Theise
Theise
Theise
Submitted by
March 2010
COMPARATIVE ANALYSIS OF DESIGN AND
PERFORMANCE OF BANK PROTECTION WORKS OF
JAMUNA RIVER AT TITPOROL AND DEBDANGA.
Submitted by
March 2010
Bangladesh University of Engineering & Technology (BUET)
Department of Water Resources Engineering
March 2010
We here by recommend that the project prepared by MD. ANISUR RAHMAN entitled
“COMPARATIVE ANALYSIS OF DESIGN AND PERFORMANCE OF BANK
PROTECTION WORKS OF JAMUNA RIVER AT TITPOROL AND DEBDANGA.”
be accepted as fulfilling this part of the requirements for degree of Master of
Engineering in Water Resources Engineering.
This is to certify that this project work has been done by me and neither this project nor
any part thereof has been submitted elsewhere for the award of any of my degree or
diploma.
Bangladesh is located in the lower part of the delta having flat land formed mainly
by the sediments deposited by the world’s three largest rivers (GBM), the Ganges,
the Brahmaputra and the Meghna. Huge quantity of flow during monsoon coming
from the upper catchments (located outside the country) makes its way towards the
sea through these rivers. Due to the geographical location of Bangladesh on the
globe, the river systems here are morphologically very much active which is
evident from the continuous wide spread bank erosion, channel shifting and
sedimentation processes. Bangladesh, on average, is losing more than 8,000 ha of
land annually due to bank erosion. Although there is a tendency of decreasing rate
of erosion due to implementation of large number of riverbank protective works
during the last few decades by Bangladesh Water Development Board (BWDB).
Failure of implemented bank protective works is also remarkable. Under these
circumstances, this study has been conducted to determine the probable cause of
the damage of riverbank protective works constructed on the right bank of Jamuna
River at some selected locations of Sariakandi.
BWDB implemented the protective works at Titporol and Debdanga along the
Right Bank of the Jamuna River during November 2004 to April 2005. From field
investigation on 28 June 2005 it was found that the revetment work at Debdanga
performed well during early flood after construction. But some portion of upstream
revetment at Titporol, however, damaged in June 2005 and mitigation measures by
dumping of sand filled synthetic bags were carried out to stop further collapsing of
river bank.
i
Design of the revetment has been reviewed using the standard procedure
mentioned in the Design Manual of BWDB and have been found satisfactory.
From the analysis of field condition it is revealed that the damage of revetment
works occurred due to low shear strength of soil and the presence of pore water
pressure developed behind the geo-textile filter. It is apprehended that for the lack
of free drainage, pore water pressure developed behind the geo-textile filter and
this resulted failure of bank slope. At the damaged portion, subsoil water might
have been drained from underground source or from the existing ponds behind the
revetment works as was found at the protective site of Titporol. On the other hand,
there found no such kind of underground source of water or existing ponds near
the revetment- works site at Debdanga. Therefore, the protective-works at
Debdanga performed well.
ii
ACKNOWLEDGEMENT
The author acknowledges his indebtedness to Dr. M. Abdul. Matin, Professor & Head,
Department of Water Resources Engineering for his kind supervision, encouragement and
guidance during the course of studies and research. It was great privilege for the investigator to
work with Dr. M. Abdul Matin, whose constant guidance made this work possible.
My heart-felt thanks with cordial gratitude to Mr. Md. Nazibur Rahman, Sub-Divisional
Engineer, Design Circle-V, Bangladesh Water Development Board for his continuous and
sincere co-operation at every stages of carrying out the study.
iii
ABSTRACT i
ACKNOWLEDGEMENT iii
TABLE OF CONTENTS iv
LIST OF TABLES vii
LIST OF FIGURES viii
ACRONYMS AND ABBREVIATIONS ix
TABLE OF CONTENTS
1. INTRODUCTION ........................................................................................................... 1
1.1 General.............................................................................................................................. 1
1.2 Objective of the study ........................................................................................................ 2
1.3 Organization of the Report ................................................................................................. 2
iv
3.5.10 Design of Toe Protection .......................................................................................... 39
v
Appendices
Appendix-A: Sample Design Computation
vi
List of Tables
Table 3.3 Angle of repose θ and density ρ s for various revetment cover layers
Table 3.5 Coefficients for design of various cover materials against wave attack
Table 5.3 : Data considered for the design of bank protection work
vii
List of Figures
Figure 4.2: Historical Bankline and Planform Changes of Jamuna River near the study
Figure 5.1: Study Area showing the channels of the Jamuna River
Figure 5.2: Titporol Revetment along the Right Bank of Jamuna River
Figure 5.3: Debdanga Revetment along the Right Bank of Jamuna River
Figure 5.6: Comparison of bed level in front of Titporol revetment before and after
Figure 5.7: Comparison of bed level in front of Debdanga revetment before and after
revetment
viii
ACRONYMS AND ABBREVIATIONS
ix
CHAPTER ONE
1. INTRODUCTION
1.1 General
Bangladesh is located in the lower part of the delta where the world’s three largest rivers
(GBM); the Ganges, the Brahmaputra and the Meghna meet with the sea. These three
rivers drain a total catchments area of about 1.72 square kilometers in India, Nepal,
China, Bhutan and Bangladesh. More than 90% of these catchments are located out side
the country and huge quantity of flow makes its way towards the sea through these river
systems. Due to this geographical location on globe, the rivers of Bangladesh are
morphologically very active which is evident from the continuous wide spread bank
erosion, channel shifting and sedimentation processes. River bank erosion problem here
in Bangladesh is a regular phenomenon as a result of which the country is constantly
facing immense multi-dimensional socio-economic problems. Bangladesh, on average, is
losing more than 8,000 ha of land annually due to bank erosion. Although different study
reports confirm that there is a tendency of decreasing rate of erosion due to
implementation of large number of riverbank protective works during the last few
decades by Bangladesh Water Development Board (BWDB). Failure of implemented
bank protective works is also remarkable.
Along the right bank of Jamuna river about 5.6 km area between the Hasnapara Spur 2
and Kalitola Groyne (near Titporol) was opened to a major fraction of the Jamuna flow
and at this area the right branch of Jamuna was flowing obliquely to the bank. Due to this
oblique flow during the 2003 monsoon, about 187 m of area was washed away and that
time the Brahmaputra Right Embankment (BRE) was only about 40 m away from the
Jamuna River. If bank erosion would continue at this rate, then there would a threat of
unification of Bengali and Jamuna flow at this area. This probable unification of Bengali
and Jamuna River is not only a threat to the adjacent area of Bogra district, but also is a
threat to the existence of the Jamuna Multipurpose Bridge. Under these circumstances,
Bangladesh Water Development Board has under taken a project entitled “Prevention of
Merger of the Bengali and the Jamuna River at Titporol and Debdanga under Sariakandi
Upazila in Bogra district”. Based on mathematical study carried out by IWM,
Bangladesh Water development board implemented the riverbank protective works at
1
Titporol and Debdanga in 2005. Under the same hydrological condition both the
protective works have been implemented. The work at Titporol has shown several slope
failures after implementation in 2005. But the protective work at Debdanga is
functioning well without any damage. Considering the importance of sustainable bank
protective works in Bangladesh, it is necessary to find out the causes of bank failure.
Under this circumstance, the study has been considered in the proposed research work.
This report describes the overall approach and methodology followed in the study and
the findings. The report contains one report with three appendices, Appendix-A: Sample
Design Computations, Appendix-B: Slope Stability Computation and Apendix-C : Flood
frequency Analysis.
Chapter 1 describes the background, the objectives of the study, Chapter 2 describes
review of literature, Chapter 3 represents study approach and methodology, Chapter 4
presents the hydraulic and morphological characteristics of Jamuna river, Chapter 5
presents the performance analysis and investigation of the causes of failure of revetment
and Chapter 6 describes discussion and conclusions.
2
CHAPTER TWO
2. REVIEW OF LITERATURE
2.1 General
River bank erosion is a major problem in Bangladesh. Bank erosion occurs in stable as
well as unstable river channels. The need of river training and bank protection in
Bangladesh arises from the fact that most rivers of the country are unstable, i.e. they are
not in a state of equilibrium with the governing physical processes. Unstable river
undergo permanent and rapid changes of the water and sediment regime and, hence,
adjust in depth, slope, width and plan form accordingly. Both, river training and bank
protection measures, which are strongly interrelated, have the objective to ensure a safe
and efficient transport of water and sediments (suspended material and bed load) through
a certain defined stretch of the river
Erosion : Generally the removal of the bed material from the land surface is called the
erosion. But when the term is related to the bank revetment works, it may be defined as
the removal of the bed materials and the sediment particles from the river bed causing
reshaped of the particular places. Someone has defined erosion as the mechanism of the
detachment of the sediment particles and other materials from the land surface. There is
another definition which defines erosion as the combination of the processes of the
detachment and the transportation of soil particles and other materials from the bank and
bed of the rivers. (ASCE, 1970). Thus, we can say,” Erosion is the loosening or
dissolving and removal of the earth or rock materials from any part of the earth’s
surface”. However, the bank erosion may be defined as detachment or the dislodgement
of the soil particles from the bank by the action of flow, waves, tidal fluctuations and
other hydrological factors governing the condition of a channel.
Scour : Scour may be defined as the removal of the soil particles from the soil water
interface by current or wave induced share forces within the waterway; possibly in
combination with hydraulic gradient forces or by rainfall run-off above the water line. It
may also be defined as the abrupt decrease in the river bed elevation of channel due to
3
erosion of the bed materials by the flow. The local drop or depression found in bed of a
river or channel found in the region of a structure is known as the local scour. This
happens due to the interaction between the velocity of flow and the bed materials causing
them weak and as a result there is a modification in the pattern. The knowledge of the
maximum depth of scour around hydraulic structures is essential from the point of view
of safety of the structures, excessive scour can undermine the foundations and lead to the
failure of the structures.
According to the experiments carried out by Rouse in 1940, it has been found that (a) the
depth of scour Ds in a uniform material is dependent solely upon the thickness (T) of the
sheet of falling water, velocity of the jet V, the fall velocity of the sediment particles W,
and the duration of the scouring action; (b) the relative rate of scour produced by a given
jet Ds/T at a given stage (time) depends only upon the ratio of jet velocity to fall velocity
V/w (UN-ECAFE, 1953).
An empirical formula for maximum scour at noses of guide banks as given by Inglis and
Joglekar is, Ds = 1.3 (Q/F)1/3 2.1
where Ds is the depth of scour in feet, Q is the maximum discharge in cubic feet per
second, and F is the Lacey silt factor. In addition to the above formula, Inglis and
Joglekar gave the following formulae :
Scour at straight groynes, facing upstream, with steep sloping head (1.5 to 1.0).
Ds = 1.8 (Q/F)1/3 (2.2)
and with long sloping heads (1 in 20)
Ds = 1.3 (Q/F)1/3 (2.3)
For other groynes along the river bends,
Ds =(0.81 to 1.8)x(Q/F)1/3 (2.4)
Hossain (1981) also carried out laboratory experiments on maximum scour depth at the
nose of solid groynes. He arrived at the following equations :
For single groyne, ds = 0.205 (b/w) + 0.196 (2.5)
For double groyne, ds = 0.583 (b/w) + 0.046 (2.6)
where b/w is the groyne projection ratio and ds is the maximum scour depth.
4
2.3 Types of Scour
Normally there are two type of scour, such as natural scour or general scour and local
scour.
1. Scour at bends/Bend scour. This type of scour happens due to helical flow patterns
associated with bend. The maximum scour occurs on the outside of a bend somewhat
downstream of its apex.
Local scour – it is structure-induced scour and usually happens due to the concentration
of flow lines in the vicinity of structure combined with local shear induced turbulence
caused by the roughness of the structure.
The erosion of any place may take place in the following two ways:
1. The displacement of the solid particles from adhesive / cohesive contact when
taken place causes the erosion.
2. The transportation / washing away of the solid particles from the particular
position/ site/place taken place causing the underlying materials attack or wash
and finally there happens an erosion. According to the California Highway
Practice, 1970, this type of erosion is nothing but a natural consequence of the
flow passing through a solid boundary (California Highway Practice, 1970).
The resistive force developed along the contact surface of a moving body with a
stationary body is called friction. When the two bodies are solids, friction is
essentially a function’ of the texture of the surface in contact and the pressure
between them, but when one of the bodies is a fluid, the conflict along its
bounary disturbs its motion for some distance away. The zone in which this
disturbance is significant is called the turbulent boundary layer, The thickness
5
and dynamic character of this layer depend on roughness of the solid boundary
and velocity of the passing fluid and is independent of pressure.
The solid particles on the river bed or on the bank of the river become exposed due to the
flow of water or fluid within very short period of time due to the following reasons:
Due to any of the reasons mentioned above, a loosely bonded particle of the bank or bed
materials are projected from upper layers of the bank or bed. The impinging particles of
water tend to take part in vibration and loosen the solid particles. The retreating particles
introduce sucking called cavitations which is very much harmful and destructive.
Therefore we may conclude that erosion of particular location of the river bed or bank is
related to the roughness of the river bed or bank and the velocity of the flowing fluid of
the channel. Here to mention that, more the roughness more the erosion and more the
smoothness less the erosion. That is why a smooth bank or bed does not go under more
erosion by the tangential flow at even tremendous velocity. River bank is to be classified
as an ideal smooth bank when it will be able to endure fairly high velocity. But when the
roughness of the bank or bed becomes significant, the tolerable velocity decreases. No
doubt the erosion will take place more or less that depends on the bonding forces of the
particles forming the bed or bank materials. The cohesive materials of the bank or bed go
under less erosion and vice versa. Vegetables on the bank or bed play an important role
on the decease of river erosion.
When solid particles from the river bed or river bank are washed away or removed from
the particular places causing the depression in elevation with respect to the surrounding
places are called scour and erosion. This happens when the soil particles composing
together from the particular places are acted upon by the forces sufficient to cause them
to move. Here we may give an example for a straight channel of uniform depth. If • is
6
the unit weight of water, R is the hydraulic radius and S is the slope of the river, then
according to the tractive forces theory, the average tractate force / the acting force of
running water on the river bed is equal to •RS. If the river be is non-cohesive, there is a
definite relation between the critical tractive force and the mean diameter and specific
gravity of the river bed material. Reference may be made to Table 2.1 wherein the
critical tractive force for different sizes of bed material is given.
For a meandering river, the actual tractive force will be larger than the mean value given
by the expression RS on account of the shifting of maximum velocity towards the
concave banks as well as the development of secondary currents (UN-ECAFE, 1953).
The percentage of increase of tractive force, or the decrease of river bed resistance as
compared with a straight channel, suggested by Lane (1951) for different degree of
sinuosity of channel (is given in Table 2.2).
The statement made above refers to the erosion of the non-cohesive bed of rivers. On a
slope, as pointed out by Fan (1947), the resistance of a sediment particle to motion is
reduced by the sliding force of the particle itself on an inclined plane due to its own
weight. If K is the ratio of tractive force required to move a particle on the inclined plane
to that required to move the same particle on a level bottom, Fan (1947) shows that for
non-cohesive materials K is affected by the angle of repose ø of the material and the
angle of side slope ө and can be expressed by the following equation :
2
sin ø
K = 1− sin 2
θ
Another important factor to be considered is the distribution of tractive force on the side
slope of river bank which has been worked out by olsen and Florey (1952). The tractive
force is a place somewhere between o.2D to 0.3D above the bottom and decreases to
zero at water surface wher D is the depth of water.
The above derivation is for non-cohesive material only. If the material is cohesive, size
of particles may be reduced. For irregular and turbulent flow, such as that below a dam
or weir, or wave wash, the actual force may be much greater than the tractive force thus
computed.
7
The tractive force which when becomes just sufficient to set the bed materials of a
certain kind into motion is called the critical tractive force. There is a particular
relationship between the mean diameter plus the specific gravity of the particles of the
river bed material and the critical tractive force. The velocity of flow is automatically
shifted towards the concave banks due to the development of the secondary current in the
meandering section of any river. For this reason, the actual tractive force will be greater
than the average value of the tractive force which is obtained by the formula •RS.
The above mentioned theory, description, or statement / remarks etc. is only applicable
for particles of non-cohesive materials of the river bed. In 1947 Fan mentioned that the
resistance of a sediment particle to motion is decreased by the sliding force of the
particle itself on the sloping surface or on an inclined plane due to its self weight. Fan
also mentioned that the ratio of the tractive force required for the motion or movement of
the particles on the inclined plane to the tractive force required for the movement of the
particles on a level surface(level bottom) for non –cohesive materials is affected by the
angle of repose of the material and angle of the side slope. Olsen and Florey have
pointed an important factor on the distribution of tractive force on the side slope of the
river bank. According to them the tractive force is zero at the bottom of the side slope
increase to a maximum at a place some where between 0.2d to 0.3d above the bottom
and decrease to zero at water surface where d is the depth of water. Here to mention that
there are cohesive and non –cohesive materials which may be available in the formation
of river bed. But the above mentioned theory, formula and derivation are only applicable
for non-cohesive materials. The shape and size of the particles of the bed and bank
materials may be reduced if they are cohesive in nature. The flow below a dam/
embankment/ sluice/ or the flow over the wire is irregular type of flow. For such kind of
irregular or turbulent flow the tractive forces which are calculated from the above
mentioned theory / formula becomes much less than the actual forces developed there.
A bank may fail owing to any one or a combination of the following reasons
1. Washing away of the soil particle of the bank by current or by the waves which is
called erosion.
8
2. Sliding due to the increase of the slope of the bank as a result of erosion and
scour.
3. Undermining of the toe of lower bank by current, wave, swirls or eddies followed
by collapse of overhanging materials deprived of support which is called scour.
4. Sloughing or sliding of the slope when saturated with water, this is usually the
case during flood of long duration.
5. Sliding due to seepage of water flowing through bank into the river after receding
of the flood, the internal shearing strength is considerably decreased owing to
saturation and the stability is further decreased by the pressure of the seepage
flow.
6. Piping in the sub layer due to movement of ground water to the river which
carries away sufficient material with it.
7. Scouring of bed and bank by eddies with horizontal axis when flow occurs over a
reef or submerged structure
The various hydraulic actions responsible for the erosion in particular location can
usually be classified as follows (After California Highway Practice, 1960)
2.7.1 General
In order to prevent or minimize the loss of valuable land, several stretches of the river
banks might need suitable protection against erosion. Design of protection structures
requires control of erosion by installing properly designed filters covered usually by
large sized materials (stones, concrete blocks etc.) to absorb the energy of moving water.
9
Due to non accountability of different field circumstances and construction limitations
experienced in field protection structures fail to perform satisfactorily in many occasions.
Reconstruction or repairing works of the same structures creates huge economic burden
for the country.
At present, geo-bags are used in bank protection works as stable protective elements. In
Bangladesh, geo-textile materials were employed on a larger scale since 1994.
Commonly sand filled geo-bags are used in emergencies but not for long lasting and
large scale bank protection. To develop low cost alternatives for riverbank protection,
Bangladesh Water Development Board took up an experimental river erosion mitigation
project named Jamuna Meghna River Erosion Mitigation Project (JMREMP). The
project looks back on eight years of implementing effective and low cost river training at
several places of the Jamuna and Meghna River. Sand filled geo-bags were identified as
a low cost effective solution during a pre- feasibility study in 2000 and subsequently they
were used by BWDB in 2001. However, very limited investigations are carried out to
evaluate the performance of geo-bags as a protective measure for rivers in Bangladesh.
Therefore, engineers should gather knowledge regarding the performance of geo-bags as
a bank and bed protection materials when applied to different field conditions.
A river bank is nothing but a sloping surface of the adjacent land coming into contact
with the surface of the still or flowing water. Protection means to save something from
the attack of someone. Therefore bank protection means to save the river from the action
of wave or current of the river water. It includes all type of protective works having the
target of maintaining the stability of the land against the action of flow of the river water
or surface runoff. This definition includes all types of protection works such as
protection works of shore lines done along the sea shore and lakes against waves action
and drift, protection work along the navigation canals against wave action generated due
to passing of the various type of vessel and the protective work along the embankment
and banks of the rivers for the purpose of the flood control.
A river bank may be divided into three parts, such as 1. Embankment portion which is
called the sloping surface of the embankment facing the river. 2. Upper bank – that part
10
of river bank which is located above the low water level and below the shore level. 3.
Lower bank which portion of the river bank is in below of water surface. Depending on
the different parts of the river bank different methods of protection measures are taken.
Type of works to be done for the purpose of training a river or a reach there of is
dependent on the objective and engineering principle chosen to be adopted in the river
training program. The usually adopted type of river training works are described below
with specially reference to their important features and applicability.
Bank revetment
Revetment is a bank protection measure where the bank slope is covered with erosion-
resistant materials. The function of the revetment is thus to reduce the hydraulic load
acting on the soil and possibly to facilitate stabilization of the soil. This type of river
training works involve a protective cover of a suitable hard material applied on the slope
and toe of the river bank so that the bank soil protected from the actions of erosions of
erosive forces of flowing water and dynamic actions of waves. Revetments are of the two
types:
1. Open joint type in which joint gaps between individual hard materials remains open
allowing free flow of water.
Open joint type revetments are suitable in situations where river water level and or
phreatic surface within the bank fluctuates significantly either seasonally or diurnally
because in combination with suitable filter media, releasing of hydrostatic pressure is
effectively possible.
2. Close joint type in witch joint gaps between individual hard materials are sealed
using cement, bitumen, or asphaltic materials. Close joint type revetments do not
allow free flow of water and are liable to develop detrimental head of water behind
witch materials eventually lead to blow-offs.
11
Close-joint type revetments are suitable only in situations where fluctuations in water
level (river or phreatic) are either insignificant or not of concern e.g. upper bank above
HWL.
Different types of materials are used for slope protection. In general, revetments can be
grouped as:
1. Self-adjusting structures- including all kinds of rip-rap protection using stones,
concrete blocks etc. but also hand-laid bricks and concrete block layers without
interlocking, slurry-filed geotextiles bags etc.
12
2. Flexible structures- including articulated blocks and slabs, wire mesh netting and
other forms of mattresses, tubes, gabions and interlocking block layers.
3. Rigid structures- including asphalt and concrete paving, concrete filled pillows,
grouted mattress, grouted rip-raps, soil stabilization etc.
Figure 2.2 : Different types of self adjusting revetment: multi-layer riprap and
geobags (left), loose cement concret blocks (right)
Based on construction materials, revetment may be of different types: rock riprap, rubble
riprap, wire riprap, wire enclosed rocks, pre-formed blocks, grouted rocks or paved
lining etc. Material units used for cover or armour layer are (a) stones or boulders, (b)
concrete blocks or slabs, (c) various bituminous system and (d) fabric containers.
Different practices of using these materials are explained below.
Rip-rap is one of the most common types of armour layer which comprises randomly
placed quarried rocks. It is made up of durable stone, sizes ranging typically from 10 to
50 cm. Usually it is specified by mass and typically ranges from 10 to 500 kg. Rip-rap is
usually placed in one, two or three layers. Their shape depends on rock quality and can
be rounded, cubical, tabular, angular or elongated. Angular, near cubical shaped stones
are most suitable for riprap construction. In Bangladesh, boulders are normally collected
from the north-eastern rivers. Quarried stones are not locally available at large quantity.
Potential sources of quarried rocks are Chittagong Hills and Madhayapara hard rock
mine. Currently, assorted sizes of quarried rocks are available from the later source
13
Protection Measures with Pitched Stone
The pitched stones or hand-pitched stones are entirely placed by hand as a single layer.
Normally, approximate single-sized stones are used only. The shape of stone should be
nearly cubical and the ratio of length to breadth should not exceed two. Pitched stone can
only be installed above water level, therefore, as part of other revetment system only.
Pitching is normally placed on an appropriate filter layer. In case of high current or wave
attack, pitched stones are placed on cement mortar.
In order to increase the stability of rip-rap or pitched stones a cement grout or bitumen
can be applied. This system is more stable to wave and current attack than loose stones.
The permeability is reduced in this case and the stability may be adversely affected by
excess pore pressure during ground water flow towards the bank.
Pre-cast concrete blocks generally without reinforcement are designed for a specified
minimum strength with durable aggregates. Different shapes like cubical, cuboidal etc.
are used. Loose blocks can be hand-placed above the low water level and dumped below
low water level like rip-rap (Figure 2.3).An approximate sub-layer or filter layer is
required. The concrete blocks are widely used in Bangladesh due to the fact that the
blocks can be made on-site using local non-technical persons with reasonable cost.
14
Protection Measures with Concrete Slabs
Use of plain concrete or reinforced concrete slabs, pre-cast or cast in situ, is also another
option for cover layer. Their applicability is, however, limited since they are normally
built or placed in the dry condition. The use of pre-cast concrete slabs above low water
level is more convenient if placed on an appropriate filter. Concrete slabs have been used
on embankment slope in Bangladesh against wave erosion.
Placing of concrete blocks under water may become more systematic and quicker in case
the individual blocks are held together by cables of steel or synthetic material reducing
the risk of localized failure. All types of cable-connected blocks are to be anchored by
appropriate means at locations beyond potential failure planes which may be difficult to
achieve in practice.
Bituminous systems provide a flexible cover layer which can withstand substantial
hydraulic loadings. Special attention must be given to the durability of the bituminous
revetments because bitumen hardens with exposure to ultra-violet radiation and the
atmosphere. Furthermore, abrasion can be caused by floating material, which can even
create damages to the revetments above the waterline. Asphalt mixes may be designed to
15
be either permeable or impermeable. Bituminous system is not used for bank protection
in Bangladesh except that use in the Jamuna Multipurpose Bridge Project.
Jute bags, synthetic bags or geo-textile bags (woven or non-woven needle punched)
filled with ballast can be placed in one to three layers directly on a slope or along the toe
of the eroding slope or on eroding river bed. The type of ballast is to be selected
considering the required flexibility and permeability of the cover layer. Sand and gravel
filled bags are relatively flexible.
The non-woven needle punched geo-bags filled with sand is one of the very useful
material for protection of riverbank. Fine sand available in the river bed filled in geo-
textile bags of desired size, can be effectively used for protection of vulnerable stretches
of the riverbank. In this approach, the slope below low water level is protected with
geobags. The slope above it is protected with cement concrete blocks mainly to resist
wave action. The bags are filled with sand and weight 78 kg, 126 kg and 250 kg.
Currently, geobags are used both for emergency and permanent protections.
LWL
Natural bank soil
sand-filled
geotextile bags
Scour
Falling apron hole
16
Groynes are structures built at an angle to the river bank to deflect flowing water away
from downstream critical zones to prevent erosion of bank. Groynes may be constructed
for multiple purposes to control erosion, to maintain safe navigation channels as well as
to reduce the flow velocity.
A typical groyne has three parts: head, shank and root. The riverside end of the groyne is
the head of the groyne, the joint of the groyne with the existing bankline is the root of the
groyne and the longer part joining the head and root of the groyne is the shank. Groynes
can be classified based on
A groyne placed at right angle to the bank (perpendicular flow attack) is termed as
deflecting groyne. Groynes inclined in upstream direction are called repelling groynes
and groynes pointed downstream to attract the flow towards the structure’s head and thus
to the river bank is called attracting groynes
Different types of groyne heads of impermeable and unsubmerged groynes are described
below and shown in sketches of Figure. 2.5.
Figure 2.5 Different types of groyne heads: (a) straight, (b) bell-head, (c) T-head,
(d) L-head, (e) Hockey-head
Permeable groynes are generally constructed of timber, steel or reinforced concrete piles
driven or sunk into the river bed in one or several rows. Impermeable groynes can be
built of local soil, stones, gravel and rock with suitable slopes at the shanks and the head
or even vertical walls at the shanks, using sheet piles. In case of an appropriately sloped
17
earth dam, the shank and the head have to be protected by a cover layer placed on a
suitable filter layer
These are structures extending from the bank into the river to intervene with flowing
water in ways conductive to predetermined purposes in relation to the river training.
Spurs/Groynes can be constructed of timber, steel or concrete piles and of earth, rock
boulders, solid constructions are sometimes called Groynes.
Submersible spurs which are of low height and often go under water. Top profiles
of submersible spurs may be either horizontal or sloping towards the river.
Non-submersible spurs/groynes which are of adequate height such that they will
not be submersed even at the highest flood condition in the river.
Submersible spurs have the inherent disadvantage that they induce scour at the
downstream due to flow of water over them. They are sometimes dangerous to river
traffic. However, they are suitable for rivers carrying floating debris.
Surface erosion of river banks occurs if the driving erosive forces are exceeding the resistive
forces of the individual grains or of the conglomerates in case of cohesive materials. The main
impacts responsible for surface erosion at river banks are :
18
d) Surface runoff and
e) Mechanical action (desiccation, ship impact, activities of humans and animals)
Mass damage of river banks can be divided into slip failures, block failures and flow slides,
initiated by different processes, which are illustrated in Figure 2.6. The actual damage of a river
bank may not follow immediately after an impact. In some cases the failure process takes several
days. On the other hand, a damage may occur without warning at almost any time if active
surface erosion and toe scouring is prevalent or an additional (surcharge) load is applied to the
bank. The risk of mass damage is increased during heavy rain and during quick fall of river
stages after flood. Potential failure modes of slip and block failures are listed below:
Surcharge loading
Surface drainage and infiltration
Block failure
High pore
water pressure
Slip failures
• In case of non-cohesive material and a shallow bank angle the failure surface is usually
approximately parallel to the slope angle. Water seepage can substantially reduce the
stability of the bank, whereas vegetation will normally help to stabilize against failure.
• Steep or almost vertical banks of non-cohesive material can fail along a plane or slightly
curved surface. This is often the case when the river water level is low relative to the
total bank height.
19
• If relatively deep tension cracks have developed on the surface of the river bank, failure
occurs by sliding and/or toppling. This failure mode is little affected by the groundwater
table, but is more likely if the crack fills with water.
• Deep-seated rotational failure is possible in cohesive soil where the banks are steep and
moderately high. If the soil is relatively homogeneous, the failure surface may follow a
circular arc.
• It is also possible that layers of weak material affect the actual shape of the failure
surface, which may then include logarithmic spirals or even planar sections. Both types
of failure can extend beyond the toe of the bank. The stability is significantly affected by
the position of the water table and if the tension cracks are filled with water.
• If the outside of an eroding meander bend lies at the edge of the river valley, further
erosion can trigger a massive landslide stretching up the valley slope. Tension cracks,
bulging above the toe or noticeable movement are signs of potential failure.
Block failure
• If the lower part of a composite bank, which is more frequently exposed to flowing
water, consists of more erodible material such as sand and/or gravel, the upper part can
be under-cut and falls as a complete block down the slope.
Typical modes of river bank failure are exemplary shown for slip failures and for block failures
and for block failures in Figure 2.7(adapted from Hermphill and Bramley, 1989.).
20
a) Slip Failure
Water table
Before After
b) Block Failure
Cohesive
Sand/gravel
Before After
Figure 2.7 :Typical Riverbank Failure
Apart from the causes described above, the following causes may also be worth
mentioning here in connection with the cause of bank failure.
21
2. Bulk head and groynes may fail by outflanking due to :
a) insufficient protection at the wing
b) embayment by the river further upstream.
Long groynes may be breached at their middle portion owing to infringement by a sharp
turning current.
3. Paved slopes themselves may fail because of one or more of the following causes
22
CHAPTER THREE
A number of field visits have been made in order to know the existing condition of the
protective works at Titporol and Debdanga as well to collect the recent data particularly
needed for the technical evaluation of the implemented bank protective works at
Titporol. During field visit the following data has been collected:-
The data that are collected during the field visits are analyzed and used for the review of
design procedure of the protective works employed by BWDB. The review of design
includes the evaluation of cover layer design, filter design and toe protection design
under hydraulic load and anticipated scour condition. Pilarczyk formula has been used
for assessment of cover layer of protective works and available scour formula has been
used for scour calculation. Observed scour values has been compared with the predicted
scour. In addition, attempts were also made to identify the possible causes of the failure
of the slope and other damages by checking the stability of slope. As the stability of
slope depends very much on the properties of the underlying soil of the slope, the bore-
log data collected from the field office has been analyzed. Slope stability analysis has
been carried out using a computer program widely used in considering different options
such as with different river stages and seepage forces, different slopes of protective
works etc.
23
3.3 Interaction with BWDB field Officials
An interaction meeting was made during the course of this study with BWDB field
officials in order to get necessary information regarding the constructional matters such
as time and schedule of construction and their views of the failure of the protective work
at Titporol.
Similar analyses mentioned above were made for the protective works at Debdanga
considering it as an implemented effective protective works. Comparison of the results of
two cases is made to identify the cause of failure at Titporol.
In this chapter, the hydraulic and hydrologic design parameters for bank protection has
been discussed. These design parameters are the design discharge, the maximum,
minimum and standard low water levels, flow velocities, wind speed and wind waves,
design scour depths and morphologic changes. Design procedure for the protective
works and different approaches for the geobag as a river bank and bed protection
material has also been described here.
The Design Water Level (DWL) is related to PWD and can be derived from the design
discharge. For the design of bank protection structures maximum, minimum, bankfull
and average low water levels are very important. The design water levels shall be
determined at the proposed protection work site, from the water levels available at some
reference stations and the water surface slope applicable for that sequence. The design
water levels may be taken from available analyzed records or may be computed from
available data.
Low water level (LWL) is particularly important for fixing under water river training
works. Mainly the average high water occurring in the lean period (December-March)
shall be the focus for fixing the DWL.
24
The design water levels (DWL) are the calculated water levels based on measurements
for the last four decades. It is assumed that the observed water levels are not influenced
by wind set-up and transverse gradients in a cross-section. The design Water Level
(DWL) is related to PWD dat6um and can be derived from the design discharge. As the
relation between discharge and water level varies due to rapid morphological changes,
only a stage discharge relation (rating curve) established at the location of the planned
structure from long term monitoring of daily averaged water levels and corresponding
discharges can be used.
River levels for both flood and dry season play important role in flood control and river
training works. For the design of bank protection structures maximum, minimum, bank
full and average low water levels are very important. The design water levels shall be
determined at the proposed protection work site from the water levels available at some
reference stations and the water surface slope applicable for that sequence. The design
water levels may be taken from available analyzed records or may be computed from
available data. The following end results of hydrological analysis are important for the
design :
Flood Levels: 1:100 year
1:50 year
Bankfull level
Low Water Level: Lowest during dry season (December-March)
Low water level (LWL) is particularly important for fixing under water river training
works. Design Low Water Level (DLWL) in consideration of construction and
maintenance of bank protection works shall ventilate the safe construction window
available in average years. Mainly the average high water occurring in the lean period
(December-March) shall be the focus for fixing the DLWL.
25
• Adaptation of too low value of LWL may create a situation that the level
may not be achieved in the year of construction.
• The idea for time available for construction (construction window) should
there ore be such that some flexible planning may be done well before the
construction period.
Selection of DLWL:
(I) Non Tidal Area
• In order to make a balance between too high and too low value of
LWL it is proposed that Upper Quartile value of annual LWL or a
value in consideration of construction window be adapted as the
DLWL for the design.
The design flow velocity may be determined by average flow velocity and from physical
model investigation. Average flow velocity in a cross-section of a channel is estimated
with a regime equation. Design flow velocities from measurement in a physical model
investigation depend on the approach flow and on the alignment of bank protection
structure. In general the depth averaged flow velocity is used as design flow velocity.
The design flow velocity may be determined according to the following approach:
26
Design flow velocities are obtained either from field measurement or in a
Physical model investigation. These design flow velocities depend on the
approach flow and on the alignment of bank protection structure
The flow velocity obtained from the method stated above shall be verified with the
observed measurements available for that stretch of the river. For the design of drag
force on the piles and bed protection around the piles of permeable groynes a designed
flow velocity is defined as the upstream flow velocity which is not influenced by the
permeable groynes. In general the depth averaged flow velocity is used. The flow attack
on a revetment depends not only on the discharge and the water level but also on the
alignment of approach channel. The design flow velocities are the maximum flow
velocities measured on the physical model with the extreme alignments of the approach
channel.
Flow fields in the main rivers, however, are determined by the bed topography, which is
determined by the morphological processes. Average velocities in the Jamuna can be
evaluated using at-a-station relationships (Klaassen & Vermeer, 1988), resulting in:
A similar approach for the Ganges on the basis of the at-a-station relations derived
in RSP special report 7 (1996), results in:
u=0.36 Qo.15 (3.2)
where u is the average velocity and Q is the bankfull discharge.
For a bankfull discharge of 44,000 m3/sec, this results in an average velocity of about
1.5 m/sec for Jamuna and 1.8 m/sec for the Ganges. Locally the velocities may be much
higher due to dimensional effects. Maximum velocities measured were (FAP-24):
27
3.2 m/sec in an eroding outer bend in Jamuna near Kamarjani,
3.7 m/sec near a protrusion in the Jamuna near Bahadurabad
4.0 m/sec near a protrusion in the Ganges, upstream of Gorai offtake.
The design discharge can be taken from available analyzed records or may be computed
from available data. The discharge of a specific river is obtained from the analysis of
hydrological data, especially through extrapolation of stage-discharge relations at water
level stations, where the discharge measurements have also been executed regularly. The
analysis of flood discharges and the associated recurrence periods result in a probability
function which can be used to define the design discharge with a return period of 100
years. The discharge of the design flood is defined in order to estimate the Design Water
Level (DWL) with a return period of 100 years. Additionally, a bankfull discharge is
defined for the estimation of the design cross-section.
According to Inglis,
For example if
Maximum Discharge, Qmax = 1,010 m3/s
Q bankfull = Qdominent = 2/3*Q max = 673.33 m3/s.
Waves are defined by wave height H, wave period T, wavelengths L, and direction.
Waves are often generated far from the place where they are observed. However, related
wind speed and duration can be derived from the observed wave height, wave period and
wave direction.
Waves at the river training sites should either be generated by wind or by water vessels.
wind waves would usually govern the design of protection work at the slopes of river
training works. With respect to protection, two aspects of waves have to be considered:
28
Run-up of waves against slope which might overtop the upper limit of
protection,
Erosive forces of breaking waves against the slope causing erosion.
The principal factor affecting the design of slope protection is wave action. The
mechanics of wave generation are extremely complex, and the forces causing erosion
during wave attack on an earth slope are both varied and complex. The described ranges
of riprap design assume that the wave height is a direct measure of the erosiveness of the
wave.
To evaluate wave height the following factors that create waves are to be analyzed:
Design wind direction,
Effective fetch,
Wind speed and duration.
The mechanism of formation of wind generated wave and its relation/dependence with
duration of wind, wind speed, fetch length, depth of water and other phenomena have
been described in following sections.
Generation of Waves
The generation of waves depends on fetch length, wind speed and duration of wind. The
fetch length is the length of the water surface, for example of a lake the wind is blowing
across. It is the length of the water surface, where the wind can transfer energy to the
water. Wave generation can be limited by (i) the duration of the occurring wind (duration
limited) or (ii) the length of the water surface of lake (fetch limited). When a wind blows,
with essentially constant direction, over a fetch for sufficient time to achieve steady-
state, it is termed as fetch-limited values. The second idealized situation occurs when a
wind increases very quickly through time in an area removed from any close boundaries.
In this situation, the wave growth can be termed duration-limited. It should be
recognized that this condition is rarely met in nature; consequently, this perdiction
technique should only be used with great caution.
29
Wind must blow for a certain time to develop the full wave height for the given fetch
length. Only after some time of blowing across the surface, sufficient energy is
transferred into the water surface to generate the full wave height. In rivers, lakes and
estuaries fetch determines the wave condition and not the duration of the wind (Alam
and Fontijn (2006). So design of bank protection works in rivers will be governed by
fetch limited wave conditions.
Wind direction: wind direction can be obtained by determining the point on the shoreline
over the longest stretch of open water from the embankment. The direction should be
weighted with other topographic conditions or climatic information.
It is hardly possible to give firm prediction of the directions in which wind generated
waves would be progressing. For design purposes it will have to be presumed that they
will approach any river training work perpendicularly.
Effective Fetch (F e )
Early studies on wind and wave development assumed the fetch to be the greatest
straight-line distance over the open water. Subsequent studies by Seville (1952) showed
that the shape of an open water area affects the fetch, the smaller the width to length
ratio, the smaller the effective fetch. Effective fetch can be determined from
climatological data or from site conditions.
Reliable estimate of the maximum wind speed that would exist over a length of time at a
given site is practically impossible. However, value of wind speed can be obtained from
climatological data of the area for cyclones, thunderstorms or norwester and design wind
speed may be selected for appropriate return period.
30
Significant Wave Height (Hs)
The significant wave height is the average height of the highest one-third of the waves
for a specified period of time.
Waves in Rivers
The flow velocity of rivers has to be taken into account for calculating the wave height.
The wind is transferring energy into the water based on the sheer stress. If the wind
speed and the flow velocity have the same direction the sheer stress becomes less,
because the difference in velocity is less. To calculate the wave heights for these cases
the flow velocity has to be deducted from the wind speed if they both have the same
direction. If the direction is opposite to each other the velocities need to be summed up.
For the major rivers in Bangladesh the highest measured flow velocities over longer
areas are 2 m/s close to the surface even though locally peaks of up to 4m/sec can be
observed. Assuming a fetch length of 5 km and a wind speed of 17 m/sec acting for 20
minutes on the river surface, the generated wave will have a height of 0.7 m and a wave
period of 2.1 sec. Superimposing the flow velocity u = 2 m/s and the wind speed of v =
17 m/s, calculated wave height is 0.8 m with a period of 2.2 sec. If the wind comes from
the same direction then the wave height is 0.6 m with a period of 2.0 sec.
φ K K
2
0.035 ⋅ u τ
≥ ⋅ sc h
D (1)
n
∆ m ⋅ 2g K ⋅Ψ s cr
Where,
Dn (m) nominal thickness of protection unit (cover layer)
31
• m (-) (ρ s -ρ w )/ρ w = relative density of submerged material
φ sc
(-) stability factor
Kτ (-) turbulence factor
Kh (-) Depth factor, dependent on the assumed velocity profile
friction
The stability parameter φ sc depends on type of application , some guide values are given
in Table 3.1.The degree of turbulence can be taken into account with the turbulence
factor K τ and some guide values for K τ are given in Table 3.2.
32
Table 3.2 Turbulence Intensity Factor K • (current)
Turbulence Intensity K • (-) K • (-)
Gabions, Mattresses Others
Normal turbulence in rivers 1.0 1.0
Non-uniform flow with increased 1.0 1.5
turbulence, mild outer bends
High turbulence, local 1.0 2.0
disturbances, sharp outer bends
Source: FAP-21, (2001)
With the depth parameter K h, the water depth is taken into account, which is necessary to
translate the depth averaged flow velocity into the flow velocity just above the
revetment. According to Pilarczyk (1998) for a non developed velocity profile K h factor
can be expressed as
K h = (h/D n +1)-0.2 and (2)
K h = 1.0. (For very rough flow, h/k s <5)
A set of values for angle of repose θ for different materials used as protection element on
different types of filter material and their density have been proposed in the Guidelines
and Design Manual for Standardized bank Protection Structures, FAP 21 is shown in
Table 3.3.
Table 3.3 Angle of repose θ and density ρ s for various revetment cover layers
Revetment type Angle of Density of
Cover layer Filter repose protection material,
material θ [°] ρ s [kg/m3]
Randomly placed, broken rip-rap Geo-textile 25 2650
and boulders Granular 30
CC blocks (cubical shape), Geo-textile 30 2250
randomly placed in multi layer Granular 35
CC blocks (Cubical shape), hand Geo-textile 20 2250
placed in single layer Granular 25
Gabions/mattress filling by stones 45 2650
Source: FAP-21, (2001).
Some guide values for critical Shield’s parameter Ψ cr are given in Table 3.4.
33
Table 3.4 Critical Shield’s Parameter Ψ cr
4.1× 10 −5 ⋅ S s .V5
W =
(S s − 1)3 ⋅ cos3 α (3)
2× 10 −5 ⋅ S s ⋅V
6
W = (4)
(S s − 1)3 ⋅ sin 3 (70 − α )
(4) PIANC
2
0.70 V
=
D ( )⋅
(5)
n
g (S s − 1) ⋅ cos α 1 − tan 2 α / tan 2 θ
(5) JMBA
2
0.7V 2 1
= ⋅ ⋅
D 2 ⋅ (S s − 1)⋅ g log(6h / D ) [1 − (sin α / sin θ ) ]
2
(6)
n 2 0.5
34
(the equation has been developed for cc blocks)
The wave attack is considered significant for the revetment cover layer above low water
level and some parts of the falling apron. The minimum dimensions for the stability of
the cover material under wave attack can be determined by Pilarczyk formula and
Hudson’s formula. The Pilarczyk formula being more universal with its breaker
similarity index is preferred.
H s ⋅ξ z
b
Dn ≥
∆ m ⋅ Ψ u ⋅ φ sw ⋅ cos α
(7)
where,
Dn [m] characteristic size of the revetment cover layer (single unit
35
size for loose elements, thickness for mattress systems)
Hs [m] significant wave height
• m [-] relative density of submerged material = (ρ s -ρ w )/ρ w
g (m/s2) acceleration due to gravity (= 9.81)
Ø sw [-] stability factor for wave loads
Ψ u [-] system specific stability upgrading factor
α [°] bank normal slope angle
ξz [-] 1.25 ⋅ T m
wave breaker similarity parameter= tan α ⋅
Hs
Tm [s] mean wave period
b [-] wave structure interaction coefficient, dependent on
roughness and porosity of protective material
The formula is valid for ξ z <3 and cotα ≥ 2. The material and armour layer unit specific
coefficients to be applied for design against wave attack are summarized in Table 3.5.
Table 3.5 Coefficients for design of various cover materials against wave attack
Revetment type Stability Stability Interaction
factor for upgrading coefficient,
incipient factor,
motion Ψ u [-] b [-]
φ sc [-]
Randomly placed, broken riprap 2.25-3.00 1.00-1.33 0.50
and boulders
CC blocks, cubical shape, 2.25-3.00 1.33-1.50 0.50
randomly placed in multi-layer
CC blocks, cubical shape, hand 2.25 2.00 0.67 ≤ 1.00
placed ,single layer (geotextile
filter)
CC blocks, cubical shape, hand 2.25 1.50 0.67 ≤ 1.00
placed in single layer, chess
pattern (geotextile on sand)
CC blocks cable connected 2.25 1.80 0.67
Wire mesh mattress 2.25 2.50 0.50
Gabions/mattress filling by stone 2.25 2.50 0.50
Source: BRTC, BUET, (2008)
36
3.5.7 Thickness and Grading of Riprap
Riprap Thickness
Opinions of different authorities regarding the thickness of slope pitching are given
below:
1. U.S. Army Corps of Engineers (1991), recommends that thickness of
protection should not be less than the spherical diameter of the upper limit
W 100 (percent finer by weight) stone or less than 1.5 times the spherical
diameter of the upper limit W 50 stone, whichever results in greater thickness.
5. The thickness of stone pitching and soling for permanent slopes required at
head, body and tail of guide bank for river flowing in alluvial plains as
recommended by Gales (1938) is given in Table 3.7.
37
Table 3.6 Spring’s Table to Compute Thickness of Stones on Slope
River bed Thickness in inches for river Remarks
materials as slopes in inches per mile
classified by 3 9 12 18 24
Springs
Very Coarse 16 19 22 25 28 The stone pitch prevents
Coarse 22 25 28 31 34 sand underneath from
Medium 28 31 34 37 40 being sucked out by high
Fine 34 37 40 43 46 velocity. More rationally
Very Fine 40 43 46 49 52 stone pitch thickness
should be based on
velocities.
Source: BRTC, BUET, (2008)
Geotextile is use as a filter material. The main design parameters for geotextile filters are
the retention criterion and the permeability criterion, which define the capability of the
material to retain the subsoil without clogging and to allow undisturbed water transport
through the membrane.
38
3.5.9 Determination of the Range of Grain-size Distribution
Range A: 40% or more of the soil particles are smaller or equal to 0.06 mm.
Range B: 15% or less of the soil particles are smaller or equal to 0.06 mm.
Range C: between 15% and 40% of the soil particles are smaller or equal to 0.06 mm
At the estimated maximum scour depth, the launching apron is assumed to cover and
stabilize the bank-sided river profile, preventing from further erosion of the bank. The
method had been widely used on sand bed streams. Scour depth R can be calculated by
Lacey’s formula which is given below
Scour Depth R = 0.47 * ( Q d / f )1/3 (10)
Where,
Q d = Design Discharge
f = Silt factor = 1.76 x √ D 50
39
2.25T
2.25T
40
XR
ASL
41
CHAPTER FOUR
4. HYDRAULIC AND MORPHOLOGICAL
CHARACTERISTICS OF JAMUNA RIVER
The Jamuna River is one of the three main rivers of Bangladesh and one of the largest
in the world, with its basin covering areas in Tibet, China, India and Bangladesh. The
Jamuna flows south, ending its independent existence as it joins the Padma River near
Goalundo. Merged with the Padma, it meets the Meghna River near Chandpur. Its waters
then flow into the Bay of Bengal as the Meghna River.
Actually Jamuna is the downstream course of the Brahmaputra which took place after the
Earthquake and catastrophic Flood in 1787. Presently the Brahmaputra continues
southeast from Bahadurabad (Dewanganj Upazila of Jamalpur district) as the Old
Brahmaputra and the river between Bahadurabad and Aricha is the Jamuna, not
Brahmaputra. It originates in the Chemayung-Dung glacier, approximately at 31°30'N
and 82°0'E, some 145 km from Parkha, an important trade centre between lake
Manassarowar and Mount Kailas. The Brahmaputra is known as the Dihang in Assam
Himalayas before it comes into the Great Plains of Bengal. It enters Bangladesh through
Kurigram district (at the border of Kurigram Sadar and Ulipur upazilas). The total length
of the Tsangpo-Brahmaputra-Jamuna river up to its confluence with the Ganges is about
2,700 km. Within Bangladesh territory, Brahmaputra-Jamuna is 276 km long, of which
Jamuna is 205 km.
42
Average annual flow at Bahadurabad is estimated to be 501 million acre-feet. August has
always been the month when widespread flooding has been most likely. Floods from
May to July are usually due to the Brahmaputra-Jamuna and Meghna. From August to
October due to the combined flows of those rivers and the Ganges. As a rule, the flow of
the Brahmaputra-Jamuna is more erratic than that of the Ganges. The gradient of the
Jamuna averages 1:11,850 which is slightly more than that of the Ganges. The Jamuna
discharges a large volume of water and at the same time brings in huge amounts of
Sediments. During the rainy season it brings down something like 1.2 million tons of
sediment daily, and the annual silt runoff at Bahadurabad is estimated at 735 million
tons. It has four major tributaries: the Dudhkumar, the Dharala, the Teesta and the
Karatoya-Atrai system. The first three rivers are flashy in nature, rising from the steep
catchments on the southern side of the Himalayas between Darjeeling in India, and
Bhutan. Of all the distributaries, the Old Brahmaputra is the longest and was actually the
course of the present Brahmaputra some 200 years ago.
The satellite imageries of Jamuna river for last thirty years show that the river channel
near the right bank is highly active and there is a strong trend of rightward shifting of the
bankline. The average easting for right bank of the Jamuna river moved about 1.5 km to
the west between the year 1973 and 1961. The bank line shifting of the study area from
year 1973 to 1996 is shown in Figure 4.1. The unsteady movement of the Jamuna right
bank particularly along the study area has also been observed in the recent years.
The river's average depth is 395 feet (120 m) and maximum depth is 1,088 feet (332 m).
The width of the river varies from 3 km to 18 km but the average width is about 10 km.
In the rainy season the river is nowhere less than five kilometers broad. The river is in
fact a multi-channel flow. Channels of many different sizes, from hundreds of meters to
kilometers wide, and of different patterns including braiding, meandering and
anastomosing pattern in the country. It is, through most of its course within Bangladesh,
studded with islands (Chars) many of which are submerged during the rainy season and
makes a single water channel. Thus, by breadth alone, this river qualifies as one of the
largest in the world. The width/depth ratios for individual channels of the Brahmaputra
43
vary from 50:1 to 500:1. The gradient of the river in Bangladesh is 0.000077, decreasing
to 0.00005 near the confluence with the Ganges.
44
4.4 Morphological characteristics of the Jamuna River
The Jamuna is a dynamic river of highly braided nature. The river has got multiple
numbers of inter connecting channels. The direction of individual channel varies from
+90 and -90 degrees where the larger channels show smaller deviations from the river
axis. Numerous alluvial chars (permanent and moving) are the characteristics of the
river. The presence of bars obstructs flow and scour occurs, either lateral erosion of
banks on both sides of the bar, scour of the channels surrounding the bar, or both. This
erosion enlarges the channel and reduces water levels.
The position of braids is likely to shift during floods, resulting in unexpected velocities,
angle of attack and depths of flow. Lateral migration of braided streams takes place by
lateral shift of a braid against the bank. The morphological behaviour of the Jamuna river
near the study area could be understood from the plan form changes as shown in Figure
4.2. The slope of the river decreases in the downstream direction. Near Bahadurabad it is
about 8 cm/km, while near the confluence with the Ganges near Aricha it is about 6
cm/km.
Within the braided belt of the Jamuna, there are lots of chars of different sizes. An
assessment of the 1992 dry season Landsat image shows that the Jamuna contained a
total of 56 large island chars, each longer than 3.5 km. There were an additional number
of 226 small islands/ chars, varying in length between 0.35 and 3.5 km. This includes
sandy areas as well as vegetated chars. In the Jamuna the period between 1973 and 2000,
chars have consistently appeared in the reaches opposite to the Old Brahmaputra off-
takes, north and east of Sirajganj and in the southernmost reach above the confluence
with the Ganges. In entire Bangladesh during 1981 to 1993, a total of about 729,000
people were displaced by Riverbank Erosion. More than half of the displacement was
along the Jamuna.
45
Figure 4.2: Historical Bankline and Planform Changes of Jamuna River
near the study area (Titporol and Debdanga during 1973-2003
46
CHAPTER FIVE
5. PERFORMANCE ANALYSIS AND DISCUSSION
The study area is located along the right bank of the Jamuna river at Titporol and
Debdanga under Sariakandi Upazila in the Bogra district. Titporol is situated at
immediate upstream of the existing Kalitola groyne and Debdanga is at 1.0 km
downstream of the existing Mathurapara Hard Point. The location of the study area is
shown in Figure 5.1. During monsoon 2003, about 187 m of area at Titporol was washed
away and that time the Brahmaputra Right Embankment (BRE) was only about 40m
away from the Jamuna. If the bank erosion would continue like this then there would
threat of unification of Bangali and Jamuna flow at this area. On the other hand, near
Debdanga, the right anabranch of the Jamuna river was flowing along the river bank.
During the same monsoon period there was a report of about 140m bank erosion at this
area and the Jamuna was only 210m away from the Bangali. Under this situation BWDB
implemented two protective works at Titporol and Debdanga in accordance with the
recommendation of model study.
47
Figure 5.1: Study Area showing the channels of the Jamuna River
48
Figure 5.2: Titporol Revetment along the Right Bank of
Jamuna River
49
5.3 Model studies for the protective works
BWDB commissioned IWM to carry out modelling study for planning and designing of
bank protection work at Titporol and Debdanga in November 2004 through a formal
contract signing. Beforehand, on request of BWDB, IWM submitted a Technical report
in January 2004 containing the planning options for the crucial bank protection works at
Titporol and Debdanga. This report was prepared based on the morphological model
developed for Special O&M programme at Site B1 structures of RBPP using the
measured 2004 pre-monsoon bathymetry data.
The sustainability of river training works largely depends on river response due to the
intervention in terms of river bank erosion/accretion, channel and char development
especially in the braided river. Hence, to predict such developments and to suggest
appropriate protective measures three two dimensional morphological models were
developed using data of 2004, 2005 and 2006. Two-dimensional mathematical modeling
tool MIKE 21C of DHI Water and Environment has been employed for the modeling
study. The sustainability of the implemented works largely depends on future
development of rivers in terms of bank erosion/accretion, channel and char development
etc. 2-dimensional modelling enables to predict such developments and to suggest
appropriate protective measures in case of any adverse developments during the on-
going implementation of the protective measures through adaptive and monitoring
modelling. MIKE 21C operates on curvilinear computational grid, incorporates fully
unsteady flow, bed load as well as suspended load via a non-equilibrium transport
formulation taking into account both phase lag and time lag effects in adaptation of
suspended load. The suspended load is calculated with the inclusion of the inertia of the
sediment in suspension i.e. adaptation and the effect of the helical flow is accounted
through profile function.
The following modelling study have been carried out regarding the implementation of
Titporol and Debdanga revetments
50
pre-monsoon 2004, this model was updated with the bathymetry measured by IWM in
pre-monsoon 2004 under RBPP, Site B1 Project. In the updated model, morphological
verification was done against pre-monsoon 2004 bathymetric data (hind cast). The model
also included the proposed river training work to compare the impact of proposed river
training works on existing water environment. The model was also employed to
investigate the hydro-morphological condition under different application scenarios.
The design of protective works at Titporol and Debdanga was prepared by Design
Circle-VI of BWDB in accordance with the findings and suggestions given by IWM. As
mentioned earlier. Typical design of Titporol revetment is shown in Figure 5.4. The
detailed design parameters and design data are provided in Table 5.1 and 5.2
respectively.
51
Table 5.1: Detailed design parameters
SL. No. Parameter Value
1 Total length of revetment 2000 m
2 Slope of slope pitching by CC Block 1:2
3 Length of Launching Apron 26.00 m
4 Volume of Launching Apron 52 m3/m(excluding void)
5 Size of CC Block of slope pitching 40 cm x 40 cm x 40 cm
6 Average thickness of Launching Apron 2.00 m
7 Thickness of Geo-textile Filter 3.00 mm
8 Thickness of sand filter 100 mm
52
Figure 5.4: Typical Design Section of Titporol Revetment drawn at Km. 17.00 (C/S No. 20)
53
5.5 Performance of the protective works
From field investigation on 28 June 2005 by IWM it was found that the revetment work at
Debdanga performed well during early flood after construction. Some portion of upstream
revetment at Titporol, however, damaged in June 2005 and mitigation measures by dumping
of earth filled synthetic bags were carried out to stop further collapsing of river bank. From
BWDB field source it is seen that the Titporol revetment from M -60.00 to M 0.00, from M
0.00 to M 360.00, from M 1125.00 to M 1600.00 (total 895 m) was damaged. Typical
failure is shown in Figure 5.5.
15
10
5
Bankline after damage
0
0 5 10 15 20 25 30 35 40 45 50 55 60
Distance in m
From some surveyed cross-sections, taken by IWM after the repair works it was observed
that the apron launched to lowest bed level of -3.00 m PWD while the same was initially
laid at level of about +7.50 m PWD. The anticipated scour level was -13.92 m PWD.
54
would not be constructed, then the bed scour would more aggravate the situation. This
indicates the effectiveness of the revetment.
The vertical change of the channel bed after one monsoon (2005) near Titporol and
Debdanga revetments were also investigated using the data collected by IWM. It is seen that
near downstream tip of Titporol revetment, river bed erosion occurred due to monsoon of
2005. At Debdanga revetment, deposition occurred after 2005 monsoon indicating the
effectiveness of the revetment at this reach.
Based on pre-monsoon 2005 data before Based on pre-monsoon 2006 data after
construction of Titporol revetment construction of Titporol revetment
Figure 5.6: Comparison of bed level in front of Titporol revetment before and after
construction of the revetment indicating sedimentation near Titporol revetment
Based on pre-monsoon 2005 data before Based on pre-monsoon 2006 data after
construction of Debdanga revetment construction of Debdanga revetment
Figure 5.7: Comparison of bed level in front of Debdanga revetment before and after
construction of the revetment indicating sedimentation near Depdanga revetment
Source: IWM 2005
55
5.6 Investigation of cause of failure of Titporol Revetment
The river bank protection works may be failed due to rapid change of hydro-morphological
conditions of the river at or near the implemented structures and also by the failure of the
riverbank /embankment itself due to sliding of soil mass along a curved surface. In order to
find the probable cause of failure of the protective work at Titporol, investigation has been
carried out through checking the adequacy of the design of the revetment, slope stability
analysis and analysis of field condition during damage of revetment.
On the basis of collected field data and observation, necessary design parameters have been
analyzed. C.C. block sizes were determined by applying commonly used formulas and
those were compared with the sizes actually provided in the implemented revetments at
Titporol and Debdanga. Adequacy of blocks in terms of size and weight has been assessed
by following this procedure. Combining this with the conditions observed in field,
performance of the bank protection works have been evaluated in terms of (i) effectiveness-
whether the protective works implemented could effectively cease bank erosion, stabilize
the slope, withstand flow slides and protect toe (ii) appropriateness- if the types of
protection works were appropriate for the situation in consideration of river size, current,
waves, etc.
Major design criteria to determine the size of the CC blocks are the resistance against
probable maximum near bank velocity and wave action, while the depth of probable
maximum scour governs the volume of launching apron materials. The maximum size of
river bank soil material determines the type of geo-textile fabrics to be used as filter.
There are several formulas for evaluating the size of CC Blocks against near bank velocity
and significant wave height. However, Neills’ method and JMBA equations are preferred
for estimating normal size of cc blocks, while Pilarczyk equation is used against wave
action for the purpose. The following equations are used for checking the design. The detail
computation is given in appendix-A.
56
A. Determination of size of Revetment: Design against current attack
b) Using JMBA Equation (This equation has been developed for CC Block)
0.7V 2 2 1
Dn = x x
2( Ss − 1) g log(6h / d ) [1 − (sin α / sin θ ) 2 ]0.5
2
0.035û 2 ϕ sc K τ K h
Size (dia of stone) of revetment cover layer material, D n =
∆ m 2g K s ψ cr
ρ w = Density of water
Ss =Specific gravity
Kτ = Turbulence factor
57
B. Determination of size of revetment against wave attack
Pilarczyk equation
H sξ z
b
Size of revetment material, D n =
Λ mψ uφsw cos α
where,
An earthen embankment usually fails, because of the sliding of a large soil mass along a
curved surface. It has been observed that the surface of slip is usually close to cylindrical,
i.e. an arc of a circle. The method which is used for computing stability of slope of an
embankment is known as Slip Circle Analysis. The minimum factor of safety against slip
circle failure is considered as 1.5. The computation of slip circle analysis has been made
using computer software “XSTABL” which shows a factor of safety less than 1.5 (1.13).
The geotechnical data used for the analysis are moist, saturated, cohesion(c) and angle of
internal fraction. The details computation of slope stability is given in Appendix-B.
58
It has been reported by the field engineers of BWDB that there were a good numbers of
ditches/ponds within 50m of the revetment at Titporol. From the analysis it is revealed that
the damage occurred during last week of June 2005 due to low shear strength of soil and in
presence of pore water pressure. It is apprehended that for lack of free drainage, pore water
pressure developed behind the geo-textile and this resulted failure of bank along a curved
surface. At the damaged portion, subsoil water might have been drained from underground
source or from the existing ponds behind the revetment as was found at the site.
From analysis of morphological data surveyed by IWM it has been seen that about 120m to
150m bank erosion occurred along the right bank of Jamuna River at Antarpara at the
upstream of Titporol revetment. This morphological change may be another reason for the
damage of the upstream tip of the revetment during 2005 monsoon.
5.7.1 Introduction
In order to find the probable cause of failure of the revetment at Titporol a comparative
study of the implemented revetment works at Debdanga and Titporol was as made in terms
of design and the performance during the first monsoon they faced in 2005. The
observations made during the study through data collection and analysis, field visits and
interaction with field officials, review of the design of the corresponding works are
discussed in this chapter. The different design and implemented parameters of the
revetments are compared and shown in Tables 5.3 and 5.4.
59
Table 5.3 : Data considered for the design of bank protection work
No Data At Titporol Remarks
1 Highest Water Level (HWL) 19.07 m (PWD) Same data was
2 Lowest Water Level (LWL) 10.00 m (PWD) used for
3 Design discharge 71774.50 m3/s Debdanga
4 Design Velocity (from model study) 3.00 m/s protective work
5 Silt factor, f 0.80
6 Design Bank Slope above LWL 1:2
7 Wind Fetch Length 10.0 km
8 Wind Velocity 20.00 m/sec
9 Significant wave height 0.90 m
10 Wave period 3.40 sec
11 Duration of Wind 1.75 hour
For the purpose of evaluation, the present study compared the different parameters
measured in pre-monsoon 2005 (without project condition) and post-monsoon 2005 (with
project condition) along the right bank of Jamuna River near Titporol and Debdanga areas.
Based on the measured data, following are the assessments on the performance of the
Titporol and Debdanga revetments during monsoon 2005.
1. Both revetments, at Titporol and Debdanga, were fully capable to stop the bank
erosion at the right bank of Jamuna river within the area. It is seen from the findings
of model study that if the revetments were not constructed (hypothetical situation),
60
then the bank would more aggravate the situation. This indicates the proper
functioning of the revetments.
2. About 120m to 150m bank erosion occurred along the right bank of Jamuna river at
Antarpara at upstream of Titporol revetment. As a result, the upstream tip of Titporol
revetment was damaged during monsoon 2005. But no such things occurred for the
Debdanga revetment.
3. Prominent sedimentation occurred in front of Debdanga revetment whereas
significant sedimentation occurred near the upstream tip of Titporol revetment.
4. Slip circle failure occurred at different locations of the Titporol revetment. But no
such damage occurred at the Debdanga revetment.
5. There were a good numbers of ponds/depressions behind the Titporol revetment, but
there were no such features behind the Debdanga revetment.
Design of the revetment has been reviewed using the standard procedure mentioned in the
Design Manual of BWDB and have been found satisfactory. The detail design calculation is
provided in the Appendix-A.
The factor of safety against slip circle failure has been found to be 1.13 which is less than
the minimum allowable factor of safety of 1.5. The detail slope stability computation is
given in Appendix-B.
From the analysis of field condition it is revealed that the damage that occurred during June
2005 due to low shear strength of soil and in presence of pore water pressure. It is
apprehended that for lack of free drainage, pore water pressure developed behind the geo-
textile and this resulted failure of bank slope. At the damaged portion, subsoil water might
have been drained from underground source or from the existing ponds behind the
revetment as was found at the site during field visit.
61
CHAPTER SIX
6. CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusions
i) This study has been done on protection works implemented at Titporol and
Debdanga along the right bank of the Jamuna River. However similar comparative
study can be carried out for other similar locations on the river.
ii) More field data can be collected in order to carry out such comparative analysis.
iii) Available low cost bank protection method can be used for future comparison of
traditional revetment works.
62
REFERENCES
1. Ahmed, N (1989), “Study of Some Bank Protection Works in Bangladesh.”
M. Engineering thesis, Department of Water Resources Engineering, BUET,
October 1989. PP. 2-13
4. FAP 21, (2001), “Guidelines and Design Manual for Standardized bank Protection
Structures.” WARPO, December, 2001. PP. A-1, A-2, B-1, B-2, 24-25
7. BRTC, BUET, (2008) “Guidelines for River Bank Protection,” BWDB, JMREMP,
Consultancy Report, BRTC, BUET, May, 2008. PP. 158-162, PIRDP-1,4
63
Appendix-A : Sample Design Computation
Peak discharge of Jamuna river at Bahadurabad = 1,02,535 m3/s (25 year flood as found
from the frequency analysis shown Appendix-C).
Thickness, t = 0.06Q1/3 m
= 0.06 x (71774.50)1/3 m
= 2.50 m
For medium bed material, provided thickness for river slopes per mile is 31 inch (≈ 0.80 m).
In accordance with Gale’s, for a river having discharge from 1.50 million to 2.50 million
cusec, the thickness, t = 3’-6” = 3.50 feet = 1.066 m.
From the above analysis the reasonable thickness of the rip-rap has been found to be 1.00 m.
1/ 3
Qd
Lacy’s regime normal scour depth, R = 0.47
f
1/ 3
71774.50
= 0.47 m
0.80
= 21.00 m
64
Therefore, Depth of scour, D = LWL – ASL
= 10.00 – (-12.43) m
= 22.43 m
b) Using JMBA Equation (This equation has been developed for CC Block)
0.7V 2 2 1
Dn = x x
2( Ss − 1) g log(6h / d ) [1 − (sin α / sin θ ) 2 ]0.5
2
65
c) Pilarczyk (1990/200) equation:
0.035û 2 ϕ sc K τ K h
Size (dia of stone) of revetment cover layer material, D n =
∆ m 2g K s ψ cr
h = Water Depth = ( HWL - LWL ) = 9.07 m
ρ s = Density of protection material = 2250 Kg/m3
ρ w = Density of water = 1000 Kg/m3
Ss =Specific gravity = 2.25
∆ m = relative density of submerged material = 1.25
g = accleration due to gravity = 9.81 m/sec2
φ sc = Stability factor for current = 0.65
ψ cr = Critical shield stress parameter = 0.05
Average flow velocity, u = 3.00 m/sec
Kτ = Turbulance factor = 1.0
α = Slope angle of bank = 26.56° = 0.464 rad
θ = Angle of repose of protection material = 40° = 0.698 rad
Ks = Slope factor = √ [ (1- (sin α / sin θ )2] = 0.718
0.035û 2 ϕ sc K τ K h 0.0818
Dn = =
∆ m 2g K s ψ cr 0.8808
= 0.0928 m
By trial and error, when assume Dn = 0.0930, then calculated Dn becomes 0.0928 m
Thus required Dn = 0.093 m = 93 mm
Required size of revetment material single unit (CC block) under current attack has been
found to be 306 mm considering maximum of the above three values.
Size provided in implementation of revetment = 400 mm, hence the design is O.k under
current attack.
66
2. Determination of size of Revetment: Design against wave attack
H sξ z
b
Size of revetment material, D n =
Λ mψ uφsw cos α
Significant wave height, Hs = 0.90 m
Unit weight of revetment material, ρ s = 2250 Kg/m3
Unit weight of water, ρ w = 1000 Kg/m3
Relative density of submerged material, ∆ m = 1.25
Acceleration due to gravity, g = 9.81 m/sec2
Stability factor for incipient motion,φ sw = 2.25
Stability upgrading factor, ψ u = 1.50
Slope angle of bank, α = 26.57° = 0.464 rad
Mean wave period, Tm = 3.40 sec
Wave similarity parameter, ξ z=tanα (1.25Tm/√ Hs) = 2.24 (<= 3, O.k.)
Wave structure interaction coefficient, b = 0.67
67
Appendix-B : Slope Stability Computation
******************************************
* XSTABL *
* *
* Slope Stability Analysis using *
* Simplified BISHOP or JANBU methods *
* *
* Copyright (C) 1991 *
* Interactive Software Designs, Inc. *
* All Rights Reserved *
* *
* *
* GEOSYNTEC *
* Norcross, GA 30093 *
* *
* Ver. 4.01 1055 *
******************************************
1 type(s) of soil
68
Water Surface No. 1 specified by 2 coordinate points
*******************************************************
PHREATIC SURFACE,
**********************************
1 12.00 38.80
2 153.00 48.00
ANGULAR RESTRICTIONS:
69
Factors of safety have been calculated by the MODIFIED BISHOP
METHOD.
1 12.00 38.88
2 17.00 38.83
3 21.98 39.31
4 26.88 40.31
5 31.64 41.82
6 36.22 43.83
7 40.56 46.30
8 44.62 49.23
9 48.35 52.56
10 51.70 56.27
11 52.95 58.00
70
Figure B-1: X-Section of riverbank showing Critical slip circles
71
The following is a summary of the TEN most critical surfaces.
• * * END OF FILE * * *
72
Appendix-C : Flood Frequency Analysis
Observed Data:
Year Value (m3/s) Sample size = 31
1976 61700 Qmean = 64221
1977 67300 Standard deviation,• = 16650.6
1981 34200 yn = 0.53702
1982 53500 •n = 1.11558
1983 50900 Recurrence XT y Frequency
75700
K• QT (m3/sec)
1984 Interval, T Factor, K
1985 61900 2 -0.521 0.367 -0.15 -2541.1888 61679.85
1986 42000 2.33 -0.613 0.579 0.04 624.710937 64845.75
1987 67200 5 -1.014 1.500 0.86 14378.815 78599.85
1988 68700 10 -1.340 2.251 1.54 25581.3298 89802.37
1989 70500 20 -1.652 2.971 2.18 36327.051 100548.09
1990 64400 25 -1.751 3.199 2.39 39735.737 103956.77
1991 61500 50 -2.057 3.903 3.02 50236.2848 114457.32
1992 65600 100 -2.360 4.601 3.64 60659.2982 124880.33
1993 59100 200 -2.662 5.297 4.27 71044.2798 135265.32
1994 39100 1000 -3.362 6.909 5.71 95100.1907 159321.23
1995 84200
1996 83485
1997 79270
1998 102535
1999 61915
2000 69320
2001 49230
2002 69728
2003 65684
2004 96106
2005 58767
2006 47666
2007 30329
2008 62379
2009 86939
73