Sediment Control at Intakes

Download as pdf or txt
Download as pdf or txt
You are on page 1of 78

SEDIMENT CONTROL AT INTAKES

A Design Guide
This book, resulting from an intensive study by experts in the,field
"'1C CENTRE OD
of civil engineering hydraulics, gives a detailed and complete
account of the problem of sediment ingress at water intakes. As I ffi
well as supplying information on the sources and nature of the i. i -?" - ' '

materials involved and basic sediment theory, the guide provides


the engineer with practical advice, case studies and sample
calculations to aid in the design of efficient structures.
Of especial interest to practising engineers in the water, hydro-
power and agricultural industries, this book also provides a
valuable source of reference for researchers, academics and
students in water science and civil engineering disciplines.

Related titles of interest


BHRA publishes a wide range of literature of interest to the water
AT INTAKES
engineer For full details of the following, or other publications, contact:
Publication Sales, BHRA - The Fiuid Engineering Centre, Cranfield, ~|
Bedford MK43 OAJ.

Civil Engineering Hydraulics Abstracts


(ISSN 0305 9456) 1986
Water Quality Modelling in the Iniand Natural Environment
(1SBM 0 947711 16 3) 1986
A DESIGN GUIDE
Hydraulics of Floods and Flood Control
(ISBN 0 947711 08 2) 1985
Water for Energy Edited by P. Avery
Ni
(ISBN 0 947711 14 7) 1986
Hydraulics for Pumping Stations
(ISBN 0 947711 07 4) 1985
Hydraulic System Contamination Bibliography
(ISBN 1 85166 147 6) 1988
Pipe Protection Bibliography
m (ISBN 1 85166 077 1) 1986

THE FLUDEHGMEEMHG CEIfTHE

ISBN 0 947711 47 3
SEDIMENT CONTROL AT INTAKES

.':: V r:ir
CONTRIBUTORS: SEDIMENT CONTROL AT INTAKES
P. Ackers Binnie & Partners - A DESIGN GUIDE
P. Avery BHRA
M.E, Bramley Binnie & Partners - C.I.R.I.A.
P.A. Hammett Sir Alexander Gibb & Partners
P.N. Hooper lames Williamson & Partners
P. Novak University of Newcastle-upon-Tyne
A.R. O'Hea Sir William Halcrow & Partners
Edited by P. Avery
K.V.H. Smith University of Southampton (BHRA Process Engineering Division)
A.J.H Winder Watson Hawks iey

The work in prcpiirint: this report was undertaken as part of


Grant No. MEF.67'0114. a contract within the Department of Trade and
Industry's "Support for Innovation" scheme. The financial support from the Ul II wk
THE RJJDENGIWEBWG CEWTRE
Department of Trade and Industry is gratefully acknowledged.
CONTENTS
First published in 1989 by

BHRA, The Fluid Engineering Centre


Cranfield, Bedford MK43 OAJ, England.
Tel Bedford (0234) 750422 Telex 825059 Fax (0234) 750074 PREFACE vii

NOMENCLATURE viii
ISBN 0 947711 47 3
1. THE PROBLEM WITH SEDIMENT CONTROL AT INTAKES 1
BHRA 1989 1.1 Row Diversion and Sediment Regime 2
1.2 Problems Caused by Abstracted Sediment 3
1.3 Examples of Problems Caused by Sediment 4
1.4 Changes in River Regime 8

2. DATA REQUIRED FOR DESIGN 9


British Library Cataloguing in Publication Data
2.1 Purpose and Location of a River Intake 10
Sediment control at intakes.
2.2 Topographic and Geomorphologic Data 10
1. Water pumps. Intakes. Design
2.3 Hydrometric Data 12
I. Avery, P.
2.4 Operational Data 13
621.252
2.5 Sediment Data 13
2.6 Data Collection Techniques 16
ISBN 0-947711-47-3
3. INTAKE LOCATION AND ALIGNMENT 21
3.1 River Bend Hydraulics 22
3.2 Alignment of Intake 23
3.3 Hydraulics of 90 Diversion Angle 23

4. RANGE OF INTAKE STRUCTURES 25


4.1 Types of Intake Structure 26
4.2 Choice of Intake 39

5. DESIGN OF INTAKES 41
5.1 Design Procedure 42
5.2 Geometric Recommendations 45
This publication is protected by international copyright 5.3 Floating Debris 46
law. All rights reserved. No part of this publication may be 5.4 Fish 48
reproduced, stored in a retrieval system, or transmitted in any form
or by any means, electronic, mechanical,
photocopying, recording or otherwise, without the
permission of the publishers.
6. SEDIMENT EXCLUDING DEVICES 51 Preface
6.1 Curved Channel Sediment Excluder 52
6.2 Vortex Tube Sediment Extractor 58
6.3 Side-Sluice Sediment Excluders 62
6.4 Tunnel Excluders 69 This design guide is the result of a collaborative effort by a number of experts in
6.5 Approach Flow Control 71
6.6 Sediment Excluding Basins
the field of civil engineering hydraulics. The Department of Trade and Industry
72
supported BHRA in setting up the panel of contributors, coordinating progress
7. SETTLING BASINS 79 throughout the life of the project and editing the document.
7.1 Fall Velocity of Discrete Particles 82
7.2 The Ideal Basin 83
The aim of the panel was to produce a guide which would enhance the engineer's
7.3 Settling Capability of Real Basins 84 understanding of the sediment problems associated with intake design, and the
7.4 Hydraulic Factors Affecting Basin Layout 89 considerations involved and the options to afford a measure of sediment control at
7.5 Operation of Basin and Removal of Sediment 91 intakes.
7.6 Examples of Settling Basin Design 95
7.7 Design Procedure 99 The efforts of those involved are rewarded by the publication of this book, which
is a unique compendium of hitherto uncollated advice and information concerning
8. EXAMPLES OF MODIFICATIONS TO INTAKES TO 103
IMPROVE SEDIMENT CONTROL this subject. Of special interest to engineers working in the water, hydropower
and agricultural engineering industries, we hope readers will find this guide
APPENDICES 111 helpful.
1. Descriptions of typical intake features 112
2. Sediment transport theory 114
3. Example calculation 133 Pauline Avery, November 1988
4. Model studies 137
HOMEHCLflTURE Vd
Re - Reynolds number ()
R - hydraulic radius of flow section (Area/Hetted Perimeter) (hydraulic
A - mean plan area of settling basin
mean depths)
A - value of F at which motion first starts - related to efficiency of
s - specific gravity of solids
transport process
sw - specific gravity of water
b - width of channel
S - hydraulic gradient {or i)
C - local concentration of sediment (at elevation y above bed)
SF - shape factor (= a//bc where a, b, c are mutally perpendicular dimen-
Ca - concentration of sediment (at reference elevation a above bed)
sions, a being the smallest)
CQ - incoming concentration
tp - retention time
Cr - concentration of suspended sediment removed
t - settling time
Cv - volumetric sediment concentration
v - shear velocity (= A /Q or /gdS)
d - water depth
Y - mean velocity of flow
D - particle diameter (D^Q - median diameter)
Vn - normal channel velocity
V - dimensionless grain size
- bed load efficiency factor
Vs50 - settling velocity of median size sediment
eb VQ - limiting velocity for zero sediment transport
- sediment mobility number ( = /Y)
V - Froude number ( -
u
)
Vs - flow velocity at incipient deposition
Fr w - individual particle terminal velocity
/gd
H - width of settling basin
- acceleration due to gravity
x - effective roughness coefficient
- weight of solids passing/unit width/unit time
X - mass rate of sediment transport per unit width
- immersed weight of solid passing
Y - mobility number
- transport parameter
V
H - head over weir
Ycr
z
- critical mobility number
- Rouse number (- w/BKv # )
I - dimensionless grain parameter
Z - relative grain size ( = d/D)
K - von Karman constant (0.4)
a - coefficient relating roughness, k s , to median sediment diameter
ks - roughness (equivalent sand gradient diameter)
tana - solid friction coefficient
L - length of settling basin
6 - ratio of sediment diffusion coefficient to momentum coefficient
m - performance parameter (m - 0 for 'best' basins; m = 1 for 'very poor1
(assumed 1.0)
basins)
y - unit or specific weight of solids
n - transition parameter (n : 0 for coarse sediments; n = 1 for fine Y - submerged unit weight of solid phase
sediments) S - reference laminar sublayer thickness
nm - Mannings n A - bed effective roughness
q - discharge intensity (volume/unit width/unit time) E - sediment diffusion coefficient
qt - sediment transport rate (submerged weight/unit width/unit time) rt - sediment removal efficiency
Q - discharge or flow rate x - friction factor (8gdS/V2)
Qc - canal flow rate (other subscripts defined in text) \i - kinematic viscosity of the fluid
r - radius of curvature (subscript i - inner wall; subscript o - outer p - density or unit mass of fluid
wall) p - density or unit mass of solids
s
i - bed shear stress
critical sheaf stress
Einstein transport function

Chapter 1

THE PROBLEM OF SEDIMENT CONTROL AT INTAKES


1. THE PROBLEH OF SEDIKEHT COMTROL ftT INTAKES slower flowing area of the bend. This means that the water on the outer, or
concave, side of a bend is clearer than that on the inside, a feature made
1.1 Flow Diversion and Sediment Regime much use of in the location and design of intakes.

The means of taking water from a river are numerous and varied. Whether Even when water diversion takes place via an intake structure on the
for irrigation, water supply or hydropower the water being removed must pass outside of a bend, the curvature of diverted flow may develop in an opposite
through some man-made structure. No matter how slight, any interruption to direction to that of the river if an inappropriate design is used, in effect
the river course will cause changes to the river regime. The change being creating an artificial bend flow with the intake on the inside of the bend.
dealt with here is that to the sediment regime of the river. It is often the
case that the proportion of sediment abstracted with relation to the total Clearly, if a river carries sediment as bed load, suspended load, or
river sediment load is greater than the proportion of water abstracted. both, it is impossible to abstract water devoid of sediment and, essentially,
the hydraulics of flow diversion rule that the bed load will be drawn towards
The sediment load of a river may be classified broadly as bed load or the intake structure or point of diversion.
suspended load. The bed load is made up of particles which move by sliding,
rolling and saltation; the deciding factor being that bed load material pro- So it is not surprising that numerous problems have been experienced
perties are such that the particles cannot stay in suspension for long. The world-wide with resulting high expenditure on remedial works or maintenance.
suspended load travels predominantly in suspension because the particle sizes This illustrates the need for careful planning in the design of an intake
and densitites are such that the turbulence in the flow does not allow them to structure from the early collection of data, consideration of requirements,
settle out. The wash load is defined as those very fine particles which may selection of site, etc., to the thoroughly studied design of the intake struc-
only settle out in completely still water. The sediment transport mechanisms ture. The aim of this document is to enhance the understanding of the sedi-
are fully described in Appendix 2, section A.2.2. ment problems associated with intake design, the considerations which must be
taken into account and the options available to the designer to afford a
The nature of the transported sediment depends largely on the supply, measure of sediment control at the intake.
from catchment erosion, landslips, etc., but also on the character of the
river channel. Typically, near the river source, where velocities are 1.2 Problems Caused by abstracted Sediment
highest, a large quantity of sediment may be carried by the river. This may
be partly in the form of bed load, i.e. coarse sand, gravel, or larger stones If the problem of sediment ingress at an intake works is not considered,
and even boulders and partly as suspended load. As the river progresses to major difficulties will often result due to sediment being transported by the
lower, flatter ground its velocity decreases and larger particles are deposit- diverted flow, or by deposits caused by a reduction in the sediment carrying
ed. As the river flows through sandy or silty lowland a greater proportion of
capacity of the diverted flow.
the sediment transported will be of much smaller particle sizes, i.e. fine
sand, silt and clay particles, and much of this will generally be in suspen- .Small particles which remain in suspension, i.e. if the flow velocity is
sion. When the river is in flood the flow volume and velocity increase and kept sufficiently high, may cause damage to any part of the intake works,
with them the sediment carrying capacity of the river also increases. particularly where machinery such as pumps (water supply) and turbines (hydro-
power) is involved. Exposure of such machinery to small abrasive particles
As will be explained later (section 3.1), the bends in a river are areas moving through the impeller or turbine runner at high velocity causes damage
of particular concern in terms of the process of erosion and deposition. This over short periods of time; seals and bearings may suffer severe wear, so
is due to the helicoidal pattern set up in the flow cross-section causing the that efficiency is reduced and complete failure may eventually follow - in
bed load to be carried away from the outer bank and swept towards the inside, either case maintenance is necessary, requiring high expenditure in terms of
^ ^^ ^ ^ ^^ ^* ^P MM W MM ! flM flHi
trol. Control includes the exclusion or limitation of sediment entry and also
replaced parts, man-hours and in lost water or power supply. Screens, pen-
the removal of sediment that gets through the intake. Once the designer is
stocks and moving parts, such as gates and valves, are also susceptible to
aware of how much sediment can be accepted, all possible design considerations
extensive damage from sediment in motion or from sediment deposits.
should be made to ensure that this quantity is not exceeded.

Wail, channel or pipe protection measures designed to resist exposure to


The following examples are of intakes where serious problems arose
water flow may not be sufficient to protect such surfaces against particle
because of the quantity or control of sediment.
movement, and both cosmetic and structural damage may result.

(1) The Layong Intake, part of the Sungai Tutong Scheme in Brunei, (Fig.
In addition to the costs of damage due to sediment carried through
1.1) had an original capacity of 0.75 m^/s but this has been increased
intakes, there is the problem, particularly in the case of water supply, of
to 1.095 m^/s. Raw water pumped to the treatment works was found to
removing the particles at treatment works, thus adding to the already expen-
sive process of providing potable water to the consumer. contain sand and silt.

In cases where diverted flow velocities are not high enough to keep all
particles in suspension - particularly in the gravity systems experienced
mainly in irrigation schemes - the main problem is to avoid a greater intake
Probable flood level
of sediment than can be transported by the canal system. An excess of
sediment will lead to deposition in the canal system, at first locally just Normal water level
downstream of the intake but gradually extending downstream to affect the
whole canal system in time. The canal system will have strictly limited scope
for adjusting slopes to accommodate the excess sediment supply, so freeboard y River pumping station
will be lost as the bed rises and the capacity of the channel will decrease. Section A-A
The process of clearing and removing sediment is very expensive, especially if
a proportion of fine cohesive material results in very solid deposits.

1.3 Examples of Problems Caused by Sediment

Bearing all these problems in mind, it should nevertheless be realised


that it is seldom necessary to exclude all sediment from intakes and man-made
waterways. This may be virtually impossible and would not be economically
viable for many projects. Sediment in small quantities is often of value to
land fed by irrigation systems. In cases where the abstracted water passes
through machinery a small concentration of fine sediment is often per-
missible. But in the ease of the high standards required for domestic water Deposition of ^
silt/fine sand
supply, considerable expense is involved in removing sediment and so the
quantity allowed through the intake should be as small as possible.
intake forebay
The emphasis should be on CONTROL of the quantity of sediment which is
permissible and this guide contains advice on methods of achieving this con-
Fig. 1.) Sungai Tutong scheme - River intake and pumping station
S':Xr

At low river level, investigation of the river bed immediately in front


of the intake showed some deposits (0.3 m deep) of silt/fine sand on top
of the stone pitching on both sides of the inlets to the pump sumps.
The silt/sand deposits, which occurred directly in front of the inlets,
are likely to have been drawn into the sumps.

Erosion of the pump impellers and casings were also found to be


excessive due to the abrasive action of the sand and the low pH of the
water, fts a result, the running speed of the new pumps installed has
been restricted to 735 rpm (from 1450 rpm), to diminish this.

Fig. 1.3 Storm damage t o Hoe intake

(3) In 1973 in a paper entitled "Sediment problems of hydro-power plants".


H R Sharma stated that the cost of abrasion repairs in one power house
could be as much as $30,000 per annum.

CO The turbines of the Florida ftlta Plant in Chile (95 m head) were com-
pletely worn out after 2,000 hours of operation due to the presence of
sand in the water.
Pig. 1.2 Cruachan aqueducts - Noe main intake and tunnel portal
(5) Kuchamedon et al stated in 1975 that 50 million m^ of sediment are
(2) removed from USSR irrigation canals every year.
The Hoe Intake, part of the Cruachan catchment area in Argyll, Scotland
(Fig. 1.2) was commissioned in 1966. In 1977, a heavy thunderstorm in
Many additional examples where complete shutdown of plants is threatened
the vicinity caused approximately !,000 tons of boulders to be washed
because of intake blockages can be found in the l i t e r a t u r e .
into the intake choking it completely (Fig. 1.3). To remove the debris
a tracked shovel was driven two miles through the Hoe tunnel to the
intake.
1.4 Changes in River Regime

Records of the stage/discharge relationships of a river throughout the


seasons and over as many years as possible are necessary before any river
T
controlling or diverting (in part or in whole) structure is constructed
(Chapter 2 ) . These are essential to ensure that the required volumes and
levels are available. In some cases a weir is constructed across the river to Chapter 2
maintain and control levels - this is discussed in chapter 4 - but the design
will nonetheless depend on good stage/discharge data. DATA REQUIRED FOR DESIGN

With seasonal changes such as floods, snow melt-water, periods of


drought, etc., the river discharge can vary with a ratio of maximum to minimum
discharge of anything up to 1,000 or more, affecting sediment capacity and the
sediment properties to a similar extent. The designer must study records of
changes in river regime due to seasonal changes very carefully and should be
aware of any other diversion, control or discharge structures planned for the
stretch of river and its catchment under investigation, and take account of
all these in his design calculations. Awareness of the life stage of the
river - young, mature or old - is also important, so that trends already
experienced or likely to occur can be considered.
2. DftTft REQUIRED FOR DESIGN Topographic surveys of the river channel will be required both upstream
and downstream of the proposed intake site. The length of reach over which

2.1 Purpose and Location of a River Intake cross-sections are taken will depend on the initial assessment of the extent
of hydraulic effects of the intake, river training works and bank protection

The purpose of a scheme and the distance of the river from the point of works; particular attention Bust be given to accurate detail of the survey if

delivery of water will determine the length of the reach of the river on which hydraulic model tests are envisaged. Sections across the flood plain are

the intake can be located economically. A pumped water supply scheme is less often necessary where designs are required for rivers subject to peak floods

likely to be affected by topography than a gravity scheme in the same in excess of bank full capacity. It is essential that the surveyors should
conditions. However, matters such as the required fall through the intake may also record the level of the natural water surface, location, date and time at
be worth considering when the site for the structure is selected. least daily during the course of all survey work on the site.

The site for an intake will be chosen on the basis of the local topo- A survey of the site of the intake structure itself should be carried
graphy, geomorphology and on the geometry of the river channel close to the out at a scale of 1:100 (or 1:200 for larger structures) with a contour in-
site. The requirements of the former subjects are dealt with in Section 2.2 terval of 0.5 metres or closer; the survey should extend onto the river bed.
below. The purpose of the system and the area selected for location of the In precipitous country, larger contour intervals (say 2 m) may be adequate for
intake are the first major items of data required for design. large intakes.

2.2 Topographic and CeomorphoIoRic Data Topographic and georoorphologlc data which should be obtained, if avail-

able, would include:


The initial selection of the site for an intake structure may often be
made by study of available maps and/or aerial photographs. However, field Existing reports and river studies.
inspection and local survey will be necessary in appraising the suitability of Topographic study of catchment and site - emphasis on features such as
a particular site and in the design of training works and of the form of the landslides and glaciers which affect stability of catchment.
entry to the intake. Aerial photographic study of catchment and site.
Geological survey including records of river bed material (grading,
Study of aerial photographs both current and previous issues, or failing mineralogy, etc.).
that, inspection of the site on the ground or from the air will give some Survey of soils and vegetation in catchment - and any likely future
indication of the stability of the main river channel. Check should be made changes which may affect erosion.
of evidence of previous channels in the river bed, any tendency of the channel
to braid instead of maintain a single channel at low flows and the possibility Other features of the catchment which may influence the river morphology at
of the river changing course within its flood plain. The presence of one or the site, and therefore should be part of the survey, include:
more old river channels does not necessarily mean that such changes of course
have occurred recently and enquiry should be made of the local inhabitants as River control features (rook bars, gorges, etc.).
to the behaviour of the river. Protection of the existing river banks may be Reservoirs and lakes upstream (existing and projected) which may act as
necessary to guard against the river adopting a previous course after a major silt traps.
flood; the stability and possible protection of the bank into which the intake Trash producing industries (e.g. quarrying} and communities upstream.
is to be set must also be considered. If possible the river should be examin-
ed during floods.

10 11
'KHMfl^Mt ^^^^^^^Mft ^^^^^^^fll ^^fl^^^^M ^^^^^^^^H ^^^^^^^^^ ^^^^^^^^M ^^^^^^^^h ^^^^^^^^M ^^^^^^^^H ^^^^^^^^M ^^^^^^^tt ^^^^^^^^M ^^^^^^^^_ ^^^^^^^Mt ^^^^^^^^ta ^^^^^^^^M ^^^^^^^^b ^ ^ ^ ^ ^ ' ] ! - ^

2.3 Hydrometrie Data low weir to raise minimum water levels and, the possible incorporation of
gates in the weir to clear sediment from the forebay of the intake.
A knowledge of the flow at the site of the proposed intake is essential,
and a stage-discharge relationship must be established. 2.1 Operational Data

Essential hydrometrie data include: It is essential to establish the relationship between the discharge
abstracted at the intake and the range of flows in the river through all the
Magnitude of floods seasons of the year for effective design of an intake structure. The pro-
Flood hydrographs visions of the design are generally to abstract the specified quantity of
Frequency of floods water of the required clarity from the source, however laden with sediment,
Flow duration curve and to keep the forebay of the structure free of deposits. The ability of the
Stage-discharge curve design to achieve this depends largely on the sediment carrying capacity of
Flow-sediment content relationship the main water course ?nd the p;:pcrtion of its flow which is being
Average daily/weekly/monthly flows abstracted.
Meteorology - rainfall, wind speed, etc.
The pattern of intended operation of the intake throughout a typical
There are standardised measurement procedures available from literature year must be assessed so that the conditions at which the intake is at great-
(fief. 2.1 (a)), including recommendations about the choice of site for obtain- est risk from sediment ingress or from blockage due to shoaling in the river
ing good hydrometrie data: often, the optimum site for an intake structure is can be considered. The effects of large abstractions of water over periods of
not a good site for standard hydrometrie measurements. a few hours, as might be required where a reservoir is being supplied, must be
taken into account. The resulting patterns of low and high discharges into

It is best to establish these hydrometrie data by direct measurement the intake must be compared with the range of river flows and water levels at
using the normal methods of stream gauging, at a site close to the proposed all periods of a typical year. This will enable the designer to be aware of
intake location (so that no major tributary flows intervene). Where time is occasions of high abstraction when the water supply is most heavily sediment
not available for gauge measurements due to urgency of design, an attempt can laden and to make special arrangements to avoid or deal with these
be made to calculate a stage-discharge curve on the basis of a survey of the occasions. He will also be able to make operational arrangements for flushing
shape and gradient of the river channel, the assumed effective roughness of and cleaning operations when the river is capable of disposing of deposits.
the bed and open channel hydraulic formulae. Velocity measurements on a
surveyed cross-section will provide some check on the assumed roughness of the Account must also be taken of stages of development of the scheme which
bed. It must be recognised that this is an inadequate method of obtaining a may require lower abstractions in its early years with much higher discharges
good approximation to the stage-discharge curve at the site. planned for the same structure later in its lifetime.

An assessment of the variation of river discharges and levels from year 2.5 Sediment Data
to year (including droughts) and with the seasons of the year, and correlation
with the seasonal demands of the intake may be essential. (See Section 2.4 A full definition of sediment transport mechanisms is given in Appendix
below). Where relevant this information will establish the probability of 2, section A2.2.
being unable to divert sufficient water into the intake to meet demands at low
river flows. If this arises because of low levels rather than insufficient The bed load originates from the bed or banks of the river channel or
flow, provision of storage may have to be considered with construction of a its tributaries further upstream and the rate of transport is dependent on the
velocity and turbulence of flow at any time. Reliable measurement of bed load

12 13
ter surface levels tor various river tlows
is very difficult and it is common practice to derive an approximation by the
methods described in section 2.6 below. Some assessment of the order or
magnitude of the movement of the bed of the river is required in the choice of
elevation of a sill at the entry to the intake structure to exclude the bed
material, and in the provision to be made at any cross-river weir to deal with
deposition behind the weir. Clearly, abstraction of the near-bed flow which Level
contains these high concentrations of sediment should be avoided.

The suspended sediment may originate from the same source as the bed
load or it may be the wash load, i.e. much finer material washed into the
river by heavy rainfalls eroding the catchment. This material will be drawn
into the intake and will either remain in suspension or settle out slowly in
relatively still conditions. The concentration of suspended solids at a range
of discharges of the river must be determined; knowledge of the gradings of
these materials is also necessary to the design. Methods of sampling the
Velocity Fatal sediment concentration
suspended sediment are described in Section 2.6 below.

These samples must provide a flow-sediment content relationship (Section


2.3) for the supply to cover all seasonal variations. Fig. 2.1 Distribution of suspended sediment and velocity

The recorded measurements of sediment concentration in the river The grading of the sediment at each level can also be informative as to the
adjacent to the proposed site of the intake may be assembled to give a series
problem of l a t e r disposal of that part of the river sedUaent drawn through
of curves of sediment concentration against the reduced levels of the points
the intake. A typical preaentation of comprehensive data for a single river
of sampling. The resulting curves for various flows at the time of measure-
discharge i s illustrated in Fig. 2.2.
ment will indicate the quantities of sediment liable to enter the intake for a
proposed elevation of the intake sill. The result of such an assembly of data
is illustrated in Fig. 2.1.

I!
' 1i

[I

14 15
I Where it is necessary to measure directly the flow in the river to
calibrate the level gauge, a cross-section should be chosen where the river
reach is fairly straight with a stable channel and with streamlined flow under
most stages of river discharge. The velocity of flow is taken at a number of
depths on a vertical line on the cross-section, the depth of water at that
location being recorded, and preferably at equal horizontal intervals across
the river. The set of readings must be repeated for various water levels to
obtain a level/discharge rating curve. A propeller actuated meter is commonly
used but there are other types of meter that may be appropriate - cup meter,
electromagnetic meter, drag meter, pitot tubes, etc.. To carry out gauging
during periods of high river flow, wading through the river with a velocity
meter becomes impracticable and recourse has to be made to an anchored boat or
to the erection of an overhead cableway from which to take measurements.

2.6.2 Bed Load Data

Other than the case where a reservoir downstream is trapping the sedi-
3 t S 6 7 8|o-O4 -12 ' -20 ' -26 -36 -U. -52 -60
Velocity ft Concentration g/litre ment carried by the river, only an approximation to the actual bed load of the
river can be obtained. There are two methods of doing this. The first is to
attempt to measure directly the entry of material into a trap placed on the
Fig. 2.2 Distribution of suspended sediment and velocity for river bed, but, although a number of pieces of equipment have been used (see
Missouri river at Omaha Fig. 3.26 of Ref. 2.2), results are of uncertain reliability. The average of
many measurements is required because of a high degree of variability in bed
The constuction of an intake structure abstracting a high proportion of the load transport. The second method is to deduce the bed load as a proportion
supply will of course modify the pre-construction sediment concentration and of the suspended sediment load (see Table 3.2 of Ref. 2.2).
distribution. So it is very important to consider the effects the intake
structure and the flow abstracted by it will have on the source. The ability 2.6.3 Suspended Sediment Data
of the river to dispose of flushed or sluiced sediment and the effects that
this will have downstream must also be considered. An adequate estimate of the suspended sediment load can only be obtained
by taking samples on a regular pattern across the river and at a range of
2.6 Data Collection Techniques depths; the sampling must then be repeated for various discharges of the
river. Suspended sediment concentration may not only vary with changing river
2.6.1 Collection of Hydrometric Data discharge but also for any given discharge because of other factors. This
variability should be identified. A number of forms of sampling equipment are
The procedures and equipment required to record hydrometric data are in use which are illustrated in more detail in other publications (see Figs.
readily available in literature such as Ref. 2.1. The points given below are 3-11 to 3.-17 of Ref. 2.2) but the main points of the devices are described
applicable to most situations. below. All these sampling instruments must be positioned in the stream so
that the following requirements are met: the flow velocity at the intake to
the sampler must be representative of the velocity in the part of the cross-

16 17
section of the river at which measurements are being taken; the disturbance to (ii) Time-Integrating Point Sampler
the streamlines of flow must be minimal; the intake of the sampler must be
correctly orientated vertically and horizontally. Depending on the size of The sampler illustrated at C on Fig. 2.3 overcomes the problem of
the river, the operator may be wading, in a boat or on a cableway. fluctuations of concentration by filling over a short but significant
time interval but is representative of that point in the cross-section
only. Samples are collected at selected depths at stream verticals
representing areas of equal water discharge in the cross-section.

Wading Hand line (lii) Depth-Integrating Sampler


rod

The sampler (illustrated at A and B on Fig. 2.3) is lowered to the


stream bed and then raised to the surface at a constant rate. The
resulting sample is a discharge-weighted mean concentration for the
vertical location. The rate of raising of the device must be chosen by
D Operating f[-Cable
line- trial and error so that the sampler is not completely filled on its
return to the surface. The verticals across the river are selected as
in (ii) above; the rate of raising on a vertical must be constant but
tubing need not be the same as on other verticals.

Streamlined weight (iv) Single Stage Sampler


Cabte Sampling stage
Where a sample is required at high flows a simple unmanned type of
Bung sampler is illustrated at F on Fig. 2.3. The bottle is mounted at a
3 predetermined location and level (above the current water surface) and a
2 sample is taken slowly on the rising stage and can be collected later.
1
(v) Pumping Samplers

A refinement of the automatic sampler is a programmed device to


Fig. 2.3 Suspended sediaent sampling equipment abstract samples at pre-set time intervals or to be activated when the
water level reaches a certain stage.
(i) Instantaneous Sampler
The concentration of suspended sediment and the velocity of flow vary with the
This is the simplest type but it makes no allowance for turbulent cross-section and the sampling technique to be adopted should enable the total
fluctuations of concentration. It consists of a horizontal tube open at load to be estimated with the minimum of samples.
both ends which is lowered to the required depth and then the ends are
closed (see D and E of Fig. 2.3).

18
19
The data normally obtained from a set of sediment samples would include:

distribution of suspended sediment load across the river section


particle size and grading curves
specific gravity of particles

In addition the site data obtained would also include: Chapter 3

INTAKE LOCATION AND ALIGNMENT


temperature of the river water
discharge of the river and the distribution of velocity across the
section

Use of the sediment data gathered by the techniques outlined above is describ-
ed in Appendix 2, section A 2.9.

References

2.1 Chapters (a) A8 and (b) C2. Ij2 Applications in Hydraulics, Book 3,
Techniques in Hater Resources Investigations of the US Geological
Survey. US Govt., Printing Office, 1970.

2.2 "Sedimentation Engineering" - Editor Vito A. Vanoni - published by the


American Society of Civil Engineers (ASCE manuals and reports on engine-
ering practice - No 54), 1975.

2.3 "Drainage Basin Form and Process" - Gregory, K.J-. and Walling, D.E.
London : Edward Arnold, 1973-

20
3. INTAKE LOCATION AND ALIGNMENT The cross gradient due to the bend causes super-elevation which sets up secon-
dary flow as shown in Fig. 3.1. The spiral flow current carries the less
The choice of site for an intake 13 governed by many factors which are heavily sediment laden surface flow to the outer or concave side of the
outlined in Chapter 5. However, particularly with regard to river intakes, channel and the more heavily laden lower layer of flow together with any bed
there are useful guidelines for the location and alignment of an intake struc- material in movement, to the inner or convex side.
ture which enhance the ability to control sediment ingress to the intake.
This indicates that near a bend the outer bank is the appropriate posi-
3.1 River Bend Hydraulics tion for siting an intake where sediment ingress is to be avoided. Further,
the effect of sediment exclusion is known to be more pronounced when the
Close to its source a river is generally of less capacity and higher intake is located towards the downstream end of the bend where secondary
velocity and is less accessible than when it reaches the flatter lowland currents have become fully effective. The river should be well established
regions. It is in the forner regions that a river is most susceptible to and the banks stable in the region where the intake structure is proposed. If
surrounding geology and will wend its way via the easiest route to the sea, most of the sediment in the river is carried in suspension, this method of
lake or other river into which it discharges. In the lowland areas bends are avoiding sediment ingress is not so effective as where most movement is in the
major features of rivers. When considering a site for an intake the effects form of bed load, but it nevertheless will encourage the removal of cleaner
of bends on the velocity distribution across the river section are very impor- water.
tant.
3.2 Alignment of Intake

The intake should be aligned to the main flow to produce a suitable


Pig. 3.Ha) Secondary motion curvature of flow into the intake. This means essentially that the flow
Aft Secondary flow //
X _ - pattern f, generated In a bend to the direction should be changed as little as possible. In this way the flow will
I Looking downstream) right behave as though the intake was on the outside of a curve and so bed load will
be swept away from the vicinity. If, on the other hand, the flow is diverted
by a large angle the flow patterns will be disturbed and bed load will be
attracted towards the intake.

References reviewed in 1981 (Ref. 3.1) recommended diversion angles of


between 10 and 45, i.e. the angle between the intake centre line and the
main direction of river flow. The ideal angle for a particular intake depends
on the ratio of abstracted to river flow, the widths of the river and intake
velocity bed flow forebay and other factors. It is fair to say that a diversion angle of less
than 15 is favoured and the optimum angle could be achieved by observing flow
High velocity surface flow patterns in an hydraulic model.

3.3 Hydraulics of 90 Diversion Angle


Fig. 3.Hb) Erosion of the outer bank of a bend caused by the faster moving
component of the stream, and deposition of silt on the inside where the slow It is repeatedly found that despite publications with recommendations to
moving bottom water rises the contrary, intakes are often constructed in river banks with the intake
axis perpendicular to the river bank. This means that the flow has to be
diverted through 90 to enter the intake structure. The associated problems
of flow separation and recirculation and sediment deposition often occur, if
no attempt is made to improve the curvature of the flow approaching the intake
these problems are insoluble. Methods of improving the flow curvature are
included in later chapters. The drawing in Fig. 3.2 illustrates how the near-
bed flow is attracted towards a 90 intake. The arrows indicate the direction
Chapter 4
of the near bed flow and S 1 and S 2 the stagnation points. Guide vanes or RANGE OF INTAKE STRUCTURES
walls can improve the flow curvature into the intake.

Fig. 3.2 90 Offtake

Examples of problems caused by bad siting of an intake often appear in l i t e r a -


t u r e . There are instances where complete re-location or re-construction of
the intake has been carried out. The location and alignment of the intake are
the e a r l i e s t most important considerations in i t s design.

References

3.1 ftvery, P. Sediment Exclusion at Intakes - A Review, BHRfl Report


RR 1725, August 1981.

24
n. RANGE OF INTAKE STRUCTURES Fig. 1.1 shows a typical river bank intake for a water supply pumping station,
where sediment transport in the river is normally not significant. The face
1.1 Types of Intake Structure of the intake is aligned with the bank. The Intakes, which are at bed level,
have coarse screens, bulkhead gates and fish electrodes. Behind the coarse
The prime purpose of an intake is to allow abstraction of water from the screens there are settling chambers to trap coarse material that might enter,
source with as little sediment as possible, thereby minimising maintenance and and band screens and finally the pump chambers. The maximum abstraction rate
operational costs, and providing some measure of protection against damage to, is 350 Ml/day. The bed material in the river was composed of approximately
or blocking of, the conduit by incoming sediment, ft number of distinct types five equal fractions of the following size ranges: 175-75 mm, 75-33 mm, 33-
of intake structure have been developed, the selection of which is likely to 17 mm, 17-5 mm and below 5 mm. A gravel trap across the river just upstream
depend on the location, scale and function of the project. of the intake was found through model studies to be efficient in protecting
the intake from the ingress of bed load.
Intakes can be from a water course, reservoir or sea. The main emphasis
in this guide is on water course intakes where sediment control is often a 1.1.2 Side Intakes with Cross Heirs
major design constraint.
For rivers and streams where a substantial proportion of the flow is to
Typical features of intake structures are described in Appendix 1. 1 be diverted, a cross weir of some sort is an essential feature to ensure
available water is not lost to the intake at low stages.
4.1.1 Bank Intakes
Fig. 4.2 shows a layout for a side intake with cross weir, supplying a
These are structures located on a river or canal bank, side of reservoir free flowing pipeline. The arrangements work well provided solid and floating
or a coastal site. They are generally adopted for locations where only a debris are not present in large quantities, and if screen cleaning can be
small portion of the flow passing the intake is to be abstracted and where undertaken regularly. The average stream flow is 50 Is and the design
fluctuations in water level are not large. Bank intakes are appropriate for intake flow (typically five times the average) is 250 is . There is about
irrigation, water supply and power functions. 10 nP of sediment storage available in the head pond, which would require
L
Bulkhead gate and shovelling to clear, this in turn may require suitable access arrangements.
screen guides Sediment sizes encountered are in the range of coarse sand to gravel plus a
few cobbles.
Band screen

Sectional plan
Fig. River bank intake structure

26 27
A modified version of the side intake with cross weir is shown in Fig. 4.3.
The screens have been eliminated, and a skimmer wall excludes floating
debris. Typical flow rates here are an average stream flow of 110 Is and
-1
design intake of 1000 Is The sediment sizes are in the coarse sand and
gravel range.

Skimmer wall Internal weir


Burn Spillweir
A Shut-off
penstock

Section A-A

Spill weir.
Scour
penstock

Section 8-B
Penstock Plan of No.2 Intake Pipe from intake

Plan

Section C-C

Original ground Screen Original ground


line ' line

Fig. 4.3 Screenless side intake

Where an heavy sediment load is carried along in suspension or near the river
bed additional arrangements may be necessary to minimise sediment influx and
Penstock
to avoid blockage of the intake by shoaling in the river. Fig. 4.U shows such
Elevation A-A
an arrangement, typically for an irrigation or power canal intake
(Ref. 4.1). The function of the main features of the intake are described in
Fig. 1.2 Side intake with cross weir Appendix 1. In this case the desilting canal is used for intermittent return
of desilting flow to the river.

29
H.1.3 Low Head River or Canal Diversion Works

Barrage
fls previously mentioned weirs
Diversion weir are constructed across rivers
Head regulator at intakes to maintain
Sand sluice
adequate levels and volumes
Desilting canal
of water for abstraction.
For irrigation systems barr-
ages are often constructed
across main water courses at
the headworks of new
canals. The barrage and
Splitter piers associated canal headworks
are designed to divert water
from the main water course,
Fig. 4.5 Typical canal diversion works to control sediment entry
into the irrigation canal
Skimmer wall Gate
system and to control the distribution of water within the canal system.

Inlet s i l -
7 -To supply
Fig. J|.5 shows a typical canal diversion works.

Some situations may require continuous sluicing of bed load, and a successful
Sluice
arrangement is shown in Fig. 4.6 (Ref. 4.1). Here the intake to a power canal
has a series of undersluices set in the face of the intake sill, which extract
Section A-A
the lowest layer of flow with the highest concentration and coarsest sediment,
and return it continuously to downstream of the river control structure.

Further examples of the methods used to control sediment entry to canals


at diversion works are described in Chapter 6.
Fig. 4.4 Side intake on river with heavy bed load
1.1.4 Bottom Intakes

Bottom intakes have been developed for glaciers and mountain torrents,
where site conditions may be extremely difficult for access and construction,
and where boulders and rock debris have to be passed with minimum obstruction.

Fig. 4.7 shows the arrangement for a 'Tyrolean' type of bottom intake
for a hydro power scheme in the French Alps (Ref. 4.2). It comprises a
collecting chamber across the bed of the stream covered by a coarse screen.
The total stream flow passes over the chamber, and the screen admits fine
^^^^^^^^^^^^ ^^^^^^^^^^^fc ^^^^^^^^^^^^ ^^^^^^^^^^^M ^^^^^^^^^^^_ ^^^^^^^^^^^ ^^^^^^^^^^^ta ^^^^^^^^^^^M- ^^^^^^^^^^^ta ^^^^^^^^^^^^ ^^^^^^^^^^^H ^^^^^^^^^^^ .^^^^^^^^^^^B ^^^^^^^^^^^h ^^^^^^^^^^^A ^^^^^^^^^^^^ ^^^^^^^^^^^^ ^^^^B^^^^^^h ^^^^^^^^^^^^ ^^^^^^^^^^^^ \

River tsar
Barrage Screen

im sill

Cross-section

Plan

Rushing Sluice gate


sluice

Rushing canal
Section A-A To
settling
basin

Sectton A-A

Fig. 4.6 Side intake with unders lit Ices

debris with the water entering the chamber. Excess inflow i s s p i l l e d at the Fig. 1.7 Tyrolean intake
downstream s i l l . The conduit from the c o l l e c t i n g chamber is designed f o r
debris which has entered to be carried with the floy to a s e t t l i n g basin Smaller versions of bottom intake are shown in Figs. 4.8 and 4.9. These
constructed a short distance downstream. Here clean water is skimmed o f f f o r designs were developed for diversion of side streams to hydro power reser-
the power scheme, and arrangements made f o r periodic scouring of settled voirs. Steeply sloping screens with round bars were found beneficial in
debris. Typical intercepted discharge is 3 m3 s " 1 . minimising blockage and loss of water. Light weight debris may be carried
through the aqueduct to the reservoir; heavier material trapped in the screen
chamber, is scoured by a hand operated valve. The maximum design intake flow
of these examples are 150 Is" 1 and 750 Is" 1 respectively.

33
wl
(.1.5 Frontal Intakes

Screens
Pipe

Winter Intoki
Scour

Section D-D

Fig. 1.8 Pipeline bottom intake

Fig. 4.10 Design of frontal intake (fro* Cecen - see Ref. 3.1)

A frontal intake designed for abstraction of clear water froa mountain streams
13 shoun in Fig. 4.10. This design has been used successfully in eajor
systeas aainly in Turkey. Since the abstracted water is taken from the upper
layers of the river while the lower layer is continually flushed past the
intake, i t i s particularly applicable where the majority of sediment carried
by the river i s bed load, and where a large proportion of the flow continues
down the original water course.

Fig. 1.9 Photograph of bottom intake to tunnel


BBB BIB flH flHB HH B^B B^B ^^B B^B B^B ^^B B^B ^^B ^^B ^^B ^^H ^^M ^ ^ H

4.1.6 Submerged Intakes

Bar screen(12mmx 50mm steel bors spaced every 150mm)

Flexible joint
Position of
Concrete block anti-vortex assembly

Stone pitching

Fig. 4.11 Submerged intake

A simple submerged intake is shown in Fig. 4.11 (Ref. 4.3). It comprises a


bellmouth and bend set in a block of concrete on the bed of a river or reser-
voir, with connecting draw-off pipe. The inlet is protected with a bar screen
and set high enough above the bed to allow sediments to pass on either side of
it if used in a river. Such an arrangement may be used for the drainage
outlet of a small reservoir.

A more elaborate concept with vertical shaft, for larger flows is shown
in Fig. 4.12. A framed structure extends to above top water level, allowing
access for maintenance of the screens and shaft. The structure was planned as
a cooling water intake in a shallow estuary. Here the screen sills must be
set above the highest level to which estuarial sediments might be expected to
rise. Each of two intakes here provides 36 ra's"1 of cooling water. Fig. 4.12 Submerged shaft intake

Both intakes (Figs. 4.11 and 4.12) are located nearly at bed level and allow a
maximum range of water levels to be utilised.

36
37
4.1.7 Tower Intakes 4.2 Choice of Intake

These are generally utilised where there is a large water level varia- In selecting a particular type of intake where there is a requirement
tion such as in storage reservoirs or tidal waters, and access for operation for the exclusion of sediment, factors to be considered should include the
of gates or valves is essential. They are free standing structures, set in function of the intake, the scale of the works In terms of flow quantity and
deep water, usually with an access gantry above top water level connecting to range of depth, particular features of the site and method of construction.
the shore.
Principal functions for intakes include water supply, power generation,
Some tower intakes are "wet" in that water fills the inside of the tower cooling water, irrigation, stream diversion and drainage. The types of in-
to approximately the same level as outside, under normal operating condi- takes shown in Figs. 4.1 to 4.13 indicate generally the function appropriate
tions. The towers are dry only when the gates are closed and the interior de- to the type of Intake, and some of the variations that may be adopted for the
watered. A "dry intake tower" is illustrated in Fig. 4.13. The tower houses exclusion of sediment. Chapter 6 outlines further devices which may be in-
a vertical withdrawal conduit arranged with Individually gated branch inlets, corporated into intake designs to exclude and control sediment.
at several different levels, allowing draw-off from any selected level. The
bottom outlet valve is designed to scour out any sedinent collecting adjacent The data collected prior to site selection (as described in Chapter 2)
to the intake. and the sizing of the intake (Chapter 5) will provide limitations to the
choice of intake structure type. Knowledge of the nature of the river, Its
range of levels, discharges and sediment load, will indicate whether the
intake should be a bank or bed intake for example. The required operating
TWL criteria of the intake structure and associated distribution works provide
further restrictions to choice. Whether or not there will be sufficient water
to spare for flushing and sluicing operations will affect the choice of
methods for dealing with excess sediment.

Where an intake is to be constructed for a large flow, such as for power


generation or cooling water, the intake location and hence its type, is most
likely to be determined by the overall layout adopted for the hydraulic
systen. In shallow water such as an estuary, a submerged vertical intake may
be the only possible choice, and In such a situation, a particularly careful
investigation of the sedimentation of the area would be an essential pre-
requisite.

A new intake site may be on a stream, river, canal, lake, new or exist-
ing reservoir, or coastal location, ftccess to the site, and whether con-
struction is to be in the dry, in cofferdam or through water, will all be
considerations in assessing the most appropriate type of intake.

In all cases model studies provide valuable information on the per-


Fig. 4.13 Dry intake tower for water supply formance and suitability of the intake structure design.

39
References

Jf.1 Mosonyi, E. Hater Power Development, Vol. 1, Hungarian Academy of


Sciences, Budapest, 1957.
1""
4.2 Ract-Madouse, X. et a l . Some recently built high altitude bottom in-
takes, La Houille Blanche, N. 6, Dec. 1955, p.853. Chapter 5

4.3 Linsley, R.K. and Franzini, J.B. Water-Resources Engineering, 3rd DESIGN OF INTAKES
Edition, McGraw-Hill Kogakusha Ltd., Tokyo 1979.

40
5. DESIQ1 OF INTftKES (iii) The concentration of sediment at these flows and the reduction in con-
centration as the level of the entry sill of the offtake is raised above
This chapter incorporates general considerations for intake design in the bed {see Section 2.5 and Figs. 2.2 and 2.3).
addition to those specific to sedioent control.
The effects that the abstraction will have on the sediment concentrations and
5.1 Design Procedure distribution must also be considered.

5.1.1 Requirements 5.1.3 Location of Intake

The basic requirements for the supply of water must be defined: The considerations limiting the location and alignment of the intake are
described in Chapter 3 above.
(i) Maximum and minimum abstraction rates during the different seasons of
the year. 5.1.1 Type of Intake

(ii) Maximum sediment content and the nsaximum grain size of particles that After consideration of the requirements of the intake and of the limita-
can be accepted into the supply system. tions of the supply, a choice of the type of intake best suited to the par-
ticular location must be made. Descriptions of the various forms of intake
(iii) The acceptability of partial or total failure of the supply and the and the features of each type are given in Chapter t above.
corresponding return period in years.
5.1.5 Dimensions of Main Structure
(iv) The location and relative level of the point of delivery of water from
the intake must be defined so that a choice is made between gravity and The size of the structure is defined by the combinations of maximum
pumped delivery and also between open channel and pressure pipe con- demand and minimum supply water level at each season of the year. The area of
veyance of the water. the entrance to the structure through which the supply is passed is sized to
give an acceptable value of maximum velocity - say 1-2 m/s in a gated opening
(v) Provision for future expansion of the demand. with open channel flow; maximum velocities for iined power tunnels and gates
in these tunnels would be higher and restricted by head loss, lining material,
5.1.2 Limitations roughness, etc.. Where the openings are not gated, where it is intended to
install screens, where it is desirable to avoid attracting fish or where
The above requirements must then be compared with the limitations of the turbulence is to be avoided {e.g. pump intakes) appropriate maximum velocities
source of water: will be adopted.

(i) Maximum and minimum water levels in the river or reservoir at the The level of the sill at the entrance to a free surface conduit can be
accepted probability of restricted supply {see 5.1.1 (iii) and (iv) set lower than that of the entry sill (see 5.1.6 below} but consideration
above). should be given to lowering the bed of the forebay so that heavier particles
carried over the entry sill will be dropped before the water enters the main
(ii) The corresponding maximum and minimum flows available. intake structure. In general the sill of the openings will define the level
of the floor of the structure except where pumps are to be installed; in the
latter case the floor inside the structure must be lowered to ensure that

42
sufficient submergence for the pumps is provided under all operating con- 5.1.7 Arrangement and Orientation
ditions. Entrances to submerged power tunnels may have to be sufficiently
deep to avoid vortex formation. The entry sill should be orientated parallel to the flow of the river to
discourage deposition of bed load against the face of the sill. Where a weir
The sizing of the intake to a tunnel or pipe system will be restricted or barrage has been constructed across the river to raise water levels it is
by acceptable velocities in the culverts. These velocities are associated often necessary to install gates in the weir immediately downstream of the
with screen area or velocity through gate openings. intake to act as scour sluices and physically prevent build up of sediment in
front of the intake (see Fig. 4.4). In such a case river training works, in
Detailed hydraulic calculations will be required to confirm the per- the form of divide walls or groynes, across the front of the intake may be
formance of any tentative intake design, and to ensure that it has an adequate constructed to constrict the channel to the scour sluices and thus generate
capacity. velocities locally exceeding those of the river prior to construction of the
weir.
5.1.6 Entry Sill

Although the entry weir is set parallel to the direction of flow of the
An entry sill is usually provided upstream of the intake structure in river the axis of the forebay and the main intake structure should be set at a
order to prevent as much as practicable of the transported sediment in the smaller angle to the river (as described in Chapter 3 above).
river from reaching the intake. The level of the entry sill is critical. It
should be set at such a height above the bed of the river that the bed load Where the intake takes the form of a tower in a reservoir the entry sill
and the higher concentration of the suspended sediment are excluded at high usually will be that of the intake gates. The lowest gate should be set clear
river flows (see Figs. 2.1 and 2.2 in Chapter 2 ) : but at the same time the of the anticipated deposition of sediment in the bottom of the reservoir, but
crest of the sill must be low enough to enable the water demand to discharge Delow the normal operating drawdown level. If the operating range of the
over the sill at low river levels. In the latter circumstances the sill
reservoir is considerable it is normal practice in water suppy systems (but
should be of sufficient length to keep the velocity to the required value.
not hydro-electric) to introduce alternative intake gates at vertical inter-
vals of 10 to 20 metres providing adequate submergence so that water clear of
If these requirements conflict then a decision must be made:- both sediment and floating debris can be selected.

(i) Accept a lower intake supply at low river levels, Where appreciable quantities of sediment are drawn over the entry sill
but retained by a downstream sill, provision can be made for installation of
or (ii) accept a higher sediment intake at high river flow, scour sluices and a discharge channel from the forebay to control build up of
sediment.
or (iii) consider river training works to lower the river bed of the
channel local to the intake, 5.2 Geometric Recommendations

or (iv) construct cross-river works to raise the river water level at The boundaries of flow formed by the walls and floor of the approach to
low flows, the ma lit intake structure should be aligned with the intention of preventing
turbulence and also of preventing separation of the streamlines from the
or (v) a combination of the above choices. boundary surfaces and inducing reverse eddies. The use of curved walls and
nosings to piers will reduce turbulent effects. The prevention of separation
of streamlines from boundaries is more of a problem. A rule of thumb used for

44 45
canals is that the radius of curvature of the axis of flow should be at least screens, usually of 50 - 150 mm bar spacing, which may be mounted on the entry
two and a half times the width of the water surface of the channel. In prac- sill if the river velocity is sufficient to assist in self cleaning. If the
tice such a large radius can seldom be provided at the approach to an intake debris has to be cleared manually then the screens are usually installed at
and the wall at the inside of the bend will be curved to suit the space avail- the openings to the main intake structure where access is easier and safer.
able. A nosing of radius 0.2 times the channel width will avoid the worst Skinner walls used with scour sluices as indicated in Fig. 5.1 have been used
effects of flow separation. successfully. The sluice will have a gradient of about 1 in 3 with a gate on
its downstream face to prevent the guides and spindle from choking. A radial
To ensure even distribution of water across the face of the intake gate for flushing should have its trunnion below the centre of curvature of
structure, floor baffles, often set at an angle to the flow, or vertical the gate face to ensure a clean and rapid break from the silt packed against
columns may be necessary. Even distribution of the water is of greatest it. The trash sill if deeper could incorporate a crest gate for small
significance where the intake contains pumps, or where the intake is set flushes. Max. design
flow httao
directly across a supply canal much narrower than the structure. Re-entrants
Skimmer wall
and areas where slow moving water will cause the deposition of sediment should
be avoided.

Some geometric recommendations are illustrated in Chapters 1 and 6.

For a large intake structure there is no substitute for hydraulic model


tests to ensure satisfactory hydraulic performance, and the study of flow Scour sluice
patterns which influence particle movements should be emphasised.
Control sluice
gate
5.3 Floating Debris
Upstreom elevation A-A

Attention must be given to prevention of floating material from reaching


the intake, and to disposal or dispersion of debris trapped against screens or
baffles at the intake. Build up of debris, once started, proceeds at an
Supply
accelerating rate and can throttle the water supply unless timely steps are Scour sluice gate
taken to obviate the threat.

In reservoirs where water level ranges are large but velocities are low,
a floating boom is a possible remedy. Such devices are often impracticable in
rivers liable to high flows as they can be swept away during floods. They are
also vulnerable to damage from logs or ice runs.

In rivers it is more common to attempt to deflect the debris at the


entry sill to the intake. Where there is sufficient depth of water a deep
skimmer wall mounted on splitter piers provides a submerged orifice for entry
of water at all flows and forms a wall of sufficient height to prevent over-
topping at high flows. In other rivers it is often necessary to resort to Fig. 5.1 Intake layout using baffle

46 47
The amount of debris liable to become trapped against the screens must (ii) artificial creation of turbulence downstream of the intake to attract
be assessed. Regular manual raking - restricted to shallow intakes, say 5 m fish there rather than to the intake itself. An extension of this
deep from deck level - may be sufficient, but large quantities of matted system is the use of vertical louvres leading past the intake to a fish
vegetation may require the use of powered mechanical rakes which may be bypass at the downstream end. For this to be effective the velocity of
arranged to operate automatically, by a timed or level sensing device. If approach to the louvres needs to be approximately in the range 0.7 to
virtually continuous accumulation of debris is anticipated then consideration 1.2 m/s.
should be given to drum screens, if the range of water levels is not too wide,
or to band screens where the range is greater. Both these devices can be made (iii) use of electrodes suspended in the water above the entry sill and a
self cleansing but require manual removal of the accumulating material. ground conductor on the sill to form an electric screen. If the
Disposal of trash should take into consideration other users downstream. velocities are high in the region of the intake the stunned fish may
still be drawn towards the intake.
In cold climates ice presents special problems. Frazil ice may accumu-
late on screens and completely block them. Also, the build up of thick sheet The velocities and mesh sizes given above are only approximations and it is
ice can exert considerable pressures on the vertical faces of structures, and strongly advised that designers consult the fisheries authorities who will be
the usual remedy is to construct the nails with sloping faces so that the ice associated with the intake structure and associated waterways.
at the edge of the river is forced upwards, fit time of thaw the presence of
heavy ice flows in the river can pose a threat of considerable physical damage
by impact; in such cases the geometric arrangement of the works must take
account of the probable directions of movement of large pieces of ice in order
to reduce the risk of direct collision.

5.1 Pish

In a number of countries both commercial and sporting interests are


sensitive to the possible destruction of fish as a result of them being drawn
into water intake works. A number of methods have been tried to prevent entry
of fish and these include:

(i) use of fine mesh screens, in addition to the coarse screens which pre-
vent entry of debris and protect the fine screens. These are very
liable to blockage by the smaller debris and leaves so a cleaning gantry
may be required. The velocity through the screens must be kept low so
that fish are not trapped against the mesh. A maximum velocity of about
0.5 m/s is usually adopted. The mesh, where young fish have to be
preserved, can be as fine as 4 ran spacing.

Where the fish screens are remote from the river channel it is
important to avoid fish becoming trapped at the screens by ensuring that
a route exists for them to return to the main river channel.

49
Chapter 6

SEDIMENT EXCLUDING DEVICES


6. SEDIMENT EXCLUDING DEVICES There are two main types of curved channel sediment excluder: Type 1 in which
the flow from the curved channel into the canal is through a relatively short,
6.1 Curved Channel Sediment Excluder gate-controlled outlet, as in Fig. 6.1.1, and Type 2 in which the flow from
the curved channel into the canal is over a relatively long skiraning weir, as
This device uses the established practice of relying on the favourable in Fig. 6.1.2.
effects of the curvature of a channel to reduce the amount of sediment enter-
ing the headworks of canal systems. The secondary currents in a curved
channel which cause a large proportion of bed load to move towards the convex Sluice flow
bank or wall of the channel are described in Chapter 3.
Canal flow
This classification is intended to include those sediment excluders for Sluice channel
canal intakes where a separate curved channel is constructed from the river or
main water course to the canal headworks which conveys the canal flow plus a 3-5m sluice gates
sluicing flow. So it has the limitation of requiring sufficient quantites of
water to allow some to be 'wasted' for sluicing purposes. This description is
not intended to cover those cases in which the curvature of the whole river or
main water course is enhanced by training works of one kind or another.

Skimming weir

Osborne canal
\
[Xl Crest of diversion weir

"Left bank guide woU

[Prototype dimensions)

Upstream :
F i g . 6.1.2 20 n 3 /s Headworks
Curved sluice channel

The principle of the curved channel excluder is that sediment close to the bed
moves to the inside or convex wall of the channel, the outlet to the canal
being on the outer or concave wall of the channel. The amount of sediment
entering the canal is thereby reduced. Some simple points must be taken into
Fig. 6.1.1 Separate curved sluice channel for sediment exclusion account: (a) the approach to the curved channel must be such that the desired
(Osborne Canal)
helicoidal flow pattern does develop in the channel, (b) the canal entry (c) The bed level of the curved channel will probably be determined by the
arrangements do not cause sediment to be thrown up into suspension, and river bed level at the sluice gates; or possibly by the level at which
(o) velocities in the curved channel maintain the forward movement of sediment it is considered that sediment will be swept away to avoid blockage of
towards the sluice gates. the outlet of the sluice channel.

Since the flow in the curved channel decreases as It approaches the (d) The sluice flow is usually fixed as a proportion of the canal flow,
sluice gates, it is necessary to reduce the width of the channel in that depending on sediment size and concentration. Values of around 30J and
direction to maintain the velocity so that sediment is neither deposited on 50$ are typical.
nor entrained from the bed. The converging channel is shown in Fig. 6.1.2. A
converging curved channel has some useful characteristics; (a) provided the (e) Some guidance about plan geometry can be obtained from Table 1 (from
curvature is not so severe that separation occurs, the flow is essentially Ref. 6.1.1) which gives the leading characteristics of five existing
irrotational, i.e. velocity is inversely proportional to the radius of curva- curved channel sediment excluders. It is suggested that the ratio of
ture and (b) the line of maximum velocity shifts towards the convex bank where centre-line radius to the width of the curved channel at its mid-length
it may help in conveying concentrations of sediment towards the sluice gates. should be between 3:1 and 5:1.

Factors affecting design are suranarised below: (f) It has been suggested that to achieve reasonable performance the Froude
number in the curved channel should be in the range 0.5 to 0.8 at the
(a) The Intake location should be in accordance with Chapter 3 of this upstream entrance to the canal head regulator.
Guide. Obviously the intake should be located favourably with respect
to river curvature and it is essential that the river or main channel (g) The conditions downstream of the sluice channel in the main water course
water enters the curved channel in a streamlined and controlled way. must be such that blocking of the channel Is avoided at all times.
This division of flow is as important as that between the curved channel
and the canal and it is essential to ensure that a disproportionately The properties of the typical sediment will be an important consideration. At
high sediment concentration is not attracted to the Intake works. this point some consideration must be given to criteria determining the
velocity of the flow in the curved channel. Obviously if a flow equal to
The total flow entering the intake structure will be of the order something like 30J to 50J of the canal flow is to continue through to the
of 1.3 to 1.5 times the specified canal flow, Q o . The width of the sluice gates, then quite high velocities are to be expected in the curved
channel at the entrance to the intake structure should relate to the channel. Possible criteria to consider are:
width of a channel in the main water course which would convey Q . The
Regime method, which relates to channels with mobile boundaries, pre- (a) Since it is intended that the curved channel does not accumulate sedi-
dicts a required width (in metres) of 5/Q~ <Q C in m^/s). it is ment and since its floor is a plane surface, it could be supposed that
suggested that from this width the curved channel is narrowed in a the curved channel velocity would be satisfactory if the shear stress
streamlined way until it reaches the dimensions required for the sluice imposed on the sediment particles by the flow exceeded the critical
channel. shear stress, for example as determined by Sheild's criteria (Appendix
2, section A 2.3). An example of such procedure is given later. How-
(b) The water level in the curved channel will be determined by other civil ever, calculations on some existing curved channels indicate that the
works requirements. If the natural river levels are not suitable, it imposed shear stress is several times the critical shear stress, provid-
may be necessary to construct a structure across the river to achieve ing a reasonable safety factor. The ratio of imposed shear stress to
the necessary water levels in the curved channel.

55
critical shear stress may still remain a useful index of performance. )
_ 5 6.1.2
Some recent experimental work suggests that a value of i, near the >

entrance to the curved channel of about 0.1 may be appropriate.


Then from Fig. A2.2 the dimensionless c r i t i c a l shear s t r e s s T,,, can be
read off where:
(b) The reality of the situation is that a sufficient velocity is required
to ensure that a fairly high concentration of sediment is carried along
6.1.3
in the flow. For example, if the concentration of sediment in the flow
at the entrance to the headworks is 500 ppm, the concentration adjacent
to the sluice gates could be 1000 ppm. The sediment carrying capacity
(since v* - A Q g/Y )
of the flow can be checked by sediment transport formulae, or alter-
natively as in the example by a semi-empirical relationship which deter-
Knowing the numerical value of the left-hand side of equation 6.1.3, the
mines sediment transport as a function of major parameters of the flow
value of the shear velocity v* can be calculated.
and sediment particles.

Finally the mean velocity corresponding to critical shear stress


The velocity corresponding to critical shear stress t,) on the sediment
can be calculated from an equation such as:
particles can be determined as follows:-
V = 2.5 vM In (- 6.1.4
(a) The average velocity, V r , in the vertical at any radius r is assumed to
follow the relationship V r proportional to 1/r. This will result in a
where a, the effective roughness of the bed is usually taken as the 65%
logarithmic distribution of the average velocities in the verticals in a
size of the bed material but is subject to a correction factor if the
section across the curved channel. If Q is the discharge at a par-
bed is not completely rough.
ticular section in the curved channel, then

(c) The velocity determined using equation 6.1.4 can be compared with the
6.1.1
f = V , in (ro/r.) actual velocity in the curved channel.

where d is flow depth, V^ is the average velocity in a vertical close to


(d) Alernatively the value of T S calculated from equation 6.1.3 can be
the inner wall of curvature r^ and r 0 is the radius of curvature of the
compared with possible necessary values as suggested previously. Assum-
outer, or concave, wall. From equation 6.1.1 the value of V^ can be
ing that a preliminary design has been completed, assisted by the proce-
determined, and from the proportionality relationship the average
dures outlined above, it would be advisable to carry out a model study
velocity in any vertical can then be determined.
of the excluder. This is because the regime of flows in the river, and
the variability of the hydrograph, are very significant factors in the
(b) It is now possible to check the average velocity required to maintain
performance of a sediment excluder. In some cases it is possible that
critical shear stress on the sediment particles using the Shields'
for extended periods, flow is run into the canal headworks even though
relationship (Appendix 2, section A 2-3) and a logarithmic mean velocity
little water Is available for sluicing. Simulation of the river flows
equation for flow in the channel. The most practicable method of using
and the operation of the sluice gates will give some insight into sedi-
the Shields' criterion is probably that given in the ASCE Sedimentation
ment exclusion performance under these conditions. Performance under
Manual (Ref. 2.2), the relevant graphical relationship being reproduced
flood conditions can also be investigated.
in Fig. H2.2 (of Appendix 2). Firstly, it is necessary to calculate a
dimensionless grain parameter I, where I is defined below

57
The method used in the example in Appendix 3, for determining sediment con- Vortex tubes are most appropriate where substantial bed load is to be

centration i s the graphical relationship developed by Colby (Ref. 6.1.2} as excluded and have limitations with respect to sediment distribution and

reproduced in Fig. k2.k. suspended sediment. However, appreciable trapping efficiencies have been
recorded at sites where suspended local is predominant.

Table 1 - Leading characteristics of some curved channel sediment excluders


6.2.1 Design of the Vortex Tube

Headworks Canal Sluice Sluice Centreline Average Angle A considerable amount of research on the theory and performance of
flow flow width radius dept to intake vortex tubes has been carried out by the Hydraulics Research Station,
m-Vs m 3 /s m m m centreline
Wallingford. This was based initially on the analysis of previous research
Courtland 11.3 5.66 6.0 - and on hydraulic models of vortex tubes. This work was later confirmed and
Superior 2.26 1.13 7.0 - refined by field measurements on prototype vortex tubes constructed recently
Republic 3.39 1.69 - in Indonesia and Nepal. The measured trapping efficiencies at these
Bartley 1.70 1.06 3.6 11.0 1.73 60 extractors were about 76% and 50% respectively, with median bed material sizes
Woodston 1.19 0.99 1.5 7.6 2.59 90 of 0.38 mm and 0.2-ram.

6.2 Vortex Tube Sediaent Extractor For the hydraulic design of the vortex tube it is recommended that Refs.
6.2.1 and 6.2.2 should be used to calculate the optimum tube dimensions; and
A vortex tube sediment extractor is a device for the continuous removal the method given in Ref. 6.2.3 should be used to predict the trapping
of sediment moving near the bed of a channel. It consists of a horizontal efficiency.
tube or duct installed normal to and below the bed of the channel, which
extracts a small proportion of the flow near the bed, where generally there is
The design objective is to remove enough sediment from the canal, such
a higher concentration of bed material load. A horizontal axis vortex is
that, the remainder does not exceed the sediment transport capacity In the
generated in the tube and the flow and sediment are conveyed laterally to a
downstream canals. The design method involves the assembly of canal and
settling basin or discharge channel. The vortex tube can be located in the
sediment data and the selection of trial tube dimensions and limits on accep-
approach to the canal headworks or sufficiently far downstream of the head-
table extraction ratios (vortex tube discharge to approach canal discharge}.
works to ensure that the equilibrium of the sediment distribution in the canal
This Is followed by trial hydraulic calculations from which acceptable com-
is established.
binations of tube dimensions, operating head and extraction ratio can be
selected. The trapping efficiencies are then computed to determine the
The comments in the section apply to vortex tubes located in canals optimum design, capable of removing the required amount of sediment from the
downstream of the headworks, where the flow is subjected to more controlled canal to satisfy the downstream canal sediment transport capacities.
conditions and vortex tube performance is more predictable than upstream of
the headworks. A possible disadvantage is that the intake has to be sized for
abstracted flow plus waste flow.

The flow enters the vortex tube tangentially and generates a forced
vortex along the axis of the tube. The flow through the tube is controlled by
a gate at the downstream end where it discharges into a disposal channel. A
typical example of a vortex tube installation is shown in Fig. 6.2.1.

59
.Canal drain culvert gate
Crump weir of increasing the Froude number, by raising the vortex tube, is that, although
Flow it does not affect the overall available head, it reduces the possibility of
backing up in the vortex tube disposal channel, at times of high downstream
Vortex tubes river levels.
Section A-A
6.2.3 Location of Vortex Tube
Chandra canal Vortex tube silt extractor gote
Ideally, the location of the vortex tube should be as near the headworks
as possible to optimise the removal of the sediment. To ensure reasonable
equilibrium of sediment distribution in the canal profile, it is recommended
in fief. 6.2.1 that the distance (in metres) downstream of the headwork should
not be less than: 10/Fr x Vn/Vs50.

where:- Fr - Froude number in approach channel


Vn - normal channel velocity at vortex tube <m/s)
Vs50 - Settling velocity of median size sediment (m/s)

In many cases where there is insufficient head for the disposal of trapped
sediment it may be necessary to move the vortex tube further downstream.

6.2.4 Disposal of Trapped Sediment


Vortex tube

In common with most sediment removal devices, the main problem encoun-
tered with vortex tube sediment extractors is the difficulty in the disposal
of the trapped sediment. For the disposal of the sediment back to the river
or escape channel, by gravity, it is necessary to provide for the following
head losses:-

- difference in head between canal full supply level and minimum canal
supply level.
operating head for vortex tube.
Pig. 6.2.1 Vortex tubes
head losses through the downstream of vortex tube control gate,
head loss in disposal channel.
6.2.2 Level Setting for Vortex Tube

These losses should not exceed the difference between the full supply level
The level at which the tube is set in relation to the bed of the canal,
in the canal at the vortex tube and the maximum river level at the disposal
effects the Froude number in the canal at the vortex tube. Generally for
channel outfall, when the canal is operating. The disposal channel should be
finer sediment low Froude numbers give higher trapping efficiencies, while
as short as possible and the outfall located at a point in the river, where
with very coarse sediment the Froude number can be increased. One advantage
there is a high concentration of flow; for example on the outside of a bend.
Backing up of the flow within the disposal channel, should be avoided.

60 61
Because of high sediment concentrations, the ideal solution for the PRINCIPLE OF OPERATION
design of the disposal channel is to provide for supercritical flow. Failing
this, the channel design should be checked for sediment transport capacity A side-sluice sediment ex-
(see Appendix 2 ) . cluder works by dividing the
already diverted flow into
In many cases it is inevitable that compromises will have to be made. two parts, one on either side
Where possible, provision should be made for high extraction ratios, for of a vertical dividing
flushing out the sediment during periods of river flows in excess of canal wall. One part is returned
demand. Alternatively a flushing sluice can be located in the canal upstream to the river through a sluice
of the vortex tube. channel and the other is
passed into the canal and
The alternative to gravity disposal is to provide a settling basin and used. Fig. 6.3.1 illustrates
mechanical or pumped removal of sediment. The smaller flows from the vortex the principle.
tube extractor, compared with a canal settling basin, tend to favour the use
of compact settling tanks and pumped disposal of the sediment. Fig. 6.3.1 Flow diagram for right bank Intake

Where there is a shortage of water supplies for continuous ejection, an The twin objects of the design procedure are (a) that the sluice water should
alternative solution is to provide settling tanks with intermittent pumped carry a higher concentration of sediment, and the canal water a lower concen-
disposal of the sediment and re-use of the vortex tube flow. For this purpose tration, than the water initially diverted, and (b) that the water initially
it is necessary to locate the vortex tubes upstream of a canal fall. This diverted should have a lower concentration of sediment than the river it-
arrangement serves the same purpose as a dredged settling basin but eliminates self. There is no point in achieving one of these objectives if it is at the
the more skilled requirements for the operation and maintenance of the expense of the other, because the overall reduction in concentration is the
dredgers, and controls the size of material trapped. product of the reductions at the two divisions of flow.

6.3 Side-Sluice Sediaent Excluders DESIGH OF INITIAL DIVERSION

Applications: The design of the initial diversion is important when only part, Q^, of
the river discharge, Q r , is being diverted, because their relative sediment
Side-sluice sediment excluders are used to abstract quite large flows concentrations will be affected.
from rivers carrying appreciable quantities of sand and gravel. They require
a small head (fall provided by a weir across the river) to operate and are The discharge Q d is the sum of the canal and sluice discharges Q c and Q 3
easily cleared if blocked during a flood. and can for design purposes be taken as:

In common with all sediment excluders, side-sluice sediment excluders do Q d = 1.5QC 6.3.1
not affect suspended sediment. In what follows, sediment should be taken to
mean sediment travelling along the bed. where Q c is the maximum canal discharge.

The discharge Q d , when in the river and approaching the headworks,


occupies a certain width, b d , out of the total width of flow b r , and it is
sufficiently accurate to assume that

62
6.3.2
Od
, Qr-Od
The total width of flow at any discharge depends on factors such as how long \ \\\
that discharge has been flowing or whether It is rising or falling, but it may i Bed
^ current /
roughly be taken as / Dividing
'' streamline
bd;
b = 5 /Q (b in metres, Q in m 3 /s) 6.3.3 5 br

although any strong evidence in a particular case may indicate the use of a FIG. 6-2-4.
factor other than 5.

Combining equations 6.3.2 and 6.3.3 gives Fig- 6.3.2 Fig. 6.3.3
Plan view of approach to right-bank intake
b, = 5 (b . in metres, Q and Q r in 6.3.4
Now equation 6.3.4 shows that as Qr increases so b^ decreases, ft layout
For a chosen value of Q r , the dividing streamline upstream of a given that f a l l s in Fig. 6.3.2 when Qr is only Just greater than Qd will do so at
diversion structure can be sketched in. If the dividing pier is further out any higher value of Qr also. It is therefore sufficient to set the centreline
into the river than b d , the dividing streamline will curve away from the bank of the pier out from the bank a distance, b g l equal to b^ when Qd = Q , giving
and vice versa.

The basic hydraulic principle at work is that bed currents cross a


b 6.3.5
dividing streamline from the outside to the inside, thereby directing the bed e =
load either towards or away from the intake, depending on the direction of
curvature of the dividing streamline. At the same time surface currents cross The above treatment is based on an important proviso: for a right-bank
the streamline from the inside to the outside. The two currents are equal, intake the width of flow in the river is always measured from the right
which is what makes the line an average streamline in plan. bank. In other words, i t is assumed that whatever the discharge, the flow
'hugs' the bank in which the intake is located - this is generally the case
In Fig. 6.3.2 more surface water is diverted than bottom water, with the only where the intake is located on the outside bank of a bend in the river
result that sediment concentration in the diverted water is reduced. In Fig. (3.1).
6.3.3 the exact opposite is true.

64 65
DESIGN OF SLUICING ftRRftHGEHEMT A well proportioned structure arises if

The same p r i n c i p l e s apply to t h e design of t h e second d i v i s i o n of t h e 6.3.7


flow i n t o t h e s l u i c e water and canal water, F i g . 6 . 3 . t .
and 6.3.8

Gate silt resulting in a frontal width for the gates equal to half or a third of that
B \ C Bank top
resulting from a thin pier and no bank outstand. A complete design along
Weir
Crest Pier these lines is shown in Fig. 6.3.5.
line tor width
measurement is where
weir crest level Canal supply culvert
intersects bank
Front elevation Guide

Sluice gate Primary canal

i Primary canal

supply culvert

1st f
division L/^ Plan

Fig. 6.3.1 Right bank intake definition sketch


Plan

If, looking downstream, the point B is further from the bank than point
A, then there will be no value of the sluice discharge, Q s , for which the
dividing l i n e will not curve the right way. I t is sufficient to make the line Fig- 6.3.5 Weito irrigation project - proposed diversion weir and headworks
AB p a r a l l e l to the bank, giving

b
6.3.6
\ -

The canal gates have to resist the pressure of the river in flood and are a
costly item. Their width can be reduced by moving B closer to the bank, which
means increasing the width of the pier (equation 6.3.6), by keeping the
angle a down to around 20 or 30, and by moving C out from the bank while
shaping the bank to remove the resulting sharp corner at C.

66 67
Summary of Widths (units in metres and nrvs) The whole of the floor between the pier and the bank and in front of the
gates may be set at the same level as the floor of the sluice channel. Some
designers feel the need to have the floor in front of the gates at a higher
Q e = canal capacity
level, with the resulting step, shown in Fig. 6.3.5, called a skimming weir.
= 6 6.3.5 It is important to realise that a step in the floor can only help in as much
as it produces the correct curvature in the dividing streamline, on which
b
6.3.9 alone the direction of the bed currents depends, and the effect of a step in
P =3 /Q
c to
this respect is not at all obvious. If the direction of flow near the floor
6.3.8 is up and over a step, sediment soon forms a ramp against the step which the
following sediment is able to ascend. This happens, for instance, on a short
b - b - b 6.3.10 stretch of step near point B, if the curvature of the sluice channel is in-
g e p
verted too soon after passing point B.

FLOATING TRASH
HEIGHT OF PIER

If the pier is built up to bank top level, floating trash in times of An intake designed to exclude bed sediment in the way described auto-
flood will get trapped by it in front of the gates. If the pier is built up matically attracts any floating debris that may be in the river. The two are
to weir crest level only, it will be completely effective when no water is inseparable parts of the same process. This can be a problem in rivers with a
passing over the weir, and will continue to be effective even when moderately lot of trash, and easy access from above should be provided to the areas in
submerged. At higher flows the pier, which is now only an extension upstream front of the intake.
of the weir crest, becomes largely invisible to the water, which sees only the
solid outstand from the bank formed by the gates. Since this outstand at high The hydraulic details of the design of such an excluder must be sorted
flows intercepts a much higher discharge than is actually taken into the out using physical or mathematical models.
gates, very favourable conditions of the Fig. 6.3-3 type cause vigorous scour-
ing in front of the gates and prevent any accumulation of sediment. Mote that 6.^ Tunnel Excluders
these are the requirements of a plain intake, which is what the present intake
becomes at high river flows. The principle of tunnel excluders and ejectors is similar to that of
frontal intake structures in that the flow approaching a barrage is split
FLOOR LEVELS horizontally and the clearer water flows through the upper chamber into the
canal, while the sediment laden water flows into the lower channel and is
The floor level of the sluice channel and its narrowest width should be passed back into the stream. A typical tunnel excluder is shown in Fig.
such that a discharge Q s equal to Qc/2 or more can pass down it without back- 6.1.1. This type of device has proven very efficient for large and small
ing the water up to such a level that it starts to flow over the weir. This diversions as a sediment excluder - used upstream of the diversion point - and
calculation must take account of the tailwater level to be expected at the as an ejector - used downstream of the diversion point.
sluice channel outfall. If the tailwater level resulting from <JS being dis-
charged into the river is so close to weir crest level that the sluice channel
must be made wide and deep in order to pass the flow, shear stress or sediment
transport calculations are necessary in order to cheek whether sediment will
settle on the floor of the sluice channel.

68
Crest
6.5 Approach Plow Control

Barrage
Although approach flow control is clearly not a sediment excluding
device, it is well established that control of approach flow can achieve
sediment exclusion from intakes. The aim is to provide the favourable effects
of channel curvature at the approach to the intake to prevent bed load sedi-
Row Undersluices"
ment from entering the intake.

Ran of tunnels
If natural channel curvature is insufficient or ill-defined at the
intake structure it may be induced by the construction of guide walls, curved
vanes, etc. upstream of the intake. Examples of walls and vanes to control
Existing pond flow are numerous and a number are included in Ref. 3.1.

Main canal
When a barrage is constructed across a waterway and intakes to canals
are to be positioned on both banks, the effects of unfavourable curvature are
almost bound to influence at least one of the water systems to be supplied.
Cross-section of tunnels To improve the situation a central island and guide banks may be used to
provide favourable curvature to both banks - Fig. 6.5.1.

Fig. 6.4.1 Tunnel type sediment excluder

Existing banks
Literature reviewed in Ref. 3.1 give a number of examples of tunnel excluders
and some design guidelines:

Concave/convex Recommended
(i) Entry to the tunnels should be bellmouthed to avoid turbulence where the right guide bank concave/convex
left guide bank
flow is split.
Central island

(ii) Approach flow direction should be straight and the site should be care- Right bank canal Left bank canal
fully chosen to ensure that the tunnels can cope with incoming dis- Barrage
charges.

(iii) The approach, sizing and entry conditions should be physically modelled
to ensure that the tunnels will operate efficiently. Fig. 6.5.1 Example of alignment for guide banks using
central island to induce channel curvature

(iv) Downstream conditions must be carefully considered to ensure that the


downstream ends of the diverting tunnels are kept clear of deposits at Literature on approach flow control suggests that intakes should always be
all times. constructed where suitably curved approach flow is available, and if this is
impossible, curvature should be brought about artificially. However, it has
been shown that where tunnel excluders (section 6.<4) are used, maximum
efficiency is obtained when the approach flow is straight. So control of the

70 71
approach flow must be considered fully, in light of the other aspects of the The total flow diverted is Q
design. When guide walls, vanes, etc. are to be used their location, extent, = 22 m-Vs, with Q c = 16 m^/s
curvature, etc. should be investigated using a model study where the whole being taken in the canal.
range of operating conditions can be studied. For instance the ratio of flow For these conditions, sedi-
approaching the barrage and the intake depending on gate openings can greatly ment will enter the intake
influence the quantity of sediment conveyed in either direction. for flows between Q o = 24
m-vs (the incipient motion
6.6 Sediment Excluding Intakes discharge) and Q = 45 nrvs.
This corresponds to entry of
The sediment excluding features of a number of intake structure types sediment over 96 days of the
and the features of some sediment excluding devices have been described in year. Initially, the sedi-
chapter t, and earlier in this chapter. Many intakes incorporate combinations ment that does enter deposits
of these methods to exclude and control sediment. A few intake designs in the mouth of the sluice-
developed with sediment exclusion and control as a main criteria are described way. Subsequent sediment is
here. moved to the right side at
the entrance of the channel
Existing problems due to leading to the
sediment entry at the intake Fig. 6.6.2 Final layout of Intake excluder. However, the
on the Kander River, second curve re-distributes
Switzerland instigated model the sediment reasonably evenly over the whole width so that flow and sediment
tests to redesign the intake conditions are uniform at the excluder. This double use of curvature within
(Ref. 6.6.1). A number of the intake (see Fig. 6.6.2) was found to be the most effective means of
alternatives were considered attaining uniform flow conditions. Considering scale effects, the material
and tested, ft maximum river deposited in the canal for different flow rates shows that only silts and fine
10 Grain size,mm discharge of 185 nP/s was sand fractions will escape the excluder in the prototype. A trash rack is
assumed and the sediment located at the entrance of the channel leading to the excluder. The gate at
Fig. 6.6.1 Grain size distribution curves grain size distribution is the end of the sluiceway will be used to enable sluicing of material from in
shown in Fig.6.6.1. The front of the intake and in the sluiceway, and to control water levels within
resulting intake design, shown in Fig. 6.6.2, combined the use of approach the intake so that material deposited in the secondary intake channel can be
flow control and a tunnel excluder just upstream of the canal proper to con- flushed through the excluder.
trol sediment.

The performance of the excluder can be improved by employing a regula-


tion scheme where the excluder is shut for flows less than Q - 20 nrvs i.e.
when no sediment transport will occur. It was considered that management of
the intake would be simplified if the excluder was left functioning for all
flows greater than Q = 20 nW/s. For all low flows, the water level will be
kept at the crest level of the weir.
For flows greater than Q = 45 nr/s, the spiral flow created by the
groyne in the river bed and the curvature of the right bank keeps the trans-
ported bed load away from the intake.

Submerged wait Diversion


Two mandatory requirements of the project were that a residual flow of dam StiUing basin
0.7 nrVs be left in the river for low flow periods, and that a fish ladder of River-
r
Entrance \
acceptable design be incorporated into the weir or intake. The fish ladder Sill \ .
can be seen in Fig. 6.6.2. The flow down this structure simultaneously satis-
fies the residual flow requirement. IK Flushing gate
\ \ \ \ \ Intake NM \ Sediment canal
Another example of new intake design is that described in Ref. 6.6.2. ^y v
^3T-Curved drop structure
The aim of this work was to develop sediment control facilities for diversion
\ \
headworks in mountain streams. The design which was developed is shown in ' ^-L-z
Xiif^-Tunnel
Fig. 6.6.3. Instead of horizontal separation of the flow, a slot in the
bottom of the intake is used to separate sediment and water. Underneath the Previous design

slot the sediment is collected in a sediment trap and flushed downstream where
it is returned into the river. The hydraulic principle is similar to that of
a vortex tube sediment extractor used sometimes in irrigation canals. How-
ever, the application in connection with a diversion and the integration in an
intake structure was new and created additional challenge.
Submerged wall and Diversion
This design for sediment control at diversions proved to be a well trash rack entrance sill J .tilling basin

functioning alternative to existing conventional devices. It can be applied Riveri


whenever the plan area is limited and sufficient head is available.
Considering these presuppositions it is obvious that the method is mainly
Sediment canal
restricted to diversions in mountain rivers with high slopes and steep banks.
Curveidrop structure

ill |JS Tunnel

Final design

Fig. 6.6.3 Subersach diversion

74 75
For proper functioning several construction rules and operational in- 6.1.3 Russein, A.S.A. Experimental study of a curved channel sedi-
structions have to be observed carefully. With respect to the sensitivity of ment excluder - experimental work carried out at Chilworth
the new intake type against unfavourable flow pattern in the vicinity of the Hydraulics Laboratory, University of Southampton.
intake and against wrong operation of the flushing gates, the carrying-out of
hydaulic model tests, by which the design was developed, is highly reeonmend- 6.1.4 Developments in Hydraulic Engineering, Volume I, Novak, P.
ed. Editor (Chapter 5, K.V.K. Smith, Developments in the Design of
Irrigation Structures), Applied Science Publishers, London,
The Brander No 1 intake for the Awe Project of the North of Scotland 1983.
Hydro-Electric Board was designed to control sediment ingress. The intake
incorporates a sediment settling channel and a water operated sector gate for 6.2.' Sanmuganathan, K. "Design of vortex tube silt extractors",
scouring. Fig. 6.6.4 shows scouring in operation. Report OD/6, Hydraulics Research Station, Harch 1976.

6.2.2 White, S.M. "Design manual for vortex tube silt extractors",
Report OD37, Hydraulics Research Station, December 1981.

6.2.3 Atkinson, E. "Field verification of a performance prediction


method for sediment extractors". Report OD 90, Hydraulics
Research, August, 1987.

6.3.1 Blanchet, C. "Importance of location of water intake on a river


for prevention of entry of sediment" (in French). Paper R22 in
Vol. 2 {Q30) Trans. l.C.I.D. 9th Congress, Moscow, 1975.

6.6.1 Whittaker, J.G. "A solution for sediment control at


intakes",
Proc Channels and Channel Control
Structures 1st International Conference,
Southampton, England, April 1984.

6.6.2 Scheuerlein, H.J. "fl new design of sediment control


facilities for diversion headworks in
Fig. 6.6.1 Cruachan aqueducts - Brander (1) intake - scouring mountain streams".
Proc Channels and Channel Control
References Structures 1st International Conference,
Southampton, England, April 1984.
6.1.1 ICID Ninth Congress on Irrigation and Drainage, Moscow 1975.

6.1.2 Colby, B.R, "Discharge of sands and mean velocity relation-


ships in sand bed streams", US Geological Survey Professional
Paper H26-A, Washington 1964.

76 77
1
Chapter 7

SETTLING BASINS
7. SETTLING BASINS effects. Discussion is limited to horizontal flow basins as typified by the
enlarged canal cross-section in irrigation systems, or the rectangular tank
Where a river carries a substantial quantity of sand and finer particles common in water supply schemes (Fig. 7.1). The basic settling theory is also
in suspension, this material cannot practically be excluded at the intake. A applicable to the more sophisticated vertical-flow and radial-flow tanks found
settling basin, in which the velocity of the offtaking flow is reduced to in water supply installations (Ref. 7.1)
enable the sand and heavier silt load to settle out under gravity, is used in
such circumstances to reduce the sediment load in the downstream system.
Water supply system

The river intake installations where settling basins are commonly used
are as follows :-

(i) In irrigation schemes, to reduce the sediment load to a level which can
be mainly transported by the distribution system through to the fields,
thus mitigating the problems of sediment deposition in the canals. (The Sediment collection
channel Supply Plan
sediment transport capacity of irrigation canals is often restricted due
to the flat water surface slopes needed to maintain command.)

(ii) In power schemes to reduce the sediment load to acceptable levels from Inlet zone Outlet zone
the viewpoint of (a) sediment transport capacity of the supply system to
Elevation
the power station and, (b) damage to machinery.
Sediment storage zone
(iii) In water supply schemes to reduce where necessary the sediment load in
the supply by gravity sedimentation as a preliminary to subsequent Irrioation system
chemically-assisted sedimentation and/or filtration. Contro! gates / Scour

The application of gravity sedimentation in water supply schemes is - ^ n - " ^ ._


limited because impurities such as algae, vegetative debris, fine silt and
colloids do not settle out adequately under gravity without the use of floecu-
Inlet zone Outlet zone
lants. Note also the separate requirement for bankside storage in river Settling zone
Plan
abstraction schemes (typically 3-7 days' capacity) where the river may be
subject to industrial pollution - such reservoirs cannot strictly be termed
Sediment storage
settling basins although as a secondary function they may settle out consider- zone Supply
_l canal
able quantities of sediment.

In this chapter the basic theory and approach to settling basin design Elevation
is sumnarised, and the practical aspects involved in their layout are review-
ed. The scope is limited to gravity (as opposed to chemically-assisted)
sedimentation - i.e. to settling out discrete particles which retain their Fig. 7.1 General layout of settling basins
individual settling characteristics without interference or flocculant

80 81
7.1 Fall Velocity of Discrete Particles good review of the settling of suspensions from the viewpoint of secondary
treatment processes in the water supply industry. Hindered settling occurs
Fall velocity, w, In quiescent water characterizes the ability of when discrete particles settle in close proximity to one another, and their
different sized particles to settle out under gravity {and indeed is conmonly velocity fields interfere. Interference effects become significant at sus-
used as standard measure of particle size for fine sediments). Fall velocity pended sediment concentrations in excess of 2,000 mg/1 {Ref. 7.2), when the
for discrete particles is dependent on particle size, specific gravity, par- reduction in fall velocity for a coarse silt suspension might be about 10*
ticle shape, and the viscosity of water. Fig. 7.2 shows fall velocity in (Ref. 2.2). In general, the effects of hindered setting are not significant
water, w, plotted against particle diameter, D, for reference quartz in terms of the ranges of sediment concentrations and the degree of accuracy
spheres. Various equations exist which give approximate solutions for fall considered here for settling basin design.
velocity of single particles - the Rubey equation (Ref. A2.11) is commonly
used for particles with the shape of natural sands. The ASCE Manual on Determination of Settling Characteristics of Suspended Sediment
Sedimentation Engineering (Ref. 2.2} can be consulted for further details.
The significant effect of water temperature on fall velocity - particularly in Design fall velocity for discrete particles may be measured directly by
relation to design for the tropical environment - should be noted. timing fall through a known depth. It is usual, however, to use published
data (such as Fig. 7.2 or Ref. 2.2) to estimate fall velocity when the basin
is principally concerned with settling out sand and coarse silt sizes. In
water treatment processes, where finer particles need to be removed, it is
usual to carry out a settling column analysis to determine the settling
1000 characteristic curve for the suspension (Ref. 7.2).

=100
7.2 The Ideal Basin

Re = 10

Han arec . A

Re-0-01 y
0-1 1 10 100 1000
Foil velocity w , cm / sec

Fig. 7.2 Fall velocity of quartz spheres in water


Depth
d L -v"
P*"
*
Discharge Q
7
Settling path of particle

Effect of Concentration on Fall Velocity


Retention t i m e , U - Ad / O
Settling time , t j d/w
With increasing concentration of sediment, actual particle fall velocity
will differ from that for discrete particles due to interference of other
particles. Flocculation can occur in high concentrations of silt, clay and Fig. 7.3 Ideal settling basin
organic particles when the particles coalesce to fall in a group at a higher
velocity (ftefs. 2.2, 7.2). Removal of such fine particles is however gener- The ideal horizontal settling basin, Fig. 7.3, demonstrates the basic
ally outside the scope of gravity sedimentation - Miller {Ref. 7.3) provides a theory of sedimentation developed by Hazen (Ref. 7.4). The following assump-
tions are made:- uniform distribution of flow and suspended solids at entry Hazen (Ref. 7.1) attempted to account for the effects of both turbulence
to settling zone (plug flow); quiescent flow (i.e. no turbulence); solids and short-circuiting by a general classification of basin performance in his
entering deposition zone are not resuspended. Consider a sediment particle formula:
entering the basin at point X :
n = 1 - [1 (=*) 7.1
Settling time, t s = d/w 7.1
Retention time, = basin volume/discharge = dA/Q 7.a where m is a performance parameter varying from m = 0 for "best" basins to
m = 1 for "very poor" basins. Hazen's equation is shown graphically in Fig.
where yo = basin flow depth; A = mean plan area of basin; Q = discharge. For 7.1 and is s t i l l commonly quoted today. The disadvantage of Hazen's formula
quiescent settling, all particles of settling velocity w are removed when is that several different physical effects are combined into a single para-
retention time equals settling time: meter, m. It is better for the designer to consider each effect separately as
follows.
i.e. dA/Q = d/w, or Q/A = w 7.3
Quiescent setting Best performance, m-0
1 m--1 Vetter's equation
A similar formula can be obtained for a vertical flow tank. In general for
both ideal and real basins, the ratio wA
g- can be regarded as a dimensionless 100
indicator of the physical ability of a basin of plan area A to remove par- Shaded zone-.
covers Camp's
ticles of fall velocity w at supply discharge Q. solutions _ 80
fRg.7.5) |

It follows, in the ideal case, that for discrete particles: i.\o -


X 8.
8 | >
(a) removal is independent of basin depth and flow-through velocity, *" 8
& 20

(b) for a given discharge and suspended sediment load, removal is a function
0-8 >6 2-4. 3-2 W) 4-8 5-6
of basin surface area (note in this context the use of compact double-
storey tanks in water treatment).
Fig. 7.4 Performance curves for settling basins of varying effectiveness
The ratio Q/A is termed the "surface loading" or "surface overflow rate".
Effects of Turbulence
7.3 Settling Capability of Real Basins
Camp (Ref. 7.5) based his classic approach to settling basin design on the
In practice, real settling basins act less efficiently than the ideal work of Dobbins (Ref. 7.6). After making simplifying assumptions that (a)
due principally to the effects of (a) turbulence in flow through the basin fluid velocity, and (b) the turbulent mixing coefficient are the same through-
leading to retarded settlement, and (b) short-circuiting and currents within out the fluid, Camp derived a relation for:
the basin. Sediment removal efficiency, n, for a given particle size is
measured as C r /C o , where C r = concentration of suspended sediment removed, and 7.5
C o = incoming concentration.
where vt is the shear velocity and w
v can be regarded as a dimensionless
*
indicator of the effect of the fluid turbulence on a given particle size.

84 85
_^B ^^^ ^^^ ^^^ ^^^ ^^^ ^^^ ^^M ^^M ^^n ^^H ^^H ^^H ^^H BBB MB HHI H^B IHI
MB Hi BBB MMi Hi BlBB ^^B ^^B B^BI B^B B^B ^^B ^^B ^^M ^^ ^^m ^^*
Camp's solution to equation 7.5 is shown graphically in Fig. 7.5 (the horizon- The practical upper limit of flow-through velocity for basins designed to
tal axis has been redrawn in terms of ^ - ) . s e t t l e out sand particles is generally taken as 0.3 m/s.

Shear velocity, v, = /g BS 7.6 Other recent approaches {Refs. 7.7, 7.8) have taken into account more
accurately the effects of turbulence on settling. Camp's solution however
where R = hydraulic mean depth, and S = hydraulic gradient which is calculated remains widely accepted. In irrigation design practice, Vetters' equation
from a boundary resistance equation such as Manning's and essentially depends (Ref. 7.9) i s often used and is virtually identical to the equation proposed
on flow-through velocity. by the USBR (Ref. 7.10}:

100 wA
n = 1 - e' Q 7.7

This i s simply t h e " b e s t performance" s o l u t i o n of Hazen's e q u a t i o n , ( i . e .


curve for m = 0, Fig. 7.1). Note that Vetter's equation also corresponds to
the turbulent side of Camp's solution (Fig. 7.5) and thus to implicit con-
ditions of turbulence. Note also that Camp's solution could be plotted on
Fig. 7.4 as a series of curves, similar to the Hazen performance curves. It
can thus be appreciated that each of the common settling efficiency formulae
(equations 7.4, 7.5 and 7.7) i s valid when the implied physical conditions are
appropriate.

0-01 2-0 10-0 Effect of bed scour

Increasing turbulence
Once particles have settled out, they must not be scoured from the basin
Fig. T.5 Camp's solution for settling basin efficiency floor by excessive flow-through velocity. The shear stress on the floor must
therefore be less than the c r i t i c a l shear stress required to i n i t i a t e move-
The following table demonstrates how, for fixed flow depth and surface load- ment.
ing, flow-through velocity has to be reduced if the same scale of turbulence
effects on settling is to be maintained when design particle size is reduced: 7.8

Design particle size, D (ran) 0 . 06 0. 02 where T = bed shear stress, y = specific weight of water, and sw = specific
Mean flow-through velocity, V (m/s) 0. 2 0.02 gravity of water. The c r i t i c a l shear stress to initiate motion can be
Hydraulic gradient, S 3.6 x 10" 6 3. 6 x 10-8 obtained for the DCQ sediment deposit size from Shield's diagram (Fig.
Shear velocity, v* (m/s) 8. 1* x 10" 3 8. 1* X 10-* 112.2). Equating the two shear stresses enables the c r i t i c a l flow-through
Fall velocity at 20 C, w (m/s) 3. 5 x 10"
3
3. 5 x ID"" velocity to be obtained. Various equations giving particular solutions for
w/v* 0 . 42 0 . 42 c r i t i c a l flow-through velocity relating to the different bed conditions in
Shield's diagram have been published (Refs. 7.5, 7.11).
Notes: d = 2.0 m and V r d 2 / 3 S*/nm where Manning n = 0.015.

86 87
Short circuiting and basin stability Factors which cause short-circuiting of flow, with consequent reduction
in hydraulic efficiency a r e : -
In the ideal basin, flow is steady and uniform (plug flow), and a l l
fluid particles are detained in the settling zone for the retention time, currents set up by poor inlet and outlet conditions, and basin
t R . In practice, even with well designed basins, flow is non-uniform and some shape,
parts of the basin volume are ineffective. Lengths of stream paths of indivi-
dual fluid particles vary - some reach the outlet in less than the theoretical dispersion in the horizontal plane due to turbulence,
retention time, while others take longer to do so. The flow-through curve for
a tank (Fig. 7.6) provides a convenient indicator of hydraulic behaviour and wind-induced surface currents,
efficiency. I t i l l u s t r a t e s the departure from ideal (plug) flow caused by
short-circuiting of individual flow paths through the basin. The objective of density currents induced by thermal effects.
good hydraulic design is to achieve conditions most closely relating to ideal
flow. The hydraulic design of the inlet and outlet layout and basin shape are the
most significant from the design viewpoint - these are discussed in section
7.^. The other effects become increasingly important as flow-through momentum
Rug flow { i . e . velocity) is reduced - hence their greater relevance in the water t r e a t -
Theoretical ment sector. Clements {Ref. 7.13) notes that the effects of wind can be
detention
period,trj Dispersion index - to-. / 1 . reduced if i t is possible to align the basin along the prevailing wind direc-
90'
tion.
Observed recovery of
tracer substance 7.1 Hydraulic Factors Affecting Basin Layout

Settling Zone

Having determined the surface area of the basin, the designer must
consider i t s plan shape. Camp (Ref. 7.14) demonstrated that the hydraulic
behaviour ( i . e . shape of flow-through curve) of long narrow tanks is superior
to that of wide low velocity tanks, also that tanks with higher (but never-
theless very low) values of Froude number have better flow patterns and give
Fig. 7.6 Flow through curve - (Dead spaces and short-circuiting in a settling less dispersion (Fig. 7.6).
basin are reflected in the concentration and tine of recovery of tracer sub-
stances) . In practice therefore, a minimum length to width ratio L/W of 2-3 is
generally adopted from hydraulic considerations - note that in the water
Non-dimensional plots of the flow-through curve (Refs. 7.5, 7.12) enable supply and power sector, the site layout frequently requires basin length to
comparison of the hydraulic behaviour of different basins. The degree and be minimized. In irrigation systems, it is generally more practicable to
nature of short-circuiting determines the shape of the flow-through curve achieve a better L/W ratio (say 8-10) by local widening and deepening of the
(Refs. 7.2, 7.5, 7.11}. Also, if the flow-through curve for an existing or canal cross-section. Basin shape can be improved by subdivision with longi-
modelled basin does not reproduce itself reasonably well in repeated t e s t s , tudinal divide walls, which may also be required due to operation considera-
then flow through the basin is unstable, and performance will be e r r a t i c . tions (section 7.5).
Inlet Zone 7.5 Operation of Basin and Removal of Sediaent

The need for a good inlet design cannot be overemphasised - poor inlet The method of removing sediment deposits from the basin must be con-
design is probably the factor most responsible for "poor" basin performance in sidered early in the design procedure {section 7.7) since it may well be a
Hazen's classification (Fig. 7.4). To achieve good hydraulic efficiency and critical factor governing the layout of the whole intake site. Removal of
effective use of the settling zone, the inlet strictly needs to distribute sediment may be either continuous (i.e. carried out during normal basin opera-
inflow and suspended sediment uniformly over the vertical cross-sectional area tion without interference to supply from the basin) or intermittent (in this
of the settling zone. context, carried out as a separate operation while supply from the basin is
temporarily stopped). When considering removal of sediment, it is important
Clements (Ref. 7.12) has shown that horizontal velocity variations to provide for removal from any part of the system where sediment is likely to
across the width of a rectangular tank affect the hydraulic efficiency con- deposit (e.g. dredger access to inlet transition) - conversely, it is equally
siderably more than velocity variations in depth, provided always that bed important that adequate sediment carrying capacity is provided anywhere else
scour is avoided. Principal attention therefore needs to be given to uniform in the system (e.g. basin approach channel, sluiceway channel) where sediment
inflow distribution in the horizontal plane. Methods which are commonly deposition is to be avoided.
adopted to achieve good flow distribution are:
In continuous systems, sediment is generally removed by hydraulic
submerged weir, dredger (Fig. 7.12), or by mechancial scraper to a central sump where it is
withdrawn either under basin hydrostatic head or by pumping (Fig. 7.13).
gradual open channel expansion (Fig. 7.1), possibly using guide Mechanical scrapers are generally only used in rectangular tanks for water
vanes, supply systems. In intermittent systems, the basin is emptied or drawn down,
and sediment is removed by hydraulic sluicing, by mechanical or manual excava-
troughs with slots or orifices in walls or bottom, and tion; or by pumping to a suitable disposal point. Environmental or river
shoaling considerations may dictate the extent, method and frequency of dis-
baffle walls. posal of sediment back into the parent river.

Orifices or baffled inlets are generally only used when extremely low flow- Wherever possible, gravity sluicing should be adopted since this is more
through velocities are needed for water treatment. As a general rule, the effective and obviously cheaper than other means. Lack of adequate head may
inlet layout should either follow an existing proven design, or be model however prohibit it, or limit its effectiveness.
tested.
Provision of head for gravity sluicing can be an important secondary
Outlet Zone advantage of low-lift pumped abstraction from the river.

The operating water level of the settling basin is generally controlled A sediment storage zone must be provided beneath the settling zone in
at the outlet, usually by a weir which may be designed to operate as submerged any intermittent removal system. It is common for settling basins to be
in order to conserve head. In irrigation and power systems, conventional constructed in parallel so that any one basin can be taken out of service
undershot lift gates are also commonly used. If it is narrower than the intermittently for removal of sediment, while continuous delivery is maintain-
basin, the outlet control requires an appropriate approach transition to avoid ed by the others. The system of separate sedimentation channels shown in Fig.
short circuiting and to maintain an even flow distribution. The outlet con- 7.7 is common in irrigation systems.
traction may be more abrupt than the inlet expansion (Fig. 7.1).
Gravity Sluicing
B
0.01
si 7.9
For effective gravity sluicing, the system is required (a) to erode
progressively all deposits from the storage zone (Fig. 7.7), and (b) to convey where g s i = transport rate of solids by immersed weight per unit width; eb r
bed load efficiency factor (Fig. 7.8); and tan o = solid friction coefficient
this material at a high transport rate through the basin and associated
(Fig. 7.9). Individual particle terminal velocity should be estimated for D,-o
sluiceway channel to the disposal point - generally the parent river. To
sediment deposit size. Note however that Bagnold's theory is inapplicable for
achieve this, the fall through the removal system should be such that it
DgQ less than 0..015 mm.
operates at supercritical flow in the sluicing mode (Refs. 7.15 (R32, R35),
7.16). Further requirements are the provision of low level scouring sluices,
and careful hydraulic design to ensure that no control to sluicing flow exists
downstream of the storage zone. The mechanism of removal is shown in Fig. 7.7
- the design ease being removal of the last remaining deposits.
D.mm
0-01
0-1
B Settling basin Outlet looe 0-3
/ Isolating gates / 1-0
Inlet zone 1-5 2 3 4 5 6 8 10
-I 11 I / Supply
Supply canal
canal Mean flow velocity of fluid ,V, [fpsl
[5 V * 5 * ^ (in flum section)
(in flumed

Sediment sluiceway
Inlet channel Pig. 7.8 Bagnold's bedload efficiency factor
Inlet gates Sediment flushing gata \ Outlet control
\ Han \ ^ /
viscous conditions
<0-25

Section A-A

Section B-B

fa) Start (blParl sluiced


Sediment sluicing (Diagrammatic)
Wholly inert id!
conditions
0-3
Fig. 7.7 Settling basin with parallel sedimentation channels 0-1 0-2 (K. M1 2
Oimenslonless bed shear stress,
D(ps-Plg
For removal of coarse s i l t s and sand, it is suggested that sediment transport
rates are estimated using Bagnold's total load formula (Ref. 7.17 and equation
7.9) since this is valid for high transport rates rather than the early stages Pig. 7.9 Bagnold's solid friction coefficient
of sediment movement.:
In many instances when designing a settling basin for low head river diversion
works, there are generally conflicting design requirements such a s :

92
To canal

Lack of sediment storage capacity.


To disposal
Channel
Seasonal limitation of sluicing supplies. Rushing sluice
Section A-A
Difficulty in sluicing settling basin during long periods of
high flood levels.

In many cases sluicing has to be confined to the lower range in flood


levels, and sediment storage has to be sacrificed to produce sufficient sluic- Canal
ing power. The reduction in the storage capacity results in the more frequent Settling basin
flushing sluices
shut down of the canal distribution system. Some of the difficulties can be
overcome if the settling basin and flushing sluices are duplicated - as in
Fig. 7.7. However, this adds considerably to the cost and requires additonal
water supplies which may not always be available, and significantly it does
not solve the problem of not being able to sluice the basins during long
periods of high flood levels in the river.

An alternative means of overcoming these conflicting requirements is to


provide for both settling-scouring operation and continuous ejection opera-
tion. During periods of low river flows and low sediment concentration, the
normal operation of allowing the sediment to deposit in the settling basin, is
followed by the scouring operation of the desanding sluices to remove the Plan
sediment. However, during periods of high sediment concentrations and high
flood levels this method of operation may not be possible, and to remove the
lower sediment laden water the desanding sluice gates are operated to allow Fig. 7.10 Settling basin
for the continuous ejection of the sediment back to the river. When consider-
ing continuous sediment ejection, it is important that the ejection flow 7.6 Examples of Settling Basin Design
should pass through tunnels starting off parallel to and under the canal
flow. An example of a settling basin and desanding sluices designed on this Relatively little prototype data has been published on the performance
basis are shown in Fig. 7.10. of settling basins outside the water supply field. Ref. 7.15 contains some
general details, including recent data on the desilting works at Imperial Dam
on the Colorado River (Fig. 7.11) - possibly the largest in existence.

94
All American
Canal hecdworks
Colorado
River
Gila Valley Canal
headwords (See Fig.7.11bl
/
1 The Imperial Dam works are designed for removal of suspended sediment
particles larger than 0.05 mm from the supply to two large irrigation
canals. Basin size was designed using Vetters' equation (equation 7.7). The
right bank (All-American Canal, constructud pre-1910) desilting works comprise
Rotating scrapers an unusual layout of three basins, each containing 2l| no. 38 m diameter rotary
Gila Valley Canal
desilting works scrapers. On each scraper, diagonal scraper blades move sediment continuously
Desilting towards the central pedestal where the sediment is drawn off by hydrostatic
works head through a sludge pipe and discharged into the sluiceway channel. The
Gravity main canal surface area of each basin is 33,500 m , giving a surface loading of
Influent channel 3.1 x 10"^ m/s at the design discharge of 113 m^/s. Design flow-through
Row
By-pass and effluent channel velocity i s 0.08 m/s. The left bank {Gila Canal) desilting works, with a 355
To sediment lagoon in river m long basin of trapezoidal cross-section, are more typical of the normal
irrigation system layout. Outlet and gravity sluicing control is provided by
Sludge pipes a double bank of gates. For the Gila basin, the surface loading is about 2.1
All American Canal x 10~3 m/s at the design discharge of 57 nrvs. The basin invert gradient is
Plan about 0.15J. Further details of the scheme are given in Refs. 2.2 and 7.18.

During the period 1957 to 1972, the Ail-American works removed an aver-
age of 411,000 tonnes of sediment each year, representing a removal efficiency
Fig. 7.11(a) Imperial daa desilting works
of 61*. The cost of removal has been estimated at t0.40/tonne (1972 prices,
(Ref. 7.15)).
C
Gila canal Alternative
pipeline

rfi Gate structure Basin 2

Axis of
non-overflow - -t Gate structure Basin 1
structure
Slurry pipeline Plan
from dredger

Fig. 7.12 Settling basin with dredger

Fig. 7.1Kb) Desilting works in Gila Valley Canal


Fig. 7.12 shows a continuously dredged basin designed for 90J removal of
fine sand (0.06 mm) from the head reach of an irrigation canal. The surface

96
97
loading is 0.95 x 10' 3 m/s a t the design discharge of 40 m 3 /s. Flow-through 7.7 Design Procedure
velocity i s 0.2 m/s. Two dredgers are provided - the main duty dredger
returning the dredged slurry to the parent river, the other acting as standby The suggested design procedure is as follows:
and able to pump to an alternative disposal lagoon.
(a) Review estimated suspended sediment inflow rates passing headuorks
(particle size distribution, concentration, variability).

Deposits flushed Deposits scraped (b) Review tolerable suspended sediment inflow to system downstream of
(a) basin.

Fig. 7.13 Representative designs of horizontal flow settling tanks (c) Decide on minimum particle size for (say) 90J removal. Estimate basin
surface loading using Vetter's equation (equation 7.7) and calculate
In comparison with irrigation works, Fig. 7.13 shows typical cross- basin surface area. Subdivide suspended sediment inflow into particle
sections of horizontal-flow settling tanks used in water supply systems and size bands. Estimate removal efficiency for each band, size distribu-
Fig. 7.14 shows an enclosed layout used in a hydropower scheme (fief. 7.19). tion of sediment deposit, size distribution and concentration of out-
Other examples are given in Refs. 7.1, 7.11, 7.20 and 7.21. Design surface flow. Compare basin outflow with tolerable inflow to downstream
loading can range from 3.0 x 10~5 to 3.0 x 10"^ m/s (0.1 to 10 m/hour) depend- system. Repeat if necessary.
ing on the size of particles to be removed.
(d) Review local topographic and environmental factors. Consider available
A-Zone for trapping sedimert head through settling basin system in relation to sediment disposal
Barrage B-lnlet to desilting tunnels
location and system downstream of basin. Review any relevant local
C-Grovel by-pass
D-Settling basin experience with settling basins. Review economies of increased main-
Flow
tenance costs versus degree of sediment removal. Make broad decisions
Shaft intake on basin size, method of sediment disposal, extent of standby capacity,
to turbines and general layout of works.
Settling basin
General arrangement (e) Review constraints on basin depth and flow-through velocity for pre-
liminary basin layout from considerations of:

practicality of construction and operation,


low-
basin turbulence,
Rushing pipes Bushing tubes !
bed scour.
Supply to
low- powerhouse Flushing
trench
153 (f) Review, using Camp's solution (Fig. 7.5) and a conservative estimate of
Rushing pipes turbulence function, the initial estimate of basin area. Optimise basin
Plan of settling basin Section A A
length and width within constraints of L/W ratio, minimum basin depth
and maximum flow-through velocity.
Fig. 7.14 Layout of settling basins in hydroponer 3che

98 99
(g)

(h)
Calculate final design deposition rate, size distribution of sediment
deposit, and size distribution and concentration of suspended sediment
in outflow as in Step (c).

Estimate additional sediment storage requirement below settling zone


1 7.9

7.10
Vetter, C.P. Technical aspects of the silt problem on the Colorado
River, Civil Engineering, Vol 10, No 11, November 1940, pp 698-701.

Pemberton, E.L. and Lara, J.M. A procedure to determine sediment de-


position in a settling basin, Sedimentation Investigations Technical
based on design deposition rate, method/frequency of sediment removal, Guide Series, Section E, Part 2, Bureau of Reclamation, US Department of
and (where appropriate) available sluicing discharge. For gravity the Interior, August, 1971.
sluicing, finalize hydraulic design of sediment removal system. Fix
invert levels of basin. 7.11 Fair, G.M, Geyer, J.C. and Okun, D.A. Hater and wastewater engineering,
Hiley, New York, 1968.
(j) Finalize hydraulic design of inlet and outlet zones. Review need for
model test of basin hydraulic layout. 7.12 Clements, M.S. Velocity variations in rectangular sedimentation tanks,
Proc ICE, Vol 34, 1966, p 171.
References
7.13 Clements, M.S. and Price, G.A. Some lessons from model and full scale
7.1 Skeat, H.A. (editor). Manual of British water engineering practice, tests in rectangular tanks, Hater Pollution Control, Vol 73, 1974, No 1.
Institute of Hater Engineering Societies, 4th edition, Vol III, Heffer,
Cambridge, 1969. 7.14 Camp, T.R. A study of the rational design of settling tanks, Sewage
Works Journal, Vol VIII, No 5, September 1936, p 742.
7.2 Tebbutt, T.H.Y. Principles of water quality control, 2nd edition,
Pergamon Press, Oxford, 1977. 7.15 International Commission on Irrigation and Drainage, Transactions, 9th
Congress, Moscow, 1975. Vol II, Question 30, Sediment-controlling
7.3 Miller, D.G. Sedimentation, a review of published work, Water and Hater irrigation intake structures.
Engineering, Vol 68, 1964, pp 52-58.
R9 Hariri, D. and Parhami, F. Desilting structure for the Moghan
7.4 Hazen, A. On sedimentation, Trans ASCE, Vol LIII, 1904, p63. canal, Iran.
R29 Moser, T.H. and Sears, H.D. Sediment control at Imperial Dam.
7.5 Camp, T.R. Sedimentation and the design of settling tanks, Trans ASCE, R32 Shimura, H. and Kinose, K. Method of improving the performance
Vol 111, 1916, Paper no 2285. of a settling basin at a water intake of an alluvial fan river.
R35 Toru, K. Design of an irrigation water intake.
7.6 Dobbins, H.E. Effect of turbulence on sedimentation, Trans ASCE,
Vol 109, 1944, p629. 7.36 Ackers, P. Sediment transport problems in irrigation systems design,
Chapter 4 - Developments in hydraulic engineering - Volume 1, Novak, P
7.7 Ruh-Ming Li and Hsich W. Shen. Solid particle settlement in open- (editor), Applied Science Publishers Ltd., 1983.
channel flow, Proc ASCE, Vol 101, HY7, July 1975, Paper no 11460.
7.17 Bagnold, R.A. An approach to the sediment transport problem from
7.8 Suraer, B.M. Settlement of solid particles in open-channel flow, Proc general physics, Geological Survey, Professional paper 422-1, US Govt
ASCE, Vol. 103, HY11, November 1977, Paper no 13368. Printing Office, Washington, 1966.

100 I0I
7.18

7.19
Donniny, F.E. Design of d e s i l t i n g works for i r r i g a t i o n systems, Proc
ASCE, Vol 92, IR1*, December 1966, Paper no 4998.

McKeogh, E . J . Sedimentation control in Himalayan r e s e r v o i r s , Hater


1
Power 4 Dam Construction, Vol 33, No 10, October 1981, pp 41-43.

7.20 Babbitt, M.S., Doland, J . J . and Cleasby, J.L. Water supply engineering, Chapter 8
6th e d i t i o n , McGraw-Hill, New Vork, 1962.
EXAMPLES OF MODIFICATIONS TO INTAKES TO IMPROVE
SEDIMENT CONTROL
7.21 Twort, A.C., Koather, R.C. and Law, F.M. Hater supply, 2nd e d i t i o n ,
ftrnold, London, 1974.

102
8. EXAMPLES OF MODIFICATIONS TO INTAKES TO IHPHOVE SEDIiCMT CONTROL
Euphrates (low 2000 m J / s
Many factors which must be considered in the design of intake structures Canal flow C! m 3 /s
Water level 32-Sm
with regard to sediment control including types of intake structures and
sediment excluding and controlling devices have been cited in previous chap-
ters. Often, however, the problems associated with sediment Ingress into
) Existing bank line
intakes are not appreciated until the system is in use. This chapter gives
examples of studies which have been carried out in an attempt to alleviate
sediment problems at existing intakes. Sheet piling divide wall
'Top EL3t'00m

(1) The Greater Kussaylb irrigation system in Iraq was completed in 1956. Material accumulated on bed
Problems of sedimentation and poor distribution were investigated and
designs for the remodelling works were completed in 1982 by Binnie and
Partners (Ref. 8.1).

The principal maintenance problem lies in the smaller canals where


the sediment carrying capacity is less than the main canal. Sedimenta-
tion is particularly severe in the head reaches of the branch canals.

It was proposed that the amount of fine sand and coarse silt Fig. 8.1 Layout and flow pattern at remodelled headworks
entering the system would be controlled by:
ft 1:50 scale model was constructed to examine the performance of the
remodelling the headworks approach channel geometry so as to existing headworks and then various modified layouts.
abstract preferentially the surface flow in the river (Fig.
8.1). The layout shown in Fig. 8.1 was found to give best all-round
performance and was adopted for detailed design. The vertical walls
provision of a settling basin in the Main canal downstream of will be constructed mainly in sheet piling. Its principal features are:
the headworks.
realignment of the upstream river bank to produce a smooth
Stopping the finer material from entering the canal system would be approach alignment with no indentations,
uneconomic. The head available at the intake was less than 0.5 m - this
limited the options for remodelling. No alternative location was avail- a submerged s i l l projecting Into the river to train the bedflow
able for constructing a new headworks. "^ away from the approach channel,

a clearly defined approach channel which maintains the velocity


of approach to the head regulator sufficiently high to avoid
sediment deposition. The entrance is aligned with the river
flow,

104 105
a divide wall to prevent circulation of flow in the approach In the first 16 months after recommlssioning the upper gabion weir
channel, was cleared of gravel once and both weirs were in satisfactory condition
having passed some high floods. No gravel choking at the Intake had
a curved nose at the junction of the downstream bank wall with been experienced. The air release chamber was operating satisfactorily.
the sill to reduce local turbulence and scour downstream.
The improvements are shown in Fig. 8.2.
The settling basin was designed using Vetter's equation as a single in-
line basin, to remove 90? of fine sand with a main canal flow of
40 m^/s" 1 . It is estimated that a maximum volume of 250,000 m^ of
sediment will be removed from the basin annually. Bar screens Spillway and cut-off wall

The operation and control of the system with respect to supply and
Intake chamber
sedimentation has been investigated thoroughly and appropriate recom- scour valve
mendations made (fief. 8.1).
Filter rose on
intake pipe
take weir crest
(2) Sediment transport and deposition problems at headworks in Indonesia
have been the cause for 65t of the rehabilitation works necessary in Stop log slots
hydraulic structures. Ref. 8.2 discusses a number of case histories
indicating how problems caused by changes in river regime and sediment
transportation have been overcome.

(3) At the Glenfinnan hydro-electric power schemes' Stalker's Intake it was


found that the intake chamber was completely filled with moraine which
had washed down from the hillside above. The improvement work comprised
removal of the original intake structure and blanking off the pipe,
Original intake
construction of an overfall intake chamber that would allow stones and chamber removed
gravel to pass over it without choking, and connection of the new
intake chamber to the turbine pipe. In aadition, two gabion weirs were
constructed at locations upstream of the intake to form gravel traps. Bar screen with leaf screen panels
Intake chamber I new)
Original concrete
With the new intake arrangement, entrained air was being drawn forming intake
weir
into the pipeline, resulting in rough running of the Turbine at high
flows and, in particular, causing trouble with the domestic water supply -Disc flushing valve
through a range of flows. A separate air release chamber adjoining the
intake was added to solve this problem.

Cross-section of Intake chamber A-A


Fine screen panels were introduced for SiaoJiial attachment over
the gravel screen bars to prevent blockage of ;he intake rose by leaves
and grass during the autumn/winter season. Fig. 8.2 Glenfinnan power schemes - Stalkers intake Improvements

106 107
(4) Irrigation water for the Ganges-Kobadak Project is withdrawn from the 8.3 Sarker, A.B., Klaassen, G.J. and Radu, M.Th. "Reducing sediment prob-
Ganges River about 1 mile downstream from the Hardinge Bridge. Since lems of the Ganges-Kobadak Irrigation Project (Bangladesh) - Proc 1st
the start of the project, in the early 60's, problems have been Inst Conf on Hydraulic Design in Water Resources Engineering : Channels
experienced due to the presence of sand banks In front of the intake and and Channel Control Structures, University of Southampton, April 1981.
excessive sedimentation in the intake canal which necessitates extensive
maintaining dredging works.

The sedimentation problems were studied recently (Ref. 8.3),


within the framework of a study for the rehabilitation and the improve-
ment of this irrigation project. It was found, from a number of
different data, that bars with a width of more than 1 km are traversing
through this particular stretch of the Ganges River, and that the quan-
tities of sediment to be dredged yearly could be correlated with this
bar movement.

After a careful analysis of the technical and economical aspects


of the various possible measures, recommendations were made for the
improvement of the present sediment handling procedure by gradually
dredging the channel through the sand bar, if any, and for increasing
the width and depth of the settling basin to encourage sedimentation
and, therefore, to reduce the sediment entering the main irrigation
canals.

In addition to this new model tests are recommended to study the


effectiveness of floating guide vanes in the Ganges River in front of
the intake, taking into account the various river bed configurations
which have been observed in the past.

References

8.1 Bramley, M.. and Thompson, G. "Design of the Remodelled Headworks and
Canal System for the Greater Mussayib Project" - Proc 1st Int Conf on
Hydraulic Design in Water Resources Engineering : Channels and Channel
Control Structures, University of Southampton, April 1981.

8.2 Memed, H. "Some methods to overcome the sediment transport problems on


headworks in Indonesia" - Proc 1st Int Conf on Hydraulic Design in Water
Resources Engineering : Channels and Channel Control Structures,
University of Southampton, April 1984.

109
108
T
APPENDICES
APPENDIX 1 : DESCaiPTIOMS OF TYPICAL INTAKE FEATURES Under Sluice, Scour or Bottom Outlet A low level channel controlled by a
sluice gate for passing bedload from the headpond through the weir, usually
located adjacent to the intake s i l l . Under sluice also passes near-bed flow
Forebay Open area of water upstream of the intake structure. with higher suspended sediment concentration.

Floating Boom Situated upstream of or in the forebay for diverting flotsam. Settling Basin or Sand Trap A basin incorporated in or close to the intake
structure, where low velocity causes some or a l l of the sediment load in
Weir Low, overflow type control structure constructed across the river to suspension to s e t t l e out of the flow.
divert flow into the intake.
De-Silting Channel or Canal Channel by which deposited sediment can be
Barrage Gated control structure. scoured out.

Headpond Pond formed by river weir or excavated or a reservoir to provide De-Silting Sluice Gate Controls flow to the de-silting channel.
head at the intake - can also provide primary sedimentation.
Transition Gradually varying channel cross-section within the intake struc-
Divide Groin/Hall Runs upstream from the weir or barrage to separate flows ture, to control the inlet velocity and maintain streamline flow in the water-
in the approach channel. ways associated with the intake.

Inlet or Entry Sill A step across the face of the intake structure or settl- Intake Gate or Regulator ft gate capable of operating against unbalanced head
ing basin to permit abstraction of less heavily sediment laden near-surface that can be used for shutting off and controlling the flow from an intake
flow. structure into the conduit.

Skimmer Wall A hanging wall across the face of the intake structure to Conduit or Channel Pipe, tunnel or canal for conveying the supply of water
prevent ingress of flotsam. from the intake.

Splitter Piers Form a series of narrow openings In the face of the intake
structure so that large flotsam is excluded from the face of the screens
(- also provide ability to shut off individual pump supplies for maintenance
if walls connect piers to sump walls).

Coarse Screen or Trash Rack Set in the face of the intake structure to
exclude surface and sub-surface trash, frazil ice, etc. usually with provision
for clearing by raking.

Sluice Gate Generally behind the coarse screen in the inlet to the intake
structure, for closing off the structure from the source, i.e. river or reser-
voir.

112 113
ftPPEHDIX 2 : SEDIMENT TRANSPORT THEORY material at very low velocities, some particles begin to
move with the increase of velocity by sliding, rolling or
hopping along the bed (bed load); at still higher velocities
A2.1 Scope particles from the bed are thrown into suspension by
turbulence {suspended load). The suspended load also in-
There is a very large quantity of literature on sediment transport cludes finer particles in near permanent suspension brought
theory and related issues such as alluvial channel resistance and river and in from the catchment (wash load). Bed load and suspended
canal regime. The scope of this Appendix is limited to a summary of those load may occur simultaneously, but the borderline between
aspects most relevant to sediment control at intakes, selecting a few of the them is not well defined.
many available formulations that will be of value in considering the general
transport of bed materials and solids in suspension in the river, through the Total load : From the view point of transport of
intake works and hence the distribution system. Local Influences generated by sediment, the total load comprises bed load and suspended
the three-dimensional nature of the flow, at the river/intake interface are not load, the latter including wash load. From the view point
covered. The situations referred to in the basic theories of sediment trans- of origin of the sediment, the total load comprises the bed
port are with steady uniform flow in straight channels. material load (including the suspended portion) and the wash
load {see Fig. A2.1).
A2.2 Terminology and Definitions
Bed material : The material, the particle sizes of
Recent International and British Standards (fief. A2.1) are available and which are found in appreciable quantities in that part of
the following extracts from BS 3680 10B (1980) provide the accepted the bed affected by transport.
terminology and definitions, illustrated also in Fig. A2.1.
Bed material load : The part of the total sediment
transport which consists of the bed material and whose rate
of movement is governed by the transporting capacity of the
r-Along the bed- -Bed loadi
channel.
(Bed material load L^Jotat load
(Transport) Suspended load : That part of the total sediment
Total load >-ln sgspension-i transported which is maintained in suspension by turbulence
(Origin!
s-Suspended load- 1 in the flowing water for considerable periods of time with-
-Wash load in suspension-
out contact with the stream bed. It moves with practically
the same velocity as that of the flowing water. It is
generally expresses in mass or volume per unit of time.

Bed load : The sediment in almost continuous contact


Fig. A2.1 Modes and definitions of sediment transport with the bed, carried forward by rolling, sliding or hopp-
ing.
"For a proper comprehension of sediment movement and related
terms, the flow of water over an a r t i f i c i a l l y flattened bed Hash load : That part of the suspended load which is
of sediment may be considered. From no movement of bed composed of particle sizes smaller than those found in

M4
I 115
appreciable quantities in the bed material; i t is in near where the left-hand side is the Einstein (Ref. A2.2) transport function 4>, q t
permanent suspension and, therefore, is transported through is the sediment transport rate as submerged weight per unit width per unit
the stream without deposition. The discharge of the wash time and ys is the submerged unit weight of the solid phase.
load through a reach depends only on the rate with which 1/2
these particles become available in the catchment and not on A.6
the transport capacity of flow. It is generally expressed
in mass or volume per unit of time," These parameters, D , Y, Z, s and 0 are found in various combinations in many
theories of sediment transport.
A2.3 Fundamentals of Bed Material Hoveaent
There is a special case of the general function in eqn A.5, namely the con-
The basic quantities which influence the process of sediment transport dition at which bed material is just on the point of movement with q t = 0.
in two-dimensional, free surface flow are the unit mass of fluid p, the unit
mass of solids p kinematic viscosity of the fluid \>, particle diameter D, 0 = f <D gr , Y c r , Z, s) A.7
water depth d, shear velocity / ( T /p) or /(gdS) denoted v. and acceleration
due to gravity g. Dimensional analysis yields the following grouping of these In practice, the relative density is fully accounted for in the parameters D
gr
basic quantities: and Y, and so may be omitted. This 'initial otion1 condition is then usually
expressed in the form
Dimensionless grain size
1/2 3/2
Y D 3 1/3
Yor = fo (Ycr Dgr ,' Z)
' A.8
1/3
x D A.I
The ratio, Z, of depth to sediment diameter is of very minor significance
Mobility number except in relatively shallow flow, and so equation A.8 simplifies to the
Shields function (Ref. A2.3).
Y= " (s-1) gD
A.2

(s-1) g D = f A.9
Relative grain size or dimensionless flow depth
This is shown in Fig. A2.2 with various sources of data and an additional set
A.3 of grid lines related to constant sediment properties.

Relative mass density, solid/liquid

A.M

Hence, any property related to the movement of bej material in steady, uni-
form, two-dimensional flow is a function of the&8 Coir dimensionless groups,
in particular, the dimensionless sediment transport parameter.
1/2
V
-T7-r-5 = f (D_ Y, Z, s) A.5

116 117
A2.5 flckers and White Method
o AMBER >06
LIGNITE 1-27 A2.5.1 Theoretical basis:
GRANITE 2-7
o BARITE i.K
Fully developed SANO ICASEYI 25
turbulent velocity profile + SANO 1 KRAMER) 2-6S
This theory first considered coarse sediment and fine sediment
s 1-0 SAND (US.WESI 2-65 separately, and then sought a transitional function between them. The tran-
SANO [GILBERT! 2-65
SAND 1 WHITE) 2-51 sitional sizes include the sands and silts of great practical interest in
a* Turbulent boundary layer a SANO IN AIR 2-M
rivers and irrigation systems.
If Value of jo-1(X^ - i ) g O
> SANO SHOT 7-9

10 20 60 100 200 tOf 600 1000


c S g-i i ft coarse sediment is transported mainly as a bed process. If bed fea-
tures exist, i t is assumed that the effective shear stress bears a similar
relationship to mean stream velocity as with a plane grain-textured surface at
Shields curve
rest.
0-02
1 10 100 1000
Boundary Reynolds No,
A fine sediment is transported within the body of the flow, where it is
suspended by the stream turbulence. As the intensity of turbulence is de-
pendent on the total energy degradation, rather than on a net grain resis-
Fig. A2.a Shields function for incipient motion tance, for fine grained material the total bed shear stress is effective in
causing sediment transport.
When Y exceeds Y c r , then sediment transport becomes established and the trans-
port function 0 increases progressively (i.e. q t increases) as the power of Sediment mobility is described by the ratio of the effective shear force
the stream increases diagonally upwards from the initial motion condition. on unit area of the bed to the immersed weight of a layer of grains. This
The assessment of sediment transport thus requires knowledge of how 0 varies mobility number is denoted F , and a general definition is:
from 2ero as one moves upwards and to the right in Fig. 112.2.
ii-n
A.10
gr
(12.4 Available Formulae for Bed Material Load /{gD(s-i)} /32 log (ad/D)

For coarse sediments {n = 0 ) , the expression reduces to the form


Ref. h2A reviewed available formulae (as in 1975) in the context of
laboratory and field data and concluded that few were of acceptable 1 A.11
F =F
gr eg
accuracy. The best were the methods of flckers and White (Ref.ft2.5)and of {gD(s-l) /32 log (od/D)

Engelund and Hansen (Ref. A2.6), the former being slightly more accurate and The coefficient a relates the roughness, k g1 in the rough-turbulent formula-
the latter being the simpler. Only these two formulations will be quoted, but tion for channel resistance to the median sediment diameter, D. For fine
other methods will be mentioned later. These include a purely empirical sediments (n = 1 ) ;
method for sand bed rivers and procedures whereby field information on
material in suspension may be used to assess the unmeasured transport close to ft.12
F - F = /Y
the bed.

The transition parameter n was evaluated by analysing flume data for a range
of sediment sizes and was expected to lie between 0 and 1, showing a con-
tinuous variation through coarse material and transitional sizes to fine non-
cohesive sand or silt.

118 1 19
Sediment transport is based on the stream power concept; in the case of The type of relationship suggested was a power function of G with {F
coarse sediments the product of net grain shear and stream velocity is used as - A ) , where ft is the value of F at which motion f i r s t s t a r t s .
the power per unit area of bed, and for fine sediments the total stream power
is used. The useful work done in sediment transport in the two cases takes - 1) ft.17
account of the different modes of transport assumed, and in relation to the
stream power gives an expression for the efficiency of the the transport A2.5.2 Calibration and Evaluation
process. Efficiency is expected to be dependent on the mobility number,
c
gr-
Clearly there would be a value of F below which no sediment would move Using very many sources of data, a computer-based optimisation routine
and efficiency would be zero. fts F_r rises above the limiting value, ft , it was developed to obtain best-fit values for the four empirical parameters, J,
is expected that the efficiency would increase. ftgr, m1 and n. The variation of these with the dimensionless grain-size D ,
was determined and algebraic functions were fitted to the data for computa-
In order to separate the primary variables, the efficiency (which is tional convenience.
dimensionless} was combined with the mobility number, F , to yield a general
transport parameter: For transitional sizes, 1 < D < 60 :

Xd
ft.13 n = 1.00 - 0.56 log D, A.13
'gr

where fl = _ o ^ i + o . 1it A.19

X = gs/pgvd A. 14 A.20

X is the mass rate of sediment transport per unit width expressed as a ratio log J = 2.86 log D gr - (log D g r ) 2 - 3.53 ft.21
of the mass rate of fluid flux per unit width, which it is convenient to think
of as a special form of sediment concentration, g s is the weight of solids in For coarse sediments, D > 60:
motion, per unit time.
n = 0.00 A.22
The general transport function tested was expressed as:
A g r = 0.17 ft.23
G
gr = f 1 V ft.15
m' = 1.50 A.24
i . e . sediment transport is a function of sediment mobility and grain size, fts
the definitions of G gr and F depend on the transition parameter, n, i t was J = 0.025 A.25
also necessary to confirm that
A round figure value for a of 10 was estimated, not significantly different
n = fo A.16 from the value of 12.3 with the roughness defined by median grain size in the
rough-turbulent equation. The transport function is plotted in Fig. A2.3.
Referring back to the functional relationship of eqn ft.5, i t will be seen that
this framework is a special case, reducing the five-variable form to a three-
variable form which was more amenable to analysis.

120 121
5. Convert G g r to sediment flux X (the effective mass concentration in the
fluid flux), using equation A.13, and the weight rate of transport using
equation A.14.

A2.5.4 Graded Sediments

Although the theory was originally checked against flume data for sedi-
ments with narrow gradings using the D 5 0 size, it has since been compared with
the data from the laboratory and the field for graded sediments. For sedi-
ments with only a modest range of particle sizes, say D g ^ / D ^ <5, it was found
that the total transport of bed material could be related to the 35 percentile
5 10 50 100 and consequently the particle size D should be interpreted as D^g for these
Dimension less grain size , Dgr
applications.

With widely graded sediments it is possible to follow the procedure


Pig. A2.3 Ackers and White sediment transport function recommended by Einstein (Ref. A2.7) and others, in which the bed material
grading curve is used to consider, say, ten size fractions separately. In
A2.5.3 application of Method this ease, however, the assumed threshold conditions for each size fraction
should take into account the shielding of the finer fractions and the addi-
To calculate the sediment flux the following properties of tile system tional exposure of the coarse material when surrounded by these finer frac-
must be known: (1) Particle size, D; (2) relative density of sediment, s; tions. The calculation procedure in this case is as follows:
(3) mean velocity of flow, V; (H) shear velocity, v*, based on measurements
of the velocity distribution or the depth/slope relationship {v, = /(gdS)); 1. Consider first the coarsest fraction; from the upper and lower bounds of
(5) depth of flow, d; {6) kinematic viscosity of fluid, v; and {7) accelera- the fraction determine the mean diameter of the fraction; from the
tion due to gravity, g. grading curve determine the percentage of the bed material sample in the
range.
The calculation proceeds as follows:
2. Follow steps 1-5 of the basic procedure given above except that at step
1. Determine the value of D from known values of D, g, s and v 2 a modified value of A, say A 1 , should be taken. A 1 is given by the
(equation A. 1). expression A ] /A = ( t V ^ ) " 0 ' 2 .

2. Determine the values of n, A, m1 and J associated with the derived D 3. Factor the transport of this fraction by the sample percentage/100.
value (equation A. 18 - ft.25).
4. Proceed to the next coarsest fraction, and so on to the finest. Note
3. Compute the value of particle mobility, F (equation A.10). that any material below 0.04 me in the bed sediment grading curve is to
be excluded, because it will travel as wash load to which the method
k. Determine the value of G from equation A.17 or from Fig. A2.3 which does not apply.

represents a graphical version of the new sediment transport function.

122 123
5.

A2.6
Total the factored transport rates for all fractions,
estimated bed material load.

Engelund and Hansen method


This is the total
1
meters of between 0.1 and 0.8 mm, depths from 0.3 m and velocities up to
3 m/s.

Colby's method consists of the simple use of curves based on the corre-
lation of sediment transport data obtained from field measurement. These
In terms of goodness of fit to laboratory and field data, there is not a prediction curves are shown in Fig. A2.1, LIIC! provide the bed material load in
great deal to choose between the Engelund and Hansen method (Ref. A2.6) and kg/day per metre width for known values of depth and mean stream velocity.
the Ackers and White method 1 (Ref. A2.5), the latter being marginally The application of this method is illustrated in section 6 . 1 .
superior but more complicated algebraically.

Even the best of the currently available theories are only able to
assess transport rates from the hydraulic parameters of a channel to a two-
fold accuracy for about two-thirds of the data, however. This disappointing
but nevertheless realistic conclusion arises from scatter in the data and from
the sensitivity of transport rate to the/hydraulic variables, as well as from
imperfections of the theories.

The Engelund and Hansen formulation is remarkably simple:

v 5/2
J
* = 0.1Y A.26
where
\ = 8gdS/V A.27

(* and Y are defined by eqns A.6 and A.2 respectively)

In their original paper, Engelund and Hansen gave a graphical solution which
facilitated the determination of both the water flow and sediment transport
rates. The function is dimensionally homogeneous, and utilises the median
grain size D 5 Q as the characteristic sediment diameter. It has a tendency to
overestimate transport at low shear rates, but performs well with laboratory
flume data and field results.

A2.7 Colby method 0-3 1 3


V m /sec

The method proposed by Colby (Ref. A2.8) was not included in the com-
parisons made by White, Milli and Crabbe (Ref. A2.H), because it consists of Fig. A2.4 Colby's curves for sand transport
empirical sediment transport curves for sands rather than formulae capable of
numerical testing. It is nevertheless a basically sound method of assessing
the bed material load in sand bed rivers and canals, with median grain dia-

124 125
= ratio of sediment diffusion coefficient to momentum coefficient (assumed
A2.8 Transport in Suspension
1.0}

In a steady uniform flow carrying sediment in suspension, under


This function is given in Fig. A2.5. This is taken from Ref. A2.10, which
equilibrium conditions the concentration at any level will be constant and the
net vertical flux at any level will be zero. The settling out of sediment due provides further information on the subject.

to the weight of the particles will be balanced by the upward flux due to
turbulent diffusion from lower layers with a higher concentration. Surface

The settling rate per unit plan area is Cw where C is the local con-
centration and w is the terminal velocity of the individual particles. Thus

dC
Cw A.28

where the second term is the vertical diffusive transport due to the concen-
tration gradient dC/dy, t being the diffusion coefficient for sediment.

Rouse (Ref. fl2.9) developed the above concept by combining it with the
basic theory of turbulent flow in an open channel. The assumption was made Bottom Relative concentration ,
that the diffusion coefficient for sediment is proportional to the momentum
diffusion coefficient in the mixing-length theory of turbulent flow. Utilis-
ing also the corresponding flow parameters, e.g. the bed shear stress and Fig. A2.5 Graph of Rouse suspended load distribution equation for

overall resistance of the fluid phase, Rouse derived the following function a/d = 0 . 0 5 and several values of z

for the vertical distribution of suspended material.


Crucial to any consideration of sediment behaviour is a knowledge of the
fall velocity of discrete particles. This may be determined from the Rubey
y (d-a) A.29
equation {Ref. A2.ll) or a graphical presentation as in Fig. A2.6 {from Ref.

where the Rouse number z, is given by: A2.12). The Rubey equation is for natural particles:

?
SKv,
A.30 w = /gd 3
gd3 (s-1)
t -.-*> V , ft.31

In the above: In the above, \J is the kinematic viscosity of the fluid which is a function of
water temperature, as follows:
C - concentration at elevation y above the bed
Cg = concentration at a reference elevation a above the bed
Temp C : 0 5 10 15 20 25 30
6
d - flow depth M, 10" mVs: 1.79 1.53 1.32 1.15 1.02 0.91 0.81
w = fall velocity
vM = shear velocity (= /gdS) Fig. ft2.6 includes the influence of shape, with the shape factor designed
K - von Karman constant (0.4) as SF = a//(bc) where a, b, and c are the mutually perpendicular dimensions (a
being the smallest).

126 127
Methods of estimating the suspended-sediment discharge in the unsampled
layer near the bed and the bed load discharge have been proposed. The Colby
and Hembree method (Ref. A2.13) is known as the Modified Einstein Method and
is based on observed suspended load samplus, me&ii velocity, depth, cross-
section, and size composition of the bed nedlment. Its name derives from the
fact that it uses a modified version o:' the Einstein bed load function {Ref.
A2.14) in estimating the unmeasured .jedunair; discharge. Sediment discharge
data based on suspended-sediment samples and estimates of the unmeasured
discharge are much more reliable than those based on the formulae alone. The
methods of estimating unmeasured sediment discharge are presented in detail in
Ref. ft2.10.
o.
SF-0-5
100 SF-0-7 A2.10 Sedinent Transport Through Lined Channels. Pipes and Culverts

10 50 SF-0-9
Rill velocity, cms/sec Most treatments of sediment transport in the civil engineering context
relate to channels with an alluvial bed. Bed material transport is then not
restricted by supply : if the local stream power exceeds that needed to convey
the incoming sediment load, the stream will pick up extra material from the
bed to achieve a new balance. Over the concrete bed of an artificial channel

Fig. A2.6 Relation between sieve diameter, fall velocity and or within a pipeline, this balance does not exist if there is no deposition to

shape factor (SF) for naturally worn quartz sand particles form a cover of alluvium over the solid boundary. Several stages of sediment

falling in distilled water transport may occur:

A2.9 Calculation of Sediment Discharge froa Stream Measurements 1. Supply less than the transporting capacity of the flow over a clean
solid surface.
The development of the depth-integrating sampler (see Chapter 2 and
Section A Ref. A2.10) made it possible to sample the suspended load of streams 2. Supply Just in balance with the transport capacity so there is incipient
on a routine basis. In many countries there are national agencies responsible deposition.
for collecting data of this kind as well as hydraulic data on streams but
where data is not already available, it may be collected specifically for a 3. Supply exceeding the incipient condition so that deposition occurs.
project using standard methods (e.g. Ref. A2.1). Suspended load samplers This deposition, through changing the available cross-section and in-
cannot sample the flow within a layer several centimetres thick at the bed and creasing roughness, will alter the flow conditions and may reduce the
do not include any of the bed load. For fine sediments, e.g". those making up "^ sediment transport capacity below that of stage 2 and so deposition may
the wash load, which are nearly uniformly distributed over the depth, the accelerate.
concentrations measured are essentially equal to the discharge concentra-
tion. However, for the bed sediment that is coarser than the wash load, the J*. Transport capacity over the deposited bed in balance with the supply.
concentrations in the samples are usually considerably less that the true mean
concentration in the flow. To get the true sediment discharge it is necessary Condition 2 above is of crucial importance. Where the flow and/or sediment
to add the contribution of the unsampled suspended load and the bed load to supply varies in time, the system has either to be designed securely in the
the measured sediment discharge.

I28 129
range of condition 1 so that the lined channel, pipe or culvert remains per- References
manently free of deposition; or it should be designed for condition 4, with
allowance made for the influence of deposition on the capacity of the system A2.I British Standards. Methods of measurement of liquid flow in open
to transport both fluid and sediment. Conventional sediment transport theory channels BS 3680, Part 1, 1983, Glossary of terms; BS 3680, Part 10,
and methods apply to the l a t t e r zone, but special equations apply to the Sediment transport; Part 10B, 1980, Measurement of suspended sediment;
former zone, in effect providing the limiting condition under which the maxi- Part 10C, 1980, Bed material sampling.
mum supply of sediment may be conveyed without deposition over a solid sur-
face. A2.2 Einstein, H.A. The bed load for sediment transportation in open channel
flow, Ser Tech Bull US Dept Agric Soil Conserv, 1950, Ho 1026.
The subject of pipe-line transport i s reviewed in Ref. A2.10, and also
by Graf (Ref.A2.15) who has himself developed the limiting function A2.3 Shields, A. "Anwendung der Aenlichkeitsmechanik und der
Turbulenzforschung auf die Geschiebebewegung", Mitteilungen der
= 10.39 P A.32 Preussischen Versuchsanstalt fur Wasserbau und Schiffbau, Berlin,
Germany, translated to English by H P Ott and J C van Uchelen,
where C y is the volumetric sediment concentration, R is the hydraulic radius California Institute of Technology, Pasadena, California, 1936.
of the flow section and V s is the flow velocity at incipient deposition.
A2.4 White, W.R., Milli, H. and Crabbe, A.D. Sediment transport theories, a
A more recent treatment of the subject by Hovak and Malluri (Ref. A2.16) review, Proc Instn Civil Engineers, Part 2, June 1975, 59, 265-92.
yields as the limiting condition for depositionr
2/3 A2.5 Ackers, P. and White, W.R. Sediment transport: new approach and
flii/a&i!^ analysis, Proc ASCE, November 1973, 99 (HY11), 2041-2060.
Ml.6 n

Laboratory experiments by May (Ref. A2.17) gave the limiting velocity Vs as a A2.6 Engelund, F. and Hansen, E. A Monograph on Sediment Transport in
function of Cv, for pipes of diameter D . Alluvial Streams, Teknisk Vorlag, Copenhagen, 1967.
D 2 0.6 Vs 3/2 V
o "^
= 0.0205 [f) (jf) { A.34 A2.7 Einstein, H.A. Bed load transportation in Mountain Creek, US Dept of
g fs-1) D s
Agriculture, Washington DC, 1941.
in which Vo, the limiting condition for zero transport, is given by Novak and
Nalluri's formula H2.8 Colby, B.R. Discharge of sands and mean velocity relationships in sand-
bed streams, US Geological Survey, Professional Paper 462-A, US
VQ = 0 . 6 1 (s-1) (5 "0.27 A.35 Government Printing Office, Washington, 1964.

These various methods give rather different results, especially when applied A2.9 Rouse, H. Modern conceptions of the mechanics of fluid turbulence,
to large systems well beyond the range of experimental data. Designing for Trans ASCE, Vol 102, paper 1965, 1937, pp 463-543.
operation without any deposition should recognise the uncertainty in the
methodology and so include something in reserve. 12.10 Vanoni, V.A. (ed), Sedimentation engineering, ASCE Manuals and reports
on engineering practice no 54, ASCE, 1975.

130 131
A2.11 Rubey, N.W. Settling velocities of gravel, sand and s i l t p a r t i c l e s , ftPPEMPIX 3 : ElflHPLE CALCULATIOH
ftmer J of Science, 5th series, vol 25, no 148, 1933, p 325-338.

R2.12 interagency Committee. Some fundamentals of particle size analysis, a For the purpose of this calculation a converging curved channel excluder
study of methods used in measurement and analysis of sediment loads in is assumed, with a layout generally as shown in Fig. 6.1.2.
streams, Report No 12, Subcommittee on sedimentation, Interagency com-
mittee on water resources, St Anthony Falls Hydraulic Laboratory, Data Canal Flow: 20 arVs
Minneapolis, Minn, USA, 1957. Sluice Flow: 7 nrVs at low flow conditions and average water
availability
A2.13 Colby, B.R. and Hembree, C.H. Computations of t o t a l sediment discharge, River Bed Sediment: A median diameter of 0.5 mm is assumed
Niobrara River near Cody, Nebraska, Hater Supply Paper 1357, US with a typical size distribution
Geological Survey, Washington DC, 1955.
t Main dimensions have to be assumed, and the design checked.
A2.14 Einstein, H.A. The bed load function for sediment transportation in
open channels, Technical Bulletin 1026, US Dept of Agriculture, Soil (a) Depth of flow: For low flow conditions the canal headgates will
Conservation Service, Washington DC, 1950. be open and i t may be assumed that the depth of flow i s con-
trolled by flow through the canal head regulator.
A2.15 Graf, W.H. Hydraulics of sediment transport, McGraw-Hill, New York,
1971. The discharge intensity, q, at the head regulator is 20/(4
x 5), i . e . 1 m3/s/m. The minimum upstream level in the sluice-
A2.16 Novak, P. and Nalluri, C. Sewer design for no sediment deposition, Proc way will be determined by the head required to pass t h i s d i s -
Instn Civil Engineers, Part 2, Vol 65, September 1978, p 669-671*. charge through the canal head regulator. Assuming t h i s is
equivalent to the discharge over a broad-crested weir.
A2.17 May, R.W.P. Sediment transport in sewers, Report IT222, Hydraulics
Research Station, Hallingford, February 1962. i.e. q = 1.706 H 3/2

H = (1/1.7O6) 2 / 3 = 0.70 m

For a skimming weir height of 0.5 m, depth of flow in the curved


channel will be approximately 1.2 m. This would have to be
maintained by controlling the sluice gate opening.

132 133
(b) Assumed inner curve radius = 20.85 m ( r i ) v. = 0.016 m/s
assumed outer curve radius = 29.It ra ( r Q )
Then,
12.
Q/d = In rWr.) V = 2.5 v . In 6.1.4

Opposite central headgate pier, Q = 17.0 nrVs approximately. To e s t a b l i s h value of effective roughness, a:

YY~ 20.85 In (29.15/20.85) The reference laminar sublayer thickness, & = : - (from roughness theory)

= 2.03 m/s so here S - 0.5 x 10 m


V =
o
Supposing the roughness height k of the sand to be 0.57
(c) The average velocity in the curved channel a t the upstream
entrance to the canal head regulator i s approximately k/6 = jL51 =

V = 27/(1.2 x 8.3) = 2.71 m/s From a graph of effective roughness coefficient, x, against k/S ,

Froude number Fr = 2.71 /T72g = 0.79 x = 1.6

0
CRITICAL SHEAR STRESS APPROACH Effecive roughness, i = - = 0.36 x 10~3 m

To c a l c u l a t e mean velocity corresponding to c r i t i c a l shear s t r e s s , use equa- From equation 6.1.4, where R = {8.3 x 1.2)/(8.3 + {2 x 1.2)) = 0.93 m
tion 6 . 1 . 2 .
V = 0.41 m/s
> / Y - 1} g d >
S 5

From this calculation it is clear that the actual velocity in the curved
sluice channel is far in excess of that required to produce critical shear
= - 5 3 x 10 ~ (0.1 x 1.65 x 9.81 K 0.5 x 10" 3 )* conditions on a flat bed of 0.5 mm sand.
0.73 x 10" c

Check corresponding to (b) above on the section immediately adjacent to


I = 19.^9 the sluice gates:

Then from Fig. A2 (and equation 6.1.3) Approximately, discharge = 7 m 3 /s

Average velocity = 7/(1.2 x 8.3) = 0.7 m/s

It is seen that the actual velocity is about 1.7 times that correspond-
Hence, ing to critical shear stress on a flat bed of 0.5 mm sand.
v , 2 = 0.032 x 9.81 x 0.5 x 10" 3 x 1.65

134 135
A value of T,, can be computed based on an average velocity of 0.7 m/s, a APPEHD1X 4 : MODEL STUDIES
depth of flow of 1.2 m and a roughness height k = 0.157 nm; R = 0.93 m.

A value of v* = 0.027 m/s i s calculated giving Model studies of intakes, particularly of major structures, are often
necessary because the flow pattern and its interaction with sediment is com-
T,, = 0.09 plicated and defies theoretical treatment; transfer of knowledge and
experience from one case study to another is, of course, very valuable but
Of course, T , will vary throughout the sluice channel and this i s only one frequently not sufficient, because of special features of design and its
representative point. More investigation is required for t h i s c r i t e r i o n . interaction with the river banks and bed.

CHECK BY COLBY'S SEDIMENT TRANSPORT CURVES (FIG A2.4) Mathematical models are not yet able to simulate the complex local flou
patterns associated with intakes, and therefore, modelling is almost exclu-
Since the excluder in Fig. 6.1.2 was model tested and found to work quite sively carried out with scaled physical models.
satisfactorily, i t seems that the c r i t i c a l shear s t r e s s will always be exceed-
ed by a c e r t a i n factor. In fact, there may be some build up of sediment in A4.1 Objectives of Model Tests
the s l u i c e channel so that the bed i s not f l a t . An a l t e r n a t i v e approach i s to
check the sediment transporting power of the sluice flow. By interpolation The type of model to be used will be primarily determined by the objec-
from Fig. A2.t, for a flow depth of 0.82 m and a mean velocity of 1.07 m/s, tives of the model tests. These may be:
the sediment transport rate for sand having a mean diameter of about 0.4 ran i s
about 20 kg/day/metre width of channel. This corresponds to about 2700 ppm in (a) Improvement of the overall or detail performance of the intake structure
the sluice channel and a sediment concentration of 700 ppo in the approaching in relation to minimising head losses, avoiding separation, stagnancy,
flow of 27 nrP/s, which i s acceptable. vortex formation, etc..

Calculations and model studies for flood conditions and for the whole (b) Improvement of intake design in relation to sediment intrusion, design
range of sediment sizes are necessary for a thorough check on performance. of sediment exclusion devices and settling basins, design of intakes in
relation to the immediate vicinity of the intake, etc..

(c) Interaction between conceptual design and siting of the intake and the
river morphology, sediment, ice and floating debris transport, river
flow and river training works.

A4.2 Types of Models

In principle three types of models may be considered in isolation, but


more frequently jointly:

(i) two or/and three dimensional undistorted scale models of intake struc-
tures and of their details U 1 . 1 U ) , above).

136 137
{ii} three dimensional models of intakes with part of the adjacent river bank (d) as much sedimentological information as possible {sediment size distri-
and bed; these models frequently do not reproduce the whole river width bution, bed and suspended sediment load transport, etc.). This infor-
and extend only a relatively short distance upstream of the intake mation is essential for models in At.2(11) and (iii) particularly, and
(near-flow-fieid modelling). The river is often modelled with a fixed is also valuable for models in AH.2(i) so that local flow patterns and
bed but with moveable material added to represent areas where scour and velocities from the model can be applied to the knowledge of sediment
deposition are likely. The models are almost invariably undistorted but transport data and some Judgement made as to the effects of the intake
occasionally there is a small distortion between vertical and horizontal on the sediment regime and the effects of sediment on the intake per-
scales (At.Kb), above). formance, even if sediment motion is not modelled. Bed levels and bank
materials are also required;
(iii) three dimensional moveable bed models where a substantial river reach
upstream of the intake as well as the intake itself are modelled; these (e) information on river operations, ongoing or planned (dredging, naviga-
are often distorted models with only the main features of the intake
tion, etc.);
reproduced (A4.1(c), above).
(f) winter regime, ice formation and ice floes movement, etc.;
The first two categories of models may be designed as hydraulic models with
water as fluid - this is the normal procedure - but in some instances aero-
*4.4 Scaling Laws
dynamic modelling, with air flowing in a pressurised system and the water
surface reproduced by a smooth fixed or adjustable surface, may be advan-
The selection of scaling laws presents the most difficult problem with
tageous .
regard to model studies of intakes.

At.3 Data Required


The general approach to model design is the same as for modelling on
scale models the performance of hydraulic structures (models A4.2(I), above)
The information required for model studies of intakes can be classified
and/or for models of rivers and open channels (models At.2(iii), above). The
as follows:
main difficulties in modelling arise from special problems related to intakes
e.g. vortex formation and in the case of models combining the structure and
(a) design parameters of the intake itself;
river approach (models At.2(ii), above). The scaling problems related to
vortices are complex and not fully resolved as gravity, viscous forces and
(b) proposed operation of the intake (water levels; discharges, etc.);
surface tension are all involved (resulting in incompatible scaling equations
according to Froude, Reynolds and Weber laws). This means that often somewhat
(c) the maximum amount of hydrological Information which is available on the
larger than usual safety factors have to be chosen when extrapolating results
river on which the intake is to be situated (low, high flows, flow
from small model tests obtained by the use of Froude scaling laws neglecting
duration curves, etc.). This information is essential particularly for
effects of viscosity and surface tension.
models in Al*.2(ii) and (iii) but clearly the levels and velocities in
the river have a major effect on intake operating conditions. If the
The modelling of intakes with part of the adjacent flow field (fW.2(ii))
intake is to be situated in an estuary, the wave climate is also
is difficult because a compromise has to be reached between the Froude scaling
necessary;
laws applicable to the nondistorted models of the structure and the laws
governing the design of open channel flow models which need not always follow
the Froude scaling laws, and/or which frequently would lead to distorted
The outcome - apart from conventional moveable bed models - could be

138 139
a nondistorted or only slightly distorted model with possibly a fixed bed tigator is satisfied that model results show good agreement with prototype
where the sediment behaviour is simulated by feeding sediment (possibly light- data. The investigator must be confident of his ability to interpret the
weight material) into the model and using it as a 'tracer1. Aerodynamic results of this type of model to provide prototype informatior
models are another possibility. The choice of the model should be made only
after a careful appraisal of advantages and disadvantages of the various It is important for the intake designer tc discuss with the model inves-
approaches. tigator the requirements of the model study (number of alternatives to be
investigated, etc.) so that the most appropriate mudel type can be selected
Where modelling sediment movement in the main water course and the and necessary data collected to enable the model investigator to evaluate the
performance of the sediment controlling features of the intake are require- costs, time-scales and limitations of the model study.
ments of the model study the aim is to select scaling laws to ensure that
hydraulic and sediment transporting and depositing conditions in the model, Modelling assists the designer to perfect his ideas and the more designs
correctly reproduce those conditions in the prototype. Many researchers have that are proved and improved by the use of models, however simple, the better
produced theories for selection of scaling laws and some of these are describ- will be designs in the future.
ed in Refs. A4.1 and At.2. Different criteria apply to suspended load and bed
load. M.5 Exaaples of Intake Model Studies

With respect to velocities, flow patterns, etc., standard Froude scaling (1} The design of loose boundary physical models of rivers in equilibrium is
criteria are applicable and it is these parameters in the intake which govern described in Ref. At.3 as applied to an intake structure on a river.
the deposition and movement of sediment in the structure. If a material is The selection of model scales developed at HRS is described and a model
selected for a mobile-bed model or as a tracer in a fixed bed model which can study of an intake to an irrigation project on the Sabi river in
be proved to represent transport of sediment in the main water course, then it Zimbabwe is described. This was a distorted (6:1) model with a mobile-
is reasonable to assume that this material will react to the intake structure bed. The study enabled the optimum intake geometry to be selected and
and its sediment controlling features in the same way as the prototype struc- the operation of the gates best suited to controlling sediment at the
ture will. intake during seasonal changes to be recommended.

Movement of material in the main water course of the model can be check- (2) The model study of the Ambergate River Intake is described in Ref. M.k.
ed by comparison with the prototype and if the intake structure is already in
existence proving of the model can be even more accurate. The model material This model study was undertaken at BHRA to investigate the effects
properties {size, density, shape, etc.) can be changed during proving and in of increasing the abstraction from the River Derwent (England) on head
the case of distorted mobile-bed models, scales and slopes can also be alter- loss and silt movement and deposition. The increase in abstraction was
ed. This proving operation may be very expensive and time consuming and so from 91 Kid"1 to 318 M i d " 1 .
the maximum amount of prototype data (A4.3) at the outset of model design is
essential to save time at the proving stage, ft thoroughly proven mobile-bed A fixed-bed model was constructed to represent 400 m of the river
model is a very valuable design tool. at a linear, undistorted scale of 1:20. The scaling criteria were
established from the Froude law. Diakon MG102, an acrylic polymer, of
The advantages of using a light weight tracer material in a fixed-bed, particle size D = 600 \m and specific gravity of 1.19, was added to the
undistorted model are that less cost and time are required for design and model to represent bed load. Each aedimentation test was run for a
testing the model. If verification is possible and considered desirable, the period of one hour, since equilibrium was established in this time, for
single possible adjustment is to change the sediment material until the inves- a combination of river flow rates and abstraction rates.

140 I4I
The only data available concerning the river sediment was that it References
was a fine sand ( D C Q = 100 inn). Calculations were made on this basis
using Shields and White theory and it was decided that Diakon was A4.1 Avery, P. "The problems of sedimentation modelling with particular
probably the most suitable of available materials. The clients were reference to river Intake models", proceedings Conference on Hydraulic
satisfied that the sediment deposition on the inside of the bend and Modelling of Civil Engineering Structures; BHRft, University of Warwick,
across the intake mouth were representative of prototype observations. Coventry, 1982.

(3) A mobile-bed model was used, as described in Ref. Af.5 to determine the A4.2 Novak, P. and Cabelka, J. "Models in Hydraulic Engineering - Physical
required sediment intake and corresponding intake layout at a large Principles and Design Applications". Pitman, London 1981, 159 pp.
diversion canal from a river. In this case a limited quantity of sedi-
ment was to be diverted into the canal to ensure the canal remained fli.3 White, W.R. "A novel loose boundary model for investigating sedimen-
stable. The design of the Jonglei Canal in the Sudan is discussed in tation problems at an intake". Proceedings Conference on Hydraulic
the reference. Modelling of Civil Engineering Structures; BHRA, University of Warwick,
Coventry, 1982.
CO A hydraulic model was used to compare three water intake lay-outs
designed to avoid bed-load transport into the intake (Ref. A4.6). The A4.1* Avery, P. "Model Study of Ambergate River Intake". BHRA Report RR
design of the model, into which 3ediment was fed at its upstream end and 1653, March 1981.
collected and analysed at its downstream end, and the tests on the three
intake designs with their sediment excluding devices are described. AH.5 Klaassen, G.J. "Sediment Intake to Reduce Degradation in an Offtake
Canal". Hydraulic Engineering in Resources Development and Management,
(5) The model studies referred to in Chapter 8 example (1) (Ref. 8.1) used a Proceedings XVII Congress IAHR, Italy, 1979, Vol 6.
fixed-bed model with polystyrene (Dcg = 0.4 mm) to simulate the
distribution of suspended sand over the vertical. A Rouse similitude A4.6 Jacobsen, J.C. "A Comparison of Three Hater Intake Lay-outs to Avoid
criterion: ratio of particle fall velocity to shear velocity was used Bed-Load Transport into the Intake". Hydraulic Engineering In Resources
as the model scale law. Development and Management, Proceedings XVII Congress, IAHR, Italy,
1979, Vol 6.
The use of model studies has been recommended throughout this guide and most
of the examples cited in earlier chapters have used some form of physical
model - the studies are described in the references given at the end of the
relevant chapters.

142 143

You might also like