Ddis

CT3410
Design of Open-Channels
and Hydraulic Structures
Dictaat behorende bij het 3^ jaars vak Waterbeheersing
oktober 2002
ir. R A n k u m
T U Delft
Faculteit Civiele Techniek en Geowetenschappen
Afdeling Watermanagement
Sectie Land- en Waterbeheer
DESIGN OF OPEN-CHANNELS
AND HYDRAULIC STRUCTURES
Lecture Notes
College CT3410
"Waterbeheersing"
ir, P. Ankum
Oktober 2002
N.B.
Deze studiehandieiding dekt de tentamen-stof niet geheel. Tevens kunnen sommige
onderwerpen als 'leesstof' worden beschouwd.
Aangeraden wordt om het college te volgen, waar de relevante onderwerpen
worden behandeld. Ook het 'operationeel maken' van bestaande kennis staat op
het college centraal.
T A B L E O F CONTENTS
1. WATER CONTROL SYSTEMS
1.1. Introduction
1.1.1. Importance of lowlands
1.1.2. Lowland development
1.1.3. Terms in lowland development
1.1.4. Literature
1.2. Flood Control Systems
1.2.1. Floods in deltas
1.2.2. Catchment management
1.2.3. Channel improvements
1.2.4. Flood embankments
1.2.5. Flood diversion
1.2.6. Hydrology of floods
1.3. Drainage Systems in Polders
1.3.1. Polder development
1.3.2. Main drainage system
1.3.3. Polder water level
1.3.4. Water balance
1.4. Irrigation Systems
1.4.1. Irrigation development
1.4.2. Main irrigation system
1.4.3. Flow control methods
1
2
3
^
"7
9
9
10
H
12
17
19
22
25
27
29
31
2. OPEN-CHANNEL FLOW
2.1. Flow of Water
2.1.1. Uniform flow
2.1.2. Bankfull capacity
2.1.3. Non-uniform flow
2.2. Flow of Sediment
2.2.1. Sediment
2.2.2. Bedload transport capacity
2.2.3. Suspended-load transport capacity
2.2.4. Discussion on the sediment transport capacity
2.3. Scouring and Sedimentation in Cross Sections
2.3.1. Tractive force
2.3.2. Critical tractive force on the bed
2.3.3. Discussion on scouring
2.3.4. Discussion on sedimentation
37
42
43
48
49
51
54
55
57
58
59
ii
D E S I G N OF O P E N - C H A N N E L S A N D H Y D R A U L I C STRUCTURES
3. DESIGN OF OPEN-CHANNELS
3.1. Design of Earthwork
3.1.1. Design procedure
3.1.2. Side slopes
3.1.3. Freeboard
3.1.4. Embankments
3.1.5. Construction drawings
3.1.6. Tender documents
3.2. Hydraulic Design
3.2.1. Three degrees of freedom
3.2.2. Discussion on the design discharge
3.2.3. Discussion on the design method
3.3. Morphological Design Method
3.3.1. Methodology
3.3.2. Width-to-depth ratio
3.3.3. Design on basis of the morphological method
3.3.4. Example of the morphological method
3.3.5. Discussion on the morphological method
3.4. Regime Design Method
3.4.1. Introduction
3.4.2. Kennedy equations
3.4.3. Lacey equations
3.4.4. Simons & Albertson equations
3.4.5. Discussion on the regime method
63
63
64
65
66
67
71
71
72
74
76
80
82
83
86
86
87
88
90
4. SUPER-CRITICAL FLOW IN OPEN-CHANNELS

4.1. Hydraulics of Super-Critical Uniform Flow
4.1.1. Critical gradient of open-channels
4.1.2. Hydraulic parameters of aerated flow
4.1.3. Uniform super-critical flow in chutes
4.2. Design of Chutes
4.2.1, Water depth in chutes
4.2.2. Energy head in chutes
4.2.2. Transition length of chutes
93
96
97
98
99
101
5. HYDRAULIC STRUCTURES
5.1. Types of Hydraulic Structures
5.1.1. Non-regulating structures
5.1.2. Regulating structures
5.1.3. Maimer of regulation
5.2. Structural Design of Hydraulic Structures
5.2.1. Scouring protection
5.2.2. Piping below the floor
5.2.3. Uplift of the floor
5.2.4. Construction drawings
105
106
109
110
112
113
115
T A B L E O F CONTENTS
6. DESIGN OF CONVEYANCE STRUCTURES

6.1. Types of Conveyance Structures
6.1.1. Culverts
6.1.2. Aqueducts
6.1.3. Siphons
6.2. Hydraulics of Conveyance Structures
6.2.1. Energy losses
6.2.2. Hydraulic design of conveyance structures
119
121
122
123
125
7. DESIGN OF PASSIVE REGULATORS

7.1. Weirs
7.1.1. Broad-crested weir
7.1.2. Sharp-crested weir
7.1.3. Control notch
7.2. Long-Crested Weirs
7.2.1. Types of long-crested weirs
7.2.2. Duckbill weir
7.3. Escapes
7.3.1. Types of escapes
7.3.2. Side-channel spillway
7.3.3. Siphon spillway
7.4. Baffle Distributor
7.4.1. Working principle
7.4.2. Hydraulics of baffle distributors
7.4.3. Design of baffle distributors
127
128
130
132
133
135
135
140
141
143
145
8. DESIGN OF MANUAL REGULATORS

8.1. Overflow Gates
8.1.1. Drop-leaf gate
8.1.2. Vertical gate for overflow
8.1.3. Romijn weir
8.2. Underflow Gates
8.2.1. Vertical gate for submerged underflow
8.2.2. Vertical gate for free underflow
8.2.3. Radial gate
8.2.4. Crump-deGmyter gate
8.3. Hydraulics of Underflow Gates
8.3.1. Types of
flow
8.3.2. Hydraulics of the vertical gate under free
8.3.3. Hydraulics of the vertical gate under submerged
8.3.4. Hydraulics of the radial gate
147
148
148
150
151
152
153
flow
flow
154
155
160
163
9. DESIGN OF HYDRO-MECHANICAL REGULATORS

9.1. Hydro-Mechanical Regulators for Drainage
9.1.1. Flap gate
9.1.2. Tidal sluice
9.2. Hydro-Mechanical Regulators for Upstream Control
9.2.1. Begemann gate
9.2.2. Vlugter gate
9.2.3. AMILgate
167
168
170
172
174
D E S I G N OF OPEN-CHANNELS A N D H Y D R A U L I C STRUCTURES
9. DESIGN O F HYDRO-MECHANICAL REGULATORS (Continued)

9.3. Hydro-Mechanical Regulators for Downstream Control
9.3.1. AVIS gate
9.3.2. A V I O gate
176
180
10. DESIGN OF DIVERSION STRUCTURES

10.1. Objectives of Diversion Structures
10.1.1. Need for diversion structures
10.1.2. Operational objectives of irrigation diversion structures
10.1.3. Operational objectives of flood diversion structures
10.1.4. Sensitivity of diversion
10.2. Design Considerations and Design Criteria
10.2.1. Design considerations
10.2.2. Hydraulic formulae
10.2.3. Side-spillway effect
10.2.4. Minimum width of the structure
10.3. Options for Diversion Structures
10.3.1. Regulators
10.3.2. Configuration of diversion structures
185
187
189
190
196
197
201
202
205
206
11. DESIGN EXAMPLES OF DIVERSION STRUCTURES

11.1. Design of Free-Offtakes
11.1.1. Overflow free-offtake
11.1.2. Gated free-offtake
10.2. Design of Passive Diversion Structures
11.2.1. Overflow offtake with orifice-control
11.2.2. Proportional diversor
11.2.3. Proportional diversor with orifice-control
11.3. Design of Gated Diversion Structures
11.3.1. Gated offtake with weir-control
11.3.2. Overflow offtake with gate-control
209
211
215
219
223
227
231
12. DESIGN OF DROP STRUCTURES AND STILLING BASINS

12.1. Introduction
12.1.1 Types of energy dissipators
12.1.2. Design of fall structures and stilling basins
12.2. Hydraulics of Stilling Basins
12.2.1. Froude number
12.2.2. Hydraulic jump
12.2.3. Standard stilling basins
12.3. Types of Stilling Basins
12.3.1. Stilling basins with baffle blocks
12.3.2. Bucket-type stilling basins
12.3.3. Miscellaneous stilling basins
12.3.4. Selection of stilling basins
13. REFERENCES
233
235
238
239
241
243
245
247
249
255
1. WATER CONTROL SYSTEMS

1.1. Introduction
1.1.1. Importance of Lowlands
Lowlands. Some 1200 million ha of the 1500 million ha of croplands on earth are flat, and
are 'lowlands'. A major part of these lowlands consists of deltas and low-lying coastal areas
and belongs to the world's most densely populated areas. About one third of the world's
population lives less than 100 km from the coastline. Many deltas have average population
densities from 200 to more than 500 persons per km'^. For instance, the Netherlands has
about 15 million inhabitants on only 3 million ha (500 per km"^), with most of the population
in the low-lying western part (900 inhabitants per km^).
Deltas possess a high potential for agricultural and industrial development, as these
areas have natural advantages with respect to soil, topography, water and location. Extensive
civil works have to be implemented to make full use of these potentials.
It is estimated that the food production of some 900 million hectares of these flat areas
are hampered by frequent floodings, high groundwater levels, and regular water shortages.
Here, the crop production can be improved considerably by flood control, drainage and
irrigation.
Multi-discipline aspects. The development of lowlands is often more difficult and costly than
the development of 'upland floodplains'.
In lowlands, not only embankments ("dijken") are needed to protect the areas against
periodic floods by the sea and/or the rivers, but also the quality of the water creates
problems as brackish ("brak") water may affect the crops. The soils in lowland areas may
suffer from acidity ("verzuring") after reclamation or may consist of thick layers of peat
("veen") which offer a poor enviroiraient for agriculture. These soils require normally a strict
water management regime.
The Netherlands probably has the most developed coastal lowland area in the world.
It has complete water management ("waterbeheer") and flood control ("hoogwaterbescherming ").
D E S I G N OF O P E N - C H A N N E L S A N D H Y D R A U L I C S T R U C T U R E S
1.1.2. Lowland
Development
Lowland. 'Lowland' ("laagland") is defined as: "land affected by fluctuating water levels and
where human activities are already existing or are being proposed" (Miura 1994).
It means that lowland consists of tidal lands, deltas, inland marshes and flood plains.
Moreover, shallow seas and lakes may belong to lowlands when impoldering is considered.
Lowland development. 'Lowland development' is a term which covers a wide field of
technical activities. It is used for the development and improvement of land for agricultural,
residential, industrial and other purposes. Lowland development requires an integrated
approach between hydraulic engineering, hydrology, soil science, agricultural engineering,
economy, sociology, etc.
Classification. Lowland development can be classified into three methods, see figure 1.1:
tidal land reclamation ("landaanwinning"), i.e. reclamation of tidal forelands and coastal
marshes;
impoldering ("inpoldering"), i.e. reclamation of shallow seas and lakes;
flood control ("hoogwaterbeheer"), i.e. the protection of lowlands against river floodings.
Often, the lowlands have become a set of polders where all elements of the hydrological
cycles are manipulated, except of course the rainfall. I n these polders, the water levels are
controlled artifically by 'water control' ("waterbeheer"):
drainage ("ontwatering"), i.e. to evacuate the excess water from the soil;
irrigation ("bevloeiing"), i.e. to supply water to the agriculture.
But water control is also widely applied in the upland floodplains.
TIDAL LAND
RECLAMATION
IMPOLDERING
(of lake and sea)
FLOOD CONTROL
P O L D E R S
DRAINAGE
Field Drainage
Systems
IRRIGATION
1
Main Drainage
Systems
Tertiary Irrigation
Systems
Main Irrigation
Systems
Figure 1.1. Lowland Development and Water Control Systems.
1. W A T E R C O N T R O L SYSTEMS
1.1.3. Terms in Lowland
Development
Reclamation, The term 'reclamation' ("drooglegging, ontginning") is also used in the concept
of lowland development. Reclamation is defined by the International Commission on
Irrigation and Drainage (ICID 1996) as: "the act or process of reclaiming swampy, marshy,
deteriorated, desert and virgin lands, and making them suitable for cultivation or habitation;
also conversion of foreshore ("buitendijks land") into properly drained land for any purpose,
either by enclosure ("bedijking") and drainage, or by deposition of material ("ophogen")
thereon".
Tidal land reclamation. A silting-up of the tidal forelands ("buitendijkse aanslibbing") is
taking place along many sea coasts in deltaic areas because of deposition of sediments, see
figure 1.2.
Tidal land reclamation ("landaanwinning", "indijken") can be carried out when the
foreshore ("buitendijks land: kwelders, schorren, gorzen") has reached a certain elevation.
Typical engineering problems are:
- suitable elevation of tidal foreland for reclamation;
- silting-up of the outfall chaimels;
- the defence of the land against storms;
- the layout of these progressive impoldered areas.
PLAN OF TIDAL LANDS
Figure 1.2. Tidal lands with creeks.
D E S I G N O F O P E N - C H A N N E L S A N D H Y D R A U L I C STRUCTURES
Impoldering. Reclamation of shallow seas and lakes through impoldering is a very special
and expensive type of lowland development, and has been implemented so far in a few
countries with serious land problems (e.g. the Netherlands, Japan). Typical hydraulic
engineering problems are:
- the defence of the low-lying land against inundations;
- seepage ("kwel") of hrackish ("brak") yjatot;
- pump-lift drainage ("bemaling").
Flood control. Lowlands are generally located in deltaic areas which have been formed morphologically by river deposits ("rivierajzettingen") during flooding of the delta. As deltas
belong to the most productive and the most densely populated parts of the earth, flood control
is needed in many cases.
Flood control ("hoogwaterbescherming") is needed to protect the areas against external
floodings, i.e. water from outside the area. Construction of embankments (dikes, levees) is
often a part of the required measures. Typical hydraulic engineering problems related to flood
control, are:
- construction of embankments ("dijken") makes the flood levels to become higher,
- the internal drainage to the river is obstructed,
- the process of gradual land raising ("colmatage") by periodic sedimentation is halted.
Flood control methods include the construction of embankments (dikes, levees),
straightening of the river channel, widening of the river chaimel, flood diversion ("afleiding")
into another channel, flood retention ("berging") in a reservoir.
Polders. Both tidal land reclamation and impoldering create new land that can be called a
'polder'. But also the embanking of areas as part of flood control works can create a 'river
polder'. A polder is defined as (Segeren 1982): "a polder is a level area which has originally
been subject, permanently or seasonally, to a high water level (groundwater or surface water)
and is separated from the surrounding hydrological regime, to be able to control the water
levels in the polder (groundwater and surface water)".
Thus, a polder shows the following characteristics:
- a polder does not receive any foreign water from a water course, but there is only water
inflow by rain and seepage, or by irrigation intake;
- a polder has an outlet structure (sluice or pump) that controls the outflow;
- the groundwater levels and the surface water levels in the polder are independent from the
water levels in the adjacent land. These water levels are artificially maintained in order to
optimize the objectives of the polders, on basis of land elevation, soil properties, climate,
type of land use, etc.
Drainage. Swamps and other water-logged areas may have a poor internal drainage system
to evacuate rainfall. They require a good land drainage system of sub-surface and/or surface
drainage systems. Typical hydraulic engineering problems related to land drainage, are:
- subsidence ("klink") of the drained land,
- irrigation of the drained areas might become necessary during part of the year.
Drainage systems. Drainage systems consist of two components, see figure 1.3:
field drainage systems ("ontwateringssysteem"), also called 'land drainage systems'. Field
or land drainage is the collection of the excess water from the land by means of a network
of field drains. Two principal types of field drainage may be distinguished: (i) a surface
drainage systems of ditches ("greppels"), (ii) a sub-surface drainage system of pipe drains
("drainage buizen").
main drainage systems ("afwateringssysteem"). The function of the main drainage system
is to collect the water from the field drainage system, to transport the water to the outlet,
and to evacuate the water from the area.
PUMPING STATION
Figure 1.3. Main drainage system and field drainage system.
Irrigation. Many lowlands, but also many upland plains, have a rainfall pattern that does not
supply the agriculture with timely and sufficient water. Irrigation of the agricultural land is
good solution when rivers are still abundant with water.
The term 'irrigation' is often used in two meanings:
the actual watering of the crop, by means of basin irrigation ("terras bevloeiing"), furrow
irrigation ("voren bevloeiing"), sprinkler irrigation ("beregening"), etc.
the combination of the artificial supply of water for the use of agriculture and the
systematic distribution over the area, with the superfluous water to natural drains, after that
as much as possible, benefit out of the water is extracted.
The second meaning of irrigation is considered here, with emphasis on the hydraulic aspects
of the water control system, see figure 1.4. Typical hydraulic engineering problems related
to irrigation are:
- discharge diversion ("aanvoerverdeling") throughout the whole irrigation scheme;
- water level control ("peilbeheer") at the diverting structures ("verdeelwerken")
- matching the total water need of the scheme ("watervraag") with the available supply in
the river ("waterbeschikbaarheid").
Figure 1.4. A n irrigation scheme with a water control system.
1.1.4.
Literature
Publications on tidal land reclamation. There are no handbooks at professional level on the
general aspects of tidal land reclamation. However, two publications discuss the technical and
historical aspects of deltaic areas:
- Volker, A . (ed.). 1993. Hydrology and water management of deltaic areas. Ministry of
transport, public works and water management, CUR-report 93-5, the Netherlands.
- Ven, G.P. van de (ed.). 1993. Man-made lowlands, history of water management in the
Netherlands. ICID, the Hague, the Netherlands.
Publications on Flood Control. There are also no handbooks at professional level on flood
control. However, a wide range of books on hydrology, hydraulics, river engineering,
irrigation and dams exist.
Publications on Polders. There is only a limited number of handbooks, textbooks and manuals
available on impoldering:
- Kley, J. van der, and H.J. Zuid weg. 1969. Polders en Dijken (in Dutch). Agon Elsevier,
Amsterdam;
- Wagret, P. 1972. Polderlands. Methuen, London;
- Luijendijk, J., E. Schultz and W.A. Segeren. 1988. Polders. In: Developments in
hydraulic engineering, volume 5 (P. Novak ed.). Elsevier, London/New York;
- Schultz, E. 1992. Waterbeheersing van de Nederlandse Droogmakerijen (in Dutch). PhDthesis, TUDelft.
- Ankum, P. 1994. Methods of Development. In: N . Miura et al (ed.), Lowlands, Development and Management. Balkema, Rotterdam, the Netherlands: 65-91.
- Ankum, P. 1996. Polders, Drainage and Flood Control. Communications of the Water
Management Department, no.69. Delft University of Technology, Delft, the Netherlands.
Publications on Drainage. There are several publications on drainage, most of them
concentrating on the field drainage aspects. The most comprehensive publication is:
- Ritzema, H.P. (ed.). 1994. Drainage principles and applications. Lecture Notes of the
International Course on Land Drainage, ILRI publ. no. 16, Wageningen, the Netherlands.
Publications on Irrigation. There is only a limited amount of handbooks on the civil
engineering aspects of irrigation:
- USER. 1974. Design of Small Canal Structures. United States Department of Interior,
Denver, USA.
- Kraatz, D.B. and I.K. Mahajan. 1975. Small hydraulic structures. Irrigation and Drainage
Paper 26, FAO, Rome.
- Bos, M . G . 1989. Discharge measuring structures. International Institute for Land
Reclamation and Improvement, Wageningen, the Netherlands.
- Ankum, P. 1995. Flow Control in Irrigation and Drainage. Conununications of the Water
1.2. Flood Control Systems

1.2.1. Floods in Deltas
Prevention of inundations. Lowlands of deltas have been formed by river deposits ("rivierajzettingen") during 'floodings' or 'inundations' ("overstromingen"). However, these
inundations have to be prevented when the deltas are becoming densely populated.
Inundations. The source of 'inundations' ("overstromingen") can be either:
poor internal drainage ("slechte interne drainage"). The inundations occurs because the
rainfall cannot be discharged properly;
external flooding ("overstroming door buitenwater"). Inundations occur because water
enters from outside the area. External flooding can enter from two sources:
o sea floods from the sea,
o river floods from the rivers.
Sea floods. The 'sea floods', or 'storm surges' ("stormvloeden"), are normally generated by
landward winds. The maximum flood levels are dependent on the amplitude of the tide, but
also on other factors like the direction and duration of the wind.
For instance, the sea tide in the North Sea differs in amplitude along the coast, with the
maximum average tide in the Southern part of 2.00 m + M S L and 2.00 m-MSL. The
maxmium sea levels during storms are much higher and the sea dikes are presently designed
for sea levels of 5.00 m + M S L , i.e. a return period ("herhalingstijd") of 1:10,000 years.
Moreover, a freeboard ("waking") of 10.00 m is added for wave run-up. Flood protection
against sea floods is not discussed here.
River floods. The 'river floods' depend on the physical characteristics of the river basin (size,
slopes, soils, etc.) and the storm rainfall ("zware regenval"). Whether these river floods will
inundate the area, depends on the capacity of the river channel. Furthermore, inundations
may also occur during a coincidence of sea floods and river floods.
time
FLASHY FLOODS
Figure 1.5. Flashy floods and gentle floods.
time
GENTLE FLOODS
Type of river floods. The river flood in a delta can be either (i) flashy flood, or (ii) a gentle
flood, see figure 1.5.
Flashy floods (Indonesian: "bandjirs") rise quickly, have sharp peaks and have a rapid
recession ("uitputting"). The water may rise as much as a few meters within 24 hours, or
even within a few hours. Flashy floods occur in small river basins with surface runoff
("oppervlakte afstroming"), especially in basins with steep slopes, impervious soils and heavy
(tropical) rains of short durations.
Gentle floods show slow rises (0.05 to 0.10 m/day) and flat flood peaks. Their duration
may extend over several months. Gentle floods occur in large river basins and with prolonged
rainfalls. Examples are the large rivers during the wet monsoon in South Asia (Ganges) and
in South-East Asia (Irrawaddy, Chao Phya, Mekong).
Flood damage. I t is often difficult to assess the benefits ("baten") of flood control projects.
In fact, the benefits are the 'non-occurrence of damage' ("niet-optreden van schade").
Damage by floods in general comprises in the first place, direct damage which can be
readily evaluated, i.e. loss of crop, damage to infrastructure and houses, etc. These costs are
often quite low. For instance, damage to the paddy means that it needs a second planting, but
land preparation is not required again.
Next comes indirect damage, like loss of income, interruption of communications,
caring of refugees, etc. Indirect damages are not so easy to estimate in costs.
Thirdly, there is 'intangible' damage ("negatieve effecten"), like loss of human lives,
public morale, and above all the limitations to land potentials.
Return period. Hence, floods are often responsible for the unfavourable economic conditions
in an area. Complete prevention of flooding is physically impossible, and flood protection can
only be attained to the degree that is technically and economically feasible.
A commonly used criterium is that of the minimum total of armual costs and damage.
In practice, it means that flood control is obtained against floods of a certain average 'return
period' ("gemiddelde herhalingstijd"). For instance, typical agricultural projects are protected
against the 1 : 20 years' flood, urban areas in developing countries ("ontwikkelingslanden")
are sometimes protected against 1 : 50 years' floods, urban areas in the Netherlands are
protected against river floods of return periods of 1 : 1800 years, etc., etc.
Flood control. Flood control is "the provision of a specific amount of protection from floods"
(ICID 1967).
More practical, flood control is the reduction of floods to a certain extent by measures
in the upper portion of a river basin and by measures on location (Volker 1991).
Flood control ("hoogwaterbescherming") is needed to protect the areas against external
floodings,
Methods of flood control. The following methods of flood control can be distinguished:
catchment management, by land management and by construction of a flood control
reservoirs ("dalafsluiting") in the catchment area;
channel improvements ("bed verbetering"), by enlarging the discharge capacity of the
river;
flood embankments ("dijken") along the river, also called 'levees' or 'dikes';
flood diversion ("afleiding") through another chaimel, i.e. a 'flood diversion channel'.
1.2.2. Catchment
Management
Land management. Floods can be reduced by agricultural measures in the catchment area,
such as afforestation ("her-bebossing"), terracing ("terras aanleg"), contour ploughing
("contour ploegen").
In general, these measures have some beneficial effect by (i) reducing the flood peaks,
and (ii) retaining a part of the sediments ("erosiecontrole"). But often, the effect is quite
small for major floods because only a part of the catchment area can be treated.
Flood control reservoirs. The construction of a reservoir ("dalafsluiting") for flood control
is normally not the appropriate solution for flood control.
Suitable dam sites are rare, and often there is still a considerable inflow into the river
below the dam.
Furthermore, reservoirs with flood control as a single purpose are seldom economically
feasible. Multi-purpose reservoirs may have conflicting operation requirements: irrigation and
hydropower require a full reservoir, but flood control requires an empty reservoir.
1.2.3. Channel
Improvements
Channel improvements. Channel improvements for flood control consist of, see figure 1.6:
enlarging the cross-sectional area ("dwarsdoorsnede") by deepening and/or widening the
chaimel bed;
shortening of the alignment ("trac"), by making cut-offs and by straightening of the
channel.
\ 7
ALTERNATIVE I
Enlarging Cross-section
A L T E R N A T I V E II
Enlarging Gradient of Channel
Figure 1.6. Flood control by channel improvement and shortening alignment.
Enlarging the cross-sectional area. The cross-section of the channel can be increased by
deepening of the channel, and/or widening of the channel bed. These measures are usually
limited in duration.
It should be recalled that the rivers created these lowlands through sedimentation. Some
equilibrium is reached between the natural cross-section and the frequency of floodings. An
enlarged cross sectional area will change the sediment-transporting capacity, and often new
sedimentation will start. It means that the life-time of the enlarged cross-sectional area is
often short.
10
DESIGN OF OPEN-CHANNELS A N D H Y D R A U L I C STRUCTURES
Shortening the alignment. Shortening of the alignment by making cut-offs through meanders
will steepen of the gradient ("verhang") of the chaimel. A steeper gradient means that the
flow-capacity of the channel is increased. Moreover, the sediment-transporting capacity is
increased as well. A check should be made, whether the channel bed remains stable against
erosion.
The shortening of the alignment appears to be quite effective in comparison with other
flood control measures. However, the possibilities for straightening depends on the present
alignment and the rate of meandering. Generally, only a small decrease of river length can
be obtained, and thus the gradient s is increased by a small ratio. Furthermore, the flowcapacity of the channel is increased by the root ("vierkantswortel") of the gradient, so byV^'.
1.2.4. Flood Embankments

Flood embankments. The application of embankments ("dijken") go far back in history, and
they are the most simple way of flood control. Embankments are also called 'levees' or
'dikes'. The design of embankments concerns two basic questions, see figure 1.7:
what is the height ("hoogte") of the embankment?
what is the mutual distance ("onderlinge afstand") between the embankment?
Height of embankments. The height of the embankment is determined by: (i) the return period
of the flood, (ii) the design flood level, (iii) and the freeboard above the flood level.
Discharge through the flood plains. The discharge through the flood plains ("uiterwaarden")
will often be small in comparison with the discharge through the channel. The flood plains
have often a high roughness, i.e. a low Strickler coefficient k, because of vegetation. It means
that the flood level is often hardly be reduced by making a wide flood plain.
The construction of embankment at some distance from the chaimel will increase
storage. This effect is also limited to flashy floods. The effect is negligible for genfle floods.
ALTERNATIVE I
Constructions of Embankments
A L T E R N A T I V E II
Construction of Flood Plains
Figure 1.7. Flood control by construction of embankments.
Scouring of embankments. Wide flood plains provide a protection against channel erosion to
the embankments. Meandering rivers do not have stable banks and scouring at the outer bends
can be expected. The constructing of embankments close to the channel will sooner or later
require expensive river training works to protect these embankments.
Geologically seen, meandering is a very fast process and the whole alignment is
changed e.g. every 300 years. The speed might be much lower from a civil engineering point
of view, when the river bend advances locally e.g. some 20 m every 10 years.
11
Effects. The constraction of flood embankments has often extensive effects on the river, and
its population living on the natural levees. A l l these effects should be considered when
designing flood embankments:
Higher flood water levels. The construction of embankments along a river will lead to
higher flood water levels. Firstly, the cross sectional area is reduced, so that the water
level will increase for a given discharge. Secondly, the function of the embankment is that
no spillage will occur any more, so the discharge will increase and at the same time the
flood water level.
No further delta formation. The natural process of delta formation will be halted by the
construction of flood embankments. Experiments i n Burma, Cambodia and Rumania with
the operation of sluices for the controlled admission of silt-laden waters have not been very
successful.
Sediment deposition on flood plains. When flood plains are designed, all the sediment
will be deposited at the flood plains. When no flood plains are constructed, the river will
transport the sediment to the sea. So, the flood control engineer has to make an important
decision for the future generations: a sediment deposition of e.g. 0.01 m per year, will
raise the flood plains above the surrounding land with 1.00 m per century.
Impact on population. The population likes to live on the natural levees along the river,
as the soils and drainage conditions are better here. Construction of embankments, and
certainly the application of flood plains will create problems for the population.
Drainage of surrounding land. Construction of embankments will block the drainage
outfalls from the backswamps to the river. Sluices have to be constructed, but they will
often not solve the problems as they have to be closed (river in flood) when the drainage
is required (after rainfall). It is often better to construct a separate 'internal' drainage that
debouches ("stroomt uit, loost") directly into the sea.
1.2.5. Flood Diversion

Flood Diversion. 'Flood diversion' ("afleiding van hoogwaters") is a widely used method in
a delta. The floods are diverted from one channel into another with a higher capacity. A
prediction must be made on the morphological effect of this diversion. A chaimel with a
lower discharge has an exponential reduction of the sediment transporting capacity. So, the
splitting of a river into two channels means that the cumulative sediment transport capacity
is reduced. Flood diversion is also a matter of hydraulic engineering, concentrating on the
design of the flood diversion channel ("afleidingskanaal") and the design of the flood
diversion structure ("verdeelwerk"), see figure 1.8.
'offlaking channel'
D i v e r s i o n Ciiannel
'OFFTAKE'
'CONTROL'
:;0IlgQiH^;:JJi.V.Cr.;
I III I
Figure 1.8. Flood diversion structure for flood control.
12
1.2.6. Hydrology of Floods

Background. Design of drainage systems and flood control works rely on an accurate estimate
of the flood discharges. Sometimes, the assessment of the peak discharge is only sufficient,
e.g. for the determination of the capacity of the channel. Often, the whole hydrograph with
its volume is required, in order to perform studies on flood routing, damage of spillage, etc.
Many methods exit to estimate the size of the design flood. Examples are the synthetic
unit hydrograph method, the curve number method. Other methods are regionally developed
and are based on experience. These methods are not discussed here, as they belong to
hydrology.
Introduction. Hydrological data can be processed in several different ways, see figure 1.9:
complete series, when all data are used, e.g. all peak floods during the years,
annual series, when only the annual extremes are used, e.g. the annual maximum peak
flood,
partial series, when all data above a certain base are used, e.g. all peak floods above 250
m^/s.
The selection is based on the purpose of the analysis, the availability of data, and on personal
preference.
magnitude
Annual Maximum
above base value
\
base
value'
year 1
year 2
year 3
year 4
year 5
year 6
year 7
Figure 1.9. Complete, partial () and annual () series.
Complete series. Hydrological data can involve an enormous amount of data, see also figure
1.9. For example, discharges that have observed hourly during the last 10 years, means some
90,000 data. These data are not independent, since the flow is only gradually changing.
Processing all these data, i.e. the 'complete series', is normally not practical or
necessary when studying the extreme values. Monthly and annually rainfall data require
normally the complete series of data. Often 'normal probability paper' ("normaal
waarschijnlijkheidspapier") is used in the statistical analysis.
13
Annual series. 'Annual series' ("jaar serie") are normally used for the determination of the
extremes, i.e. the maximum and/or minimum flow with a certain return period. Now, only
the maximum or minimum event for each year is extracted from the complete series.
Annual series are used e.g. for determining peak floods, daily and hourly peak rainfall.
Often 'Gumbel probability paper' ("Gumbel waarschijnlijkheidspapier") is applied in the
statistical analysis.
Partial series. A year may contain two or more high peak floods which are higher than other
yearly maxima, and which are independent events, see also figure 1.9. So, an analysis based
on annual series ignores the second- and lower-order events of each year which may be even
greater than the annual floods of other years.
'Partial series' ("partile serie") will overcome this limitation, by evaluating all
extremes above some arbitrary base value, e.g. 250 m^/s. The partial series are plotted on
single or double logarithmic paper as no statistical paper is available for these series.
In principle, the use of partial series is more correct than the use of annual series.
However, the difference is very small when the extreme values are calculated from a data
series of more than 1 0 years. Partial series are needed;
when only a few years of data are available, e.g. < 5 years.
for the determination of events of a small return period, e.g. for return periods T <2
years. Now, it is practical to use not all data of e.g. 20 years, but to select at random a
shorter period, e.g. 3 - 5 years.
Short return periods from annual series. Events with short return periods, e.g. the twice-peryear flood: T = 0.5 year, cannot be read from 'Gumbel probability paper' or from 'normal
probability paper'. Gumbel probability paper shows return periods of T = 1.01 years and
higher, normal probability paper shows return periods of T = 2 years and higher.
However, it is also possible to convert the return period of the aimual series to the
'true' return period as follows from the partial series. Therefore, the following formula can
be used (Chow 1964):
where:
ânnual
^partial
return period, according to the annual analysis,

^^^^ period, according to the partial analysis.
Example. The once-per-year flood of T = 0.5 years, i.e. Tpartiai = 0-5 year of the partial
series, equals the flood of r^jûal = 1 / ( 1 - e ) = 1.16 year.
The determination of the once-per-year flood from the annual series through the above
formula is less accurate. This is because most of the actual flood data in that range are not
used in the annual analysis.
Statistical analysis. The statistical analysis, to determine the magnitude of the event (e.g. the
peak discharge) with a certain return period (e.g. T = 20 years) is as follows:
take N years of data;
select the type of data, i.e. annual, partial, complete;
arrange the data in the descending order, so from large to small;
assign the order number m to the data, i.e. m = 1 to the largest, etc.;
estimate the frequency of all data, i.e. assign the 'estimated' return period T or the
'estimated' probability of non-exceedance P' to all data;
14
select the type of distribution, e.g. log-Gumbel, log-normal, linear-normal, by plotting the
data on different probability paper, and by testing on a straight distribution;
plot the data on the selected probability paper;

draw a straight line on 'best fit' through these points;
extend this straight line to the range of the desired return periods;
read the magnitude of the event Qj for the required return period T.
Estimation of the plotting position. The data are first arranged in descending order of
magnitude and assigned to an order number m, where m = 1 for the largest value. These data
have to be plotted on paper. The question is: how to plot?
The magnitude of the event, e.g. discharge is m?/s, is available and is related to the
vertical axis. The parameter for the horizontal axis is either:
the 'return period' ("herhalingstijd"), T i n years,
the 'probability of exceedance' ("kans van overschrijding"), Pin %,
the 'probability of non-exceedance' ("kans van onderschrijding"), P ' in %.
The return period. There are many formulas available to estimate the return period of an
event. The term 'estimated' is important, because the hydrologist will determine 'true' return
period only after the plotting process. The 'estimated' return period is commonly estimated
by the formula (Chow 1964):
T
^ ^^
m
where:
T is the 'estimated' return period of the event, in years
A'^ is the number of years of record,
m is the order number of event, where m = 1 for the largest value.
For annual series, the highest value of the rank m is determined by the number of
years of record, and equal m = N. Thus, the lowest plotting position T depends on the
number of years of record only. For instance, an annual series of N = 10 years leads to m
= 10 for the lowest event, so T = (N+l)/m = (10-l-l)/10 = 1.10. Thus, the 'estimated'
return period T will always be > 1.
For partial series, the highest rank of m is determined by the choice of the "base
value". For instance, the rank of event m could become e.g. m = 35, while the number of
years is only
= 4, thus T = (4 + 1) / 35 = 0.14 years. Thus, T w i l l also become < 1.
Probability of exceedance. The probability of exceedance ("kans van overschrijding") follows
from:
p
_I_ X
100%
for r > 1 year
where:
P
is the probability of exceedance in %,
T
is the 'estimated' return period of the event, in years.
For instance, a remrn period of T = 20 years equals the probability of exceedance P = 0.05
= 5 %. And also, the return period of T = 2 years equals the probability of exceedance P
= 0.50 = 50 %, which means that "on average, the flood happens every second year".
The probability of exceedance ("kans van overschrijding") for extreme values has only
a meaning for P > 50%.
15
<
,00
' Return Period,

50
50
,0
i n years
.
5
<
Return Period,
.
.0
in yesr
>
80 100 8 0
20
Return Period,
100 50
aO
10
i n years
5 3.33 3.5 2
2
2.5 3,33 5
10
^"
20
"V
50 100 200
3
a.5|
3
1.9
3
3.5|
10
20
SO
30 40 SO 0 70 00
0
gs
8 99
P r o b a b i l i t y o f Non-Oocurenoe, i n %
30
40 50 50 70 50
90
95
98 99
P r o b a b i l i t y of Non-Occurence, i n *
Log - Normal Probability Paper
Linear - Normal Probability Paper
2.51
2
3
2.5
2
tl,
1.3 M
2 a.6 3
5 6 7 8 10 15 20 30
SO 70 100 150 200 ' " , 5 ,
Return Period, i n years
>
Linear - Gumbel Probability Paper
Figure 1.10. Four types of probability paper.
,.5
2.53
4 3 6 78 10
1f20
Log - Gumbel Probability Paper
30^_,ji0 J O I O O ^ I S O 200
DESIGN OF OPEN-CHANNELS A N D H Y D R A U L I C
16
STRUCTURES
Probability of non-exceedance. The probability of non-exceedance ("kans van onderschrijding") follows from:
P' = 100% - P
and thus:
P'
100%
for r > 1 year
where:
P'
is the probability of non-exceedance in %,
P
is the probability of exceedance in %,
T
is the 'estimated' return period of the event, in years.
For instance, a return period of T = 20 years equals the probability of non-exceedance P'
= 0.95 = 95 %.
It is often more practical in analysis of extremes to use just the 'return period', or
alternatively the 'probability of non-exceedance' P'. The return period has also the advantage
that it can be used for frequent events. The twice-per-year flood has a return period of T =
0.5 years, while the probability of non-exceedance would read a meaningless P' = -100%.
Type of distribution. It is necessary that the type of distribution of the event is known. This
can be assessed by plotting the data on different types of probability paper, and to choose that
distribution for which the points can be represented by a straight line on the corresponding
paper.
Some of the possible combinations are, see table 1.1 and figure 1.10:
for a horizontal axis:
- Normal (Hazen) probability distribution
for "complete series"
- Gumbel probability distribution
for "annual series"
- logarithmic scale
for "partial series"
on the vertical axis:
- linear scale
"x"
- logarithmic scale
"In x" or "log x"
Drawing of the straight line. Statistics 'by plotting data' is essentially based on drawing a
straight line. It is only a straight line that can be extended to the range of the required return
periods.
The advantage of the statistics 'by plotting data' is that an optical insight is obtained in
the 'best f i t ' and thus in the accuracy of the analysis. Moreover, the hydrologist can make
his personal decision to underweight certain points. For instance, the points of the lower
return periods in the armual analysis tend downwards because of many data in this range have
been filtered out. Also, a very high flood in a series of N = 15 years, may not have the
'estimated' return period of e.g. T = 16 years, but a real return period of T = +100 years.
Table 1.1.
Type o f
Series:
Type o f
Data:
Type o f
Paper:
The use of plotting paper.

COMPLETE
SERIES
annual r a i n f a l l
monthly r a i n f a l l
l i n e a r - N o r m a l paper
log
-Normal p a p e r
PARTIAL S E R I E S
ANNUAL S E R I E S
extremes T < 2 y e a r s
s h o r t extreme s e r i e s
maximum f l o o d s
minimum d i s c h a r g e
daily rainfall
( l i n e a r - l o g paper)
log
- l o g paper
linear-Gumbel paper
log
-Gumbel p a p e r
17
1.3. Drainage Systems in Polders

1.3.1.
Polder
Development
Different professional disciplines. Many professional disciplines are involved in design of

polders.The first activities on polder development are covered by the hydraulic engineers.
They design and construct a ring dike ("ringdijk") and one or more pumping stations. The
main channels to the pumping stations are constructed by underwater dredging. Finally, the
water is removed by the pumping stations.
The second group of activities are taken over by drainage engineers and by agricultural
engineers. They make the new land f i t for agricultural use.
Finally, a third group of activities is covered by rural plarming engineers and other
specialists. The polder is developed with the necessary infrastructure (villages, roads, etc.)
and landscaping. Also the non-technical matters are covered, such the selection of
"immigrants" and the setting up of a Public Administration.
A n example of the empoldering of a shallow lake, the Hachiro Gata polder in Japan,
is presented in box 1.1, see figure 1.11.
Box 1.1. Hachiro Gata polder.
The Hachiro Gata polder is located in shallow lagooi^ i n Japan and was completed in
l972.The design of the Hachiro Gata polder comprises:
the size of the polder is 16,000 ha, and the depth is some 5.00 m below sea level;
a part of the lagoon remains open-water, i.e. the 'belt canal* ("randmeer, boezem"}, and
has a function on (i) water storage for irrigation, (ii) prevention of groundwater drawdown in the old-land, and (ill) for collection of surface drainage water from the old there is a separated ("gescheiden") irrigation and draiimge system in the polder to allow
the cultivation of paddy ("rijst");
* ie drainage modulus of the paddy fields is based on a design storm ("ontwerpbui") of
225 mm in three consecutive days ("opeenvolgende dagen"), when a water level rise on
the paddy fields of O.iO m Is allowed. This leads to a drainage modulus of 5 1/s.ha or
IpiiijiIJilîllli^
a sub-drainage system under the paddy fields at a depth of 0,50 - 0.70 m and at a
distance of 10 m has been designed, to leach ("uitspoelen") the soil as to ptsvent
problems with acid-sulphate clay ("katteklei").
Components. The water management system in a polder exists of different components, see
figure 1.12:
field drainage system ("ontwateringssysteem"), to maintain the groundwater table under
the root zone;
main drainage system ("afwateringssysteem"), to divert the drainage water from the field
drainage system to the outlet;
sluices and/or pumping stations ("lozingssysteem"), to evacuate the water from the main
drainage system to the bordering canals;
belt system ("boezem") surrounding the polder.
18
Figure 1.11. Hachiro Gata polder, Japan.
19
Figure 1.12. Water management system of a polder.
1.3.2. Main Drainage System

Function. The function of the main drainage system ("afwateringssysteem") is to collect the
water from the field drainage system, to transport the water to the outlet, and to evacuate the
water from the area.
Types. The main drainage system of a typical new polder in the Netherlands consists of: (i)
canals, and (ii) main ditches and plot ditches.
The canals are often navigable and have (stagnant) water depths of some 1.50 - 2.50 m.
The gradient of the canal bed is normally horizontal. The main ditches and plot ditches run
along the plots, e.g. the rectangular fields of about 500 m X 1200 m.
Polder water level. The main drainage system flows mainly because of rainfall. So most of
the time it contains stagnant water. The water level of this stagnant water must be carefully
controlled. A lower or a higher water level might have negative effects on the crops and on
land subsidence. This target water level of the open-water drainage system of a polder can
be called the polder water level ("polderpeil"), see figure 1.13.
Thus, water management in a polder is mainly determined by water level control
("peilbeheer") in the drainage system, and to a lesser extent by discharge control
("debietbeheer") like in irrigation schemes.
20
pumping station
Figure 1.13. The 'polder water level' is the target the open-water level.
Capacity. The required capacity of the drainage system depends not only on the rainfall but
also on the seepage, e.g. 1 mm/day (0.11/s.ha) or higher, and on lockage water from polder
shiplocks, e.g. 1 mm/day (0.1 1/s.ha).
Furthermore, the required capacity depends also on the percentage of open-water, which
is kept low in the modern polders ( 1 - 2 % ) with electric pumping stations, but might be much
larger (5 - 15%) in older peat polders with pumping by windmills.
Table 1.2 presents some typical values of canal capacities in the main drainage system
for moderate climates (Luijendijk 1982). These discharges are of the order of 13 mm/day (1.5
1/s.ha) for the smaller ditches and of 11 mm/day (1.3 1/s.ha) for the larger canals. For urban
polder areas, values for the main system of 20 nun/day (2.3 1/s.ha) and more are often
applied.
Table 1.2. Design parameters of main drainage systems for moderate clunates.
Polder type
old polders
old polders
old polders
new polders
urban polders
Soils
peat
clay
clay
clay
Land use
grass
grass
crops
crops
Open
water
5
3
5
1
3
Polder
water level
m-terrain
Canal
capacity
mm/day
- 10
- 10
- 10
- 2
- 8
0.20
0.40
0.80
1.40
1.50
8
8
8
11
15
0.50
0.70
1.00
1.50
1.80
12
12
12
14
30
21
Operational objectives. The operation of gravity ("zwaartekracht") drainage systems is

normally not required during the periods of rainfall. The field drains just confluence into the
smaller drains, and these smaller drains discharge into the larger drains. There is no need to
maintain a certain minimum water level in the drains, so that they may fall dry during periods
of no rainfall.
However, water level control ("peilbeheer") in the drainage system might be needed
during the dry periods, for instance to store groundwater in the soil. This groundwater can
be used by the crop by "capillary rise" ("capillaire opstijging"), or will avoid subsidence
("klink").
Lavout. The main drainage system of a polder can be designed in a grid pattern, or in a
branching pattern, see figure 1.14.
Moreover, the question may arise in polders with irrigation water needs, whether
separate irrigation and drainage systems ("gescheiden stelsel") are constructed, or that the
mixed irrigation and drainage system ("gemengd stelsel"), where the irrigation supplies are
taken from the drainage system.
A grid pattern can be applicable for flat polders. It has the advantage that the canals
can be used as boundaries between land holdings. Examples of grid patterns can be found in
the polders of the Netherlands.
A branching pattern ("vertakt") of the main drainage system might be applied for
sloping polders. A branching drainage system is also applied in polders with sulphate-acid
problems to flush the acid water. Examples can be found in Indonesia.
Irrigation. Irrigation supply to the crops with a grid drainage system is possible by means of
capillary rise of the groundwater ("capillare opstijging"). The appropriate groundwater level
is maintained by bank-infiltration ("oever infiltratie") from the open-water system.
The other possibility for irrigation in a grid drainage system is by means of pumping
by the individual farmers on an 'on-demand' supply basis, where the open-water level in the
polder is maintained by subsequent intake into the polder.
A branching drainage system must be applied in polders with a separate irrigation and
drainage systems, such as in Japan, Korea and Taiwan. Here, gravity ("zwaartekracht") is
used to irrigate from the higher irrigation system to the fields, while the water levels in the
drainage system are kept sufficiently below terrain level.
IIIIIIIIIIHIIIIIIIIIIIIIII llllllllllllllllllllllllll
lllllllllllllllllllllllllllllllllllllllllllllllllllllll III
Grid Drainage System
lllllllllllllllllllllllllillllliliiiiiiiiiiiiiiiiiiiiiiT
Branching Drainage System
Figure 1.14. Pattern of main drainage system in polders.
Separate Irr. & Dr. System
22
1.3.3. Polder Water Level

Determination. The polder water level, i.e. the stagnant water level, is often established on
agricultural considerations to obtain the highest economic return of the crops (Kley 1969).
The relevant parameters are:
type of the soil, like peat that requires shallow groundwater,
proposed land use, and the yield curve ("opbrengstcurve") for different open-water levels,
see figure 1.15 for example,
elevation of the terrain.
But the polder water level has also to meet interests of navigation, recreation and nature
preservation ("natuurbeheer"). Moreover, it should prevent or reduce the irreversible
subsidence ("klink") of the soil.
A n example of the determination of the polder water level based on the highest crop
yields of figure 1.15 is given in box 1.2 and in table 1.3.
Percentage of the optimum yield
/
80%
/
/1
60%
40%
/
/
20%
y
0.00
0.20 0.40
4 /
'
%)
^/
\
\
ss
<*
0.60
to
0.80
1.00
1.20
1.40
1.60
1.80
2.00
2.20
Open-water level below terrain, in m ________
Figure 1.15. Example of crop yields for different open-water levels.
Box 1.2. Determination of the polder water level.

A certain polder has a size of 1200 ha. A lower area al 0.00 m + M S L of 900 ha of peat
soils win bo used for gras.s, a higher area at 1.00 m + M S L of 300 ha of clayey soils will
be used for crop cultivation.
The yield curves for different (stagnant) open-water levels ("polderpeilen") In the
channels below ten-ain level, are presented in figure 1.15. The economic value of arable
land ("bouwlattd") is twice as high as the value of grassland. Thus, the annual yield of
amble land is $ IX, against aii annual yield of grasjsland of % X The calculation is done
by means of table 1.3.
Each assumed open-water level gives a certain relative yield of the grassland and of
the arable land. It shows that the total aimual yield of $ 1470X is maximum for one polder
water level of 0.40 m-MSL.
As an alternative, two different polder water levels can be used: 0.40 m-MSL for the
grassland and 0.60 m-MSL for the arable land. The total annual yield would be S 900X
+ S 600X = $ 1500X. The extra construction costs would include a gated drop structure
between the two polder sections.
23
Table 1.3. Example: Calculation of the polder water level.

Polder
water level
m-MSL
0.60
0.50
0.40
0.30
Grass land (900 ha)

Water level
m-terrain
0.60
0.50
0.40
0.30
yield in $
0.75
0.90
1.00
0.93
X
X
X
X
900A:
900X
900Z
900Z
Arable land (300 ha)

Water level
m-terrain
1.60
1.50
1.40
1.30
yield in $
1.00
0.98
0.95
0.90
X
X
X
X
600X
600X
600Z
600X
Total polder
(1200 ha)
yield in $
1275X
1398Z
1470Z
1377Z
Soils. Generally, peat soils ("veen") require high polder water levels of some 0.30 - 0.60 m
below terrain level to prevent the subsidence of the soil. Thus, grass and other crops only
with shallow roots can be cultivated.
Clayey soils permit deep polder water levels of some 1.20 - 1.80 m below terrain. It
means that dryland crops with a thick root zone can be cultivated.
Even greater depths of 1.50 - 2.00 m below terrain have to be selected in polders with
brackish seepage and with low rainfall. Capillary rise of the brackish water can be avoided
by over-irrigating the crops and allowing a percolation flow to the drains.
Bv-law. There are conflicting interests in the determination of the polder water level: the
highest crop yield for the whole polder may not give the highest yield to each individual
farmer, and may not not meet the requirement for subsidence, or may not allow navigation
in the chaimels, etc.
Thus, the ultimate polder water level will be a compromise between the different
parties. This selected polder water level is laid down in a by-law ("peilbesluit") of the Water
Board ("waterschap").
Very often a summer polder level ("zomerpeil") and a deeper winter polder level
("winterpeil") is applied, as to provide for more drainage in the wetter winter season.
Box 1.3. Polder sections.
In alternative I , the drainage water is discharged from the higher to the lower sections
by means of gated dxop strctHfes. Only one pumping 5tatloii, with a capacity Of Qi +
02 + <23 is required, but at the highest operation costs, whieh are related to ( ^ i + O2
+ 3) X
, where B and Q are the required pumping head and discharge of the
different polder sections, respectively,
9 In alternative 11, the drainage water is gradually pumped to ie higher polder sections.
The total operation costs wl be lower than for the above alternative and Will be
related to ^1 1 + ^ 2 2 + ^3 s In alternative in, the minimum operation.costs and the minimum pump capacity costs
can ht obtained by drainir^ the three polder sections directly to the external water
system. The lower and the higher sections are located along the external water system,
and can discharge directly > The tniddle polder section In figure L 1 6 should be
coMeoted with the external water system direcay, or by means of a 'belt canal'
("ringkmaal"). Three pumping stations are required faere^ with capacities related to:
Ql, Q2 and 3* respectively. The total operation costs are at aminitnura and are related
to; Hi 0 1 + Il2 Q2 ^ H3 3<
24
Polder Sections. Polders may be divided into compartments or sections having different
polder levels ("polderafdelingen"), depending on the topography as to limit the variation in
groundwater depth to e.g. 0.50 m. Different polder levels are also applied when soil
conditions differ, e.g. peat soils with high polder levels 0.50 m below terrain, clay soils
with deep polder levels at 1.50 m below terrain. So is the Wieringermeer Polder (the
Netherlands) divided into four sections with different levels, and the Flevo Polder (the
Netherlands) into two sections. An example of a polder with three polder sections is presented
in figure 1.16 and in box 1.3 (Kley 1969).
Pump-lift
P O L D E R
QI + Q2 + Q3
Outside
Water Level
H1
Q2
Gravity
Lower
Section
Alternative
Q3
Gravity
Middle
Section
Higher
Section
I : One pumping station (maximum energy requirements).

Q1 + Q2 + Q3
Pump-lift
P O L D E R
Alternative
I I : Three pumping stations (minimum energy requirements).
Pump-lift
Outside
Water Level
Pump-lift
Q1
Q2
H1
H2
Lower
Section
Alternative
Outside
IW a t e r L e v e l
Middle
Section
Pump-lift
" Belt"
Canal
Higher
Section
III: Three pumping stations & high " Belt Canal"
Figure 1.16. Pumping stations in polders with three 'sections'.
Q3 H3
Outside
ater L e v e l
25
1.3.4. Water balance

Units of the water balance. The terms of a water balance are often expressed in a water
depth ("waterschijf") per time unit: in 'mm/dav', like the evaporation and rainfall data.
However, the unit '1/s.ha' for water depth is alternatively used in the design of drainage
systems. The conversion between these two units is:
1 mm/day = 0.116 1/s.ha,
1 1/s.ha
= 8.64 mm/day.
Water balance equation. The water balance equation of a polder reads:
The inflow consists of precipitation P, seepage from the sub-soil S, and intake water / for
irrigation supplies and lockage water from ship-locks.
The outflow consists of open-water evaporation E^, evapo-transpiration of the
vegetation ET^, and the pumping and sluicing discharges Q.
The difference between the in- and outflow over the considered period AT determines
the changes in storage ("berging") A F in the soil and in the open-water. This time interval
AT can be taken as e.g. one year for determination of the aimual pumping costs, or as one
day or even shorter for the determination of storm-drainage requirements of the canals and
the pumping stations.
Seepage. The assessment of the seepage ("kwel") to new polders is quite difficult as the
seepage depends on the permeability of the sub-soil, the thickness of the aquifers and on
several other geo-hydrological parameters.
The seepage can be divided into (i) a seepage through the embankment, and (ii) a
seepage from the deeper soil layers.
The total seepage of existing polders can be calculated by means of the water balance
equation. The typical seepage value of polders in the Netherlands is 1 mm/day (0.1 1/s.ha),
but values as high as 13 - 16 mm/day (1.5 - 1.9 1/s.ha) are also encountered (Schultz 1992).
Irrigation needs. Many polders require a regular intake of fresh water during a part of the
growing season. For instance, grass-polders in the Netherlands may require an irrigation
supply upto 0.6 1/s.ha during dry years. Some 1.5 1/s.ha is required in the paddy-polders in
Indonesia.
The quality of deep seepage water is often poor, as it can be brackish or even salty.
Thus, the salt water balance of the polder should be checked. Polders in (semi) arid climates
may require a considerable irrigation supply together with a good drainage system, to control
the saline conditions of the root zone of the crops.
Water storage. Water storage in open-water channels ("open-water berging") depends on the
percentage of open-water in the polder and the allowable rise in water level.
For instance, a modern polder may have 1 % of open-water, with an allowable rise of
water level of 0.20 m. The rate of open-water can also be expressed in a water depth
("waterschijf") over the whole polder. So, the polder in the above example has an open-water
storage of 0.01 x 0.20 = 0.002 m = 2 mm.
Water storage in the soil depends on factors like the "drainable pore space" ("bergings-
26
coefficient") of the soil and the allowable rise of the groundwater table (Luijendijk 1982). So,
polders with deeper open-water level during the dry season will store more water in the soil
during heavy rainfall, see table 1.4.
Required pump capacity. The water balance with a time interval of A r = 1 day can be used
for a rough estimate of the required pump capacity (Kley 1969). An example is presented in
box 1.4 and in table 1.5.
Table 1.4. Typical values of water storage in the soil.

Water storage in soil
Open-Water Level
below terrain level
0.20
0.40
0.60
0.80
1.00
1.20
1.40
m
m
m
m
m
m
m
20
45
75
105
peat
light clay
heavy clay
mm
mm
mm
mm
4 mm
12 mm
25 mm
40 mm
60 mm
85 mm
115 mm
2 mm
6 mm
12 mm
20 mm
28 mm
38 mm
50 mm
Table 1.5. Water balance of a polder for the design rainfall.
Rainfall
mm/day
Percolation (Rain-day 1)
Percolation (Rain-day2)
Percolation (Rain-day3)
Total percolation
mm/day
nun/day
mm/day
mm/day
Percolation to open-water
Rainfall on open-water
Seepage
Total flow to open-water
Pumping from open-water
Day
no.1
Day
no.2
Day
no.3
Day
no.4
Day
no.5
Day
no.6
18
33
12
6
11
11
4
15
4
4
17
6
11
4
21
mm/day
mm/day
mm/day
15
3
1
19
1
1
14
5
2
1
mm/day
mm/day
1
-1
8
-8
19
-14
21
-14
15
-14
5
-14
Open-water storage
Open-waier elevation
mm
m-fMSL
0
0.00
0
0.00
5
0.05
12
0.12
13
0.13
4
0.04
Water storage in soil

Groundwater elevation
mm
m+MSL
0
0.10
12
0.29
0.54
19
0.40
iiiiBii
0.16
0
0.10
Box 1.4. Water balance of a polder.
27
A polder may have an open-water ratio of 10% and the groundwater level is initially
0.10 m above the Magnum open-water level, i.e at 0 10 m + M S L . bee also Uible 1.5.
The water storage in ihe soil is 7 mm per 0.10 m gnmndwater rise, i.e. 'drainable
pore space" {"bergingscoeîcient") of 0.07. A seepage flow to the polder amounts to 1
mm/day (0.1 1/s.ha). The design rainfall may consist of a ihree-day storm of 18 mm/day.
33 mm/day and 12 mm/day, respectively.
It is assumed that the rainiall of one day percolates during 3 days, i.e. during the day
of rainfall and the two following day.s. This percolation and the seepage is discharged to
the open-water, where the pump with a capacity ot 14 mm/day, is removing the water.
An open-water storage of 4 mm in day no.3 makes a rise of the open-water level of
0.004/10% = 0,04 m. The groundwater storage of 28 mm in day no.3 was created by the
rainfall storage (18+33) minus the percolation (6+17). This 28 mm give a groundwater
rise over the land area 90% of 0.028/(90% xO.07) == 0.44 m.
Thus, it follows trom ihe calculations in table 1.5 that the average groundwatej- level
will rise by 0.54 - 0.10 = 0.44 m, and the open-water level by 0.14 m for an assumed
pump capacity of 14 mm/day (1.6 1/s.ha).
1.4. Irrigation Systems

1.4.1. Irrigation
Development
Dual-management. Many large-scale irrigation schemes are under 'dual-management'. The

Government, or another entity, is charged with the operation, management and maintenance
of the main infrastructure of the scheme. This part of the scheme belongs to the general
interest of the population, and requires technical, financial and organizational capabilities that
are beyond the capacity of the individual farmer.
However, such an entity cannot be charged with the operational aspects at the lowest
level, such as the operation and maintenance at farm level. This remains the task of the
individual farmer, or may become the task of a group-of-farmers.
Irrigation svstems. Thus, many irrigation schemes are divided into two parts, with each their
own management responsibilites, see figure 1.17:
the 'main irrigation system', ("hoofdsysteem", "dienstleidingen"), i.e. the part that is
under the control of the Government or another entity, through e.g. an ' O & M Agency'
("beheersdienst",
"irrigatiedienst");
the tertiary unit, ("tertiaire vak", "bevolkingsvak"), i.e. the part that is under the control
of a number of farmers, to be grouped in e.g. 'Water Users Associations'
("watergebruikersorganizatie").
28
Figure 1.17. A n irrigation sclieme under 'dual-management'.
Terminology. The terminology of irrigation system as used here, is shown in figure 1.18. The
feeder canal ("onverdeelde hoofdkanaal") diverts the irrigation water from the headworks
("watervang") to the irrigation scheme. The primary canal ("hoofdkanaal") transports the
water to the different sub-areas of the irrigation scheme. Often, the gradient of such a
primary canal is flat, as to maintain the highest head.
Secondary canals ("secundaire kanalen") take off through secondary offtakes
("verdeelkunstwerken",
"secundaire aftappingen") from the primary canal, and may divert
into other (sub)secondary canals. These (sub)secondary canals distribute the water through
the irrigation scheme to the tertiary offtake ("tertiaire aftapping").
The main irrigation system ("hoofdsysteem",
"dienstleidingen") consists of the
headworks, the feeder canal, the primary canal, the (sub)secondary canals, and all the
structores on these canals. Thus, the tertiary offtakes belong to the main system.
The tertiary unit ("tertiaire vak", "bevolkingsvak") is the irrigation area supplied by
one tertiary offtake. The tertiary unit consists of tertiary canals ("tertiaire kanalen",
"bevolkingsleidingen") and quaternary canals ("quartemaire kanalen") with their structures.
The tertiary unit is under responsibility of the water users association. The individual farms
receive their water from the quaternary canals.
Tertiary offtake. The structure that diverts the water from the main system into the tertiary
unit is called 'tertiary offtake' ("tertiaire aftapping") . Other names are 'turnout', 'outlet',
'inlet' and 'mogha'. The tertiary offtake is the official point at which water passes from the
control of the O & M agency to the water users association. But it is also the site of struggle
between the water users and the staff of the O & M agency.
29
1. W A T E R C O N T R O L SYSTCMS
PRIMARY CANAL
FEEDER CANAL
secondary offtake
Headworks
(SUB-)SECONDARY CANAL
Crsecondary
secondary
offtake
offtake
t&tiary offtake
U
tertiary unit
C/5
O
C/3
C/3
Figure 1.18. Terminology of irrigation canals.
IA.2. Main Irrigation System

Function. As elaborated above, the main system is managed by the O & M Agency and not by
the farmers. The function of such a main irrigation system is to supply irrigation water to the
tertiary units according to the 'operational objective'.
Main irrigation systems can only be well designed when the ultimate operational
objectives have been specified accurately. Otherwise, controversy will remain during the
operation of the project, whether the system is performing 'to the requirements'.
Moreover, the technicians (civil engineers, agriculturists, economists, etc.) involved in
the design of a new irrigation system, and the ultimate users of this system (local
administration, O & M Agency, fanners, etc.) have to agree jointly on these operational
objectives.
Irrigation requirement. The 'irrigation requirement' ("irrigatiebehoefte")
of the crop
depends on many factors, such as the water requirement of the crop, the effective rainfall,
and the application efficiency. The crop water requirement is normally calculated by a
formula, such as the Penman-Monteith formula (Ankum 1996). The irrigation requirement
is normally expressed in the unit of a 'water depth per tune', such as mm/day or 1/s.ha.,
whereas 1 1/s.ha = 8.64 mm/day.
Canal capacity. The capacity of an irrigation canal is the maximum discharge, in rn^/s, that
should be accoimnodated during the peak period. It is also often expressed in 1/s.ha, to make
it independent of the command area
("verzorgingsgebied").
30
There are several methods to determine the canal capacity of an irrigation system.
Traditionally, the capacity of irrigation systems is based on experience and not on
calculations. For instance, typical canal capacities in Indonesia were taken at 1.4 1/s.ha for
the irrigation need of paddy ("rijst") during the wet-season.
At present, more emphasis is given to the calculation of the canal capacity. The required
canal capacity depends on the peak irrigation requirement of the crop and the water losses in
the system.
In India, the canal capacities were intensionally taken below the peak irrigation
requirement of the crop. It is a dilemma that a supply of the full crop water requirement to
a (small) area means that the remainder area will not receive any irrigation water at all. Thus,
it was decided to supply many farmers with less than the required irrigation supplies. Such
an irrigation method is called 'protective irrigation', or also 'extensive irrigation'. Typically,
the canal capacities are only 0.3 - 0.6 1/s.ha to irrigate the dryland crops ("gewassen") in a
large command area. Thus, the cropping pattern ("gewaspatroon") is less than 100%, which
means that a part of the area might be fallow ("braak") during the irrigation season.
Importance of the off-peak period. It should be acknowledged that the operation of the main
system during the peak discharges differs from the operation during off-peak discharges
with less flow through the main system.
Often, the operation during the peak period is simple, as the canals are just flowing at
their maximum discharge. The operation might become cumbersome during the off-peak
period, see figure 1.19. These off-peak conditions determine normally the operational
objective of the main system.
I Q in 1/s.ha
^ Peak
Period
Off-Peals: Period
primary canal
Rotation over
^Secondary Unit
Q in 1/s.ha
Peak
Period
Off-Peak
time
Adjustable Flow
Period
Tertiary
offtake
time-
Inteimittent How
Rotation over
Tertiary Units
Tertiaiy
offtake
Q in 1/s.ha
Peak
Period
Off-Peak
Period
timeQ in 1/s.ha
Peak
Period
Off-Peak
Period
time
Rotational How
Q in 1/s.ha
^ Peak
Period
Off-Peak
Intemiittent How
Period
time
Rotational How
Figure 1.19. Example of flow diversion during the 'off-peak period'.
31
Diversion structures. Water control systems in irrigation have to divert the larger discharges
into smaller portions during the peak period. Moreover, the diversion of the required flow
is more cumbersome during the off-period period when the canals are flowing at a lower
discharge than their capacities.
Therefore, diversion structures ("verdeelkunstwerken") are needed at all bifurcations
("splitsingen"), see figure 1.20. These structures may have one or more of the following
components, see also the next chapters (Ankum 1995):
discharge regulator ("debiet regelaar"), i.e. to regulate the discharge,
water measurement structure, ("meetkunstwerk"), i.e. to measure the discharge,
water level regulator ("peilregelaar"), i.e. to regulate the water level.
1111 111!
1111 I I I !
I I I i T T
LEGEND
>
O
Irrigation Canal
I I I I I I
Regulator
Tertiary Offtake
Figure 1.20. Diversion structures at all bifurcations in a main irrigation system.
1.4.3. Flow Control Methods

Introduction. Many irrigation schemes were initially constructed as simple flooding systems,
by which an unregulated supply of water from the river was spread over the fields. A basic
improvement was made by splitting the flow in a proportional way, as to relate the irrigation
flow to the size of the command area.
A next step in the development of flow control methods was the transition of these main
irrigation systems into upstream controlled systems by placing gates in the proportional
control structures, see flgure 1.21. The simple use of these gates on an 'on/off basis
32
improved the efficient use of water in many cases. The introduction of water measurement
structures with the above gates allowed the systematic distribution of water by an O & M
Agency.
Since the introduction of the downstream control concept in the 1930s, 'selfmanagement' of main irrigation systems are possible. These systems react on changes in
water demand without system management, see also figure 1.21
Since the 1960s, the technology of telemetry on the individual canal reaches (volume
control, E L F L O control) and since the 1980s also of computerized control on the main
system as-a-whole, allows for a further optimization of the distribution of irrigation water.
These methods are not discussed here, and can be found in the literature (e.g. Ankum 1995).
Controller
: I
i:
Controller
Figure 1.21. Control methods i n irrigation.
Proportional control. Irrigation systems under 'proportional control'

("proportionele
verdeling") divide and distribute the water according to a fixed ratio ("vaste verhouding"),
see figure 1.22. Proportional control is a pure type of discharge control: the prmiary purpose
of a structure is to maintain a certain ratio-in-discharges to a downstream destination. An
irrigation system under proportional control is very simple in construction and in operation.
Proportional control can well be applied in modern irrigation schemes. For instance,
when an irrigation scheme has one cropping pattern and a uniform growing stage for all
tertiary units, there is no need to divert more water to certain tertiary units. Thus, irrigation
water can be released from the headworks and can be proportionally distributed to all tertiary
units on basis of the command area. A condition is of course, that the lower discharges do
not give managerial problems within the tertiary units, e.g. a too low water level for gravity
irrigation, or a too low discharge to be managed by the farmers.
Proportional control is the most simple control method in irrigation. However, it cannot
be used efficiently for different crops with different water needs, since the flow cannot be
regulated.
33
Inflow in canal
outflow offtake 1
outflow offtake 2
outflow offtake 3
Figure 1.22. Proportional control in a main irrigation system.
Structures under proportional control. Although there is no active discharge regulation is

needed, structures are still required at every offtaking canal. Otherwise, the downstream
water levels will influence the proportional distribution. This happens because of the different
depth-discharge relations of the canals {Q = k A i?^^^ s^^'^).
The most simple strucmre in proportional control is the Fayoum weir (Kraatz 1975,
Bos 1989, Ankum 1995), see figure 1.23.The Fayoum weir as a divisor for proportional
control, consists of two similar type of overflow structures to the continuing canal and to the
offtaking canal. A fixed division of the discharges is obtained when both crest-heightp above
the canal bed are at the same. The widths
and b^ of weirs determine the ratio of the
discharge into each canal.
(not to scale)
CROSS SECTION B - B
Figure 1.23. Irrigation structure under proportional control.
Upstream control. The above proportional control may not meet the expectations of a main
irrigation system, and a need for a more active regulation of flow may arise. A logic decision
is to equip the above proportional distribution structures with gates, so that regulation of the
discharge becomes possible. Thus, a system under 'upstream control' ("bovenstrooms
peilbeheer") is created, see figure 1.24.
However, new drawbacks on the flow control method are created by solving the imperfections of the proportional control:
34
a central system management is required, where the O & M Agency decides on the water
releases to the tertiary irrigation units, and on the cumulated discharges through the whole
main system;
it appears that efficient operation of an upstream controlled system is very cumbersome,
because of the response time due to the filling and the emptying of the 'dynamic canal
storage'.
Canals under upstream control. Structures under upstream control maintain the upstream
water level at the target water level, which means that the upstream water level is constant
for each discharge, see also figure 1.24.
However, such a water level control will not lead to a workable water distribution
through the irrigation scheme: it will effect a proper diversion of discharges.
Gate
Figure 1.24. The principle of 'upstream' control.

Upstream control is a 'control-of-discharges'. A main irrigation system under upstream (water
level) control is essentially a discharge controlled system. Water is released from the
headworks and is distributed over the smaller canals and ultnately towards the tertiary units,
see figure 1.25. Discharge regulation and discharge measurement is required at all offtaking
canals, as to control the discharge that leaves the canal. The continuing canal is transporting
the remainder part of the discharge.
TTTTTTrm
1 mVs
nTTTTT'liI
1 canal capacity in mVs

\ tertiary offtake: 0.1 mVs
Figure 1.25. Diversion of discharges through an 'upstream' controlled system.
35
Structures under upstream control. Irrigation structures under upstream control are quite
complex. These structures may have one or more of the following functions, see figure 1.26:
(i) discharge regulation, (ii) discharge measurement, and (iii) water level regulation.
Moreover, there are many alternatives to achieve these functions, such as 'fixed',
'manual', or 'hydro-mechnical' regulation, 'underflow' or 'overflow'. These matters will be
discussed in the following chapters.
^^get
i_dischê
Discharge
Measurement^
Discbarge
Measurement
t
T T S . Discharge
Sêgulator
\ S^Discharge
[SRegulator
'target
'water level
I
Water lever
J Regulator
' target
'water level * Water lever
'
^ Regulator
regulation
OFFTAKE TO A SECONDARY CANAL
regulation
OFFTAKE TO A TERTIARY UNIT
Figure 1.26. Irrigation structures under upstream control.
Downstream control. Downstream control is the control of the water level regulator, based
upon changes in the water level immediately dowtistream of it, see figure 1.27.
The regulators in the main system maintain a constant water level at the downstream
side of the structure, without regarding discharges. Such a regulation of structures means that
more supply is given to a canal reach when the water level drops. The effect is that the
discharge at each regulator is automatically adjusted to the accumulated downstream demand
for irrigation water.
Gate
Figure 1.27. Principle of 'downstream' control.
36
Downstream control as 'self-management'. The term main system management refers to the
management of the system-as-a-whole, and deals with e.g. determination of the target water
levels and/or target discharges. The main system management can either be (Ankum 1992*^),
see figure 1.28:
no management of the main system, which occurs with proportional control because any
adjustment of structures is impossible;
central management of the main system, which occurs with upstream control because a
'water operation centre' should manage the inflow and the total outflow of the whole
system;
self management of the main system, which occurs with downstream control as the water
control system adjust the inflow to a changing outflow.
t j^ TTTTt
tertiary offtakes
proportional control
"NO" MANAGEMENT
i\\\V\Vv>
i u t i '* 'A;
tertiary offtakes
tertiary offtakes
upstream coniiol
downstream control
CENTRAL MANAGEMENT
SELF MANAGEMENT
(by a Water Operation Centre) (by the water level regulators)
Figure 1.28. Overall 'system management' of main irrigation systems.

Structures under downstream control. The water level regulators at the secondary offtakes
are located in the incoming canal reach, see figure 1.29. These regulators are not located in
the continuing canal reach, like in upstream controlled systems. A discharge regulator at a
(sub)secondary offtake is absent, and it is not matching with downstream control.
However, the tertiary offtakes have to be equipped with discharge regulators, as here
water changes 'management', i.e. from the ' O & M Agency' to the 'Water Users Association',
see also figure 1.29.
Water level i water level ,
Water level ,
Regulator -
j^^^^j^-^
OFFTAKE TO A SECONDARY C A N A L
Regulator
;
regulation
OFFTAKE TO A TERTIARY U N I T
Figure 1.29. Irrigation structures under downstream control.
2. OPEN CHANNEL FLOW

2.1. Flow of Water
2.1.1. Uniform flow
Background. Uniform flow ("eenparige stroming") in an irrigation canal, a drainage channel
or a river can be described by one of the three formulae, see figure 2.1:
the Strickler formula:
Q = kA R^'^ s^''^,
the Manning formula:
Q l l n A R^'^ s^'^,
the Chzy formula:
Q = CA R^''^ s^''^,
in which:
o the velocity of the water
v = Q IA
o the water cross-sectional area
A = {b+my)y,
o the hydraulic radius
R = A / {b +
2yV(l+m'^)}
where: Q is the discharge in m^/s, v is the velocity in m/s, A is the wet cross-sectional area
in m^, R is the hydraulic radius in m, s is the energy (chaimel) gradient, b is the bed width
in m, y is the water depth in m, m is the side slope (lyert '^Hor)' ^
Strickler
coefficient in rn^^^/s; n is the Maiming roughness coefficient in s/m^^^, C is the Chzy
roughness coefficient in m^''^/s.
CROSS SECTION
LONGITUDINAL PROFILE
Figure 2.1. Parameters in the Strickler formula.
Relation between Chzy. Manning and Strickler formulae. Basically, the Chzy, the Manning
and the Strickler formula are the same equations. The difference lies in the determination of
the 'roughness coefficient' ("ruwheidscoefficient"):
Chzy presented his formula in 1769. His roughness coefficient, C in m^''^/s, is not a
constant for a specific chaimel but depends also on the water depth.
The roughness coefficient should be calculated through another formula. Examples
are the Kutter formula, the Basin formula, the Powell formula, the Nikuradse formula, etc.
37
38
Manning presented his formula v = lln R^'^ s^''^ in 1889. He concluded that the
roughness coefficient, n in s/m^^^, is independent of the water depth in the charmel.
The coefficient, however, does depend on the general characteristics of the chaimel,
such as soil, vegetation and (meander) form.
Strickler presented his formula v = k R^'^ s^''^ independently in 1923. The Strickler
coefficient, k in m^^-^/s, is related to Manning through n = Ilk, and to Chzy through C
= k R^'^. The 'roughness' ("ruwheid") coefficient k as used in the Strickler formula is in
fact a 'smoothness' ("gladheids") factor. A n increasing roughness will decrease the value
of k. Therefore, it is better not to use the term 'roughness' coefficient but to use the term
'Strickler' coefficient instead.
The Strickler coefficient is independent of the water depth, but depends strongly on
the maintenance ("onderhoud"). A well-maintained channel may have a Strickler
coefficient of A: = 30 - 40 m^'^ls.
Discussion on of the Chzy. Manning and Strickler formulae. The use of the Chzy, Manning
or Strickler formula is mainly based on tradition:
the Chzy formula is often applied by river engineers and in hydro-dynamic models;
the Manning formula is widely used in the English-speaking countries by irrigation and
drainage engineers;
the Strickler formula is widely applied by irrigation and drainage engineers from the
European continent.
The Manning and the Strickler formula have the practical advantage over the Chzy formula
that they use a coefficient which is not dependent of the water depth. The Strickler formula
is recommended here.
Strickler coefficient. The coefficient k in the Strickler formula has a constant value, that does
not depend on the water depth y. However, this Strickler coefficient can change considerably
between cross sections and over the tune. The factors that have the greatest influence upon
the Strickler coefficient k are described below:
surface roughness, depending on the size and shape of the grains of the bed material;
vegetation, the Strickler coefficient may vary during the year which is due to seasonal
growth of aquatic plants and grass in the channel or on the banks. Thus, maintenance has
a major effect on the roughness;
silting and sedimentation, the flow through a cross section will decrease after siltation and
before it is removed during maintenance. This effect can only be expressed in the Strickler
formula by adjusting the Strickler coefficient to a lower value;
channel irregularity, comprising abrupt variations in cross sections along the channel
length, by e.g. sand bars, obstructions, holes in the channel bed;
channel alignment, meandering streams may decrease the Strickler coefficient k with up
to 30%. However, the effect of the curvation is influenced by the flow velocity.
Design Criteria. It is not srniple to determine the 'correct' Strickler coefficient k. Different
engineers will obtain different results.
Therefore it is essential that the value of the Strickler coefficient after it has been
determined, is well communicated with others and that other engineers may present their
comment at an early stage, before extension design work is done. Often the best procedure
will be to present and discuss a special report 'design criteria'
("ontwerpvoorschriften"),
where the procedure for design is well explained.
Values. The Strickler coefficient k of drainage channels is for a major part dependent on the
vegetation in the channel, which is determined by the maintenance ("onderhoud"), die
39
2. O P E N C H A N N E L F L O W
season, the soil and the water depth. Often a Strickler coefficient k = 40 m^'^/s is selected
for channels with frequent and good maintenance. But normally a Strickler coefficient k =
30 m^''^/s is selected for drainage chaimels as maintenance will be less frequent. A channel
full with weeds ("planten") may have Strickler coefficients as low as = 5 - 10 w}^^/s.
Some values are presented in table 2.1.
Table 2.1. Strickler coefficients k in ml/3/s.
STRAIGHT ALIGNMENT
range
normal
D e s c r i p t i o n of channel
ml/^3/s
LINED CHANNELS
Steel
Brickwork
Concrete
ml?3/s
75
55
65
100
65
75
MEANDERING CHANNELS
range
normal
mli^3/s
ml/^3/s
- 110
- 80
- 90
EXCAVATED CHANNEL
clean, short grass
some weeds
brushwoods
standing timber
40
30
20
15
25 - 45
20 - 35
10 - 20
5-15
NATURAL CHANNELS
clean, short grass
some weeds
brushwoods
standing timber
35
30
15
10
25 - 45
20 - 35
10 - 20
5-15
FLOOD PLAINS
short grass
crops
brushwoods
heavy timber
30
20
15
10
20 - 35
15 - 25
10 - 20
5-15
30
25
10
5
20 - 40
15 - 30
10 - 15
5-10
Programme 'Profile'. The computation with the Strickler formula is somewhat cumbersome:
the discharge Q can be calculated directly when the other parameters are known. But, the
water depth y or the bed width b can only be calculated by iteration. The formula is easily
progranunable. The PC-programme 'Profile' can be downloaded from the internet site of
the Section Water Management (http://www.landandwater.tudelft.nl). The programme is
'public domain', and can be copied freely. The output of the programme 'Profile' should be
printed as a "File". This file has to be entered as an "ASCII-text file" into e.g. Word. A
sample of the output is shown in box 2.1.
Box
2.1. Results from the computer programme 'Profile'.

FlhEtm-tE
DATE : 07-13-2002:
iiiiiiiiiii i i ^ l i ^ l i i w i i l
i i i i i i i " I H I l ks ( i / x i )i i i i l B B P i i i i i
m/s
N/m"2
m
m " l ^ 3 / s l i l i H i (v:ni} 1 0 " '3 (b/h)
m"3/s
25.00
25.00
25.0
15.0O
ISO.00
150.00
1,15
0.98
0 86
2 "G
2.40
2 .46
40.00
4C.00
40.00
40.00
40.
40.00
15 00
20,00
25.00
20.00
00
20.00
2.00
2.00
2.00
2.00
^.OU
2 00
1.00
1.00
1.00
1.00
1.
1 50
20.33
28.92
7.24
Li),t$
8 12
1*26
1,1
1.0&
2.13
a.04
2.44
9.65
S,48
27.11
E
K/m"3
12. 31
11. S5
10, 61
20. 86
1$.9S
35.25
35. 94
40
Composite sections. Above, trapezoid-shaped channels have been discussed. Flow through
'composite sections' ("samengestelde doorsneden") is also possible. These channels with
composite sections can be divided into, see figure 2.2:
channels with 'berms' ("bermen") are characterized that the berms run parallel to the
main channel. Thus in the hydraulic calculations, the berms have the same gradient
as
the gradient
of the main channel;
channels with 'floodplains' ("uiterwaarden") are characterized that the main channel is
meandering between the embankments. Thus in the hydraulic calculations, the floodplains
have a steeper gradient S2 as the gradient
of the main channel.
gradient SI
SI
^>nXV
NO F O R E L A N D S
WITH FORELANDS
FORELANDS A N D
M E A N D E R I N G CHANNEL
Figure 2.2. Composite cross-sections with berms and with forelands.

Flow over berms and floodplains. The flow through the main channel is normally higher than
the flow over 'forelands' ("buitendijksland"),
i.e. the berms and the floodplains, because:
o the main channel is hydraulically smoother as there is less vegetation, so the Strickler
coefficientfc^hannel> ^forelands:
o the main channel has a deeper water depth, so the water depth ychannel > ^forelands;
o the hydraulic radius of the main channel is reduced by a smaller wetted perimeter ("natte
omtrek").
On the other hand, the bed width ^channel
*e main channel is often smaller than the width
^forelands of the forelands, so that b^^^^^^ < b^^,^^^^^,.
Strickler formula twice. Flow through channels with forelands ("buitendijks
approxunated by applying the Strickler formula twice:
^
channel
,
f channel +
2/3
2yJ\+m^\
land") can be
1/2
{'^channel
-yg'^ ^ horelands
^ (3^ - yp)]
,
.1/2
,
2/3
fforelands]
f forelands + ^ (y - y^) / 1 + m ^ |
where: Q is the total discharge through the channel and over the forelands in m^/s, ^channel
is the bed width of the channel in m, ^forelands *s the width of the forelands, y^ is the depth
of the channel bed below the forelands in m, y is the water depth above the channel bed in
m, m is the side slope (lyg^^ : mpjoj.) of the channel banks and of the embankments, ^channel
is the Strickler coefficient of the channel in m^'^ls; fcforeiands *e Strickler coefficient of
the forelands in m^''-^/s.
forelands
The gradient -Schannel *e channel equals the gradient ^forelands

run parallel with the channel.
^^ê of berms which
The gradient ^channel of tê channel is less than the gradient .Sforeiands
case of the
channel is meandering through the floodplains. It is obvious that the calculation of the latter
case is subject to many approximations which makes the calculation less accurate.
41
Hydraulic exponent N. The Strickler formula can also be re-written into (e.g. Chow 1959):
Q = cy^
where: Q is the discharge in m^/s, c = k X s^'^ is a coefficient in m^'^/s, y is the water
depth in m, and N is the hydraulic exponent.
The hydraulic exponent A'^ depends on the water depth y, so that the aboye equation can
only be applied for specific calculations. Examples are the 'sensitiyity' at a proportional
distribution, with small values of Ay.
The value of the hydraulic exponent depends on the side slope lyg^^ : Wg^j. and on
the width-to-depth ratio n = biy between the bed width b and the water depth y, and follows
from, see box 2.2:
5
n + 2m
n +m
\fl+rn^
3 n^ +
^ i\J I 2\f\^^~^
+ m'
Typical, the hydraulic exponents A^ range from:
o A'^ = 2.01 for small channels with side slopes m = 1 and the width-to-depth ratio n = I,
o N = 1.95 for channels with side slopes m = 1.5 and the width-to-depth ratio n = 2,
o A^ = 1.83 for larger channels with side slopes w = 2 and the width-to-depth ratio n = 5.
Box 2.2. Derivation of the hydraulic exponent N (see also Chow 1959).
The Strickler lormula Q - kA R"'^ s^- can be re-writtcn as a function of the water
depth V (Chow 1959): Q = c y^, where A' is the 'hydraulic exponent'.
Processin} of the equation with the hydraulic exponent:
Taking the logarithm on both sides of the Q - c
gives: in - In c + TV In y,
and differentiating with respect to>', leads to:
i i l i l ^
îm,andso:iKl!l)
^ .
MiilBiiiiliiiiiii^
Processing of the Strickler equation:
The hydraulic radiuj, R in the Strickle! fornnila can be exprej,sed in R ~A
where P is the perimeter of the wet cross-scctitm, so Q = k A'''-'' P'-'^
s^'-.
Taking the logarithm on both sides o f the Strickler formula gives:
In Q = In {k .s''") + "X:, In A - '1^ In P,
and differentiating with respect to leads to:
d(lnO)
A J_ I d
1 1 ^
'!'
3 A
dr
3 P iXy
11\
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
^ = / ) V +/>M--, so that ^
dfcjlZir)
h.2,ny
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
-~ / > . 2 r v l ' ^ ^ ^ s o t h a t i ^
d(/. ^ 2 y v ' T : ; ; ; ^ )
^ r ; ^
iiliiiiiiiliiiiii^
IHlBiBliillil^
d{lnO)
5 h+2my
dy
3 f j y ^ jny'3
Solving ,V from the two equations:
i i B i i i i i i f " ' " ^ ^
5 J _^
^ ^
y-j-i m-
2 v ' l -> nr
, 2 j ' v ' l < nr
5 n + 2ni ^ 4
^îiiiiir*'"''''"'"'"^
J\_
42
2.1.2. Bankfull Capacity

Natural levees. Figure 2.3 shows the cross section of a river reach in a delta. The capacity
of natural rivers in deltaic areas is limited, and the higher discharges will overtop the river
banks. During these overtoppings, the heavier sediments, such as sand and silt, are deposited
first on the riverbanks, i.e. on the 'natural levees' ("oeverwallen"), and the finer particles,
clay, are transported further to the 'backswamps' ("komgronden").
The relative height of natural levees in the upstream parts of a delta may be as much
as 2.00 - 3.00 m, and the width as much as 2 - 3 km. Natural levees offer the best
possibilities for human settlement, as they are better drained than the backswamp. Also, the
depth of flooding is smaller than in the backswamps. Even in developed and protected deltas,
most of the population is still living on these ridges.
Backswamps consist of heavy soils with a low permeability. They are often
waterlogged, even during dry periods. In some deltas, the formation of peat takes place in
the backswamps.
'Back Swamp'
Hca'^y soils
Figure 2.3. Natural river with 'natural levees' during overtopping.
Bankfull capacity. The discharge which just fills the channel is often called the 'bankfull
discharge' ("kapaciteit van het rivierbed").
The 'bankfull capacity' or the 'bankfull discharge' of a channel is defined as the
"discharge corresponding to the state at which the river berms are just about to be
submerged" (ICID 1967).
Field data show that the crest of the natural levee is raised to the level of a flood with
a remrn period ("herhalingstijd") between 0.5 to 1.5 years, depending on the nature and the
duration of the floods. Floods with remrn periods of 10 to 100 years are too rare to build up
the levee to that elevation. It means that the 'bankfull discharge' of a natural river reach
may equal to the 0.5 to 1.5 years flood (Henderson 1966).
Calculation of the bankfull capacity. The calculation of the bankfull capacity of a crosssection can be done with the Strickler formula for a known Strickler coefficient k in m^'^ls.
The following procedure may be followed, see figure 2.4:
determine the gradient of the chaimel from the longitudinal section. Therefore, draw a
straight line through the lowest points of the bed, i.e. the 'thalweg' ("talweg"). Read the
gradient s from this line;
schematize the bottom of the channel-bed by drawing a horizontal line in the cross-section.
Read the bed width b from this bottom-line;
draw the maximum water level in the cross-section at which 'bank-overtopping' begins.
43
Read the maximum water depth y between the 'bottom-line' and this water-line;
make a schematization of the side slopes, and draw two side-lines of the channel, starting
from the bottom-line with width b. Read the width w of the water-line between these sidelines. Calculate the side slope m (lyert ^Hor) ^^^^ ^he equation:
w
= b + 2my,
or directly f r o m :
2y
deepest point in the cross section
GRADIENT s FROM 'THALWEG'
b, y AND m FROM CROSS SECTION
Figure 2.4. Determination of the 'bankfull capacity'.
2.1.3. Non-uniform flow

Types of backwater curves. The water levels in drainage channels and in rivers often form
backwater curves ("stuwkrommen"). The water levels may form either, see figure 2.5:
'positive' backwater curves, when there is a backing up ("opstuwing") of the water level,
'negative' backwater curves, when there is a drawing down ("afzuiging") of the water
level.
uniform
flow
uniform
flow
enei^jine
"Control"!
CREST
-^REST
POSITIVE B A C K W A T E R C U R V E
NEGATIVE BACKWATER CURVE
Figure 2.5. Positive and negative backwater curves.
44
Energy depth. Backwater curves are determined by the energy depth at the 'tail end' or
'downstream end' of a chaimel reach ("pand"). This 'downstream end' can be formed by:
the control section ("stuwgedeelte") of a structure, such as a 'fixed weir' ("vaste stuw"),
a 'control notch' ("vernauwing"), a screen ("vastscherm"), or a 'gate' ("schuif", "deur"),
see figure 2.6;
the downstream canal reach. In the following, the conditions at the downstream canal
reach are also called: 'control'.
WEIR
NOTCH
SCREEN
MANUAL
GATE
HYDRO-MECH.
GATE
Figure 2.6. The 'control section' of a structure determines the energy head ^^control-
Backing up / drawing down? The water line in a reach forms a curved line for non-uniform
flow:
there is a backing up ("opstuwing") when this energy depth Hôatml
the control above
the bed level is more than the uniform energy depth ^ûniform' thtis when AH = ^control"
ûniform is positive.
there is a drawing down ("afiuiging") when AH = //control " 'ûniform is negative.
Obviously, there is uniform flow for A H = /^control " -ûniform ~ ^
Use of programmable pocket calculator. The calculation of a backwater curve by hand is a
laborious matter, specifically when the cross section is trapezoid.
It is more practical to prepare a computer programme. Such a programme can be based
on the 'standard step' method, and will discussed below.
Parameters. The water depth
at a location x = L from a 'control' is determined by the
following parameters, see figure 2.7:
the energy depth H^ in m at the control,
the discharge Q i n m^/s,
the characteristics of the channel reach, i.e. the bed gradient 5", the bed width b in m, the
side slope m (lyert '^Hor)' ^îd the Strickler coefficient k in m^^^/s.
45
2. OPEN C H A N N E L F L O W
to be calculated
"Control"
CROSS SECTION
L O N G I T U D I N A L PROFILE
Figure 2.7. Parameters for the calculation of backwater curves.

Vertical 1. The calculation is done step-by-step from vertical 1 to vertical 2, see figure 2.8.
Subsequently, vertical 2 becomes the new vertical 1, and the next vertical 2 is calculated, etc.
The first vertical 1 is taken just upstream of the 'control'. There is no friction loss
between the 'control' and vertical 1, so
= H^.
The water depth y^, the velocity
and the gradient s of the energy line have to be
calculated at vertical 1:
known are: the discharge Q in w?/s, the bed width b in m, and the side slope m (lyert
^Hor)- Thus, water depth y^ and the velocity
at vertical 1 can be calculated from the
two equations through iteration:
2g
_Q_
where
(b + my^) y^
The iteration is done as follows:

o assume V j = 0 , calculate y^ = H^^
I 2g and v^ = QI \_{b + m y{) y{\.
check
the
'new'
with
the
'assumed'
v^. I f there is still a difference, repeat the
o
iteration with the new v^. Otherwise, the value of the velocity Vj and the value of the
water depth yy have been found;
known are: the velocity Vy in m/s, the water depth y^ in m, the bed width b in m, the side
slope m (Ivert " ^Hor) ^"^^ ^he Strickler coefficient k in m^^^/s. Thus, gradient
of the
water line at vertical 1 can be calculated from the Strickler equation:
(h
+ myj)yj
where : R
k^R'^'
ly,
{ + m-
Difference in energv head. Vertical 2 is defined on basis of the change AH in the energy
depth, and on the shape of the backwater curve, see also figure 2.8:
for a backing up ("opstuwing") when the energy depth //control the control is larger than
the uniform energy depth //uniform- -^2 ~ ^ l + ^> ^^^h AH < 0.
for a drawing down ("afzuiging") when the energy depth //control the control is smaller
than the uniform energy depth //uniform- ^ 2 ^ ^ 1 +
"^^th AH > 0.
The value of difference AH is often taken at AH = - 0.01 m for backing up, and AH = +
0.01 m for drawing down.
46
Vertical 2. As the new H2 at vertical 2 is known now, the water depth

the velocity V2
and the gradient S2 of the energy line can be calculated accordingly to the above calculations.
Mutual distance. The mutual distance Ax between vertical 1 and vertical 2 follows from the
standard-step equation (e.g. Chow 1959):
s, + s.
where: Ax mutual distance in m between vertical 1 and vertical 2, AH = H2- H^ in m,
is the energy gradient at vertical 1, ^2 is the energy gradient at vertical 2, and S is the
gradient of the channel bed.
Distance from the control. The total distance from the 'control' at .x = 0 can be calculated
by, see figure 2.9:
^2 = Xj + Ac, or in general : ^2 = 2 x + Ac
Check the calculated distance X2 with the required distance L to the control:
when L >
calculate next vertical, by considering the vertical 2 becomes the new
vertical 1;
when L < X2, the calculated water depth ^2 is the required water depth
= J2
the
distance x = L from the control.
An example of a backwater calculation is shown in box 2.3.
47
vertical "2"
vertical " 1 "
to be calculated
Figure 2.9. Calculation of total distance from the 'control'.
Box 2.3. Example of a backwater curve.

A lined urban drainage channel is constructed with the following characteristics:
bed width b = 6.00 m, side slope m ^ 1 (1 j ^ ^ , ^ : Ivcrt)'
gradient S = 0.24 X 10"-,
and a Strickler roughness coefficient k ~ 55 m^'^/n.
'i he channel is designed for a normal discharge of Q = ^ . 4 3 nv'M. A corresponding watei depth of v = 1.60 m and velocit)' v = 0.93 m/s can be calculated with
the Strickler foinuila.
i The channel has been equipped with an automatic water level regiilator (gate)
so the water level ol v = 1.60 m is always maintained at thai location. The bed level,
at this rcgulattir, is at 19.84 m4-Datum.
^ A computer programme may calculate a backwater curve for a discharge of Q ~
5 ni'/s. at the lollowing distances from the regulator:
"X"
in. in
waterdeptb
in m
(1)
aoo
400
eoo
800
1000
120
1400
1.60
1.S6
1.51
1.48
1.44
1.41
1,38
1,34
1,31
1.28
1,2S
1^00
1800
2O00
i--eioclty t r a c t i v e
m m/s
0.41
0.42
0.44
0,4S
0.47
0.48
0.45
0.51
0.52
0.S4
0.S$
force
S i i i i i f t ^ i i ^
iîiliiiiiiiliiil
0.84
0.89
0.95
1,01
1,0^
1.14
1.20
1,27
bedlevel
m+Datum
19,84
19.89
19. 94
19 . 98
20,03
20,08
20,13
20,18
22.22
20,27
20.32
waterlevel
m+Datum
21.43
21.45
21.46
21.46
21,47
21.49
21,51
21. 52
S I , 53
21,55
21.57
48
2.2. Flow of Sediment

2.2.1.
Sediment
Sediment transport. It is not only that the 'flow of water' through a channel should be
guaranteed, but also the 'flow of sediment' must be sufficient.
The main rule for the design of stable chaimels is that: "all sediment which enters the
channel reach should be transported through without sedimentation".
Thus, the 'relative transport capacity' is of importance. When the relative transport
capacity reduces, it will lead to sedimentation.
Classification of Sediment. A classification of sediment transport can be made on basis of the
transport mechanism of the transported material, see figure 2.10 (e.g. Jansen 1979):
'Bedload' ("bodem transport") is the movement of particles in contact with the bed by
rolling, sliding and jumping;
'Suspended-load' ("zwevend transport") is the movement of particles that has no contact
with the bed, and the weight of the particles is continuously compensated by the turbulent
action of the water.
Also a classification of sediment transport can be made on basis of the origin of the
transported material:
'Bed-material load' ("bodem-materiaal transport") has its origin in the bed, which means
that the transport is determined by the bed and flow conditions;
'Wash load' ("spoeltransport") has its origin outside the bed and is supplied by erosion
in the catchment area. This material has no direct relation with the bed-material and is
generally fine material (D < 0.060 mm).
Importance of washload. Typical figures from deltaic areas in Indonesia, Nigeria and the
Netherlands show that only a small portion ( + 10%) of the annual sedmient transport is
transported as 'bedload', and the remainder as 'washload'.
TOTAL
SEDIMENT
TRANSPORT
TOTAL
SUSPENDED LOAD
WASH LOAD
(direct erosion material)
1-C
SUSPENDED
BED-MATERIAL
BEDLOAD
(from bed)
1
I
Figure 2.10. Sediment transport classification.

Concentration. The concentration of suspended-load in a sediment-water mixture is usually
expressed in 'concentration'. However, the concentration can be expressed in various ways,
of which the most commonly used are:
o concentration by weight, that is the dry weight of solids per unit weight of mixture,
expressed in kg/m^ or mg/l and also often in ppm;
o concentration by volume, the absolute volume of solids per unit volume of mixture,
expressed in ppm.
49
Ppm. It is customarily to express the sediment concentration in a concentration by weight in

mg/l (=ton/m^) or in 'ppm' (parts per million). One per cent (1%) equals 10,000 ppm.
The conversion of 'mg/l' to 'ppm' assumes that 1 ppm equals 1 milligram sediment per
1 liter of water, i.e. 1 mg sediment per 1000 g water, so 1 mg/l = 1 ppm.
Fall velocitv. The fall velocity of a sediment particle is an important parameter in studies on
suspension of sediments. The fall velocity w in m/s can be calculated with the Rubey formula
for a sediment diameter D < 0.2 mm (e.g. Jansen 1979):
40 X 10'
20 + t
1)
where w is the fall velocity of sediment particles in m/s,
is the diameter of the particles
in m,
= 1000 kg/m^ is the density of clear water,
= 2600 kg/ur' is the density of
sedmient, g = 9.S m/s^ = 9.8 N/kg is the acceleration of gravity, v is the kinematic
viscosity, and t is the temperamre in C .
The Rubey formula becomes for a water temperature t = 20 C :
w
with:
w ~ 0.87 X 10^ D j
A warning should be given on the use of this formula, as many factors such as the shape of
the particles, make that this formula is quite inaccurate!
Dry Density. Sometimes, it is necessary to convert the weight of the sediment into a volume.
This may happen to estimate the life of a reservoir. Thus, the 'dry density' must be known.
The dry density p is the mass of dry sediment per unit volume, in kg/m^. It depends
on the grain size, the condition of settling and on the time. A n empirical relation for
estimating the dry density of deposits in a reservoir follows from the literature:
P j = Po + 5 log r
where: pj is the dry density after T years in kg/nP, p^ is the initial dry density (for T = I
year) in kg/m^, B is the consolidation coefficient, see table 2.2, and T is the consolidation
time in years.
Table 2.2. Dry density of sedmient deposits.
Submerging
of t h e
Sediment D e p o s i t s
Dry d e n s i t y a t T = 1 y e a r
sand
silt
clay
kg/m3
kg/m3
kg/m3
a l w a y s submerged
o f t e n submerged
sometimes submerged
normally dry
1500
1500
1500
1500
500
750
950
1250
1050
1185
1275
1320
Consolidation coefficient
sand
silt
clay
B
B
B
0
0
0
0
90
45
15
0
250
170
100
0
2.2.2. Bedload transport capacity

Einstein-Brown formula per unit width. The Einstein-Brown formula has been derived
empirically for bedload transport, although it might also be valid for suspended-load. The
formula reads (e.g. Henderson 1966, Raudkivi 1993):
R s
40
G
or:
gD'
1)
50
_ 4 ^ ^ ^ _ i _
- 1 2
^3,3
,,_gl_
Z)2
^3,3
jy2
J
.
where:
is the sediment transport per unit width in nr/s, R is the hydrauhc radius of the
cross-section in m, s is the gradient of the channel, G 2/3 is the particle fall function, D
is the grain size of sediment in m,
= 2600 kg/n? is the density of the sediment in kg/m?,
p = 1000 kgltc? is the density of water, g = 9.8 m/s^ = 9.8 N/kg is the acceleration of
gravity.
Thus, the 'relation' between the sediment transport q^ per unit width, the hydraulic
radius R and the gradient of the chaimel is:
q, ~ R^ s\
Einstein-Brown formula for the whole cross-section. The Einstein-Brown formula can be
transformed into the bedload transport
for the whole cross-section by multiplying with the
width b. Thus, the total bedload transport
in vc?ls becomes according to the EinsteinBrown equation:
s ~
bR^s^.
Box 2.4. Further processing of the Einstein-Brown formula.
There arc two formulae important for the "relative transport capacity' QJQ of the
bedload through a cross-section:
* the Kiti.stein-Brown formula for the total bedload transport Q, in nr'ls reads:
Q, hR^s\
the Strickler formula for the flo\^ of water reads: Q
A R^'^ s^^^.
Thus, the 'relative transport capacity' SlQ of the bedload becomes:
~
- r r ^ r .
and sa: A
- 4 /f-.vQ
A R^^ s^^
0
\
where
is the bedload transport per cross-section in nrVs, Q is the discharge of water
in m- /s. b is the bed width in m. A is the wet cross-sectional area in n r . R is the
hydraulic radius i n m, and s is the gradient of the channel.
The hydraulic radius R for wide channels becomes equal to the water depth y:
which means that the hydraulic radius becomes: R Âlb. For narrow channels of a bed
width h = 30 m. side slopes m = 2 and a water depth v = 2 m to 5 m, the hydraulic
radius R can be calculated at R ^(A/bf^-^.
However, the hydraulic radius R is
as.sumcd here at R ^(A/bf^'
in order to obtain a 'practicaP relation with the veiocit}'
and the gradient. So:
O
A
The velocity r folIov\s from the Strickler formula v - R-'-'
so that the hydraulic
radius can be substituted in the above formula through R ~ t-^'-
^liiiliiplii^^
^^M^^^^^^^^^.. ...^^^^^^^
......
...........
It means that the 'relative transport capacity' Q^fQ of thf^ bedload through a crosssection is related to the velocit>' v in m/s and the gradient s by means of:"
51
Relative transport capacity. Channels should be designed in such a way that all bedload which
enters the channel reach should be transported toward the end without sedimentation. Thus,
the 'relative transport capacity' "QJQ",
which is the total bedload transport s per total
discharge Q, is of importance. When the relative transport capacity QJQ reduces, it will lead
to sedimentation.
Further processing of the Einstein-Brown formula, see box 2.4, leads to the relation
between the 'relative transport capacity' QJQ of the bedload through a cross-section, the
velocity v in m/s and the gradient 5 by means of:
n
12
i
Q
It means that the factor "v X s" should not decrease in downstream direction to prevent
sediment deposits. It will be shown below that the criterium for bedload transport is equal to
the criterium for the transport of washload.
2.2.3. Suspended-load transport capacity

Suspension as an energy problem. Stodying the suspended-load at different depths of rivers
and canals, it was found that particles smaller than 0.06 mm generally are uniformly
distributed over the water depth (Vlugter 1962). It would appear that these particles are not
'sinking' but are 'floating'.
Several researchers (deVos 1925, Vlugter 1941 and 1962, Bagnold in 1956, Wang in
1959, Schoemaker 1983^) considered suspended-load transport as a problem of energy. They
considered that "the fluid must lift the solids at the rate that they are falling under gravity".
The most systematic derivation of the energy equation has been done by Vlugter
(Vlugter 1941), and is followed here. He compared the flow of water without sedmient, 'clear
water', with the flow of a water - sediment mixture, see figure 2.11. Through an energy
balance, the criteria for suspension are finally reached.
VELOCITY
(Clean Water)
v sin a
LOSS OF POTENTIAL ENERGY

(WATER WITHOUT SEDIMENT)
LOSS OF POTENTIAL ENERGY

(WATER & SEDIMENT MIXTURE)
Figure 2.1L Loss of potential energy by water 'without' and 'with' sediment.
52
Loss of potential energy in clear water. When 'clear water' is flowing with a velocity v^,
it will lose potential energy. The loss of potential energy, in Watt or in Nm/s, by flowing
clear water with a volume
in m'^ depends on the drop in vertical distance "v^ sin a", and
reads:
Pw^^wôSino;
where:
= 1000 kg/m^ is the density of clear water, g = 9.8 m/s"^ = 9.8 N/kg is the
acceleration of gravity,
is the volume of water in the channel in m^, a is the gradient of
the chaimel in radials,
is the velocity of the clear water in m/s.
This loss of potential energy equals the losses by friction at the bed and the side slopes.
It means that the friction losses of clear water, in Watt or in Nm/s, can be calculated from:
Pw S
^0 sin Oi
Water - sednent mixture. When 'sediment-loaded water' is flowing with a (higher) velocity
V j j j , there are three energy components, in Watt or in Nm/s:
the potential energy that comes available by the drop in vertical distance "Vj^^ sin a":
Pw S
% sina +
g
sina
the energy needed for suspension equals the energy that comes available for 'sinking'
sediment, thus:
(Ps - Pw) ^
^
the friction loss of the water - sediment mixture is higher than the friction loss of clear
water:
g
sin a, because of the higher water velocity V j ^ > v^.
The friction losses are related to 'square' of the velocity (conform the Strickler formula
V s^^-^), so the higher water velocity v^^ of the water - sediment mixture makes that the
friction losses increases by a factor " (v^ I vj^ ", so:
Pw S ^w
sina X (v^, / vj^
where: p ^ = 1000 kg/m"* is the density of clear water, pg = 2600 kglvc? is the density of
sediment, ^ = 9.8 m/s^ = 9.8 N/kg is the acceleration of gravity,
is the volume of
water in the channel in m-^, Vg is the volume of sediment in the water in m^, w is the fall
velocity of the sedmient in m/s,
is the velocity of the clear water in m/s, V j ^ is the
velocity of the water - sediment mixture in m/s, a is the gradient of the channel in radials.
Energv balance for water - sediment mixture. The energy balance for the water - sedmient
mixture reads:
Potential energy = energy needed for suspension + friction loss
^w S ^w
sina +
g
sina =
= (Ps - Pw) ^
w + Pw g ^w
sina ( V j ^ / vj^
A fiirther processing of this equation is shown in box 2.5, and leads to the formula of
Vlugter for the concentration C by weight:
=
J
^
sin a
where: w is the fall velocity of the sediment in m/s,
is the velocity of the clear water in
m/s,
is the velocity of the water - sediment mixhire in m/s, a is the gradient of the
channel in radials, and A^ = (p^ - p^)/ pg = 0.6.
53
Box 2.5. Further processing of the energy balance.
The enerfo balance ibr liie water - sediment mixture rcadj,:

=
(P,-PJ.^';IV
-f
p^, g { i ; , sma
(rj/v,,r
The balance is divided by "p^^ g 1 '^^ \\^^ sina":

K, .sina tp.^ ^ w" is brought lo the left:
and the teim "p^ i ;
Kv
i\ - pTV
p.
'ft)
P.
sm
p,
As \ - (p^/ p^, and as the term "p^ i ; / p.^ y^," equals the concentration C of
tiic sediment by weight, finally the tbrmiila of Vlugter is reached:
P.
A, w
Discussion on the formula of Vlugter. The meaning of the formula of Vlugter lies in the
term:
There are three options for this term:

when the term (Ag w) / (v^ sin a) = I, the concentration C becomes ' o o ' . Thus, the
sediment with a fall velocity w remains in suspension, at any concentration C. It means
that the washload can be transported through the chaimel reach;
when the term (Ag w) I {v^ sin a) > 1, the concentration cannot exist and deposition of
the sediment will happen;
when the term (Ag w) I {v^ sin a) < 1, the sedmient with a fall velocity w remains in
suspension, while even the value 'v^ sin a' can be decreased. This leads to the 'de Vos
& Vlugter criterium' for suspended sediment transport. This is in line with the results of
Bagnold and Wang (e.g. Raudkivi 1976, Chien 1999).
Criterium of 'De Vos & Vlugter'. The above rule on suspended sediment transport was first
found by De Vos in 1925 and later re-presented by Vlugter in 1941:
sediment with a fall velocity w remains in suspension for: v X sin a > Ag X w
as the energy gradient ^ = sin a and Ag = 0.6, the De Vos - Vlugter criterium becomes:
"washload with fall velocity w remains in suspension for v * > 0.6 w", where v is the
velocity of the water in m/s, s in the gradient, and w is the fall velocity of the sediment
with a diameter
in m.
Criterium Schoemaker. Schoemaker presented an energy model in line with the above
thoughts of de Vos and Vlugter, and derived the relation (Schoemaker 1983*'):
E = p gvs
where: E is the energy dissipation of the mixmre in Watt/m^, p = 1000 kg/m^ is the density
54
D E S I G N O F O P E N - C H A N N E L S A N D H Y D R A U L I C S TRUCTURES
of water, g = 9.8 m/s^ = 9.8 N/kg is tlie acceleration of gravity, v is the velocity of the
water in m/s, and s is the gradient of the channel.
Thus, the 'criterium of Schoemaker' reads: "the transport capacity for suspended
sedunent is guaranteed when the value of E = p g v s does not decrease in downstream
direction".
2.2.4. Discussion on Sediment Transporting Capacity

Relative transport capacity. It has been discussed above, that the main rule for the design of
stable channels is that: "all sediment which enters the channel reach should be transported
through without sedimentation". Thus, the 'relative transport capacity' is of importance.
The 'relative transport capacity' "QJQ" is the total bedload transport Q^ per total
discharge Q. When the relative transport capacity QJQ reduces, it will lead to sedimentation.
Criterium for bedload. The relative transport capacity QJQ of the bedload through a crosssection depends on the velocity v in m/s and the gradient s by means of (as derived above):
n
io i2
~ V ' s '
Q
It means that the factor "v x s" should not decrease in downstream direction to prevent
sediment deposits.
Criterium for washload. Also the relative transport capacity QJQ of the washload through
a cross-section depends on the velocity v in m/s and the gradient s by means of (as derived
by Vlugter and Schoemaker):
Qs
V s
Q
Also here, the factor "v X s" should not decrease in downstream direction to prevent
sedmient deposits.
Criterium of Schoemaker. Apparently the 'criterium of Schoemaker' can be applied for the
relative sediment transport capacity for both bedload as washload.
Thus, "the transport capacity of all sediment is guaranteed when the value of:
E = p g V s does not decrease in downstream direction", where E is the energy dissipation
of the mixture in Watt/m^, p = 1000 kg/m^ is the density of water, g = 9.8 m/s^ =
9.8 N/kg is the acceleration of gravity, v is the water velocity in m/s, and s is the gradient
of the channel.
Workable form. Sometimes, a more workable form of the criterium of Schoemaker is
obtained by replacing the velocity v by the ratio "Q/A", so that:
E
= Pg ^
s
A
where E is the energy dissipation of the mixmre in Watt/m^, p = 1000 kg/m^ is the density
of water, g = 9.8 m/s^ = 9.8 N/kg is the acceleration of gravity, Q is the discharge in m^/s,
A is the wet cross-sectional area in m^, and s is the gradient of the channel.
55
Meaning for the cross-sectional area. The above relation can be made more operational for
the design by considering the relation emerging from the Strickler formula:
so that:
E ~ p g X R^'^ s^'^ X s
and finally the design condition for maintaining the 'relative transport capacity' E of the
sediment through the alignment:
R ~
s-'l'
with the hydraulic radius i? in m and the chaimel gradient s. The meaning of this condition
is that the hydraulic radius i? of a sediment-loaded channel should increase for a decreasing
gradient s, see figure 2.12.
Cross-Section with a smaller R

for 'relative transport capacity'
Cross-Section with a larger R

for 'relative transport capacity'
at a steeper gradient
at a flatter gradient
Figure 2.12. Changing cross sections for stable sediment transport at changing gradients.
2.3. Scouring and Sedimentation in Cross Sections

2.3.1.
Tractive
Force
Total tractive force on reach. The 'total tractive force' ("totale schuijkracht") is the pull
of flowing water on the wetted area of a reach of a channel ("kanaalpand"). I n uniform flow,
the tractive force is equal to the component of the gravity force parallel to the channel
gradient. Thus, the tractive force on a channel reach follows from, see figure 2.13:
F = p
gALs
where: F is the total tractive force on bed and slopes of the reach in N , p = 1000 kg/m-^ is
the density of water, g = 9.8 m/s^ = 9.8 N/kg is the acceleration of gravity, A is the wet
cross sectional area in m^, L is the length of the channel reach in m, and s is the (energy)
gradient.
56
Tractive Force on Reach
Distribution of Shear Stress
Figure 2.13. Distribution of the shear stress or 'tractive force'.

Unit tractive force. The 'unit tractive force' is the average value of the tractive force per
unit wetted area. It is also called the 'average shear stress' ("gemiddelde
schuifspaming").
The unit tractive force can be calculated from:
= p gALs/(PL)
=
pgRs
where:
is the unit tractive force per unit wetted area in N/w?, P is the wetted perhneter
in m, is the hydraulic radius in m.
The tractive force is not uniformly distributed along the wetted perimeter. Thus, this
average value of the unit tractive force cannot be used in the design of stable channels, as the
actual tractive force on the bed is higher than the average value.
Distribution of unit tractive force. USBR calculated in 1952 the distribution of the (unit)
tractive force in trapezoidal cross sections through membrane analogy and analytical methods.
The resuhs are extensively copied in the literature (e.g. Chow 1959, Henderson 1966). A
typical distribution over the bed and the slopes is shown in also figure 2.13.
Peak value in the cross-section. It has been found by USBR that the peak value of the unit
tractive force is dependent on the ratio b/y and on the side slopes m. The peak value of the
tractive force at the bed T^^^, and of the side T^^^^ follows from the general formulae:
^bed = bed P gysmd
Tgide = ocîde p gys
where aêd
"side
coefficients, see table 2.3. It can be seen that the coefficient a^gj
for the tractive force at the bed can be approximated by a^^^^ = 1.
Table 2.3. Coefficients for the peak tractive force on trapezoidal cross sections.
b / y RATIO
0.5
1
2
3
4
5
7
10
VERTICAL SIDE SLOPE

on b e d
on s l o p e
^bed
^side
0 .20
0 . 40
0 . 68
0 . 87
0 . 94
0.30
0 .46
0 . 68
0 . 73
0 . 74
S I D E SLOPE I V : 1.5H
on b e d
on s l o p e
bed
"side
0.72
0.80
0.89
0 . 94
0 . 97
0.98
0 . 99
1.00
0.67
0.70
0 . 73
0 . 74
0 . 75
0.76
0.76
0.76
S I D E SLOPE I V : 2H
on bed
on s l o p e
bed
"side
0.72
0.80
0 . 89
0 . 94
0 . 97
0.98
0.99
1. 00
0.70
0.73
0 . 74
0 . 75
0 . 76
0 . 77
0 . 77
0 . 78
57
2.3.2. Critical Tractive Force on the Bed

Critical tractive force. The tractive force on the bed should remain below a certain critical
value, i.e. the 'critical tractive force' T^p as to prevent scouring. Thus:
T = p g y s and T < T^^
where: T is the tractive force on the bed during the bed-forming discharge in N/m-^, T^^. is
the critical tractive force for that soil in N/m^, p = 1000 kg/m^ is the density of water, g =
9.8 m/s2 = 9.8 N/kg is the acceleration of gravity, y is the water depth during the bedforming discharge in m, and s is the energy (water) gradient.
Values. The critical tractive force depends on factors like soil type, grain size, sediment
content of the water, etc. Values for the critical tractive force are presented in table 2.4.
The difference between 'sediment-loaded water' and 'clear water' is based on the
observation that sediment-loaded water makes a 'cementation' between the grains, so that they
can resist a higher shear stress. Similarly, 'colloidal soils' (clay, loam) can resist a higher
shear stress.
The critical tractive force of grains sizes of Z) > 5 mm may also follow from (e.g.
Chow 1959):
r = 0.9 D
where: T^j. is the critical tractive force in N/m^, and D is the diameter of the grains in imn,
see table 2.5.
Reduction of tractive force. When the tractive force of a chaimel is too high, a reduction may
be obtained by:
reducing the water depth y by making the width b of the channel bed wider, thus
increasing the ratio b/y. Normally, the hydraulic radius will increase and more earth work
is required;
reducing the (energy) gradient s of the channel. It means that expensive drop structures
("stortdammen") are required, to dissipate the excessive energy in the stilling basin
("woelbalc").
Table 2.4. Critical tractive force T,
SEDIMENT--LOADED
WATER
MATERIAL ON BED AND BANKS
F i n e sand
Sandy loam
S i l t loam
O r d i n a r y f i r m loam
Alluvial silts
Fine gravel
Stiff clay
Alluvial silts
Mixture loam-sand-gravel
S h a l e s and h a r d p a n s
Coarse gravel
Mixture loam-sand-gravel
Cobbles and s h i n g l e s
non- c o l l o i d a l
colloidal
colloidal
colloidal
colloidal
colloidal
CLEAR WATER
Tcr
V*
Tcr
V*
N/m2
m/s
N/m2
m/s
3
3
5
7
7
15
22
22
32
32
32
39
53
7
7
4
3
3
6
5
5
2
7
7
1
7
0 . 061
0 . 061
0 . 073
0 .085
0 . 085
0 . 125
0 . 150
0 . 150
0 . 179
0 . 181
0 . 181
0 . 198
0 .232
1
1
2
3
2
3
12
12
18
32
14
21
44
3
8
4
7
4
7
7
7
6
7
7
0
4
0 . 036
0 . 042
0 . 049
0 . 061
0 . 049
0 . 061
0 .113
0 .113
0 . 136
0 . 181
0 . 121
0 . 145
0 . 211
58
Table 2.5. Size range of particles.

S i z e range
large boulders
small boulders
coarse cobbles
small cobbles
coarse gravel
f i n e gravel
c o a r s e sand
f i n e sand
1
0.50
0.15
0.05
0.01
0.002
coarse s i l t
fine s i l t
coarse clay
fine clay
inm
5
1
0.50
0.15
0.05
0.01
S i z e range
i n mm
2
0.15
0.060
- 10
- 2.0
- 0.150
0.020
0.004
0.001
0.0002
0.060
0.020
0.004
0.0010
2.3.3. Discussion on Scouring

'Shear stress' or 'tractive force'? It is obvious from above that the term 'tractive force' is
actually wrong as the unit is 'N/m^'. The term 'shear stress' is more appropriate, and is also
used in some literamre.
However, the term 'tractive force method' is popular in design practices, and usually
the expression 'tractive force' is also used instead of 'shear stress'.
Tractive force on side slopes. The literatare gives much attention to the calculation of the
stability of side slopes (e.g. Chow 1959, French 1994). This is related to the phenomenon
that the peak value T^^^^ of the tractive force on the side slope is lower than on the bed, while
also the permissible tractive force on the side slopes is lower. However, it is also concluded
in the literature that the rolling-down effect from the side slopes is negligible for chaimels in
cohesive material. It means that the calculation on the side slope stability is generally not
relevant, and that the stability of the bed is normally the determining factor in the design of
stable channels.
Permissable velocitv. In the past, it was believed that the permissible velocity v can be used
as a design condition. It was thought that such a permissible velocity v would prevent the
scouring of the bed, as well as the sedimentation of washload.
Since +1930, the 'tractive force concept' ( ' T = p gy s") is widely accepted as a tool
to describe the physical process of scouring, while it is more and more accepted that the
'energy concept' ("E = p g v s") describes the process of sedimentation.
As an example, different channels with a permissable velocity v = 0.6 m/s are
presented in table 2.6. Apparently, the design discharges (Q = 100 m^/s, 10 m^/s, 1 m^/s
and 0.1 m^/s) have an effect on the tractive force T and on the sediment transport capacity E.
It can be concluded that the 'permissable velocity concept' is not in line with the
modern judgement on the tractive force T and the sediment transport capacity E. Thus, the
'permissable velocity method' should not be considered anymore, as a better methods are
available.
59
Table 2.6. Example of charaiels with a 'permissable velocity': v = 0.6 m/s.

DATE
FILENAME
Q
m" 3 / s
100.00
10 . 00
1 . 00
0 . 10
h
m
1.93
1 . 45
0 . 58
0 . 24
ks=(1/n)
m"l/3/s
b
m
40.00
40 . 00
40.00
40 . 00
132 . 50
8 . 50
1.75
0 .25
m
( v ; mh)
2
2
2
2
. 00
. 00
. 00
. 00
s
10 "
0
0
0
3
-3
. 10
.20
. 80
.50
08-17
- 1996
n
(b/h)
V
m/s
T
N/m"2
42 . 6 7
5 . 86
3 . 03
1 . 06
0.60
0.60
0.60
0.60
1,90
2 . 85
4 . 54
8 . 10
E
W/m"3
0 .59
1.19
4 .67
20 . 15
2.3.4. Discussion on Sedimentation

Side slope and siltation. Experience shows that in channels with considerable suspended-load,
the side slopes become steeper than initially constructed, see figure 2.14.
It is observed that washload is not deposited at the bed of the chaimel, but at the side
slopes. This phenomena is widely reported in the Indian literature (e.g. Garde 1977), but not
in the USA literamre. It can be understood by considering:
the initial sedimentation ("afzetting") takes place at the side slopes. Here, the sedimenttransporting capacity "p gvs" is the lowest at the side slopes, because the velocity v is
here the lowest,
the initial 'scouring' ("uitschuring") takes places at the bed. At the bed, the scouring
forces (tractive force) "p gys" are higher than at the side slopes.
Instabilitv. The result of the depositions on the side slopes makes that the steep slopes become
steeper than constructed, e.g. up to lygj.^ : 0.5jjoj. (m=0.5).
These steeper side slopes are not stable and will slide down after some time. The flow
is obstructed and will scour the opposite bank, thus starting degeneration and meandering.
The energy approach can be used for the design of irrigation systems with washload.
The criterium that 'the energy dissipation E = p gvs should not be decreased below an intial
value somewhere in the system' is simple to apply. It gives also an insight in the effect of the
design variables on the washload transport. The energy approach also illustrates the
unsolvable problem of sedimentation of washload on the side slopes, which ultimately leads
to a capacity reduction in irrigation canals and to unstable banks in rivers.
just after construction
sedimentation starts at slopes ultimate siltation of channel
Figure 2.14. Deposits of sediment on the side slopes of a channel.
Sand trap i n an irrigation svstem. The 'criterium of Schoemaker' reads: "the transport
capacity for washload is guaranteed when the value of E = p g v s does not decrease in
downstream direction". The criterium of Schoemaker can be applied in the design of
irrigation systems with washload. Normally, such a system is equipped with a sandtrap. The
60
sandtrap should not only prevent the entrance of bedload into the irrigation system, but should
also control washload.
An example is presented in figure 2.15. The earlier-designed primary and secondary
canals will provide the value of E = p g v s for each reach, i.e. 0.48 W/m^, 0.72 W/m^,
0.36 W/m^, 0.52 W / m ^ 0.47 W / m ^ etc. The sandtrap has to be designed so that its energy
dissipation is less than the lowest value in the whole system, i.e. 0.36 Watt/m'^, to avoid
sedimentation in one of the canal reaches.
Size of the washload. The size of the sediment that is allowed in the irrigation systems cannot
easily be determined with the above approach. It seems true that the fall velocity w can be
determined from the condition for a infinite concentration of washload according to Vlugter:
w
P - P
with: A , =
V sma
Ps
hence the fall velocity w:
=
and:
sma = s
0.6
while E = p g V s, the equation for the fall velocity w can be re-written into:
0.6 pg
where: w is the fall velocitv of the sediment in m/s, pg = 2600 kg/m^ is the density of
sediment, p ^ = 1000 kg/m^ is the density of clear water, p = 1000 kg/m^ is the density of
water-sediment mixmre, v is the velocity of the water-sediment mixmre in m/s, a is the
channel angle in radials, s is the gradient of the channel, E is the energy dissipation of the
mixture in Watt/m^, and ^ = 9.8 m/s^ = 9.8 N/kg is the acceleration of gravity.
Thus, the fall velocity w of the washload that can be transported through the canal
system of the above example can be calculated at:
w = E / [0.6 p g] = 0.36 / [0.6 X 1000 X 9.8] = 0.061 X lO'^ m/s.
and a corresponding diameter
follows from the Rubey formula:
= (w/ 0-87 X 10^)1^2 ^ 8 4 ^ io-6
= 0.008 mm.
However, a warning should be given on the validity of the value fall velocity as calculated
with the Vlugter formula.
Figure 2.15. The 'criterium of Schoemaker' applied on an irrigation system.
61
To change the energy dissipation. Changing the canal parameters, such as the bed width b,
the channel gradient s, the side slope m and the Strickler coefficient k, haye all there effect
on the yalue of the energy dissipation E. Howeyer, some of the parameters have more effect
on the energy dissipation than others.
As an example, a canal reach with a capacity of = 10 m^/s is designed for a Strickler
coefficient k = rn^'^ls at a bed width = 5 m, gradient s = 0.2 X 10'^, and side slopes
Ivert %or- ^he corresponding water depth y = 1.82 m, and the yalue E of the energy
dissipation amounts to E = 1.25 Watt/m^.
The change of these design variables on the transport capacity of washload is shown in
figure 2.16. The energy dissipation, and thus wasload transport capacity, increases for smaller
a bed width b, a steeper gradient s, steeper side slopes m, and a higher Strickler coefficient
k.
y = 2.22 m
7 = 1.82 m
y= 1.34 m
b= \Qm
2.5 m
= 5m
1.16 watt/m'
E^ 1.28 watt/m'
E= 125 watt/m'
OPTION 1: the bed width 5

/> = 5 m is changed.
7 = 2.17 m
y= 1.82 m
5= 0.1 X 10 '
5= 0.2 X 10 '
E = 0.49 watt/m'
^ = 1.25 watt/m'
Y= 1.52 m
5 = 0.4 X 10 '
^ = 3.21 watt/m^
OPTION 2: the canal gradient 5 = 0.2 x 10"^ is changed.

y-
1 VERT
E-
7 = 1.82m
L58m
4oR {m = 4)
I V E R T : 2HOR
1.10 watt/m'
OPTION 3: the side slope
{m = 2)
E= 1.25 wat^m'
I V E R T ' 2HOR
7 - 2.22
m
y-'-'--^_y
IVERT : IROR ( =
^ = 1.35 w a t t ^ '
is changed.
y = 2.57 m
7 = 1.82 m
v= 1.27 m
k - 20 ni'"/s
i-= 40 m'"/s
1 = 80 m'"/s
^ = 0.75 watten'
E= 1.25 watt/m'
^ = 2 . 0 6 watt/m'
OPTION 4: the Strickler coefficient i r = 40 m''Vs is changed.

Figure 2.16. The effect of parameters b, s, m and k on the energy dissipation E.
1)
62
3. DESIGN OF OPEN-CHANNELS
3.3. Design of Earthwork
3.3.1. Design Procedure
Need to design. Deltas are created by sedimentation during the frequent floodings. But, the
human occupation requires these floodings are controlled and that the land is adequately
drained. So, the hydraulic engineer has to play an active role in the water management of
deltaic areas, and many flood control, irrigation and drainage works have to be implemented.
Improvements of the major rivers in these coastal areas will mainly consist of the
construction of embankments, removal of sedmient and vegetation, straightening of bends,
slope protection, etc.
However, situations may arise where it is necessary to reshape ("normaliseren") an
unstable river to new dimensions, or to 'rehabilitate' ("herstellen") the unstable river to its
old dimensions. When flood diversion is applied, the hydraulic engineer has to design the
dunensions of the new flood channel. Furthermore, many new open-water drainage channels
and irrigation canals have to be designed, as to manage the water in the deltaic area.
Present design. The present design procedure of flood channels, drainage channels and
irrigation canals is still rather empirical. The design can be divided into:
the determination of the 'alignment' ("trac") and the location of structures. This is not
discussed here;
the preparation of the 'design criteria' ("ontwerpvoorschriften"), such the side slopes, the
freeboard, the dunensions of the embankments;
the 'hydraulic design', such as the 'morphological method' and the 'regime method'.
3.1.2. Side Slopes

Soil mechanics. The side slope ("talud helling") lyert ^Hor depends on the principles of
soil mechanics, but it is often selected on basis of regional experiences.
Typically, the side slope is taken as a function of the design discharge. These
assumptions are usually laid down in 'design criteria' ("ontwerpvoorschriften"). For example,
m = 1 for < 0.5 m^/s, m = 1.5 for 0.5 < < 4 m^/s, and m = 2 for g > 4 n?/s.
63
64
3.1.3.
Freeboard
Need for freeboard. The 'freeboard' ("waking") is the vertical distance between the design
water level and the top of the banks or the embankments. The required freeboard is often
established on basis of experience or on 'good engineering practice'. The magnitude of the
freeboard should be specified in 'design criteria'
("ontwerpvoorschriften").
Freeboard is needed for several reasons, but mainly to protect the embankments against
breaching:
the water levels may become higher than the design water levels, because of inaccuracies
in the hydraulic design, waves, operational mistakes, etc.;
the top of the embankment may become lower, because of inaccuracies in the
construction, settlement ("zetting") of the earthfill, settlement of the sub-soil, etc.;
a higher freeboard with the same top width of the embankment creates a longer length for
seepage ("dijkskwel").
Freeboard in irrigation. The freeboard of irrigation canals is often based on the design water
depth y, or on the design discharge Q. Examples are:
the USA standards for irrigation canals refer often to the recommendations of the USBR
(e.g. Chow 1959): F = cVy in m, with values of 0.68 < c < 0.87 for 0.6 m^/s < Q
< 85 m^/s, respectively.
Thus, an irrigation canal of = 0.68 m^/s with a water depth y = 0.70 m requires a
freeboard of F = 0.57 m. And a canal of = 85 in^/s with a water depth y = 2.80 m
requires a freeboard of F = 1.46 m.
Indonesia applies freeboards of 0.40 m to 1.00 m on irrigation canals for design discharges
of 2 = 0.5 m^/s to 15 m^/s, respectively (DHV 1986).
Freeboard in drainage. The freeboard of drainage channels is more difficuh to assess than
the freeboard of irrigation canals. Irrigation canals have design discharges that are also the
'maximum' discharges. Drainage charmels may have a design discharge, but this discharge
is related to a 'return period', e.g. the 20-years flood, and higher discharges may happen
as well.
Strictly-speaking in drainage, discharges higher than the design discharge are allowed
to breach the embankment. However, a somewhat higher freeboard is often applied in
drainage engineering to create an additional protection.
For instance, a drainage chaimel with a freeboard of F = 2.00 m above the water depth
>'5-years = 3.00 m during the design flood of s-years "^^y acmally acconmiodate the 20years' flood.
Example of the freeboard in drainage nad i n flood control are:
Indonesia applies freeboards of 0.40 m to 1.00 m on the main drainage systems for design
discharges of s.years = 0 - 1 "i^/s to 40 m^/s, respectively (DHV 1986).
the Netherlands applies river embankments with a freeboard of 0.50 m plus the 'wave runup' ("golfoploop") above the design flood i250-years -
65
3. D E S I G N O F O P E N - C H A N N E L S
3,1.4.
Embankments
Spoil banks. When excavation ("ontgraving") exceeds filling ("ophoging"), extra earth has
to be disposed. Normally it is very costly to transport this soil by mechanical means to other
locations. A n option is to provide more width on embankments when applied. Otherwise,
'spoil banks' ("stortplaatsen") must be designed, see figure 3.1.
The spoil banks are heaps of soil, and they should be discontinued at suitable intervals
as to allow cross drainage of surface water.
Borrow pits. When filling exceeds earthwork in excavation, the earth has to be brought from
somewhere. The pits, which are dug for bringing earth, are known as 'borrow pits'
("grondhaling").
Those pits can be located outside the channel (external pits), but they can also be
located within the chaimel or in the foreland or flood plain (internal pits). Again, it is very
costly to bring soil from a distance.
External pits are not preferred, as land has to be expropriated ("onteigend") and is
reshaped into ponds that collect rainfall and may cause mosquito nuisance. Internal pits can
be applied in the larger channels or in the flood plains. These internal borrow pits may not
endanger the stability of banks. Moreover, they may not be constructed over the total length
but only in compartments as to avoid flowing water that may cause scouring.
Spoil banks may not hamper the drainage

Figure 3.1. Spoil banks and borrow pits.
Borrow Pits within the channel
66
3.1.5. Construction Drawings

Role of Drawings. Many people are involved in realization of open-channels. Drawings play
an essential role in all stages of the works:
the topographer prepares topographical maps, including contour lines ("hoogtelijnen");
the geologist prepares geological and or soil-technical maps;
the design engineer designs the works, and prepares draft drawings for the longimdinal
profile ("lengteprojel") and for typical cross-sections ("dwarsdoorsneden");
the draughtsmen prepares the constmction drawings for the tender documents ("bestekken");
the contractor carries out the work on basis of these drawings.
Longitudinal Profile. The 'longitudinal profile' ("lengteprojel") of anopen-charmel, together
with a few typical cross sections, show basically all design information, see also figure 3.2.
These (draft) drawings are made by the design staff. They show all technical data, such as:
the design parameters (discharge, roughness, side slope, gradient, water depth, bed width);
elevation of the groundlevel at the axis;
design levels of the embankments, waterline and the bed;
any strucmres.
Cross sections. The longimdinal profiles should be supplemented by a large number of 'cross
sections' ("dwarsdoorsneden"), e.g. at intervals of 50 or 100 m.
Draughtsmen have to prepare these drawings on basis of the information of longitudinal
profiles and the typical cross sections. The cross sectional drawings are used for the 'bill-ofquantity' ("hoeveelhedenstaat", "resultaatsbeschrijvingen") of the tender documents, and for
the implementation in the field.
3.1.6. Tender Documents

Information. A contractor ("aannemer") will be invited to construct the designed openchannel with the related structures. Tender documents ("bestek") are required for the
preparation of the Contract between the Goverimient and the Contractor.
Tender documents. The tender documents ("bestek") will be prepared by the Programming
and Implementation Unit of the Project, using standard bidding documents ("standaard
bepalingen"). The Design Unit has to provide the following information:
Description of the Works ("Omschrijving van het Werk"),
Construction Drawings ("Bouwtekeningen"),
Bill of Quantities ("Staat van Hoeveelheden",
"Inschrijvingsstaat").
Cost estunate. Moreover, an engineering cost estimate ("directiebegroting") has to be made
for this project, i.e. the esthnated costs of the construction contract, and any additional costs,
such as for the detailed design and the construction supervision (toezicht op de uitvoering).
The engineering cost estimate will be based on 'unit prices' ("eenheidsprijzen"), as
collected from different projects in similar regions. An example of the unit rates is presented
in table 3.1.
lOm-l-O
9m+.
O
8m+.
crq
Oh
lm+.
ha-
is:
6nHHt
5m+
cT
O
reference level 2 m +
02
O
O
&
Profile Number
CD
Distance, i n m
CD
T o t a l Distance, i n m
G r o u n d Level, i n m +
250
250
250
250
o
O
T3-
O
O
TT
250
cr
o
O
cvi
E m b a n k m e n t Level, i n m +
W a t e r Level, i n m +
B e d Level, i n m +
250
250
250
cx5
T7T
Q = 200 m3/s
Design Parameters:
m = 2
k = 35 ml/3/s
Q = discharge b = bed width
b
=
35
m
s = 1.0 o/oo
y = water depth s = bed gradient
y
=
2.61m
v = 1.90 m/s
K = roughness m = side slope
?5
o
vo
vd
0
0
CO
cn
Q = 350 m3/s
m = 2
k = 35 ml/3/s
b = 65 m s = 0.5 o/oo
y = 2.78 m v = 1.78 m/s
-J
68
Table 3.1. A n example of 'unit prices'.

Earthwork and Roads:
o Land clearing
O Large scale excavation and disposal, within 50 m
O Dike construction, when soil directly available
O Transport of material, per 1000 m
O Grass planting on slopes
O Village road, unpaved, 3.00 m wide
O Inspection road, with stones, 2.00 m wide
Materials:
O Masonry lining ("gemetselde bekleding") of canals, 0.30 m thick
O Masonry with brick blocks ("metselwerk")
O Stone pitching ("gezette natuursteen") in concrete
O Reinforced concrete ("gewapend beton"), 110 kg steel/m^
O Mass concrete ("stamp beton") and work floors, 280 kg cement/m^
O Vertical gate with hoisting device ("windwerk"), upto +1.00 m wide
O Vertical gate with hoisting device, 1.50 m - 2.00 m wide
$
$
$
$
$
$
$
0.50
2.00
3.00
5.00
1.00
5.00
50.00
/m2
Im?
/m3
/m3
/m2
/ml
/ml
$
$
$
$
$
$
$
100.00
200.00
100.00
200.00
100.00
500.00
900.00
/m^
/m3
/m3
/m3
/m3
/m2
/m2
Description of the Works. The 'Description of the Works' is particular to each Contract and
must be specially written. It should include (i) a general description of the Project ("Het
Project"), (ii) the work to be performed under the Contract ("Het Werk"), (iii) site location
("Locatie") and information as to its accessibility ("bereikbaarheid"), and (iv) other relevant
information.
Construction Drawings. The 'Construction Drawings' will be made by the Design Unit on
basis of the pre-design drawings.
Table 3.2. Bill of Quantities for 'General Items'.

Unit
Quantity
Rate
in US$
Lump
Sum
("stel
post")
Description of the Work Item

Mobilization ("Mobilisatie",
"Voorbereiding"), including but not
limited to preparing, shipping and
erecting construction plant and camp for
the personnel of the Contractor
Providing offices for the Engineer as
specified, including furniture and
operation & maintenance ("Directiekosten ")
Demobilization ("Demobilisatie"),
including but not limited to removing and
shipping construction plant and camp,
and cleaning the sites
Lump
Sum
Lump
Sum
Total of the Bill:
Amount
in US$
3. D E S I G N O F
69
OPEN-CHANNELS
Bill of Ouantities. The 'Bill of Quantities' ("inschrijvingsstaat") itemizes the work in

sufficient detail between different classes of works. The Bill of Quantities is divided here into
the following work items:
Bill of Quantities for 'General Items' ("Wericen van Algemene Aard"), see table 3.2,
Bill of Quantities for 'Earthwork' and 'Roads' ("Grondverzet en Wegen"), see table 3.3,
Bill of Quantities for 'Structures' ("Kunstwerken"), see table 3.4 and table 3.5.
Table 3.3. Bill of Quantities for Earthwork and Roads.

Unit
Clearing and grubbing ("verwijdering
struikgewas") the site of the canals
m^
Excavate soil, depth of 1.00 m to 3.00 m,

and dispose within 50 m
Excavate fill material from approved
borrow pits ("grondput", "cunet"), hml
("grondvervoer") up to 50 m, deposit
("aanbrengen"), shape ("verwerken") and
compact ("verdichten") to fill
("ophoging")
Place, shape and compact of fill material
("grond in ophoging") from excavated
soil
Rate
inUS$
ha
Excavate topsoil ("ontgraving bodemlaag") to maximum depth of 0.25 m and

stockpile ("opslaan") for reuse within 50
m
Excavate soil ("grond ontgraving"), depth
of 0.25 m to 1.00 m, and dispose
("storten", "verwerken") within 50 m
Quantity
Overhaul of material ("grondvervoer")

per 100 m distance
Replace topsoil ("aanbrengen van
bodemlaag") to 0.20 m thickness, and
plant grass ("gras inzaaien") on slopes
("taluds") and crests ("kruin") of
embankments
Supply, place and compact stone-fill
material ("steenmengsel") as road
pavement ("wegverharding")
m2
m2
Total of the Bill:
Amount
in US$
70
Table 3.4. Bill of Quantities for ... cross regulators.

Unit
Quantity

Excavating soil ("grondontgraving") for
the structure
...X
Blinding layer concrete ("beton voor

fundering"). Class B-IV2
m3
... X
Reinforced Concrete Class A-IV2,

including shuttering ("bekisting") and 110
kg steel/m^
m3
... X
Masonry ("metselwerk") of solid concrete

blocks
...X
Supply and fix vertical gates ("vlakke

schuiven") with hoisting device
("windwerk"), upto +1.00 m wide
No.
...X
Supply and fix vertical gates with hoisting

device, 1.50 m - 2.00 m wide
No,
... X
Place and compact backfill ("grond

aanvulling en verdichting") for structure
m^
Rate
in US$
Amount
in US$
...X
Total of the Bill:
Table 3.5. Bill of Quantities for ... inflow structures.

Unit
Quantity
m3
...X

Excavating soil ("grondontgraving") for
the structure
Blinding layer concrete ("beton voor
fundering"), Class B-IV2
Reinforced Concrete Class A-IV2,
including shuttering ("bekisting") and 110
kg steel/m^
Masonry ("metselwerk") of solid concrete
blocks
Place and compact backfill ("grond
aanvulling en verdichting") for structure
m-^
3
... X
m^
... X
m3
...X
m^
Rate
in US$
... X
Total of the Bill:
Amount
in US$
71
3.2. Hydraulic Design

3.2.1. Three Degrees of Freedom
Present literature. The present literature on design of open-channels does not provide a proper
methodology. Generally it overlooks the statement (Ankum 1996): "channels have three
degrees of freedom (bed width, water depth and chaimel gradient), which require three
equations for its solution".
Here, several design methods will be discussed, i.e. the 'Strickler method', the
'permissible velocity method' and the 'tractive force method', and it w i l l be shown that they
provide less than the three required equations. Furthermore, the 'regrnie method' is discussed
which provides even more than the three required equations. The 'morphological method'
(Ankum 1996, Ankum 2000) provides just the three equations to solve the three degrees of
freedom.
Furthermore, it should be acknowledged here that the maximum discharge is not
necessary also the 'bed-forming discharge' which determines the morphological stability of
the channel.
Three unknown parameters. The hydraulic design of open-channels depends on several
assumptions laid down in 'design criteria', such as the Strickler coefficient k in m^''-^/s, the
side slope 1^^^^ : Wj^gj., etc., etc.
For a given design discharge, the hydraulic design of the open-channel should provided
three unknown parameters:
the bed width b,
the water depth y,
the gradient s of the bed.
It should be emphasized that the channel gradient s is not equal to the gradient of the
surrounding land, although some relation may exist. It means that the chaimel gradient s has
to be designed, and that it is not a 'given' parameter as is suggested by several authors
(Jensen 1983, James 1988, Chaudhry 1993).
3.2.2. Discussion on the Design Discharge

Difficuhies. Design of flood channels and drainage channels is more difficult than the design
of irrigation canals.
Irrigation canals have a well-defined design discharge, which is the maximum canal
discharge. Normally, the 'headworks' ("watervang") of the system prevents inflow of bedload
into the canals. Also the transport of washload through the canals can be controlled by
constructing a 'sandtrap'
("zandvang").
Flood and drainage channels have flucmating discharges, and the 'bed-forming
discharge' is normally lower, e.g. i-year ^^^^
(maximum) design discharge, e.g.
020-yearsinflow of bedload and washload cannot be controlled. Moreover, the sediment
transport becomes quite difficult for decreasing channel gradients.
72
Design discharge. Thus, the discharge in a flood and drainage channels is varying and is not
constant like in irrigation canals. The capacity of the flood and drainage channels should be
set at a 'maximum discharge' Q^^y. of a certain return period. However, the hydraulic design
should also focus on the 'dominant discharge' dom' which is the bed-forming discharge. The
term 'design discharge' is confusing, and should be used only for the 'capacity' or for the
'maximum discharge' of the open-channel.
Dominant discharge. Thus, the design methods of open-chaimels make often use of the
concept of a 'dominant discharge' ("bedvormende afvoer").
The concept of 'dominant discharge' originates from the regime design method and was
put forward by Inglis in 1946 (e.g. Garde 1977): "the dominant discharge is that hypothetical
steady discharge which would produce the same result in terms of average chaimel dunensions
as the actoal varying discharge". The dominant discharge dom "^^y
taken at half to twothird of the maximum discharge according to Inglis, or may be set at a certain remrn period
of e.g. 1.4 years (e.g. Garde 1977).
What is the dominant discharge? The following typical values for the dominant discharge are
normally used:
for irrigation canals, the dominant discharge is often equal to the ' f u l l supply discharge',
i.e. the maximum discharge or the design discharge;
for rivers and drainage channels, the dominant discharge may be taken as the discharge
with remrn period ofe.g. r = 0 . 5 t o l . 5 years. It means that scouring may occur during
the higher discharges, and that sedimentation occurs during the lower discharges.
for rivers in the lower delta with frequent overtoppings of the riverbanks, the dominant
discharge may equal the 'bankfull discharge'. The frequency of overtopping may have
remrn periods of T = 0.5 to 1.5 years, for instance (Henderson 1966).
It is obvious that the dominant discharge of a river may change considerably when the
hydrology of the river is changed. For instance, the reduction of peak floods by a reservoir
or by a flood diversion will have an impact on the dominant discharge.
Design criterion. The concept of 'dominant discharge' means that scouring is allowed for the
larger discharges and that sedimentation may occur during the lower discharges. The designer
may translate the concept of dominant discharge into the design criterion: "no scouring nor
sedmientation may occur during the dominant discharge dom "
3.2.3. Discussion on the Design Method

Strickler method. It is sometimes suggested that the design of open-channels can be done by
the 'Strickler formula' only (e.g. Chow 1959).
It is tme that the Strickler formula is a valuable equation for uniform flow. But the
formula describes only the relation between the parameters. In general, the design discharge
is known, while also an assumption can be made on the Strickler coefficient and the side
slope of the channel. As stated above, the (re-)construction of the channel by the contractor
requires three parameters, i.e. the bed width b, the design water depth y and the gradient of
the bed s, while only one equation is available.
The Strickler method is recommended in literature for designing 'non-erodable
channels'. In this method, the designer simply computes the dimensions of the chaimel by a
uniform-flow formula and then decides the final dimensions on basis of 'hydraulic efficiency',
or empirical rule of best section, practability, and economy (e.g. Chow 1959).
73
Thus, the Strickler method provides only one equation and fails to solve the three
degrees of freedom.
Permissable velocitv method. The survey of Fortier and Scobey in 1926 on the permissible
velocity v became the basis for the 'permissible velocity method' (e.g. Chow 1959). It was
thought that such a permissible velocity would prevent scouring of the bed, as well as
deposition of sediment. The permissible velocity method uses only two equations: (i) the
permissible velocity v for that soil type, and (ii) the Strickler formula. The channel gradient
^ has to be assumed.
The limitations of the permissible velocity method is shown by the following example,
forwhich a permissible velocity v = 1 m/s is applied. A flood channel with a capacity of
100 m^/s, side slopes 1 ^ : 2^^^, a Strickler coefficient ^ = 40 m^'^/s and an assumed
gradient ^ = 0.3 x 10"^, leads to a bed width = 50 m and a water depth y = 1.87 m.
Another channel with capacity of 10 n?/s would have a bed width
= 5 m and a water
depth y = 1.31m for an assumed gradient ,s = 0.7 X 10'^. Evaluation of these designs by
the physics of scouring, shows that the larger channel has a tractive force T = 5.5 N/m^ at
the bed and the smaller channel T = 9.0 N/m^ while e.g. a permissible tractive force T^fiUcal
= 6 N/m^. It means the larger channel is stable at this velocity, but that the smaller channel
will scour at the permissible velocity v = 1 m/s.
Thus, the permissible velocity method ignores the physical process of scouring which
is well described by bed tractive force. Furthermore, the method provides only two equations
instead of three. Therefore, the permissible velocity method should not be used anymore,
although it is still recommended in recent literamre (Jensen 1983, James 1988, Chaudhry
1993, ASCE 1995^).
Tractive force method. The 'tractive force method' was developed by Lane in the 1950s, and
focuses on the threshold of bed scouring (e.g. Chow 1959, French 1994). The tractive force
method uses two equations: (i) the tractive force formula, and (ii) the Strickler formula. Also
here, the channel gradient s has to be assumed.
For instance, a flood channel with a capacity of 100 m^/s will be constructed with side
slopes lyert 2hor ^^'^ ^ Strickler coefficient ^ = 40 m^'''^/s. The soil has a permissible
tractive force of ^critical = 6 N/m-^. When a channel gradient ^ = 0.2 X 10'^ is assumed,
a bed width = 25 m and a water depth y = 3.00 m can be calculated. The assumption of
another gradient 5 = 0.4 X 10"^ would lead to other dimensions: a bed width ^7 = 60 m and
a water depth y = 1.50 m.
The tractive force method is not practical for the design as it uses two equations to
solve the three degrees of freedom. Moreover, it ignores the process of sedimentation.
Two methods. The present methods for the design of stable channels can be divided into two
approaches:
the 'morphological method' uses hydraulic theories to define a stable channel, such as the
uniform flow formula for the flow of water, the tractive force formula to prevent scouring,
the sediment transportation formulae for the flow of sediment. The 'morphological method'
is recommended for further use.
the 'regime method' is based on the belief that design rules can be derived from
observations on stable chaimels.
The regime method is mainly a product of the Anglo-Indian school of hydraulic
engineering. It was developed on irrigation and drainage projects throughout the Middle
East, India and Egypt with canals in fine-grained soils, of less than 1 mm particle size and
for capacities up to 400 mr'/s. The 'regime method' is discussed here because of its widely
use, but is not recommended here.
74
3.3. Morphological Design Method

3.3.1.
Methodology
Three unknown. The morphological design method acknowledges that three parameters has
to be determined: (i) the bed width b, (ii) the water depth y, and (iii) the gradient s of the
chaimel. Thus, three equations are required.
Two discharges. Furthermore, the morphological design method acknowledges that there are
two discharges relevant in the design (Ankum 1996, Ankum 2000):
the 'dominant discharge' ("dominante afvoer"), also called the 'bed-forming discharge',
("bedvormende apoer") for the stability of the channel in order to avoid scouring and
sedimentation on an annual basis;
the 'design discharge' ("ontwerpapoer"), also called the 'maximum discharge' or the
'capacity', for the water transport capacity of the channel, in order to avoid overtopping
of the banks.
Equation 1. The morphological design method uses the Strickler formula as its first equation
to describe the flow of water:
Q = kA R^'^ s^'\ and
Q
with the wet cross-sectional area A:
A = (b + ym)y
and the hydraulic radius R:
R
vA
4
b + 2y \f+m^
in m^, R is the hydraulic radius in m, s is the water level (energy) gradient, b is the bed
width in m, y is the water depth, in m, m is the side slope (lyert '^Hor)'
^
Strickler coefficient in m^^^/s.
The values of the Strickler coefficient k and the side slope ("talud helling") m of the
channel should be considered as assumptions in the 'design criteria'
("ontwerpvoorschriften").
Equation 2. The second equation is related to the flow of sediment. It is assumed that there
are two different situations, see figure 3.3:
or, the channel is subject to scouring during the dominant discharge. It means that the
channel has to be checked on the criterion of the critical 'tractive force' ("wandschuifspanning") J^r to prevent scouring. Scouring can be prevented e.g. by reducing the
gradient s;
or, the channel is subject to sedimentation during the dominant discharge. It means that
the channel should be checked on the sedmient transporting capacity QJQ. Sedmientation
can be prevented e.g. by increasing the gradient s.
These two different situations cannot occur at the same tune, as a channel cannot 'scour' the
bed and deposits its sediments at the same time. This would be reflected in the values of the
allowable 'tractive force' T^j. and of the 'sediment transporting capacity'
Therefore,
75
only one equation can be used in tiie design.

Furthermore, it is aclmowledged here that there are several gradients s without scouring
or sedimentation, because the process of scouring (T = p g y s) is described by other
parameters than the process of sedimentation (E = p g v s).
The above assumption for equation 2 is in accordance with findings by Hjulstrm
(1935), see also figure 3.3. However, he related the sediment transport to the velocity v of
the water, which is considered now as out-of-date by recent research.
water velocity
vin cm/s
Diameter D in mm
Figure 3.3. Scouring and deposition criteria, according to Hjulstrm (1935).
Alternative for equation 2. The operational objective of main drainage systems in polders is
to maintain the target water level ("streejpeil"), i.e. the polder water level ("polderpeil").
However, it is only the water level at the outlet that can be maintained at the polder water
level during all discharges, see figure 3.4. At all other locations, the water level during the
design discharge will differ from the target level as a water level gradient s is needed. It
means that the actual canal levels at some distance L from the outlet will deviate from the
polder water level with a difference Ay = L X s, where s is the gradient of the water line.
The design of open-chaimels in polders may not use the flow of sediment as the
condition in equation 2, but use another condition that is related to water level variations Ay
during the design discharge. Thus, a low gradient s is selected often as the equation 2 in
polder areas.
For instance, the velocity v during the design discharge is set in the Netherlands at a
low value v = 0.25 m/s. This makes the gradient s of the water level during the design
discharge quite, in the order of 0.05 - 0.10 m per kilometer (s = 0.05 x lO"'^ to 0.10 X
10"^). It means that the water level deviations from the (stagnant) polder water level Ay are
only small for changing discharges.
76
gradient s during Q^^

polder water level
location
sluice
TERRAINLEVEL
r
- t n
water line for Omax
^
^
This water level is kept

y
at target level by the operator/
Figure 3.4. Water levels in the open-water system of a polder.
Equation 3, The third equation is related to the geometry of the cross-section. This is
expressed by the relation hly, between the bed width b and the water depth y.
This width-to-depth ratio n = b/y is a logic condition, as a design water depth e.g. y
= 0.25 m should not be designed in a chaimel with a bed width of b = 25 m. However,
satisfying values for the width-to-depth ratio n = b/y cannot be found in the literamre.
In general, smaller design capacities require also smaller width-to-depth ratio n = b/y.
For instance, Q ~ 1 m^/s requires /y = 1, and <2 = 10 m^/s requires b/y ~ 5, etc.
3.3.2. Width-to-depth Ratio

Considerations. The following considerations can be used in selecting the proper width-todepth ratio n = b/y, between the bed width b and the water depth y:
minimum wet cross-sectional area will lead to lower earthwork. It will be calculated
below that the minimum wet cross-sectional area will lead to a small values of ti. thus to
narrow channels;
minimum maintenance costs will be obtained, for instance, when the channel can be
cleaned by a machine in a single run. This is better possible for a small values of ti. thus
for narrow channels;
minimum construction costs will not be obtained by the mimmum earthwork only, but
also by the type of earthwork: earthwork on deep channels will involve higher 'unit rates'
("eeti/ieidsprijzen") because of the deeper layers are harder and the deeper layers will
require more vertical lift.
Thus, the lower 'unit rates' will be obtained by a large value of n. thus by wider
channels;
minimum water level variations in the channel are attractive for several reasons: (i) the
stability of the embankments is better, (ii) the maximum water level, and so the
embankments, are lower, (iii) navigation is possible during many discharges, etc.
Minimum water level variations are obtained by a large value of n. thus by wider
channels;
77
maximum ciiannel storage is important when peak floods have to be reduced by 'flood
routing', or when storage have to be created when a (tidal) sluice is closed during a certain
time.
The maximum channel storage is obtained by a large value of n, thus by wider
chaimels;
minimum seepage into the drain ("kwel") is hnportant in polder areas with a 'Dutch
profile'. A drain with a small width-to-depth ratio n = b/y will cut deep through horizontal
impervious layers as is observed in the polders of the Netherlands. Thus, the seepage will
be limited by taking large values of n, thus by wide channels.
Minimum wet cross-sectional area. Sometimes, designers try to design at the 'minimum wet
cross-sectional area ^ ' . It can be proven, see also box 3.1, that the minimum cross-sectional
area for any side slope m is guaranteed when: "the trapezoidal shape is formed by tangents
("raaklijnen") to the semi-circle, having its centre in the water surface", see figure 3.5.
The corresponding width-to-depth ratio n = b/y, between the bed width b and the water
depth y for a minimum wet cross-sectional area reads, see also box 3.1:
b/y
= 2\Jm^ + 1 - 2m
with the bed width b, the water depth y and the side slope lyert '^HorNo practical value. Thus, the minimum wet cross-sectional area is obtained for the following
width-to-depth ratio n = b/y, between the bed width b and the water depth y:
o for a side slope m = I
: the width-to-depth ratio n = b/y = 0.83,
o for a side slope m = 1.5 : the width-to-depth ratio n = b/y = 0.61,
o for a side slope m = 2
: the width-to-depth ratio n = b/y = 0.47.
These values have no practical value for the design of unlined open-channels, as the value
of the width-to-depth ratio n = b/y is too low, and thus the cross section is too narrow, to
meet the other considerations.
BASIC GONIOMETRY
tg a = 1 / m
tg Vi a = (1- cos a) / sin a
h/2 b/2
Figure 3.5. 'Best' hydraulic cross-section.
Cross-section for lined channels. The approach of the 'minimum wet cross-sectional area',
as discussed above, is applicable for the design of lined channels ("beklede kanalen"). So,
it is even possible to calculate the minimum cross-section for a variable side slope m, see also
box 3.1.
The optimum cross-section for a trapezoid is 'one half of a hexagon' ("de helft van een
gelijkbenige zeshoek"), see also figure 3.5: the three sides of the cross section are equal long,
the bed width b is twice the water depth y, and the side slope amounts to lygj.^ : 0.58jjoj. (m
= 0.58 or o: = 60), and the width-to-depth ratio n = b/y = 1.2.
78
Box 3.1. Determination of tlie widtli-to-deptli ratio n = b/y for the minimum cross-section.
Minimum ci'o.ss-section for a known side slope /?
The determination of the width-to-depth ratio n - b/y for the minimum cross-section
is usually transformed into the question: "what width-to-Ueplh ratio n = b/y gives the
maximum discharge O f o i a fixed cross sectional area AT This is solved in the
following way (Chow 1959):
the Strickler formula reads: Q = kA R-'^ s ' = k A-^ P'-''' v''-.
* the hydraulic radius is: R = A/P ^
^
' so that the discharge is: Q = k A^'"^ P'"''^ s^'\
the bed width is: b = A/y - m y, because of the equation: A = {h -f my)y,
the wet pcrimetei P ^ b + 2 y \ [ttr- + 1 j , so P = A/y - m y + 2 > \ \m'' + 1 ].
Considering that the parameters Q and .4 are fixed, the criterium becomes one of
minimising the icim P. Thus: d/Ydy = -A/y- - m + 2 v'{w- -i- i ] 0^
As A = {b + niv)y, this equation becomes -h/\> - m - m + 2 \'[m- + IJ = 0, and
finally % ^ 2 ^
+ I j - 2m.
Are the bed and the slopes the 'tangents' to a semi-circle?
The side slope lY^,,f : m^^^ has an angle a with the horizontal. Thus:
tg a = 1 / m, or u = arctg ni, sin a = \ / yn" + 1 , and sin a = m / v'w ' " 1
The widlh-io-dcpth ratio n - b/y for the minimum cross-section reads:
2{\'my
+ 1 -OT}- 2 -
\ sin
mmmmmmmmmmmmmmn^mm
ft sma
.
Basic goniometry learns that tg
a = (1 - cos a) / sin a, so that: tg ' A a =
b/y.
This condition means that the ptiint M in the middle o f the water line is also the
centre of a circle, which tangcntj arc the bed and the side slopes of the cross-section.
What -side slope m gives the most optimum croj>s-section?

The above width-to-depth ratio b.'y - 2 y [nr + I] - 2m leads lo an expression for the
iil^HlliiiBBiHiiiBiHlHHH^H
The cross-sectiiinal area: A - b\ + my- = 2v" v"(/;z- 4- 1] -
so that:
IBIHIppBll^llllBIIBH^
The wet perimclei P =^ b ^ 2y^'\m- + I] ^ 2y{2^[m.
{2(m'
+ 1 - wf
+ \]-m},
so that:
4/1 (2v':^rTT - / w )
liHl^BBpiiiHiiHii^^
and f i n a l l y : ? - 2\'A {2^'^P~n
- m)
The minimum value is found for d/Vd/// = 0, so:
iliiiiiiiS
and so for a = 60.
2w
for : ///
79
Maximum value for E = o g v s. The geometry of a cross-section (bed width b, water depth
y and side slope lyert ^Hor) ^^^^ ^1^ influence the 'deVos value' of the energy dissipation
E = p gvs.
The determination of the width-to-depth ratio n = b/y for the maximum value of E =
p g V s can also be transformed into the question: "what width-to-depth ratio n = b/y gives
the maximum velocity v for a fixed cross sectional area AV, and follows the calculation of
box 3.1.
Thus, the maximum value of E = p g v s fox a given discharge Q and a given gradient
s, follows also from the above formula of the minimum cross-sectional area:
b/y
= 2y/m^ + 1 - 2m
Again, this formula has no practical value for the design of unlined open-chaimels, as the
value of n = b/y is too low, and thus the cross section is too narrow, to meet the other
considerations.
Width-to-depth ratio b/v. The width-to-depth ratio n = b/y, between the bed width b and the
design water depth y, is often assessed on basis of practical considerations. Considerations
may include wider channels have less water level variation, deep channels may cut through
impervious horizontal layers, deep channels require less expropriation, as well on economic
considerations.
Different relations have been developed for irrigation and drainage channels in different
countries:
in USA, the USBR-formula is used: b/y = 1.65 Q^-^^;
the Indonesian design standards are based on the Kennedy equation, but applied together
with the tractive force concept.
Examples of often-used width-to-depth ratio n=b/y in irrigation are presented in table 3.6.
Width-to-depth ratio b/y in the design. It is obvious that the width-to-depth ratio n = b/y
cannot be defined on strict objective grounds. Therefore, it is advisable to set a range of the
width-to-depth ratio n = b/y in the design criteria, instead of just one value. Some guidance
can be obtained from the USBR-formula: b/y = 1.65 Q^-^^. For instance: Qdom = 30 m^/s
needs a range in the width-to-depth ratio n = b/y of 3 < n < 5.
In this way, the design remains flexible and can start the design with an assumption on
the gradient s, and can finally check this assumption of the width-to-depth ratio n = b/y.
Table 3.6. Values for the width-to-depth ratio n = b/y in irrigation canals.
Design
Discharge
m3/s
0.5
1
2
5
10
20
50
100
"best-"
Hydraulic
b/y
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
USA
Indonesia
Standard
standard
b/y
b/y
1.4
1. 7
2.0
2.6
3.1
3.8
4.9
6.0
1. 2
1. 5
2.0
2.9
3.9
5.8
10 . 0
R e g i me
c a n a 1 *
cohesive
sand
sand+cohes
b/y
b/y
b/y
5 .5
4.2
6.4
7.5
7. 9
8. 9
10 .4
11.8
5.9
6.1
6.9
7. 9
8.4
9.5
11. 7
12 . 8
6.5
6.8
7.6
8.9
9.4
10.6
12 . 5
14 .4
Simons & Albertson equations, with m = 1 for = 0-5 m^/s, with m = 1.5 for 1 m^/s
< < 5 m^/s, with m = 2 for ^ 10 m^/s.
80
3.3.3. Design on Basis of the Morphological
Method
Objectives of the channel. The objectives of the channel should be agreed between the 'client'
(~ "opdrachtgever") and the designer before the hydraulic design starts. These objectives
include:
the 'dominant discharge' ("bedvormende apoer") dom
which the channel will not
scour or deposits its sediment. This is e.g. the twice-per-year flood Q 5-year
the onceper-year flood i-year!
the maximum discharge Q for which the embankments may not overtop. This is e.g. the
once-per-20-years flood 020-years
the once-per-50-years flood so-years!
the maximum value of the tractive force during the dominant discharge, i.e. the 'critical
tractive force' r^jg^ in N/m-^, as to prevent scouring;,
the minimum value of 'sediment transport capacity' E.^,^ in Watt/m^ during the
dominant discharge, as to prevent sedimentation;
the water elevation at the 'tail-end' ("benedenstrooms") of the new channel reach during
the dominant and the maximum discharges.
This elevation has to be tied to the 'reference level' ("referentiepeil"),
such as 'NAP'
in the Netherlands, the 'mean sea level' ("gemiddeld zeeniveau"), the 'project datum'
("projectpeil"),
etc.
Furthermore, information may be required on the water elevation at the 'head-end'
("bovenstrooms"),
the terrain elevations, allowable variations of the water line, other
functions of the channel, etc., etc.
Unknown. There are basically, three unknown parameters in the hydraulic design of openchaimels:
the bed width b,
o the water depth y^^^ during the 'dominant discharge' ("bedvormende apoer"), which
is the basis for the design of a stable chaimel, and
o the water depth y^^îx during the maximum discharge, which is the basis for the design
of the earthwork and the embankments ("dijken").
the gradient s of the bed, which equals the gradient ("verhang") of the energy line and of
the water line.
Furthermore, the elevation of the new channel has to be tied to the reference level ("referentie
peil"), so that the contractor knows at what elevations are required.
Three equations. There are three equations available to solve the above unknown parameters:
equation 1 describes the flow of water by means of the Strickler formula:
Q ^ kA R^'^ s^'\ and
Q =
with the wet cross-sectional area A:
vA
A = (b + ym)y
and the hydraulic radius R:
R
^
b + 2y \Jl+m^
equation 2 is related to the flow of sediment. It is assumed that there are two different
situations, and that only one equation can be used in the design:
o or, the channel is subject to scouring during the dominant discharge. It means that the
channel has to be checked on the criterion of the critical 'tractive force' ("wandschuifspanning") T^^^ to prevent scouring
81
T = pgys
and T < T^^^
o or, the channel is subject to sedimentation during the dominant discharge. It means that
the channel should be checked on the sediment transporting capacity Fîn to prevent
sedimentation:
E
= p^v^,
or: E
= P ^ ^ ^ '
'
^mi"
the alternative equation 2 is applied when flow of sedmient is not relevant. For instance,
water level changes Ay should be limited, which means that the gradient s should be kept
at a low value;
equation 3 is related to the geometry of the cross-section during the dominant discharge,
or the maxhnum discharge. This is expressed by the width-to-depth ratio n = b/y,
between the bed width b and the water depth y;
where: Q is the discharge in m^/s, T is the tractive force in N/m^, E is the energy dissipation
in Watt/m^, v is the velocity in m/s, A is the wet cross-sectional area in m^, R is the
hydraulic radius in m, s is the water level (energy) gradient, b is the bed width in m, y is the
water depth in m, m is the side slope (lyert ' '"Her)' ^ the Strickler coefficient in m^'^/s,
p = 1000 kg/m^ is the density of water, and ^ = 9.8 m/s^ = 9.8 N/kg is the acceleration
of gravity.
Assumptions in 'design criteria'. Several design parameters have to be confirmed with the
'client' (~ "opdrachtgever") in design criteria ("ontwerpvoorschriften"):
the Strickler coefficient Ic in m^'^/s,
the side slope m,
the width-to-depth ratio between bed width and water depth n = b/y,
and other parameters like: freeboard, crest width of embankments, width of berms, side
slopes of the embankments at land-side, minimum radius of a bend, location of borrow pits,
location for spoil, etc., etc.
Actoal design. The hydraulic design may provide different cross-sections that are valid. It is
even not justified to ami at the most 'optimum' design, as so many assumptions have been
made. The design can be done in the following steps, see figure 3.6:
step 1: determine the elevation above the reference level at the tail end of the channel;
step 2: draw from this reference level the straight water line during Qôm "^^^ upstream
direction. The uniform flow during the dominant discharge avoids backwater effects that
may influence locally sedimentation and scouring.
The gradient s has to be estimated sufficient flat, as to prevent erosion and to match
with the head-end water elevations.
The gradient s has to be esthnated sufficient steep, as to prevent sedimentation and to
avoid drop structores at the head-end;
step 3: design the water depth y^^^ and the bed width b of the channel by incorporating
the width-to-depth ratio n = b/y. Cheek the sediment transporting capacity E^^^ and the
critical tractive force T^^^x' ^tid retorn to step 2 i f necessary;
step 4: calculate the water level y^^^ during Q^^x with the Strickler formula. Also here,
a 'straight' line may be taken by ignoring the backwater effects in the tail-reach of the
channel. Check at the head-end, whether:
o the calculated water elevation is not higher than the available water elevation (water
cannot flow into the new channel),
o the calculated water elevation is lower than the available water elevation, because a
drop structore is required.
82
3.3.4. Example of the Morphological Method

Flood diversion channel. A new flood diversion channel has to be designed between a 'flood
diversion structure' and the sea. The channel will have a length of 15 km.
The conditions at the tail-end are determined by a sea-level at 0.00 m"^, while the tide
in the sea can be ignored.
The conditions at the head-end are determined by the flood diversion structure. This
structure should always flow by 'free flow', so that the water elevation during max
become higher than at 14.00 m"^. Lower water levels are allowed.
The dominant discharge is Q^^^ = 1 0 0 m^/s, and the maximum discharge is max ^
250 m^/s.
The maximum value of the tractive force during the dominant discharge, i.e. the
'critical tractive force' T^^.^ = 30 N/m^, as to prevent scouring. The minimum value of
'sediment transport capacity' E^^^ = 10 Watt/m-^ during the dominant discharge, as to
prevent sedimentation.
Assumptions in 'design criteria'. It has been agreed with the 'client' f "opdrachtgever")
that the following parameters are valid in design criteria:
the Strickler coefficient k = 35 m^'^/s,
the side slope m = 2,
the width-to-depth ratio n = b/y between bed width and water depth during the dominant
discharge dom = 100 ni^/s is set at the range 6 < n < 9.
Some related parameters of the Strickler values have been calculated with the 'Profile'
computer program, are presented table 3.7. They can be used in the design of the flood
diversion chaimel.
Table 3.7. Some Strickler values for the flood diversion channel.
Q
m^3/s
y
m
ks=(1/n
m^l/3/s
b
m
m
iv.mh)
s
10 "-3
n
(b/y)
m/s
T
N/m^2
100.00
100.00
100.00
2 . 88 35 . 00 20 . 00
2 . 57 35 . 00 20.00
2 .37 35 . 00 20 . 00
2 . 00
2 . 00
2 . 00
0 . 50
0 . 75
1 . 00
6 . 94
7 . 78
8 .44
1.35
1. 55
1. 71
250.00
250.00
250.00
4 . 77 35 . 00 20 . 00
4 .28 35 . 00 20 . 00
3 . 95 35 . 00 20. 00
2.00
2 . 00
2.00
0 .50
0 . 75
1 . 00
4 .19
4 . 68
5.06
1. 77 23 .41
2 . 05 31.47
2.27
38.79
E
W/m^3
14 .14
6.61
18 . 91 11.39
23 .23
16.74
8 . 70
15.06
22.22
Design process. The water elevation at the tail-end of the channel is known, as it equals the
sea elevation of 0.00 m"^.
A gradient s is assumed for dom- The water depth y^^^ and the bed width b can be
calculated by the Strickler equation (use e.g. program 'profile'), by considering the width-todepth ratio ti = b/y^Q^, while the critical tractive force T^^. and the minimum sedunent
transporting capacity ^ ^ j j ^ are checked. Finally, a check is made on water elevation at the
head-end during m a x '
^^ô figure 3.6.
Thus, the design of the chaimel follows a trial-and-error approach by assuming different
gradients of the water line during dom ^îd max-
83
Choose a gradient of ^ = 1.00 "/QQ = 1.00 X lO'^:

o Calculate the corresponding water depth y^^ = 2.37 m and bed width = 20 m for
n = 8.4, r = 23 N/m2 and E = 11 W/mT
o Calculate the water y^^y^ = 3.95 m.
o The water elevation at channel tail-end for dpm
0.00 m + , so the bed elevation at
the channel tail-end is 0.00 - 2.37 = -2.37 m"
Thus, the bed elevation at the channel head-end is -2.37 + 15x1.00 = 12.63 m +
and the water elevation at channel head-end for âx: 12.63 -I- 3.95 =16.58 m .
Conclusion: not successful, as the water elevation at îax îgêr than the objective of
14.00 m + .
Choose a gradient of ^ = 0.75 "/^^ = 0.75 X lO'^:
o Calculate the corresponding water depth y^^^ = 2.57 m and bed width = 20 m for
n = 7.8, r = 19
and E = 11 W/mT
o Calculate the water y^^^ = 4.28 m.
o The water elevation at channel tail-end for e^nm is 0.00 m + , so the bed elevation at
the channel tail-end is 0.00 - 2.57 = -2.57 m ^ .
Thus, the bed elevation at the channel head-end is -2.57 + 15x0.75 = 8.68 m ,
and the water elevation at channel head-end for Q^^y^: 8.68 + 4.28 =12.96 m +
Conclusion: correct design, as the energy elevation allows for free flow!
20
15
m^p-sr^
10 m"'
5 m*
Cm* -
Figure 3.6. Longitudinal profile of the flood diversion channel.
3.3.5. Discussion on the Morphological
Method
Range for the gradient. The above described morphological method for the design of openchannels acknowledges the three design parameters and provides three design conditions.
Furthermore, it is recognized that 'sedimentation' and 'scouring' of the cross-section are two
different phenomenons that cannot occur at the same time. Thus, gradient s of the channel
should be larger than the minimum gradient s^^^ as to avoid sedimentation, and should be
smaller than the maximum gradient s^^^^ as to avoid scouring.
84
Formula for the minimum gradient. The formula for the minimum chaimel gradient ^ j ^ j ^ can
be derived by using the energy dissipation formula p g v s > fi'min- see box 3.2:
\8/ll
\2/ll
4/11
(n + 2s]
+ 1 )'
k'Q
{n + mf'^^
with the minimum channel gradient s^,^, the minimum value for the energy dissipation E^,^
in Watt/m-^ to avoid sedimentation, the discharge Q in m^/s, the Strickler coefficient k in
m^^^/s, the width-to-depth ratio n = bIy, the bed width b in m, the water depth y in m, the
side slope m (lyert '"Hor)density of water p = 1000 kg/m-' and the gravity acceleration
g = 9.8 m/s2 = 9.8 N/kg.
Pg
Box 3.2. Formula for the minimum channel gradient s^,^ to avoid sedimentation.
The formula for the minimum channel gradient .v^^-jj to avoid sedimentation can be
derived for a bcd-to-depth ratio n = biy from the Strickler formula:
BIIHi^^^HlllHIHIIIIlBiBB
with the wet cross-sectional 'dTO'dA:
A ~- (d ^ my)}' ~ {n - /?;);'and with the hydraulic radius R\
. _
b + 2 r v ' i ^ m-
{"^
{n+ 2\\
-> nr )
Hence:
{n ^ rn)
X s
\n + 2v' 1 +
^BIBillllilllBBP^
in *
{n^2^,;^
mf''
2 V
nr]
X s
so the equation for the water depth}' reads:
{n
^ m
m)5/8
\h
and the wet cross-seettonal area:

Xs
liliiiipliliiH
Sedimentation will be avoided when pgva^
'mill
E^^^^, with the velocity v ^ Q/A. so:
^ A
a
Hence.
min
in
2v L + m
and finally the formula for the minimum channel gradient >s,^ to avoid sedimcniation
liiiiiiiiliiiiiii^^
{/I * 2 ^ +
P S%
1)
2\\
3. D E S I G N O F
85
OPEN-CHANNELS
Formula for maximum gradient. The formula for the maximum gradient s^^^ to avoid
scouring can be derived by using the tractive force formula p gy s < Tjâx. see box 3.3:
16/13
\6/13
(n + m) 10/13
4/13
VJ
( + 2 \ / l + m M
with the maximum gradient ^jâx- the maximum value for the tractive force r ^ a x in N / m to
avoid scouring, the discharge Q in m^/s, the Strickler coefficient k in m^'^/s, the width-todepth ratio n = hly, the bed width b in m, the water depth y in m, the side slope m ( l ^ r t
'"Hor)' the density of water p = 1000 kg/m^ and the gravity acceleration ^ = 9.8 m/s .
Example. The meaning of these limits for the gradient s can be shown in the following
example. A flood channel has a dominant discharge dom =
tn^^^' ^ Strickler coefficient
= 40 m^^^/s, side slopes l^grt :
^ maximum tractive force Tj^^x = 6 N/m^, and a
minhnum energy dissipation ^min = 1 - 2 5 Watt/m^. Thus, the minimum and the maximum
channel gradient s can be calculated for a given width-to-depth ratio n = bIy, see figure 3.7.
Box 3.3. Formula for the maximum channel gradient jax to avoid scouring.
The formula tor the maximum channel gradient A-JJ^^J. to avoid scouring can be derived
tor a bed-to-depth ratio n - bfy by using the tractive force formula pgys<
T^^.^, and
by substituting the above value of}-, so:
X .y <
r
max
(77 +
5.8
r
max
and finally the formula for the maximum channel gradient .Vj^..^ to avoid scouring:
\16/
T
i7T
O.lxIO-^
0.2x10-'
0.3x10"'
0.4x10'
channel gradient s
Figure 3.7. Design example of a stable channel bed.
0.5x10'
86
3.4. Regime Design Method

3.4.1.
Introduction
India. Tlie 'regime method' ("regiem methode") for the design of water courses was first
developed in India for irrigation canals. The reghne method is based on the belief that design
rules can be derived from observations on stable channels. The term 'regime' suggests an
equilibrium between scouring and sedimentation on a yearly basis.
Many authors have developed their own reghne formulae. Well-known is Lacey, who
provided in 1937 a set of three equations (e.g. Raudkivi 1976). Simons & Albertson presented
in 1963 a set of six equations, which were reduced by Graf (1971) to three equations.
The original regime method was related to the more or less constant discharges as
happens in irrigation canals. The concept of the 'dominant discharge' has made it possible
to apply the regime method also on the design of drainage channels with varying discharges.
3.4.2. Kennedy equations

Keimedv Equation. The regime method started with the observations of Kennedy on the
Upper Bari Doab irrigation canal, India, in 1895. It was considered that this canal was
'stable', i.e. no maintenance was needed.
For regime conditions, Kennedy concluded that the velocity v in m/s, and the water
depth y in m, were related by v = 0.55 y^'^^.
Investigations in India and elsewhere have shown that the constants in the Kennedy
equation vary for various canal systems, see table 3.8.
Table 3.8. Variations of C and n in the Keimedy equation v =
Cy^.
CountryI n d i a (Upper B a r i Doab, Kennedy)

Indonesia: f l a t
(Haringhuizen)
Indonesia: steep (Haringhuizen)
I n d i a (Pennar r i v e r )
I n d i a (Chenab)
India (Krishna)
Egypt
Thailand
A r g e n t i n a ( R i o Negro)
Burma (Shwebo)
USA ( I m p e r i a l v a l l e y )
C
C
C
C
C
C
c
c
c
c
c
- 0
0
0
0
= 0
0
= 0
= 0
= 0
= 0
= 0
=
=
55
49
55
60
57
53
25 t o 0 35
35
52
55
62 t o 1 19
n
n
n
n
n
n
n
n
n
n
n
0
0
0
0
0
= 0
0
0
= 0
0
= 0
=
=
64
64
64
64
57
52
64 t o 0 73
66
44
57
61 t o 0 64
Haringhuizen equations. Haringhuizen developed regime equations for irrigation eanals on

Java, Indonesia, in the 1920's. Haringhuizen distinguished between 'flat' and 'mountainous'
areas. For 'mountainous' areas, he has re-written the Kennedy equation into: y = 2.54 v^-^^
and added a second equation between the velocity v and the discharge Q.
87
3. D E S I G N OF O P E N - C H A N N E L S
The Strickler equation is the third equation needed for the determination of the three
design parameters y, b md s.
The terms 'flat' and 'mountainous' can be better translated in 'flat alignment' and 'steep
alignment', as to indicate the lower and upper range of equilibrium. Thus, the Haringhuizen
equations read:
flat alignment
v = 0.42 ^'^^^, and
y = 3.00 v -J^,
steep alignment :
v = 0.46 '^^^, and
y = 2.54 v^-^^,
while
:
A = Q / v = (b + my) y.
in m^, s is the energy (channel) gradient, b is the bed width in m, y is the water depth in m,
m is the side slope (lyert ' '"Hor)-
3.4.3. Lacey Equations

Lacey equations. In the 1930s, Lacey performed a systematic analysis of the available stable
channel data in an attempt to improve the Kennedy equation. He established three equations
for regime channels which are presented in literamre in different forms. For design purposes
it is advantageous to write them as (Raudkivi 1976):
P = 4.83 Q 1/2
wetted perimeter
A = 2.28/^^3 gS/
wet cross sectional area
s = 0.315" x / ' ^ Q-1/6 X 10-3.
chaimel gradient
where Q is the dominant discharge in m^/s, and is the Lacey silt factor. Note that the
numerical coefficients are not dimensionless.
Shnplified Lacey equations. The above Lacey equations are not very practical for the design
engineer. They can be shnplified, considering the wet cross sectional area A = (b + my)y,
and the wetted perimeter P = b + 2yV[l
-h m^], see box 3.4.
After the above processing, the simplified Lacey equations can be written for different
side slopes lygj-t : ^Hor
equation 1:
for m = 1
:
y = 1.32 Q^'^ - V[1J4 Q-1.25 f^'^ Q^'^}
for m=1.5
:
y = 1.15 Q^''^ -V[1.32 Q - 1.08
g5/6j
for m=2
y = 0.98 Q^'^ -V[0.96 Q - 0.92
Q^^^]
equation 2:
b = 4.83 Q^'^ - 2yV[l+m^],
equation 3:
s = 0.315 X
Q-^'^ X lO'^.
where Q is the design discharge in m^/s, is the Lacey silt factor in nmi^^-^, s is the (energy)
gradient of the channel, b is the bed width in m, y is the water depth in m, m is the side
slope (lygrt : WHor)Box 3.4. Further processing of the Lacey equations.
The Lacey equations can be simplified, considering the wet cross sectional area
A ~ (b + my)y, and the wetted perimeter P ~ b + 2 v
+ nr].
Thus,my~ +hy-A
= 0, inwhichb --^ P-2y v'f 1 4- m'], and aftei substitution
of b: (2 \'"fi + nr\ - m)y''- -Py+A
= 0 . The value of > can be solved:
V = {/* -r v[/'" 4 A {2s [\ I

V
- m)} / {AV[\+m^ 2w}.
After substitution of P and A of Lacey's equations:

= {4.83 (?^'- + \ [4.83- (2 - 4 x 2.28 ^'-^ (f^^' (?.v-( I
{4v[1
88
Silt factor. The 'silt factor' is based on the assumption that stable canal sections may exist.
Lacey suggested that = 1.59 VD^, where
is the sediment size in mm, see also table
3.9.
In practice, i t is difficult to assess the value of D^. Measurements in existing canals
may lead to a wide range of silt factors, as sediment deposits through the bed may vary
considerably in sizes. Moreover, the sediment size of new chaimels is difficult to assess.
Therefore, the silt factor is often directly calculated from the channel characteristics
through = 2 . 5
/ R, where v is the velocity in m/s and R is the hydraulic radius in m.
Validity. Criticism of the Lacey's method centres on the limited range of validity:
the Lacey equations have been derived from regional data and for velocities between 0.3
m/s and 1.2 m/s;
the Lacey equations are based on sediment observations in Punjab, India, for canals with
suspended-load and where the concentration ranges between 1000 - 2000 ppm;
the size of sediment has an effect on the wetted perimeter ("natte omtrek"), i.e. finer
sedmients give more narrow channels with steeper side slopes;
basically, the Lacey equations determine the gradient s of the channel for a known
discharge Q and a known silt factor . In river work, all three parameters Q, f and s are
known, and they may not satisfy the calculated channel gradient.
the Lacey equations provide for three equations to calculate three unknown parameters: the
gradient s, the bed width b and the water depth y for an assumed side slope m. The
Strickler formula is not yet used, so that also the Strickler coefficient k can be calculated.
This is not logic, as the Strickler coefficient k is related e.g. to the maintenance of the
channel.
Table 3.9. Lacey silt factor.
channel
material
average
large stones
large boulders
small boulders
gravel
medium s a n d
s t a n d a r d Kennedy s i l t ( I n d i a )
lower M i s s i s s i p p i s i l t
3.4.4. Simons & Albertson
diameter
{Dm <= 600 mm)

{Dm <= 150 mm)
{Dm = 40 mm)
{Dm <= 10 mm)
{Dm =
1 5 mm)
{Dm =
0 40 mm)
{Dm =
0 06 mm)
silt
factor f
40
20
10
5
2
1 0
0 4
Equations
Types of channels. A n important step toward generalization of the reghne method is the smdy
of Shnons and Albertson (Henderson 1966). They made a collection of field data from Indian
and North American sources and distinguished five types of channels:
coarse non-cohesive bed and coarse non-cohesive banks;
sand bed and sand banks;
sand bed and cohesive banks, with low sediment load;
sand bed and cohesive banks, with heavy sediment load (2000 - 8000 ppm);
cohesive bed and cohesive banks.
89
Basic equations. A separate equation could be fitted to each of these types. So, Simons and
Albertson prepared a set of equations:
b' = 0.9 P, where P = K^VQ,
b'
= 0.92
W-0.61
R = K2 e-^^
y = 1.21/?, f o r i ? < 2 . 1 3 m,
or:
y = 0.93 A + 0 . 6 1 , for i? > 2 . 1 3 m
V = ^ 3 (10.76
= K^gysivb
/ v)^-^'^.
where: Q is the design discharge in m^/s, v is the velocity in m/s, R is the hydraulic radius
in m, P is the wetted perimeter in m, s is the (energy) gradient of the channel, b' is the
'average' bed width in m, y is the water depth in m, v is the kinematic viscosity in m-^/s.
It is obvious that this set of formulae is not very practical. Moreover, there are 6
equations against 3 unknown variables, i.e. b, y, s. Graf ( 1 9 7 1 ) skipped three of the above
equations and has re-written the others. Moreover, the equations can be processed into a more
practical shape, see box 3.5.
Box 3.5. Further processing of the Simons and Albertson equations.

Henderson simpHficd tiie Simons & Albertson equations by skipping three of them,
and re-writing the others into (Graf 1971):
equation 1:
v = 1,21 A^ C^""!^
^'^^
^ 2.13 m
or
y s= 0.93
Q^^'^^ + 0.61
for R > 2.13 m
equation 2:
A ly ^ K).9
' equation 3:
v- - A'4 g y s (v 0.9 A j
/ u)""^*^.
Even this set of formulae is not veiy practical ibr the design engineer:
equation 2 can be further processed as A Iv = 0.9
\Q can be rc-written through
Aly^{h^my)mio:b^03K,s'Q~my;
equation 3 can be converted into: 5 s= v- (A', ey)'* (v 0.9 A, \-Q )""-^' i f ^ " " =
r ^ 2 - 0 . 3 7 - ( U 0 ^ - 2 + ( , . 3 7 j ^ . - 1 ^-1 ^ - 1 ().<?-'-^Â^-^? xP'^ =
' - 1 V I -l n Q-0,37 LT -0.37 0-37 _
QiAM2
^0.62 ^ ^ - l
-1' -r^,^M.63 3
,-2.63 ^ 9-2
9-U37lj.b3 ^
-0.37-1.63 ^0..37
-2 ,^).37 ^ '
{03*- K{~ A V ' g'^ t)^'"") Q^^-^-y--^''^ ^ A5 (>^' 2 ^.'2/0
Shnplified Simons & Albertson equations. Thus, the three simplified Simons & Albertson
equations depend on the design discharge Q in m^/s, the temperature r in C, and three
coefficients Ky, K2 and A5 see also table 3.10:
equation I :
y = 1.21 K2 Q^'^^
for A < 2 . 1 3 m
or
y = 0 . 9 3 K2
+ 0.61
for A > 2 . 1 3 m
equation 2 :
b = 0.9 K^VQ-m
y, and check R = (b+my)y I {b + 2yV(l +m^)}
equation 3:
s = A'5
^'^ y'^'^^.
where: Q is the design discharge in m?ls, R is the hydraulic radius in m, s is the (energy)
gradient of the channel, b is the bed width in m , y is the water depth m m, m is the side
slope (lyert * '^Hor)' ^
^he kinematic viscosity in m^/s for f C.
QP'^^
90
Kinematic viscosity. Tlie kinematic viscosity u depends on the temperamre:

i; = 40 X 10"^ / (20 + t) m^/s for t C. The effect of the temperamre on the kinematic
viscosity is normally small. Thus, v = 1.15 x 10'^ m^/s for ? = 15 C is assumed in the
following.
Table 3.10. Coefficients and exponents for Simons & Albertson equations.
GRAVEL BANKS
& GRAVEL BED
SAND BANKS
& SAND BED
0.29
0 . 33
0.33
0.29
n.a.
Kl
K2
K3
K4
3 .17
0 . 25
5.46
n.a.
6.34
0.57
4.24
0.33
4.71
0.48
4.88
0.54
3.08
0.37
4.88
n.a.
3.98
0 . 41
n.a.
0. 87
KB
n.a.
60
X 10-S
COHESIVE BANKS
COHESIVE BANK
& SAND BED
&C0HESIVE BED
+low sedim.
+heavy s e d i m . +heavy sedim.
67
X 10"^
n.a.
58
X 10"S
3.4.5. Discussion on the Regime Method

Many attempts. Many attempts have been made to develop the regime method to the channel
design. Many authors have developed their own regune formulae, such as Inglis in 1946,
Lane in 1953, Henderson i n 1961, Simons & Albertson in 1963, Blench in 1966, etc., etc.
General validity. It is still questionable whether the regime equations are of a general validity
("geldigheid"). Obvious, there are the following disadvantages:
there are large discrepancies between the results from various equations when applied
under similar conditions, see box 3.6;
the channels designed with the reghne method have flatter gradients than the tractive
force theory would allow. This w i l l lead to the construction of several drop structures in
the channel, that could have been avoided by using the tractive force method;
the regime method provides normally for more equations than required for solving the
three design parameters, i.e. the bed width b, the water depth y and the gradient s. It
means that the Strickler coefficient k can be calculated from the Strickler formula. This
is contradictious to common understanding as the Strickler coefficient depends on physical
parameters, such as soil type and maintenance.
The limitations of the regime method can be shown in the examples in box 3.6 and box 3.7.
Conclusion. There seems to be no agreement in the various regime equations. It seems
obvious that the reghne method can only be used for very special conditions for which the
equations and its coefficients have been developed.
91
Box 3.6. Example of regime method for the design of a drainage channel.
A drainage channel with a design discharge of H) nyîs and for side slopes ol m - 2
(I y^.j^: 2^^,,,.) have to he designed with the regime method. There are different regime
equations, each requiring additional information:
for the Hiiringhui/en equations, a Strickler coefficient k = 40 m''-/s is added.
Moreover, there are the options of (i) a flat alignment, and (ii) a steep alignment;
loi the Lacej equations, there ate the options ot (i) medium sand with an average
diameter ot Z^,,^
1.5 mm. (ii) silt of the Punjab, India, with diameter D^^ ~ 0.40
mm. and (iii) line sill with diameter
0.06 mm;
for the Simons & Albertson equations, there arc the options for (i) sand bed and
sand banks, (ii) sand bed and cohesive banks, with low sediment Icad, and (iii)
cohesive bed and cohesive banks.
The results are presented below:
U 3S 0 X H E C o w e E P T
Bed VJater Grawidth (3ptb dient
s
b
y
m m i n m 10-3
acc.
StrlcklejT f o m u l a
CoexVelo- T r a c t i v e
Force
fioient city
T
k
V
N/m2
itt^/^/s
m/s
7 . 50
S.90
1.50
3 .45
0.23
0.29
40
40
0.63
0,70
3.4
4.2
11.02
9.58
6 .08
0.95
1.27
2.06
0 .67
0.21
0.05
36
4S
55
0.81
0.65
0,48
6 .24
2,6:2
l.Ol
sand banks & sand bed

14.88
cohesive banks & sand bed -0.74
cohesive banks and bed
9.05
1.58
1.33
1.14
0.07
0.13
0,16
3S
4$
65
0,35
0,56
0,77
i.OB
f l a t alignment
steep alignment
liACEY -FORMtJliAE
medium sand (Dro-l.5 mm)
Indian s i l t (Dzii=0.40 mm)
fine s i l t
{Dm=>0.06 mm)
1,79
Box 3.7. Example of regime method for the design of a flood chaimel.
A flood channel with a^capacity of 100 m^/y side slopes 1^^.^ : 2,^^. and a Strickler
coefficient k = 40 m^'~''/s has to be designed. The cohesive bed material allows a
maximum tractive force of 7^^.,|,i^.j^, = 6 N/m^,
The design according to the Lacey equations leads to a bed width b ~ 32 m, a
water depth v - 3,63 m and a gradient s ~ .3 x 10"^. The resulting bed tractive
force r =f 1.1 N / m - is much lower than the maximum tractive force. The Strickler
coefficient A- - 59 m^'^h can be calculated from the Strickler formula, and does not
match with the required Strickler coefficient.
The ilesign according to the Simon<; & Albertson equations leads to a bed width
b = 30 m, a water depth3^ - 2,61 m and a gradients ~ 0.08 x
Now, the bed
tractive force T = 2.1 N/m- and the Strickler coefficient k = 71 m^'^/s, which aie both
liliHiBiSIHHiiiliH
92
4. SUPER-CRITICAL FLOW IN OPEN-CHANNELS

4.1. Hydraulics of Super-Critical Uniform Flow
4.1.1. Critical gradient of Open-Channels
General. Normally, hydraulic engineers are designing unlined open-channels for sub-critical
flow only and not for super-critical flow. Super-critical flow would lead to excessive scouring
and would require protection of bed and slopes by lining. The hydraulic engineers apply flow
formulae such as the Strickler formula, and design criteria such as maxhnum permissable
tractive force T in N/m^ to avoid scouring.
Only in specific cases, such as in chutes, the engineer will allow a super-critical flow
in open-channels. Chutes are open-channels with super-critical flow, see figure 4 . 1 . Usually,
they are designed in concrete with vertical side slopes.
energy level
Sloping Glacis
LONGrrUDINAL SECTION
CROSS-SECTION
Figure 4.1. A chute with super-critical flow.
Critical flow. Critical flow in a channel is reached when the channel gradient equals the
'critical gradient's^, see figure 4.2. Steeper gradients lead to super-critical flow, and flatter
gradients to sub-critical flow. Open-channels at just the critical gradient should be avoided
as such a flow is characterized by flow instabilities such as waves.
93
94
jnergyjevel
7>7c
Channel Gradient s smaller than Critical Gradient Sr : Sub-Critical Flow
Channel Gradient s is equal to Critical Gradient Sc : Critical Flow
Channel Gradient s larger than Critical Gradient 5^ : Super-Critical Flow

Figure 4.2. Type of flow in an open-water channel, depending on channel gradient.
Although the hydraulic literature extensively discusses the critical flow in open channels
(e.g. Henderson 1966, Chanson 1999), no practical feeling is given to design engineers about
the critical gradient
in a typical channel.
Critical depth. The critical water depth y^, normal to the bed in open-channels with trapezoidal
cross-sections can be derived from the specific head equation. The specific energy head H of
channels under an angle 6 in radials is defined by the 'specific energy equation for sloping
glacis:
H
= y cos e +
with the water depth y normal to the bed in m, the discharge Q in m-^/s, the wet crosssectional area A in m^, the velocity distribution coefficient o; = 1, the channel angle 6 in
radials, and the gravity acceleration ^ = 9.8 m/s^.
The criterium that the maximum discharge Q for a certain energy head H occurs during
critical flow can be converted into the inverted statement: the minimum value of the energy
head
is reached during critical flow. This leads finally to the equation (e.g. Montes 1998,
Ankum 2002):
b
y
95
4. SuPER-CRmcAL F L O W I N O P E N - C H A N N E L S
with the critical water depth

in m, the bed width b in m, the side slope m (lêrt % o r ) '
the discharge Q in m^/s, the channel angle 6 in radials, and the gravity acceleration ^ = 9.8
m/s2.
The value of the critical water depth y^ normal to the bed can be calculated by iteration:
enter
= 0 in the second part of the equation and calculated a new y^, which can be entered
as the new input.
Critical gradient. The critical gradient
of an open-channels can be calculated for a given
discharge Q when the dimensions of the cross-section are known. There are two equations
available: (i) the above formula for the calculation of the critical depth y^ for a certain
channel angle 6 in radials, and (ii) the Strickler formula. Thus, the equation for the critical
gradient
can be found (Ankum 2002):
4/3
b + 2y^ sj I +
(b + my^) y^
10/3
with the discharge Q in m^/s, the Strickler coefficient k in m^'^/s, the critical water depth y^
in m, the bed width b in m, the side slope m (1^^^ : môr), and the gravity acceleration g =
9.8 m/s^.
Tvpical values. The two above equations for the critical depth y^ normal to the bed and for
has been used for channels with side slopes lyg^j: 2^^^^ and a Strickler
coefficient k = 40 m^'^/s. The critical gradient
for these channels have been plotted in
figure 4.3.
It can be seen that critical flow for design discharges 1 < < 10 ni-'/s occurs at
channel gradients s ~ 10 X 10'^. Generally, the high tractive forces T = p g y s will
prohibite such a design because of the excessive tractive forces.
100
Legenda
Beei width b = 100 m
Siip(sr--cri t cal flov
:
*
Dea wiain d =
ou
Bed width b =
Bed width b =
Bed width b =
10 m
5m
1m
.,
<
>
a
Sub-critica flow
0,1
10
100
1000
discharge Q in nn3/s
Figure 4.3.
The critical gradient s^ for channels with side slopes l^gj.^ : 2^^^^. and a
Strickler coefficient k = 40 m^/^/s (Ankum 2002).
96
4.1.2. Hydraulic Parameters of Aerated Flow

Aerated flow. Chutes are used to destroy the excessive energy of the water in canals on steep
slopes, see figure 4.4. They are designed at chute gradients s = iand that are steeper than
in order to force super-critical flow.
The high speed flow in chutes is characterized by super-critical flow and by aeration
which increases the wet cross-sectional area. The typical appearance of 'white water' is an
air-water mixture by air entrainment after a certain transition length.
energy level
Figure 4.4. Development of the aerated flow i n a chute.

Transition flow. After passing over the weir crest, the water is accelerated by gravity over
the chute. First, the flow does not yet entrain air and the flow remains 'black water', see also
figure 4.4. In this transition length, a mrbulent boundary layer is generated by bottom friction
and develops in the flow direction. When this boundary layer reaches the free water surface,
the flow becomes aerated and the 'white water' flow begins.
Parameters of the air-water mixture. The water is mixed with air by which its volume is
increased, see figure 4.5. The amount of air is usually defined in terms of the air
concentration ratio u, which is the ratio between the air volume V^j^ to the volume of the airwater mixmre Immixture
(^-S- Chow 1959):
V
nuxture
The depth y of the air-water mixmre normal to bed depends on the air concentration ratio u.
The 'pure' water depth yâêr
calculated from:
)'water = ( ! - ) > '
The air-water mixture has an apparent velocity v which follows from:
yb
The (apparent) velocity v of the air-water mixmre is smaller than the real water velocity
^water'
V
water
1!
/
y water
hence:
V
water
97
4. S U P E R - C R I T I C A L F L O W I N O P E N - C H A N N E L S
with tlie real water velocity v^^^^J. in m/s, the (apparent) velocity v of the air-water mixture
in m/s, the depth y of the air-water mixture in m, the pure water depth y^^ter
air concentration ratio u.
Hager (1991) developed the following empirical expression for the air concentration u
on basis of the work done by Straub and Anderson in 1958:
u = 0.75 (sm0)-'^
with the air concentration ratio u and the angle 6 of the chute.
AIR - WATER MIXTURE
PURE W A T E R
Figure 4.5. Air-water mixture of the aerated flow in a chute.

Roughness of chutes. Initial research by Vreedenburgh in 1924 found that the roughness of
chutes is smoother than of channels with the same bed and slope material (Vreedenburgh
1924). He studied the flow of 'pure' water in chutes, and he found that the Strickler
coefficient k was increased to 146% and 157% for channel angles 6 = 14 (s = 250 X 10"^)
and 6 = 37 (s = 750 x 10'^), respectively. A similar effect of steeper angles 6 is confirmed
by recent research work (e.g. Wood in 1991).
However, the present literature does not give a clear insight on the ratio between the
Strickler coefficient ^âter during super-critical flow and the Strickler coefficient k of the subcritical flow (Ankum 2002).
4.1.3. Uniform Super-Critical Flow in Chutes

Vreedenburgh formula. Vreedenburgh (1924) recognized that the air-water mixture i n a chute
has a greater water depth y than the water depth yâter
'P^^^' water in the chute. He
proposed that the uniform flow in a chute would follow the Strickler formula, but with a
different (Strickler) coefficient than for sub-critical flow. Although the steeper gradient s
would increase the Strickler coefficient for 'pure' water, he introduced a smaller Strickler
coefficient k j for the air-water mixture. This reduced coefficient k j would lead to the larger
water depth in the Strickler formula because of the larger volume of the air-water mixhire.
Vreedenburgh measured the water depth y of the aerated flow in existing chutes for a
known discharge Q and a known chute gradient s. He calculated the (apparent) Strickler
coefficient k-^ and compared it with the Strickler coefficient k for similar lining in channels
with low gradients. Vreedenburgh (1924) found the relation for 0 < .y < 750 X 10'^ (0 <
e < 37):
k j = k (1 - sind) with the angle 6 = atan s
Thus, the water depth y of aerated flow can be calculated by the Vreedenburgh formula:
Q
k (1 - sinQ) A R^'^ sin^'^Q
98
with: A = b X y
R = ^ ^ y
. q ^ ^^^^ ^
with the discharge Q in m^/s, the (sub-critical) Strickler coefficient k in m^'^/s, the wet crosssectional area A of the air-water mixture in m^, the hydraulic radius R of the air-water
mixmre in m, the chute width b in m, the aerated water depth y in m, the chute angle 6, and
the chute gradient s.
Present literature. The present literamre does not provide a simple formula for a super-critical
uniform flow, like the well-known formulae for the sub-critical uniform flow. Instead, it
recormnends a somewhat cumbersome approach with a set of formulae.
These methods (e.g. Hager 1991) determine the water depth y of the air-water mixture
from the pure water depth ywater
considering the air concentration ratio :
y
with: M = 0.75 (sm0)0-^5

(1 - u)
The pure water depth yâter follows from the Strickler formula with the Strickler coefficient
^water
P^^^ water during super-critical flow:
With:
^,,er
y,,,er.
^water
T ^ T V ^
^ 6 = ataU .
with the discharge Q in rn^/s, the Strickler coefficient k^^^^j. for pure water in m^^^/s, the wet
cross-sectional area ^âter f ^ê pure water in m^, the hydraulic radius iâter f P^^^ water
in m, the chute width b in m, the pure water depth y^^^^j. in m, the chute angle 6, and the
chute gradient s. The Strickler coefficient ^âter
Pu^'c water is not further specified in
literamre, but is not equal to the Strickler coefficient k for sub-critical flow in m^^^/s.
Discussion. The approach by Vreedenburgh and the approach in the present literamre agree
quite well for chute gradients s < 700 X 10'^ (Ankum 2002).
It is obvious that the approach by Vreedenburgh is more direct than the approaches
followed by the present literamre. Thus, it is recommended here to use the Vreedenburgh
formula (1924).
4.2. Design of Chutes

4.2.1. Water Depth in Chutes
Calculation of the water depth. It is recommended here to use the Vreedenburgh formula
(1924), which is basically the modified Strickler formula, for the aerated uniform flow in
chutes:
Q
= k (I - smQ) A R^l^ sin^'^Q
with: A = b X y,
R ^
b x y
^ ^
^
b +2 y
with the discharge Q in m^/s, the (sub-critical) Strickler coefficient k in m^'^ls, the wet crosssectional area A of the air-water mixmre in m^, the hydraulic radius R of the air-water
mixture in m, the chute width b in m, the aerated water depth y in m, the chute angle d, and
the chute gradient s.
99
The calculation of the water depth y of the air-water mixture in a chute follows from
the above formulae:
Q = k(l-smQ)
sin^/20
which can be re-written into:
.5/3 - Q
1
^
b'l^ k(l-smQ)
(b^2yf'
sm'l^Q
Hence, the water depth y of the air-water mixture follows from:
Q "'^ (b + 2yf^
^
& (1 - sm0)'^ sm"'30
This equation is solved by iteration: assume y = 0, calculate the new y^gw ^^'^ enter this y
again the equation, until y^gw = y- A n example calculation of the water depth of the air-water
mixmre of a chute is presented in box 4 . 1 .
Box 4 . 1 . Calculation of the water depth of the air-water mixtore in a chute.
A drainage channel with a design discharge of 5 n?js follows a steep alignment. A
concrete chute is used lt> dissipate the energy. This chute is 300 m long and the energ>^
head drops bv z - 15.00 m over this distance. Thus, the gradieiu of the chute is s 50 X Ur-\ and 0 ^ atan 0.05 = 2.86^ ( = O.OS radials).
The concrete has a (sub-critical) Strickler coefficient k - 70 m ^ % . The crosssection of the chute is designed at a width of /) = 2.00 m and with vertical walls.
The water depth y of the air-water mixture followi* from the equation:
<?"-^ ( b ^ 2 r r
.
5^Û2^2xyr
, o.343(l M - ) " ^
r'^/j(l~Mn)"Sin''-^
70"^ 2 (1-0.05)"'*0.05'"
Tbe equation is solved by iteration: assume y ~ 0, calculate the new value y^^j^^ =
0.343 m, and enter this y again into the equation and calculate the next y^j.^,. Finally,
a water depth of the air-water mi.xture of a chute is calculated a t j - 0.392 m.
When the gradient of the chute is increased to.? = 500 X 10"^ (B = 26.6% a water
depth y = 0.271 m can be calculated.
4.2.2.
Energy Head in Chutes
Real velocitv of the water. The (real) velocity of the pure water v^^^gj. is higher than the
(apparent) velocity v of the air-water mixmre, see figure 4.6. The real water velocity v^g^gj.
is based on the the air concentration u, which has been experimental determined by Hager
(1991):
V ,
with: u = 0.75sin'^ê
and:
1 - u
and the (apparent) water velocity v of the air-water mixtore:
0 = atan 5
b Xy
where v^^jgj. is the (real) velocity of the pure water in m/s, v is the velocity of the air-water
mixtore in m/s, the air concentration ratio M, is the discharge in m^/s, b is the width of the
rectangular chute in m, y is the water depth of the air-water mixmre in m, the chute angle
6, and the chute gradient s.
100
LONGITUDINAL PROHLE
air - water mixture
J'VATER
CROSS SECTION
CROSS S E C T I O N
A I R - W A T E R MIXTURE
PURE WATER
Figure 4.6. The energy head H of the (pure) water flow in a chute.
Box 4.2. Calculation of the energy head / / in a chute.

The real water velocity \ ' ^ ^ ^ of pure water in the chute of the above examj^lc with a
gradient of.? = 50 x 10"' ( = 2.86) depend.*; on the (apparent) water velocit>'t-:
V
= 6.38 m/s
by
2 X 0.392
and the air contration ratio H:
ti
0,75 sin"''6> = 0.75 sin" " 2 . 8 6
0.079 = 7.9%
and can be calculated by:
"'^lIBIillllillP
= 6.93 m/s
V
water
\ - u
1 - 0.079
And the encrg>' head i f can be calculated from the following equation:
H
v{\
- u) cose ^ '^""^''^ - 0.361 ^ 2.45 = 2.81 m.

2^
When the gradient o f the chute is increased to s
500 X 10"-* {0
26.6), a real
water velocity v^^..^^^.^ - J5.7 m/s can be calculated, and an energy head H ~ 12.7 m.
101
Energy head. The energy head H in the chute is not related to the apparent velocity v of the
air-water mixture but on the real water velocity v^gter of Pure water. Thus, the energy head
H above the bed of the chute is defined by the 'specific head equation', see also figure 4.6:
H
= V
cos e +
/water
Iv
^
?
with:
and the pure water depth yâterV

= ( 1 - w) y
with:
-'water
9 = atan .y
2g
= 0.75sm'^ê
and:
9 = atan 5
with the energy head H above the bed of the chute in m, the pure water depth yâter
the (real) velocity vâtgj. of the pure water in m/s, the air concentration ratio u, the water
depth y of the air-water mixmre in m, the chute angle 6, the chute gradient s, and the gravity
acceleration ^ = 9.8 m/s^. An example is presented in box 4.2.
4.2.3. Transition Length of Chutes

Performance as 'chute' or as 'drop structore'? Although an open-channel at a steep slope is
constructed, it is not yet guaranteed that it functions as a 'chute' with sub-critical uniform
flow. The condition for sub-critical uniform flow is that the sloping energy level is parallel
to the bed level of the chute. Thus, there is dissipation by friction on the sloping glacis, see
figure 4.7.
For instance, steep and short flumes, or narrow flumes, are not functioning as a 'chute'
with energy dissipation on the sloping glacis. They are too short or too smooth, so that the
friction loss is very small and should be ignored. Thus, they function as a 'drop structore',
where the energy dissipation takes place only in the stilling basin by the hydraulic jump and
eddies.
DROP STRUCTURE
Energy dissipation by:
' turbulence in stilling basin
Figure 4.7. Energy dissipation by a 'chute' and by a 'drop structure'.
102
Analysis. The analysis, whether a fall structure performs as a 'chute' or as a 'drop stmctore',
can be simplified by considering 'energy lines'.
A chute has an energy line parallel to the sloping glacis, at a vertical distance H. This
vertical distance H can be calculated with the above formulae.
However, when this calculated energy line of the glacis is located higher than the
available energy line of the upstream canal reach, it means that an umealistic simation is
created, see figure 4.8. Thus, the following conclusions can be drawn (Ankum 2002):
uniform flow in the chute is not yet reached,
the gradient of the energy line in the chute is less than the gradient of the glacis,
the energy loss is smaller than expected in a chute and is still in the 'transition'. It is better
to ignore the friction losses, and to decide that the strucmre functions as a 'drop stmcture'
only.
Figure 4.8. 'Chute-flow' with friction losses is not yet reached.
Analysis step-bv-step. The following steps can be taken to obtain a first impression, see
figure 4.9 (Ankum 2002):
draw the (horizontal) energy line of the upstream canal reach in the longimdinal section
of the chute,
calculate the energy head H above the sloping bed of the chute,
find the location on the chute where the energy head H coincide with the horizontal energy
line of the upstream reach. This intersection point between the two energy lines may be
called point A ,
the preluninary elevation of energy line in the chute follows the horizontal line until point
A, and continuous with the sloping energy line on a distance H above the bed of the chute,
friction losses occur only beyond point A ,
i f point A is not existing within the range of the chute, the friction losses have to be
ignored and the strucmre does not function as a 'chute'. The uniform chute-flow has
apparently not yet been reached and the flow may still belong to the transition part.
103
energy loss in
stilling basin
Stilling Basin
Figure 4.9. Checking on the 'chute-flow' with friction losses.
Transition length. The first part of the chute is a transition between the 'critical flow' at the
control and the 'uniform and super-critical flow' in the chute. The water depth should be
calculated with a backwater curve technique, which is not discussed here.
The horizontal length L^^^^ccs f the transition zone before a uniform flow is reached in
a chute depends on, see also figure 4.9 (Ankum 2002):
-^trans
hence:
T
trans
H ^
E_
with the (horizontal) transition length Lf^g^g in m, the energy head

at the control above the
crest, the energy head H above the sloping bed of the chute, and the chute gradient s. An
example of the calculation is presented in box 4.3.
Box 4.3. Calculation of the transition length L^x&m
^ chute.
The transition length

of the chute of the above examples can be calculated for
the following design parameters:
the energy head at the control notch at the beginning of the chute follows from the
formu!a:^ = 2X\ b hJ"^, and amounts to
1.16 m for b ^ 2.00 m and Q = 5
m^/s. A sill o(p = J.OU m is applied to avoid backwater curves in the upstream
' the energ\' head H above the sloping bed of the chute amounts to H - 1.78 m;
the gradient ol the chute is 3 = 20 x 10"^.
Thus, the transition length /-t,.,,^ = ( / / ^ /,,)/.?
(1.78 - 1.16)/(20 x 10'-^) = 31.00 m.
The actual transition length will" be somewhat longer, but it is obvious the chute with
the length of 300 m will functum as a ehute with energy losses by friction.
104
5. HYDRAULIC STRUCTURES
5.1. Types of Hydraulic Structures
5.1.1. Non-Regulating
structures
Classification. 'Non-regulating structures' ("niet-regelende kunstwerken"} are structures that

transport the approaching flow, and do not have the primary funtion of regulating the water
level or the discharge.
Non-regulating strucmres in flood and drainage channels can be divided into two
different groups depending on the headloss:
conveyance structures ("doorvoerkunstwerken").
Conveyance stractures are required at
crossings of flood and drainage channels with roads and irrigation canals, see figure 5.1.
These strucmres are designed at a low headloss for all discharges.
Examples are bridges, culverts ("duikers"), aqueducts and (inverted) siphons ("sifons,
grondduikers"). But also, 'gated regulators' ("regelkunstwerken met af sluitmiddelen") with
low headlosses are designed as a 'conveyance strucmre' during the maximum discharge.
PLAN
CROSS SECTION
Figure 5.1. 'Conveyance structure', with low headloss for all discharges.
105
106
drop structures. Drop structures ("stortdammen") are required to dissipate the excessive
energy at steep alignments to avoid erosion in unlined open-charmels, see figure 5.2. These
structures are designed at a high headloss for all discharges. As an alternative, 'chutes'
("hellende goten") can also be applied, which are just lined open-water channels with
super-critical flow. Drop structures are discussed in chapter 12. Chutes have been
discussed in chapter4 on 'Super-critical flow in open-channels'.
PLAN
CROSS SECTION
Figure 5.2. 'Drop structure', with high headloss for all discharges.
5.1.2. Regulating structures

Function. 'Regulating structures' ("regelkunstwerken")
the water level, or
the discharge.
are structures that may regulate:
Water level regulator. The structure that regulate the water level is called the 'water level
regulator' ("peilregelaar"), also called 'check structure' or 'cross regulator' ("stuw"), see
figure 5.3. The function of the water level regulator is to maintain a certain target water
level, i.e. the 'setpoint-in-water-level' ("streejpeil"), at every discharge;
istafff
target gauge.
water
level
a/
'target
iwater level
I
regulatim
A
"Water level
J Regulator
PLAN
CROSS SECTION
Figure 5.3. 'Water level' regulator, with the water level as target.
Regulation of the water level. The regulation of the water level is theoretically easy. The
discharge formula of the water level regulator is not relevant for the regulation as the gate
will be regulated simply by opening or by closing until the water level reaches the 'setpoint'
C "streefwaarde", "streefpeil"), see figure 5.4.
107
5. H Y D R A U L I C STRUCTURES
GATE
STAFF GAUGE
Water level
above
"Set-Point"
-Id
Water level
at
"Set-Point"
Water level
below
"Set-Point"
10
^777777777777777777777777777777777777777.
Figure 5.4. Regulation of the water level.

Discharge regulator. The structure that regulate the discharge is called the 'discharge
regulator' ("debietregelaar"), also called 'offtake' ("afiapping") or 'intake' ("inlaat"), see
figure 5.5. Discharge regulators are structures that regulate the discharge from one channel
to another. The function of the discharge regulator is to maintain a certain discharge, i.e. the
'setpoint-in-discharge' ("streefdebiet"), at every water level.
Discharge regulators are widely applied in irrigation and may occur at different places
in the main irrigation system: at the 'headworks' ("watervang"), at the head-end of secondary
and sub-secondary irrigation canals ("secundaire inlaat"), and at the 'offtakes' ("aftapping")
to the tertiary irrigation units.
Discharge regulators are also used in flood control, when a flood diversion structure
has to divert a certain flood volume into a flood diversion charmel.
Discharge regulators are used less in drainage, where 'water level' control is normally
applied.
Figure 5.5. 'Discharge' regulator, with the discharge as target.
108
Regulation of the 'on/off discharge. Regulation of the discharge is more difficuh than the
regulation on the water level. Regulation on discharge implies that also the passing discharge
should be known. This is smiple for an intermittent water supply, with only an on-off
("aan/uit") function, see figure 5.6. When the regulator is fully closed, the discharge is zero.
When the regulator is fully open, the maximum discharge is passing as per design.
AQ in mVs
AQ m m^s
100%
time
INTERMITTENT FLOW
time
^
ADJUSTABLE FLOW
Figure 5.6. Regulation of 'on/off discharge and of 'different' discharges.
Regulation of the 'different' discharges. Regulation of the different discharges, e.g. 75%,
30%, 55 % of the maximum flow, is more combersome as also discharge measurement should
be done. Several regulators, such as the vertical gate, have a complicated relation between
the gate opening and the discharge. The regulation and the measurement of the discharge can
be done in two ways, by:
regulation gates with an accurate rating curve, such as the Romijn weir and the CrumpdeGmyter gate;
regulating gates with a separate water measurement structure 'in series' ("achter
elkaar"), see figure 5.7. The regulating process of the gate is translated into the control
of the 'tailwater level': the target discharge is obtained by adjusting the discharge regulator
on basis of the tailwater level.
DISCHARGE MEASUREMENT
Figure 5.7. Regulation of the discharge, through the water level.
109
5. H Y D R A U L I C S T R U C T U R E S
5.1.3. Manner of Regulation

Types of regulators. The water level and the discharge regulators can be divided into different
groups, depending on the 'manner' of regulation, see figure 5.8:
passive regulators ("passieve regelaars"). Passive regulators are applied to maintain a
certain water level or to divert the flow. They also called 'fixed regulators' ("vaste
regelaars"). Passive regulators do not have gates, but their function is based on the
'overflow' ("overlaat") or on the 'underflow' ("otiderspuier") principle.
Examples are the 'broad-crested weir' ("brede overlaat"), the 'sharp-crested weir'
("scherpe overlaat"), the 'long-crested weir' ("latige overlaat") or 'duck-bill weir' ("eendebekstuw"), the 'controlnotch' ("sluis"), the 'side-channel spillway'
("zijdelingseoverlaat")
and the 'siphon spillway' ("hevel overlaat");
manual regulators ("handbediende regelaars"). Manual regulators are applied to maintain
a certain water level or to divert the flow. They have manual-operated gates
("handbediende schuiven"). Manual regulation is applicable as long as the frequency of
gate setting is limited. Manual regulation becomes labour-intensive when the flow is
regularly changing.
Examples are the drop-leaf gate ("klepstuw"), the flap gate ("duiker deur?"), the
vertical gate ("vlakke schuif"), the Crump-deGmyter gate, and the radial gate ("segment
deur");
Decrement Ay
, target
staff
gauge
i
target
^2
M A N U A L REGULATION
PASSIVE REGULATION
float
motor
counter weights
sensor
target
t-
HYDRO-MECHANICAL
REGULATION
ELECTRO-MECHANICAL
REGULATION
Figure 5.8. Options for (local) regulation of structures.
hydro-mechanical regulators ("hydro-mechanische

regelaars").
Hydro-mechanical
regulators are applied to maintain a certain water level. A higher water level leads to
higher water forces on the gate, by which the gate opens.
110
Examples are the Begemami gate, the Vlugter gate, the A M I L gate, the AVIS gate and
the AVIO gate.
electro-mechanical regulators ("electro-mechanische regelaars"). Electro-mechanical
regulators are applied to maintain a certain water. A programmable logic controller (PLC)
processes the water level measurements by a sensor and instructs a motor that adjusts the
gate. Examples are vertical and radial gates with PD, PI and PID-controllers. They will
not be discussed here.
Tvpes. There are many types of regulators. They can be divided into two groups depending
on the flow type, see figure 5.9:
Overflow gates ("overlaten"), such as the 'drop-leaf gate', the 'vertical gate for overflow',
and the 'Romijn weir'.
Underflow gates ("onderspuiers"), such as the 'vertical gate for underflow', the 'radial
gate' and the Crump-deGruyter gate.
These types are discussed in the next chapters.
GATE FOR OVERFLOW
GATE FOR UNDERFLOW
Figure 5.9. 'Overflow' and 'underflow' type of the vertical gate.
5.2. Structural Design of Hydraulic Structures

5.2.1. Scouring Protection
General. A l l stmcmres in open-water channels need protection against scouring ("ontgrondingen") at the downstream side and against effects by groundwater. These stmcmres
need to be protected against:
scouring of the downstream channel bed, by constmction of a stilling basin ("woelbak")
and of coffers ("koffers");
piping of groundwater ("onderloopsheid"), by constmction of a sufficient long 'critical
seepage path';
uplift by groundwater pressure, by constmction of sufficient thick floors.
111
Stilling basins. It is a good principle to provide any structure where super-critical

("schietend") flow may occur, with a 'stilling basin' ("woelbak"), see flgure 5.10. This
stilling basin will force the hydraulic jump within the structure, so that excessive scouring of
the downstream channel bed can be avoided.
The most simple stilling basin is just a lowering of the floor at the downstream end of
the strucmre. A n over-depth of e.g. 0.3 X the water depth y is often sufficient.
STILLIN [G BASIN
STILLING BASIN
1
COFFER
COFFER
COFFER
COFFER
Figure 5.10. 'Coffers' and stilling basins for protection of structures.
Coffers. 'Coffers' ("koffers") or 'cut-off walls' are provided at the end of floors and of
stilling basins, see also figure 5.10. Usually, coffers are also applied at the upstream side of
a floor.
Coffers are basically a 'beam' under the floor, to a depth of e.g. 4 x the floor
thickness. The construction cost of coffers is only minor, while the benefits are high. The
function of coffers is multifold:
coffers protect the downstream channel bed against concentrated groundwater flow, and
thus a higher outgoing velocity, see figure 5.11;
coffers protect the floor against any scouring-holes ("ontgrondingen") in the downstream
bed. These scouring-holes makes that the foundation soil below the floor is removed,
which makes that the floor collapse. Moreover, these scouring holes will be propagated
into the upstream direction, thus shortening the 'critical seepage path';
coffers increase the critical seepage path of the groundwater flow.
Low velocity
of groundwater
CORRECT DESIGN:
WITH COFFER
(to avoid a 'blow-out')
Danger of 'Blow-out' of soil

(high groundwater velocity)
WRONG DESIGN: WITHOUT COFFER

(danger of 'Blow-ouf)
Figure 5.11. Groundwater flow and the coffer at the end of a stilling basin.
112
DESIGN O F OPEN-CHANNELS A N D H Y D R A U L I C STRUCTURES
Discussion on rip-rap. It is still a topic between hydraulic engineers whether bed and slope
protection has to be applied at the downstream end of a structure.
It is believed that 'rip-rap' ("losse steenstorting") or 'lining' ("bekleding") is essential
to protect the downstream charmel bed against scouring. The length of such a protection may
be in the order of 4 x the downstream water depth y.
Others believe that a well-designed stilling basin will reduce the ttirbulence of the water
and will dampen any waves ("golven") of the water, so protection is not needed.
5.2.2. Piping below the Floor

Piping. Flood diversion structures should be checked on the danger of 'piping'
("onderloopsheid"). Seepage ("kwel") will damage a structure when the foundation soil is
removed by the flow and holes are formed. With 'piping', a direct connection is created
between the upstream and the downstream part of the structore, resulting in more
groundwater flow and dangerous losses of foundation soil. Ultimately the structore will
collapse. The material transport and the resulting piping can be prevented by, see figure 5.12:
lengthening the distance of seepage flow, so that the gradient of the groundwater flow
becomes flatter. This can be effected by e.g.:
o longer floors. However, the lengthening of the downstream floor will increase the uplift
pressure,
o cut-off walls and sheet piles ("schermen en damwanden")
the use of filters. Good filters will prevent the removal of foundation soil. They can be
applied at the end of the structore as part of 'rip-rap'. However, the torbulence in the
water may damage filters here. Filters have to be applied in 'weep holes' (" drainage
openingen") of lined slopes ("talud bekledingen").
The danger of piping can be checked by the preparation of a flownet, and by several
empirical methods such as the Lane's method.
Connection
must be constructed
Figure 5.12. Protection of structures against 'piping' ( D H V 1986).
113
Lane's method. The Lane's method compares the 'creep ratio' below the structore with the
allowable creep ratio for that soil. The creep ratio is the critical seepage path under the
structore, divided by the drop in water level z over the structore. The 'critical seepage path'
is the sum of, see figure 5.13:
the vertical path distance E L^^^^ along the structore,
one-third of the horizontal path distance E L^q^. >
two tunes the short-cut path distance E Z-short-cut *at shortcuts through the soil.
The background is that the vertical path has three trnies more resistance against flow than the
horizontal flow. Slopes steeper than 45 are regarded as vertical, and those less than 45 as
horizontal. Thus, the 'creep ratio' C L follows from:
Yl^ven.
"7
XlAior,
2 5^ />short -cut
where: C L is the creep ratio, E L^^^^ is the sum of the vertical lengths in m, E L^g^. the
sum of the horizontal lengths in m, t Z-short-cut the sum of the shortcut lengths in m, and
z is the headloss over the structore in m.
Typical values for the allowable creep ratio are: C L = 2 for hard clay, C L = 3 for soft
clay, C L = 4 for fine gravel, C L = 5 for coarse sand, C L = 7 for fine sand, and C L = 8
for sih. Thus, clay allows for the shortest critical seepage path.
5.2.3. Uplift of the Floor

Uplift. Floors subject to 'uplift' pressure ("opwaartse druk") because of groundwater below
the floor, see figure 5.14. Normally, the uplift pressure can be reduced by:
making a longer floor at the upstream side of the structore;
applying 'weep holes' with filters.
The uplift pressure can be calculated by drawing a flownet or with the assumptions used
by the Lane's critical seepage path.
114
Lane's method. The Lane's method calculates the uplift pressure

in a point x by assuming
that the drop in pressure over the structore is proportional to the critical seepage path, thus:
10000
floor
d)
E^vert.
l/3E^hor.
2 ^3 L^,
where: P^^ is the uplift water pressure in point x in N/m^, y^^^^. is the water depth above the
floor in m, d is the thickness of floor at point x in m,
is the relative critical seepage path
from downstream to point x in m, E L,^Q^^ is the sum of the vertical lengths in m, E L^^^ is
the sum of the horizontal lengths in m, E i-short-cut the sum of the short-cut lengths in m,
and z is the headloss over the structore in m.
Minimum thickness of floor. The floor of the stmctore has to resist the uplift pressure of the
groundwater. The minimum floor thickness d follows a stability calculation: "the uplift
pressure P^ should be less than the weight of the floor plus the weight of the water above the
floor". A unit weight of 2000 kg/rn^ is often assumed for the floor. Thus:
20000 X d + 10000 X y^^^^ > P^
where: d is the thickness of the floor in m, yfjoor is the water depth above the floor in m,
is the uplift pressure in N/m^.
The above formula has to be applied for the stability of masonry ("metselwerk") floors.
However, some reduction in floor thickness can be allowed for reinforced concrete floors,
where the structore as-a-whole resists the uplift force.
Ouick check on uplift. A quick check of the minimum thickness of masomy floors assumes
proportional decrease of the water pressure under the floor, without considering the vertical
seepage paths. Moreover, an empty stilling basin, thus yfloor = 0. is assumed. It means that
the weight of the floor is determined by:
20000 X d
>
10000 d
[
T L
E^hor.
x / ,
hence by:
>
x z
^hor.
Graphically, it means that the thickness d of the floor should be larger than the distance
between the floor level and the (straight) water pressure line, see also figure 5.14.
maxunum
water level
Drop
z
mmunum
iwater level
Figure 5.14. Quick check for the uplift water pressure under a stilling basin.
115
5.2.4. Construction Drawings

Types of drawings. There are different types of drawings during the design of structures. The
hydraulic engineer will use 'principe sketches' of the structures, with its hydraulic
dimensions giyen in letters, such as the width b, the water depth y, the energy head H, and
the sill height p.
Sometimes, there are '3D-drawings' with the artist hnpression of the stmcmre, see
figure 5.15. Often, these 3D-drawings are not very usual for the engineers, as these drawings
hide too many details.
The design engineer will produce 'engineering drawings' of the strucmre, to be used
by the drawing office, see figure 5.16. This drawings have a plan ("bovenaanzicht"), one or
more longitudinal sections ("lengtedoorsneden"),
and one or more cross sections
("dwarsdoorsneden"). The plan and sections are closely linked with each other, and the
design engineer draws them simuhaneously.
The draughtsmen in the drawing office prepares the 'construction drawings' to be used
for the tendering ("aanbesteding") and for the constmction by the contractor, see figure 5.17.
A l l relevant details are given in this drawing. Also the measures and the levels are shown.
An 'as-built drawing' is often prepared after the stmcture is completed by the
contractor. A l l revisions during the constmction are shown.
Figure 5.15. Example of a '3D-drawing' of a structure.
116
Preparation of the 'engineering drawing'. The 'engineering drawing' of a stmcture can be

drawn gradually in a certain sequence, see also figure 5.16.
Firstly, the plan is drawn, on basis of the data from the channel design. The location
of the stmcture is selected. The vertical walls of the stmcture are drawn, with its transition
to the incoming charmel.
Secondly, the longitudinal section of the above is drawn. This is done at the same scale
as the plan and precisely below the plan. The floor is drawn, and the water level regulator.
Also the longitudinal section of the stilling basin is drawn. The 'coffers' or cut-off walls at
the up-and downstream end of the floor are drawn.
Thirdly, the plan is adjusted to the above elements in the longitudinal section. The
vertical walls of the stmctures are extended into the downstream direction so that they include
the stilling basin. Also the transition of the vertical walls to the continuing chaimel is drawn.
Fourthly, a check is made whether everything is drawn. A bridge is often needed. The
hatching ("arcering") of cross sections is made for soil, concrete, etc. Also the hatching of
the slopes is made. The water is not drawn in the plan. The maxhnum water level is drawn
in the longimdinal and in the cross sections.
-2.20 m^
-3.70 m^
LONGITUDINAL SECTION
Figure 5.16. Example of an 'engineering drawing' of a stmcture.
117
5. H Y D R A U L I C S T R U C T U R E S
SECTION A A
Figure 5.17. Example of a 'eonstruction drawing' of a structure (Nedeco 1986).
118
rrT.T.i.i.r.i.i.i.i.t.i.i.r.CT
aan
0 ^
'7:
Figure 5.18. Example of a 'eonstruction drawing' of a structure (Nedeco 1986).
10 m
6. DESIGN OF CONVEYANCE STRUCTURES

6.1. Types of Conveyance Structures
6.1.1. Culverts
Culverts. A 'culvert' ("duiker") is a structure that conveys a drainage channel or an irrigation
canal underneath a road, a railway or another water course, see figure 6.1, figure 6.2 and
figure 6.3.
There are several possible flow conditions through a culvert, depending on the
discharge, the cross section, the length of the culvert and the downstream flow conditions.
The flow conditions in the 'barrel' ("koker") can be:
an open canal flow, with a free water surface,
a pipe flow, with the barrel completely filled.
Furthermore, the water in the culvert may flow as:
free flow ("volkomen, schietend") or 'inlet control', for large headlosses, which is
basically contradictious to the requirement of a conveyance structure (i.e. low headloss);
submerged flow ("onvolkomen, verdronken") or 'tailwater control'.
Design. The design of culverts is essentially based on the headloss calculation as presented
below, thus for 'tailwater control'. The design velocity may range from 0.5 to 2.0 m/s and
determines the headloss z over the strucmre.
However, 'inlet control' is also possible for a large headloss with supercritical flow m
the barrel, see figure 6.1. The inlet control is not discussed here, and should be considered
from a hydraulic point of view as a 'drop struchire'.
The choice of a single or multi-barrel culvert depends on the bed width of the canal,
the water depth of the canal invert and the available headloss for the culvert.
irrigation canal
drain
1L
!:::>
r
^
I
'//////////////////////////A
culvert
Figure 6.1. A culvert as a cross-drainage structure, with 'inlet control'.

119
1 drain
120
3.00 m
?o
00
GO
CO
1 m
1.65
5.00 m
plan
2.00 m
1 1
1
m
l o n g i t u d i n a l section
Figure 6.2. A culvert in a canal.

I
min. cover 0.60 m

or 15 D
frtf
-
section A
!*r
I 11T"in masonry j
i|!
section
submerged composite culvert

C;oI
T
D<81
'l reinforced concrete cover

I masonry bottom and walls
section
submerged composite culvert - alternative
section D
f
section
Figure 6.3. Types of culverts, and types of culvert-flow
section
(NEDECO
1986).
6. D E S I G N O F C O N V E Y A N C E STRUCTURES
6.1.2.
121
Aquaducts
Aqueducts. A n 'aqueduct' ("aquaduct") is another type of a conveyance strucUire, that is

design at a low headloss. Aqueducts are used to divert irrigation canal over a drain or even
a river, see figure 6.4. The greatest part of the live load is the water in the aqueduct. Thus,
the velocity v should be kept high to minimize weight. On the other hand, the velocity v
should be kept low as to minimize the headloss z. The velocity v should be at least twice the
canal velocity to discourage deposition of sediment, but generally should not exceed v = 3.0
m/s The box section of the aqueduct acts as the spanning beam, so an economic design
depends on a high ratio bIy, where b is the width of the stmcmre and y the water depth.
Ratios of bIy = 1 to 3 are often used.
Figure 6.4. A n aqueduct (NEDECO 1986).
122
6.1.3,
Siphons
Inverted Siphons. A n 'inverted siphon' ("sifon, grondduiker") is also designed normally at

a low headloss. Inverted siphons are mainly used in irrigation canals for river crossings where
an aqueduct is unsuitable, see figure 6.5.
The decision whether to use an aqueduct or an inverted siphon in an irrigation canal
depends on a number of factors, the main one being sediment transport. The siphon will trap
all sediment at low flow, as the velocity in the siphon will become much lower than in the
canal. Special actions may be required, such as wash-out facilities, or simply gates to close
one of the barrels during low discharges.
To prevent siltation and blockage, the velocity v should be kept high. However, a high
velocity leads to a high headloss. Velocities should be at least twice the normal canal velocity
and in any case not less than 1.5 m/s. Maxhnum velocity should not exceed 3 m/s.
The top of the siphon entrance is set below the normal water surface. This will
minimize possible reduction in the siphon capacity caused by the entering of air into the
siphon. The depth of submergence is known as the water "seal". The required height of the
seal depends on the slope and the size of the barrel, and can be taken at 1.5 times of the
entrance loss, with a minimum of 0.15 m.
'Trashracks' ("krooshelcken") have to be installed at the entrance of siphons as to avoid
the involuntary entrance by persons or animals and to avoid clogging by floating debris.
However, the blocking of the trashrack will increase the entrance headloss.
The inverted siphon will be under internal pressure and hence joints between the pipe
elements should not be used. Instead a reinforced concrete collar may be used, preferably cast
in-situ.
Figure 6.5. Siphon, to transport water below another water course ( N E D E C O 1986).
123
6. D E S I G N OF C O N V E Y A N C E STRUCTURES
6.2. Hydraulics of Conveyance Structures

6.2.1. Energy Losses
Friction Losses. The 'headless' ("verval") Zf in a conveyance structure due to the friction
("wrijving") can be calculated with the Strickler formula: v = ^ R^^^ s^^'^, where s = z I L
is the hydraulic gradient over the structure, thus by:
where Zf is the headloss by friction over the struchire in m, L is the length of structure in m,
V is the velocity in the structure in m/s, R is the hydraulic radius in m, k is the Strickler
coefficient in m^''^/s, g = 9.8 mls^ = 9.8 N/kg is the gravity acceleration.
The Strickler coefficient k may have values of A; = 50 m^'^ls for stone masomy, to k
= 70 m^''^/s for concrete.
Transition Losses. The headloss
over inlet ("intree") transition and the headloss
over
the outlet ("uittree") transition for sub-critical ("stromend, niet-schietend") strucmres can be
calculated by (Chow 1959):
(V - v ) '
(V -
v,)2
C,n
Yg
'
^
2g
where:
and Cgyj are headloss coefficients over inlet and outlet, v is the velocity within the
stmcmre in m/s,
is the velocity in the upstream canal in m/s, v^j is the velocity in the
downstream canal in m/s.
In practice, the above formulae are often simplified and are related to the velocity head
("snelheidshoogte") in the stmcture:
2i
în
-in
and
ôut
-out
"
The headloss coefficient depends on the hydraulic shape of the transition and on whether
it is an irdet or an outlet, see table 6.1 and figure 6.6.
It is recoimnended to use vertical-walled transitions of either (i) rounded-flared wall
transition {R > 0.5 y), (ii) with straight-line walls under 45 to the canal centerline, or (iii)
rectangular walls under 90. A gradual transition, e.g. 'warped walls' ("vallende muren") can
be applied when the transit losses have to be at a minimum. However, the constmction of
warped walls is somewhat more labour-intensive.
Normally the headloss is so small for these types of transitions that it may be neglected.
However, a minimum headloss of 0.05 m is taken at stmcmres with transitions to cover
possible additional losses, such as turbulence caused by gate slots ("sponningen"), friction
through the stmcmre.
124
WARPED
TRANSITION
inlet loss 0.1 v72g
outlet loss 0.2 vV2g
ROUNDED
TRANSITION
inlet loss 0.2 vV2g
STRAIGHI-LINE
TRANSITION
inlet loss 0.3 vY2g
RECTANGULAR
TRANSITION
inlet loss 0.5 v72g
outlet loss 1.0 vY2g
Figure 6.6. Transition head losses of inlets and outlets.
Table 6.1. Headloss coefficients Cj for inlets, and c* for outlets.

Type
of
transition
warped transition
rounded transition
straight-line transition
rectangular transition
headloss
over inlet
qn
qn
qn
qn
=
=
=
=
coefficients
over outlet
0-1
0.2
0.3
0.5
Table 6.2. Headloss coefficients in bends (crenel) ^^'^
ôut =
=
Com =
Cout =
0.2
0.4
0.5
1.0
elbows (Cgiô^).
angle a of B e n d and
a = 30
a = 45
a = 60
Elbow
o; = 9 0
BEND:
^bend =
^bend = 2Z)*
^bend =
-bend = 0.1
-bend = 0.1
-bend = 0.1
Cbend = 0.1
Cbend = 0.1
Cbend = 0.2
= 0.1
Cbend = 0.1
Cbgnd = 0.2
Cbend = 0.1
Cbend = 0.2
Cbgnd = 0.3
ELBOW:
Circular
Rectangular
-elbow = 0.1
Celbow = 0.2
cîbow = 0.2
Cgibow = 0.3
Ceibow = 0.5
Cgibow = 0.6
cîhov^ = !!
Celbow = 1-4
D is the diameter of the pipe.
125
6. D E S I G N O F C O N V E Y A N C E STRUCTURES
Bend and Elbow Losses. 'Bends' ("bochten") and 'elbows' ("knikken") in siphons or pipes
cause a change in the direction of flow and consequently a change in the general velocity
distribution.
Also the headloss zênd ^ ^^^'^ ("bocht") and the headloss Zgibow
elbow ("knik")
may be expressed as a function of the velocity head in the pipe:
^bend
^bend
'
êlbow
-elbow
Values of headloss coefficients c^^^^ in bends and headloss coefficients cî^^^ in elbows are
presented in table 6 . 2 .
6.2.2. Hydraulic Design of Conveyance
Structures
Froude number. Conveyance structures should have sub-critical flow ("normal" flowing
water) as to the limit the headloss. I t means that they have a Froude number Fr < 1. In
design, however, the Froude number is often Imiited to
< 0.5 to avoid standing waves
at the water surface and to avoid flow to become critical because of decreased canal
roughness.
The Froude number Fr can be calculated for structures with vertical wall and an openwater surface by the formula:
Fr
= ^
with
V = ^
where Fr is the Froude number, Q is the discharge in rn^/s, v is the velocity within the
structure in m/s, y is the water depth within the structure in m, b is the (open-water) width
of the structure in m, g = 9.8 m/s-^ = 9.8 N/kg is the acceleration of gravity.
The open-water surface width b for full-flowing siphon pipes equals to zero, thus the
Froude number is not defined. Typical values of the maximum velocities are presented in box
6.1.
Box 6.1. Maxhnum velocity in conveyance structures with free water surface.
The maximuu design ve]ocit>' in conveyance structures with an open water surface is
limited by the Froude number Fr < 0.5 to avoid standing waves. Hence, the design
\elocity v within the structure is limited by t- < 0.5 {gvf^'^. where y is the water depth
in the structure in m.
Thus, the lollowing maximimi design velocities have lo be maintained:
' } " 0.5U m
< . ' . . . . . > . , < . . . . . > . < . . . . . , , . . ^'jjjjj^ ~* i '1 ni/i>
y = 0.75 m :
Vj!^x = ^-^ "^^^
>' 1.00 m: .
^'max ~ ^ *^
y = 1.25 m:
v^^,^ = 1.8 m/sy = 1.50 m:
= 1.9 m/s
y = 2.00 m:
1-^,^!^^ = 2.2 m/s
) = 2.50 m;
v^.^ s? 2,5 m/s
" y = 3.00 nr.
f,^^,^j^ = 1.7 ni/s
126
Total headloss. The total headloss Zstructure

^ conveyance stmcture is the sum of the
entrance ("intree"), friction, bend and/or elbow and the exit ("uittree") losses, and varies with
the velocity v in the stmcture:
^structure
2 e L
c- +
"
k'^
^^'^
,2
+ m^c.
.+
bend
nx c
,u
elbow
+c .
out
where: L is the length of stmcmre in m, R is the hydraulic radius in m, k is the Strickler

coefficient in m^^-^/s, m is the number of bends, n is the number of elbows, Cjjj and
are
headloss coefficients over inlet and outlet, v is the velocity within the stmcmre in m/s, g =
9.8 m/s^ = 9.8 N/kg is the gravity acceleration.
Hydraulic design. Some examples of the entrance and the exit of conveyance stmcmres are
presented in figure 6.7. In fact, these transition types can be applied for all kind of stmcmres,
such as drop stmcmres, water level and discharge regulators.
It is advised to constmct a simple 'stilling basin' ("woelbak") at every outlet transition
to prevent scouring when super-critical flow ("schietend water") may happen during e.g. mismanagement of the canal. Such a stilling basin can be achieved by lowering the bed by some
Vs X the water depth y.
Furthermore, it is strongly advised to constmct a 'coffer' ("koffer") at the end at any
flood, as to prevent the concentration of groundwater flow at the bed level.
7. DESIGN OF PASSIVE REGULATORS

7.1. Weirs
7.1.1. Broad-Crested Weirs
General. The rectangular 'broad-crested weir' ("brede overlaat") is an overflow structure
with a horizontal crest, while the streamlines are straight and parallel, see figure 7.1.
There are no movable parts on the strucmre. The broad-crested weir can be used for
water level regulation, but also for 'proportional discharge regulation' ("proportiotiele
verdeling"). Furthermore, the broad-crested weir is applied in irrigation canals as a stmcmre
for discharge measurement ("qfvoermeting") under small headloss, see figure 7.2.
Discharge Formula:
Q = 1,7 b
Conditions:
H<KLorH<3z
Discharge Formula:
Q 1.9 b tf/^
Conditions:
H>KLorH<<z
BROAD-CRESTED WEIR-FLOW
SHARP-CRESTED WEIR-FLOW
Figure 7.1. Broad-crested and sharp-crested weir flow.
127
128
Hydraulics. The basic discharge formula of a broad-crested weir with a rectangular control
section under free flow is, see also figure 7.2:
Q
= cb H 1.5
with:
= c, X 1.71
-g
3 \
1/2/
where Q is the discharge in m-'/s, c is the weir coefficient in m^'^/s,
is the discharge
coefficient, b is the width of the crest in m, H is the energy head aboye the crest in m.
FRONT VIEW
Figure 7.2. Hydraulic parameters of the broad-crested weir for discharge measurement.
Discharge coefficient. The discharge coefficient

depends on the curye of the streamlines
aboye the crest, and amounts to
= 1 for the broad-crested weir. Thus, weir coefficient of
the broad-crested weir is c = 1.7 m^^'^/s. Thus, the hydraulic design equation of a broadcrested weir with a rectangular control section is often simplified into:
Q
= c b H^-^ with c = 1.7 m^'^/s.
Validity. The tailwater leyel may not rise too high to allow a 'free flow' ("volkomen
stroming"). The free flow (non-submerged, modular) condition is satisfied for a headloss z
oyer the strucmre of z > Vs H.
7.1.2. Sharp-crested weir

Hydraulics. The basic discharge formula of a 'sharp-crested weir' ("scherpe overlaat",
"meetschot") with a rectangular control section under free flow is shnilar to the formula of
the broad-crested weir:
_2
g
=
1.71
3 \ 3
where Q is the discharge in m^/s, c is the weir coefficient in m^^^/s, c^ is the discharge
coefficient, b is the width of the crest in m, H is the energy head aboye the crest in m.
Q
= cb
with:
Broad- and sharp-crested weir flow. The difference between broad-crested and sharp-crested
is a matter of curyed streamlines aboye the crest, see also figure 7.1.
A broad-crested weir-flow occurs when the energy depth H is less than Vi tunes the
crest length L . For instance, a concrete crest of L = 0.30 m functions as a broad-crested weir
for H < 0.15 m with the (critical) water depth y = %H = 0.10 m.
129
7. D E S I G N O F PASSIVE R E G U L A T O R S
Discharge Formula:
Q1.9btf'^
Conditions:
z H
UNROUNDED CREST
Discharge Formula:
Q2.1btf'^
Conditions:
Discharge Formula:
Q2.3btf/2
Conditions:
z > K H
Z > X H
CYLINDRICAL CREST
OGEE CREST
Figure 7.3. Types of sharp-crested weirs.
Weir coefficients. Typical values for the weir coefficient c are, see figure 7.3:
c = 1.9 m^''^/s for um-ounded crests,
c = 2.1 m^''^/s for cylindrical crests ("cylindrische kruin"),
even as high as c = 2.3 m^^^/s for the specially designed 'ogee-crests' (USBR 1973).
Headloss. Unrounded crests require a large headloss z for 'aeration of the nappe' ("beluchting
van de straal"), thus z > > H. Cylindrical crests and ogee-crests require a minimum headloss
^min ^
H.
The radius R of cylindrical crests may not be taken too small, because of the subatmospherical pressure under the nappe. Usually, cavitation is prevented with a radius 0.1 H
< R < 0.1 H for concrete weirs, and 0.3 H < R < 0.1 H for masonry weirs.
Cipoletti weir. The 'Cipoletti weir' is a special type of sharp-crested weir and has been
developed as measuring device ("meetinrichting") in irrigation, see figure 7.4.
The Cipoletti weir has a trapezoidal control section: the crest is horizontal and sides are
sloping outwards by I j j ^ j . : 4ygj.j. The hydraulic design equation of a Cipoletti weir is (Bos
1989, Kraatz 1975), see also figure 7.4:
Q = 1.9 bH^-^
where Q is the discharge in m^/s, H is the energy head upstream in m, b is the width of the
crest in m.
Application. The sharp-crested weir is normally used in rivers at headworks for irrigation
("watervang") and at flood diversion structures. The advantage is a shorter width b of the
weir, because of the higher weir coefficient c. This makes the costs of the weir lower.
However, a sharp-crested weir requires more headloss for free flow conditions.
130
4V E R T
FRONT VIEW
IHOR
Figure 7.4. Hydraulic parameters of the Cipoletti weir for discharge measurement.
7.1.3. Control
Notch
General. A 'control notch' ("vernauwing") is used to control upstream water levels in a

channel, and to maintain the uniform water depth in the channel for any flow, see figure 7.5.
It is usually applied at entrances of super-critical flow strucmres, such as drop strucmres and
chutes, thereby preventing increased 'shear stress' or 'tractive force' ("wand schuifspanning")
in the channel which otherwise could cause erosion.
The advantages of the trapezoidal control notch are: it is a smrdy strucmre that can be
constructed at low costs, and which does not need much maintenance, it does not need
regulation, and the strucmre does not have a sill or an orifice, and therefore passes sedunent
and debris.
Figure 7.5. Control notch for water regulation at the uniform canal depth.
131
Hydraulics. The depth of the notch equals the bed leyel of the channel, see figure 7.6. The
side slopes of the notch are trapezoidal-shaped. The bottom width, top width and the side
slope of the notch are determined. The relation between the discharge Q and the energy head
U of the control notch under free flow conditions, follows from the formula:
Q = 1.8 {B +
%mH)H^'^
where Q is the discharge in m^/s, H is the (uniform) energy head in the channel in m, B is
the bed width of the notch in m, m is the side slope of the notch walls (lyert '"Hor)Design. The trapezoidal-shaped opening of the control notch is designed to avoid backing up
("opstuwing") or drawdown ("afiuiging") for discharges in the range o f e . g . 20% and 100%
of the design flow.
The design steps are:
o determine the range of discharges for which the control must work, e.g. 20% - 100% of
design discharge;
o calculate the energy (-water) depths in the upstream channel for these two discharges;
o determine the design values B and m from the two equations:
0100% = 1'8
+
^ ^100%) ^^100%^,;f
Q
= l.S{B + V3mH 20%) H
A design example of a control notch at a drop strucmre in a drain is presented in box 7.1.
niVERT IHOR
Figure 7.6. Hydraulic parameters of the 'Control Notch'.

Box 7.1. Design example of a control notch.
The "control" of a fall structure is designed, to prevent racing of the water to the fall.
The drain has a water depth of v - 1.00 m ( / / - 1.01 m) for the design discharge of
(?joyr^ = 2 nrVs, and o f j = 0.41 m {H = 0.41 m) during Qop^vf = 0.4 nrVs.
The control notch shcmld prevent any drawdown or backing up within this range.
This is obtained when the ener' depth at the control notch equals the uniform energy
depth in the drain.
The two design parameters B and m of the notch follow from the two conditions:
2.0 = l.S {B + \ m LOO) 1.00^'^ ox B + 0.67 m = 1.11
0.4 s= 1.8 (B +
m 0.41) 0.41^'^ or B + 0.27 tn --= 0.85
Thus, m ^ .26 / 0.40 = O.t, and B ^ l.U - 0.67 x 0.66 = 0.67 m.
As a check, the energy depth at the control notch during 60% of the design
discharge ((2(,(j,; = 1.2 vr?h) is compared with the energy depth {ll^yy^ - 0.75 m) of
the uniform canal ilow. The energj' depth H ai the notch follows from:
HîQI
f l . 8 X 0.67 + 1.8 X % x 0.66 H\ )''-^ - {Q / [1.21 -f 0.79 H] )
and amounts to H ^- 0.76 ra lor
== 1-2 m''/s. So, ihcie are also no backwater
132
7.2. Long-Crested Weirs

7.2.1, Types of Long-Crested Weirs
General. 'Long-crested weirs' ("lange overlaten") are passive regulators, and are frequently
applied for water level control at flood diversion structures and in irrigation. A long-crested
weir maintains the water level within certain narrow limits.
Types of long-crest weirs. Long-crest weirs have been developed into different types, as to
match the length of the weir with the width of the canal, see figure 7.7:
the diagonal weir, which has an angle "a" of more than 0 with the cross section of the
canal;
the duckbill weir ("eende-bek overlaat") or horse-shoe weir, which has a U-shape and has
an angle "a" of more than 45;
the Z weir, which is a combination of the normal weir and a weir with an angle of "a"
with the cross section of 90.
Of these, the duckbill weir is the most commonly used because it is providing optimum
discharge in relation to the amount of construction works.
The provision of a gate is advantageous on larger strucmres to enable evacuation of the
upstream reach.
NORMAL
WEIR
DIAGONAL
WEIR
DUCKBILL
WEIR
Z
WEIR
Figure 7.7. Types of long-crested weirs.
Hvdraulics. The calculation of the discharge over a weir is based on the general weir formula
for free flow:
Q = cb
i?3/2^
where: Q is the discharge in m^/s, c is the weir coefficient for free flow in m^''^/s, b is the
width of the weir in m, H is the (upstream) energy level above the crest level in m. The
tailwater level y^, should be lower than % H to satisfy the free flow condition.
7.2.2.
133
Duckbill Weir
Duckbill weir. The duckbill weir is a water level regulator without moving parts and can be
used for passive water level control (Kraatz 1975), see figure 7.8 and figure 7.9.
The duckbill weir is a weir with a certain length of the crest. Such a long-crested weir
controls the water level at a certain target level within relatively narrow Ihnits. Certain
fluctuations of the upstream water level with varying canal discharges LQ = max ' minThese fluctuations are unavoidable and can be restricted by developing a large crest width b.
This range of water level flucUiations is called the 'decrement' ("afwijking", "speling") AH.
The decrement is found with all passive and hydro-mechanical water level regulators.
The narrower the decrement AH, the greater must be the crest length b. Usual the
decrements are in the order of AH = 0.05 to 0.10 m, but they depend on the level
requirements of the offtaking discharge regulators. The provision of a vertical gate is
advantageous on larger strucmres to restrict the length of the stracture, for emptying of the
upstream reach, or for flushing of sediment, see also figure 7.9.
The calculation of the weir length b depends on the decrement AH for the varying
discharge AQ = max ' min ^"^^^ ^ long-crested weir and is based on the general weir
formula for free flow (Q = c b H^'^). Thus:
/g2/3 _ Q2/3|3/2
c
AH'I'
where: Q is the discharge in m^/s, c is the weir coefficient for free flow in m^^^/s, b is the
width of the weir in m, AH the decrement for the varying discharge max ' min
^
Figure 7.8. Duckbill weir for a 'passive' water level regulation.
134
PLAN
Decrement
AH
Decrement
y////////////M
L O N G I T U D I N A L SECTION
VARL>iNT W I T H G A T E
Figure 7.9. Long-crested weir for 'passive' upstream control.
Weir coefficient. The weir coefficient c depends on the size and shape of the crest, as well
as on the angle a. of the crest with the cross section of the canal:
for a chaimel flow rectangular on the weir direction, so the angle a = 0 , the coefficient
c of the normal broad-crested weir amounts to c = 1.7 m^^^/s;
for a channel flow not rectangular on the weir direction, so the angle a > 0 , there will
be an energy loss because of the sudden deflection of flow {"verandering van stroomrichting"). This can be incorporated in the weir formula by reducing the weir coefficient
c while calculating with the original energy depth H above the crest. For instance (see also
Kraatz 1975):
o the diagonal weir, with an angle 0 < o; < 4 5 , can be designed on a coefficients c
= 1.7ml/2/stoc = 1.5 m ^ % ;
o the duckbill weir, with an effective angle 45 < a < 90, ean be designed on a
coefficients c = 1.5 m^''^/s to c = 1.2 m^^^/s;
o the Z weir, with an angle a 9 0 , can be designed on a coefficients c = 1.2 m^^^/s.
Application. The great advantage of long-crested weirs is their simplicity in construction and
maintenance, and their reliability in operation. Mismanagement of the strucmre is impossible.
A disadvantage of long-crested weirs is its trap sediment ability which prohibits their
use when water is permanently charged with silt. Another disadvantage is that the length of
the long-crested weir will become very long when a small decrement MI is allowed. Longcrested weirs have a headloss z during the design discharge Q in the range of z > Vs H^^^.
135
7.3. Escapes
7.3.1. Types of Escapes
Function. 'Escapes' are canal protective stmctures, that are used to protect canals and
stmctures against damages caused by excess amounts of water, see figure 7.10. The capacity
of such an escape should be adequate to discharge any excess amounts of water resulting:
- from floods entering the canal,
- during operational mistakes.
In fact, they protect the water level in the canal against rising to such a level that the banks
will be endangered or overtopped.
Tvpes. Protection can be realized with different types of stmcmres:
- manual regulator, i.e. a manual-operated gate, as discussed in the previous chapter.
- 'side-channel spillways' ("zijdelingse overlaat") or lateral spillways, are constracted in
the bank of the canal, at low cost but often over a considerable length;
- 'siphon spillway' ("hevel"), to evacuate a large flow for a fairly constant water level;
- hydro-mechanical regulators, i.e. the gates under upstream control, as discussed in the
next chapter.
3 mVs
5 mVs
ImVs
1 mVs
HUHHH
1 m7s^
2mVs
1 mVs
1 mVs
TTTHTTTT
m
1 mVs,
1 mVs
TTTfTTTTT
VVV
TITITTTTT
LEGEND
3
canal capacity in m^s
discharge in m^s
A
110 mVs
11 mVs
escape
tertiary offtake
Figure 7.10. 'Escapes' are sometimes needed in irrigation systems.
7.3.2. Side-channel spillway

General. 'Side-channel spillways' ("zijdelingse overlaten") are passive water level regulators,
figure 7.11. They are constmcted as a long, protected, lowered part of the canal embankment
for the purpose of spilling excess flow above a certain water level. The crest of the sidechannel spillway is slightiy above the 'target water level' ("streejpeil"), and its length
determines the rise of the water level above this target water level.
136
Figure 7.11. A side-channel spillway.
137
Decrement. The 'decrement' ("afwijking") of the side-channel spillway is the difference

between the design water level, i.e. which is just under the crest of the spillway, and the
water level in the ongoing channel during the flood, i.e. sufficient above the spillway crest.
It should be noticed that the water depth along the spillway decreases (!!) in upstream
direction due to the increase of the velocity head at a constant energy head, see figure 7.12.
Spillway length "L"
yjy//////////////^^^^^
Figure 7.12. Side-channel spillway as a 'passive' regulator.
Design. A design formula of the side-channel spillway does not exist and a numerical method
has to be used to calculate its length L , see figure 7.13. Basically, the weir formula {dQ =
c db if''^) is valid. The energy head H refers to the flow over the crest, and is rectangular
to the flow direction in the canal. Therefore, the energy head H equals to the canal water
level y - p above the crest, as the total energy head \P-I2g is lost by the sudden flow deflection
("plotselinge verandering van stroomrichting").
The calculation is done in steps of db into an upwards direction, and the outflowing
discharge E dQ is calculated until it equals the required value of fiood ' Qo- '^^^ corresponding value E db equals the spillway length L. A n example of the calculation is presented in
box 7.2.
Spillway length "L"
^^^^^^^^^
AL
Figure 7.13. Parameters for the step-by-step calculation of the side-channel spillway.
138
D E S I G N O F O P E N - C H A N N E L S A N D H Y D R A U L I C S TRUCTURES
Box 7.2. Example calculation of a side-channel spillway.
Fomuilac. Discharge over spillway length "XU':

A C c M a-, Discharge at vertical "2":
Water depth al vertical "2":
' a
X
and rewritten f o r the use of the programmable calcula^tor or computer:
the energy head H;
H - p + [Q^, IA^, p. I / 2g
the discharge
at S(A/.):
=
+^ c AL iy^ - p f the water depth y2 at X(AL): (h-^ m
(H / 2
Spillway. The normal di,schargc of <2 = 7 m'^s flov^'s through a channel with bed width
of b^j - 5 m, side slopes
: 1.5^^,^ Strickler coefficient k - 40 m^'-Vs and a gradient
of .V = 0.22 X 10"-^. A full supply level o f y = 1.54 m can be calculated with the
Flood waters can enter the channel at drainage inlet structures and a total
discharge during flood of = 12 m-^/s can be expected in the channel.
The flood waler should be evacuated over a side-channel spillway. Downstream
of this spillway the water level may rise 0.10 m higher than the f u l l supply level or y^,
- 1.64 m. The corresponding channel discharge amounts 7.83 ra7s, as Can be
calculated with the Strickler formula. The remainder discharge should be spilled over
the spilKvay. The 'step method' calculates a total spillwav length of L 100 m (steps
IB^^^BIIIIIBIIIBIBBiBBHIBBB
'Comrolled' spillwav. The spillway length can be reduced by choosing the site of the
spillway up.stream of an obstruction in the channel.
For instance, a siphon of l.HO x 1.80 m^ (A ^ 3.24 m-) with a head loss of
1.45 v^/2^^ is located in the channel, The nonnal headloss in the siphoji tor j2 ~ 7 nrVs
amoimts 0.35 m. The headloss increases during flood {Q = 7.83 m^/s) to 0.43 m, thus
an increase of 0.08 m. The water level downstream of the spillway increases in this
case toy\, = 1.72 m for Q =^ 7.83 m.
A spillway length of L = 3 6 m can be calculated, similar to the example shown
above (steps of
- 2 m).
Flood screen. The construction of a flood screen downstream of the spillway which
obstructs specially the higher discharges, can reduce the spillway length once more,
Assume in the above example that the headlosses at the flood screen increases
with 0.1 m during the discharge of = 7.83 m. it means that
^ 1.82 m. A
spillway length of L^18 in can he calculated.
139
7. D E S I G N O F P A S S I V E R E G U L A T O R S
Hydraulic performance. The hydraulic performance of a side-channel spillway can best be

understood by considering the energy line, see figure 7.14. There is sub-critical flow in the
canal, thus the tailwater (downstream) conditions in the canal determine the simation at the
spillway. The tailwater and the energy line in the downstream canal reach follows from the
uniform flow and can be calculated through the Strickler formula. The energy line in the
downstream canal is sloping down because of the friction losses.
The energy line in the canal at the spillway location equals the energy line of the
downstream canal reach. The spillway length L is relative short, and the headloss due to
friction can be ignored here, and can be taken as horizontal.
The discharge at the upstream canal reach is much larger than in the downstream reach,
because of the outflow. The energy line is the same, so the velocity head v^llg makes that
the upstream water level is lower than the downstream water level.
The velocity v in the upstream canal reach is also much higher than in the downstream
reach, so the friction losses in the upstream canal reach are higher than in the downstream
reach. Therefore, the upstream energy gradient s is steeper than in the downstream reach.
Figure 7.14. Backwater curves by a side-channel spillway.
Flood screen. The length L of the side-chaimel spillway can be reduced considerably by
applying a 'flood-screen' ("bandjir-scherm"), see figure 7.15.
A flood-screen is a beam above the target water level in a canal, with the function to
create an 'orifice flow' ("onderspuier") with headloss for discharges higher than the design
discharge (Eijsvoogel 1932). It can be constructed as a simple beam of concrete, or it can be
constmcted in combination with a bridge. The effect of a flood-screen on the design of the
side-channel spillway is also presented in box 7.2.
FLOOD
W I T H SET UP O F W A T E R LEVELS
W I T H FLOOD SCREEN
Figure 7.15. Side-channel spillway with a "flood screen".
140
7.3.3. Siphon spillway

General. 'Siphon spillways' ("hevels") are also passive water level regulators as they prevent
the water level to raise above the target water level ("streejpeil"), see figure 7.16.
Small excesses of water levels will initially cause the siphon to act as a weir. I f the
water level continues to rise, the discharge through the siphon will attain a velocity sufficient
to 'prime' ("aan de gang brengen") the siphon. Pruning is initiated as the water flow carries
an air-water mixmre out of the barrel ("koker"). A full prime occurs when all the air is
removed from the barrel and pressure over the crest becomes sub-atmospheric.
A flow deflector is usually constructed in the downstream leg of the siphon. The flow
deflector will initiate the pruning at a low discharge by deflecting the flow to the top-side of
the barrel. The mrbulence thus created and the rnipact of the flow will mix the air in the
barrel with the outflowing water.
Once fully primed, siphonic action continues until the water level drops below the inlet
elevation or below the inlet of the siphon-breaker, and allowing an inflow of air to restore
atmospheric pressure at the crest, ending the flow abruptly.
Hydraulics. The capacity of the siphon follows from the equation, see also figure 7.16:
Q = \iA sj 2g z , with z <
where: Q is the discharge in m^^/s, /x is a coefficient, A is the cross-sectional area of the
barrel in m^, z is the headloss over the siphon in m,
= 10 m is the atmospheric pressure,
g is the acceleration of gravity g = 9.8 m/s^ = 9.8 N/kg. The coefficient n depends on the
energy losses due to bends, friction, the entrance and the exit. For a first estimate, fx = 0.1
can often be used, thus z =
l.Ov^llg.
on breaker
WEIR F L O W I N SIPHON
i Siphon breaker
P R I M I N G O F SIPHON
i Siphon breaker
DESIGN F L O W I N SIPHON
Figure 7.16. Types of flow in a siphon spillway.
Application. Spillway siphons are very effective in rapidly removing a large volume of water
from a canal, but may be more expensive to construct than other escapes. A distinct
advantage is its ability to discharge a large volume of water with a small rise in water
surface.
A disadvantage to the use of a siphon spillway in a small canal is the abmpt starting
and stopping of the discharge, which may produce surges or bore waves. It may also cause
erosion in the downstream canal.
141
7.4. Baffle Distributors

7.4.1. Working Principle
General. Baffle distributors, also called 'Neyrtec distributors' or 'orifice modules', are
designed to supply constant discharges that are independent of the upstream water levels,
see figure 7.17. These constant discharges are obtained by its hydraulic function, and not by
moving parts. The baffle distributor can be used for an on/off operation by opening or
closing the slide plate or 'shutter'.
GATE CLOSED
G A T E OPEN: C O N S T A N T D I S C H A R G E
Figure 7.17. Baffle distributor for discharge regulation and measurement.
Application. Baffle ("scherm") distributors were initially applied by French engineers only.
At present, baffle distributors are world-wide applied and are an international standard for
constant discharge control at offtakes in irrigation systems.
The advantages of the baffle distributor is its shnple operation. A l l that is required for
discharge regulation is to open or close the different slide gates or 'shutters' ("luiken") of the
compartments. The discharge regulation is independent of upstream water level variations.
The main disadvantage of the baffle distributor is its cost: the strucmre is expensive,
especially when it has to be hnported from France with 'hard' currency.
Furthermore, the headloss through the baffle distributor is larger than the combined
headloss of a vertical gate for discharge regulation with a broad-crested weir for
measurement.
Working principle. The baffle distributor consists of a weir, and one or two flxed metallic
"baffle", see figure 7.18.
At low upstream levels, the strucmre operates as a sharp-crested weir with overflow
("overlaat"). When the upstream water level rises, the water level above the weir reaches the
142
bottom edge of the baffle, and the stmcture operates as an orifice with underflow
("onderspuier"). The contraction of the jet becomes more pronounced as the upstream water
level increases, thus reducing the discharge coefficient C-Q and maintaining an almost constant
discharge. It should be noted that the baffle is not located above the crest but above the
sloping face.
OVERFLOW
UNDERFLOW
Figure 7.18. Working principle of the single-baffle distributor.
Double-baffle distributor. Essentially two types of distributors are available, the single-baffle
and the double-baffle distributors.
The double-baffle distributor has two baffles: a low and high baffle, see also figure
7.19. The function of the low baffle is shnilar to the baffle of the single-baffle distributor,
and forces an orifice flow for the higher water levels.
As the upstream water level rises further, the low upstream baffle is overtopped and the
second baffle comes into action to Ihnit the flow. Thus, the upstream low baffle has the dual
function of (i) contracting the jet at low heads and (ii) acting as a weir at high heads.
Water passing over the low baffle causes an additional contraction of the jet. As a
result, the double-baffle distributor maintains an almost constant discharge over a considerable
variation in upstream water levels.
OVERFLOW
UNDERFLOW
Figure 7.19. Working principle of the double-baffle distributor.
UNDERFLOW
143
7.4.2. Hydraulics of Baffle Distributors

Hydraulics. The baffle distributor acts as a short-crested weir with rectangular control
section, when the water surface is not in contact with the baffle(s). The discharge equation
for such a weir reads:
Q = 1.S5 bH^-^
When the discharge approximates the design discharge, the water surface touches the baffle
and orifice flow conunences. The free flow discharge through the orifice equals:
Q
= C^bw
; 2gy,
, with:
C,
1 + )i
i
where Q is the discharge in vc?ls, b is the width of the gate in m,
is the discharge
coefficient, w is the vertical distance between the baffle and the weir in m, Ji is the upstream
water depth above the sill in m, /x is the contraction coefficient of the jet and depends on the
relation y j / w , and g = 9.8 nxls^ = 9.8 N/kg is the gravity acceleration.
The headloss z through the baffle distributor depends on the orifice flow:
Z > 0.4 yîn + V with: V = yâx " }'min
where V is the variation is upstream water level between the minimum yîn and the maximum
water level y^^^.
The design discharge without any deviation is reached at the 'nominal' water level.
Usually, a flucmation of 5% from the design discharge is allowed.
The rating curve of the single- and the double-baffle distributors are presented in figure
7.20 and figure 7.21. The permissible water levels are presented in table 7.1. A n example
of the design is presented in box 7.3.
4
150
20 20
58
31
water depth
above sill
in cm
92
90:
30i
55
235
ia5
27
14ffl
54,5
486
8 ^
50 50
80-H
1^
7a_
125
75:
permissable
water level
variations
25
45
TO]
15
1UI
65:
40
U5
13
20
215
39,5
20
_37
'-^
59
9(fl
XX,
(CC,
Type of Baffle Distributor
- 10 % - 5 %
+ 5 % + 10 %
Deviation of Discharge ( i n %)
Figure 7.20. Depth-discharge curve of the single baffle distributor.
Flow
144
4
water depth
I225 above sill
permissable
water level
variations
- 10 % - 5%
Type of Baffle Distributor
5 % * 10 %
Deviation of Discharge (in %)
Figure 7.21. Depth-discharge curve of the double baffle distributor.
145
7. D E S I G N OF PASSIVE R E G U L A T O R S
7.4.3. Design of Baffle Distributors

Series of distributors. Five different series of baffle distributors are available. They are
identified in terms of design discharge per unit width, see also table 7.1:
- series X : 10 1/s per 0.10 m width,
- series X X : 20 1/s per 0.10 m width,
- series L
: 50 1/s per 0.10 m width,
- series C : 100 1/s per 0.10 m width,
- series CC : 200 1/s per 0.10 m width.
The series can be given an index ' 1 ' or ' 2 ' , denoting single or double-baffle type,
respectively.
An offtake with baffle distributors is identified by the series symbols, followed by the
design flow. So, an offtake with a distributor Li-2250 has a design discharge of 2250 1/s, and
a single-baffle distributor series L is used of a total width b = 2250 / 500 = 4.50 m.
Table 7.1. Permissible Water Level of Baffle Distributors.
Baffle Distributor
Baffle Distributor
in series with AVIS/AVIO gate
range in
upstream
water level
in m
minimum
headloss
in m
range in
upstream
water level
in m
minimum
headloss
in m
0.04
0.06
0.11
0.16
0.27
0.07
0.12
0.22
0.35
0.60
Type
Flow
per 1 m
in vc?ls
XI
XXI
LI
Cl
CCl
0.100
0.200
0.500
1.000
2.000
0.05
0.08
0.15
0.23
0.37
0.05
0.08
0.15
0.24
0.38
X2
XX2
L2
C2
CC2
0.100
0.200
0.500
1.000
2.000
0.14
0.22
0.42
0.67
1.06
0.05
0.08
0.15
0.24
0.38
do
do
do
do
do
not
not
not
not
not
use!
use!
use!
use!
use!
Box 7.3. Example of a design of an offtake with a baffle distributor.

A n offtake has to be designed for a constant discharge o l 0.25 m-'/s and al a point in
a canai where the levels vary from 23.1$ to 23.12 m"^. The water level in the offtaking
canal, downstream of the offtake, will be always lower than 23.02 m"^.
An offtake with X X I distributors of total width of 1.25 m wide suitable for the
discharge of 0.25 m^/s. The lower and upper limit of the upstream water level of the
X X I dis,tributor can be read in table 7.1 and values 0.22 to 0.30 m above the crest,
while the d(wnstream water level should be tower than 0.14 m below the crest.
The sill crest oi the distributor can be designed at 23.1:^ - 0.22 - 27.90 m"^.
Then, the maximum up.stream water level would be 22.90 + 0.30 = 2.3.20 m + . while
the maximum downstream water would be 22.90 + 0.14 ~ 23.tN m ' .
146
Regulation of the discharge. The discharge through a baffle distributor is regulated by

opening or closing the shutter. The shutter ("scherm") can be locked in place, either fully
opened or fully closed, see figure 7.22.
The X and the XX-series distributors are fitted with flat-slide gates. They are suitable
for individual offtakes.
The L , L L and the C-series are fitted with sector-type gates, so as to reduce the
operating forces by the high water forces. The CC-series are controlled by independent
vertical gates. This equipment is prhnarily intended for major offtakes.
Compartments of Distributors. Since the discharge of one baffle distributor can only be
operated by the shutter on an on/off basis, the only way to vary the discharge is to combine
several baffle distributors of different widths within one offtake by creating 'compartments',
see also figure 7.22.
The discharge is set to any required value by opening or closing a combination of
different sized shutters of the compartments. It is not advisable to combine compartments of
different series (e.g. X X and C) in one offtake.
Figure 7.22. Discharge regulation with the (single) baffle distributor.
8. DESIGN OF MANUAL REGULATORS

8.1. Overflow Gates
8.1.1. Drop-Leaf Gates
General. The 'drop-leaf gate' ("klepstuw") pivots along an underwater axis, and is of the
overflow type, see figure 8.1. The load on the operating mechanism comes mainly from the
hydrodynamic thrust on the leaf and is quite complex. Drop-leaf gates are widely used in the
Netherlands.
Hydraulics. The rating curve of the drop-leaf gate is not available for practical use, as it
depends on too many parameters, such as the shape of the rounded crest, the angle of opening
and the submerged ratio. In fact, the knowledge on the precise rating curve ("Q-h kromme")
is not very hnportant as the operator regulates the gate on the basis of water levels and not
on discharges.
Design. The hydraulic design of the structure is also not based on the rating curve of the
drop-leaf gate. The design discharge has to pass over a fully lowered gate, which means that
the conveyance formula ("vernauwingsformule"),
z = a v^/2g, is valid for submerged
conditions, and the weir formula ("overlaatformule"),
Q = 1.1 b ifi''^, for free flow
conditions.
GATE AT A LOW DISCHARGE
GATE AT THE DESIGN DISCHARGE
Figure 8.1. Drop-leaf gate for manual regulation.

147
148
8.1.2. Vertical Gate with Overflow

General. Vertical gates can also be manufactured as overflow gates, see figure 8.2. However,
structural provisions have to be made to allow the gate-leaf to be lowered, either by applying
a second gate or by designing a drop in the canal aligimient.
Hvdraulics. The hydraulic design equation of a vertical-gate-with-overflow equals the rating
curve of the sharp-crested weir. However, the struchire is not standardized. A discharge
coefficient c = 1.9 m^^'^/s is recommended for a first estimate. Thus:
Q = 1.9 bH^ '
where Q is the discharge in w?/s, H is the energy head upstream in m, b is the width of the
crest in m.
The tailwater level must be kept quite low to allow the aeration of the nappe. This is
satisfied for a tail water level of at least 0.05 m below the crest level. It means that the
headloss z over the structure has to meet z > H.
Application. The vertical gate with overflow is simple in reading but requires considerable
headloss. Moreover, the rating curve is not accurate. The structure is not recommended for
discharge measurement. As such, the vertical gate with overflow is often not the best solution
for an offtake strucmre.
GATE FLOW AT Q<
OIAX
DESIGN A T M A X . DISCHARGE Q
Figure 8.2. Vertical gate with "overflow" for discharge regulation.
8.1.3. Romijn Weir

General. The Romijn weir has been developed in Indonesia during the 1930s for use in
relative flat regions and for adjustable supply to the tertiary irrigation units (Bos 1989, Kraatz
1975). The Romijn weir has also been applied in the Netherlands and many other countries
where Dutch engineers have worked. The strucmre is also called the "Hobrad weir", being
the abbreviation of horizontal broad-crested adjustable weir (STOWA 1994).
The Romijn weir is basically a broad-crested weir, see figure 8.3. The broad-crested
weir is mounted in a steel frame, and can be moved up- and downwards to decrease or
increase the discharge. A (bottom) gate is optionally added in the steel guide-frame for
flushing sedmient that has been deposited in front of the gate.
149
8. D E S I G N O F M A N U A L R E G U L A T O R S
Hydraulics. The hydraulic design equation of a Romijn weir equals the equation of the
broad-crested weir with a rectangular control section {Q = \.1 b H^'^). It means also that
the headloss z oyer the strucmre has to satisfy z ^ Vs H.
The maximum head H oyer the weir table depends on the length L of the table as
straight flow-lines aboye the table has to be guaranteed for the broad-crested weir flow. In
practice, it means that H < 0.30 m for the standard length of L = 0.30 m.
Consequently, standard widths b of the Romijn weir of b = 0.50 m, 0.75 m, 1.00 m
and 1.25 m will provide for design discharges upto Q = 0.30 m^/s, 0.45 m^/s, 0.60 w?/s
and 0.75 m^/s, respectively.
cm
gauge
counter
gauge
counter
gauge
liter
gauge
H+V
^////////////////A
G A T E CLOSED
GATE A T T H E M A X I M U M DISCHARGE
Figure 8.3. Hydraulic parameters of Romijn weir.
Design. The hydraulic design of the strucmre is based on a known design discharge Q. The
above standard widths b will determine the required numbers of bays and the total width of
the strucmre.
Furthermore, the design should also incorporate any variations V in upstream water
levels at which the design discharge Q has to be supplied, see also figure 8.3. The variation
y may originate from the "decrement" Ah of the water level regulator, and/or from backwater
effects during non-uniform flow. The variation V might be small at offtakes in combination
with water level regulators, but might become considerable for "free" offtakes with water
level regulators at some distance. Thus, the sill level of the offtake has to be designed at
2 X
+ V) below the maxhnum upstream water level. The resulting headloss z over the
structure is z > V3 i / + V.
Discharge measurement. The discharge measurement and the weir-setting is shnple, although
it needs three gauges, see also figure 8.3:
a water level gauge in the canal, the counter-gauge, which shows a reading in centhnetres;
another fixed gauge on the frame, the centimetre-gauge, which is identical to the above
counter-gauge;
a gauge that moves with the weir, the litre-gauge, with a logarithmic scale of the i/-values
according to the weir formula: H =(Q I 1.7 bf-^''.
150
Application. The Romijn weir has a good sediment and floating debris capacity, is easy in
operation and reading. It has a double function of both for regulation and for discharge
measurement. However, the structure with gate is quite expensive and is not widely
understood in many countries.
8.2. Underflow Gates

8.2.1. Vertical Gate for Submerged Underflow
General. The 'vertical slide gate' ("vlakke schuiven") consists of a simple rectangular gateleaf ("blad") that is moved up- and downwards in grooves ("sponningen"), see figure 8.4.
Slide gates are used only for the smaller openings (b < +3 m) in strucmres and to a
maximum of 3 m height, as the hoisting device ("windwerk") becomes otherwise too heavy
to overcome the friction in the grooves.
The advantages of the vertical gate with underflow as discharge or as water level
regulator are: the stmcture is relative cheap, the stmcmre is simple and smrdy, it provides
a reasonable constant discharge for varying upstream water levels in the parent canal.
Moreover, sediment can be transported through the stmcmre.
The disadvantages of regulators with vertical gates are: the gate does not pass floating
debris, and the discharge caimot be measured accurately.
Design. The design of a stmcmre with a vertical gate for water level regulation, is normally
based on the criterium: the headloss during the design discharge ioo% should be low. Thus,
the vertical gate should be under submerged ("verdonken") conditions. A higher headloss
would make that the design water level in the upstream chaimel reach is rising with the same
amount.
The minhnum headloss z of the stmcmre will occur not only at submerged flow, but
also when the vertical gate is lifted out of the water, see also figure 8.4. So, the flow through
the stmcmre is not a 'gate flow' anymore, but becomes a 'conveyance flow' like at a bridge
or at a culvert.
<p
GATE F L O W AT Q< Q^^^
cp
Figure 8.4. Design of a vertical gate under 'submerged underflow'.

Conveyance flow. The gated discharge regulator under (submerged) conveyance flow
("vernauwingsstroming") follows the rating curve ("Q-h kromme") of conveyance stmcmres,
such as bridges, aqueducts.
151
The headloss z in these structures reads:

Q
and: a
with: V
a:_ + a out
Ig
yb
with the entrance coefficient ot-^^ ( ~ Vs), the exit coefficient 0;^^^ ( ~ % ) , the velocity v within
the structure in m/s, the discharge Q in m^/s, the (average) water depth y within the
stmcture, and the width b of the stmcmre. Thus, the rating curve becomes:
=
^
by
{ 2 ^
a
The conditions for the conveyance flow are, see also figure 8.4:
the gate is lifted out of the water, so w > y;
the tailwater level is quite high: the headloss z < % H.
Q
8.2.2. Vertical Gate for Free Underflow

Design. A regulator with a vertical gate can also be applied at a location with a high headloss
z during the design discharge <2ioo%' ^^^^
happen in a chaimel that follows a steep
alignment ("tracee") and where drop stmcmres ("stortdammen") are required to avoid
scouring of the channel bed.
Often, the design of discharge regulators, and also water level regulators, should meet
the criterium of low costs of gate and stmcmre. This is achieved when the water depth y in
the stmcmre and the width b of the stmcmre are at a minimum. Both can be met by designing
a stmctore where the gate can be lifted out of the water during the design discharge. This
means that the free underflow ("volkomen onderspuier"), Q H^^^, of the gated discharge
regulator is transformed into a free overflow ("volkomen overlaat"), Q Z/^''^, see figure
8.5.
)/////////////^
GATE FLOW A T Q< Q^^x
Figure 8.5. Design of a vertical gate under 'free underflow'.
Free overflow. The regulator with a vertieal gate is designed with the gate lifted out of the
water. For a high headloss, the flow becommes a free overflow ("volkomen overlaat"). The
corresponding the rating curve ("Q-h kromme") is either:
of a broad-crested weir, when there is a sill ("drempel"), with the discharge formula:
152
Q = l.l
bW^,
of a control notch, when there is no sill, with the discharge formula:
Q =
l.SbH^^^,
where Q is the discharge in m^/s, b is the width of the struchire in m, and H is the
(upstream) energy head in m.
The conditions for free overflow are, see also figure 8.5:
the gate opening w is sufficient high, i.e. w > %H, in order that the water does not touch
the gate;
the 'tailwater depth' ("achterwater diepte") is sufficient low, i.e. the headloss z > % H,
in order to maintain a free flow.
8.2.3. Radial
gate
General. 'Roller gates' ("roldeureti") are essentially equal to the vertical slide gates, but they
have rollers or wheels in the groove to limit the friction, see figure 8.6.
'Radial gates' ("segmentdeureti"), also called 'Tainter gates' or 'Segment gates', are
used as underflow gates as an alternative for the vertical gates, see also figure 8.6. The
choice of a radial gate above a vertical gate may depend on a variety of factors and each type
has its own advantage.
Radial gates are not often used for small discharges. However, they have the advantage
over sliding gates that the lifting force is only required against the weight of the gates, as all
thrust on the gate leaf due to water pressure is passing through the axis. Moreover, lifting
force is almost constant for all gate openings. Counterweights in order to counter balance the
self weight, may also be used to further reduce the lifting force.
Hence, radial gates can be used with smaller lifting force for all heads than with the
sliding gates. Even hand-operated hoisting mechanism ("windwerken") may suffice for smaller
works, whereas the vertical lift gates of the same size might need a power mechanism.
The radial gate is often more costly than vertical gates. The rating curve of a radial gate
is even more complex than that of the vertical gate, which make both less suitable for
discharge measurement.
hoisting device
SLIDE GATE
hoisting device
ROLLER GATE
Figure 8.6. Types of vertical and radial gates.
hoisting device
RADIAL GATE
153
8.2.4. Crump-deGruyter Gate

General. The Crump-deGruyter gate is a 'streamlined' vertical gate, see figure 8.7 (Bos 1989,
Kraatz 1975). The gate is equipped with a flow-streamliner which avoids the curvation of
the streamlines and the contraction of flow under the gate. Thus, the contraction coefficient
ix = 1.0.
Flow-streamliner. The diameter d of the flow-streamliner must satisfy d > 2 w for a
contraction coefficient p ~ I.
Otherwise, the formula for the contraction coefficient IJL depends on the gate opening w
(STOWA 1994):
\2
II
0.51 + 0.1
Staff
Gauge
23 -
GATE A T A L O W DISCHARGE
^
- 4.7
w
Staff
Gauge
C3
G A T E A T T H E DESIGN D I S C H A R G E
Figure 8.7. Crump-deGruyter gate, for manual regulation.
Hvdraulics. The hydraulic design equation of a Cramp-deGruyter gate under free flow
conditions is basically the orifice formula of the vertical gate, see also figure 8.7:
Q = ^ b w sj2g{H- w)
= 4.43 \i b w ^JH-W
So, the discharge formula of the Crump-deGrayter gate under free flow and for / i = 1
becomes:
Q = 4.43 b w
^H-w
where Q is the discharge in m^/s, H is the energy head above the sill in m, b is the width of
the throat in m, w is the gate opening in m.
The conditions for the Cmmp-deGruyter gate are:
the contraction coefficient n 1 for w < 0.5 d, where d is the diameter of the flowstreamliner;
the free flow condition is satisfied for a headloss z > 0.3 H;
the gate opening w <% H to allow for underflow.
The maximum discharge depends on maximum gate opening w < % H. Hence:
Q^^^ = 4.43 b % H (H - %Wp-^ = 1.11b H^'^.
which appears the overflow formula.
154
Application. The Crump-deGmyter gate is an improvement of the vertical gate, as the rating
curve is more simple. The Cmmp-deGmyter gate is recommended when an underflow
regulator with measurement provisions is needed. Its hydraulic functions are good and the
stmcture is reasonable in costs.
The disadvantages of the Cmmp-deGmyter gate are: the headloss is relative large, the
discharge measurements needs two readings: upstream water level and gate opening. The
stmcmre faces problems with floating debris.
8.3. Hydraulics of Underflow Gates

8.3.1. Types of Flow
Types of flow. The flow through a gated discharge regulator is quite complex, and the rating
curve is not straight forward. Basically, there are five different flow types through a stmcmre
with a gated discharge regulator, see figure 8.8 and figure 8.9:
free overflow. The water level or 'nappe' ("straal") does not touch the gate, and the
tailwater level is sufficient low to allow for a free overflow, like the flow over a broadcrested weir.
conveyance flow. The water level does not touch the gate, and the tailwater level is so
high that the flow is submerged, like the flow below a bridge.
free underflow. The flow under the gate is not influenced by the tailwater level. The
outflowing jet is open to the atmosphere because the hydraulic jump is formed at some
distance from the gate. The rating curve will be discussed below.
partially-submerged underflow. The flow under the gate is super-critical (Fr > 1).
However, the flow is influenced by the tailwater level as the jet is overlaid by the
hydraulic jump. The rating curve will be discussed below.
fully-submerged underflow. The flow under the gate is sub-critical (Fr < 1), and as
such, is influenced by the tailwater level. Also this rating curve will be discussed below.
w ~ % yi
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
FREE
"WEIR" FLOW
Fr > 1
SUBMERGED
"CONVEYANCE" FLOW
Fr < 1
Figure 8.8. Types of "weir flow" through a stmcture with a vertical gate.
155
4 yi
FREE
ORIFICE FLOW
Fr > 1
SUBMERGED
ORIFICE FLOW
Fr > 1
SUBMERGED
ORIFICE FLOW
Fr < 1
Figure 8.9. Types of "orifice flow" through a structure with a vertical gate.
8.3.2. Hydraulics of the vertical gate under free flow

Bernoulli equation for free flow. The Bernoulli equation for a vertical gate under free flow,
can be written, see also figure 8.9:
H
y,^
2gy:
2gy;
hence for y2 = /x w the equation of the rating curve reads:
q = ^jlgy^{H-y^)
= li w ^/ 2g (H - \iw)
where q is the discharge per unit width in m^/m.s, y2 is the water depth of the jet above the
sill in m,
is the energy head above the sill in m, w is the opening of the vertical gate in
m, fi ~ 0.6 is the contraction coefficient of the jet, and g = 9.8 m/s^ = 9.8 N/kg is the
gravity acceleration.
General discharge equation. The above discharge equation is often not used as it cannot be
applied for the submerged flow. Therefore, the Bernoulli equation is processed in an
alternative way, and a more general discharge equation is obtained (Henderson 1966):
yi
-yi
2g
2g y2
2g y,
= \i w
2g
y y
^2
y y
\iw
\iw
1 +Hw
yi
156
Thus, the general vertical gate formula reads:

Q
= Cj^bw
{ W J i . with:
1 + II
^
where Q is the discharge in nr/s, b is the width of the gate in m, Cj^ is the discharge
coefficient, w is the opening of the vertical gate in m,
is the upstream water depth above
the sill in m, / i ~ 0.6 is the contraction coefficient of the jet, and g = 9.8 m/s^ = 9.8 N/kg
is the gravity acceleration.
Discharge coefficient. The discharge coefficient C^ for free flow can be calculated with the
above formula for the ratio y^/w. The calculated values for the higher ranges correspond
reasonable well with experimental data. However, the calculated values of Cjy for the
practical range y^/w < 4 do not correspond well with the experimental data, as the calculated
values are some 5% too high.
It is recommended to use the experhnental values of the discharge coefficient Cp as
presented in figure 8.10 (Chow 1959, Henderson 1966, Bos 1989).
157
Discharge coefficient for overflow. Two types of flow may occur when the gate opening w
~ 2/3 y^: (i) the underflow when the nappe touches the gate, and (ii) the overflow when the
nappe does not touch the gate. Also the overflow can be calculated by the vertical gate
formula, but only for w = % y^!
The discharge coefficient
follows from the actoal discharge according to the (broadcrested) weir formula and the gate formula for w = % y^:
q = IJ H^l^ ~ 1.7 yl'\
Weir formula
and together:
1-7 y i
3/2
Vsy^ ^ 2gy,
Orifice formula : q =
and so:
1.7
CD
^ 2g y^
0.58
Validity for free flow. Free flow through a vertical gate is guaranteed when the hydraulic
jump does not distorb the jet. So, the tailwater depth yg has to be lower than the "conjugate
depth", i.e. the tailwater depth needed to form a hydraulic jump with the water depth y2 of
the jet. Thus:
and
1+8/^^2^ - 1
y.
The Froude number Fr2 of the jet with depth y2:
K
y.,
= \i w
yi
2gy,
3
CD
\
\
syi
Hence, the hydraulic jump will not occur for a ratio between the tailwater depth y^ and the
gate opening w:
I ,
\
y^
ii
1 + 16 C 2 J Zi
1
-0
2
w
The above equation is confirmed by the experhnental results of figure 8.10 for y^Jw < 10.
The calculated data are presented in table 8.1.
Table 8.1. Parameters of the vertical gate under free flow.

Upstream
Water Depth
yj/w
Discharge
Coefficient
Maximum
Tailwater
yg/w
Minimum
Headloss
z/w
Minimum
Rel. Headloss
ziyi
(1.5 Overflow)
1.5
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
(0.58)
0.50
0.51
0.52
0.53
0.55
0.56
0.56
0.57
0.57
0.58
(1.00)
1.31
1.59
2.04
2.45
2.89
3.25
3.54
3.87
4.13
4.45
(0.50)
0.19
0.41
0.96
1.55
2.11
2.75
3.46
4.13
4.87
5.55
(+0.33)
0.13
0.21
0.32
0.39
0.42
0.46
0.49
0.52
0.54
0.56
158
The above equation can be simplified by assuming /x = 0.6 and Cp = 0.55, so that:
3^3
0.55^
0.3
w
0.6^
Hence, a vertical gate will flow under free flow for:
0.3
1 + 16
1 + 22.4
0.3
\0.5
1.42
0.3
yi
0.3
Free flow. The maximum tailwater depth y^/w from table 8.1 has been plotted against the
upstream water depth y^/w, see figure 8.11.
In practical terms, it means that free underflow at a vertical gate occurs when the
tailwater level yg is below the critical value:
3^3
0.37
yi
+ 0.75
y^ < 0.37 y j + 0.75 w
or
w
w
This can be rewritten as a criterium for the headloss z:
Z = y^-y^
so: z > yi- 0.37 yi - 0.75 w = 0.63 y^ - 0.75 w.
As the gate opening w is limited in practice by 0 . 1 5 y i < w < 0.5 y^, the minhnum
headloss z^[j^ of a vertical gate under free flow becomes:
y^yi < ^min < ' / ^ y i .
where the headloss zî^ ^
J'l is required for the larger gate openings w = 0.5 y^.
Figure 8.11. Conditions for the vertical gate under free flow.
159
Vertical gate for discharge regulation and measurement. The use of a vertical gate for
discharge regulation and discharge measurement, under free flow conditions, is shown in
figure 8.12. The design of the struchire is based on an upstream water level yi and a design
gate opening w = 0.5 y^. A larger design gate opening would create the danger that the
underflow would be transformed into an overflow at w = 0.61 yi.
The range for discharge measurement would be set at 025% ^ Qm%calculated
in box 8.1 that the gate opening for 025% would be at w = 0.11 y^.
The required headloss over the vertical gate for accurate measurement for the above
range would be z > 0.56 y j .
The maximum discharge for underflow would amount to 130% when the gate is lifted
to w = 0.67 y^. The discharge jumps to 150% when the nappe does not touch the gate and
overflow starts. The rating curve is presented in figure 8.13.
ZMIN ~
0.13
0.33
y,
ZMIN ~
0.21
y,
y,
0.39
'MIN
0.56
y,
y,
r
yi
yi
w=
0.67
0.67
y,
w=
yi
w=
y,
0.25
y,
w=
0.50
w=
0.11
y;
y,
7777/
150%
QDESIGN
(weir flow)
130%
QDESIGN
(orifice flow)
100% x QDESIGN
(DESIGN)
50% x QDESIGN
25% x
Q^^^^^
Figure 8.12. Discharge and minimum headloss of the vertical gate.
Box 8.1. Example calculation of the vertical gate with free underflow.
A gate with opening w - 0.5 y-j (y^'w = 2.0) requires a minimum headloss ol z =
0.21 V| for free Ilow. The corresponding discharge q amounts to <7
H' (Igy^f^-^
~ 0.51 X
X y^.'2.0 {Igy^f^
- 0.255 xb x y i ' ( 2 g y j ) ^ \ and can be set O - 100%.
Similarly, the discharge for tv = 0.25 j - ^ (>',/ = 4.0) can be calculated at Q =
0.132 X h X Vi {2gy^f-^, which is O 50?c. at a minimum headloss of z *= 0.39y^
BlHlililllii^BiilH
Also, the discharge for ti' = 0.11 y , (yîw
9.0) can be calculated at - 0.063
X b X y j ( 2 ( r j , y ' - \ which is 0 ~ 25%, at a minimum headloss o f z ~ 0.54 y j for free
II^HIililliiilH^
The maximum flow under the gate occurs for w =
(y^/w = 1.5), when the
discharge reaches Q - 0.333 x b x Vj (2g>,)^'"^ which is Q 130%, at a minimum
headloss of i - 0.13 y, for tree flow.
However, the maximum flow thnmgh the .structure will occur when the gate is
opened at w > % y | (VjAv " 1.5) and the gate (low is transformed into a weir flower = 0.387 X /; X V, ( 2 i , ' y i ) " \ which is 0 150%, at a mimmum headloss of r =
ii||iiiilfciiiiiBiM^
160
J l
\
\
r
Water depth
above the sill
L m m
Gate openmg
w = 0.6 m
Gate opening
w = 0.4 m
Gate opening
w = 0.2 m
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7
Discharge in m^s
Figure 8.13. Rating curves of the vertical gate under free flow.
8.3.3. Hydraulics of the vertical gate under submerged flow

Bernoulli equation for submerged flow. The Bernoulli equation for the submerged flow under
a vertical gate depends on the total water depth y instead of water depth of the jet
So the
Bernoulli equation ean be written as:
2g y{
hence for y2 = iiyv follows:
2
2g y2
yx-y
yi
161
Discharge coefficient formula. As the general discharge equation reads:

= Cp^
the value of Cp can be calculated out of the formula for a known value of y:
7
2 2
w'y^
2g
2g
w^lgy^,
-y
yi
yx
w^y,
yi
Momenmm equation. The value of the water depth y can be calculated by considering the
momenmm equation:
^3
y
8y2
2
gy,
2
There are two equations (Bernoulli equation and momenmm equation), one condition (y2 =
(JL w), and three known parameters (y^, yg and w), and three unknown parameters (q, y^ and
y). So the unknown value of y can be calculated by the substimtion of the term
I f .
Discharge coefficient. The analytic solution of the value y appears to be very cumbersome.
Furthermore, the calculated value of the discharge coefficient Cp differs from the
experimental value.
Therefore, it is reconunended to read the value of the discharge coefficient Cp, also for
submerged conditions, from the graphs in figure 8.10 (or figure 8.14).
Froude number. The Froude number of the jet under submerged conditions equals the above
formula for the Froude number under free flow conditions.
^1
^^2
thus:
0.33 Fr,
yx
yx
The Froude numbers depends on the ratio y^/w and the value of Cp, and has been plotted in
figure 8.14. It can be seen that the submerged conditions occur for both Froude numbers Fr
< 1, as well as for Froude numbers Fr > \.
Accuracy for submerged flow. It is obvious that the reading of the discharge coefficient Cp
for submerged flow is not very accurate. Especially, the determination of the Cp-value for
submerged flow may resuh in considerable errors i f the head differences z between upstream
water level y^ and the tailwater level yg become small. Hence, the calculation of the discharge
of a vertical gate under submerged condhions will be of a low accuracy, see box 8.2.
162
Figure 8.14. Froude numbers of flow under vertical gate.
Box 8.2. Example calculation of the vertical gate under submerged underflow.
The vertical gate of h = 1.00 m wide of the above example can al.so flow under
submerged conditions. Assume a gate opening of w = 0.20 m. The upstream water
level vj'is kept constant by a water level regulator at
= 0.80 m.
The discharge under free flow depends on the ratio VjAv = 4.0 and the discharge
coefficient
- 0.53. and ean be calculated at amounts to <2f,ce tlow " ^ D ^
{2gy^f-^ = 0,53 X I X 0,20 x (2g x O.HO)*'""* ^ 0.42 m^/s. This discharge is valid for
tailwater water levels V:, < 2.45 x w ^ 2.45 x 0.20 ^ 0.49 m.
Higher tailwater 'levels y^ lead to submerged flow and a reduction of the
discharge, following the equation: submcigcd How =
^
( ^ y i f ^ =
x I x
0.20 X {2g x 0.80p'" = 0,79 C y . The value C^y can be read from the graphs in figure
8.10, and the discharges for increasing tailwater levels can be calculated:
y^ = 0.50 m: V-^.-'H- = 2.5 and forj-^/w = 4.0 is C^^ 0.50, thus Q = 0.40 nr/s.
- Y\ - 0.55 m: v^/w - 2.8 and for y j / i r - 4.0 is Cj^ 0.40, thus Q - 0.32 nr^'s,
\'3 = 0.60 m: yyiv
3.0 and i'or y^/w = 4.0 is
^- 0.37. thus Q = 0.29 mi's,
v'^ ^ 0.65 m: v'^'u3.3 and f o r ) ' j / H ' = 4.0 is Cj^
0.32, thus Q = 0.25 nrVs.
.
- 0.70 m: y\;M- - 3.5 and for VjAv - 4.0 is
0.:^5, thus Q = 0,20 m'Vs,
y^ - 0.75 m; v^fw - 3.8 and f o r y j A v = 4.0 is C p
0.15. thus = 0.12 nrVs,
yg ^ 0.80 m: vy'W
4.0 and for VjAv = l.O is
0.00, thus Q ^ 0.00 m7s.
163
8.3.4. Hydraulics of the radial gate

Radial gate and the vertical gate. The hydraulics of the radial gate are basically the same as
for the vertical gate. Thus, the general gate formula for the radial gate reads, see figure 8.15:
= C^bw
f 2 j y 7
where Q is the discharge in m^/s, b is the width of the gate in m, Cp is the discharge
coefficient, w is the opening of the radial gate in m, y^ is the upstream water depth above the
sill in m, /X ~ 0.6 is the contraction coefficient of the jet, and g = 9.8 m/s^ = 9.8 N/kg is
the gravity acceleration.
mmmmmmmmmmmm.
Figure 8.15. Hydraulics of the radial gate.
Contraction coefficient. The contraction coefficient \K of the radial gate is larger than the
coefficient of the vertical gate, as it is influenced by the angle (in ) of the gate-lip with the
horizontal plane. Thus:
0
e
90
90
The angle Q (in ) of the gate-lip with the horizontal plane can be calculated from the gate
opening w by simple goniometry, see figure 8.16. Thus:
1 - 0.75
+ 0.36
sm(90 - t t g - a ) ,
R
or: sm(90 - 0 )
Radial Gate under free flow. The discharge coefficient Cp for the radial gate under free flow
is basically equal to the equation of the vertical gate, see also figure 8.16:
CD
1 +
\
The values of the discharge coefficient Cp for the radial gate {alR = 0.9) under free flow
are presented in table 8.2.
164
Validitv of free flow. Free flow through a radial gate is guaranteed when the hydraulic jump
does not dishirb the jet. The actoal tailwater depth can be calculated by means of the Froude
number, as shown for the vertical gate. It may be expected that the tailwater depth of the
radial gate is more or less equal to the tailwater depth at the vertical gate. I n practical terms,
it would mean that a radial gate under free underflow requires a minimum headloss z^nin ~
0.37 yi + 0.75 w, but further analysis may be required.
Table 8.2. Discharge coefficient Cp for radial gates (a/R = 0.9).

Gate Opening
w/R -= 0.1
Upstream
Level
Jl
R
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
Gate Opening
w/R = 0.3
Gate Opening
w/R -= 0.5
Tailwater
Tailwater
Tailwater
Tailwater
Tailwater
Tailwater
^=0.7
R
^
R
^
R
Zl=0.5
R
^=0.7
R
0
0.33
0.44
0.51
0.58
0.64
0.69
0.73*
0.74*
0.74*
0.75*
* Free flow
0
0
0
0.28
0.39
0.44
0.48
0.53
0.57
0.59
0.62
= 0.5
0
0.50
0.59*
0.60*
0.61*
0.62*
0.62*
0.63*
0.63*
0.64*
0.64*
= 0.7
0
0
0
0.34
0.45
0.55
0.62*
0.63*
0.63*
0.64*
0.64*
not
available
0
0
0
0.46
0.53*
0.53*
0.54*
0.54*
0.54*
0.55*
0.55*
165
Radial Gates under submerged flow. The submerged flow through the radial gate can be
calculated by means of the Bernoulli and the momentom equations, and the parameter y as
for the vertical gate.
Modeltests are only available for radial gates of fl/i? = 0.1, fl/i? = 0.5 and a/i? = 0.9
(Chow 1959, Henderson 1966). The gate alR = 0.1 is not applicable in irrigation and
drainage engineering because of the large radius {R = 10 X a). Also the experiments on the
gate a/R = 0.5 are not applicable, as the tailwater depth is taken quite high, seefigure8.17.
Only the radial gate of a/R = 0.9 is of practical use. These discharge coefficients Cp are also
presented in table 8.2.
Gate
Openings:
Tailwater
Levels:
y^miy/^^^^
w/R = 0.1
y3/R = 0.5
w/R = 0.1
y3/R = 0.7
w/R = 0.1
CD Coefficients are less relevant
AXIS OF GATE AT a/R = 0.5
Tailwater
Levels:
^
Gate
Opeiungs
^^^^^^^^^^^^^
w/R = 0.1
y3/R = 0.5
w/R = 0.1
ya/R = 0.7
w/R = 0.1
Cj, Coefficients m Table
AXIS OF GATE AT a/R = 0.9
Figure 8.17. Modeltests on the radial gate to determine the
coefficients.
166
9. DESIGN OF HYDRO-MECHANICAL REGULATORS

9.1. Hydro-Mechanical Regulators for Drainage
9.1.1. Flap Gate
General. 'Flap gates' ("-duiker-klep?",
"uitwateringsduiker") are automatic gates, and
consist of a gate-leaf that pivot about a horizontal hinge. In drainage, flap gates are opened
by the water pressure, i.e. the higher tailwater levels will close the gate, see figure 9.1. It
is also possible that an operator opens and closes the gate manually. Flap gates as discharge
regulators are widely applied at drainage outfalls.
Hvdraulics. The rating curve ("Q-h kromme") of the flap gate is not available for practical
use, as it depends on too many parameters, such as the angle of flap gate, the weight of the
flap gate and the submerged ratio.
Design. The hydraulic design of the strucmre with a flap gate will be quite a problem as the
rating curve of the flap gate is not readily available. The design discharge has to pass through
the strucmre at the maximum gate angle, when the nappe has the maxhnum contraction
GATE CLOSED
GATE AT THE DESIGN DISCHARGE
Figure 9.1. Flap gate, for automatic regulation in drainage.
167
168
9.1.2. Tidal Sluice

General. Drainage water can be evacuated from a polder by gravity through a tidal sluice
("uitwateringssluis", "suatiesluis") in the dike during low tide, when a polder is surrounded
by open water i n which tides are active. These structures should be designed at the smallest
dimensions, to limit the costs, by aiming at:
overflow ("overlaat"), instead of underflow (orifice) flow,
free flow (volkomen stroming"), instead of submerged flow.
Operation mles. Drainage outlets are operated a sort of 'mixed control' that operates with two
'setpoints' ("streejpeil"): the inside setpoint at the inner side of the strucmre, and the outside
setpoint at the outer side, see figure 9.2. The operation rules are:
the gate is closed at an internal water level ("binnenwaterstand") below the inside setpoint;
the gate is open at an internal water level above the inside setpoint, and for an external
water level ("buitenwaterstand") below outside setpoint;
the gate is closed at an internal water level above the inside setpoint, but the external
water level is above the outside setpoint.
PLAN
CROSS S E C T I O N
Figure 9.2. Tidal sluice under combined up- and downstream control.
Discharge formula. The discharge formula of a sluice and a culvert, both with a free water
level, reads:
for free flow:
Q = 1.7 b H
for submerged flow:
Q -
by
valid for z > Vs H
b y \[z~ valid for z < Vs H
^ 2 j T = 4.4
a
where Q is the sluice discharge in m^/s, b is the width of the sluice in m, H is the upstream
energy head above the sill in m, y is the (outer) water level above the sill in m, z is the
(energy) headloss over the structure in m, and o; is a coefficient (a = 1.0).
Method of calculation. The design calculations of a tidal sluice have to be done in time-steps.
For each thne step, the outfall volume should be calculated for that specific tidal water level.
This outfall volume decreases the volume of water in the polder, so the open-water level in
the polder decreases. This new water level is the input for the next time-step, see figure 9.3.
The whole calculation can be made by means of a table, see table 9.1. An example is
presented in box 9 . 1 .
169
9. D E S I G N O F H Y D R O - M E C H A N I C A L REGULATORS
i.oH
Water levels
inm+MSL
i 4 8 io
Discharge Volume
in "3600 xb" m3/hr
inner water level

(drainage system)
12 1416
1618
1820
20 Z22 24 26
Time in hrs
Water levels
4 6 8 10 12 14 16 18 20 22 24 26
Time in hrs
Discharges
Figure 9.3. Water levels and water volumes of a tidal outlet.
Box 9.1. Design of tidal sluice.

As an example, the width o(i\ tidal outlet structure is designed. This tidal sluice drains
a polder area of 6000 ha. The open-water system has an area of 12,5 ha, and is
supplied by a field drainage system at a rate of 10 mm/day.
The .sill elevation is located at the bed level of the drainage canal at 1.20 m-MSL.
The tide in the sea is a ".single-daily" tide. i.e. a cycle of 12^2 hours high water and
12^2 houis low water, see figure 9,3.
The water level in the drainage system after Hushing must be again at the target
water level at mean sea level (MSL). The water level will rise during the period that
the tidal gates are closed, e.g. during about 10 hours, thus 10/24 X 6000 / 125 x O.OlO
- 0.20 m. The total volume that have to be discharged during one sluicing-period
amomits to 25/24 x 0.010 x 6000 x 10"^ = 625,000 m^.
The flow will be mainly free flow. Submerged flow will occur only for z < V3
i.e.
for the higher outer water levels:
during the beginning of the sluicing period for an inner water level of 0.20 m + MSL,
when z < 1/3 X 1.40 - 0.47 m, thus for an outer water level of y > 0.93 m ( > 0.27
llll^^jpilillllpilliBl^^
at the end of the sluicing period for an inner water level of 0.00 m + M S L , when r
< 1/3 X L20
0.40 m, thus for an outer water level of y > OMm { >
The calculation is made by means of a simple computer program or by hand, see table
9.1, The time-step may be taken atone hour. The total volume of outflowing discharge
depends on the sluice width /?. and amounts to 34.28 x x 3600 = 123,408 x b n?.
It was already known that the total volume equals to 625,000 m"^, so that 123,408 x
/; = 625,000. It means that the minimum required width h of the sluice amounts to:
lIBiiilHiiilHiiiilliB
170
Table 9.1. Design of a tidal sluice.

Water
Timestep
in hrs
11-12
12-13
13-14
14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22
22-23
23-24
24-25
Calculation
levels
Inside
"H"
m+MSL m
m+MSL m
Head loss
Type of Flow
"z"
m
+0.19
+0.17
+0.16
+0.14
+0.13
+0.11
+0.10
+0.08
+0.07
+0.05
+0.03
+0.02
+0.01
+0.00
-0.02
-0.10
-0.38
-0.56
-0.80
-0.88
-0.96
-1.00
-0.96
-0.92
-0.82
-0.70
-0.56
-0.33
0.21
0.27
0.54
0.70
0.93
0.99
1.06
1.08
1.03
0.97
0.86
0.72
0.57
0.33
1.39
1.37
1.36
1.34
1.33
1.31
1.30
1.28
1.27
1.25
1.24
1.22
1.21
1.20
Outside
1.18
1.10
0.82
0.64
0.40
0.32
0.24
0.20
0.24
0.28
0.38
0.50
0.64
0.87
submerged
submerged
free flow
free flow
free flow
free flow
free flow
free flow
free flow
free flow
free flow
free flow
free flow
submerged
Outfall
Volume
m3
2.38X^X3600
2.51X^X3600
2.70x^7X3600
2.64X^X3600
2.61x^x3600
2.55XZ7X3600
2.52XZ?X3600
2.46x^7X3600
2.43x^7X3600
2.38X^7X3600
2.35X^X3600
2.29XZ7X3600
2.26X^X3600
2.20X^7X3600
Total Volume of outflowing discharge: 34.28XZ7X3600
9.2. Hydro-Mechanical Regulators for Upstream Control

9.2.1. Begemann gate
General. The Begemarm gate is an automatic gate for a constant upstream water level and has
been developed in Indonesia in the 1930s (Vlugter 1940, Brouwer 1987).
Principle. The Begemann gate consists of a flat gate-leaf which pivots around a horizontal
axis, see figure 9.4. The upstream water pressure tends to open the gate, the torque from the
weight of the gate and the counterweight tends to keep the gate closed.
The Begemann gate is simple in constmction, but requires a large headloss z > h. The
downstream water level may not touch the gate-leaf, as this would introduce an additional
closing torque on the gate.
171
target
Q =0
CLOSED G A T E
GATE A T T H E DESIGN DISCHARGE
Figure 9.4. Begemann gate, for regulation of the upstream water level.
Hydraulics. The depth h of the sill and the width b of the strucmre follows from the broadcrested weir formula {Q = 1.1 b H^^^) and from the b/h relation as applicable for gate-leafs.
The location of the axis determines the sensitivity of the gate for water level variations.
The height a of the axis above the target water level is recommended at a = Vi h. A larger
value of a will increase the decrement Ah. The distance p of the axis and the gate-leaf
depends on the location of the counterweight. It is recommended here to place the
counterweight above the gate-leaf, see also figure 9.4, so that the value p < Vs h. The
Begemann gate can also be transformed into a "hook" gate by placing the counterweight just
above the axis, which means that a larger value p is required for the closed position, i.e. p
^ 6h.
The closing torques by the gate weight G and by the counterweight Ware in equilibrium
with the hydraulic torque on the gate, so that for closed position is valid:
(-h
+ fl) = /? X (G + W)
The weight W of the counterweight can be calculated from the above equation, and for a =
Vi h and p = Vs h follows:
W = (1.6 pg b h^) - G
The location of the counterweight Wis determined by the angle a, see also figure 9.4, and
determines the decrement Ah. Model tests have shown (Vlugter 1940) that the angle a should
be in the range from 15 to 55. The adjustment of the counterweight during the installation
of the gate will provide the exact location.
Decrement. The decrement of the Begemann gate depends on the angle a, and can estimated
by the equation Ah = 0.4 - 0.17 tan o;, see figure 9.5.
Generally, an angle of a = 45 to 55 provides the lower decrements of Ah ~ 0.20
h. Thus, the stagnant water depth in the strucmre will he at h - Ah. The rating curve of the
Begemann gate is shown also in figure 9.5.
172
Elevation
above
downstream
canal bed
Elevation
above
downstream
canal bed
free
"submerged" flo w
flow
free
flow
"submerged"
flow
<^
âê-^
Discharge Q
BEGEMANN GATE
Discharge Q
VLUGTER GATE A N D A M I L GATE
Figure 9.5. Decrement of hydro-mechanical regulators for upstream control.
9.2.2. Vlugter gate

General. The Vlugter Gate is an automatic gate for a constant upstream v^'ater level and
operates at a very low headloss z during the design discharge Q.
The gate has been developed in Indonesia in the 1930s (Vlugter 1940) and is
successfully used in irrigation canals, in drains and in rivers. Winches ("windwerk, lier") are
usually applied to open the (automatic) gate to release any floating debris ("drijvend vuil").
Principle. The gate exists of a floating drum that functions as a gate-leaf, and of a counterweight, see figure 9.6. The front face of the drum is vertical, but the drum has a semicircular rear face. The centre of ie circle of the rear face is the same as the rotation axis of
the gate so that the hydraulic thrust of the downstream water level passes through the axis and
does not lead to a closing torque.
The counterweight is located above the drum to provide a closing torque during the
lower discharges, and to provide no torque at the higher discharges as to reduce the headloss
over the gate. The location of the counterweight Wis also determined by the angle oc, see also
figure 9.6.
The model tests of the Vlugter gate have been based on the angle a = 4 5 . The
determination of the weight W o f the counterweight is shnilar to the Begemann gate, but the
exact counterweight has to be determined in the field until the gate opens and closes
correctly.
173
target
Q =0
CLOSED G A T E
GATE A T THE DESIGN DISCHARGE
Figure 9.6. Vlugter gate, for regulation of the upstream water level.
Hvdraulics. The depth h of the sill and the width b of the strucmre follow from the
conveyance formula (z = 1.0 v^/2g) and the relation b/h 5 as applied in the modeltests.
The headloss z will occur during the design discharge Q, when v is the velocity through the
opening of the strucmre without gate. The height a of the axis above the target water level
is recommended at a = Vz h.
Decrement. The decrement Ah of the Vlugter gate has been found in the modeltests and
amounts to Ah ~ 0.15 h. Thus, the stagnant water depth in the strucmre will he at h - Ah.
The rating curve of the Vlugter gate is shown in figure 9.7.
1
0.80
0.50
0.20
i
-0.20
b-L
dischorge curve for
-0.60
r l n u / n c f r o n m
rnnnl
downstream water level
g -0.80
. -1X0
^ ^
stre am SOjtom leva l__

^
-1.60
-1.80"
0 = dcsian water levei, h,=1.' 5 m
-O.i.0
-1.60
r-
up stream woter level
-1.20
hi a 1.15 m, B = 6.00 m
0.60
I -
discharge characteristic
of automatic flap qate
bottom of downstreom conal

1
2
4
6
8
Q for 1 g a t e in- m'A
10
12
Figure 9.7. Rating curver of the Vlugter gate (Vlugter 1940).
16
18
174
9.2.3. AMIL gate

General. The A M I L gate is an automatic upstream water level regulator to maintain a
constant water level on the upstream side (Alsthom undated), see figure 9.8.
The A M I L gates have been developed by Neyrpic/Neyrtec for irrigation, and are commercially manufacmred by GEC Alsthom ACB (Establishment Bergeron, Le Niemeyer. 10/12
avenue des Olympiades, 94132 Fontenay-Sous-Boix Cedex, France. Fax 33-1-48.73.52.19).
Principle. The gate consists of, see also figure 9.8:
o a balanced radial gate-leaf,
o a float attached to the front side of the leaf,
o a "servo-tab", i.e. a reversed bend of the gate-leaf at the top of the gate,
o two counterweights, both adjustable.
The leaf of the gate is shaped to a trapezium with sides ly^î: 0.5jjoj.. This provides the best
flow conditions with a separation from all edges when the gate starts to open. To avoid
jaimning, there is a small gap between the side edges of the leaf in closed position and the
walls of the strucmre. Therefore the gate caimot be fully tight.
Figure 9.8. A M I L gate, for regulation of the upstream water level (Alsthom undated).
175
counter weights
FRONT V I E W
Figure 9.9. Hydraulic parameters of the A M I L gate.
Design of gate-leaf. The A M I L gate is designed so that all the forces acting on the gate
makes that the gate maintains the upstream water level. The hydraulic thrust on the gate-leaf
passes through the axis and does not effect the equilibrium. A condition is that the target
upstream water level ("setpoint") is at the same elevation as the axis, see figure 9.9.
Tvpes. The various types of A M I L gates are geometrically similar to each other and
designated by a dimension index D. The dimension index D is approximately the width d of
the water surface, but expressed in cm, thus D = 100 d. The A M I L gates are available in
a series of increasing widths: D80, D90, DlOO, DUO, D125, D140, D160, D180, D200,
D220, D250, D280, D315, D355, D400, D450, D500, D560, D630, D710 and D800. The
interval "log D" of these gates is constant, and amounts to A(log D) 0.05 (Ankum 1995).
Hvdraulics. The free flow through the A M I L gate occurs for a headloss z > 12 X 10""* D.
The rating curves of the A M I L gate can be derived from the manufacturers leaflet, see figure
9.10 and read (Ankum 1995):
e = 4.4 X 10-5 D^-^ z^-"^ (submerged flow), g = 3.0 x 10"^ D^-^ (free flow)
where: Q is the discharge in m^/s, D is the dimension index of the gate, z is the headloss
over the gate during the design discharge in m.
Design. The design of an A M I L gate under submerged flow for a known design discharge
Q and the corresponding headloss z, and under free flow for a known design discharge Q can
be based on the above formulae (Ankum 1995):
D 120 Q^-"^^ z'^-^^ (submerged flow),
D 160 g^ '^^ (free flow)
The hydraulic dimensions of the strucmre are also expressed in index D: the bed width b =
56 X 10""^ D, the upstream water depth h = 45 X 10'^ D. Thus, with a headloss z > 0.27 h.
Decrement. The decrement Ah of the A M I L gate amounts to Ah ~ 0.05 h. Thus, the
stagnant water depth in the structure will he at h - Ah.
176
Figure 9.10. Rating curves of the A M I L gates (Alsthom undated).
9.3. Hydro-Mechanical Regulators for Downstream Control

9.3.1. AVIS gate
General. The AVIS gate is a downstream water level regulator, to maintain a constant water
level on their downstream side. The AVIS gate has been developed in the 1930s and is
commercially manufacmred by GEC Alsthom ACB (Establishment Bergeron, Le Niemeyer.
10/12 avenue des Olympiades, 94132 Fontenay-Sous-Boix Cedex, France. Fax 33-1
48.73.52.19).
The name ' A V I S ' has a French background: A V are the first two letters of the word
"aval", which means downstream, and S is the first letter of the French word "surface". It
illustrates that the gate operates at a free surface flow.
Principle. The AVIS gate consists of a cylindrical gate-leaf with a trapezoidal section, a
framework and its bearings, and a float in the shape of a cylindrical sector, the whole
forming a rigid assembly, see figure 9.11 and figure 9.12. The gate axis is set at the target
water level, i.e. the water level at zero-flow.
177
The float is not directly floating in the downstream canal but is floating in a 'float
chamber'. The float chamber enables the float to stand continuously in calm water so that it
is not effected by mrbulent flow which passes through the regulator. The float chamber is
provided with a slot in the bottom for admission of the water. The rotating velocity of the
gate V depends on the head difference Ay between the water level in the chamber and in the
downstream canal by the formula, as well as on the (adjustable) opening: v = c Ay.
A damping tank is constructed on the upstream centre portion of the AVIS gate-leaf.
The damping tank is equipped with a wide opening under water, and a small calibrated
opening at the top as an air-vent. The damping tank, together with the float chamber, will
slow down the movement of the gate in case of sudden downstream water level changes or
oscillations. I n addition, the damping tank enables the gate to react temporally but
umnediately to upstream water level changes.
Figure 9.11. AVIS gate for downstream water level control (Alstom undated).
178
b = 0.01 B
F R O N T VIEW
Figure 9.12. A V I S gate (High Head type) for downstream water level control.
High and Low Head Tvpes. Two groups of AVIS gates are available, one for 'High Heads'
and one for ' L o w Heads'. The Low Head AVIS gate has a wider gate of lesser height than
the High Head gate with the same float, see figure 9.13. The High Head AVIS gate has a
higher height E = 0.92 of the gate leaf than the Low Head gate, which has a height E =
0.74 b.
b1.89xlO-2R
Front View of AVIS Gate (High Head)
b 2.13 X 10-2R
Front View of AVIS Gate (Low Head)
Figure 9.13. The "high-head" and the "low-head" type of the A V I S gates.
Identification of gates. AVIS gates are identified by their dhnension index R/B. The
dhnension index R equals the float radius r, but expressed in cm, thus R = 100 r. The index
B equals the bed width of the stracture, in cm, thus B = 100 b, where B 1.89 7? for the
High Head types, and 5 2.13 /? for the Low Head types.
Procurable Gates. The AVIS gates are available in a series of increasing widths:
AVIS High Head types: 56/106, 71/132, 90/170, 110/212, 140/265, 160/300, 180/335,
200/375, 220/425, 250/475 and 280/530;
AVIS Low Head types: 90/190, 110/236, 140/300, 160/335, 180/375, 200/425, 220/475,
250/530 and 280/600.
Headloss chart. The manufacturer of the AVIS gates provides a 'headloss chart' for the
selection of the AVIS gates, see figure 9.14. The chart are not very practical in the day-today design.
It appears that the relations in the chart can also be presented in the form of analytic
formulae. These formulae are easier in the design than the graphs of the chart. Moreover, the
cost aspects, which appears to be often determining, can also expressed in a formula.
179
Ql/s
Figure 9.14. Headloss chart of the A V I S gate (Alstom undated).
Alternative design formulae. The design of an AVIS gate can only be done after the design
of the canal has provided: (i) the headloss z over the gate during the design discharge Q, and
(ii) the variation Y in upstream water level.
Thus, the selection of the AVIS type is based on the following conditions:
o the gate may not be overtopped at the highest upstream water level:
High Head AVIS gate: R > 143 (V + z),
Low Head AVIS gate: R > 250 (V + z);
o the capacity of the gate should be sufficient to pass the design discharge Q at free flow:
High Head AVIS gate: i? > 66 Q^-'^^,
Low Head AVIS gate: R > 51 gO-Ô;
o the selected gate R will have an actual headloss z at the design discharge Q:
High Head AVIS gate: z = 1.0 X 10^ R''^
+ 0.05 X lO'^ R,
Low Head AVIS gate: z = 0.6 X 10^ R ^^
+ 0.05 X lO'^ R.
I f this calculated headloss z is quite different from the original adopted headloss,
another calculation run has to be done,
o the costs C of the gate, depends also on the index R, and should be at the minhnum:
High Head AVIS gate: C 8 R^-^, in US dollar of 1990,
Low Head AVIS gate: C = 9 i?^'^, in US dollar of 1990,
180
Hydraulic dimensions. The hydraulic dimensions of the stmcture are also expressed in the
index R:
o High Head AVIS gate: the bed width b of the stmcmre ^ 1.89 X 10'^ R, sill leyel h
below the downstream target water leyel: h ~ 1.00 X lO'-^ R, the gate radius 1.60 X
10'^ R, and the decrement Ah ~ 0.05 X 10"^ R.
o Low Head AVIS gate: the bed width b of the stmcmre ^> ~ 2.13 X 10"^ R, sill leyel h
below the downstream target water leyel: h 1.13 X 10'^ R, the gate radius 1.80 X
10"2 R, and the decrement Ah ~ 0.05 X 10"^ R.
The decrement Ah of the AVIS gates amounts to Ah ~ 0.05 h. Thus, the stagnant
downstream water depth in the stmcmre will be at /z + Ah.
Hand winch. The larger type of AVIS (and AVIO) gates are provided with hand winches in
case hydro-mechanical control is temporarily suspended.
When an empty canal has to be filled, the discharge through the AVIS gates may be
controlled manually, so as to avoid formation of a hydraulic jump outside the stmcture.
Leakage. AVIS (and AVIO) gates are water level regulators and not discharge regulators. So,
they are not water-tight.
If operating conditions occasionally require the flow to be shut off completely, the AVIS
gate is to be changed by the AVIO gate which is provided with an upstream orifice
(emergency) gate.
Level detection. The float chamber of the AVIS/AVIO gates is linked with the downstream
water level that is representative for the gate setting. This is the water level in the stilling
basin. Apart from this case, it is also possible to coimect a certain water level with the water
level in the float chamber by means of a pipe. It is also advisable to use such a pipe when
two or more gates are parallel. This allows a common feed of the float chambers whatever
dissymmetries may occur. The levels in the float chambers remain the same and the gates are
open to the same extent.
9.3.2. AVIO gate

General. The AVIS and AVIO gates are essentially similar gates. The AVIS gate is made up
of a radial gate which closes completely the canal, see figure 9.15. The AVIO gate is a
variant of the AVIS gate and has also a radial gate, but it closes only an orifice at an offtake
of a canal or reservoir. So the AVIO is capable of standing much higher heads than the
AVIS.
The letter S in the name AVIS has been replaced by the letter O, which is the first
letter of the French word "orifice", to get the name A V I O .
AVIS or AVIO? The choice between the open-type (AVIS) and the orifice-type gate (AVIO)
is solely determined by the maximum headloss likely to occur between the upstream and the
downstream controlled levels. The AVIO gate is capable of standing much higher heads than
the AVIS gate.
181
Figure 9.15. A V I O gate for downstream water level control (Alstom undated).
Principle. Tiie damping tank does not exist on AVIO gates because of the connection with
the metal-lined opening. Extra provisions have been made on the larger types to help the
damping of the gate movements.
High and Low Head Types. Two groups of AVIO gates are available, one for 'High Heads',
see figure 9.16, and one for 'Low Heads', see figure 9.17. The High Head AVIO gate has
a square orifice opening with a width b = w, while the Low Head AVIO has a rectangular
opening with a width b = 2w.
Identification of gates. A V I O gates are identified by their dhnension index R/S. The
dhnension index R equals the float radius r, but expressed in cm, thus R = 100 r. The index
S equals the area of the orifice of the strucmre, in cm^.
The AVIO gates are available in a series of increasing widths:
A V I O Low Head types: 28/6, 36/10, 45/16, 56/25, 71/40, 90/63, 110/100, 140/160,
160/200, 180/250, 200/315, 220/400, 250/500 and 280/630;
A V I O High Head types: 45/32, 56/50, 71/80, 90/125, 110/200, 140/315, 160/400,
180/500, 200/630, 220/800, 250/1000 and 280/1250.
182
FRONT V I E W
Figure 9.17. A V I O gate (Low Head type) for downstream water level control.
Headloss chart. The manufacturer of the AVIO gates provides a 'headloss chart' for the
selection of the AVIO gates, see figure 9.18. The chart are not very practical in the day-today design. It appears that the relations in the chart can also be presented in the form of
analytic formulae. These formulae are easier in the design than the graphs of the chart.
Moreover, the cost aspects, which appears to be often determining, can also expressed in a
formula.
Alternative design formulae. The design of an AVIO gate can only be done after the design
of the canal has provided (i) the headloss z over the gate during the design discharge Q, and
(ii) the variation V i n upstream water level. Thus, the selection of the AVIO type is based on
the following conditions (Ankum 1995):
o each gate can stand a certain maximum upstream water level:
High Head AVIO gate:
> 25 (V + z). Low Head AVIO gate: i? > 50 (V -h z);
o the capacity of the gate should be sufficient to pass the design discharge Q at free flow:
High Head AVIO gate: /? > 87 g^ "*^, Low Head AVIO gate:
> 73
o the selected gate R will have an actual headloss z at the design discharge Q:
High Head AVIO gate:
z = 9.8 X 10^
+ 0.05 X lO'^ R,
Low Head AVIO gate:
z = 2.4 X 10^
+ 0.05 X lO'^ R.
I f this calculated headloss z is quite different from the original adopted headloss,
another calculation run has to be done,
o the costs C of the gate, in U S dollar of 1990, depends also on the index R, and should be
at minhnum: High Head AVIO gate: C ~ 9 R^-^, Low Head AVIO gate: C
11 R^-^.
183
0 1/s
10
10
5 6 789100
5 6 789100
5 6 7 8 9 1000
5 6 7891000
4 5 6 7 8 910000
5 6 78910000
4 5
3 4
Ql/s
Figure 9.18. Headloss chart of the A V I O gate (Alstom undated).
Hydraulic dimensions. The hydraulic dimensions of the strucmre are also expressed in the
index R:
o High Head AVIO gate: the width b of the orifice b ~ 0.90 x 10"^ R, the height w of the
orifice w ~ 0.90 X 10"^ R, sill level h below the downstream target water level: h ~
1.07 X lO'î?, the width
of the strucmre: b^ 2.50 X 10'^ R, the gate radius 1.80
X 10"^ R, and the decrement Ah ~ 0.05 X 10"^ R.
o Low Head AVIO gate: the width b of the orifice ft 1.80 X lO"^ R, the height w of the
orifice w 0.90 X lO"'^ R, sill level h below the downstream target water level: h
1.07 X 10'^ R, the width ftg of the struchire: b^ 2.50 X 10'^ R, the gate radius 1 . 8 0
X 10"2 R, and the decrement Ah ~ 0.05 x lO'^ R.
The decrement Ah of the AVIO gates amounts to Ah ~ 0.05 h.
184
10. DESIGN OF DIVERSION STRUCTURES

10.1. Objectives of Diversion Structures
10.1.1. Need for Diversion Structures
Irrigation. Irrigation is normally a 'discharge control', where the larger main canal flow is
diverted into secondary canals with smaller discharges. Thus, a diversion of discharges is
needed at each bifurcation point ("splitsingspunt"), see figure 10.1.
Figure 10.1. Need for diversion structures in irrigation and flood control.
185
186
Drainage. Drainage is normally a pure 'water level control' without a diversion of discharges.
For instance, rainfall enters into the smaller drainage chaimels which are flowing into the
larger drainage channels, just by gravity and without any discharge regulation. Even, drainage
pumping stations are operated at 'water level control', as the pumps are started when the
water levels are above the target levels.
Flood control. Often, the capacity of a river in a delta is too small, and more protection
against floodings ("hoogwaterbescherming") is required. The construction of a new flood
diversion channel ("hoogwater afleidingslcanaal") might be an option for flood control. Such
a channel is attractive when the distance to the sea can be shortened, and thus the gradient.
Like in irrigation, a diversion of discharges is needed at the bifurcation point ("splitsingspunt"), see also figure 10.1.
Need for diversion strucmres. 'Diversion structures' ("verdeelkunstwerken") are required at
every offtaking channel when a discharge has to be diverted, as happens in irrigation and in
flood control. Otherwise, the tailwater levels will influence the diversion. This happens
because of the different depth-discharge relations of the canals (Q = k A R^^^ s^'^)\
the dimensions of the two offtaking chanels are different, and so are the values of the wet
cross sectional areas A and the hydraulic radius R;
the gradients of the two offtaking channels are different, and so are the values s;
the maintenance schedules are different for the two canals, and so there are different
Strickler coefficients k for the canals, see box 10.1;
the bed level of the ongoing channels ean be lowered (by scouring, by excavation), and
so there are lower tailwater levels.
Box 10.1. Effect of the maintenance schedule on the water distribution at a 'free offtake'.
It is normal in irrigation that different canal reaches have different maintenance
schedules. Thus, they have different Strickler coefficients k for the canals during the
irrigation season. Grass growing and sediment deposits will easily reduce the
coefficient k Xo 5{V''c or 25% of its original value. It means that the tailwater level rises
lor a given discharge.
Such a rise in tailwater level in the ongoing irrigation canal means that a larger
portion of the discharge is diverted from this canal.
Hence, an over-supply is given lo the first offtakes in the system. The headcnders of the system receive a larger portion of the flow. As a consequence, problems
of water shortage will arise at the tail-end.
Terminology. The function of a diversion stmcture is to divert a certain flow from an

upstream channel reach, i.e. the 'incoming channel' ("onverdeelde kanaal"), into a diversion
channel, i.e. the 'offtaking channel' ("aftappende kanaal") and the ongoing channel, i.e. the
'continuing channel' ("doorgaande kanaal"), see figure 10.2.
The regulator towards the offtaking channel is called 'offtake' ("aftapping"), and the
regulator towards the continuing channel is called 'control' ("stuw"), see also figure 10.2,
187
10. D E S I G N O F D I V E R S I O N STRUCTURES
'offtaking channef
t
'OFFTAKE'
'CONTROL'
1 1 1 1
^continumg
cbamet
1 1
Figure 10.2. Terminology of a diversion strueture.
Need for objectives and design conditions. The selection of the type of flood diversion
strucmre requires clear 'operational objectives' ("doelstellingen").
Furthermore, the diversion stracmre and the chaimels have an interrelation on each
other that should be specified, and should be translated into 'design conditions' ("ontwerpeisen") and 'design criteria'
("ontwerpuitgangspunten").
10.1.2. Operational Objectives of Irrigation Diversion Structures

Options for water allocation to the tertiarv unit. Splitted flow is the traditional allocation
method to a tertiary unit in many village-managed systems.
Intermittent ('on'off') flow to the tertiary units is often applied in schemes under
dryland crops. Intermittent flow to tertiary units is essential in schemes with only 'dayirrigation', when an interval of 12 hours is required during the supply period.
Adjustable flow to the tertiary units is often applied in schemes under paddy, when the
peak discharge is required e.g. during land preparation and is gradually reduced during the
off-peak season.
Options for water diversion through the main svstem. The selection of method for water
diversion through the main system has to follow the above selections on the method of water
allocation.
Thus, the following options for water diversion through the main irrigation system
during the off-peak period are possible, see figure 10.3 (Ankum 1992^):
a splitted flow to tertiary units auns usually to select also a splitted flow throughout the
whole main irrigation system in order to obtain a proportional diversion of the water.
However, it is also possible to change at higher levels to an adjustable flow, an
intermittent flow (e.g. pumping station) or a rotational flow;
188
CONVEYANCE
to the
SECONDARY UNIT
DIVERSION
at the
HEADWORKS
Q in I/s.ha
Peak
Period
AQ in 1/s.ha
AQ in 1/s.ha
*
Off-Peak
Period
Pealc
Period
Off-Pealc
tune
Off-Peak
Period
Peak
Period
Period
AQ in 1/s.ha
Off-Peak
Period
tune-
Period
Peak
Period
Peak
Period
Off-Peak
Period
time-
Intennitteat How
Q in 1/s.ha
'
Period
tune-
timeRotational How
AQ in 1/s.ha
Adjustable How
Off-Peak
Xnttxmitt^t Flow
AQ in 1/s.ha
Off-Peak
Period
time-
tuneAdjiistable How
Peak
Period
Period
latetmittent Flow
AQ in 1/s.ha
Period
IQ in 1/s.ha
Off-Peak
tune-
Intennitteait How
Off-Peak
Off-Peak
Splitted Flow
Q in 1/s.ha
Peak
Period
tune-
Splitted Flow
Q in I/s.lia
4 Peak
Period
Period
Peak
Period
time-
Splitted Flow
Period
ALLOCATION
to the
T E R T I A R Y UNIT
AQ in 1/s.ha
Off-Peak
Period
time-
Adjustable How
Figure 10.3. Flow diversion in the main irrigation system
Peak
Period
Off-Peak
time-
Adjustable How
Period
189
an intermittent flow to tertiary units dictates either rotational or adjustable flow in the
main irrigation system. A n intermittent flow in the main system would not match a
regulated intermittent flow at the tertiary offtake, and could simply be obtained by a
splitted flow at the tertiary offtakes;
an adjustable flow to tertiary units dictates also an adjustable flow at the higher levels.
Intermittent or rotational flow over (sub)secondary canals would be in contradiction to an
adjustable flow at the tertiary offtake.
10.1.3. Operational Objectives of Flood Diversion Structures

Operational obiectives. Many operational objectives for flood diversion are possible, such as,
see figure 10.4:
proportional diversion, i.e. all flow, and also the floods, are diverted 'proportionally'
("proportioneel") into a fixed ratio between the ongoing river ("doorgaande rivier") and
the flood diversion channel ("hoogwaterkanaal");
diversion of floods only, i.e. the flow above a certain threshold value ("drempelwaarde")
is diverted into the flood diversion channel. This threshold value is determined by e.g. the
capacity of the downriver (beneden rivier);
proportional diversion, but to a maximum, i.e. the river flow is diverted proportionally
until the capacity of the ongoing river is reached. The higher river flow is diverted into
the diversion channel.
/
/
Diversion
channel
Diversion
-I
200
300.
Diversion
chaimel
Ongoing
\river
Ongoing
\nver
100
J.S
ciiannel
100
200
300,
Ongoing
fiver
100
200
300
River discharge in mVs
River discharge in m7s
PROPORTIONAL
DIVERSION
DIVERSION
OF FLOODS ONLY
PROPORTIONAL
DIVERSION,
but to a maximum
= design discharge CUx.
Figure 10.4. Some options for flood diversion.
190
Diversion above the design capacity. Flood control worlcs are designed for a 'design
discharge' ("ontwerpafvoer"), e.g. the 1 : 20 years' flood. Higher discharges will certainly
occur, but are considered as an 'act-of-God' ("overmacht") for which provisions are not
taken.
The floodings during these higher discharges above the design discharge max
concentrated in a certain area. So it is possible to define an additional operational objective
for flood diversion, see figure 10.5:
protection of the ongoing river against floods above the design discharge,
protection of the diversion channel against floods above the design discharge.
200
300
400
500

PROTECTION OF
ONGOING RIVER
= design discharge

PROTECTION OF
DIVERSION CHANNEL
QVIAX.
Figure 10.5. Options for flood diversion above the 'design discharge'.
Design discharges. The flood diversion structure itself is also designed on a 'design flood'
("ontwerp ajvoer"). Higher floods may damage the strucmre. Typically, the flood diversion
structure is damaged when the abutments ("landhoofden") are overtopped and water passes
over the unprotected slopes. The strucmre will also be damaged when the stilling basin
("woelbak") does not dissipate the energy sufficiently. So, the 'return period' of the design
discharge of the flood diversion strucmre determines the risk of failure.
The remrn period of the design discharge must be selected in combination with the
freeboard of the abutment above the flood level. For instance, a 'freeboard' ("waking") of
1.50 m above the 1 : 50 years' design flood may provide the same protection against
overtopping, as a freeboard of 0.25 m above the 1 : 1000 years' flood.
10.1.4, Sensitivity of Diversion

Sensitivity. The concept of 'sensitivity' ("gevoeligheid") evaluates the performance of
bifurcations under varying discharges.
The 'sensitivity of a bifurcation' ("gevoeligheid van een splitsing") specifies the
variations in the ojftaking discharge in relation to the discharge in the continuing charmel. I n
literamre, the term 'flexibility' is incorrectly used for this concept, see box 10.2.
191
Box 10.2. Terminology: 'sensitivity' or 'flexibility' of structures?

The here defined "sensitivity" of a structure is also called "flexibility" in literature (Bos
1989). U is proposed (Ankum 1992'\ 1993^^) to defer from the term "flexibility" in this
The term "flexibility" means: adjustable to changes (Webster dictionaiy), and is
widely used in irrigation literature in relation to "demand-based" water allocation U)
the tertian' unit (Merriam 19K7). A .similar meaning of the term "llexibilit)'" is used on
the I C I D Congress of 1993 in Question nr. 44.2; "Planning and Design for Flcxlbihty",
which was elaborated as: the system should have sitfJicLmt Jlexibildy to cope with changa
Furthermore, the term "sensitivity" is used in literature for: the accuracy of
discharge measurement structures in relation to the vaiiations in water level (Bos 1989).
It recommended here to maintain the following terminolog>' in flow control:
' the term "sensitivity of a structure" is related to the variations in tbe offtaking
discharge, in relatioii to the discharge in the continuing discharge in the canal;
the term "ncxibility" should be related to the freedom of users on the frequency, rate
and duration of the water allocation of the flow control method.
The term "accuracy of a discharge measurement structure" can be used for the effect
of water level changes on the reading of discharge measurement structures.
Definition of sensitivitv. The sensitivity 'S' is defined as the ratio in the variation of an
offtaking discharge Q^, and the variation in the discharge
in the continuing channel. Thus,
the sensitivity 'S' depends on the rating curves of the regulators, see figure 10.6:
the depth-discharge relation of the offtake is a function of the energy head: t = ^ -^^t^'
or is a function of the water level y^: t = ^ )'t^ the depth-discharge relation of the control is a function of the energy head: Q^ = a H^,
or is a function of the water level:
= a y^.
Offtakmg
canal
"FREE" O F F T A K E
Figure 10.6. Rating curves at the bifurcation.
Offtaking
canal
C O N T R O L L E D OFFTAKE
192
Hence, the sensitivity S:

dQ, I dH
^ T H,
r-l
S = -
Thus, the sensitivity 5 of a stracture can be calculated

Tx
H^
C ^ H /
or:
C X y^
Sensitivity of 'fixed' regulation. The sensitivity as discussed above, is related to fixed

regulation. The following rating curves are applied, see figure 10.7:
fixed water level regulators:
o underflow,
H ~ ^ ,
o overflow,
H =
o long-crested weirs, H = constant,
fixed discharge regulators:
o underflow,
Q H^'^,
o overflow,
Q H^-^,
o 'special' orifices,
Q = constant.
discharge
W A T E R L E V E L REGULATORS
discharge DISCHARGE REGULATORS
Figure 10.7. Rating curves of regulators.
Sensitivity of 'manual' and 'automatic' regulation. The concept of sensitivity can also be
utilized for manual and automatic regulation. Now, the target value is maintained under all
circumstances. It means that the rating curves becomes most simple, see also figure 10.7:
manual or automatic water level regulators:
o all flow types,
H ~ constant,
manual or automatic discharge regulators:
o all flow types,
Q - constant.
193
Manual regulation at fixed times onlv. It is a common practice is many irrigation schemes
to adjust the regulators only e.g. once per 14 days, to match the water requests with the water
availability. It means that the regulators perform as 'fixed' regulators in the period between
the adjustments. Thus, the sensitivity depends also on the frequency of regulationl
Constant discharge diversion. A n offtake of a diversion stmcture has no sensitivity to any
variations in incoming discharges, when 5 = 0, see figure 10.8. This occurs when the
offtaking channel maintains at a perfect constant discharge Q^. Such a constant discharge
occurs:
an offtake which is designed for a perfect constant discharge, independent of the upstream
water level, e.g. by a special underflow such as the baffle distributor in irrigation,
a control that maintains the upstream water level at a kept constant level, e.g. by A M I L
gate, long-crested weir, but also by a vertical gate that is operated continuously.
'offtaking' |
channel
1
'iiiconihig' ohimnel
to
Time
'ongoing' channel
I
Q
Time
Figure 10.8. Flow diversion for Sensitivity S = 0.
Low sensitivitv of a diversion stmcmre. A n offtake has a low sensitivity to variations i n

discharge when 5 < 1. This occurs when the discharge variation in the offtaking discharge
are less than that in the continuing channel. A low sensitivity {S < 1) occurs at:
a 'free-offtake' with underflow (gate), that is not operated,
a control with overflow (weir) and an offtake with underflow (gate), that are both not
operated, see box 10.3.
Box 10.3. Example of diversion stmcture with a low sensitivity (5 < 1).
A flood diveision stiuclure is constructed with overflow (weir) to the ongoing river
channel and underflow (orifice) to the offtaking flood channel.
The depth-discharge relation to the ongoing river channel reads
= ay^^'^ with
a design water depth
^ 0.75 m. The depth-discharge relation to the offtaking flood
channel reads a = B x^-^ and design water depth y-j = 1.00 m. The sen.sitivity S can
be calculated al .V = [T H^) / (C H^) ^ (0.5 x 0.75) / (1.5 x LOO) = HX
It means that the offlake will take a very low variation dO of the changing river
BlBS^lllillllHH
194
Proportional diversion of a structure. The diversion structure is exact proportional for 5* = 1

and any percentage change in the incoming discharge causes precisely the same percentage
change in the offtaking discharge, see figure 10.9. A sensitivity S = 1 occurs:
v^'hen the same type of flow is used in the offtake and in the control, i.e. both overflow,
or both underflow. Moreover, both energy heads (H^ and H^) should have the same value,
because their products should have the same value: T X
= C x H^.
As above, the regulators are not operated here.
High sensitivitv of a diversion stmcmre. The variation in the offtaking discharge will be more
than that in the ongoing channel for a high sensitivity, when S > 1. A high sensitivity 5'> 1
occurs at:
a bifurcation without stmctures,
a 'free-offtake' with overflow (weir), see box 10.4,
a control with underflow (gate) and an offtake with overflow (weir).
'offtaking' |
channel
t
'incoming' channel
Time
o
' ngoing' channel
jf.lllTT
Time
Figure 10.9. Flow diversion for Sensitivity S = 1.
Box 10.4. Example of diversion stmcmre with a high sensitivity (S > 1).
A Hood division structure is constructed with underflow (orifice) to the ongoing river
channel and overflow (weir) to the offtaking flood channel.
The depth-discharge relation to the ongoing river channel reads Q^. - ayj-^'^ with
a design water depth N^. = 1.20 m. The depth-discharge relation to the offtaking flood
channel reads
= By^^-^ and design water depthy^ = 0.75 m. The sensitivitj' S can
be calculated at S = (TH^) J (C H^) =: (1.5 x 1.20) / (0.5 x 0.75) = 4.8.
I l means that the offtake will take a vciy high variation dQ of the changing
BilfcK^fcilili^^BjiiHiBHIIiH
Perfect sensitivity of a diversion stmcmre. A diversion stmcture has a perfect sensitivity for
S = 00, when all variations in incoming discharge is diverted into the offtake, and a constant
flow is diverted into the continuing channel, see figure 10.10. A perfect sensitivity 5=00
occurs at:
* an offtake that maintains the upstream water level at a kept constant level, e.g. by A M I L
gate, long-crested weir, but also by a vertical gate that is operated continuously,
a control which is designed for a perfect constant discharge, independent of the upstream
water level, e.g. by a special underflow such as the 'baffle distributor' in irrigation.
195
'offtaking' k
channel
Figure 10.10. Flow diversion for Sensitivity S = <.

Relevance of 'sensitivitv'. The relevance of the concept 'sensitivity' in flood control is
Ihnited. Normally, flood control involves diversion at one stmcture only.
The concept of sensitivity is most hnportant in irrigation schemes with flucmating river
discharges. Irrigation systems consist of a large number of structures, where the everchanging flow has to be diverted. These systems may have serious problems with the equal
distribution of discharge flucmations, see figure 10.11.
A solution for these equity problems can be found in 'gate-proportional' facilities within
the upstream control method. This can be obtained by two supplemental measures (Ankum
1995):
the constraction of the same type of flow for water level regulators and discharge
regulators, i.e. both overflow, or both underflow. The selected combination determines the
"sensitivity" of the diversion stracmre;
the operational order not to regulate these stractures during an 'irrigation cycle', i.e. the
period between the system adjustments to new discharges, e.g. 14-days.
Inflow in canal
outflow offtake 1
outflow offtake 2
outflow offtake 3
Sensitivity of diversion: S = 0
Offtake 3
Inflow in canal
outflow offtake 1
outflow offtake 2
outflow offtake 3
Sensitivity of diversion: S = 1
Figure 10.11. The concept of 'Sensitivity' in irrigation.
196
10.2. Design Considerations and Design Criteria

10.2.1. Design
Conditions
Backwater effects. A diversion structure has an influence on the water levels in the upstream
channel. The 'depth-discharge relation' ("Q-h kromme") of the struchire (orifice or weir
formula) is different from the depth-discharge relation of the channel (Strickler formula). It
means that the structure creates backwater effects on the upstream channel: there is either
a 'setting-up' ("opstuwing") or a 'drawing-down' ("ajzuiging") of the water level, and the
uniform flow is only obtained in specific cases.
Thus, an additional design condition ("ontwerp eis") of the diversion structure can be
included in the design: "there is uniform flow during the 'bed-forming' or 'dominant'
discharge ("bedvormende, dominante ajvoer") for which the channel bed is supposed to be
in equilibrium", see figure 10.12 It means that:
o the dominant discharge dominant ^^^^ create no backwater curves on the upstream
channel;
o the lower discharges Q < dominant "^^^^ ^^^^ * ^ 'positive' backwater curve ("opstuwing"), with sedimentation before the stmcmre because of a lower energy dissipation
E = p gv s;
o the higher discharges Q > dominant ^^^^ ^^^^
^ 'negative' backwater curve or 'drawdown' ("afzuigen"), with scouring before the stmcture because of a higher tractive force
T =
pgys.
Uniform
Flow
Free flow at Q20
No backwater effects at QDOMINANT
DESIGN CONDITION:
"BACKWATER EFFECTS"
DESIGN CONDITION:
"FREE FLOW"
Figure 10.12. Two design conditions for a diversion structure.
Super-critical flow. The diversion at a stmcture is also determined by occurrence of:

sub-critical flow ("onvolkomen, niet-schietende stroming"). For instance, a sub-critical
flow towards the ongoing channel makes that any changes in tailwater levels ("achter
waterstanden") lead to changes in the diversion;
super-critical flow ("volkomen, schietende stroming"). A super-critical flow makes that
any changes in the tailwater levels do not have any effect on the diversion of flow, see also
figure 10.12.
197
Thus, an additional design condition ("ontwerp eis") can be included in the design: "there
is always 'free flow' ("volkomen stroming") towards the diversion channel and/or towards
the ongoing river".
10.2.2. Hydraulic Formulae

Weir for water level control. A weir is used for 'passive' regulation of the water levels, see
figure 10.13. The structore is normally applied for free flow only. The discharge formula of
a weir under free flow reads, see also section 7.1 of chapter 7:
2
2
g
= c. X 1.71 m^f^ls
'^ 3
3*
where Q is the discharge in m^/s, c is the weir coefficient in m^^^/s,
is the discharge
coefficient, b is the width of the crest in m, H is the energy head above the crest in m.
The width b of weirs can be reduced by applying a sharp-crested type with a higher
weir coefficient c in the discharge formula.
For example, a broad-crest weir may have a weir coefficient c = 1.7 m^^^/s. And a
sharp-crested weir may have a weir coefficient c = 1.9 m^'''^/s for unrounded crests, c = 2.1
m^^-^/s for cylindrical crests, or even c = 2.3 m^^'^/s for the specially designed 'ogee-crests'
(USBR 1973). It means that the width of an ogee-crest is only 75 % of the width of the broadcrested weir.
The value of the weir coefficient c influences also the required headloss z for free flow.
For example, a broad-crested weir requires a small head of some z ~
H. Cylindrical crests
and ogee-crests require a minhnum headloss z ~ Vi H. Unrounded crests require a larger
headloss z for aeration of the nappe ("beluchting van de straal"), thus z > > H.
Thus, an additional design condition ("ontwerp eis") can be included in the design:
"there should be a minhnum width b of the weir to reduce the costs, by selecting a weirshape with a high weir coefficient c".
Q
= c b
with:
c =
SHARP-CRESTED WEIR
Figure 10.13. Design conditions for a weir.
198
Orifice for water level control. A n orifice opening, such as a 'flood screen' ("bandjirscherm"), may be used for 'passive' regulation of the water level. There is submerged flow
for the lower discharges, as the orifice should not create much headloss. During the higher
discharges, however, the orifice obstmcts the flow and the upstream water level will rise
considerably, see figure 10.14.
The headloss formula of an orifice under submerged flow, 'conveyance flow', reads,
see also section 6.2 of chapter 6 and section 8.3 of chapter 8:
z
= a
2g
with: v -
y b
with the headloss z in m, a coefficient a, the velocity v within the orifice, the discharge Q
in rc?ls, the (average) water depth y within the stmcmre, and the width b of the stmcmre,
and the gravity acceleration ^ = 9.8 m/s^ = 9.8 N/kg.
The above formula can be re-written into the discharge formulae for the control during
submerged flow for an open-water level in the orifice, so y j < w:
- b y ,
sf2i7
a
for orifice flow, so y^ > w:
Q
2 - ^ b w \j2gz
where Q is the discharge in m^/s, a is a coefficient, b is the width of the orifice in m, y^ is
the upstream water level above the sill in m, w is the height of the orifice-opening above the
sill in m, and z is the headloss in m.
The coefficient a depends on the entrance and the exit losses, and can be taken at o; =
0.5 to 1.0, for instance. In principle, the coefficient a should be kept as low as possible to
avoid high headlosses. The lower coefficients a are obtained by allowing an open-water level
in the orifice, as long as regulation is not required.
Thus, an additional design condition ("ontwerp eis") can be included in the design: "the
height w of the orifice opening should allow an open-water level in the orifice, as long as
water level regulation is not required, so y^ < w".
199
Orifice for discharge control. An orifice, such as a 'gated offtake', is used for regulation of
the discharge. The structure is normally applied for free flow only, see figure 10.15.
The discharge formula of an orifice under free flow reads, see also section 8.4 of
chapter 8:
Q = C^bw
^2gy,
where Q is the discharge in m^/s, b is the width of the opening in m,
~ 0.50 - 0.55 is
the discharge coefficient, w is the height of the opening in m, y^ is the upstream water depth
above the sill in m, and g = 9.8 m/s^ = 9.8 N/kg is the gravity acceleration. The discharge
formula can be approxunated by:
Q = 23bw
H^'^.
The minimum headloss z for free orifice flow is quite complex, see section 8.4 of chapter 8.
Normally, a minhnum headloss z ~ Vi B. is required for free flow. The above formula is
valid for H > '^1^ w.
However, the design of an orifice for discharge control should not apply the above
formula for an orifice-type of flow. The dimensions of the orifice are determined by the
design (maximum) discharge. A n orifice-type of flow means a vertical contraction by the
'front wall' ("borstwering"), and requires a wider width b of the opening and a more
expensive stmcture.
NOTCH-TYPE OF FLOW
Figure 10.15. Design conditions for an orifice as discharge regulator.
200
It is better to apply the formula of a 'control notch' ("vernauwing") for the design of
an orifice for discharge control, and to allow an open-water flow through the orifice during
the maxhnum discharge, see also figure 10.15.
The discharge formula of a 'control notch' under free flow and with vertical wall is
closely related to the formula for a broad-crested weir and reads, see also section 7.1 of
chapter 7:
Q =
l.SbH^'^
where Q is the discharge in m^/s, t> is the width of the opening in m, H is the energy depth
above the sill in m . Normally, a minimum headloss z ~ Vs H is required for free flow. The
above formula is valid for H <
w.
Thus, an additional design condition ("ontwerp eis") can be included i n the design: "the
height w of the orifice opening should allow an open-water level in the orifice during the
maximum discharge, so H < ^/j w".
Sill level of the orifice. The energy depth H and the opening w are both related to the height
p of the sill ("drempel") above the riverbed, see figure 10.16. A high sill level for > >
means that the energy head H and the opening height w are smaller, so the width b of the
orifice should be increased at higher costs. The minimum costs of the structure are often
obtained for a sill at bed level, sop = 0.
Thus, an additional design condition ("ontwerp eis") can be included i n the design: "the
width b of the orifice is at a minimum by selecting the sill of the orifice at bed level, so p
= 0".
ORIFICE WITHOUT SILL

Figure 10.16. Design conditions for the sill level of an orifice.
201
10.2.3. Side-spillway
effect
Side-spillwav. The hydraulic design of the side-channel spillway has been discussed in section
7.3. of chapter 7. Two features are important, and are called here the 'side-spillway effect',
see figure 10.17:
there is a constant energy head in the longimdinal channel direction, as there are no
energy losses other than friction losses. The energy head is determined by the energy head
ôngoing
downstream (ongoing) channel reach.
It means that in upstream direction: (i) the water level drops, and (ii) that the velocity
head v^llg increases ("toeneemt");
there is a loss in energy by sudden flow deflection ("plotselinge verandering van stroomrichting"). It is assumed that these 'bend losses' equal 1.0 v^llg.
It means that i n upstream direction: (i) the total velocity head v^/2g is lost, and (ii) the
energy depth
over the spillway crest decreases ("afneemt").
The 'side-spillway effect' makes that the hydraulic calculation has to be done step-by-step in
upstream direction, and that the spillway becomes quite long.
Bend loss
CROSS-SECTION
Figure 10.17. The 'side-spilway effect' at diversion structures.
'No side-spillwav effect'. The above effect of side-spillway can be ignored, and there is 'no
side-spillway effect' when, see figure 10.18:
o there is a gentle flow deflection, so that there are no bend losses a v^/2g because the value
of a 0 ,
o the energy depth
over the spillway crest is much larger than the bend loss a v^/2g, so
the bend loss a v^/2g can be ignored, as
> > a v^l2g,
o the variation Mi^ of the energy depth
over the spillway crest is small and can be
ignored, so AF/j < <
The result of 'no side-spillway effect' makes that the hydraulic calculation is most simple,
and that the energy head at the offtake equals the energy head iôngoing
the downstream
(ongoing) channel reach.
Thus, the energy depth
above the offtake-crest is: Z/j = //offtake 'P^ where p is the
elevation of the offtake-crest above the bed.
202
v72g
H.
H,ONGOING
f side-spillwa^
CROSS-SECTION
QT + QC
Figure 10.18. No 'side-spillway effect' at diversion structures.
10.2.4. Minimum Width of the Structure

Check. A check should be made on sub-critical flow ("stromend water") within the entrance
of the diversion structure, see figure 10.19. I f there is no sub-critical flow but super-critical
flow ("schietend water"), there will be excessive friction losses within the structure. This
leads to a high tractive force T, because ofT = p gy s. Furthermore, the headloss by the
structure will increase, which has effects on the backwater effects in the incoming chaimel.
Moreover, the hydraulic calculation becomes quite complex.
The above check on the sub-critical flow can be made by determining the water line in
the diversion strucmre, and to calculate the corresponding Foude number Fr from Fr =
^/(gy)^'^- It is also possible to use broad-crested weir formula, as will be shown below.
'ofthkmg
ckmiieh
'OFFTAKE'
1
1'CONTROL'
'mcoming
dmnnef
'contim
Kim
'ENTRANCE'
Figure 10.19. The minimum width b^^^ within a diversion structure.
203
Water line within the structure. The course ("verloop") of the water line within a structure
is based on the following considerations, see figure 10.20:
there is a constant energy head in the longitudinal flow direction, and the energy head is
determined by either:
o for a free-offtake: the energy head equals the energy head H^^^^^^g of the downstream
(ongoing) channel reach,
o for a controUed-offtake: the energy head equals the energy head
+ p at the control.
the water level is determined by the velocity head v^/2g. So, the water level will change
for a changing discharge.
A n example of the water line calculation within a diversion stmcmre is presented in box 10.5.
offtaMng
chaimel
-VB72g
lvA72g
i
=
'GATES' m i l ]
QT + QC
QC
PLAN
'
> \ - lyc = % H e ^
'WEIR'
jC
QC
QT + QC
QC
QC
Figure 10.20. The water level within a diversion stmcture.
Direct calculation. The minimum width ftj^jj^ within a diversion stmcmre to avoid supercritical flow can also be calculation dirctly by using the broad-crested weir formula. Flow
over the broad-crested weir has the 'critical depth' with a Froude number Fr = \. Thus, the
following relation is valid:
with the discharge Q through the stmcmre in m-'/s, the minhnum width b.^,^ within the
stmcmre, and the energy head H m.m. The energy head E follows from the downstream
conditions at the control, e.g. H = p + H^. A n example of the minhnum width within a
diversion stmcmre is presented also in box 10.5.
204
Box 10.5. Example of the water line calculation and the minimum width within a stmctore.
The scttin. A flood di\'crsion structure has to divide a 20-years flood C?2{) - '^^^^ "^^'-'^
into an offtaking discharge Q^ - 250 m~Vs and a continuing discharge C?^. - 150 m'/s.
The structure has vertical walls at a mutual distance h = 30 m, and with a weir as
the control. The weir-crest i s p = 1.50 m above the bed. and is b = 15.75 m wide. The
rating cuive ol the shatp-cfestcd weir is (? = 2.1 ft //^'-. The river bed within the
structure is at 10.00 m+.
Design. The energy depth during the design flood of
= 150 nv^s is at H = p +
= 1.50 + 2.74 = 4.24 m above the bed.
The water depth >( in point C is at % //^ above the crest. So, the water level at
point C is at 10.00 4 1.50 + % x 2.74 - 13.33 01+.
The water depth Vj^ in point B is determined by the energy depth H = 4.24 ni above
the bed. the discharge
= ^^0
and by the width ft between the vertical walls.
The iterative calcula"tion is as follows:
assume i' = 0.00 m/s, soy^j = 4.24 - v;/2^ = 4.24 m. and v = 150 / [30
= 1.18 m/s,
assume i- = 1.18 m,/s, soyj^ - 4.24 - v-J2g - 4.17 m, and v - 150 / [30yg] - 1.20 m/s,
assume v - 1.20 m/s, sovg - 4.24 - r~f^ - 4,17 m, and v = 150 / i30yj3] ^ 1.20 m/s.
So. the water level al point B is at 10.00 + 4.17 = 14.17 m"*".
The water depth at point A is determined in a similar way for a discharge
+
= 400 m'/s. It can be anticipated that the water level at point A will be lower than at
point B, as the term v-/2g is now larger. Thus, a water depth y ^ = 3-51 m and a
velocity v - 3.79 m/s is calculated. So, the water level at point A is at 10.00 + 3.51 =
nrVs
B^HlllHIIIIBIIiHIIIIlillW
Check. A check should be made on sub-critical tlow, as super-critical flow with
excc.ssive friction losses has to avoided. The Froude number in point C follows from
/.V = v/ig\f-^ = 3.79/(9.81 x3,5i)^'"'' = 0.65. So, there is sufa-crltlcal flow at point A.
The minimum structure width ft^,,,, foIlo\\s from the broad-crested weir formula;
C> = 1.7 ftji //''-. Hence. 400 = 1.7 ftj^ji,j 4.24^-. So, the minimum width ftj,,i,j =
26.95 m, which is less than the distance ft =f 30 m. It means: there is sub-critical flow
iBiBiiiiiBiiiiiiiii^
205
10.3. Options for Diversion Structures

10.3.1.
Regulators
'Free' and 'controlled' offtakes. Diversion structures may have one or two 'regulators'
("regelaars"), i.e. an 'offtake' ("aftapping") and/or a 'control' ("stuw"), see figure 10.21:
a free-offtake ("vrije aftapping") has only a 'regulator' in the offtaking channel, and no
'control' in the continuing chaimel;
a controUed-offtake has both a 'regulator' in the offtaking channel, as well as a 'control'
in the continuing channel'.
offtakmg
offtaldng
cbamiel
t
/OFFTAKE'
'OFFTAKE'
'incoming' ffunnel
channel
'continuing'
i lunncl
I
'uimmuwg
'incoming' cimnricl
'CONTROL'
"CONTROLLED" O F F T A K E
"FREE" O F F T A K E
Figure 10.21. A diversion structure as a 'free' offtake and as a 'controlled' offtake.

Open offtake. It is also possible that design an 'open offtake' without a diversion structure,
see figure 10.22. However, a reliable and accurate diversion is impossible without regulators,
as the diversion will depend on the tailwater levels ("achterwaterstanden") in the continuing
channel and in the offtaking channel. It means that the diversion changes on e.g. changing
cross-sections and channel roughness. This matter is discussed at the beginning of the chapter.
OPEN OFFTAKE
without Control
Figure 10.22. No diversion structure is unsuitable for flow diversion.
206
What is regulated? The regulators can either be (see chapter 5):

a 'water level' regulator ("peilregelaar"), or
a 'discharge' regulator ("debietregelaar").
Furthermore, it is also possible to regulate on the 'ratio-in-discharge'
("proportioneel").
How is regulated? The regulation of the regulators can either be (see chapter 5):
Passive regulation ("vaste regeling"),
Manual regulation ("hand-bediende
regeling"),
Hydro-mechanical regulation ("hydro-mechanisch"),
Electro-mechanical regulation ("electro-mechanisch"), which is not discussed here.
Basically, there are only two manner of regulation, i.e. (i) 'passive', and (ii) 'gated' which
covers manual, hydro-mechanical and electro-mechanical regulation.
What flow tvpe? The flow type at the regulators can either be (see chapter 5):
Overflow ("overlaat"),
Underflow ("onderspuier).
10.3.2. Configuration of a Diversion
Structure
Configuration. The operational objectives of a diversion strucmre should match the type of
'offtake' and the 'control', i.e. the 'configuration' ("configuratie"). The options are:
the type of regulator (none, water level, discharge, ratio-in-discharge),
the manner of regulation (passive, manual, hydro-mechanical),
the type of flow (overflow, underflow).
Not all configurations appear to be very practical.
Practical configurations for free-offtakes. There are only two practical configurations for freeofftakes, see figure 10.23:
overflow free-offtake ("overlaat
vrije-aftapping"),
gated free-offtake ("regelbare
vrije-aftapping").
The design of these configurations will be discussed in the next chapter.
'WEIR' with
^GATES' with
free flow
free flow
mcomins
channel
ongoing
OVERFLOW FREE-OFFTAKE
Figure 10.23, Types of free offtakes.
incoming
t
ongoxbg
GATED FREE-OFFTAKE
207
Practical configurations for 'passive' diversion structures. Three configurations are practical
for 'passive' diversion struchires, see also figure 10.24:
overflow offtake with orifice-control ("overlaat-aftapping met onderspui-stuw"),
proportional divisor ("proportionele verdeler"),
proportional divisor with orifice-control ("proportionele verdeler met onderspui-stuw").
The design of these configurations will be discussed in the next chapter.
W E I R ' with
free flow
ongomg
imomiag
'FLOOD
SCREEN'
OVERFLOW OFFTAKE
with Orifice-Conh-ol
WEIR' with
'WEIR' with
^
s^free flow p H B i d ^
free flow
ongoing
j'nuvmag
incoming
tmg
ongomg
channel
'WEIR' with
T L O O D SCREEN
'WEIR'
PROPORTIONAL DIVISOR
PROPORTIONAL DIVISOR
with Orifice-Control
Figure 10.24. Types of passive diversion structures.
Practical configurations for 'gated' diversion structures. Also three configurations are
practical for 'gated' diversion structores, see also figure 10.25:
gated offtake with weir-control ("regelbare-aftapping met overlaat-stuw"),
overflow offtake with gated-control ("overlaat-aftapping met regelbare-stuw"),
gated offtake with gated-control ("regelbare-aftapping met regelbare-stuw").
The design of the first two configurations will be discussed in the next chapter. The third
configuration is discussed below.
Gated offtake with Gated-control. A 'gated offtake with gated-control' ("regelbare-aftapping
met regelbare-stuw") hnplies the application of two regulators with gates, both at the offtake
and at the control, see also figure 10.25.
In irrigation, the configuration of two regulators with gates is widely applied. The
regulator at the offtake is equipped with a measuring device ("meetinrichting"). This can be
208
achieved by hydraulic adjustments of the gate (e.g. Romijn weir, Crump-deGruyter gate,
etc.), or by constructing a separate measurement structure in series (e.g. broad-crested weir,
Cipoletti weir). The resuh is that the regulator at the 'offtake' becomes a discharge regulator,
and the regulator at the 'control' is a water level regulator. This can be translated into clear
instructions for the operator.
A 'gated offtake with gated-control' is normally not a good configuration in flood
control, as discharge regulation of floods is troublesome. It is difficult to measure the
discharge with a vertical or radial gate, and the application of separate measurement
structures in flood control is prohibited by its high costs.
So, the gated offtake of the flood diversion may become a (upstream) water level
regulator. Such a configuration of two water level regulators will confuse the operator, as he
does not know which of the regulators need adjustment to obtain the target water level. It
means, there is an instability of diversion: many gate settings, and so flood diversions, are
possible to maintain the target water level.
However, there might be solutions to overcome the above problems with the
configuration of 'gated offtake with gated-control' in flood diversion:
the assessment of a rating curve for the discharge regulator, which is accurate and simple
for the operator;
the application of Crump-deGruyter gates at the discharge regulator;
to operate the discharge regulator as a 'downstream' gate, i.e. the target water level is
at the downstream side of the gate. The effect is that the downstream water level is related
to the diverted discharge by the Strickler formula. It is obvious that such relationship is
quite inaccurate for changing river dimensions AFp"^^ and for a changing roughness
coefficient k.
A G A T E S ' witli
free flow
mcoming
channel
IE
otigomg
chaimel
'WEIR'
GATED OFFTAKE
with Weir-Control
' W E I R ' with

free flow
^GATES' with
free flow
ongimg
channel
mumiing
iamclM
/coming
.y/^^/^jj./
'GATES'
OVERFLOW OFFTAKE
with Gated-Control
Figure 10.25. Types of gated diversion structures.
mgomg
cbmmi
'GATES
GATED OFFTAKE
with Gated-Control
11. DESIGN EXAMPLES OF DIVERSION STRUCTURES

11.1. Design of Free-Offtakes
11.1.1. Overflow Free-Offtake
Configuration. A n 'overflow free-offtake' ("overlaat vrije-aftapping") consists of one weir
located at the offtake and no structure at the control, see figure 11.1. The 'side-channel
spillway' belongs to this type and has been discussed in sub-section 7.3.2 of chapter 7.
offtaking
cliannei
diversion depends
on water level
in ongoing channel
WEIR' with
free flow
mcoininf
100
200
incoming
300
cliannei
discharge in mVs
CONFIGURATION
PERFORMANCE
Figure 11.1. A n 'overflow free-offtake'.
209
210
Performance. A n 'overflow free-offtake' is applied when the water depth in the ongoing
channel should not exceed a certain value y^^^, see also figure 11.1. The weir is a 'passive'
water level regulator, and diverts a discharge when the water level exceeds the crest level.
It should be noted that this configuration is based on 'water level control', and that the
distribution of flow depends on the water level ^target
ongoing channel. However, the
discharge at this target water level ^target
determined by the depth-discharge relation
according to the Strickler formula. Thus, the actual discharge changes for changing values
of the wet cross-sectional area A and the hydraulic radius R, and for changing values of the
Strickler coefficients k during the growing season.
It has been shown in sub-section 7.3.2 of chapter 7, that the side-channel spillway is
normally not a good structure for diverting flow, because:
the length of the offtaking weir may become unrealistic long;
the target water level yongoing cannot be maintained, as an additional head is required for
flow over the offtaking weir;
the diversion of the discharge of the side-channel spillway depends on hydraulic changes
in the ongoing chaimel, such as changing roughness and changing cross-sections;
the side-channel spillway creates a negative backwater curve (drawdown) in the upstream
channel.
Design variables. The design requires that two water levels in the ongoing channel should be
known: the target water level ytarget
which diversion starts, and the maximum water level
>'max during the diversion of the maximum (design) discharge. Furthermore, the crosssectional areas of the channel are required, so that the velocity v can be calculated for a
known discharge Q. It is assumed that the energy drop at the offtake is always sufficient
for free flow.
Two design variables have to be designed, see figure 11.2:
(i)
the width
of the weir at the offtake to the offtaking channel,
(ii) the crest-height p^ of the weir at the offtake.
Moreover, the following hydraulic parameters are also unknown:
(iii) the energy head //max during the maxhnum discharge in the ongoing chaimel,
(iv) the water depth y^ in the channel, at an upstream distance x from point Z,
(v) the water velocity
in the chaimel, at an upstream distance x from point Z ,
(vi) the energy head /T^ above the weir-crest at the offtake, at an upstream distance x from
point Z ,
(vii) the offtaking discharge x over a width Aft of the weir, at an upstream distance x from
point Z .
Design conditions. There are seven design conditions required to determine the two design
variables and the five hydraulic parameters, see also sub-section 7.3.2 of chapter 7:
condition 1. Diversion into the offtaking channel starts only when the water level becomes
higher than the target water level ytarget
crest-height p^ of the weir at the offtake
follows from:
Pt ~ >'target
condition 2. The energy depth in the channel during the maximum discharge follows from:
^max = W
+ ( W ) /2ând remains constant in upstream direction because the friction losses can be ignored.
condition 3. The water depth y^ in the channel at a location x during the maximum
(design) discharge follows from:
= ^max - (Vx)^/2g
211
11. D E S I G N E X A M P L E S O F D I V E R S I O N STRUCTURES
t
A4>
Cross-section A-A: THE OFFTAKE
^
y ONGOING
PLAN
(not to scale)
ONGOING
H|
Cross-section B-B: 'NO' CONTROL
Figure 11.2. Design variables of the' overflow free-offtake'.
condition 4. The energy head //^ above the weir-crest at a location x depends on the 'bend
loss' due to a sudden flow deflection ("plotselinge verandering van stroomrichting"). It is
assumed that this 'bend loss' equals 1.0 v^/2g, so that at location x:
=)'x
condition 5. The discharge formula for the offtaking weir of a width Ab at a location x
is:
= cAb H^'^ for z > T^^X
condition 6 and condition 7. The width b^ of the offtaking weir depends on the cumulative
offtaking discharge D AQy^, and by checking it with the required offtaking discharge Q^.
So: ftt = E Aft, when E A x = Qv
Design process. A n example of the design process of an 'overflow free-offtake', i.e. the sidechannel spillway, is presented in sub-section 7.3.2 of chapter 7.
Operation. Operation of the overflow free-offtake is not needed, and is even not possible.
Any approaching flow above a certain value, is diverted over the spillway. However, it is
unavoidable that the continuing flow to the ongoing channel is increased as well.
11.1.2. Gated free-offtake

Configuration. A 'gated free-offtake' ("regelbare vrije-aftapping") consists of a gated offtake
and no strucmre at the control, see figure 11.3.
Performance. A 'gated free-offtake' is applied when the water depth in the ongoing channel
should not exceed a certain value ytarget'
figure 11.3. The offtake is a manual or a
hydro-mechanical water level regulator, and diverts a discharge when the water level exceeds
the target level ytarget'
figure 11.4.
212
offtakmg
chamiel
diversion depends
on water level
in ongoing channel
/''
.0
OATES' wii
free flow
offtaking
ciiannel
mcotiung
dmnd
100
200
incoming
300
channel
discharge in m^s
CONFIGURATION
PERFORMANCE
Figure 11.3. A 'gated free-offtake'.
Figure 11.4. A n artist impression of a 'gated water level regulator' (Nedeco 1973).
11, D E S I G N E X A M P L E S O F D I V E R S I O N STRUCTURES
213
Like the 'overflow free-offtake', the diversion of discharges depends on the water level
3'target
*e ongoing channel. However, the discharge at this target water level ^target is
determined by the depth-discharge relation according to the Strickler formula. Thus, the
discharge changes for changing values of the wet cross-sectional area A and the hydraulic
radius R, and for changing values of the Strickler coefficients k during the growing season.
The 'gated free-offtake' is an alternative for the side-channel spillway. The 'sidespillway effect' can be ignored because of the energy head H is quite large. It means that the
width ftt of the gated free-offtake is much smaller than the length of the side-channel spillway.
The disadvantages of the 'gated free-offtake' are:
the diversion of discharge by the free-offtake depends on hydraulic changes in the ongoing
channel, such as changing roughness and changing cross-sections;
the gated free-offtake creates a negative backwater curve (drawdown) in the incoming
charmel.
Design variables. The design requires that the maximum (design) water level ytarget
^he
ongoing channel should be known. Furthermore, the cross-sectional areas of the channel are
required, so that the velocity v can be calculated for a known discharge Q. It is assumed that
the energy drop at the offtake is always sufficient for free flow. There are no 'side-spillway
effects'. The sill level of the offtake is at the channel bed level, so
= 0.
Two design variables have to be designed, see figure 11.5:
(i)
the width b^ of the gates at the offtake to the offtaking channel,
(ii) the maximum orifice opening
at the offtake.
(iii) the energy head /^target in
ongoing channel for the target water level,
(iv) the energy head
at the offtake.
Cross-section B-B: 'NO' CONTROL

Figure 11.5. Design variables of the 'gated free-offtake'.
Design conditions. There are four design conditions required to determine the two design
variables and the two hydraulic parameters:
condition 1. The energy depth in the channel for the target water level follows from:
^target
>'target + (^target)^/^^
and remains constant in upstream direction because the friction losses can ignored.
214
condition 2. The energy head

above the sill of the offtake equals the energy depth in
the chaimel as there are no 'side-spillway effects':
~ -^target
condition 3. The depth-discharge relation of the offtake during the maxhnum offtaking
discharge follows the discharge formula of the 'notch':
t = 1.8 ftt //t^^^ for k
condition 4. The (minimum) orifice opening

depth at the offtake:
at the offtake depends on the critical water
Figure 11.6. A n example of a 'gated free-offtake' on a meandering river.
11. D E S I G N E X A M P L E S OF DIVERSION STRUCTURES
215
Design process. A n example of the design process of a 'gated free-offtake' is presented in

box 11.1. A n example of a 'gated free-offtake' is shown in figure 11.6, where it is applied
for flood diversion from a large meandering river within a flood plain.
Operation. Operation of the gated free-offtake is required, and is based on 'water level
control'. The gates are opened or closed in such a way that the target water level in the
ongoing channel is maintained. The acmal discharge in the ongoing channel is not known, but
can be estimated with the Strickler formula.
Box 11.1. Design example of the 'gated free-offtake'.
The setting. A 'gated freeôfftake' has to be designed in an existing river. The design
liver discharge amounts to
^ 500 m^/a. Tl\c capacity of the continuing down-river
is 300 m-^/s, with a uniform water depth y = 3.9.3 m and a velocity v = 2.32 m/s. So,
the offtaking discharge C>t - 200 m^/s.
Design. The design of the gated free-offtake weir is based on the following conditions:
" condition 1. The enercv depth in the river follows from:
ûuget = >vm.i + 0'turgct>'/2^ = ^-^'^ + ^"^22/2^ = 4.20 m
condition 2. The energy head H^ above the sill of the offtake equals the encrg>'
depth in the river as there are no 'side-spillway effects';
condition 3. T h e depth-discharge relation of the offtake during the maximum
ofltaking di?chargc foHows the discharge formula of the 'notch*: Q^ ~ L8
H(''~
or: 200 = 1.8 l\ 4.20^'-. So, the width l\ of the gated offtake is calculated at ft^ -
lIMBIB^liHiBiiilililB^^^^
condition 4. The mmimuni otificeôpening \s\ at the offtake depends on the critical
water depth at the offtake: w\ =^ -/, H^ = V3 x 4.20 = 2.80 ni.
11.2. Design of Passive Diversion Structures

11.2.1. Overflow offtake with Orifice-control
Configuration. An 'overflow offtake with orifice-control' ("overlaat-aftapping
met
onderspui-stuw") consists of a 'weir' as offtake and a 'flood screen' ("bandjirscherm") as
control, see figure 11.7.
Performance. The 'overflow offtake with orifice-control' has passive regulators. A t lower
discharges when the 'flood screen' is not touched by the water, the water is flowing
undismrbed into the ongoing channel. At higher discharges, the orifice flow-type and the
corresponding contraction, generates extra headlosses. Thus, the control becomes a water
level regulator and forces the water to flow also over the overflow offtake. However, the
'passive' flood screen cannot avoid that the discharge
towards the ongoing channel
216
increases at the same time, due to the increasing (upstream) water levels, see also figure
11.7.
The 'side-channel spillway with flood screen' ("zijdelingse overlaat met bandjir
scherm") belongs to this configuration, and has been discussed in sub-section 7.3.2 of chapter
7. Certain aspects of this struchire will be discussed here.
offtakmg
charmel
diversion depends
on water level
in ongoing channel
offtaking
channel
'WEIR' with
V free flow
ongmng
mt^c&diuei
incomiug
clmnnel
'FLOOD
SCREEN'
100
200
incoming
300
channel
discharge in m^s
CONFIGURATION
PERFORMANCE
Figure 11.7. A n 'overflow offtake with orifice-control'.
Design variables. The design requires that the water level ytarget in the ongoing channel at
which diversion starts, is known, as well as the corresponding headloss z^ over the control.
The 'control' is constructed in an existing channel. The sill of the control is at the
charmel bed level, so
= 0. The cross-sectional areas of the chaimel are required, so that
the velocity v can be calculated for a known discharge Q. It is assumed that the energy drop
Zt at the offtake is always sufficient for free flow, and that there are no 'side-spillway effects'.
The maximum water level y^^y^ or the corresponding discharge c max during the
diversion of the design (maximum) discharge may be required for a final check of the design.
Four design variables have to be designed, see figure 11.8:
(i)
the width b^ of the orifice at the control to the continuing chaimel,
(ii) the width b^ of the weir at the offtake to the offtaking charmel,
(iii) the orifice opening
at the control,
(iv) the crest-height p^ of the weir at the offtake.
(v) the energy head //target during the target water level in the ongoing channel,
(vi) the upstream energy head //ôm during the dominant discharge in the incoming chaimel,
(vii) the energy head
above the sill of the offtake during the dominant discharge,
(viii) the offtaking discharge t,dom during the dominant discharge,
(ix) the continuing discharge c.dom during the dominant discharge.
Design conditions. There are nine design conditions required to determine the four design
variables and the five hydraulic parameters:
condition 1. The orifice-opening w^, of the control is slightly above the target water level
3'target
which diversion should start:
^c ~ >'target'
217
Cross-section B-B: THE CONTROL
Figure 11.8. Design variables of the 'overflow offtake with orifice-control'.
condition 2. The energy depth in the channel at which diversion should start, follows
from:
^target = ^'target + (^target)
î
condition 3. The orifice opening

at the control depends on the headloss z under
submerged flow during the target water level in the ongoing chaimel:
^target =
^'^S
o with V = QJib^x
^target)'
open-water flow at the target water depth ytargef with
^target < ^^c '
O with V = c/(^c ^ ^c) for orifice flow, with ytarget > ^c !

condition 4. The crest-height
of the weir at the offtake follows from:
Pt ~ -^target ^ ^target
condition 5. There are no backwater effects by the stracmre during the dominant
discharge, so the upstream energy depth H^QJ^ follows from:
^dom "~ ûniform
condition 6. The energy head H^ above the sill of the offtake, equals the energy depth
in the channel as there are no 'side-spillway effects':
So, the energy head Ht dom above the sill of the offtake during the dominant discharge is:
^t.dom = -^dom 'Pt
condition 7. The depth-discharge relation over the weir at the offtake during free flow is:
t = Ct ftt Hi^'^' for z > 7 Hi
So, the offtaking discharge t^dom during the dominant discharge follows from:
et,dom =
c^h^ôm-Ptr
condition 8. The depth-discharge relation of the control for orifice flow is:
Qc =
{ l i j ; ^
b^
{ i n ; foxH^ >
^l^w,
see also figure 8.10 of chapter 8 for the value of Cj^;
So, the continuing discharge c,dom during the dominant discharge follows from:
dom
218
condition 9. The width b^ of the weir at the offtake to the offtaking channel, can be solved
from the continuity of flow
dom ~ t,dom
c,dom
checks. The following checks can be made on the above design:

o the discharge
into the ongoing channel during the design discharge 020
m-^/s can be calculated by determining first the upstream energy head H^^^ from the
continuity of flow during the design (maxhnum) discharge:
and by entering the upstream energy head H^^^ into the discharge formula:
cmax
=
^2gH^^
;
o the assumed values of C^ of the orifice formula can be checked with the values of Cj)
in figure 8.10 of chapter 8.
Design process. A n example of the design process of an 'overflow offtake with orificecontrol' is presented in box 11.2.
Operation. Operation of the 'overflow offtake with orifice-control' is not needed, and is even
not possible. At lower discharges, the water is flowing undistorbed into the ongoing chaimel.
At higher discharges, the control forces the water to flow over the overflow offtake.
However, the discharge towards the ongoing channel increases at the same time.
Box 11.2. Design example of the 'overflow offtake with orifice-control'.

The setting. An "overflow offtake with orifice-control' has to be designed in an existing
river. Tlie design discharge amounts to
=
"^^/s, and the maximum capacity of
the down-river is 2 0 0 m^/s.
The flood diversion should start at a water depthyt^êi -"^^^^
ongoing
channel with a cturesponding discharge 0 = 120 m- As antTvelocitx' v = 2.04 m/s. The
headloss of the eontiol should be limited to 2 = 0.10 m during the ()ni?oini; " ^^0
m^/s, and follows from 2 = 0.25 v'^lTg.
The dominant discharge of the upstream river is 300 nrVs, with an uniform water
idepthy ~ 3.93 ni and a velocity v - 2.32 m/s,
Design. The design of the overflow offtake with orifice-control is based on the
lollowing conditions:
condition 1. The oriflce-opening it'^, of the control is slightly above Viaj-gct- ô:
* condition 2. Tlie energ>' depth in the river at which diversion should start, follows
iiln:
vliiliiBiiî^Blii^Blii^
^target = >'tan.ci + 0'targct)^''^ ^ 3 J 3 -f im^jlg

^ 3.34 m.
conditiiin 3. The'headloss i = O.IO m over the control under submerged flow, so;
= 74 v-dg with y = Qlib^ X y
)
I - = (8.1,'x O . I O ) ' ' - ^ 2.80m/s,
hence: the width of the orifice b^ =^= 120 / (3.13x2.80) ^ 13.69 nt.
" condition 4. The crest-height
of the weir at the offtake follows from;
* condition S. Tslo backwater cun'e at Ojom'
^ d o m - ^ûniform ^ ^-'^^ ^ ^ ^ ^ S
(to hi' cont'mwtd in fhr next box)
= 4.20 m.
219
Box 11.2 (continued).
Design example of the 'overflow offtalce with orifice-control'.
condition 6. The energ>' head //mom

^^^^ ^"^^^
off^'iê during the
dominant discharge equals, as there are no "sidc-spilKvay effects';
Ik...
- ^/dom - Pi = ^-20 - 3.44 = 0.76 m.
condition 7. The offtakmg discharge over the wen- at the otitake during the
dominant discharge is;
t.dnm = ^\ f\ ( ^ o m " PO''" ^ 2.1 />, 0.76^^ - 1.39
condition 8. The continuing di-scharge through the control during the dominant
C^,.... V 2 i ^ - " ^'-^ ^ ^^-^^ ^ ^ - ' ^ v 2 i ^ n 3 0 - 156 m >

with Cj-) ~ 0.4 during the dominant discharge, or with C'^ = 0,5 during the
maximum discharge;
condition 9. The diverted discharges into the diversion channel and the ongoing
river during the dominant discharge equals Q^^,^^, so:
300 = 1..39/>j + 156
Hence the width /?, of the offtaking weir is /?{ = 104 m.
Check. The discharge Qc.mju
^^""^ ongoing channel during the design discharge (J;,^
= 500 vs?is follows Irom" the two discharge formulae;
500 = 2.1 X 104

-44 .
- 3.44}'m
0.5 x 13.69 X 3.13 /2^;'- x //
_ 0.5 X 13.69 X 3.13 ^Tg
I 2.1 X 104
r j f
2.1 X 104
//nx
3.44 . (2.29 ~ 0.434
f
This equation is solved by iteration: assume //^^^^ - 4.20 m and calculate //^.^^ 4.69 m. And again: assume H^.^.^^ = 4.69 m and calculate H^^.^ = 4.66 m. And Anally:
assume //, ,,^ = 4.66 m and ca culatc //,n.j,. - 4.66 m. So, H^.^
- 4.66 m.
The discharge Cî,g^,i,jj, into the ongoing channel during the design discharge foUow.s
iiiph"
: liiiiiiiiiiHHiiBlH^
<^.,u.n>: - ô ^ K v " 2 T ^ .
0.5 X 13.69 X 3.13
So that n,.,,it)g ~
which is acceptable.
11.2.2.
Proportional
x 4.66
Diversor
Configuration. A 'proportional divisor' ("proportionele verdeler") consists of two weirs, one

as the offtake and one as the control, see figure 11.9.
Performance. A proportional divisor is applied when an incoming discharge has to be splitted
into an ongoing chaimel and an offtaking flood channel by a fixed ratio, see figure 11.10.
The weirs are 'passive' regulators, and the discharge is regulated into a fixed ratio. The
water level cannot be regulated. Thus, a proportional divisor is a pure type of discharge
control: the primary purpose of the strucmre is to maintain a certain ratio-in-discharges to
a downstream destination.
220
offtakmg
chamiel
CONFIGURATION
PERFORMANCE
Figure 11.9. A 'proportional divisor'.
Figure 11.10. 'Fayoum weir' as a proportional divisor in irrigation.
Design variables. It is assumed here, that the 'control' is constructed in an existing chaimel.
The depth-discharge relation of the ongoing channel should be known. Furthermore, the
cross-sectional areas of the channel are required, so that the velocity Vojjgoing ^^^^ the energy
head //ongoing ^^"^
calculated for a known discharge . It is assumed that the energy drop
Zf at the offtake is always sufficient for free flow. There are no 'side-spillway effects'.
(i)
the width
of the weir at the control to the continuing channel.
(ii) the width

of the weir at the offtake to the offtaking channel,
(iii) the crest-height
of the weir at the control, and
(iv) the crest-height
of the weir at the offtake.
(v) the energy head i?ongoing
the ongoing channel,
(vi) the energy head
above the weir-crest at the control, and
above the weir-crest at the offtake.
Figure 11.11. Design variables of the proportional divisor.

Design conditions. There are seven design conditions required to determine the four design
variables and the three hydraulic parameters:
condition 1. The ongoing discharge ongoing follows from the fixed proportional ratio n
between the offtaking discharge t and the approaching discharge incoming = t + c
reads:
n
100%
where n is a constant, e.g. n = 50% for an equal splitting. So:

ongoin. = (100% "
>< incoming
condition 2. The energy depth in the channel for a discharge ongoing follows from:
ôngoing ~ >'ongoing + (ôngoing)
. ,
condition 3. There is just 'free flow' over the weir of the control during e.g. the diversion
of the dominant discharge:
Zc > yH^ becomes or:
+ p^ = /ôngoing + 7
condition 4. There should be 'uniform flow' during the dominant discharge in the
incoming channel. So, the energy depth at the control equals the uniform energy depth
ûniform 1" ^1^^ incoming channel during the dominant discharge domHQ
Pc ~ ûniform
, ,
i.
condition 5. The depth-discharge relation over the weir at the control during free flow is:
e =
b, H^^'\ for z > y^H^
condition 6. Proportional diversion is only possible when the crest-height p^ of the weir
at the control is equal to the crest-height
of the weir at the offtake: p^ = p^
condition 7. The depth-discharge relation over the weir at the offtake during free tlow is:
t = q ftj H,^'^, for z > y^H,
222
Box 11.3. Design example of the 'proportional divisor'.

The setting. A flood diversion structure has to be designed in an existing river. The
design river discharge amounts to (?2{) =
nrVs. The dominant discharge of the river
is assumed at Q2 - 300 nr^'sK with a unifonn water depth v - 3.93 m and a velocity
I ' = 2.32 m/s. The capacity of the continuing downriver is 200 m-'/s. A discharge of 120
m-^/s in the ongoing river leads UJ a (uniform) water depth > - 3.13 m and a veiocit)-
IliwiBiilliill^^
The new flood diversion channel will transport a fixed ratio of 600T of the river
discharge. The downriver discharge
should reach the capacity of 200 m-'/s al the
design discharge C?2o- 'Hiete should be (jusi) 'free flow" over the weir of the control
during the dominant discharge Q^^^^^y
Design. The structure is designed with two sharp-crested weirs with c = 2 . 1 m^'-'^/s. at
an equal crest-height > above the bed. Free flow over the wch occurs at the headloss
lill(BlllHillllillBIIM^
'
condition 1. The ongoing discharge ongoing follows from the fixed proportional
ratio n between the ofltaking discharge
and the approaching discharge t+C?c
lilliBiilliilBilliiii^^
X 1009?. - 60%
/;
So for the dominant discharge Q^^^^, Qanôim = (100-^'' 0%) X C?ij,coming ^ -Ô
X 300 - 12m^/s,
condition 2. The energy depth in the river during the ()^. = 120 m'Vs follows from:
^ônômu = ônôm^ + i'orrôm^'^^S - 3.13 + 2.04-/2g - 3.34 m.
condition 3. Thci^- i.s just 'free flow" over the weir of the control during the
diversion of the dominant discharge:
>
/ / , becomes:
+
- /ônuoing +
^'^c r: %
+
- 3.34 m.
upstream river channel. So, the cnerg>' depth at the control equals the uniform
energy depth l/um[mra i " ^1^*^ upstream river channel during the dominant discharge
H, +
=
^ 3.93 + 2.32"/:??? - 4.20 m .
Subtracting this equation from the above equation of condhion 2: 7^
0.86.
Thus,
= 1.72 and the sill level p^ = 4.20 - 1.72 = 2.48 m.
condition 5. The depth-discharge relation over the weir at the control during tree
llliiw is:'
B
' ^
^ 2.1
H^^--. for z > V3 H^, or: 120 - 2.1 b^ \J2^'\ so that the weir a l the
control has a width b^. ~ 25.33* m.
condition 6. Proportional diversion is only possible when the crest-height
of the
weir at the control is equal to the crest-height
of the weir at the offtake:
p^. , so the offtake as a sill height p^ - 2.48 m.
* onditlon 7.1"hc depth-discharge relation over the weir at the offtake during free
= c b^ / / / - , or: 180 = 2 1
1.72^'% so that the weir at the control has a
width i), = 38.00 i, (or alternatively: b^ ^ fi x b^^ 15 x 25.33 - 38.00 m).
223
11. D E S I G N E X A M P L E S OF D I V E R S I O N STRUCTURES
Design process. An example of the design process of a proportional divisor is presented in

box 11.3.
Operation. Operation of the proportional divisor is not needed, and is even not possible. Any
approaching flow is diverted into the fixed ratio n.
Application. Applications of the proportional divisors can be found in river engineering and
in flood control. Rivers in deltas are often diverting in several river branches. Only the
application of a proportional divisor guarantees a fully-proportional diversion.
Proportional divisors are widely applied in irrigation. The most simple type of
proportional divisor in irrigation is the Fayoum weir. The Fayoum weir has been developed
in Egypt in the 1920s. The Fayoum weir is basically a broad-crested weir with a crest-height
p above the bed, and a crest-width b between vertical abutments, see also figure 11.10.
Both underflow? The same performance could have been obtained by a proportional divisor
with a configuration of two orifices with underflow.
However, such a configuration requires a higher headloss z V2 H as the energy head
H is large, e.g. z ~ 1.50 m for a canal-depth y^ = 3.00 m. Moreover, the height w of the
opening of the orifice should be taken quite small, in order to guarantee underflow also for
the smaller discharges. Thus, a proportional divisor with two orifices should not be applied.
11.2.3. Proportional divisor with Orifice-control

Configuration. A proportional divisor with orifice-control ("proportionele verdeler met
onderspui-stuw") consists of a 'weir' as offtake and a 'weir with flood screen' ("bandjirscherm") as control, see figure 11.12.
ofaking
channel
ongoing
'WEIR' with
'FLOOD-SCREEN'
100
200
incoming
300
channel
discharge in m^s
CONFIGURATION
PERFORMANCE
Figure 11.12. A 'proportional divisor with orifice-control'.
224
Performance. The 'proportional divisor with orifice-control' is composed of passive

regulators. The structure is identical to the proportional divisor at the lower discharges when
the water does not touch the flood screen. At the higher discharges, the control functions as
an orifice with a high sill.
A proportional divisor with orifice-control is applied when the discharge in the incoming
channel has to be splitted into the ongoing channel and the offtaking flood channel by a fixed
ratio, but only to a maximum discharge cap.
ongoing channel, see also figure 11.12.
Design variables. It is assumed here, that the 'control' is constructed in an existing channel.
The maximum water level yongoing
ongoing channel should be known. Furthermore,
the cross-sectional areas of the channel are required, so that the velocity Vojjgoijjg can be
calculated for a known discharge Q. It is assumed that the energy drop
at the offtake is
always sufficient for free flow. There are no 'side-spillway effects'.
Five design variables have to be designed, see figure 11.13:
(i)
the width
of the weir at the control to the continuing chaimel,
(ii) the width b^ of the weir at the offtake to the offtaking channel,
(iii) the crest-height
of the weir at the control,
(iv) the crest-height Pi of the weir at the offtake,
(v) the orifice-opening
at the control.
(vi) the energy head /ôngoing
ongoing channel,
above the weir-crest at the control,
(viii) the energy head
above the weir-crest at the offtake.

Figure 11.13. Design variables of the 'proportional divisor with orifice control'.
Design conditions. There are eight design conditions required to determine the five design
condition 1. The ongoing discharge ongoing fohows from the fixed proportional ratio n
between the offtaking discharge t and the approaching discharge incoming
t+c
reads:
11. D E S I G N E X A M P L E S O F D I V E R S I O N S T R U C T U R E S
225
100%
where n is a constant, e.g. n = 50% for an equal splitting. So:

ongoing = (100% - ) X incoming
condition 2. The energy depth in the channel for a discharge ongoing follows from:
ôngoing ~ ^'ongoing
(ôngoing)
condition 3. There is just 'free flow' over the weir of the control during e.g. the diversion
of the dominant discharge:
Zc ^
becomes:
+
= /ôngoing + T
condition 4. For instance, there should be 'uniform flow' during the dominant discharge
in the incoming charmel. So, the energy depth at the control equals the uniform energy
depth flujyfoj.j^ in the incoming chaimel during the dominant discharge domPc " -ûniform
condition 5. The depth-discharge relation of the control under free flow is:
o for weir flow, so JEL <
w^:
c =
H}'^, for
z>y^H^
o for orifice flow, so H,. >
w^,:
Qc = C^b^w^{2iJ:
= C ^ b ^ w ^ { ^
see also figure 8.10 of chapter 8 for the value of C^;
condition 6. Proportional diversion is only possible when the crest-height p^ of the weir
at the control is equal to the crest-height p^ of the weir at the offtake:
Pc =Pi
condition 7. The depth-discharge relation over the weir at the offtake during free flow is:
t = Ct ^t H?'^, for z > 7t / / t
condition 8. The proportional diversion should be transformed into an orifice flow at the
control for a certain incoming channel discharge rivermeans that the nappe touches
the flood screen. Hence, the critical water depth y^ = Ij^
equals the opening-height
during the ongoing discharge (1 - n) riverAn additional condition may added: the discharge c into the ongoing channel should not
exceed the capacity cap of the ongoing chaimel, so:
C ^ cap.
Design process. A n example of the design process of a proportional divisor with orificecontrol is presented i n box 11.4.
Operation. Operation of the proportional divisor with flood screen is not needed, and is even
not possible. Any approaching flow is diverted into the fixed ratio n upto the capacity of the
ongoing channel is reached.
Application. Applications of the proportional divisors can be found in flood control. The
proportional divisor with flood screen is also applied in irrigation at the headworks ("watervang, prise d'eau, hoofd-inlaatwerk").
However, the flood screen ("bandjirscherm") is
constructed at the offtake, as to limit the intake discharge into the prhnary irrigation canal
during river floods.
226
Box 11.4. Design example of the 'proportional divisor with orifice-control'.

The setting. The Hood diversion strucime of the above example was a 'proportional
divisijr" to divert the incoming river discharge upto
-''Ô nrVs in a fixed ratio of
6(K7 into a new flood diversion channel. So, the downriver discharge C\- should reach
the capacity of 2 0 0 nr'/s at the design discharge Q2Q.
It was calculated that the offtaking weir should have a width
= 38.00 m, the
weir at the control should have a width b^. = 25.33 m, and that both weirs should have
an equal height p = 2.48 m. The encrg>' head H above the weir crest was calculated
at Jl^ - 1.72 m during the diversion of the dominant discharge j o m '
Now, the additional objective is that the diversion of Hoods between 20 = 500
nr'/s and
= 700 wr'k should be diverted: (i) into the downriver up lo Q'^^ i= 200
m~\'s and the remainder into the new diversion channel. Therefore, the above
proportional divisor will he transformed into a 'pioporttonal divisor with orificecontrol' by constructing a flood screen above the weir at the control.
Design. The additional design variable is now the orifice opening H' at the control
between the weir crest and the flood screen.
condition I . The depth-discharge relation for overflow of the control under free flow
i i i l i i i i i i i i ^ i i i l i l i ^ p
^ 2.1 b^. U^}'-, so 0.40 X 5 0 0 = 2.1 X 25,33 fl^?''\ and

^ 2.42 m
condition 2. The proportional ^diversion is into an orifice flow at the cHintrol when
the critical water depth y^. = ^/^ H^. - 1.61 m touches the flood screen. I l means
that the opening-height w^. - 1.61 ra.
Check. A check is made whether the downriver discharge
^ 200 nT /s is maintained.
' condition 2: the discharge formula of the control during these higher discharges
depend.s on the head
above the weir-crest and is related to the orifice flow, so:
=^ 2.3 b w H^^-- = 2.3 X 25.33 x L 6 1 x H^^^93.80 H^^'" condition 3; the discharge formula of the offtake during these higher discharge
remains the overflow formula with
= H , so:
(?t
2.1 / j , H^^'^ =^ 2.1 X 38.00 X
=^ 79.80 x
Rating cuive. The rating cuive for the structure cannot calculated directly, hut should
follow an iteration by assuming the cnerg>' head / . above the crest:
c for
= 2.50 m: i\. = 93.80 x 2.50'/2 = 148 m-^s and
= 79.80 x 2.50'^'- = 315
m\/s, so that the incoming river discharge is Qf-^y^^ = 463 nr^/s,
^ for
^ 2.75 m :
^ 93.80 x 2.75^'^ ^ 156 m'Vs and
^ 79.80 x 2.75^^^- = 364
nr/s, .so that the incoming river discharge is Qij,voi ^ ^^0 nr^/s,
" foxH^ = 3.0 m:
^ 93.80 x 3.00'^- ^ 162 m^s and Ot =^ 79.80 x 3.00^'^ = 415
ur^^/s, so that the incoming river discharge is Q-~ 577 m^/s.
for H^. = 3.50 m;
= 93.80 x 3.50^'175 m^Vs and
^ 79.80 x 3.50^'- ^ 523
m^/Sf so that the incoming river discharge is J^VOF ~
It is concluded that the additional objective has been met by constructing a flood
screen above the weir at the control.
227
11.3. Design of Gated Diversion Structures

11.3.1. Gated offtake with Weir-control
Configuration. A 'gated offtake with weir-control' ("regelbare aftapping met overlaat stuw")
consists of a gated offtake and a weir as the control, see figure 11.14.
Performance. A 'gated offtake with weir control' is applied when the discharge in the
ongoing channel should not exceed a certain value targef
^1^ figure 11.14. The offtake
is a manual or an hydro-mechanical water level regulator, and diverts a discharge when the
discharge exceeds the target discharge targef
offtaldng
channel
CONFIGURATION
PERFORMANCE
Figure 11.14. A 'gated offtake with weir-control'.
Design variables. The design requires that the maxhnum discharge target into the ongoing
channel and at which diversion starts, is known. The 'control' is constructed in an existing
channel, but should flow under free flow. The cross-sectional areas of the channel are
required, so that the velocity v can be calculated for a known discharge Q.
The sill of the gated offtake is at the channel bed level, so
= 0. It is assumed that
the energy drop at the offtake is always sufficient for free flow, and that there are no 'sidespillway effects'.
(i)
the width b^ of the weir at the control to the continuing channel,
(ii) the crest-height p^ of the weir at the control,
(iii) the width b^ of the gated offtake to the offtaking charmel,
(iv) the orifice opening w^ of the gated offtake.
228
Cross-section B-B: T H E CONTROL
Figure 11.15. Design variables of the 'gated offtake with weir-control'.
Figure 11.16. A 'gated offtake with weir-control' in irrigation ( D H V 1986).
229

(v) the energy head /ôngoing during the maxhnum discharge in the ongoing channel,
(vi) the energy head
above the sill of the weir during the maximum discharge,
(vii) the upstream energy head H^^^ during the dominant discharge in the incoming channel.
Design conditions. There are seven design conditions required to determine the four design
condition 1. The energy depth //ongoing
channel during a discharge ongoing into
the ongoing chaimel follows from:
ôngoing ~ ^'ongoing
('ôngoing) ^'^S
condition 2. There is just 'free flow' over the weir of the control during the dominant
discharge:
> Vi
becomes:
= z^. So: Vi
= (H^ +
- (//ongoing)
equation 1: Vi
+ p^ = //ongoing
condition 3. The energy depth //ôm in the channel during the dominant discharge dom
follows from:
^dom = >'dom + (Vdom)^/2g
incoming channel. So, the energy depth at the control equals the uniform energy depth
-ûniform i ^
incoming channel during the dominant discharge domequation 2:
+ p^ = /ûniform
The sill level
can be calculated by subtracting this equation from the above equation 1.
condition 5. The width
of the weir at the control into the ongoing channel is determined
by the maximum discharge in the ongoing channel:
e = c, b, H,^/^
condition 6. The width b^ of the offtake is determined by the maximum discharge into the
offtaking channel, while the energy head H^ =
p^ at the crest is determined by the
capacity of the ongoing channel. So:
t = l.Sf
{H^+pJ^'^
condition 7. The (minimum) orifice opening
at the offtake depends on the critical water
depth at the offtake:
Design process. A n example of the design process of a gated-offtake with weir-control is

presented in box 11.5.
Operation. Operation o f the gated offtake with weir-control is quite effective. The discharge
c to the ongoing channel is determined by measuring the energy depth
above the weir
crest and by applying the discharge formula of the weir. The gated offtake is regulated in
such a way that the ongoing discharge c does not exceed the target discharge target *o
ongoing channel. In fact, the regulation is transformed into a 'water level control', by
maintaining the energy head //target above the crest. In practice, the upstream water level may
not exceed the ytargetApplication. Applications of the gated offtake with weir-control can be widely found in flood
control. The configuration is also applied in irrigation at the headworks ("watervang, prise
d'eau, hoofd-inlaatwerlc"), see figure 11.16, and at secondary and tertiary offtakes.
230
Box 11.5. Design example of the 'gated offtake with weir-control'.

The setting. A ^gated offtake with weir control" has to be de.signed in an existing river.
The design river discharge amomits to G*2() - ^^^^ n'''''^^'
The dominant discharge of the upstream river is 300 ixr/a, with a uniform water
depth > = 3.93 ni and a vefocitj' r = 2.32 m/s. The capacity of the down-river is 200
nrVs, with an uniform water depth y = 3.13 m and a vtitjcity v = 2.04 m/s.
Design. There arc two general discharge f(jrmuiac:
- the depth-discharge relation over the weir at the control during free flow is:
" the depth-discharge relation of the ofl'take during the maximum offtaking discharge
follows the discharge formula of the 'notch':
= 1.8/jj//,3/^for2< > %H^
Furthermore, there are seven conditions to solve the seven unknown parameters;
ctmdition 1. The energy depth in the river during the maximum discharge ongoing
- 200 nrYs into the down-river follows from:
ôngoinj.
3'on.ing + 0ongoin)"/2/ - 3.13 + 2.04^/2^ = 3.34 m .
condition 2. There is just Tree flow' over the weir of the control during the
dominant discharge;
2, > './, ^7, becomes: ' / ,
= r , . So: '/^
= {fi, + p , ) - (/ôngoing)
^^
equation 1;
+
= /^ngoing
^
+ /^c ^ 3--4
condition 3, The encrgj- depth H^^^^ in the river during the dominant discharge
dom = -'00 nvVs follows from:
^^dom = .Vdom + 0'dom)'/^ ^ 3.93 + 2.32^/2^ - 4.20 m.
condition 4. There should be 'unifoim flow" during the dominant discharge in the
upstream river channel. So, the energ>' depth at the control equals the uniform
energy depth Wynifonn i " ^^'^ upstream river channel during the dominant discharge
equSion 2:
+
//H^^rm
+ /^c =^ 4.20.
Subtracting this equation from the above equation 1: 'A H^. ~ 0.86. Thus, li^ = 1.72
and the sill level p^ = 4.20 - 1.72 = 2.48 n i .
condition 5. The width
of the weir at the control into the ongoing river is
determined by the maximum discharge i n the ongoing river;
, = c^ ^>c ^c^'''
= 2.1 b^
so b^ ="42.22 m.
condition 6. The width b^ of the offtake is determined by the maximum discharge
into the offtaking channel, while the encrg>' head H^^
H^-^
at the crestis
determined by the capacity of the down-river. So:
0 , = L8 /?'t (//,. 4'
and 300 = 1.8 ZJ^ (1.72 + 2.48)^'"
Hence, the b^ - 19.36 m,
' condition 7. The (minimum) orifice opening
at the offtake depends on the critical
water depth at the offtake:
= "l^ (M^ + p^) so: H\ = -/3 (1.72 + 2.48) = 2.80 m.
231
11. D E S I G N E X A M P L E S OF D I V E R S I O N STRUCTURES
11.3.2. Overflow offtake with Gated-control

Configuration. A n 'overflow offtake with gated-control' ("overlaat-aftapping met regelbarestuw") consists of a weir as offtake and a gated regulator as the control, see figure 11.17.
Performance. The performance of the 'overflow offtake with gated-control' is basically equal
to the 'overflow offtake with orifice-corAroV. But here, the discharge to the ongoing channel
can be regulated.
offtaking
ciiannei
diversion depends
on water level
in ongoing channel
/'
/
'WEIR' w i t h .
free flow [ B W M |
offtaldng
channel
ongoing
channel
incoming
Mnnd
'GATES'
100
200
incoming
300
channel
discharge i n m^s
PERFORMANCE
CONFIGURATION
Figure 11.17. A n 'overflow offtake with gated-control'.
Design. Also the design of the 'overflow offtake with gated-control' is equal to the 'overflow
offtake with orifice-control', see figure 11.18. Therefore, reference is made to sub-section
11.2.1 of this chapter for the design process and a design example.
Figure 11.18. Design variables of the 'overflow offtake with gated control'.
232
12. DROP STRUCTURES AND STILLING BASINS

12.1. Introduction
12.1,1. Types of Energy
Dissipators
Need for energy dissipators. 'Energy dissipators' are required wlienever an excess of energy
would damage an unlined canal. This may be required, see figure 12.1:
when the gradient of the terrain is steeper than the maximum permissable slope of the
canal. Thus, a 'fall structure' is needed;
when a 'super-critical flow' may occur in a flow control structure, such as a water level
regulator, a discharge regulator, and a discharge measurement structure. Here, a 'stilling
basin' is needed.
Fall structure
FALL STRUCTURES
IN A CANAL
to avoid erosion of the canal bed
STILLING BASIN
A T A REGULATOR
to avoid erosion behind the structure
Figure 12.1. The danger of scouring requires energy dissipators.
233
234
Manner of energy dissipation. Energy dissipation can occur in different manners: (i) due to
friction along a sloping glacis, (ii) by impact of the flow against the floor, and (iii) by
mrbulence in the stilling basin. So, different types of energy dissipators can be distinguished,
see figure 12.2.
ENERGY
DISSIPATORS
to dissipate excessive energy
STILLING BASINS
F A L L STRUCTURES
in Caiial Reaches
iiiiiiiiitfiiiiiM
Flow Control Stractwres
Di-op StnicturevS
Vertical
Drop Stmctures
Chutes
Energy
dissipation
by:\ Energy dissipation
friction on glacis
turbulence in basin
by:
impact with floor

turbulence in basin
Inclined
Drop Structures
Energy dissipation
by:
turbulence in basin
Q-Measurement
Stiuctures
Discharge
Regulators
Water Level
Regulators
Energy dissipation
by:
turbulence in basin
Figure 12.2. Classification of energy dissipators.

Tvpes of fall strucmres. There are three types of fall stmcmres to dissipate the excess energy,
see figure 12.3:
A 'chute' or 'flume' dissipates the major portion of the energy by friction over the 'sloping
glacis' and the remainder portion by mrbulence in the stilling basin. Thus, the drop of the
energy level occurs not only in the stilling basin, but also along the sloping glacis.
Chutes are normally used where the drop in elevation is greater than + 5 meter and
where the water is conveyed over long distances. The gradient of the glacis is quite flat,
but yet steep enough to maintain super-critical flow.
A 'vertical drop stmcmre' has a vertical wall between the 'control' and the stilling basin.
The small portion of energy loss occurs by impact of the jet on the floor. The major
portion of energy loss occurs by mrbulence in the stilling basin.
Vertical drop stracUires can only be recommended for the smaller drops. Uplift pressure
for greater drops will require thick floors to prevent uplift.
A n 'inclined drop stmcmre' has a sloping glacis between the 'control' and the 'stilling
basin'. The energy loss on the sloping glacis due to friction is normally small and is
ignored, so that the energy is dissipated in the stilling basin only. The slope of the glacis
is chosen as steep as 2vert : iRor' but is usually lyen : iRor or lyert ^ôr
Stilling basins at flow control stmctores. It is a good principle to provide any stmctore where
super-critical flow may occur, with a 'stilling basin'. This stilling basin will force the
hydraulic jump within the stmctore, so that excessive scouring of the downstream channel bed
can be avoided.
235
12. D R O P S T R U C T U R E S A N D S T I L L I N G BASINS
energy
level
VERTICAL DROP STRUCTURE
INCLINED DROP STRUCTURE
Energy dissipation
Energy dissipation
by:
impact with floor

turbulence in stilling basin
by:
turbulence in stillmg basin
Figure 12.3. Types of fall structures for energy dissipation.
12.1.2. Design of Fall Structures and Stilling Basins

Components of fall structures. Fall structures and all other structures with super-critical flow
have two functional parts, each having its own typical design features, see figure 12.4:
the 'control', such as a 'control notch', a 'fixed weir', or a 'gate';
the 'stilling basin' such as the USBR standard basins, SAF-basin.
Moreover, a 'sloping glacis' may be applied.
the "Control"
Figure 12.4. Components of fall structures.
the "Control"
236
The Control. The water level at the top of a fall is normally regulated by a 'control' to
prevent drawdown, which may lead to scouring of the upstream canal, see figure 12.5.
The options for the regulation at the control of a fall structure are:
'fixed' regulation, by means of a weir-crest, a notch or a screen,
'manual' regulation, by means of gate,
'hydraulic regulation', by means of a hydro-mechanical gate.
WEIR
NOTCH
SCREEN
MANUAL
GATE
HYDRO-MECH.
GATE
Figure 12.5. Control types of fall structures.

The stilling basin. The most simple stilling basin is just a lowering o f the floor at the
downstream end of the structure. A n over-depth of e.g. 0.3 X the water depth y is often
sufficient. But normally, the design of stilling basins follows hydraulical formulae.
Rectangular stilling basins. The stilling basin should be 'rectangular', which means that
vertical walls should enclose the stilling basin, see figure 12.6.
The design formulae of typical stilling basins are two-dimensional. They are based on the
unit discharge q in m^/m.s, and provide the basin depth and the basin length, as well as the
sizes of any chute blocks, baffles and the end sill. Thus, the stilling basin should also be
constracted 'two-dimensional' with vertical walls, and not 'three-dhnensional' with sloping
sides.
Model tests have shown that the hydraulic jump in a trapezoidal basin is much less
complete over the total width than it is in a rectangular basin. Side eddies may spoil the
action of the basin, and the flow over the sloping sides may by-pass the stilling basin (USBR
1973).
Onion Outlet. One of the main problems in the design of any canal system is the fitting of
'smooth' lining with a Strickler coefficient k ~ 6Q m^^^/s to the 'rough' unlined canal with
a Strickler coefficient A: 30 m^''^/s.
The same problem arises at the end of a side-lining. The shear stress along the lining
has to adjust to the permissible shear stress of unlined side slopes, and 'scour holes' quickly
develop. These holes make that the lining progressively collapses. Even an end-coffer is often
not an ultimate solution.
The principle of the 'onion outlet' is an applicable solution, see figure 12.7. The onion
outlet allows for side eddies which do not damage the side slopes because of the low
velocities.
Thus, the vertical walls of the stilling basin should be extended after the stilling basin
to form an onion outlet, see also figure 12.7. The side slopes of the canal may or may not
be protected by rip-rap for extra safety.
237
CORRECT DESIGN:
Stilling Basin
with Iî?77Cî. Sidewalls
(no by-passmg of flow)
WRONG DESIGN:
Stilling Basin
with i Z G / W t ? Sidewalls
(by-passmg of flow)
Figure 12.6. A stilling basin should be enclosed by rectangular walls.
Figure 12.7. 'Onion outlet' of a stilling basin.
238
12.2. Hydraulics of stilling basins

12.2.1. Froude number
Froude number. The Froude number describes the state of the free-surface flow, and is
hnportant for the design of stilling basins. The flow is called 'sub-critical' or 'normal
flowing' ("stromend") when Fr < 1.0. The flow is 'super-critical' or 'shooting' ("schietend") whenFr > 1.0.
The Froude number is defined as the ratio of the stream velocity v to the wave velocity
V ' ( ^ ) , and is expressed by the formula:
,
Fr =
or
Fr
where: Fr is the Froude number, v is the velocity of the water in m/s, y is the water depth
in m, ^ is the discharge per unit width in m^/s, and g = 9.8 m/s^ = 9.8 N/kg is the gravity
acceleration.
Energy
Drop
z
Length of Jump L j
Upstream
Energy Head
Tailwater
Incoming
jet.
"Conjugate"
Depth
Downstream
Energy Head
HT
YMIN
Figure 12.8. Parameters of the hydraulic jump.
Hydraulics of inflowing jet. The type of stilling basin to be designed depends on the energy
of the incoming flow, the 'jet' ("straal"), expressed by the Froude number of the incoming
jet, see figure 12.8.
The water depth yj and the velocity Vj of the incoming flow into the stilling basin can
be calculated for a known discharge q per unit width and the energy head
from the two
equations:
2
where: q is the discharge per unit width in m^/s, Hj is the energy depth of the jet above the
bed in m, yj is the water depth of jet in m, Vj is the velocity of the jet in m/s, and g = 9.8
m/s^ = 9.8 N/kg is the gravity acceleration.
Iteration. The above equations can be solved by eliminating Vj, so:
v.=
'
^
(Hj - y.)
239
The value of can be determined through iteration. First, it is assumed that y-^ = 0, and a
new value of jj is calculated. This value is inserted in the formula and again a new value of
yj is calculated. The ultimate value of y^ is found after a few iteration cycles, see box 12.1.
Apparently, the Froude number of a falling flow Q depends on the depth Hj of the fall
and on the discharge q per unit width:
a deeper stilling basin increases the Froude number, thus Fr > >.
a more widen stilling basin increases the Froude number, thus Fr > > .
Box 12.1. Example calculation of the Froude number.
A flow o f " 2.80 ra^/s drops at a structure. The structure has vertical w^lls and has
a width of ^ 2.5 m . The upstreatn water level (*eiiergy depth) k located at
0.00 m ^ , the elevation of the floor i n the stilling basin has been designed at 4.50 m"^*
The inflowing jet into the basin amounts to ^ ~ 0 / ^ ~ 1,12 m'^/s. The energy
head above the floor is
= 4,50 m. The water depth yj aixd the velocity Vj of the
incoming jet can be calculated by iteration: y>. ~ q f { v^fS^g^^fj-tj})] >:
* assume yj ^ 0, and calculate
^ 111 I { 42x9.8(450^0.00)] >
0.119 m,
' enter this value again in the above formula and calculate a new yi - 142 /
{ v/[2x9,8 (4.50^0.12)] > * 0 J 2 1 m,
* continue the iteration^ until the "input" yi equals tbe ''output" y j .
Thus, the water depth y^ of the inflowing jef at the f i l l i n g basin i$y} ^ 0.121 m, with
a corresponding velocity o f v^^qfy-.^
L12 / 0421
9.26 m/s, Tue corresponding
Froude number amounts to Fr ^Vzf '/{gy-^ = 9,26 / /(9.$x0,121) = 8,5.
Hence, the Froude number of the flow at the level of the stilling floor at
-4>50 m ^ amounts t o Fr g.s,
A similar calculation can be carried out f o r floor depths o f 3.00
and
-1.50 m"**;, when Froude numbers of F r ^1 and 3.4, respectively^ can be calculated.
When the width of the structure is reduced f r o m B 2.50 m t o 2.00 m or to
L50 m , the Froude number at floor depth of *4J m"^ will amount to Fr t . 6 and
6.5> respectively.
12.2.2. Hydraulic jump

Hvdraulics. The super-critical flow (Fr > 1) is transferred to the sub-critical flow (Fr < 1)
through a hydraulic jump, see figure 12.9. The hydraulic jump is formed when the tailwater
depth y3 is sufficient deep, thus y3 > y^^^.
The minimum tailwater depthyjj^jj^, or the 'conjugate depth', depends on the discharge
q per unit width in m^/s.m, and the Froude number Frj and the water depth yj of the jet
before the jump:
^min =
yj ( / l + 8 F r /
- 1 ) , with: Fr. =
( g y f
or: Fr. =
isy~
The acmal length Lj of an 'undisturbed hydraulic jump' is somewhat ill-defined, but can
be estimated by the formula:
= 6.9 (y^ - y.)
6.1 y.
240
Analytical formula. The minimum stilling basin depth y^:^^ to form a hydraulic jump can
be calculated analytically, see also box 12.1 and figure 12.10. Thus, the simple formula to
check i f any hydraulic lump will occur in a stilling basin with water depth ymin, is:
y ^ i = 1.04 ^0-556 ^0.21
where q is the discharge per unit width in m^/s.m, z is the energy drop between up- and
downstream in m, and y^j^^ is the minimum water depth in stilling basin required for a
hydraulic jump.
The value of the formula is the direct calculation of the conjugate water depth yjjjjjj
without a cumbersome iteration process. It is obvious that this formula cannot be used for the
design of stilling basins. Such a stilling basin would create a long hydraulic jump at a water
slope of approx. lyert ^Hor'
leads to an expensive basin. Its length can be reduced
considerably by an over-depth and/or baffle-blocks and an effective end-sill.
10
value of X = z / yc
100
Figure 12.10. Minimum water depth y j for a hydraulic jump
241
12. D R O P S T R U C T U R E S A N D S T I L L I N G B A S I N S
Box 12.2. Analytical derivation of the conjugate depth y^,^ for a hydraulic jump.
The minimum basin depth y,i can be written as a function ot the drop 2 between the
upstream energy level and the d<wnstream water level, and as a funclitm o l the
discharge q per unit vsidth, in nr^'s.m.
The ccmversion is done by using an auxiliary broad-crested weir. The approaching
cnerg> level is a disumce W., above this crest, the stilling basin Is a distance
below
this crest. Thus, there arc four basic cquaticms:
the upstream energv level above the auxiliaiy weir:
llilSlilillllilB^
the upstream energ\' level abo\e the floor of the stilling basin, at the end o l the
IBÎ(IIIBI!HlllilBilii^HMll^
the upstream energy level above the floor o f the stilling basin, at the beginning of
the jump, where y- is the water depth of the incoming jet:
the hydraulic jump depends on the \\ater depth
in stilling basin, the Froude
number F/'j., and the water depth Vj of the incoming jet:
Vi - v . v,i (1 f 8 Fr^-)''-' - '/.'vnnn. with / v ^ - i / (iT V , ^ )
The depth D^^ = y^.^^^^ + z - //., can be calculated trom the above equations, by using
three t)ther parameters p, / and x:
p-^yjll.^
//a
x^zlH.^
F.limination o f the unknown parameters leads to two equations in />, / and x:
^ Iy^?^r^{}i>;..r.^)-i'^l^îMyrm^^
= (2/3)-'/r^ -
0.30/r^
so that: p = V . / vCI + 8 x 0 . 3 0 / 0 - 72 r
'+ ^-â ->{
(^'S (2/3 E . f - ! y-r 1
or: DJll, ~ p - 1 + 0.15 / p^
while: DJU.^ = A + ?- - I , so that: x = p - r + 0.l5//r
Based on these two equations, the relation between r nudx can be fotmd by assuming
a value of / and calculating the corresponding value x.
The data rand.v have been plotted in^gurc 12.10, and the best-fit curve through
these points has the equation r ~ 1.38 ,v'^*-^. Thus, the minimum water depth in the
stilling basin at which the jump will occur can be calculated by: >',nj,///j, = 1.38
^
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
As a - 1,7 / - / . t h i s equation can be rewritten into:
12.2.3. Standard Stilling Basins

USBR tvpe I basin. The hydraulic formulae as discussed above, lead to the design of a stilling
basin for the 'undismrbed-jump'. Such a basin has a basin depth ybasin ^^^^ equals the
conjugate depth yconj- USBR smdied the undismrbed jump basin as part of their series of
model tests on basins, and denoted this basin as the 'USBR type I basin' (Peterka 1974).
The undisturbed-jump or USBR type I basin would create the hydraulic jump, which
is quite long at a water slope of approx. lygrt ^Hor'
leads to an expensive basin, see
figure 12.11. Its length can be reduced considerably by an over-depth and/or baffle-blocks
and an effective end-sill.
242
s
s
\
Figure 12.11. The principle of the 'undisturbed-jump' basin (USBR type I basin).
Improvement of efficiency. The energy dissipation in a stilling basin is improved by one or

more of the following measures, see figure 12.12:
an extra over-depth of the floor,
an end-sill,
vertical drop, by impact of the jet on the floor,
chute blocks, to spread the inflowing jet,
baffle blocks, to dissipate energy of the jet by impact.
Thus, different standard basins has been developed, such as the USBR type I I I basin, USBR
type I V basin, the SAF basin, the Bucket basin, the Vlugter basin, and the vertical drop
basin.
Moreover, the width of the whole stmcmre can be increased, as to increase the Froude
number Fr of the inflowing jet by a lower unit discharge q.
CONJUGATE DEPTH
(undisturbed-jump basin)
(= USER type I basin)
OVER-DEPTH
(bucket basin)
(= USBR type V I I basin)
OVER-DEPTH
with END-SILL
(Vlugter basin)
IMPACT
with END-SILL
(vertical drop basin)
CHUTE BLOCKS
(USBR type I I basin)
(USBR type IV basin)
Figure 12.12. Types of standard stilling basins.
CHUTE BLOCKS
with BLAFFLE BLOCKS
(USBR type H I and SAP basins)
243
12.3. Types of Stilling Basins

12.3.1. Stilling Basins with baffle blocks
type I I I . The U S B R standard basin type I I I , see figure 12.13, is suitable for Froude
numbers Fn> A.5. The water depth in the basin equals the conjugate depth yiîn, needed for
an unaided hydraulic jump. Howeyer, the length of the U S B R type I I I basin is considerably
less by the use of the blocks.
The chutes blocks, baffle blocks and the end-sill along the floor of the basin produce
a stabilizing effect on the jump, which permits shortening of the basin. The chute blocks are
triangulate blocks that split and aerate ("beluchten") the jet.
The original design formulae for the U S B R type I I I basin are related to the hydraulic
testing resuhs (e.g. U S B R 1973, Chow 1959, Henderson 1966) and are less practical for the
design engineer.
The design of the U S B R type I I I basin can be shnplified (Ankum 1991) by expressing
the formulae into a design parameter
= (q / l . l f ' ^ . Thus, the depth y^ of the floor below
the tailwater leyel and the length
of the floor follows from:
USBR
\0.21
1.38
Ha
\0.21
3.7
with a drop z between the upstream and downstream energy level, and a discharge q per unit
width in m^/s.m. The U S B R type I I I basin may only be applied for the ratio zlH^ > 1.2.
chute
baffle
Figure 12.13. The principle of the U S B R type I I I basin.
type I V . The USBR standard basin type I V , see figure 12.14, has been developed
for Froude numbers 2.5 < Frj < 4.5, when a tme hydraulic jump does not fully develop and
waves may persist beyond the end of the basin. The sole function of the U S B R type I V basin
is 'wave suppression'
("golfdemping").
The water depth y^, of the basin is +10% greater than the minhnum depth yîn. as
required for the unaided hydraulic jump. The length
of the stilling basin equals the length
of the unaided hydraulic jump. The basin is equipped with large chute-type blocks that help
to intensify the roller of the hydraulic jump.
The original design formulae for the U S B R type I V are related to the hydraulic testing
results (e.g. U S B R 1973, Chow 1959, Henderson 1966) and are less practical for the design
engineer.
USBR
244
The design for the USBR type I V can be simplified (Ankum 1991) by expressing the
formulae into a design parameter
= (q / l . l f ^ . Thus, the depth y^ of the floor below
the tailwater level and the length
yb
1.52
Ha
9.3
\0.21
V Ha
( Z
Ha
K Ha
width in m^/s.m. The USBR type I V basin may only be applied for the ratio 0.30 < zlH^
< 1.2.
chute
block
LJ
Figure 12.14. The principle of the USBR type I V basin.
SAF basin. The SAF basin has been developed at the St.Anthony Falls (SAF) Hydraulic
Laboratory, USA, in the 1940s. The basin is recommended where Froude numbers are in the
range 1.7 < Fr < 17. The principle of the SAF basin is presented in figure 12.15.
The SAF basin is a shallow basin of an extreme shortness. For instance at Fr = 9, the
length of SAF basin is
= 0.87 y^:^^, against
= 2.7 y^^^ for the USBR type I I I basin,
and against L|, = 6.1 yjîn for the unaided jump. The depth of this SAF basin is y|, = 0.85
>'min' against y^ = y^j^ for the USBR type I I I basin and for the unaided jump.
The original design formulae for the SAF basin are related to the hydraulic testing
results (e.g. USBR 1973, Chow 1959, Henderson 1966) and are less practical for the design
engineer.
The design of the SAF type can be shnplified (Ankum 1991) by expressing the formulae
into a design parameter
= (q /1.7)^'^. Thus, the depth y^, of the floor below the tailwater
level and the length
for 0.15 < zlH^ < 2.0:
yb
0.14
1.33
Ha
\-0.12
and
2.2
for 2.0 < zlH^ < 6.5:

/
yb
1.17
0.21
-0.25
and
2.4
Ha
FT.
for 6.5 < zlH^ < 13.0:

yb
-0.20
2.60
-0.25
and
2.4
width in m-^/s.m. The SAF type basin may only be applied for the ratio 0.15 < zlH^ < 13.
245
12. D R O P S T R U C T U R E S A N D S U L L I N G BASINS
chute baffle end

block block sill
^7
<>
Figure 12.15. The principle of the SAF basin.
12.3.2. Bucket-type stilling basins

Bucket basin. The 'bucket basin' is a deep basin, for which the water depth in the basin is
larger than the minimum depth for a hydraulic jump, see figure 12.16.
The energy dissipation depends not on the formation of a hydraulic jump but on the
formation of two rollers: a bucket roller on the surface ("oppervlakte neer") and a ground
roller ("bodemneer"). The intermingling movements of these two rollers are effectively
dissipating the energy.
A minimum water depth in the basin is required to avoid the sweeping out of the
surface roller. A maximum water level is required to form a proper ground roller, to avoid
the diving of the flow.
The bucket basin has been modeltested by USBR, and data have been presented in the
form of curves in relation of the Froude number Fr^., whereas the Froude number Fr^ refers
to the flow at the point where incoming jet y^ enters the bucket, at the level of the tailwater
surface (e.g. USBR 1973, Chow 1959, Henderson 1966).
The Froude number Fr^ of the incoming jet at the level of the tailwater surface, is not
a practical parameter in the design, and can be avoided by expressing the formulae into a
design parameter
= (q /
The design of the bucket basin involves determination
of the radius R of the curvation of the bucket, and the floor depth
below the tailwater level
which follow directly from the formulae (Ankum 1991):
R
Ha
0.93
+ 0.113
"
^2
0.019
Ha
yb
2.0 + 0.1
Ha
H^
width in m^/s.m. The bucket basin may only be applied for the ratio 1.5 < zlH^ < 7.
Vlugter basin. The Vlugter basin has been developed in Indonesia by Vlugter in the early
1940s (Vlugter 1941). The Vlugter basin looks like the bucket basin, but has a horizontal
floor. The stilling basin is relative deep and short, and has no blocks but an end-sill, see
figure 12.16. The glacis ("achtervlak") has a slope of 45, and is transferred into a curve
with a radius R before reaching the stilling basin.
The modeltests as performed by Vlugter provided a series of design formulae, which
are based on the design parameter
= (q / l.l)^''^. The depth yj, of the floor below the
246
tailwater level, the length

of the floor, the radius R between the glacis and the floor, and
the minimum height a of the end-sill above the floor follows from:
II
2.0
+ 0.1
1.0
1.1
-0.5
0.15
H
width in m^/s.m. The Vlugter basin may only be applied for the ratio 0.50 < zlH^ < 5. Box
12.3 presents an example calculation of a drop stmcmre with a Vlugter basin.
B U C K E T BASIN
VLUGTER B A S I N
Figure 12.16. Bucket-type basins (for z/H^ = 3).
Box 12.3. Example calculation of a Vlugter basin.

A drop stmcture with a water level drop of - 1.50 m have to designed in a drain
with a design capacity of - 2.2 nrVs. The drain characteristics arc: bed width
= 1.75 m, side slopes ly^^j.^; LSjjj,^ and a design water depth o f y = 1.16 m. The drop
structure will be equipped with a vertical gate of a width dg^^^ - 1.50 m.
I t is decided that the stilling basin will be a Vlugter basin of a width of
=
1,75 m. The design parameter H.^ depends on the discharge per unit width, which is
selected here at the gate width 6'^t,. - 1.50 m. Hence> H.^ - {Qli.lb)^'^
~
(2.2 / 1.7 L50)^/^ = 0.91 ra. Thus, z/H.^ = 1.50/0.91 = 1.66.
The design dimensions of the basin are:
- radius K
1.1 c + i / , , = 1.1 x L50 + 0.91
2.56 m 2.60 m,
length of the basin
= 1.1 2 + //.^ = 2.60 m.
depth of basin Vj, - 0.1 z + 2.0
O.l x 1.50 4- 2x0.91 - 1.97 m ^ 2.00 m.
height of end-sill a = 0.15 f / , . {HJz) = 0.15x0.91 Xs (0.91/150) = 0.11
0.15 m.
247
12.3.3. Miscellaneous stilling basins

Volume basin. Volume basins are recommended for downstream control gates (AVIO and
AVIS) to dissipate the excessive energy. Volume basins are rectangular basins of sufficient
volume, see figure 12.17.
The volume V of the volume basins depend on the energy to be dissipated, and are
proportional with the product of the design discharge Q and the (energy) headloss z. Thus
(Alsthom undated): L X b X
= 21.2 Q X z^'^. Furthermore, the optimum proportions
for the length L, the width b and the water depth y^ of the volume basins are: L = 3 x
4.5 X y^.
The original design formulae for the volume basin can be simplified (Ankum 1991) by
expressing the formulae into a design parameter
^ (q / 1.7)^'^. So, the depth y^ and the
length
of the volume basin follows from:
yb
\0.25
2.931
Ha
\0.25
13.2
H
width in m^/s.m. The range of zlH^ for which the volume basin may be used has not been
specified.
max. decrement
drop z
/POOL
/BASIN
A3.ASIN
OL
PLAN
Figure 12.17. The hydraulic dimensions of the Volume basin.
248
Vertical-drop basin. Vertical-drop basins are characterized by a free-falling jet into the
basin. The free falling jet makes an impact ("botsing") with the basin floor and is irned into
the downstream direction, see figure 12.18. The basin is equipped with an end-sill
("einddrempel").
Upto 50% o f the energy may be dissipated by the hnpact of the jet and by the mrbulent
circulation in the pool beneath the jet. The remainder part is dissipated by the hydraulic jump
in the basin.
Drop Length LD
"unaided" Hydraulic Length L j
Stilling Basin Length Lg
Figure 12.18. Straight drop structure with a "vertical-drop" basin.
Although the vertical-drop basin seems to be an efficient energy dissipator because of

the hnpact, the basin becomes very long in comparison with other stilling basins. The stilling
basin should not only provide for the hydraulic jump, but should also acconmiodate the
horizontal distance
covered by the dropping jet, which is quite long:
\0.857
5.14
For example, a discharge per unit width of ^ = 0.5 - 1.0 m-^/s.m dropping over z = 1.00
m requires a horizontal distance
of 5 . 0 0 m for the jet to reach the floor!
Modeltests as reported in the literamre (e.g. Chow 1959, Henderson 1966), are less
practical for the design engineer. These tests provided a series of design formulae, which are
based on the 'drop number' D = q^/(g z^^) and on the drop Z\y in chaimel bed.
The original design formulae for vertical-drop basin can be shnplified (Ankum 1991)
by expressing the formulae into a design parameter H^ = {q I IJ)^^^. Thus, the depth y^ of
the floor below the tailwater level and the length
0.179
1.25
11.6
width in m^/s.m. The range of zlH^ for which the volume basin may be used has not been
specified. However, the Froude number Frj of the jet after the hnpact with the floor can be
derived as well:
249
\ 0.389
Fr:
3.27
H
Considering a range for the Froude number of Frj > 4.5, the vertical-drop basin may be
applied for zlH^ > 2.3. Box 12.4 presents an example calculation of a drop strucmre with
a vertical-drop basin.
Box 12.4. Example calculation of a vertical-drop basin.

A drop structuje with a water level drop of c = 1.50 m have lo designed in a canal
with a design capacity of Q = 0.8 nv\s.
It is decided to apply a vertical-drop basin. The width b of the drop structure is
h = 1.00 m, so that tiie design parameter M j , which is the energy depth above an auxiliary broad-crested weir, follows from f/^ ^ {Qn.lhY'^
= (0.8/1.7 \mf'^
= 0.61 m.
Thus, z//-/., = 1.50/0.61 = 2.46.
The design dimensions of the basin arc:
. basin depth yt, - 1.25 / / , [ziH f-^"''^ - 1.25 x 0.61 [lA^f^^^
- 0.90 m:
b a s i n l e n g t h V ^ / , = 11.6(z/7f/-'^'"^- 1 1 . 6 ( 2 . 4 6 ) - 19.91. Thus, t j ,
19.91
X 0.61 = 12.14 =^ 12.5 m.
12.3.4. Selection of stilling basins

Most shnple basin. Anyhow, it is wise to construct always a stilling basin behind a stmcmre.
Supercritical flow may even occur at unexpected circumstances. The cost of a simple basin
will be minimal, and it will safeguard the stracmre during unexpected operation. Such a
simple basin can be created by:
a floor that is slightiy lower (e.g. 0.30 m) than the ongoing channel bed;
the horizontal floor of the basin will be ended by 'end sill';
a coffer under the end sill.
Will a lump occur? The check whether a basin will force a hydraulic jump follows from the
formula:
where q is the discharge per unit width in m^/s.m, z is the energy drop between up- and
downstream in m, and yj^j^ is the minhnum water depth in stilling basin required for a
hydraulic jump.
When is a basin certainlv needed? The need for a stilling basin follows from, see also box
12.5:
a basin is required to force a j u m p when the tailwater depth is less than the 'conjugate
depth', thus when y^, < ymin- Utilize the ratio zlH^ for the selection and the design of the
basin;
a basin is required when waves may occur, even i f a basin is not required to force a jump
CVs > >'min)- Thus, the value of the Froude number Frj have to be checked on 'waves'.
250
Parameter for selection of the basin. There are several types of stilling basins, and each of
these basins can be applied in combination with the different control sections, such as weirs,
notches and gates. I n fact, it is the incoming jet at the floor which determines what type of
basin is required.
This incoming jet has a certain flow, q in m^/s.m, and excess energie, z in m. Thus,
the hydraulics of the incoming jet for the selection of a stilling basin, can be described by
either:
the Froude number Fr^ of the incoming jet, where Frj = q I (g y-^f'^, where q is the
discharge per unit width in m^/s.m, y-. is the water depth of the jet and g = 9.S m/s^ =
9.8 N/kg is the gravity acceleration. The Froude number depends on the energy depth Hj
of the jet;
the ratio z/H^, where H^ = {q I 1.1)^'^, z is the drop between the upstream and
downstream energy level.
It appears that the ratio zlH^ is the most practical parameter in the selection of stilling basins.
However, the Froude number F / j is still needed when a check has to be made on waves.
Box 12.5. Example of an evaluation on the need of a stilling basin.
First try. A gated structure ib^

1.50 m wide aiid a discharge Q ~ 1.5 m^/s might
require a stilling basih. The upstream energy head
- 1,00 m, the tailwater depth
y-^ aa 0,90, thus the energy drop z * J0
The discharge q per unit width is ^ - LO m^/s.m, the value oiH^ ^ (ql L7)^^'^
= 0.70 tn. Tiie water depth yj of tbe (supercritical) Jet is calculated by iteration
through the equation
=^ q /
{Hy y-)f'^ ^ LO / |2g {100 * y | ) f y * ^ 0.00 m,
gives >N ss 0.23 my gives y^ ^ 0.26
gives finally/j - 0.26 m.
The minimum tailwater depth for a jump k y^^^ ^ 1,04 q^-^^^ 2^-^^ = 1.04 X
1.0^*^^ X 0.10^'^^ ~ 0.64 m. Thus, a stilling basin is not required for a hydraulic Jump.
The Froude number Fr^ of the (supercritical) j e t F ^ | ^ / igy^^f'^ - 2.4, which
is i n the ^'danger" range.
; S^c<^nd try. The opening of the stracture is made smaller to ^ = 1.00 m to decrease
the Froude number. The upstream energy head remains H> ~ IM m.
!
The discharge q per unit width is now q i s m^/s.m, the value of
{q /
1.7)^/^ ss 92 m. The water depth yi of the (supercritical) jet is calculated by iteration
through the equation ^qf{2g(Hi^
y j ) f - 1.5 / 1 % (LOO -yj)]^'^: gives finally
=^ 0.45 m,
The minimum tailwater depth for a Jump i^y^î^ ~ 104 q^-^^^
- 104 x
15^'^^^ X O.IO^"^^ ^ 0,79 m. Thus, a stilling basin Is still not required for a hydraulic
Ip^JliHIiliiliiBiM^l^
The Froude number Frj o f the (supercritical) jet Fr| ^ qi l ^ y f f ' ^ - i 6 , which
means that a stilling basin is not needed.
Froude number. The Froude number yj of the jet cannot be determined directly. The water
depth yj follows from yj -I- (^'Hy^ 2g) = Hj, or:
^ 2g (HJ - y.)
The value of yj can only be calculated by iteration for a known energy depth Hj of the jet and
a known discharge q per unit width in m^fs.m: assume y; = 0, calculate a new value of y:
251
and insert this value again in the formula, etc., etc.

Based on the Froude Number Fr^, the following stilling basins are generally recommended, see figure 12.19:
Fn = 1.
When the Froude number of the incoming jet is equal to ^ ^ j = 1, the flow is at critical
depth and a hydraulic jump cannot form. A stilling basin is not needed.
1.0 < Fr. < 1.7
The incoming jet is only slightly below critical depth, and the change f r o m this low stage
to the high stage flow is gradual and manifests itself only by a slight raffled water surface.
No stilling basin is needed, only some protection with rip-rap, or by masonry or
concrete lining.
25<Fr< 4.5
"TRANSITION" STAGE
oscillating jumps: avoid this range!
4 J < Rr < 13
"EFFECTIVE JUMP" STAGE
most effective
R > 13
"EFFECTIVE JUMP" STAGE
but rough downstream water surface
Figure 12.19. Hydraulic jump related to the Froude number.
252
1.7 < F r j < 2.5

A series of small rollers develop on the surface for the Froude number of Frj > 1.7,
which become more intense with increasingly higher values of the number.
A stilling basin is needed to dissipate energy effectively. Generally, a basin with an
end-sill is needed.
2.5 < F r j < 4.5
An oscillating form of jump occurs, the entering jet intermittently flowing near the bottom
and then along the surface of the downstream charmel.
This range is difficult for selection stilling basins as the hydraulic jump is not
well-stabilized and produces waves that travel long distance through the channel. It is
better to avoid this range by increase the Froude number through a wider strucmre. I f not
possible, the USBR type I V may be used. However, the wave action cannot be entirely
dampened.
4.5 < F r j < 13
This range requires the most economical basins, because they can be kept short. The
USBR type I I I is recommended, as well as basins with end-sills.
F>j > 13
When Frj > 13, the jump is very rough, and will make conditions rather difficult to calm
the tailwater. The stilling basin will become relatively expensive.
The Froude number Fn of the jet can be changed by another design of the strucmre:
o a wider strucmre will decrease discharge q per unit width in w?/s.m, and will lead to a
higher Froude number F r j ;
o a lower floor will increase the energy head
of the jet, and thus w i l l lead to a higher
Froude number Frj.
Ratio 7./Ha. The selection of the basin type can be based on the ratio z/H^, i.e. the ratio
between the drop z in up- and downstream level, and the parameter H^ = (q / 1.7)^^^ where
q is the discharge per unit width in m?/s.m.
The design formulae for the depth
and the length
of the different basins are given
in table 12.1, as well as the validity z/H^ for each of the standard basins.
The comparison of the depth y^ and the length
of the different basins are presented
in figure 12.20 and in figure 12.21.
Table 12.1. Selection of stilling basins.
Depth yj3 of Basin
Length Lj, of Basin
valid for drop z

" UN-AIDED" JUMP
STANDARD BASINS:
Vertical-drop basin
USBR-basin type I I I
USBR-basin type I V
SAF-basin
SAF-basin
SAF-basin
Bucket basin
Vlugter basin
Volume basin
Vâ*
1.38 (z/H^f-^^
z/H^ > 1.2
z/H^>2.3
z/H^ > 1.2
0.30<z/H^<l.2
0.l5<z/H^<2.0
2.0
<zlH^<6.5
6.5
<zlH^<l3
1.5 <zlH^< 7
0.5 <zlH^< 5
not known
1.25 (z/H^f-^^
1.38 iz/H^f-^^
1.52
{ZIHJ^-'^^
1.33
{ZIHJ^-'^'^
1.17 {z/H^f-^^
2.60 (z/HJ^-^^
0.10 (z/H^) + 2.0
0.10 (z/H^) + 2.0
2.93
{ZIHJ^-^^
9.3 (z/Hf-^^
11.6
3.7 (z/H^f-^^
9.3 (z/H^f-^^
2.2 (z/H^r^-'^^
2.4 (z/H^y^-^^
2.4 (z/H^y^-^^
not applicable
1.1 zlH^ + 1.0
13.2 (zlH^-^^
H^ is a design parameter, which follows from H^ = { q I 1.7 ) ^'^, and q = Q I B.
Figure 12.21. Lengths
of stilling basins.
254
Other criteria. Other criteria for the ulthnate selection of stilling basins may concern the
constructional matters.
The USBR and the SAF basins are attractive from the hydraulic point of view. The SAF
basin has a large range of validity, and is very short and shallow. The USBR type I V
performs well for small Froude numbers. The USBR type I I I basin is shorted than the
unaided jump. However, these basins require a precise constmction work for the chute-blocks
and baffle-blocks.
The bucket basin and the Vlugter basin are simple to constmct as they do not have
chute blocks and impact blocks. The Vlugter basin is quite deep, which may be difficuh for
construction in areas with high groundwater tables and pervious soils. Moreover, the Vlugter
basin has shown to be unreliable with tailwater levels above and below the laboratory tested
levels, and is therefore not recommended with fluctuating discharges as with weirs in rivers.
The volume basin and the vertical-drop basin require quite large dimensions. The
vertical-drop basin seems easy to construct, but has to be well checked on uplift of the floor
by groundwater pressure.
13. REFERENCES
Ankum, P., K. Koga, W . A . Segeren & J. Luijendijk. 1988. Lessons from 1200 years impoldering in
the Netherlands. In: Proc. int. symposium on shallow sea and low land, Saga: 101-108. Saga
University, Japan.
Ankum, P. 1991. Flow Control in Irrigation Systems. Delft University of Technology, Faculty of Civil
Engineering, Delft, the Netherlands.
Ankum, P. and R. Brouwer. 1993*'. Nevengeulen en sedimentverdeling (in Dutch). Communication
49 of water resources management and sanitation dept., University of Technology. ISSN 0169
6246. Delft: 77 pp.
Ankum, P. 1994^. Methods of Development. In: N . Miura et al (ed.), Lowlands, Development and
Management. Balkema, Rotterdam, the Netherlands: 65-91.
Ankum, P. 1994*'. Integrated Management. In: N , Miura et al (ed.). Lowlands, Development and
Management. ISBN 90-5410-603-4. Balkema, Rotterdam, the Netherlands: 359-379.
Ankum, P. 1995. Flow Control in Irrigation and Drainage. Communications of the Water
Ankum, P. 1996. Polders, Drainage and Flood Control. Water management dept.. Faculty of civil
engineering. University of Technology. Delft, the Netherlands.
Ankum, P. 1996. Reference Evapo-transpiration.
Water management dept., Faculty of civil
engineering. University of Technology. Delft, the Netherlands.
Ankum, P. 2000. Design of new flood channels in Lowlands. Proceedings of the International
Symposium on Lowland Technology, Saga University, Japan: 355-362.
Ankum, P. 2002. Stilling Basins and Drop Structures in Open-Channels, (unpublished yeO Water
management dept., Faculty of civil engineering. University of Technology. Delft, the Netherlands.
Arends, G.J. 1993. Sluizen en stuwen. Delftse Universitaire Pers, Delft, the Netherlands.
ASCE. 1995^. Hydraulic design offload control channels. Technical engineering and design guides
as adapted from the US Army Corps of Engineers, No. 10. American Society of Civil Engineers.
New York.
ASCE. 1995'^. Guidelines for design of intakes for hydroelectrical plants. Committee on Hydropower
Intakes of the Energy Division of the American Society of Civil Engineers. New York.
ASCE. 1995'^. Hydraulic design of spillways. Technical engineering and design guides as adapted from
the US Army Corps of Engineers, No. 12. American Society of Civil Engineers. New York.
Baban, R. 1995. Design of diversion weirs. Wiley series in water resources engineering. Chichester,
U.K.
Bendegom, L . van. 1965. Algemene Waterbouwkunde, deel I, Ua, lib. Lecture Notes, Technical
University, Delft, the Netherlands.
Bendegom, L . van. 1971. Natuurlijke en kunstmatige waterlopen: deel III, begin van beweging en
transport van sediment in waterlopen. Lecture Notes, Technical University, Delft, the Netherlands.
Bendegora, L . van. 1978. Rivieren en Rivierwerken (in Dutch). Lecture notes. Delft University of
Technology, Faculty of Civil Engineering, Delft, the Netherlands.
Boer, H.E. and J.A. Kielman. 1956. Rivieren, verdedigingswerken,
dijken, duinen en stranden,
waterschappen, polders en droogmakerijen (in Dutch). Weg- en Waterbouwkunde, deel I I I .
Kosmos, Amsterdam.
255
256
Bos, M.G, 1984. Long-throated flumes and broad-crested weirs. PhD-thesis. University of
Agriculture, Wageningen, the Netherlands.
Bos, M.G. 1989. Discharge measuring structures. International Institute for Land Reclamation and
Improvement, Wageningen, the Netherlands.
Brouwer, R. 1987. Design and application of automatic check gate for tertiary turnouts. In:
Proceedings of the 13th international congress on Irrigation and Drainage. ICID, New Delhi.
Chanson, H . 1999. The hydraulics of open-channel flow. Butterworth-Heinemann, Oxford.
Chaudhry, M . H . 1993. Open-channel flow. Prentice-Hall, Englewood Cliffs, USA
Chien, N . and Z . Wan. 1999. Mechanics of sediment transport. ASCE. Reston, USA.
Chow, Vente. 1959. Open-channels Hydraulics. McGraw-Hill, New York.
Chow, Ven te. 1964. Handbook of applied Hydrology. McGraw-Hill, New York.
Clemmens, A.J., M . G . Bos and J.A. Replogle. 1993. Flume, design and calibration of long-throated
measuring flumes. Version 3.0. International Institute for Land Reclamation and Improvement,
publication 54, Wageningen, the Netherlands.
Davis, C.V. and K . E . Sorensen. 1969. Handbook of applied hydraulics. McGraw-Hill, New York.
DHV. 1985. Training Packages, training programme on integrated irrigation development. SEDEKU
Trail-Run project. Ministry of Public Works, Indonesia.
Eijsvoogel, W.F. 1932. Het ontwerpen van een bandjirscherm (in Dutch). De Waterstaats Ingenieur,
No.10.
Framji, K . K . 1972. Design practices of irrigation canals in the world. ICID, New Delhi.
French, H.F. 1994. Open-channel hydraulics. McGraw-Hill. New York.
Garde, R.J. & K . G . Ranga Raju. 1977. Mechanics of sediment transportation and alluvial stream
problems. Wiley Eastern Limited, New Delhi.
Garg, S . K . 1978. Irrigation Engineering and Hydraulic Structures. Khanna Publishers, Delhi.
Goussard, J. 1993. Automation of Canal Irrigation Systems. International Commission on Irrigation
and Drainage, New Delhi.
Graf, W . H . 1971. Hydraulic of Sediment Transport. McGraw-Hill Book Company, New York.
Hager, W . H . 1991. Uniform aerated chute flow. Journal of Hydraulic Engineering. Vol, 117, no. 4,
April. American Society of Civil Engineers.
Hager, W . H . 1992. Energy Dissipators and Hydraulic Jump. Kluwer Academic Publishers, Dordrecht,
the Netherlands.
Henderson, F . M . 1966. Open Channel Flow. Macmillan Publishing Co., Inc., New York.
Herbermann, H . en W. Schuurmans. 1991. Profile, Users Manual (version 3.00). Delft University
of Technology, Faculty of Civil Engineering, Delft, the Netherlands.
Hilgen, J.H. 1937. Berekeningen van hellende goten, uitgaande van de formule van Gauckler-Manning
(in Dutch). In: De Ingenieur in Nederlandsch-Indi.
Hjulstrm, F. 1935. Studies of the morphological activity of rivers as illustrated by the river Fyris.
Bulletin, Geological Institute of Upsala, vol. X X V . Upsala, Sweden.
Hoi, W.H.J. 1963. Inleiding tot de Waterbouwkunde: twintig eeuwen strijd om de beheersing van land
en water in de Lage Landen, (in Dutch). Weg- en Waterbouwkunde, deel A. Kosmos, Amsterdam.
Horn, M . L . ten, e.a. (ed.). 1994. Nederlandse Ingenieurs en hun kunstwerken. Walburg press,
Zutphen, the Netherlands.
Huisman, P. 1966. Extreme Waarden en Decisieproblemen. Lecture Notes, Technical University,
Delft, the Netherlands.
ICID. 1967. Multilingual technical dictionary on irrigation and drainage. International commission
on irrigation and drainage. New Delhi.
ICID. 1993. Water management in the next century. Proc 15th congr. on irrigation and drainage.
Volume l A (question 44, R1-R35); volume IB (question 44, R36-R72); volume IC (question 44,
R73-R103); volume I D (question 44, R104-R118); volume I E (question 45, R1-R29); volume I F
(Question I F , R30-R40); volume I G (special session); volume I H (general reports); volume I I
(history special session); volume IJ (keynote addresses); workshop on micro irrigation worldwide;
workshop on subsurface drainage simulation models; 2nd workshop on crop-water models; the
impacts of irrigation and drainage on the environment. The Hague & New Delhi.
ILRI. 1964. The Design of Open Watercourses. International Institute for Land Reclamation and
Improvement, bulletin 7, Wageningen, the Netherlands.
13. R E F E R E N C E S
257
ILRI. 1982. Polders of the World: Papers, vol.1, vol.2 and vol.3. International Symposium,
International Institute for Land Reclamation and Improvement, Wageningen, the Netherlands.
ILRI. 1983. Polders of the World: Final Report. International Symposium, International Institute for
Land Reclamation and Improvement, Wageningen, the Netherlands.
James, L . G . 1988. Principles of farm irrigation system design. John Wiley, New York.
Jansen, P.Ph., L . van Bendegom, J. van den Berg, M . de Vries, and A. Zanen. 1979. Principles of
river engineering, the non-tidal alluvial river. Pitman, London.
Jensen, M . E . 1983. Design and operation of farm irrigation systems. ASAE St. Joseph.
Jurrins, M . 1976. Sedimenttransport (in Dutch). Delft University of Technology, Faculty of Civil
Engineering, Delft, the Netherlands.
Jurrins, M . 1993. Maintenance of Open Drainage systems. Lecture notes, no. 20, International Course
on Land Drainage, Wageningen, the Netherlands.
Jurrins, M . and K.P. Jain (ed.). 1993. Maintenance of irrigation and drainage systems. International
Institute for Land Reclamation and Improvement, Wageningen, the Netherlands.
Kley, J. van der & H.J. Zuidweg. 1969. Polders en dijken (in Dutch). Agon Elsevier, Amsterdam.
Kraatz, D.B. and I . K . Mahajan. 1975. Small hydraulic structures. Irrigation and Drainage Paper 26,
FAG, Rome.
Leliavsky, S. 1965^. Irrigation Engineering: Canals and Barrages. Design textbooks in civil
engineering, volume one. Chapman and Hall Ltd., London.
Leliavsky, S. 1965''. Irrigation Engineering: Syphons, Weirs and Locks. Design textbooks in civil
engineering, volume two. Chapman and Hall Ltd., London.
Leliavsky, S. 1965'^. Design of dams for percolation and erosion. Design textbooks in civil
engineering, volume three. Chapman and Hall Ltd., London.
Leliavsky, S. 1965*^. River and canal hydraulics. Design textbooks in civil engineering, volume four.
Chapman and Hall Ltd., London.
Leliavsky, S. 1981. Weirs. Design textbooks in civil engineering, volume five. Oxford & IBH
Publishing Co, New Delhi.
Leopold, L . B . , M.G. Wolman and J.P. Miller. 1964. Fluvial processes in Geomorphology. Freeman,
San Francisco, USA.
Leusen, B. van. 1993. Duikers en sluizen. Stam Techniek, Culemborg, the Netherlands.
Leusen, B. van. 1993. Kleine waterbouwkundige constructies. Stam Techniek, Culemborg, the
Netherlands.
Luijendijk, J. & E. Schultz. 1982. Waterbeheersingssysteem van een polder (in Dutch). In: PT Civiele
Techniek Nr.9/82: 21-29. NIRIA, the Hague, the Netherlands
Luijendijk, J., E. Schultz & W . A . Segeren. 1988. Polders. In: Developments in hydraulic
engineering, volume 5 (P. Novak ed.). Elsevier, London/New York
Miura, N . (ed.). 1992. Problems of Lowland Development. Proceedings I T L seminar. Institute of
lowland technology. Saga university, Japan. Saga.
Miura, N . , M.R. Madhav and K. Koga (ed.). 1994. Lowlands, development and management.
Balkema, Rotterdam, the Netherlands.
Montes, S. 1998. Hydraulics of Open Channel Flow. American Society of Civil Engineers. Reston,
USA.
NEDECO-DHV. 1986. Design Criteria: vol KP-01 Irrigation System Design, vol KP-02 Headworks,
vol KP-03 Canals, vol KP-04 Structures, vol KP-05 Tertiary Units, vol KP-06
Structural
Parameters, vol KP-07 Drawing Standards. Ministry of Public Works, Indonesia.
Prins, A. and M . de Vries. 1971. On dominant discharge concepts for rivers. Delft University of
Technology, Faculty of Civil Engineering, Delft, the Netherlands.
Raudkivi, A.J. 1976. Loose Boundary Hydraulics. Pergamon Press, Oxford.
Raudkivi, A.J. 1979. Hydrology, an advanced introduction to hydrological processes and modelling.
Pergamon Press, Oxford.
Raudkivi, A.J. 1993. Sedimentation. Exclusion and removal of sediment from diverted water.
Hydraulic structures design manual. Balkema, Rotterdam, the Netherlands.
Razvan E. 1989. River Intakes and Diversion Dams. Elsevier, Amsterdam.
258
Ritzema, H.P. (ed.). 1994. Drainage principles and applications. Lecture Notes of the International
Course on Land Drainage, International Institute for Land Reclamation and Improvement,
publication no. 16, Wageningen, the Netherlands.
Schoemaker, H.J. 1979. Side Weirs along Prismatic Canals. Publication no.218, Delft Hydraulics
Laboratorium, Delft, the Netherlands.
Schoemaker, H.J. 1983^. Monografien irrigatie fl8 en fl9 (in Dutch). Lecture Notes. Delft
University of Technology, Faculty of Civil Engineering, Delft, the Netherlands.
Schoemaker, H.J. 1983*'. Dynamica van Morphologische Processen (in Dutch). Lecture Notes. Delft
Schoklitsch, A . 1952. Handbuch des Wasserbaues. Springer Verlag, Wien.
Schultz, E., 1987. Watermanagement in newly reclaimed land. RlJP-rapport 9 cbw. Rijksdienst voor
IJsselmeerpolders, Lelystad, the Netherlands.
Schultz, E. 1987. De Nederlandse Droogmakerijen (in Dutch). Rijp rapport. Ministerie van Verkeer
en Waterstaat, Lelystad, the Netherlands.
Schultz, E. 1992. Waterbeheersing van de Nederlandse droogmakerijen {inDnioh). Dissertation. Delft
Segeren, W . A . 1982. Introduction to polders of the world. In: Final report, int. symposium polders
of the world. International institute for land reclamation and improvement (ILRI), Wageningen, the
Netherlands.
Sellin, R.H. 1969. Flow in channels. Macmillan, London, U K .
Smedema, L . K . and D.W. Rycroft. 1983. Land Drainage. Batsford, London.
STOWA. 1994. Handboek debietmeten in open waterlopen. STOWA, Utrecht, the Netherlands.
USBR. 1965. Hydraulic Design of Transitions for Small Canals. United States Department of Interior,
Denver, USA.
USBR. 1973. Design of Small Dams. United States Department of Interior, Denver, USA.
USBR. 1974. Design of Small Canal Structures. United States Department of Interior, Denver, USA.
USBR. 1974. Hydraulic Design of Stilling Basins and Energy Dissipators. United States Department
of Interior, Denver, USA.
Vanoni, V . A . (ed.) 1977. Sediment Engineering. ASCE Manuals and Reports on Engineering Practice
No.54, American Society of Civil Engineers, New York.
Varshney, R.S. 1977. Engineering Hydrology. Nem Chand & Bros, Roorkee, India.
Varshney, R.S., S.C. Gupta & R.L. Gupta. 1978. Theory and Design of Irrigation Structures, volume
II: canal & storage works. Nem Chand & Bros, Roorkee India.
Ven, G.P. van de (ed.). 1993. Man-made lowlands, history of water management in the Netherlands.
ICID, the Hague, the Netherlands.
Vlugter, H . 1940. Over zelfwerkende peilregelaars bij den Waterstaat in Nederlandsch Indie (in
Dutch). De Ingenieur in Ned.-Indie, No.6.
Vlugter, H . 1941.1216, jaar hydrodynamisch onderzoek aan waterloopkundige modellen in Ned.-Indi
en hydraulisch research werk in het algemeen (in Dutch). In: De Ingenieur in Ned.-Indi, No.9.
Vlugter, H . 1941. Het transport van vaste stoffen door stroomend water (in Dutch). In: De Ingenieur
in Ned.-Indi, No.3.
Vlugter, H . 1962. Sediment transportation by running water and the design of stable channels in
alluvial soils. In: De Ingenieur, No.36.
Volker, A . 1967. Waterhuishouding. Lecture Notes, Technical University, Delft, the Netherlands.
Volker, A . (ed.). 1993. Hydrology and water management of deltaic areas. Ministry of transport,
public works and water management, CUR-report 93-5, the Netherlands.
Vos, H.C.P. de. 1925. Transport van vaste stoffen door stroomend water (in Dutch). In: De
Waterstaats Ingenieur, No.7.
Vreedenburgh, C.G.J. 1924. De formule van Gauclcler-Manning voor buizen en kanalen (in Dutch).
In: Waterstaatsingenieur, no.9.
Wagret, P. 1968. Polderlands. Methuen, London.
WERKGROEP VADEMECUM. 2000. Cultuur-technische Vademecum (in Dutch). Cultuurtechnische
Vereniging. Elsevier. Doetinchem, the Netherlands.
Wilson, E . M . 1969. Engineering Hydrology. Macmillan.

Ddis

Uploaded by

Copyright:

Available Formats

Ddis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ddis

Uploaded by

Copyright:

Available Formats

CT3410

4. SUPER-CRITICAL FLOW IN OPEN-CHANNELS

6. DESIGN OF CONVEYANCE STRUCTURES

7. DESIGN OF PASSIVE REGULATORS

8. DESIGN OF MANUAL REGULATORS

9. DESIGN OF HYDRO-MECHANICAL REGULATORS

9. DESIGN O F HYDRO-MECHANICAL REGULATORS (Continued)

10. DESIGN OF DIVERSION STRUCTURES

11. DESIGN EXAMPLES OF DIVERSION STRUCTURES

12. DESIGN OF DROP STRUCTURES AND STILLING BASINS

1. WATER CONTROL SYSTEMS

Figure 1.1. Lowland Development and Water Control Systems.

1.1.3. Terms in Lowland

PLAN OF TIDAL LANDS

Figure 1.2. Tidal lands with creeks.

Figure 1.3. Main drainage system and field drainage system.

Figure 1.4. A n irrigation scheme with a water control system.

1.2. Flood Control Systems

Figure 1.6. Flood control by channel improvement and shortening alignment.

DESIGN OF OPEN-CHANNELS A N D H Y D R A U L I C STRUCTURES

1.2.4. Flood Embankments

Figure 1.7. Flood control by construction of embankments.

1.2.5. Flood Diversion

Figure 1.8. Flood diversion structure for flood control.

1.2.6. Hydrology of Floods

Figure 1.9. Complete, partial () and annual () series.

return period, according to the annual analysis,

DESIGN OF OPEN-CHANNELS A N D H Y D R A U L I C STRUCTURES

plot the data on the selected probability paper;

for r > 1 year

' Return Period,

Log - Normal Probability Paper

Linear - Normal Probability Paper

Linear - Gumbel Probability Paper

Figure 1.10. Four types of probability paper.

Log - Gumbel Probability Paper

for r > 1 year

The use of plotting paper.

1.3. Drainage Systems in Polders

Different professional disciplines. Many professional disciplines are involved in design of

Figure 1.11. Hachiro Gata polder, Japan.

Figure 1.12. Water management system of a polder.

1.3.2. Main Drainage System

Operational objectives. The operation of gravity ("zwaartekracht") drainage systems is

Grid Drainage System

Branching Drainage System

Figure 1.14. Pattern of main drainage system in polders.

Separate Irr. & Dr. System

1.3.3. Polder Water Level

Open-water level below terrain, in m ________

Figure 1.15. Example of crop yields for different open-water levels.

Box 1.2. Determination of the polder water level.

Table 1.3. Example: Calculation of the polder water level.

Grass land (900 ha)

Arable land (300 ha)

I : One pumping station (maximum energy requirements).

I I : Three pumping stations (minimum energy requirements).

III: Three pumping stations & high " Belt Canal"

Figure 1.16. Pumping stations in polders with three 'sections'.