow in a groyne-field ---------------------------------------------- 28
5. Morphological Interaction between the Groyne-Fields and the Main Channel

5.1. General -------------------------------------------------------------------------------------------------- 32
5.2. Estimation of erosion from groyne-fields -------------------------------------------------------- 32
5.2.1. Field studies ...............................................................................................................33
5.2.2. Analytical approach ...................................................................................................35
5.3. Effect of changing the existing groynes geometry---------------------------------------------- 39
5.3.1. Lowering the groynes.................................................................................................39
5.3.2. Lengthening or shortening of groynes .......................................................................40
5.3.3. Decreasing the groynes spacing .................................................................................40
References 41
Appendix I – Conceptual Model 44
Appendix II – The model of Sieben & Douben 47

Introduction
1. INTRODUCTION
1.1. BACKGROUND
In 1992, the World Wildlife Found published a report named ‘Living Rivers’. This plan has the
objective to recover the natural river landscape that will lead to a return of river characteristic plants
and animals. Next to this plan, many initiatives and plans from different points of view, like
navigation, nature and landscape have been published. However, safety should always have the first
priority and not to be put aside for nature development. For example, the construction of flood plain
forests will lead to an increased roughness of the river and to higher water levels upstream. It is
possible to find engineering compensation by e.g. lowering the groynes. In this case, nature and
safety would go hand in hand. However, the impact on the morphology in this case is unclear,
which means navigation might be in danger. This example emphasises that the river is a coherent
system, in which every engineering measure has implications on all the functions of the riverine
area. Consequently, the decision to execute any of the measures must be supported by good
arguments. In this way actors and stakeholders can be informed on the advantages and
disadvantages of the measure and decide if the measure is socially desirable.
Within the framework of the research project “Ruimte voor Rijntakken” (in English: “Space for the
Rhine Branches”) several measures have been devised to achieve a decrease of the water levels at
peak discharges, by means of enlarging river space. A wide range of fourteen measures is
mentioned in detail in the ‘Landscape Planning River Rhine’ (LPR) report, (see Figure 1-1). One
of those measures, is lowering of the existing groynes, say by 2 m is proposed.
Figure 1-1 River engineering measures as proposed in LPR report
The rationale behind this proposal is that; due to large-scale erosion of the low-water bed through
the past decades, the groynes that were constructed between 1860 and 1920 are higher than strictly
necessary for keeping the main channel at depth. Lowering the groynes along certain reaches of the
river, an activity requires a huge investment, would result in a reduction of the effective roughness
during high water conditions. Thus, increasing the river’s flood conveyance capacity. This
increased flood capacity would help restoring some of the riverine nature, which would in turn,
increase the effective roughness. This is important in the light of the policy objective to
accommodate higher flood discharges without raising the dikes.
1
Introduction
1.2. OBJECTIVE
If the groynes are lowered, however, the balance of hydrodynamic forces acting on the
groyne-fields will change, and there will be a large-scale morphological impact. This may involve
for example; tilting of the entire river, similar to the effect of the normalisation works in the first
half of the previous century. It may also involve the necessity of dredging or other maintenance
measures, so as to ensure desired navigable depth.
To carry on with such a plan, a thorough understanding to the effect of groynes on the morphology
of the river is necessary. The sediment exchange between the groyne-fields and the main channel
needs to be more comprehensible.
1.3. SCALE OF THE PROBLEM

Before going into the details of the problem, it is wise to classify the scale of the problem.
Following the classification of de Vriend (1999), the morphological process of a river could be
represented by a series of scale levels. Assuming that to some extent these scale levels can be
considered separately, they form a sort of cascade (Figure 1-2), in which, the micro-scale level
represents the small-scale bedforms, e.g. ripples and dunes. The meso-scale level is that of alternate
bars and cross sectional-profile evolution. The macro-scale level is that of meander formation, up
to longitudinal profile evolution of river reaches, e.g. in response to training works. Channel pattern
formation at the scale of the river basin constitutes the mega-scale.
aggregation step
extrinsic conditions Mega-scale

Spatial scale
Macro-scale
Meso-scale
Micro-scale
Predictability limit
Temporal scale
Figure 1-2 Scale cascade of morphological process according to de Vriend (1999)
In our case, we will adopt a rather simplified classification from that of de Vriend (1999). It consists
of two levels, a small-scale level, and a large-scale level. The small scale is to some extent a
combination between the micro and meso-scales of de Vriend. The large scale is equivalent to the
macro-scale while the mega-scale is beyond the scope of this study. The small-scale analysis will
be devoted to study in detail the local behaviour of a single/few groyne-fields under the effect of
the different hydrodynamic forcings, “spatial distinction”, for a period which is relatively short
“temporal distinction”. In the large-scale analysis, the impact on the whole river system will be
considered. The results from the small-scale analysis will then be aggregated to form some kind of
a forcing function to the large-scale stage. This function is supposed to be representative of the the
small-scale phenomena. An example is presented in Appendix I for a conceptual model for the
interaction between the groyne-fields and the main channel.
2
Introduction
The hydrodynamic forcing that governs the interaction between the groyne-fields and the main
channel is the resultant of two components. The first is the effect of navigation, which is a factor
that cannot be overlooked in a river like the River Rhine that is considered the backbone of the
North west European waterways network. The other is the current induced water motion. On the
one hand, the groyne-fields are filled with sediment during times of high discharge. On the other
hand, the navigation induced water motion is held responsible for eroding the groyne-field beaches;
this effect is largest during low discharges. For an overview on the different parameters that are
affecting the interaction between the groyne-fields and the main channel, see Figure 1-3.
Small-scale
Sediment exchange between the groyne-fields and the main channel
2D Approach
Discharge stage Discharge stage

LOW HIGH
dimensions
speed
distance effect of navigation effect of river flow effect of river flow effect of navigation
frequency
navigation induced large-scale eddies large-scale eddies negligible

water motion secondary flow
floodplain flow
Effect on sediment motion Effect on sediment motion

NET EROSION NET DEPOSITION
E = f1(Q) D = f2(Q)
Large-scale
Large-scale impact on the river system
Aggregation step 1D Approach
Extrinsic conditions Analysis for different scenarios
Figure 1-3 Overview on the parameters that affect the interaction

between the groyne-fields and the main channel
1.4. OUTLINES OF THIS REPORT

The aim of this report is to acquire the background knowledge required to study the morphological
interaction between the groyne-fields and the main channel. Chapter 2 describes aspects and
consideration for groynes construction, presenting the different types of groynes. Furthermore, the
characteristics of the groyne-fields along the Waal River are presented. Chapter 3 describes the
hydrodynamic and morphological impact of groynes on a river. In Chapter 4, navigation induced
water motion and its effect on the flow in groyne-fields is described. In chapter 5, a review of some
prediction attempts to the interaction between the groyne-fields and the main channel is presented.
3
Characteristics of groynes
2. CHARACTERISTICS OF GROYNES
2.1. GENERAL
River training is the stabilisation of the channel in order to maintain the desired cross section and
alignment. The practice of training a river dates back to the sixteenth century where the Yellow
River in China was trained by building embankments along its banks so that the flow would be
confined to a single deep channel, which would transport the sediment load to the sea. Modern river
training practice, however, started in Europe in the nineteenth century, driven by the demands of
the industrial revolution for the purpose of maintaining sufficient channel depth and a better course
for navigation.
In general, the objectives of river training may be summarised as:
- to increase the safety against flooding by accommodating the flood flow
- to improve the efficiency of the sediment transport
- to minimise bank erosion by stabilising the course of flow
- to direct the flow to a desired river stretch
- to reduce the probability of ice jamming
- and in most of the cases the primary objective of river training is to improve navigation by
maintaining channel depth
Natural processes and human interference may disturb the equilibrium between the sediment load
contributed to the channel and the transport capacity of the flow. Seasonal variations in the flow,
dredging of the river, construction of a reservoir, and deforestation in the catchment area are all
examples of causes of disturbance. Training structures are then necessary in order to protect the
channel against the changes that occur due to this disturbance. They could be classified into:
i. Bed fixation and bottom vanes
ii. Longitudinal structures
iii. Transverse structures
In this report, the focus will be only on the groynes as a transverse structure. For a complete review
of all river-regulating systems, see Przedwojski et al. (1995).
2.2. TRANSVERSE STRUCTURES – GROYNES

Groynes are structures constructed at an angle to the flow in order to deflect the flowing water away
from critical zones. They are made of stone, gravel, rock, earth, or piles, beginning at the riverbank
with a root and ending at the regulation line with a head. They serve to maintain a desirable channel
for the purpose of flood control, improved navigation and erosion control.
2.2.1. Types of groynes

Various types of groynes can be distinguished according to their construction, action on stream flow
and appearance. Beckstead (1975), (as reported by Przedwojski et al. 1995) considers the
following, necessary for a full description of groynes:
i. Classification according to the method and materials of construction:
Groynes may be permeable allowing the water to flow through at reduced velocities or
impermeable blocking and deflecting the current. Permeable groynes are fabricated from
piles; bamboo or timbers whereas impermeable groynes also called solid groynes, are
constructed using rock, gravel, or gabions.
4
ii. Classification according to submergence:

Groynes may be designed either as submerged or as non-submerged. Which of the two
types will be used is dictated by the design conditions. Usually impermeable groynes are
designed to be non-submerged since flow over the top of solid groynes may cause severe
erosion along the shanks. For submerged conditions, on the other hand, permeable
groynes may be designed owing to the fact that they disturb the flow much less than solid
groynes.
iii. Classification according to the action on the stream flow:
Groynes may be classified as attracting, deflecting or repelling groynes. Attracting
groynes point downstream, they serve to attract the stream flow towards themselves and
do not repel the flow towards the opposite bank. Deflecting groynes are generally short
ones and used for local protection. They serve to change the direction of flow without
repelling it. Repelling groynes point upstream. They serve to repel the flow away from
themselves.
iv. Classification according to their appearance in plan:
Groynes may be built with different planview shapes. Examples are straight groynes,
T-head, L-head, hockey shaped, inverted hockey shaped groynes, straight groynes with
pier head, wing, or tail groynes.
2.2.2. Design Considerations for Groynes

The most important considerations involved in groyne design are planview shape, length of the
groynes, spacing between groynes, orientation to the flow, crest elevation and slope, cross-section,
construction materials and scour; Alvarez (1989), Richardson et al. (1975), and Przedwojski et al.
(1995).
i. Planview shape:
Of the above mentioned types of groynes according to their appearance in planview, the
straight groyne is set at an angle from the bank and has a rounded head to provide extra
volume and area for scour protection at the outer end. The T-head groyne is normally set
at a right angle from the bank and it has a straight shank with a rectangular guide vane at
the outer end. L-head, wing or tail groynes have larger sediment deposits between
groynes, less scour at their head, provide greater protection to the banks and are more
effective in channelization for navigation when the length closes 45 to 65 percent of the
gap between groynes. Hockey-shaped groynes have scour holes that are more extensive
in area than the T-head groynes.
ii. Length of the groynes:
Groyne length depends on the location, purpose, spacing, and economics of construction.
The total length of the groyne includes the anchoring length, which remains embedded
in the bank, and the working length, which stays in the flow. The length can be
established by determining the channel width and depth desired. The working length is
usually kept between the lower and upper limits of the mean depth and a quarter of the
mean width of the free surface respectively. The anchoring length on the other hand is
recommended to be less than a quarter of the working length.
iii. Spacing between groynes:
The spacing between groynes is measured at the riverbank between their starting points.
It is related to river width, groyne length, velocity of flow, angle to the bank, orientation
to the flow, bank curvature, and purpose. However, it is often expressed as a multiple of
the groyne length. Richardson (1975) recommends a spacing of 1.5 to 6 times the
upstream projected groyne length into the flow. In order to obtain a well defined deep
channel navigation, to keep a spacing of 1.5 to 2 times the groyne length is recommended,
whereas for bank protection the ratio of spacing to groyne length is less and distances
from 2 to 6 times the groyne length are generally used, although there exists successful
5
examples of bank protection with short groynes spaced apart 10 to 100 times their length
where the banks are protected with riprap or vegetation. If the spacing between groynes
is too long, a meander loop may form between groynes. Long and far apart spaced
groynes may contract the flow resulting in channel degradation and bank erosion, and
cause a hindrance to navigation. If the groynes are spaced too close together on the other
hand, construction costs will be higher and the system would work less efficiently without
making best use of each individual groyne.
iv. Orientation of the groynes:
Groynes may be oriented perpendicular to the flow or be inclined either upstream or
downstream. Each orientation affects the stream in a different way and results in different
deposition of sediment in the vicinity of the groyne. A groyne pointing downstream is an
attracting groyne, which attracts the stream flow towards itself. Repelling groynes, which
repel the flow away, and deflecting groynes, which deflect the flow away from the bank,
point upstream.
A groyne that is oriented upstream causes more deposition than a perpendicular one at the
downstream bank and also at the area upstream where a reverse eddy is formed and
causes suspended load to settle. The amount of deposition between groynes is maximised
in case of upstream inclination due to their ability to protect bank areas upstream and
downstream of themselves. Therefore, groynes of this kind are best suited for bank
protection and sedimentation purposes. Groynes that are perpendicular to the flow have
protection over a smaller area. Downstream facing groynes are not suitable for bank
protection purposes due to their attracting effect on the flow. The flow towards the root
of the downstream groyne threatens the surrounding bank area as well as the groyne itself.
For the purpose of maintaining a deep channel to improve navigation on the other hand,
best performance is obtained by perpendicular or downstream pointed groynes.
v. Crest elevation and slope:
The crest elevation of groynes depends on the purpose and possible problems due to
overbank flow and ice. For bank protection, the crest should be at least as high as the
bank. To avoid ice overtopping the crest elevation should be higher than the expected
levels of ice. Crests may be either level or sloping downwards from the bank towards the
end of the groyne. For bank protection, sloping-crested groynes are recommended by
Alvarez (1989) with a slope of 0.1 to 0.25 due to their advantages of reducing scour at the
groyne end, less material needed for construction, faster deposits of sediment between
groynes. For navigation channel control, level crested groynes work best normal to the
flow or angled downstream, whereas, sloping crested groynes work best normal or angled
upstream, Richardson (1975).
vi. Cross-section of the groynes:
The crest widths range from 1to 6m and side slopes from 1:1.25 to 1:5. The minimum crest
width of 1m is controlled by the equipment placing the groynes and wider crests make
placing easier.
vii. Construction materials:
Examples of the wide range of materials used for the construction of groynes are timber
piles, tree trunks or branches, rock, soil gravel, sandbags, riprap, prefabricated concrete
elements, steel and wire, etc.
viii. Scour:
The expected scour depth should be taken into consideration in the determination of the
base depth of the groynes.
6
2.3. GROYNES IN THE NETHERLANDS
2.3.1. Historical background

The first major hydraulic engineering works since the Roman era were undertaken at the Rhine
bifurcation in the early 1700s. Until then, channel correction had been carried out and groynes,
dams, and revetments built, but they served only local proposes such as the protection of dike
sections. At the end of the 17th century, the Waal carried by far the greatest part of the Rhine
discharge, probably over 95%. Water supply to the Nederrijn and the IJssel was so small that
navigation became difficult.
Around 1775, the bifurcation of the Nederrijn
and the IJssel was reconstructed by digging a
new channel in the upper part of the IJssel.
Since then, the IJssel received about one third
of the discharge of the Pannerdensch Canal
which in turn received about one third of the
total Rhine discharge. The Waal carried the
remaining two thirds of the flow. Stabilisation
of the situation was achieved by moles,
constructed under the supervision of Brunings
who was one of the first engineers to carryout
accurate flow measurements in the river. Until
the 1960s, the riverbed of the Oude Rijn
continued to function as a spillway for excess Figure 2-1 Rhine branches – key map
water, causing numerous floods in the region,
see van Urk & Smit (1989).
The works at the bifurcations improved water distribution between the Rhine branches, but floods
continued to be hazardous. As administrative boards or water authorities with responsibility for
integrated river engineering did not exist until the 19th century, individual landowners built bank
revetments and groynes to protect their land from erosion and to increase sedimentation. However,
the irregular array of groynes and the presence of many sandbars in the channel, not only impeded
flow but also led to the formation of ice dams. A formal ban on the irregular construction of
groynes proposed by the States of Gelderland in 1602 and finally adopted in 1715, but it had little
effect because of the lack of supervision.
In 1809 and 1820, large areas were inundated and in 1821 a government committee was appointed
to make proposals for the solution of the problem. In 1825, the committee completed its report,
which was published in 1827. Meanwhile, prominent engineers such as Blanken, Goudriaan, and
Krayenhoff had published their different views on the best solution. Their recommendations varied
from the creation of new spillways by the partial removal of dykes, to the building of sluices, and
to the construction of a totally new canal from the entry of the Rhine in the Netherlands to the
IJssel, van Urk & Smit (1989). These drastic plans were never carried out. The solution that finally
adopted was outlined in a report submitted by inspectors Ferrand and van der Kun of the newly
formed Rijkswaterstaat (RWS). It proposed the removal of the sandbars and the constriction of the
river’s channels to accelerate flow.
As stated before, the channel constriction was the solution finally adopted to increase the discharge
capacity of the Lower Rhine. The method of channelization can best be demonstrated in a series
of maps for a stretch of the river that was relatively unaffected before work started. Because the
wider River Waal has a stronger tendency to form multiple channels than the much narrower,
strongly meandering River IJssel, the effect of channel constriction on the Waal was greater. The
various operations carried out to transform the unregulated River Waal into a single constricted
channel are shown in Figure 2-2, and Figure 2-3.
7
Figure 2-2 The Waal River near Km 899-901 showing the channel changes over period of about 200
years , source: van Urk & Smit (1989).
Figure 2-3 The Waal River near Km 892-894 showing the channel changes since the 18th century, source:
van Urk & Smit (1989).
What the old maps do not show is the major degradation process going on in the river. The
degradation process in the Lower Rhine is complicated; it is mainly due to shortening of river
bends, dredging activities, regulation of tributaries, and channel constriction through series of
groynes.
Near the entry point of the Lower Rhine to the Netherlands, bed degradation mainly occurred
between 1925 and 1960 during that period degradation was about 1.0 m, van Urk & Smit (1989).
8
The average bed-degradation in the Lower Rhine branches since the end of the normalisation works
ranges between 0.4 m to 2.2 m, see Table2-1. Visser (2000) estimated that the bed level of the
Rhine branches did not reach equilibrium yet. However, the rate of degradation will be considerably
slow when compared to that of the early period after the completion of the normalisation works
Table2-1 Average bed degradation in the different Rhine branches (after, Visser 2000)
River section Rhine Km Period Average degradation (m)
IJssel 878.5 – 1005 1938 - 1990 -0.40
Lek and Neder-Rijn 878.5 – 989 1933 - 1990 -0.90
Pannerden Canal 867.5 – 878.5 1926 - 1990 -2.20
Waal 867.5 – 952 1926 - 1990 -0.70
Boven-Rijn 857.5 – 867.5 1934 - 1990 -1.20
2.3.2. Characteristics of the groyne-fields along the Waal River

Based on field measurements during the years 1996-1997, the Characteristics of the groyne-fields
along the Waal River are estimated. With reference to the definition sketch Figure 2-4, the
dimensions of the groyne-fields could be defined as (after Schans, 1998):
- A: groyne-field length
- B: groyne-field width
- C: length along the waterline
- D: beach width
- E: distance between the normal line and the thalweg
- F: river width (between groynes)
- G: orientation of a groyne (wrt. the line ⊥ thalweg)
- H: orientation of the groyne-field (wrt. North)
Figure 2-4 Definition sketch for the dimensions of an arbitrary Groyne field, after Schans (1998)
9
Two more parameters are important for characterising the groyne-fields, viz. the beach slope and
the groyne-field bed material size (D50).
Introducing two additional dimensionless parameters, (Schans, 1998), the first is the curvature
index (CI), and the second will be called the path index (PI). The curvature index could be defined
as the ratio between the length along the waterline (C) and the groyne-field length (A), and it gives
an indication about the sinuosity of the beach profile in a groyne-field. CI has a minimum of one
and the higher it gets the more curved the water line is. The path index (PI) could be defined as the
ratio between the distance of the thalweg from the groyne (E) and the width of the river (F). PI
should always be less be than the unity, and for any value other than 0.50 the thalweg is not in the
centreline of the river, (PI) gains more importance in curved parts of the river as an indication for
the location of the deep part of the river.
C E
CI = and PI =
A F
The characteristics of the groyne-fields along the Waal River could be summarised in Table 2-2,
and the accuracy of the measurements of the different parameters is presented in Table 2-3.
Table 2-2 the characteristics of the groyne-fields along the Waal River
Parameter Mean Standard Min. Max. Median Mode
deviation
Groyne-field length (A) 198.2 37.7 50 420 200 200
Groyne-field width (B) 67.9 28.6 0 175 65 50
Length along the waterline (C) 215.1 43.5 100 480 210 200
Beach width (D) 25.1 21.2 0 150 20 0
Distance between normal line and 129.8 93.6 10 320 123.8 25
thalweg (E)
River width (F) 279.5 35.2 252 412 260 260
Orientation of a groyne (G) -8.0o 8.7o -30o 10o -5o 0
Orientation of the groyne-field (H) 86.2o 31.4o 2o 150o 91o 96o
Bed material D50 (µm) 439.5 264.5 200 1300 347.5 225
Beach slope 0.042 0.008 0.03 0.05 0.04 0.05
Curvature index (CI) 1.097 0.117 1 1.92 1.056 1
Path index (PI) 0.478 0.338 0.03 0.949 0.046 0.077
Dimensions are in meters
Table 2-3 The accuracy of the different parameters

Parameter Accuracy No. of data points
Groyne-field length (A) ±5m 799
Groyne-field width (B) ±5m 738
Length along the waterline (C) ± 10 m 741
Beach width (D) ±5m 734
Distance between normal line and thalweg (E) ±5m 736
River width (between groynes) (F) ±5m 799
Orientation of a groyne (G) ± 3o 792
Orientation of the groyne-field (H) ± 3o 798
Bed material D50 Unknown 46
Beach slope Unknown 40
10
Statistical representation of the groyne-fields length (A) Statistical representation of the groyne-fields width (B)
25% 100% 14% 100%

12%
Cumulative %
Frequency %
Cumulative %
20% 80% 80%
Frequency %
10%
15% 60% 60%
8%
10% 40% 6% 40%
5% 20% 4%
2% 20%
0% 0%
0% 0%
110
140
170
200
230
260
290
320
350
380
410
50
80
105
120
135
150
165
15
30
45
60
75
90
0
length A (m) Width B (m)
Frequency Cumulative % Frequency Cumulative %
Figure 2-5 Statistical representations for the groyne-fields length (A), and width (B), for the Waal River
Based on the previous analysis and with consideration to the representativeness of groyne-field
dimensions, it is possible to assume the following values for a simple geometry that is appropriate
to schematise the groyne-fields of the Waal River,
A ≅ 200 m
B ≅ 60 m
D ≅ 25 m
F ≅ 260 m
G ≅ 0o
H ≅ 90o
D50 ≅ 350 µm
Beach slope ≅ 1:25
CI ≅ 1.0
PI ≅ 0.5
The bed material in the groyne-fields generally consists of well-sorted medium sand with around
70% -by weight- between 250µm and 500µm. At sites which are more exposed to the current, the
bed material is less well-sorted and an armour layer of gravel may be found. At these sites, more
than 25% of the sediment is coarser than 2.0mm and around 57% is between 125µm and 500µm.
The amount of clay and silt in the river bed is insignificant in this stretch of the River Waal the
sediments contain no coarse organic material nor any significant amounts of fine detritus.
11
The effect of groynes on river
3. THE EFFECT OF GROYNES ON A RIVER
3.1. FLOW NEAR GROYNES

The flow field near groynes differs significantly in the case of a single groyne from that near a
series of groynes.
3.1.1. Flow near a single groyne

The simplest case of groyne layout is a single groyne in a straight reach. The groyne confines a
certain part of the river cross section and affects appreciably the kinematic structure of the flow in
its vicinity. Mean velocity and specific discharge increase due to the constriction. The increase in
the mean velocity leads to a rise in their gradients and more intensive generation of
macro-turbulent∗. Many researchers investigated the flow pattern in the vicinity of a single groyne
and found some significant flow characteristics that could be highlighted in the following points:
- Separation region
- Migrating horizontal large eddies
- Water level fluctuation
Geometry of separation region

The characteristics of the separation region formed behind a groyne of non-overflow type were
investigated by many researchers, e.g. Ishii et al. (1983), Chen & Ikeda (1997), and Ouillon &
Dartus (1997). Such studies gave an impression about the geometry of the separation region
downstream a groyne in a rectangular channel.
In a study for the recirculation zone induced by sandbars along the Colorado River, (Schmidt et al.
as reported by Chen & Ikeda 1997) subdivided the flow field into four main zones;
- main flow zone
- return flow zone
- shear layer
- reattachment zone
From the tip of the groyne to the opposite channel bank, the flow velocity is accelerated because
of the reduction of the channel width. This is called the main flow zone. The return flow zone is
located at the downstream side of the groyne, generally with two relatively large eddies. The centre
of the larger one is located at a distance of about 6 times the groyne length. The other eddy is
smaller, of which the centre is about one time the groyne length. A velocity difference exists
between the main flow zone and the return flow zone, which leads to the formation of a shear layer
between the two zones.
The reattachment zone is usually simplified by most of the researchers into a point. This point is
defined as the point at which the boundary streamline reattaches to the channel boundary. However,
the instantaneous reattachment point fluctuates back and forth, mainly due to the intermittence of
eddies in the shear layer and the unstable balance of entertainment and pressure gradient between
the main flow and recirculation zone. We can regard the point with a maximum instantaneous
velocity of zero as the upstream end of the reattachment zone, and the point with a minimum
instantaneous velocity of zero as its the downstream end. Chen & Ikeda (1997) observed that the
length of the reattachment zone is almost constant and over the range from 11 to 17 times the length
∗
The description of the macro-turbulence flow structure is related to the local scour problem which is out
side the scope of this study
12
of the groyne, i.e. the reattachment zone covers a distance of around 6 times the length of the
groyne.
Nevertheless, the time-averaged reattachment point could be defined as the point at which the time-
averaged velocity is zero. According to Chen & Ikeda (1997) the reattachment point is located at
a distance of about 14 times the length of the groyne. In a comparison between numerical model
investigation and experimental results, Ouillon & Dartus (1997) reported that -for the experiments-
the reattachment length is in the order of 12.5 times the length of the groyne, and it is around 11.5
times the groyne length according to Tingsanchali & Maheswarn (1990).
In an extensive experimental study Ishii et al. (1983), investigated the effect of some dimensionless
parameters on the shape of the separation region (only subcritical flow was studied). They reported
that the shape of the separation region is hardly affected by Froude number (Fr), and it had
dimensions of a length that ranged from 10 to 12 times the groyne length, and the maximum width
measured from the sidewall was 2 times the groyne length (for a fixed angle of 90o, and a relative
projected length of 10% from the channel width). Yet, with the increase of the relative projected
length to the flow from 10% to 40% of the channel width, the relative separation length decreased
from 12 to 7; but the reattachment angle of the boundary streamline by which the main flow and
the separation region are bounded remained constant at 15o.
Furthermore, changing the projection angle of the groyne to the flow from 90o to 150o i.e. pointing
downstream, the geometry of the separation region remained almost constant. Yet, by decreasing
the groyne angle from 90o to 30o i.e. pointing upstream, the relative length decreased from 14 to 11,
but both the relative width and the reattachment angle remained constant at 2, and 10o respectively.
The upstream separation angle is governed only by the groyne projection angle to the flow. It varies
from 30o to 60o, for a change of the groyne projection angle from 30o to 90o, and remains constant
at 60o for any groyne angle more than 90o.
From those results, we can deduce that the separation region has a length that could vary from 7
times the groyne length (for relatively long groynes), to 15 times the groyne length. Yet, the relative
width is less varying and has a value that is slightly less than two times the groyne length.
Depending on the relative wall roughness, the downstream reattachment angle could also vary from
15o to 10o with the higher values for relatively high wall roughness. Moreover, the upstream
reattachment angle varies from 30o to 60o with the variation of the groyne projection angle.
Horizontal large eddies

Another important aspect of the flow field near a groyne is the horizontal large eddies that shed
from the tip of a groyne. Through measuring the water level fluctuations along the centreline of the
migrating vortices, Chen & Ikeda (1997) found that there is a clear periodic water level fluctuations.
These water level fluctuations have a clear phase difference between the signals at two consecutive
points along the centreline of the migrating eddies. The lag time between the peak of these two
signals is considered the time that a migrating eddy takes to move between those two points.
Applying FFT method to analyse the periodicity and the time lag between any pair of signals, he
found that the average migration velocity of the eddies is nearly constant and takes a value slightly
(1.5%) higher than the mean flow velocity.
As the eddies move downstream they merge with each other. Thus, their length-scale increases in
the downstream direction. Since the migrating velocity is constant, the time-scale of the eddies
should also increase in the downstream direction. Chen & Ikeda (1997) showed the increasing
pattern in the time-scale of horizontal eddies, this increase indicates that there is a frequent merging
of small-scale eddies after shedding from the groyne tip, until a certain distance and then remaining
constant.
Water surface fluctuation

The water surface fluctuates as the horizontal large eddies migrate down stream, that is why the
properties of the eddies could be studied by measuring the water surface fluctuation. In general, the
13
water level increases at the upstream side of the groyne and decreases in the downstream side, and
continuously fluctuates as the horizontal eddies periodically shed from the tip of the groyne. Chen
& Ikeda (1997) studied the water surface fluctuations though plotting the root mean square of the
surface fluctuations at several cross sections downstream of the groyne. For every cross section,
he found that there is a peak, which indicates that the centre of the large eddy. However, the
influence of the groyne on the water level fluctuations extends to a distance of only 10 times the
groyne length.
3.1.2. Flow pattern in groyne-fields

Under conditions where the groynes are not submerged, the groyne-fields are not really part of the
wetted cross section of a river. Because of that, the flow pattern in the groyne-field is not directly
the result of the discharge in the main channel. Reducing the main stream velocity has no effect on
the flow pattern itself, whereas lowering the water level does, Uijttewaal et al. (2001) In the later
case the effect of the bottom slope become more pronounced, shifting the eddy centre towards the
main stream. Moreover, the flow pattern inside a groyne-field may change with the change of its
geometry, location along the river (inner curve, outer curve, or straight part), and/or the groynes
orientation, Przedwojski et al. (1995).
However, there is an indirect effect of the discharge on the flow pattern in the groyne-field. Because
of the flow that is diverted from the main channel into the groyne-fields, the water flows into the
groyne-field with low velocity through the downstream half of the interfacial section between the
groyne-field and the main channel. This water flows back to the main channel through a small
width of that section, just downstream the upstream groyne of the groyne-field, Termes et al.
(1991).
A typical result from one of the experiments done by WL | Delft hydraulics for a groyne-field in
a straight section of the river WL|Delft_Hydraulics (1987) is shown in Figure 3-1. It shows the flow
pattern in a groyne-field with dimensions 200m x 50m, at a river discharge of 1450 m3/s
(representative dimensions and discharge for the Waal River). We can clearly observe a large eddy
that covers the first three-quarters of the groyne-field develops directly downstream of the upstream
groyne where the main current cannot make a sharp bend into the groyne field. The point around
Figure 3-1 Flow pattern in a groyne-field – dimensions in metres, source: Brolsma (1988)
14
which the eddy circulate, is not really at the centre of the eddy as we may observe, it is very near
to the tip of the upstream groyne and dragged towards the main channel. Further downstream, the
main current does enter the groyne-field, consequently the stream width becomes larger. A second
-smaller- eddy develops at the upstream face of the second groyne where the outflow is hampered
by this groyne.
Based on model tests for groynes located along a river bend Klingeman et al. (1984), report that six
types of eddy patterns between groynes can be distinguished Figure 3-2.
- Type one: The circulation pattern of this
type is distinguished by the main flow that
is deflected outside the groyne field, and a
single eddy develops between the groynes.
This eddy is well developed and it could
prevent the main flow from penetrating the
groyne-field. Therefore, this pattern is
desirable for navigation purposes as a
continuous deep channel is maintained
along the face of the groyne field.
- Type two: In this type, a second eddy
appears but the main current is maintained
deflected outside the groyne-field.
- Type three: As the spacing between
groynes increase, type three-flow pattern
develops. The main current is directed into
the groyne-field, creating a much stronger
eddy near the upstream groyne, and greater
turbulence along the upstream face and at
the groyne lower head.
- Type four: in this type, the stability of the
upstream eddy is washed out, and a sing1e
strong reverse current occurs.
Figure 3-2 Types of flow pattern in groyne-field,
- Type five: In this type the flow, which is source: Przedwojski et al. (1995)
diverted by the upstream groyne, is
directed to the bank in the groyne-field.
Eddies form on both sides of this flow,
providing some protection for the bank.
- Type six: As the spacing between the groynes further increases, the downstream eddy, which
was providing the protection to the bank wanes, and the flow attacks the bank directly.
Furthermore, the maximum velocity along the bank inside the groyne-field of an aspect ratio of 2.5,
is roughly 40% of the velocity measured in a similar bend protected by riprap. This percentage is
slightly less than 40% when the spacing-length ratio decreased to 1.5, and equal to approximately
50% when the spacing increased to 3.5 times the groyne length.
In his observation on the effect of the geometry on the flow field in a groyne-field, Uijttewaal
(1999) concluded that the groyne-field length to width ratio determines the number and shape of
eddies that emerge in the stagnant flow region. An aspect ratio close to unity gives rise to a single
eddy, (Figure 3-4, upper part). A larger aspect ratio gives room for two stationary eddies, a large
one called primary eddy, in the downstream part of the groyne-field, and a smaller secondary eddy,
emerges near the upstream groyne. The extreme long groyne-field case of length to width ratio of
six, shows the penetration of the main flow into the groyne field, (Figure 3-4, lower part). The two
eddies remain in a relatively stable position, while the main flow field starts to penetrate into the
groyne field further downstream. In all cases, there is an eddy detaches from the upstream groyne
15
tip that travels along the main channel groyne-

field interface and eventually merges with the
primary eddy.
The significant difference between the mean,
and the instantaneous flow field was
highlighted by Uijttewaal (1999). In the upper
part of Figure 3-3, the mean flow field obtained
a groyne-field with length to width ratio (L/W)
= 3. In the downstream part of the groyne-field
a large eddy covering two-third of the groyne-
field area, is clearly visible. The lower left
corner contains a second eddy rotating much
slower and in anti-clockwise direction. While
in the upper left side where this second eddy
borders the main stream, no clear pattern is
observed. The indefinite velocity field in this
area is mainly due to the averaging process,
which obscures the dynamics of the
instationary flow. From an instantaneous
velocity field as shown in the lower part of
Figure 3-3, the strong time-dependent motion
of a large eddy that is advected through the
groyne field is visible. Comparing both figures
taken from the same experiment reveals the Figure 3-4 Mean flow field in case of length to
significant increase in the intensity of the eddy width ratio =1, 3, 6 consequently, after Uijttewaal
in the case of instantaneous flow field. (1999)
From the first look, a conflict might appear
between the results of the experiments done by
WL|Delft Hydraulics (1987) and those of
Uijttewaal (1999). However, through a closer
look to the original data of WL|Delft
Hydraulics (1987), we can conclude that it
compares well with the instantaneous flow field
reported by Uijttewaal (1999).
Close review to the flow field exhibits the fact
that the flow pattern when groynes are not
submerged is predominantly two-dimensional.
The small-scale three-dimensional turbulence
plays a minor role in the mass and momentum
exchange process between the groyne-field and
the main channel, Uijttewaal (1999). Whereas,
no strong three-dimensional large structures are
developing obviously due to the shallowness of
water, with the exception for the area near the
groyne head where the flow is strongly three
dimensional, Kerbs et al. (1999).
Moreover, these results emphasise the
importance of numerical simulation methods
that consider the large-scale dynamics, as the
exchange processes is highly affected by the
presence of large dynamic structures and their
Figure 3-3 Comparison between mean, and
associated mixing length scales. However, the instantaneous flow field, after Uijttewaal (1999)
effective length scale associated with the
16
mixing process is hard to determine from Reynolds-averaged simulation models since they do not
incorporate the time dependent large-scale motions as shown in the instantaneous flow field. This
feature could be dealt with by using, e.g. large eddy simulation.
3.1.3. Submerged groynes

Reviewing the available literature revealed the lack of investigations that deals with groynes during
the submerged condition. This might be due to that; the need to investigate the submerged groynes
did not arise as the groynes are mostly operating under emerged conditions. Another reason might
be the complexity and three-dimensionality of the problem, which require advanced measurement
techniques, and/or powerful three-dimensional computational abilities. However, the case of
submerged groynes was studied by e.g. Aya et al. (1997), Peng et al. (1997), Krebs et al. (1999),
and Tominaga et al. (2001).
Aya et al. (1997), reported a sharp decrease in the water level between the upstream and
downstream sides of the groynes, this means that the water surface slope between two successive
groynes is less than the slope in the main channel region.
Peng et al. (1997) compared three-dimensional numerical results with experimental results and
found that the flow pattern in the case of submerged groynes shows strong three-dimensional
features behind groynes. The recirculation size at the back of the groyne is reduced gradually as the
top of the groyne is approached. Consequently, the reattachment length decreases from bottom to
top plans. The location of recirculation centre also varies in Z-direction. It moves from the tip of
the groyne (near the bed), towards the bank (close to the top surface plane). In the upstream face
of the groyne, the flow shows an upward motion because of the blockage effect of the groyne.
It should be also noted that the when groynes are submerged the flow over the groynes acts as a
damper for the large horizontal structures, ultimately causing it to disappear when reaching a high
enough submergence level.
Another important aspect for the flow pattern in the case of submerged groynes, is the secondary
flow structure. It occurs simply because of the existences of groynes and the disturbance that it
presents. On top of the groyne, the secondary flow (in YZ-plane) has the following characteristics:
Near the surface, it is from the bank side towards the mid-channel, and near the groyne top, it is the
other way around. Behind the groyne in the recirculation region, the secondary flow also goes
towards the mid-channel near the water surface, and near the bed, it is from the mid-channel
towards the groyne field as well. Krebs et al. (1999) reported the same feature, they found that in
the case of submerged groynes, there is a near bed flow from the mid-channel towards the
groyne-field. Yet, in the case of emerged groynes, the secondary flow is negligible.
As mentioned before that the spacing between the groynes affect the flow pattern in the XY-plane;
it affects the flow pattern in the YZ-plane as well. Separation flow over the upstream groyne may
reattach the groyne-field bed and the bed shear stress recovers its large value (that is usually
reduced because of the groynes), if the groynes are spaced far apart. Too close groynes will prevent
the flow reattachment to the bed maintaining the bed shear stress at low value, see Peng et al.
(1997).
17
3.2. MORPHOLOGICAL EFFECT OF GROYNES ON A RIVER
3.2.1. General
Erosion can refer to a multitude of natural process, such as soil erosion, beach erosion, or riverbank
erosion. It may result from flow of water or air, or from wave action. For localised erosion the word
scour is often used, scour at the head of a groyne, at an abutment, or at a bridge pier, etc. A reach
of the river may also scour but then it is usually referred to as degradation.
Scour is a localised lowering of riverbed that is usually linked to structures. It could be subdivided
to:
- Constriction scour, arising from the constriction of the waterway by the presence of the
structure. It changes the cross section geometry in the area near the structure and normally it
does not extend to a longer distance.
- Local scour, resulting from the effect of the structure on the local flow pattern and the
generation of macro-turbulent in its vicinity. It is always more pronounced than the constriction
scour. The local scour maybe – in most cases – superimposed on constriction scour.
Further, the scour may be clear-water scour, or live-bed scour. Clear-water scour refers to
conditions when the bed material upstream of the scour area is at rest. Live-bed scour occurs under
conditions of general sediment transport i.e. there is a continuous sediment supply to the scour zone.
The term degradation, in contrast to scour, implies a lowering of the riverbed that extends over a
long distance. Degradation may progress in the downstream direction, upstream direction, or in both
directions. For example, the construction of a dam would cause downstream propagating
degradation, lowering the downstream water level would cause degradation that propagates in the
upstream direction. Usually, channel bed degradation is accompanied by change in the river slope.
When a series of groynes is constructed, the bed forms near the groynes change because of the
combined effect of bed degradation due to the long constriction, and the local scour at the end of
each groyne. Bed degradation due to long constriction of an alluvial channel as well as, the local
scour phenomena near a single groyne has been discussed by many researches. Yet, the overall bed
degradation caused by a series of groynes, which have characteristics of both long constriction and
a single groyne was discussed by very limited number of researchers. In the following section, we
will present the morphological impact of groynes on a river.
3.2.2. Bed Degradation Caused By Long Constriction

Channel bed degradation that is originated from the construction of series of groynes is normally
dealt with as a long constriction scour. The effect of reducing the channel width is to increase the
bed shear stress, which would result in a considerable scour within the constricted reach. If the
constriction is long and permanent, the scour develops to an extended reach of the river
(degradation) and its effect further extends to affect the entire river.
Several analytical models have been proposed to compute the constriction scour, One of the earliest
works that presented a simplified one-dimensional model theory of the equilibrium depth for a long
constriction is due to Straub (1934). Several analytical investigations were conducted on this
problem e.g. Komura (1966), Gill (1981), Bhowmik (1989), and Klaassen (1995).
Komura (1966) and Gill (1981) exclusively extended the approach of Laursen and reached nearly
similar expression for the constriction scour. Komura (1966), investigated the effect of different
sediment sizes (D) and the standard deviation of the particle size distribution (σφ), his expression
reads:
18
 B ( 7 )  τ ( 7 )  D ( −
6 3 1
7)
 1   1   1  clear-water
d 2  B2   τ 2   D2 
=
d1  B ( 6 7 )  τ ( 2 7 )
1 1
    live-bed
 B2   τ 2 
LANE’S BALANCE
Another way to evaluate the ultimate response of the river to a long constriction, or any other
change in the forcing parameters, is the use of the so-called “Lane’s balance”. It was first
introduced by Lane in 1955, see Jansen et al. (1979), Bhowmik (1989), and Klaassen (1995). The
basic theory behind it is that for any stable stream, a balance exists between the water discharge
(Q), gradient (i), sediment load (S), and the bed load material size (D). This relation takes the form,
[S . D : : Q . i ] where “::” means “is proportional to”. This relation could also be derived analytically
through the application of the four equations describing the time and space dependant behaviour
of a river. The four equations are the conservation of water mass, conservation of water momentum,
conservation of sediment mass, and the equation of sediment motion, Jansen et al. (1979). Later,
the original Lane’s balance was modified by Klaassen (1995), who derived it analytically to include
more parameters. The expressions are then read:
i. For the slope (i):
n −3 2n n n
S ⋅ Dp ⋅ B 3
= m ⋅C 3 ⋅Q3 ⋅i3
ii. For the water depth (h):

S ⋅ D p ⋅ B n −1 = m ⋅ Q n ⋅ h − n
Where, the exponents (n) and (p) depend on the sediment transport predictor. The value of (n)
ranges from 2 to 5 for high values of bed shear stress; n = 5 for the England & Hansen (1967)
formula. For the same formula, (p = 1), and for the formula of Meyer-Peter & Muller (1948) (p)
takes the following form:
3 0.047
p=
2 θ ′ − 0.047
iii. To include the effect of yearly discharge variation:

n −3 2n n n n
V ⋅ Dp ⋅ B 3
= m⋅C 3
⋅ i 3 ⋅ N ∑ Pi ⋅ Qi3
i =1
Where:
V= ∫
1 year
S .dt
N is the number of seconds in a year

Pi is the probability of occurrence of the discharge Qi
Applying this concept for the simple case of a constant water discharge and sediment discharge
leads to the following expressions for the slope and depth variations of the constricted reach:
n −1
−
d 2  B2  n
= 
d1  B1 
19
n −3
i2  B2  n
= 
i1  B1 
3.2.3. Bed Degradation Caused By Series of Groynes

The channel bed degradation caused by a series of groynes could be treated as a long constriction
scour. However, due to the formation of separation flow zone around the tip of every groyne the
effect of groynes as a constriction to the channel is slightly different from that of a solid long
constriction. Michiue et al. as reported by Suzuki et al. (1987), introduced a multiplication factor
(λ) that modifies the constriction width caused by the construction of a series of groynes into its
equivalent constriction width by a solid constriction. The multiplication factor (λ) takes the
following values:
S
for →0 λ →1
L
S B0
for → ∞ λ →( )
L B1
Where (S) is the spacing between groynes, (L)

is the groyne length (B0) is channel width, and
(B1) is the constriction width (B1 = B0 – L).
Based on flume experiments for series of
groynes, Suzuki et al. (1987) showed that when
the ratio (S/L) is very small the groynes work as
a group. On the other hand, when (S/L) is very
large each groyne works independently. When
(S/L) is between 4 and 8, the channel bed
degradation is almost the same by that of a long
constriction i.e. (λ = 1). When (S/L) is less than
4, bed degradation is larger than that of a long
constriction i.e. (λ < 1), while it becomes
smaller when (S/L) is greater than 8 i.e. (λ > 1).
Figure 3-5 Definition of the coefficient (λ), source:
Wang & Yanapirut (1988), carried out similar Spannring (1999)
experiment to study the effect of the ratio (S/L),
he covered the range from (S/L) = 1.67 to (S/L)
= 5.0. He analytically derived the static equilibrium bed degradation formula (reached before by
d2 B 6
other authors) of { = ( 1 ) 7 }. Moreover, he extended his formulation through dimensional
d1 B2
analysis and dimensionless plots to include the ratio (S/L), his formula reads:
6 −1
d 2  B1   S 
7 7
=   ⋅ 
d1  B2   L 
Spannring (1999), applied Komura’s formula of equilibrium degradation depth to calculate the
coefficient (λ) introduced earlier by Suzuki. From known values for ∆zmax resulting from numerical
computations, the corresponding values for (λ) could be determined. The resulting groyne
coefficients are in a narrow range with a mean value of λ = 0.78 and a their standard deviation is
σλ = 0.02. Further, he commented that the cross section profile could be approximated by a 4th
degree parabola.
20
3.2.4. Local Scour Near Groynes

One of the characteristic features of groynes is the local scour that occurs in they vicinity. The
process of local scour around abutments, and groynes could be divided into several phases: initial
phase, development phase, stabilisation phase, and equilibrium phase, Hoffmans & Verheij (1997).
Two important characteristics are the equilibrium depth and the time evolution of scour.
Equilibrium scour depth

In order to estimate the maximum depth of scour, many formulae have been developed. Extensive
review and detailed comparisons between the different formulations of local scour around
abutments and spur- dikes is given by; among others Gill (1972), Darghi (1982), Noshi (1997),
Klingeman et al. (1984), and Hoffmans & Verheij (1997).
Through laboratory experiments, several researchers independently reached at similar expression
for the equilibrium scour depth near groynes. The expression takes the form, d (∞) = K ⋅ q 3 see
2
for example Gill (1972), Klingeman et al. (1984)and Hoffmans & Verheij (1997). For example, the
expression of Ahmed as given by Hoffmans & Verheij (1997) reads:
2
 q 3
ys ,e + h0 = K A ⋅ K ⋅ 
`

 1− m 
A
Where:
ys,e = equilibrium scour depth below initial depth
h0 = initial water depth
m = b/B, b and B are the width of the dike and channel respectively.
K`A = 2.14g- 1/3 (≅ 1.0 m-1/3. s2/3)
KA = 2Kp Ks Kα Kµ
Kp = correction factor for the influence of channel bend, (inner = 0.85, outer = 1.1~1.4)
Ks = for the shape of structure, (vertical wall = 1.0, 1:1 sloped =0.85)
Kα = for the angle of attack, (30o to 150o = 0.80 ~ 1.10)
Kµ = for the influence of porosity (0.2 porosity =1.0, 0.5 porosity =0.9~0.6)
The previous expression for local scour estimation was based on experimental work, and
dimensional analysis techniques. It lacked the theoretical background and missed some parameters
that proved to be of importance. Laursen, who contributed many researches about local scour
problem, emphasised the importance of the distinction between clear-water conditions, and live-bed
conditions in the estimation of scour depth. A factor that was considered in the above mentioned
treatments.
Gill (1972), through analytical approach extended the formulation of Straub. He distinguished
between clear water condition, where the bed shear stress is less than then the critical bed shear
stress (τ1<τc), and live bed condition where (τ1>τc). Moreover, Gill introduced the effect of the
sediment size (D), and the sediment transport capacity formula exponent (n) (of s = m.un). His form
reads:
  1  ( 6 7 )  τ (
3
7)
α    
1
clear-water
  1 − m   c τ
ys ,e + h0 = h0 ⋅ 
( 6 )−( 3 )
  1  7 7n
α  1 − m  live-bed, τ c >> τ 1
  
From the laboratory measurements, he found that,
21
0.25
D 
α = 8.375  50 
 h0 
Hoffmans & Verheij (1997) compared a large number of scour predictors with experimental data
and proposed the following formula:
 2

 1  h 
= h0   
3
y s ,e  − 1 + K B ⋅ b ⋅ tanh  0 
1− m   b
 
With (ys,e) as the scour below the original depth and (KB) correction factor ≅1.5 for groynes with
sloped face, and m = b/B, b and B are the width of the dike and channel respectively.
Local scour depth for a series of groynes

When a bank is protected with a series of groynes, the scour depth varies with the groyne location.
The local scour depth near a groyne that is far downstream from the first groyne is different from
that of a single groyne. However, the scour depth around the first groyne is similar to that of a
single groyne. Because of the influence of the neighbouring groynes, the scour depth usually
becomes smaller than that of the single groyne. Suzuki et al. (1987), showed through laboratory
experiments, that the local scour depth around a groyne located far downstream in a series of
groynes is a function of the groyne spacing (S) to length (L) ratio, and it could be expressed roughly
in the following form:
Z s , DS S S
= 0.07 ⋅ + 0.14 for 2 < < 10
Z s ,1 L L
Where:
= scour depth around any groyne far downstream.
Zs, DS
Zs, 1= scour depth around the first groyne which is similar to the scour depth near a
single groyne and could be estimated using any of the above mentioned formulae.
When (S/L)>12, i.e. the groynes are very far apart, the group action vanishes and the scour depth
near any groyne is nearly the same as that of a single groyne.
Przedwojski (1995) investigated the bed topography and the local scour at groynes in two bends
of the Warta River. He found that the local scour depth varies with groyne location, and the
maximum depth occurs at the groyne located downstream of the bend apex. The longitudinal
variation of scour depth at groynes located along the outer bank of a bend is quite similar to the bed
level changes due to the bend curvature. The scour depth changes depend significantly on the flow
and bed topography in a given bend. Moreover, based on the analysis of the field investigations he
estimated that the variation on the scour depth along a bend takes the following relation:
n −2
h Q x 
y s ,e = β ⋅ H  0 + g cos(2π ⋅ ) + sin(α − 90)
 H Q0 L 
Where:
- β = coefficient (β = 0.275 as given by Przedwojski (1995))
- α = angle between groyne axis and flow direction
- n = exponent of sediment transport formula (s = m.un)
- H = average reach flow depth
- h0 = unperturbed flow depth
- L = bend length
- x = distance from the bend entrance
- Q0 = discharge over the bottom width between groynes
22
∆ϕ
- Qg = part of the discharge blocked by the groyne ( Qg = h0 U S sin )
2
- S = spacing between two groynes
- U = unperturbed velocity at the toe of the groyne
- ∆φ = angle between two successive groynes
Extended review for local scour problem is given by Przedwojski et al. (1995) and Hoffmans &
Verheij (1997).
Time Scale for Scour Development

For abutment with a length that is more than the water depth (same as groynes), Hoffmans &
Verheij (1997) proposed the following formula, which is valid for all scour phases:
γ
t
a 
ys
= 1− e  ts 
y s ,e
In which:
ys,e = equilibrium scour depth below initial depth [m]
ys = maximum scour depth at any time (t) [m]
t1 = characteristic time at which ym = h0 [s]
γ = constant ≅ 0.40
 h0 
and, a = ln  1 − 
 h0 + y s ,e 
From dimensional analysis and the results of many experiments, he found that:
K ⋅ h02 ⋅ ∆1.7
ts =
(αU 0 − U c )4.3
In which:
K = coefficient (K = 330 hours m2.3.s-4.3, then t1 is expressed in hours)
∆ = relative density
α = coefficient depends on turbulence intensity
U0, Uc = mean, and critical velocities [m/s]
The α-factor is given in Hoffmans & Verheij (1997) to range from 2 to 9 according to the geometry
of the groyne.
Scour Geometry
As the geometry of scour near groynes is not of much importance to the research, reference is made
to some researchers who treated this point.
Rajaratnam (1983), through experimental work described the bed shear stress distribution in the
vicinity of a groyne, he noted an increase in the shear stress value up to 5.2 times the values in the
undisturbed locations.
Kuhnle et al. (1999) studied the effect of normal/overtopping flow conditions in the laboratory. He
deduced detailed topographic maps showing the geometry of the scour/deposition near groynes, and
predicted the area and volume of scour holes. He further commented that the larger the overtopping
ratios caused the region of maximum scour to shift towards the channel bank and caused a
secondary scour zone to form downstream of the groyne.
Peng et al. (1999) studied the scour and deposition around submerged groynes through numerical
simulation and compared his results with experimental results. He noted that the scour develops
rapidly at the initial stage, 78% of the equilibrium depth was reached within 25% of the equilibrium
23
time. Moreover, he presented the transverse and longitudinal profiles of the scour hole, and noted
that the eroded sand deposited just downstream the scour hole.
Klingeman et al. (1984) tested different orientations and found that the shape and area of scour is
strongly related to the orientation of the groyne to the flow.
24
Hydraulic disturbances caused by navigation
4. HYDRAULIC DISTURBANCES CAUSED BY NAVIGATION

In a river like the River Rhine, which is considered the backbone of NW European waterways
network, navigation is very important element affecting the morphology of both the groyne-fields
and the main channel. It is ought to be included when attempting to investigate the interaction
between the groyne-fields and the main channel.
4.1. NAVIGATION INDUCED WATER MOTION

Before discussing the navigation induced water motion near groynes, some attention will be given
to the water motion around a ship and the associated water level variations.
4.1.1. Water movement around a ship

As a vessel navigates through a waterway, it generates hydraulic disturbances in the form of waves
and currents. The dominant hydraulic disturbance features associated with a moving tow are the
drawdown, return current, propeller jets, and secondary waves. The drawdown and the return
current together form the primary water movement. The ship-induced waves form the secondary
water movement. The size of the vessel with respect to the waterway along with its speed dictates
the magnitude of these forces and their effects on the environment, Bhowmik et al. (1995),
Hochstein & Adams (1989).
Figure 4-1 Ship-induced water motion in a restricted waterway, source:Przedwojski et al. (1995)
As the vessel displaces water during its forward motion, it causes a drop in the water level alongside
the barges known as the drawdown, Figure 4-1A. Drawdown begins near the bow and rebounds
near the stern producing a single wave with a duration on the order of 40 to 120 sec, depending on
vessel length. Drawdown can cause dewatering of shallow areas along the shoreline during vessel
passage.
The maximum return current is produced adjacent to the barges and typically closer to the stern.
As vessels move upstream, return currents cause a temporary increase in ambient current velocities.
25
In a tow moving downstream, the return current causes a decrease in ambient current velocities and
under certain low flow conditions can create temporary ambient flow reversals. Currents associated
with the propeller jets are highly three-dimensional and cause localised disturbances to the flow.
The characteristics of these jets are a function of the hull shape, propeller type and size, and
horsepower of the vessel. The thrust, alignment to the bank, and the rudder angles affect the
potential flow impingement on the bed or banks. Under normal underway operations, propeller jet
effects are limited to the area behind the tow in the navigation lane.
Beginning at the corners of the lead barges, waves diverge from the sides of the tow. As transverse
stern waves intersect with this diverging wave, secondary waves are formed which propagate away
from the tow at an angle toward the shoreline, Figure 4-1B. These waves are rather consistent in
amplitude and have short periods (1 to 5 sec.). For high-speed vessels, these waves can have
significant wave heights and often dominate the hydraulic disturbances produced by the vessel.
Transverse waves diminish in magnitude with distance from the stern and have wave periods on
the order of 2 to 5 seconds. The influence of waves will diminish as the return current and
water-level depression are enlarged. In this way, the smaller and faster ships are usually responsible
for bank erosion due to the secondary waves, while the larger and slower ships cause erosion due
to the return current.
4.1.2. Return current and water level depression

A ship moving along a restricted waterway with a relative velocity (Vs) will induce a return current
(E) and a water level depression (B), see Figure 4-1. In accordance with Bernoulli theorem, the
higher the return current velocity, the lower the water level will get in the vicinity of the vessel.
Using the continuity principle and the Bernoulli equation, the average values of the return current
velocity (Ur) and the water level depression (∆h), can be expressed in Schijf’s form, more details
are given by Przedwojski et al. (1995) and Groeneveld (1997).
A 
U r = Vs  c − 1
 Aw 
V2  A 
∆h = s α ( c )2 − 1
2 g  Aw 
Where:
Vs
α = 1.4 − 0.4( )
Vlim
Ac = undisturbed waterway cross sectional area
Aw = waterway cross sectional area during a ship passage = (Ac – ∆h.B - Am)
Am = cross sectional area of water replaced by the vessel
B = undisturbed waterline width
Vs = vessel speed
Vlim = vessel limit speed = (0.5 ~ 0.75) ⋅ gh 
 
These equations can be solved by iterative procedures, or by the use of design curves, see
Groeneveld (1997). The maximum values of return current velocity (Ur,max) and the water level
depression (∆hmax), are dependant on the ship type. For the pushtow units, Verheij & van der Wal
(1984) found that:
U r ,max = C ⋅ U r
∆hmax = C ⋅ ∆h
Where:
26
B L2s
C = 1.2 + 5 ⋅10 −4 ⋅ Frh ⋅ ⋅
yt h Am
Vs
Frh : Froude number related to water depth =
gh
yt : distance between vessel axis and waterway bank
Ls : Length of the vessel
A review for the different formulations to compute the return flow due to navigation traffic is given
by Bhowmik et al. (1995) and Hochstein & Adams (1989). For the Waal River conditions the
model given by Hochstein & Adams (1989) seems to be most applicable, ten Brinke et al. (1999).
Later Ten Brinke used it to compare between the measured and computed current velocities in the
groyne-fields along the Waal River. Yet, he found that the actual measured maximum current
velocities are much higher than the computed near bank current velocities (even higher than the
computed average return flow in the river). Further, he concluded that it could not be applied to
navigation traffic in the Waal River. However, the model of Hochstein & Adams (1989) reads:
(
U r = Vs ⋅ [(a − 1) B + 1] − 1
0.5
)
Where:
2.5
 n  Ac
a=  with, n =
 n −1  Am
 V
1.8 s Vs
0.3 ⋅ e ≤ 0.65
Vcr
for
 Vcr
B= for the Waal River B =1.0
1 V
for 0.65 < s ≤ 1.0

 Vcr
Vcr is the so-called vessel critical velocity and takes the following form:
0.5
 g ⋅ Ac 
Vcr = K  
 Bc 
Bc = channel top width
K = constrainment factor = f(n, vessel dimensions), for the Waal River K ≅0.70
The shape of the lateral flow velocity distribution could be represented by the following relation:
y
−
U r ( y ) = k1 ⋅ e k2
Where:
k1 = U r (0) = α ⋅U r
ys
k2 =
α (1 − e−α ⋅ f (α ) )
 Bc
0.114 + 0.715
α = max  b
1.0
f (α ) = 0.42 + 0.5ln α
Ur(y) = return flow velocity at a distance (y) from the centreline of the vessel
ys = distance between the vessel centreline and the bank
27
b = vessel width
4.1.3. Ship waves

Secondary waves are induced by ships moving with a relatively high speed, they comprise
diverging and transverse waves. The result of interference of these waves is known as interference
peaks or ship waves. The direction of propagation of the peaks makes an angle of 35 degrees to the
bank line. (Verheij & Bogaerts, 1988) as reported by Przedwojski et al. (1995), showed that the
interference peaks are very similar to wind waves, despite their origin. The wave height of those
peaks at the bank could be taken as:
0.33
 h
H i = αi ⋅ h ⋅   ⋅ Frh4 For Vs < 0.8Vlim
 ys 
Where:
ys = distance between the side of the vessel and the bank
αi = coefficient depends on the shape of the vessel
αi = 1.0 for tugs, patrol boats, and loaded conventional inland motor vessels)
αi = 0.5 for empty European barges
αi = 0.35 for empty conventional motor vessels
The above expression is based on Kelvin’s theory for deep-water conditions. The same theory used
for calculation of the wave length gives:
2π 2
Lw = 0.67 ⋅ ⋅Vs (Valid only for Frh <0.7, and Hi <0.67h)
g
And the wave period,
2π Lw
T=
g
An expression that relates the maximum bottom wave orbital velocity to wave height, length, and
period is given by (van Rijn, 1993):
!= π ⋅ Hi
U
T ⋅ sinh(kh)
2π
k (wave number) =
Lw
h = the water depth
4.2. EFFECT OF NAVIGATION ON THE FLOW IN A GROYNE-FIELD

In the previous section, the basic components of navigation induced water motion was introduced
for an ideal case of a ship sailing near the centreline of a rectangular channel section. Yet, in a
natural river, the situation is rather complex, a number of factors; for side slope effect, eccentric
sailing course, and natural irregularities has to be introduced to convert the simplified case into the
more complex one. Further, if groynes exist, the estimation of navigation-induced water motion is
even more complex.
Some attempts were made to understand the navigation induced water motion, and its effect on the
sediment motion from the beaches of the groyne fields prior to the introduction of the six-barge
pushtow units to the River Waal. Field measurements during a trial year were carried out see,
28
Havinga et al. (1984), in addition to laboratory experiments WL|Delft_Hydraulics (1987) to

investigate the effects that those units would cause. Dimensions for the different pushtow
formations are given in Table 4-1. A computation methodology for the navigation-induced water
motion in a groyne-field, including the wave propagation through the groyne-field, is described in
details by Termes et al. (1991).
Table 4-1 dimensions for different pushtow formations
Formation 2x2 3x2 2x3

Length (m) 193.0 229.5 193
Breadth (m) 22.8 22.8 34.2
Draught (m) 2.7 2.7 2.7
Figure 4-2, shows three important stages during the passage of a pushtow sailing upstream. The
return current is a maximum immediately after the bow passes a groyne. The return current is
furnished by water from the upstream groyne field and the groyne field alongside. An eddy
develops at the groyne head and the small vortex at the downstream end of the groyne field
apparently disappears entirely. As the push-boat passes by the supply flow refills the groyne-field.
When the stern of the push-boat passes the particular groyne-field, the supply flow is forced to flow
out of the groyne field by the upstream groyne, perpendicular to the axis of the fairway. The natural
eddy immediately downstream of the groyne is transported downstream by the main current.
Figure 4-2 Flow pattern in a groyne-field during passage of a pushtow unit,

source: Brolsma (1988)
29
The magnitude of the velocities in the complex ship-induced water movement depends on, ship
dimensions, draught, unit speed, distance between pushtow and groyne, dimensions of the groyne
field and the river discharge.
In a physical model study performed by WL|Delft_Hydraulics (1987) , see also Brolsma (1988),
the navigation-induced water motion for three different formations, namely (3x2), (2x2), and (1x2)
were investigated, Figure 4-3. In addition, field measurements during the trial year 1984, to study
the effect of (2x2), (3x2), and (2x3) formations on the flow in groyne-fields were carried out by
Havinga et al. (1984). From both, the following remarks could be emphasised:
- The increase in return and supply flows can be expected to cause larger velocities mainly just
downstream of the groynes. Therefore, increase in ship speed, dimensions and draught will
increase the flow velocity near groynes.
Figure 4-3 Flow velocities during a pushtow passage, source: Brolsma (1988)
- The flow velocities in both (X), and (Y) directions, increase almost linearly with the increase
of the relative navigation speed for all the tested formations.
- The model results demonstrated the effect of the passing distance, where the velocities increase
with decreasing distance between the ship and the groyne.
- The experiments showed that the maximum current velocities in small groyne-fields are less
than in large groyne-fields, because of the smaller effect of the supply flow.
- From the field measurements, the current velocity in the groyne-field was recorded during the
passage of all three types. The (2x2) formation produced a maximum return current of 1.10 m/s,
2 to 3 times higher than that with no navigation. The wide formation (2x3) had a lesser effect
than the long one (3x2), however they both produced higher current velocity than the (2x2)
formation. The long formation yielded an increase in the current magnitude of 45% than the
(2x2) formation i.e. 1.6 m/s, and the wide formation increased the current only by 15% i.e.
1.25m/s.
- The water surface decline was also reported by Havinga et al. to be almost the same for the
three above-mentioned formations. Around 27cm near the tip of the groyne, and around 20cm
near the bank.
- The field experiments further indicated that whereas pushtows tended to produce a considerable
increase in the flow in the groyne field, the largest self-propelled ship (about 2000 tons) had
very little effect.
- When the river discharge increases, the river cross section also increases. This increase causes
a relative reduction of the return current and supply flow caused by navigation. Thus, a
reduction of the navigation induced flow velocities occurs. In addition, the relative draught
(blockage) decreases, which is also a factor in favour of reducing the navigation effect.
30
From the field measurements by ten Brinke et al. (1999), the effect of a push-two combination
passing a groyne-field was also observed. It induced a water level depression of 15~20 cm, in
addition to a drawdown current of 30~40 cm/s (less than what was reported by Havinga et al. there
is no clear reason for that). This current was strong enough to resuspend the sediment of the
groyne-fields beach. An example of the effect of navigation on current velocities, water level
fluctuations, and suspended sediment concentration is shown in Figure 4-4. The effect of a pushtow
unit passage on the sediment concentration, water level depression, and the return current velocity
could be observed during the time 11:5 to 11:20. These results are in analogy with the results
obtained from the field measurements of Havinga et al. (1984).
Figure 4-4 Example of the field records showing the navigation effect on current velocities, water level,
and sediment concentration, source: ten Brinke et al. (1999).
31
Morphological interaction between the groyne-fields and the main channel
5. MORPHOLOGICAL INTERACTION BETWEEN THE

GROYNE-FIELDS AND THE MAIN CHANNEL
5.1. GENERAL
The morphological interaction between the groyne-fields and the main channel, is a typical Dutch
problem. In most rivers, this interaction is insignificant and not perceived as a problem. After the
construction of the groynes, the river experiences large-scale deepening. Meanwhile, sand deposits
between the groynes form a stable beach. However, in the Waal River the shipping density is
among the highest of all the inland waterways of the world. Ships passing through the Waal some
times have a length that is more than the distance between two groynes, and they often sail
relatively close to the groynes. This results in pronounced sediment transport from the groyne fields
to the main channel.
The morphological interaction between the groyne-fields and the main channel is a combination
between the effect of navigation and the effect of river flow. At different discharges, the relative
effect of the various forcing parameters is different. Erosion of sand is thought to take place due
to currents and waves induced by navigation traffic. Deposition of sediment probably takes place
mainly at times of high discharge, when the groynes are completely submerged. During these
events, sediment is transported from the main channel to the groyne-fields and further landward to
the floodplain, (ten Brinke et al. 1999). As long as erosion and deposition are in equilibrium on a
time-scale of a couple of years, the beaches between the groynes are in dynamic equilibrium. This
has been the case over the last several decades in the Dutch part of the Rhine River. However, this
situation may change if the balance between the hydrodynamic forces changes. This could happen
if, for example, the navigation intensity changes or the discharge distribution between the branches
of Rhine changes, ‘which is unlikely to happen’. Changing the geometry of the existing groynes
by lowering, shortening, extending, etc. is also a change that could affect the morphodynamics of
the river.
Studies to understand the interaction between the groyne-fields and the main channel are scarce in
the literature. Moreover, the focus is on understanding the effect of navigation on the morphology
of the groyne-fields only during low water conditions. This is one side of the process, which causes
erosion of groyne-fields beaches, as the effect of navigation dominates. On the other hand, t the
author’s knowledge there are no studies ‘up to my current knowledge’, that discuss the process
during high water conditions. Thus, the question of how much sediment is restored to the beaches
of the groyne-fields during a flood could not be answered yet.
The effect of navigation on the morphology of the groyne-field was investigated through field
measurements, (Havinga et al., 1984), (de Haas & van Essen, 1987a; 1987b), and (ten Brinke et al.,
1999) and through model tests (WL|Delft_Hydraulics 1987). A conclusion that they all agreed on,
is the complexity and difficulty of describing this phenomenon. In addition, there is a large
uncertainty in the estimated sediment transport during a ship passage.
5.2. ESTIMATION OF EROSION FROM GROYNE-FIELDS

In the following two sections, we will present the results of some field campaigns, as well as some
attempts to analytically describe the interaction between the main channel and the groyne-fields.
On the first section, we will give a summery of the results that were reached through field
measurement campaigns. In the second section, we will present some analytical expressions
describing the interaction between the groyne-fields and the main channel. Field data analysis is
a very important tool to study a phenomenon, and to have an overview of system behaviour under
the conditions that are prevailing during the measurements. Thought in many cases interpolating
these results could be done successfully, extrapolating them could yield significant errors. Often
32
a better understanding to the physics behind the phenomena is reached through analytical models.
In general, analytical models are rather simplified; yet, they give a comprehensive grasp of the
physics behind the phenomena. Combined with field data and physical model results they yield a
powerful tool of analysis and prediction.
5.2.1. Field studies

One of the earliest investigations for the morphological changes in groyne-fields is due to Bruin
(1977) – after Havinga et al. (1984). Who mentioned that the groyne-fields erosion for the reach
Hulhuizen-Zaltbommel of the River Waal is about 3 ~ 4*106 m3 during the period 1960 through
1976 (i.e. 17 years). This estimate was based on photos at low water conditions. This sediment
volume is equivalent to an average lateral supply of sediment to the main channel of about 7.5*10-3
m3/s.
During the field measurements carried out by Havinga et al. (1984), it was noticed that only
pushtow units cause an increase in the flow velocity that could bring the sediment inside the
groyne-fields into motion. To have an idea about the sediment transport volume from a single
groyne-field due to a ship passage, we will use the information about Hulhuizen-Zaltbommel reach
given in Table 5-1. Utilising that erosion is fully due to pushtow units, we can then deduce that a
single groyne-field loses about 0.14 m3 of sand per passage. In that specific reach, the net changes
in the river bed level between the normal lines during the period 1960 through 1976 is known to
be negligibly small. During that period, the bottom level in the reach Hulhuizen–Nijmegen (15 Km)
dropped by around 0.20 m, and in the reach Nijmegen–Zaltbommel (57 Km) a bottom rise of ca.
0.06 m occurred. Therefore, the erosion value estimated by Bruin (1977) appears to be
overestimated.
Table 5-1 information about Hulhuizen-Zaltbommel reach

Distance Hulhuizen-Zaltbommel 65 Km
Number of groyne-fields 560
Frequency of pushtow units 8/day
Groyne-fields erosion (1960-76) 4.106 m3/17 years
Havinga et al. (1984) measured the sediment concentration and the flow velocity during a ship
passage; the study area is shown in Figure 5-1. They were aiming to quantify the effect of
increasing the capacity by introducing six-barge pushtow units to the navigation fleet. Based on
these measurements, they found that for pushtow units, the sediment transport duration is about 60
seconds, and the sediment outflow is primarily through the upstream part of the groyne-field. They
estimated this length to be around 20 m, in addition to some outflow from the downstream part
directly after the passage of the unit. Further, they estimated the sediment flux from groyne-fields
to the main channel per passage for the different pushtow formations. Summery of the results from
Havinga et al. (1984) is presented in Table 5-2. These values would lead to an estimate of the
erosion volume from the groyne-fields nearly 40%, of that estimated by Bruin (1977).
Table 5-2 Sediment transport from a single groyne-field due

to navigation (the effect of different formations)
Formation Max. transport Sediment out flux
10-6 m2/s m3/passage
2x2 9.8 0.035
3x2 19.6 0.070
2x3 12.0 0.040
33
Additional conclusion from Havinga et al. (1984) is the effect of the formation geometry on the
amount of sediment transport during its passage. With reference to section 4.2, a long formation
causes much stronger return current that lasts for a longer period than a shorter formation.
Consequently, its associated sediment flux from a groyne-field is much larger than that of the
shorter one.
Moreover, they reported that the bottom material is generally at rest in the absence of navigation
where the velocity does not exceed 0.2~0.3 m/s as a threshold value. In addition, groyne-fields are
supplied with sediment in a much less quantity from the main channel during the periods when
there is no ship passages. This supply takes place across the downstream part, primarily due to the
primary eddy.
De Haas & van Essen (1987a) studied the groyne-fields in a straight reach of the Waal River near
Druten, Figure 5-1. His results conformed to the same conclusion that, there is a sediment supply
from the river in the absence of navigation. In this case, the measurements were carried out during
a relatively high discharge condition (2000 to 3000 m3/s). Consequently, the estimated recharge of
sediment was relatively high, i.e. in the order of 3.6*10-5 m3/s for each groyne. Still the effect of
navigation was higher, but no quantitative results could be obtained.
In addition to the study of a straight reach near Druten, de Haas & van Essen (1987b) studied the
groyne-fields in a curved reach near St. Andries, where the groynes are located at the outer bend,
Figure 5-2. In this case, no sediment recharge from the river, on the other hand, erosion was
observed. This is clear as the groynes are located already in the deep part of the cross section and
the helical flow structure in a bend is in favour of eroding the groyne-fields that are located at the
outer bend.
Figure 5-1 River Waal near Druten Figure 5-2 River Waal near St. Andries
The most recent field investigation is due to Ten Brinke (1999). He carried out field measurements
in the area shown in Figure 5-1. A sample result of his measurements is shown in Figure 4-4, in
which it is clear that the turbidity peaks coincides with the passage of a pushtow unit. He estimated
an average sand transport from a single groyne-field to be 4.5 kg/s (0.003m3/s), for 15% of the time.
This value is equivalent to 0.18 m3/passage (with passage time of 60 seconds), compared to the
values in Table 5-2, i.e. 2.5 times higher than the maximum value estimated Havinga et al. (1984).
Further, they used this value to estimate a sediment budget from the groyne-fields to the main
channel. For roughly 500 sandy groyne-fields along the banks of the Waal, this would result in a
total input volume of 7*106 m3/year. Compared with a yearly sand transport of 0.5*106 m3/year, this
is clearly far too much. The reason for this exaggerated estimation could be one or more of the
following reasons:
- The study was carried out during a relatively low discharge, around 1000 m3/s compared with
1467 m3/s as a mean value. This means that the effect of navigation is relatively high, as the
effect of navigation increases with the decrease of the water level.
34
- The study area is in a straight reach of the river and cannot be generalised for the entire river.
As the groyne-fields behaves differently according to their location along the river. A
groyne-field in a bend behaves significantly different from that in a straight reach. The location
in the bend, i.e. inner curve or outer curve has a great effect.
- For the estimation of the total sediment volume, the total year was used as a base time, and the
effect of the discharge stage was not considered. During relatively high discharges, the effect
of navigation could be neglected. This would lead to decreasing the time at which there is
erosion of the groyne-fields.
5.2.2. Analytical approach

The first analytical representation to the problem is due to Havinga et al. (1984), primarily to
investigate the impact of using six-barge pushtow units. It is based on a comparative approach
between the impact of the navigation induced water motion for the four-barge units, and that of the
six-barge units. They considered that the current situation is a result of the effect of the four-barge
units. Comparing, the maximum return current velocity because of the two formations, and taking
the existing bed level inside the groyne-fields as a reference level. They estimated that the use of
six-barge units would lead to bed lowering for the groyne-fields beaches in the order of one to two
meters. No comment will be given on this model as it is based on very crude assumption.
Based on the concept of mass conservation, WL|Delft_Hydraulics (1987) proposed an analytical
model to the problem. The model is divided into two different parts. The first part describes the
sedimentation of groyne-fields by normal flow. The second part treats the problem of groyne-field
erosion. Both parts are dealing with the non-submerged conditions.
Sedimentation of groyne-fields
In the first part of the model, the navigation effect is not present and the bed is assumed to be
aggrading due to an inflow sediment flux. The sediment flux is assumed to take place through
the downstream part of the groyne-field. Integrating the sediment concentration (s/q) over the
inflow length and depth, the total inflow sediment volume could be estimated. Due to the eddies
that are dominating the flow inside the groyne-field, sediment is assumed to be distributed
uniformly throughout the whole groyne-field area (Lgf x Bgf), the temporal bed variations
( Zgf/ t) could then be estimated. The assumptions behind the model are based on the results
of a physical model investigation. With reference to Figure 5-3, they are as follows:
Figure 5-3 Definition sketch for the model of WL|Delft Hydraulics (1987)
- Active inflow length is only 20% of the total groyne-field length, the last 20% of the
groyne-field length (Lgf).
35
- The inflow velocity (u) is 35% of the stream velocity (ur), i.e. u = 0.35*ur.
Integrating the inflow velocity and the sediment concentration (c) over the length and depth of
the groyne-field;
∂Z gf
Lgf a
1
Bgf ⋅ Lgf ⋅
∂t
=
(1 − ε ) ⋅ ρ s
⋅ ∫ ∫ u ( x, z ) ⋅ c( x, z ) ⋅ dz ⋅ dx
0 0
with inflow concentration (ci);
a − z gf s
ci = ρ s ⋅ (1 − ε ) ⋅ ⋅( )
a q
and, inflow discharge;
∆Q = 0.2 ⋅ Lgf ⋅ 0.35 ⋅ ur ⋅ (a − z gf ) ⋅
"
#$# % " #$# % " #$# %
active length inflow velocity water height
introducing the constant (A1);

0.07 ⋅ ur s
A1 = ⋅( )
Bgf ⋅ a q
reducing the double integration to;
∂Z gf 1
Bgf ⋅ Lgf ⋅ = ⋅ ∆Q ⋅ ci
∂t (1 − ε ) ⋅ ρ s
Consequently, the groyne-fields bed level temporal variations read:
∂Z gf
= A1 ⋅ (a − Z gf )2
∂t
This equation has the following solution, with a graphical representation in Figure 5-4.
1
Z gf = a −
1
A1 ⋅ t + ( )
a − z0
Groyne-field Bed Evolution

1
0.9
0.8
0.7
0.6
Zgf/a
0.5
0.4
0.3
0.2
0.1
0
0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25
Time (years)
Figure 5-4 Groyne-field bed aggradation in the absence of navigation
36
Note that for time (t ), the solution of this equation is (Zgf = a) i.e. the groyne-fields will
be completely filled with sediment. This is not true, but it is a direct result for the assumptions
underlying this model.
Erosion of groyne-fields because of Navigation

The second part of the model of WL|Delft_Hydraulics (1987), treats the problem of groyne-field
erosion. In which the effect of navigation as a forcing parameter that creates an additional
outflow velocity which transports sediment out from the groyne-field. The outflow velocity (v*),
is maximum (vmax) near the tip of upstream groyne (sailing wise), and reduces exponentially with
the distance towards the downstream groyne.
In the same manner like the first part, integrating the sediment transported throughout the whole
groyne-field length, and over the passage time of a vessel. Utilising England&Hanesen sediment
transport predictor formula, the total erosion volume (E) from a groyne-field during a ship
passage time (T) could be written in the following form;
Ts Lgf
E=∫ ∫ mv
δ ⋅( x − Lgf )
n
* ⋅e ⋅ dx ⋅ dt
0 0
by integration;
Ts ⋅ mv*n −δ ⋅ L
E= (1 − e gf )
δ
−δ ⋅L
Assuming that the term (e gf ) is negligible, the temporal bed variation due navigation
induced erosion will take the following form:
dZ gf f s ⋅ Ts
=− ⋅ (m ⋅ vmax
n
)
dt Bgf ⋅ Lgf ⋅ δ
Where:
fs : frequency of navigation;
Ts : vessel passage time;
δ : reduction parameter;
and by integration;
f s ⋅ Ts
Z gf = z0 − ⋅ (m ⋅ v*n ) ⋅ t
Bgf ⋅ Lgf ⋅ δ
This is a linear relation with time, and it gives rise to unacceptable results for large values of
time. This is a result to the absence of any slowdown mechanism. For example, if the outflow
velocity is a function of the groyne-field bed level, and the sediment transport formula has a
threshold value, eventually the bed degradation will converge to an equilibrium value. In the last
step, i.e. the integration of the DE, all parameters were considered constant with time. However,
if the effect of the discharge stage on the navigation-related parameters is implicitly introduced,
a better definition might be reached.
Equilibrium bed level

For reach a definition for the equilibrium bed level we will equate the erosion rate with the
deposition rate.
 ∂z gf   ∂z 
  +  gf  =0
 ∂t erosion  ∂t deposition
simplifying and introducing a constant (A2);
37
m ⋅ a2
A2 =
0.07 ⋅ Lgf ⋅ s ⋅ δ
We reach a definition for the equilibrium bed level (Zeq) that reads:
Z eq = a − A2 ⋅ f s ⋅ Ts ⋅ V*n
Furthermore, by combining the two definitions reached for the temporal variations of both
erosion and deposition, we reach at a definition that describes the long-term variations of the
groyne-fields bed level, see Figure 5-5. The definition reads:
A3 1 + A3 ( a − Z gf )  A3 1 + A3 (a − z0 ) 
⋅ ln  =t+ ⋅ ln   for A3 ⋅ (a − Z gf ) < 1
A1 1 − A3 (a − Z gf )  A1 1 − A3 (a − z0 ) 
A3 1 + A3 ( a − Z gf )  A3 1 + A3 (a − z0 ) 
⋅ ln  =t+ ⋅ ln   for A3 ⋅ (a − Z gf ) > 1
A1  A3 (a − Z gf ) − 1  A1  A3 (a − z0 ) − 1 
With the constant (A3);
A1 ⋅ Bgf ⋅ Lgf ⋅ δ
A3 =
f s ⋅ Ts ⋅ (m ⋅ vmax
n
)
Groyne-field Bed Level development
1
0.8
0.6
Zgf/a
0.4
0.2
0
0 5 10 15 20 25
Time (years)
Figure 5-5 Long-term variation in groyne-fields bed level
This representation of the combined effect of flow and navigation induced sediment motion
inherits the deficiencies of both parts. The groyne-field bed level converges to the equilibrium
bed level (zeq) which is directly defined by equating both of them.
Finally, the model of (WL|Delft_Hydraulics 1987) has some deficiencies as we mentioned before
and it might lead to unrealistic results. In addition, it is very sensitive to some parameters such as
the reduction parameter (δ ), navigation frequency (fs). However, the basic idea and the assumptions
behind it are valuable and it forms a base for further development to remove the deficiencies.
The model of Sieben & Douben

Another model is proposed by Sieben & Douben (2000). A complete derivation is given in
Appendix II. They estimated the equilibrium groyne-field bed level to take the form,
∆z
= A + B ⋅ ln θ
a
where, A and B are constants, and θ is the Shields parameter.
38
Furthermore, they attempted to estimate the effect of the interaction between the groyne-fields and
the main channel on the morphology of the main channel he arrived at the following expressions:
 c k Lg ∆z  ∂z u dqs ∂zmc  Lg 1  ∆z   ∂z gf
1 − k (1 − ) 2  ⋅ mc + ⋅ =  c − c k (1 − ) − 1 + ε   ⋅
 (1 − ε ) Bmc a  ∂t a du ∂x  Bmc (1 − ε )  a   ∂t
 c k Lg ∆z 2  ∂zmc u dqs ∂ 2 zmc  Lg 1  ∆z   ∂z gf

1 − k (1 − ) ⋅ + ⋅ =  c − c k (1 − ) − 1 + ε   ⋅
 (1 − ε ) Bmc a  ∂t 3 i du ∂x  Bmc (1 − ε )    ∂t
2
a
The first for the small-scale (simple wave model), and the second for the large-scale effect
(parabolic model). Moreover, he estimated the morphological time-scale for the groyne-field bed
variations to take the following definition:
a (1 − ε )
Tm = ⋅
k ⋅ w` χ ⋅ ( qs )
q
where, Tm [s] is the morphological time-scale. Sieben & Douben (2000) estimated that for the
groyne-fields bed variations under the Waal River conditions is about 50 days.
5.3. EFFECT OF CHANGING THE EXISTING GROYNES GEOMETRY

The possible effects of lowering the groynes, lengthening or shortening, and decreasing their
spacing were investigated by Verheij (1997). He concluded that groyne lowering has the greatest
influence at water levels between the old and the new crest heights. It would also lead to some
aggradation in the main channel. In an outer bend, the groyne lowering could result in a small
reduction of the width of the fairway and in some depth reduction for part of the channel section.
Compared to the present situation, lowering of the groynes has a negative effect on the siltation
after a high water.
5.3.1. Lowering the groynes

The effect of a lowered groyne height is negligible as long as the water level remains below the
lowered groyne height (average 50 to 70% of the time depending on the level of the crest). As soon
as the water level is higher than the lowered level the groynes will be over flooded and a part of the
total discharge will pass over the groynes through the groyne fields. This has both positive and
negative impacts.
As positive effect, the flow capacity will be better and there will be a milder velocity gradient
through the mixing zone between the main channel and the groyne-fields. It is as well, expected that
the contribution of the discharge passing over the groynes and through the groyne field, will
increase the conveyance capacity of the river. Further, the local scour holes forming near the tip of
the groynes will be less deep due to the reduced specific discharge and the reduced maximum water
depth at which the groyne is not yet submerged. Consequently, the sedimentation downstream of
the scour hole forming the so called “groyne flames” will be reduced.
As a negative effect, the discharge in the main channel will drop compared to the initial situation
(1D-effect), and the effective stream width will be larger. This results in a tendency to general
aggradation in the main channel, as the flow velocity will be lower. In an outer bend, during times
that the water level is higher than lowered groyne level, more water will flow through the groyne
fields and for a longer period, compared with the not-lowered groynes. This might lead to an
increased sediment flux towards the floodplain.
39
Additionally, as the morphological interaction between the groyne-fields and the main channel due
to navigation is not fully understood, the impact of lowering the groynes ‘from this point of view’
cannot be predicted. In the light of the above section (5.2), the groyne-fields bed generally erodes
due to navigation during low water conditions, and during high water conditions deposition occurs.
The transition of the behaviour from bed erosion to deposition and vice versa, is strongly related
to the groynes crest level. Lowering the crest level would affect the existing balance between the
two forcing parameters. The result is not really known.
5.3.2. Lengthening or shortening of groynes

Extending the groyne length results in general sense in larger water depths in the main channel
compared to the original groyne length. Lengthening can limit the width of the fairway depending
on the local situation. Shortening leads to opposite effects compared to lengthening. Lengthening
a series of groynes over a long reach is ‘to some extent’ similar to long constriction, i.e. the river
bottom will drop over that reach. This has a negative morphological impact on the upstream part;
further, the water surface slope will change. However, lengthening or shortening of one or few
groynes is a local operation and has limited impacts in both upstream and downstream of it.
Changing the dimensions of groynes will indeed change the geometry of the groyne-fields.
Consequently, the large turbulent structures that are governed by groyne-fields length to width ratio
will also change. However, with a small change in the groyne length (in the order of 10 to 20m) this
effect is negligible. In general, groynes lengthening or shortening have positive and negative
consequences. Yet, for navigation purposes groyne shortening appears to be not desirable, groyne
lengthening may be considered favourable.
5.3.3. Decreasing the groynes spacing

With reference to section 3.1.2, the large horizontal eddy structure is determined by the aspect ratio
of the groyne-field’s dimensions. Reducing the groynes spacing will permit a single ‘more stable’
eddy that is confining the streamlines to a relatively smaller width. The morphological impact on
the main channel is minor, yet it would cause some degradation. Another effect that is not clear at
the moment, is the effect on the navigation induced water motion and its accompanying sediment
transport. In the one hand, we would expect a more protection to the groyne-fields of less spaced
groynes. On the other hand, there might be a stronger water level depression, and more confined
outflow, that would then transport more sediment out of the groyne-fields.
The distance between the groynes should be specified mainly by the required stability of the
separation flow line with various discharges. A considerable reducing of the groyne spacing leads
to an increase of the water depth in the main channel by a few percent. The length of the scouring
hole will specify the minimum spacing between the groynes.
40
References
REFERENCES
Alvarez, J. A. M. (1989). “Design of groins and spur dikes.” Proceedings 1989 National
Conference On Hydraulic Engineering, New Orleans, 296-301.
Aya, S., Fujita, I., and Miyawaki, N. (1997). “2-d models for flows in river with submerged groins.”
27th IAHR Congress, San Francisco, CA. USA, 829-837.
Bhowmik, N. G. (1989). “Phsical impacts of human alterations within river basins - the case of the
kanakakee, mississippi and illinois rivers.” XXII IAHR Congress, Ottawa, Canada, B139-B146.
Bhowmik, N. G., Xia, R., Mazumder, S., and Soong, T. W. (1995). “Return flow in rivers due to
navigation traffic.” Journal of Hydraulic Engineering, ASCE, 121(12), pp. 914-918.
ten Brinke, W. B. M., Kruyt, N. M., Kroon, A., and van den Berg, J. H. (1999). “Erosion of
sediments between groynes in the river waal as a result of navigation traffic.” Spec. Publs int.
Ass. Sediment, 28, pp. 147-160.
Brolsma, J. U. (1988). Six-barge pushtow trials, PIANC, Den Haag.
Chen, F. Y., and Ikeda, S. (1997). “Horizontal separation in shallow open channels with spur
dikes.” Journal of Hydroscience and Hydraulic Engineering, 15(2), pp. 15-30.
Darghi, B. (1982). “Local scour around bridge piers - a review of theory & practice.” TRITA-VBI-
114, Royal institute of technology, Stockholm, Sweden.
Gill, M. A. (1972). “Erosion of sand beds around spur dikes.” Journal of Hydraulic Engineering,
ASCE, 98(HY9), pp. 1587-1602.
Gill, M. A. (1981). “Bed erosion in rectangular long contraction.” Journal of Hydraulic Division,
ASCE, 107(HY3), pp. 273-284.
Groeneveld, R. (1997). “Inland waterways - ports, waterways and inland navigation.” CT4330,
Delft University of Technology, Delft.
de Haas, A. D., and van Essen, J. A. F. (1987a). “Onderzoek naar de invloed van duwvaart op de
water- en zandbeweging in het kribvak durten.” DBW/RIZA 87.007, RIZA, Arnhem.
de Haas, A. D., and van Essen, J. A. F. (1987b). “Onderzoek naar de invloed van duwvaart op de
water- en zandbeweging in het kribvak st. Andries.” DBW/RIZA 87.008, RIZA, Arnhem.
Havinga, H., Slootweg, H., and Zeekant, J. (1984). “Kribvakmeting t.B.V. Zesbaksduwvaart op de
waal bij druten.” 84.9, Rijkswaterstaat, district Zuidoost, Arnhem.
Hochstein, A. B., and Adams, E., Jr. (1989). “Influence of vessel movements on stability of
restricted channels.” Journal of Waterway, Port, Coastal, and Ocean Engineering, ASCE,
115(4), pp. 444-456.
Hoffmans, G. J. C. M., and Verheij, H. J. (1997). Scour manual, A.A. Balkema, Rotterdam.
Ishii, C., Asada, H., and Kishi, T. (1983). “Shape of separation region formed behind a groyne of
non-overflow type in rivers.” XX IAHR Congress, Moscow, USSR, 405-412.
Jansen, P. P., van Bendegom, J., van den Berg, J. H., de Vries, M., and Zanen, A. (1979). Principles
of river engineering - the non tidal alluvial rivers, Delftse Uitgevers Maatschappij, Delft.
Klaassen, G. J. (1995). “Lane’s balance revisited.” 6th International Symposium on River
Sedimentation, New Delhi, India.
Klingeman, P. C., Kehe, S. M., and Owusu, Y. A. (1984). “Streambank erosion protection and
channel scour manipulation using rockfill dikes and gabions.” WRRI-98, Oregon State
University, Water Resources Research Institute, Corvallis, Oregon.
Komura, S. (1966). “Equilibrium depth of scour in long constrictions.” Journal of Hydraulic
Engineering, ASCE, 92(HY5), pp. 17-37.
Krebs, M., Zanke, U., and Mewis, P. (1999). “Hydro-morphodynamic modelling of groin fields.”
28th IAHR congress, Graz, Austria.
Kuhnle, R. A., Alvonso, C. V., and Shields, F. D. J. (1999). “Geometry of scour holes associated
with 90o spur dikes.” Journal of Hydraulic Engineering, ASCE, 125(9).
41
References
Noshi, H. M. (1997). “Abutment scour in uniform and stratified bed,” PhD., Colorado State
University, Fort Collins, Colorado, USA.
Ouillon, S., and Dartus, D. (1997). “Three-dimensional computation of flow around groyne.”
Journal of Hydraulic Engineering, ASCE, 123(11), pp. 962-970.
Peng, J., Kawahara, Y., and Tamai, N. (1997). “Numerical analysis of three-dimensional turbulent
flows around submerged groynes.” 27th IAHR congress,, San Francisco, USA.
Peng, J., Tamai, N., Kawahara, Y., and Huang, G. W. (1999). “Numerical modelling of local scour
around spur dikes.” 28th IAHR congress, Graz, Austria.
Przedwojski, B. (1995). “Bed topography and local scour in rivers with banks protected by
groynes.” Journal of Hydraulic Research, 33(2), pp. 257-273.
Przedwojski, B., Blazejewski, R., and Pilarczyk, K. W. (1995). River training techniques-
fundamentals, design and application, A.A. Balkema, Rotterdam.
Rajaratnam, N. (1983). “Flow near groin-like structures.” Journal of Hydraulic Engineering, ASCE,
109(3), 463-480.
Richardson, E. V., Stevens, M. A., and Simons, D. B. (1975). “The design of spurs for river
training.” XVIth , IAHR congress, Sao Paulo, Brazil, 382-388.
Schans, H. (1998). “Representativitet van kribvakmetingen uit 1996 en 1997 ten opzichte van de
hele waal.” ICG 98/15, Universieit Utrecht, Fysische Geografie, Utrecht.
Sieben, J., and Douben, N. (2000). “Globale inschatting van morfologische effecten in het
zomerbed door herinrichting van uiterwaarden.” 2000.027X, RIZA, Arnhem.
Spannring, M. (1999). “Degradation of the river bed after building of groynes.” 28th IAHR congress,
Graz, Austria.
Suzuki, K., Michiue, M., and Hinokidani, O. (1987). “Local bed form around a series of spur dikes
in alluvial channels.” XXII IAHR Congress, Lausanne, Belgium, 316-321.
Termes, A. P. P., van der Wal, M., and Verheij, H. J. (1991). “Waterbeweging door scheepvaart
op rivieren en in kribvakken.” Q1046, WL|Delft Hydraulics, Delft.
Tingsanchali, T., and Maheswarn, S. (1990). “2-d depth averaged flow computation near groyne.”
Journal of Hydraulic Engineering, ASCE, 116(1), pp. 71-85.
Tominaga, A., Ijima, K., and Nakano, Y. (2001). “Flow structures around submerged spur dikes
with verious relative height.” 29th IAHR CONGRESS, Beijing, China, 421427.
Uijttewaal, W. S. J. (1999). “Groyne field velocity patterns determined with particle tracking
velocimetry.” 28th IAHR congress, Graz, Austria.
Uijttewaal, W. S. J., Lehmann, D., and van Mazijk, A. (2001). “Exchange process between a river
and its groyne fields - model experiments.” Journal of hydraulic Engineering, ASCE, 127(11),
pp. 928-936.
van Urk, G., and Smit, H. (1989). “The lower rhine geomorphological changes.” Historical changes
of large alluvial rivers: Western europe, G. E. Pettes, ed., John Wiley & Sons. Ltd, 167-182.
Verheij, H. J. (1997). “Effectiviteit van kribben.” Q2360, WL|Delft Hydraulics, Delft.
Verheij, H. J., and van der Wal, M. (1984). “Prototype tests with respect to the stability of bank
protection with various types of ships in the hartel canal in 1983 (text in dutch).” M 1115, WL
| Delft Hydraulics, Delft.
Visser, P. J. (2000). “Bodemontwikkeling rijnsysteem.” Delt University of Technology, Hydraulic
and Offshore Engineering Section, Delft.
de Vriend, H. J. (1999). “Long-term morphodynamics of alluvial rivers and coasts.” Environmental
applications of mechanics and computer science, cism courses and lectures, G. Bianchi, ed.,
Springer Wien, New York, 1-19.
de Vries, M. (1996). “River engineering.” f10, Delft University of Technology, Delft.
Wang, T. W., and Yanapirut, N. (1988). “Channel bed degradation caused by constriction.” 6th
Congress, Asian and Pacific Regional Division, IAHR, Kyoto, Japan, 285-292.
42
References
WL|Delft_Hydraulics. (1987). “Kribvakerosie door zes- en vierbaksduwvaart op de waal.” Q 93/Q

576, WL | Delft Hydraulics, Delft.
43
Conceptual model
APPENDIX I – CONCEPTUAL MODEL

Description
Groynes are constructed to provide a fixed river planform and a navigation channel that is relatively
deep over a large part of its cross section. The success of the groynes in fulfilling they function
depends on the balance of the hydrodynamic forces acting on the sandy deposits in the groyne-
fields. This balance could be defined by the balance between erosion of sand that occurs due to
currents and waves induced by navigation traffic, and deposition that probably takes place during
times of high river discharge. As long as the erosion and deposition are in equilibrium on a time-
scale of a couple of years, the groyne-fields are in a dynamic equilibrium, (ten Brinke et al., 1999)
To have an insight into the interaction between the morphology of the main channel of a river and
that of the groyne-fields, we will assume that erosion of sediment from the groyne-fields occurs
primarily due to navigation induced water motion, during times of normal to low river discharge.
Whereas, the effect of navigation vanishes during times of high discharges, (WL|Delft_Hydraulics,
1987). Moreover, deposition of sand will take-place during times of high discharges with a rate that
is a function of the discharge.
Figure I-1 is a schematic illustration for the co m b in e d ef fe c t o f n a v ig a ti on a n d d is c ha r g e
basic assumption. Erosion takes place during a 1 0
ship passage; deposition takes place at the

8
Q3
times where there is no shipping with a much 4
lower rate. The combined effect could be

G F be d level (m m )
simplified by a straight line. The three line 0
- 2
drawn for three different discharges Q1<Q2<Q3. Q2

ship passage
- 4
With the current state of knowledge, a more

ship pas sage
ship passage
- 6
Q1
detailed “complex” function for the combined - 8
effect of navigation and discharge is not

-1 0
0 500 1 000 1500 200 0 25 00
tim e (S )
possible; more effort will be spent to define this Figure I-1 Combined effect of navigation and
relation through the course of the research. discharge for three different discharges Q1<Q2<Q3
For every discharge, we can estimate the
sediment transport rate from the slope of line.
S = (Smax/Qtrans).Q - Smax
Then, the relation between the discharge Q
S (Groin Field)
[m3/s] and the sediment flux (S) [m2/s] between

the main channel and the groyne-fields could
be deduced. We will assume again that this Qtrans
relation takes the form of a straight-line (Figure Smax Q
I-2). S = a ⋅Q + b
Where:
Figure I-2 S-Q Relation
S: sediment flux from the GF to MC per
meter length of the GF-MC interfacial line [m2/s]
a: constant = Smax/Qtrans
b: constant = -Smax
Smax: maximum sediment flux from the GF to MC
Qtrans: the discharge at which the sediment flux changes direction
The values of (a) and (b) could be chosen based on imperial values for (Smax) and (Qtrans), that could
be obtained from field data and/or numerical computations. This relation is a simplified way of
introducing the effect of both navigation and high river-discharge. The effect of navigation is
assumed to be eroding the groyne-field at low to normal river-discharge, and will vanish during
high river-discharge. The navigation effect will determine the value of (Smax). During the time that
there is no navigation, some of the sediment that is transported to the main channel will be restored
Appendix I 44
Conceptual model
to the groyne-fields, with a rate that is determined by the discharge value. The effect of discharge
will determine the slope of the line. The compound effect of navigation and river discharge will
determine the value of the transition discharge (Qtrans).
Application
Utilising the dimensions of the Waal River i.e. mean channel width of 250m, and mean groyne-field
width of 60m, and applying the mass balance concept to a unit length of the river, with no supply
or losses to the system. With the aid of the river discharge data (figure I-3), the morphological
behaviour of the main channel and the groyne-field only due to their interaction could be obtained,
(Figure I-4).
The value used for Smax were estimated from the field measurements results, performed by ten
Brinke et al. (1999), who estimated a sediment flux from the groyne-fields to the main channel of
4.5 Kg/s for 15% of the time, with a dry density of 1500 Kg/m3. This would be equivalent to a
uniform sediment flux along the whole time of 2.25*10-6 m2/s. The transition discharge were
assumed to be 1450 m3/s and the sediment porosity 40%.
W aa l River H y dro graph M a in ch annel bed v aria tions

fro m 19 45 to 19 98
1.00 0.25
9000
3
Q t ra ns = 1 45 0 m /s
8000
0.60 0.20
Maxi mum v ariation on a year (m)

7000
6000 0.20 0.15

D Z (m )
5000
Q (m3/s)
4000 -0.20 0.10
3000
-0.60 0.05
2000
1000
-1.00 0.00
0 Jan -4 5 Ja n-50 Ja n-55 Ja n-6 0 Jan -6 5 Ja n-70 Ja n-75 Jan-8 0 Jan -8 5 Ja n-90 Ja n-95 Jan -0 0
Jan -4 5 Ja n-50 Jan -55 Ja n-60 Jan -65 Ja n-70 Ja n-75 Jan -8 0 Ja n-85 Jan -9 0 Ja n-95 Jan -0 0 D ate
D ate
D Zmc (m ) One ye ar variation
Figure I-3 Waal River discharge from 1945 Figure I-4 Bed variations for the period from
to 1998 1945 to 1998
Discussion
The results of this simple approach showed the dynamics of the interaction between the main
channel and the groyne-fields. Provided that the channel bed level was in equilibrium at the
beginning of the time series, The range of the channel bed variations was found to be in the order
of ±0.50m for the whole time series (50 years). The variation over a year was found to have a
maximum value of 0.20m, (Figure I-4).
By calculating the water level equivalent to W a te r Le vel Inc rea se
each river discharge, the effect of the bed 20%
variations on the water level could be

estimated. (Figure I-5) shows the percentage of 10%
water level increase/decrease due to the

H eight (m)
0%
interaction between the groyne-fields and the

main channel. It is obvious that the computed
water level could vary by ±17.5% just because
-10%
of the interaction between the groyne-fields and -20%
the main channel. Jan-45 Jan-5 0 Jan-55 Jan-6 0 Jan-65 Jan-70

Time (d ay)
Jan-7 5 Jan-80 Jan-85 Jan-9 0 Jan-95 Jan-0 0
The discharge history of the river defines the

starting point of the bed level computation for Figure I-5 Increase in water level due to the
the next time step. Again, as there is no interaction between the MC and the GF
equilibrium bed levels that could be reached
after a given period (morphological time scale),
Appendix I 45
Conceptual model
the effect of the discharge history is very significant. On the one hand, a flood will have a much
higher level if it is preceded by a long period of low discharges since the channel bed would have
had a long period of aggregation as a consequence of the low discharges. On the other hand, a flood
preceded by a flood will have a relatively lower water level as the channel would experience some
degradation as a result of the earlier flood.
Through the analysis, the effect of both Smax and Qtrans were analysed. Changing the value of Smax
changes the values of the bed variations with 1:1 proportion, but the shape and the transition points
remains the same. Yet, changing the value of Qtrans would change the whole pattern of the channel
bed variations leading to changes in the bed level with a much higher magnitude.
Conclusions
This conceptual model succeeded to show the dynamics of the interaction between the groyne-fields
and the main channel. However, it requires improvement.
The values of channel bed variations, (aggregation/degradation) might be over-estimated with this
approach. Moreover, the bed variations could reach ±∞, which is not possible in reality. This is due
to the fact that in this simple approach there is no slow-down mechanism that could control the
morphological process i.e. equilibrium bed level(s).
The effect of discharge history in this approach is very significant. The absence of an equilibrium
bed level or as consequence to that, the absence of a morphological time-scale effect allows the
effect of the discharge history to extend over a very long period (infinite in this model).
The choice of the Smax might be easy and straightforward as it only affects the values of the bed
variations. Nevertheless, the choice of a value for Qtrans might change the pattern of the bed
variations so it is very important to make an accurate choice for Qtrans.
Recommendations
It is very important to define two different equilibrium bed levels for the groyne-fields. The first
is the groyne-field’s minimum bed level, after which the navigation-induced erosion would vanish.
The second is the groyne-field’s maximum bed level, after which the sedimentation due to high
discharge would stop.
The morphological time-scale associated with the groyne-fields, main channel interaction is a factor
that plays an important role on the extent of the effect of the discharge history, i.e. the influence of
that interaction on the flood levels.
Appendix I 46
APPENDIX II – THE MODEL OF SIEBEN & DOUBEN

The aim of this simplified model is to have a theoretical understanding for the morphological
behaviour of the groyne-fields. A very important limitation for this model, that it is limited to the
low-water condition, i.e. the groynes are not submerged.
Mass Balance
With reference to the definition sketch the groyne-
field (GF) bed level is in balance with the vertical
sediment flux (F), and the sediment balance reads:
∂z gf
(1 − ε ) ⋅ Agf ⋅ +F = D−E (II.1)
∂t
Where:
Agf = Lgf x Lg [m2]
ε = porosity [--]
With the vertical sediment flux (F) as
F = w`⋅ Agf ⋅ (ce − c ) (II.2)
Where:
w‘ = exchange rate between GF bed material,
and water [m/s]
Ce = equilibrium concentration in the GF
C = actual concentration in the GF
Sediment concentration
Both the equilibrium, and the actual concentration is

not easy to define or estimate for the complex flow in Figure II-1 Definition sketch for a river with
a groyne-field, but for simplicity we may assume that: groynes, after Sieben & Douben (2000)
Ce = Cre ⋅θ gfb / 2
∆z
−k⋅ qs
C=e a
⋅χ⋅ (II.3)
q
Where:
C = actual concentration in the GF
θgf = effective GF dimensionless shear stress (Shields parameter) [-]
Cre = reference concentration in the main channel
b = empirical factor
K = damping factor (for the concentration over the depth)
χ = steering factor (relates the concentration in GF to the one the main channel)
q = specific discharge [m2/s]
qs = specific sediment load in the main channel [m2/s]
The values for both parameters (Cre), and (b) must be analysed from the prototype data. For the
estimation of the steering factor (χ) along the Waal River, (Lenders et al., 1998 – after Sieben &
Douben, 2000) showed that a base sediment concentration in the GF (flow induced) Cflow = 2*10-5,
and a peak concentration (navigation induced) Cnav = 2*10-4, i.e. Cnav = 10*Cflow.
Appendix II 47
Assuming that the average ship passing time is around 100s, and an average frequency of about one
ship every three minutes, i.e. the percentage of time that there is a ship passing (p) ≅ 0.50.
Consequently, the time of high concentration (Cnav) could be estimated to take place for around 50%
of the time, and a rough estimate for χ ≅ 6, is possible.
In addition to the navigation-induced disturbances, there is another effect for the relative depth
∆z
( ) of the groyne-field; (this effect will not be included in this model)
a
Effective shear stress in the groyne-field

Two different conditions should be distinguished for the estimation of effective shear stress in a
groyne-field. The first is the normal flow condition in the absence of any navigation effect, and the
second is the combined flow with navigation-induced water motion.
For the case of normal flow only, we may assume that the flow velocity in a groyne-field is
proportional to that in the main channel with a factor α ≅ 0.1 to 0.4, this factor could be estimated
from prototype measurements, laboratory experiments or at least from numerical simulations.
For the situation of navigation-induced water motion, based on the experiments that were done by
WL|Delft_Hydraulics (1987), the outflow from a groyne-field could be averaged along the ship
passage time, in the following way:
∆h ⋅ Lg ⋅ Lgf
Qout = (II.4)
Tpassage
Where:
Qout = exchange flow rate between the groyne-field and the main channel [m3/s]
∆h = water level depression caused by a ship passage [m]
Tpassage = average ship passage time [s]
Typical values for the water level depression ∆h ≅ 0.2~0.5m, and for the passage time Tpassage ≅
100s.
The outflow velocity uout could as well, be written in the following form,
∆h Lg
uout = ⋅ (II.5)
Tpassage a − ∆z
Assuming a linear relation between the dimensionless shear stress in the groyne-field (θgf) and the
main channel (θ) i.e.
2
umc  u 
θ gf = f ⋅θ ; θ= ; f = α 2  p ⋅ ( out ) 2 + 1 (II.6)
Ch ⋅ ∆ ⋅ D50  α ⋅ umc 
Where:
Ch = Chezy coefficient for the main channel [m½/s]
∆ = sediment specific weight [-]
f = proportionality factor [-]
p = percentage of time that there is a ship passing [-]
A typical value for (f) is 0.03 for the conditions of the Waal River
Equilibrium bed level

For the groyne-field bed level equilibrium to occur, the sediment flux (F) should vanish i.e. (F!0),
which means that C = Ce, then the definition of the bed variations could be written as:
Appendix II 48
∆z
−k ( ) qs
e a
⋅χ ⋅ = Cre ⋅ f ⋅θ b / 2 (II.7)
q
We must realise that it is a dynamic equilibrium, which could be explained in the following way.
The navigation-induced water motion creates a higher sediment concentration inside the groyne-
field, while the concentration in the main channel (qs/q) is almost constant. The damping factor
represented in the left-hand side of the equation [-k .(∆z/a)] must increase, by taking a lower value
for the groyne-field bed level (∆z), (EROSION). On the other hand, in the absence of navigation, the
damping factors must decrease through a higher value for (∆z), (SEDIMENTATION).
An expression for (qs/q) could be estimated with a combination between Engelund & Hansen
sediment transport formula, and Chezy equation,
qs 0.084 ⋅ i Ch
= X ⋅ θ ( n−3) / 2 ; where X = ⋅ [− ]
q ∆ g
Finally, the equilibrium bed level could be written in the following form;
∆z
= A + B ⋅ ln θ (II.8)
a
Where:
−1/ k
 Cre ⋅ f b / 2 
A = ln  
 χ⋅X 
b−n+3
B=
−2 k
Time-scale for groyne-field bed variations

Without any sediment supply to the groyne-field (D = 0), and no sediment outflow to the floodplain
(E = 0), the mass balance could be written in the following form,
∆z
∂z −k ( ) q
(1 − ε ) + w`(
⋅ Cre ⋅ θ − e a ⋅ χ s ) = 0
b/2
(II.9)
∂t q
and, if
∆z ∆ze ∆z `
= +
a a a
then, we can rewrite;
∆z ∆z `
∂ (∆z `) −k ( e + ) q
(1 − ε ) ⋅ Cre ⋅ θ b / 2 − e a a ⋅ χ s ) = 0
+ w`( (II.10)
∂t q
Utilising Taylor expansion the following expression could be reached;
∂ (∆z `/ a ) ∆z `/ a
+ =0
∂t Tm
a (1 − ε )
Tm = ⋅ (II.11)
k ⋅ w` χ ⋅ ( qs )
q
Where, Tm [s], is the morphological time-scale for the groyne-field bed variations. Sieben, (2000)
estimated the morphological time-scale for the groyne-fields bed variations for the Waal River
conditions, to be around 50 days.
Appendix II 49
Sediment Exchange
Using the above-described model, Figure II-1, the sediment exchange between the groyne-field and
the main channel could be roughly estimated. The mass balance could be written the following way:
∂c (a − ∆z )
Agf ⋅ − F + G ⋅ Lgf = 0 (II.12)
∂t
For the value of the horizontal flux (G) to be determined, the time variation of the suspended
sediment “1st term of the equation”, and the vertical flux (F) must be first defined. The vertical flux
is defined in equation (I.2), the horizontal flux could then be written;
∂c ∂ (a − ∆z )
G ⋅ Lgf = − Agf ⋅ (a − ∆z ) − Agf ⋅ c + w`⋅ Agf ⋅ (ce − c) (II.13)
∂t ∂t
For c>>1.0, then:
∂z gf D − E
G ≅ − Lgf ⋅ (1 − ε ) + (II.14)
∂t Lgf
The sediment balance in the main channel could be now written;
∂zmc ∂B qs G V
Bmc + = − dredging (II.15)
∂t ∂x (1 − ε ) (1-ε )
for c<<1, we can use equation (I.14) to substitute for (G), then:
∂zmc ∂B qs ∂z D−E
Bmc + = − Lgf gf + (II.16)
∂t ∂x ∂t Lgf ⋅ (1 − ε )
For the basic equation, de Vries (1996) showed that it could be simplified to:
i- Simple-wave model – for small-scale morphological changes,

∂z c0 ∂z
+ ⋅ =0
∂t α 0 ∂x
c0 1 dq u
with, = ⋅( s ⋅ 0 )
α 0 1 − Fr2
du a
∂z n qs ∂z
then, + ⋅ =0
∂t a (1 − Fr ) ∂x
2
ii- Parabolic model – for large-scale morphological changes,

∂z ∂2 z
− K0 ⋅ 2 = 0
∂t ∂x
c 1 dq u
with, K 0 = 0 = 3 i ⋅ ( s ⋅ 0 )
A0 ( a ) du a
∂z n qs ∂ 2 z
then, − ⋅ =0
∂t 3 i ∂x 2
Applying the same concept to equation (1.16), we end up with two equations describing the main
channel bed variations due to the interaction with the groyne-fields. Neglecting any losses
‘dredging’, or supply ‘filling’ to the system, the two equations will take the following form:
Appendix II 50
 c k Lg ∆z  ∂z u dqs ∂zmc  Lg 1  ∆z   ∂z gf
1 − k (1 − ) 2  ⋅ mc + ⋅ =  c − c k (1 − ) − 1 + ε   ⋅
 (1 − ε ) Bmc a  ∂t a du ∂x  Bmc (1 − ε )  a   ∂t
Simple-wave model (II.17)
 c k Lg ∆z 2  ∂zmc u dqs ∂ 2 zmc  Lg 1  ∆z   ∂z gf

1 − k (1 − ) ⋅ + ⋅ =  c − c k (1 − ) − 1 + ε   ⋅
 (1 − ε ) Bmc a  ∂t 3 i du ∂x  Bmc (1 − ε )    ∂t
2
a
Parabolic model (II.18)
Appendix II 51
View publication stats

The Effect of Groynes On Rivers PDF

Uploaded by

Copyright:

Available Formats

The Effect of Groynes On Rivers PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Effect of Groynes On Rivers PDF

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

The Effect of Groynes on Rivers

Sediment management in rivers and reservoirs View project

The user has requested enhancement of the downloaded file.

Delft University of Technology

Delft Cluster project no. 03.03.04

Delft University of Technology

3. The Effect of Groynes on a River

4. Hydraulic Disturbances Caused by Navigation

5. Morphological Interaction between the Groyne-Fields and the Main Channel

Appendix I – Conceptual Model 44

Appendix II – The model of Sieben & Douben 47

Figure 1-1 River engineering measures as proposed in LPR report

1.3. SCALE OF THE PROBLEM

extrinsic conditions Mega-scale

Figure 1-2 Scale cascade of morphological process according to de Vriend (1999)

Discharge stage Discharge stage

navigation induced large-scale eddies large-scale eddies negligible

Effect on sediment motion Effect on sediment motion

Extrinsic conditions Analysis for different scenarios

Figure 1-3 Overview on the parameters that affect the interaction

1.4. OUTLINES OF THIS REPORT

2.2. TRANSVERSE STRUCTURES – GROYNES

2.2.1. Types of groynes

ii. Classification according to submergence:

2.2.2. Design Considerations for Groynes

2.3. GROYNES IN THE NETHERLANDS

2.3.1. Historical background

2.3.2. Characteristics of the groyne-fields along the Waal River

Table 2-3 The accuracy of the different parameters

25% 100% 14% 100%

Frequency Cumulative % Frequency Cumulative %

3. THE EFFECT OF GROYNES ON A RIVER

3.1. FLOW NEAR GROYNES

3.1.1. Flow near a single groyne

Geometry of separation region

Horizontal large eddies

Water surface fluctuation

3.1.2. Flow pattern in groyne-fields

tip that travels along the main channel groyne-

3.1.3. Submerged groynes

3.2. MORPHOLOGICAL EFFECT OF GROYNES ON A RIVER

3.2.2. Bed Degradation Caused By Long Constriction

ii. For the water depth (h):

iii. To include the effect of yearly discharge variation:

N is the number of seconds in a year

3.2.3. Bed Degradation Caused By Series of Groynes

Where (S) is the spacing between groynes, (L)

3.2.4. Local Scour Near Groynes

Equilibrium scour depth

From the laboratory measurements, he found that,

Local scour depth for a series of groynes

Time Scale for Scour Development

4. HYDRAULIC DISTURBANCES CAUSED BY NAVIGATION

4.1. NAVIGATION INDUCED WATER MOTION

4.1.1. Water movement around a ship

4.1.2. Return current and water level depression

4.1.3. Ship waves

4.2. EFFECT OF NAVIGATION ON THE FLOW IN A GROYNE-FIELD