Professional Documents
Culture Documents
All rights, including translation, reserved. Except as permitted by the Copyright, Designs
and Patents Act 1988, no part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, mechanical, photocopying
or otherwise, without the prior written permission of the Publishing Director, Thomas
Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London E144JD.
This book is published on the understanding that tbe author is solely responsible for the
statements made and opinions expressed in it and that its publication does not necessarily
imply that such statements and/o.r opinions are or reflect the views or opinions of the
publishers. While every effort has been made to ensure that the statements made and the
opinions expressed in this publication provide a safe and accurate guide, no liability or
responsibility can be accepted in this respect by the author or publishers.
Typeset by M HL Typesetting Ltd, Coventry
Printed and bound in Great Britain by .MPG Books, Bodmin, Cornwall
Contents
1.
Introduction
2.
Types of gates
3.
4
4
9
15
22
23
33
35
48
50
51
51
53
59
63
Valves
71
SIuice valves
Butterfly valves
Cavitation in valves
Hollow-cone valves and hoods
Hollow-jet valves
Needle valves
Pressure-reducing valves
Sphere valves
Matching terminal discharge valves and guard val ves
Summary of types of valve
71
71
75
78
84
85
85
87
89
90
17
Contents
4.
5.
93
93
96
96
97
97
Structural considerations
101
Design criteria
Structural design of radial gates
Structural design of vertical-lift gates
Stiffening members of skin plates
Composite construction
101
103
107
107
109
Operating machinery
115
Electromechanical drives
Oil hydraulic operation of gates in free surface flow
Hoist speed
116
121
126
127
Seals
Guide and load rollers
Trunnion assembly
Trunnion bearing failure and lubrication
Trunnion mounting for radial gates
Limit switches
Ropes
Chains
127
137
140
142
144
145
147
148
8.
Embedded parts
151
9.
157
158
166
171
171
171
172
172
Gate vibration
185
185
186
188
191
6.
7.
10.
11.
12.
13.
14.
15.
192
194
207
Control objectives
Operating rules and systems
Telemetry
Factors in the choice of automatic gate control systems
Fall-back system and standby facilities
Instrumentation
207
209
216
217
218
223
Hazard
227
and reliability
of hydraulic
196
199
200
201
201
201
202
202
227
232
234
235
Ice Formation
247
Ice
247
Earthquake
effects on gates
251
252
254
255
255
256
257
257
261
261
Materials
265
and protection
Materials
Steel corrosion and painting
Cathodic protection
Contents
265
266
268
[J
Contents
16.
17.
18.
Model studies
271
271
272
272
273
Environmental
277
considerations
277
281
281
285
289
Index
295
Preface
The success of the first edition of this book has encouraged me to persist in
writing a second edition. Some chapters of the book are in their original form,
with minimal additions. Other chapters have been substantially enlarged. New
chapters deal with hazard and reliability, earthquake effects and environmental
impact of gate installations. These su bjects ha ve become more promi nent since 1
wrote the first edition, and I have attempted to provide an introduction to the
particular problems at gated structures.
As in all fields of technology, the position is never static and a constant
stream of technical papers adds to our knowledge. It is impossible to deal
comprehensively with all the material which has become available; I therefore
had to be selective about what to include. Even without absorbing new
material, it is difficult for a practising engineer to cover the field of hydraulic
,gates and valves exhaustively as it incorporates aspects of many different
engineering disciplines. As a compensation, I like to think that my extensive
hands-on experience puts me in a better position to bridge the gap between
theory and practice, which is essential for the success of any engineering
endeavour.
I am indebted to all those who made valuable contributions to the book
based on their own specialist knowledge, particularly Mr Derek Wilden for
the first edition, and Drs Geoff Ballard and Mike Gardner for the second
edition. lowe much to Dr Paul Kolkman, formerly of Delft Hydraulics, for
h.is generous permission to use material from his extensive papers, fundamental
to an understanding of gate vibration. Thanks are due to Sue Lamb for tracing
the diagrams and to Mr Mark Noble for supplying material.
I would also like to thank clients for whom I have carried out projects for
permission to use information acquired in the course of this work.
Jack Lewin, 2001
Acknowledgements
My thanks are due to my wife Barbara and daughter Jaqueline, who struggled to
translate the manuscript .into a readable form.
Acknowledgement and thanks for permission to reproduce material from the
author's papers are made to CIWEM for Figs 2.8,2.16,2.56,2.57,2.58,2.60,
2.61,2.64,3.16,6.2,
7.13, 10.10 and 10.14.
Acknowledgement
similarly due to:
to reproduce
material are
Introduction
The control of rivers, canals and reservoirs requires weirs or gated structures.
From considerations of reliability and maintenance, the fixed weir is the
preferred control structure. Similarly, the fixed crest free overflow spillway is
the most advantageous arrangement for reservoirs. \X1herever weirs or fixed
crest overflow spillways cannot be accommodated, or where the backwater
stages of a flood or variable river levels are unacceptable, a device which
provides a movable crest or a submerged variable discharge opening has to be
provided. Gates and valves are therefore an essential and critical part of many
flood control schemes, of reservoir management and the control of water in
river courses.
Many types of gate are in successful operation. However, only a few of these
may be suitable or cost-effective for a specific situation. The challenge is to select
. and design a gate of the most appropriate type and size which will meet the
hydraulic, operational, site specific and economic requirements.
Gates and valves control the flow of water, so the hydraulic conditions are
basic to the success of the installation. This comprises not only the flow under or
over a gate but also the upstream and downstream hydraulics.
Since gates are designed for extreme events, personal experience of their
performance under these conditions is by definition limited. Some gate
installations have met with serious difficulties in service. Subsequent research
and published papers, sometimes presented at a specialist congress, are not
always disseminated widely enough to prevent repeated use of flawed design
features. This book attempts to combine available knowledge with practical
experience of gates and valves in civil engineering structures.
The first task is to provide a guide to selecting the right gate or valve.
Information on details of gate design is provided to help in an assessment of
the suitability of a gate for its task, and the chapter on valves is intended to assist
in identifying the correct type of valve for its duty. Ancillary equipment
required in control structures, such as screens, stop logs and handling
equipment, is covered in separate chapters.
Later on, space is given to the consideration of hazard and reliability of gates
and valves. This is part of dam safety, but it is also of paramount importance at
tidal defence barriers. Gated weirs and barrages control flood flow and their
failure can result in inundation and serious damage.
II
Hydraulic gates and
valves
Note on units
The symbols used in the equations of the text, with few exceptions, have no
designated units. Wherever this is the case consistent SI units can be used.
Note on terminology
The term 'barrage' is used in the book for structures which impound water.
The term 'barrier' defines a structure which is brought into operation to protect
a river against an extreme event, such as storm surge.
Types of gates
In this chapter, gates are divided into two groups: in free surface flow and in
conduits. While similar types of gate are fouod in both groups, the design of
gates in submerged outlets, especially at high heads, is more demanding and
requires special consideration.
The main applications,
advantages and
disadvantages of the various types of gate are summarised in the table at the
end of this chapter.
While there are many types of gate, a limited number predominate because of
their advantages. In open channels, at spillways and barrages, radial gates are the
first choice. Except for very large span gates in navigable rivers, new vertical-lift
gate installations are infrequent apart from situations involving rehabilitation
of old barrages, such as the Sukkur and Kotri Barrages 00 the River Indus in
Pakistan where old gates have been replaced by similar vertical-lift gates of
modern construction.
The bottom-hinged flap gate or tilting gate is sometimes preferred in river
courses because it is considered less visually obtrusive than a radial gate. Also its
ability to discharge debris over tbe gate may be important (a radial gate requires a
flap section to carry out the same function). In tidal barrages a tilting gate can
completely prevent ingress of saline water to the upstream pond or reach, and is
often cbosen on this account. Undershoot gates can, under drowned flow conditions, permit a lens of salinewaterto penetrate upstream against the flow of water.
In conduits, vertical-lift gates are used more frequently than radial gates,
mainly due to greater flexibility in installation (radial gates require a large
chamber to retract). The gate slots of vertical-lift gates create hydraulic
problems at high velocity flow, whereas radial gates do not require slots. In spite
of this advantage, vertical-lift gates are often selected because of the installation
problems associated with radial gates.
Detail aspects of gate design are dealt with in separate chapters because they
apply to a number of different gates. The exception is radial automatic gates
where many design considerations are specific to this type of gate.
A number of gates, such as tidal barrages or storm surge barriers, have been
developed for special conditions. So far they have not found more general
application.
Gate types which were once common, such as the rolling-weir gate, drum
and sector gates, Stoney-roller gates and others not dealt with in this chapter
have been largely superseded because of their complexity, cost and associated
civil eogineering construction costs.
!I,
Types of gates
GATE IN
ovmFLOl4
POSlfION-
~
(a)
RADIAL
ONLY
GATE
UNDER
(C)
(b)
FOR DISCHARGE
RADIAL
THE GAfE
OVERFLOW
GAfE
SUBMERGIBLE
WITH
RADIAL
Gf\lE
SECTION
Figllre2.1.
gate
Types of radia!
Overflow flaps are usually operated by an independent motor and hoist gear
or a separate hydraulic cylinder. The overflow discharge section is curtailed on
this type of gate for the nappe to clear the gate arms. Tilting crests may be
required when ice floes have to be discharged in spring or where the river carries
an abnormal amount of debris during part of the year.
The shape of the tilting crest has to be checked so that flow separation cannot
take place in anyone position when the crest section is lowered. If the crest
section is curtailed the nappe will be vented and problems due to nappe collapse
will not occur.
Radial gates have been designed to be submergible to provide both overflow
and discharge under the gate (Fig. 2.1(c)).
/"
(
PIVOT
>-:_j
FRICTION
RAISING
---SIDE-SEAL
RAISING
MASS
OF GATE
FRI(fION
I
I
I
PRESSURE
DISTRIBUTION
AT THE GATE
Figure 2.4. Distribraiou of
pressure bead for radial gate
underfree discbarge conditions
77777777777777777777/~~77777
running friction, and this also applies to bushed bearing although not to roiler
bearing pivots.
Figure 2.4 shows the distribution of pressure head for a radial gate under free
discharge conditions. If the pressure curves are available from flow-net analysis
or actual measurement on a model or prototype, the required components of the
forces may be obtained by graphical integration. An approximation assumes
hydrostatic distribution of pressure over the gate as indicated by the broken
lines in Fig. 2.5. The corresponding force components are equal to the enclosed
areas multiplied by the specific weight of water and the gate width.
This will result in an overestimate of the hydraulic forces acting on the gate.
Rouse1 has given a method of computing the hydraulic forces by taking into
account the discharge characteristics of the gate. This leads to a closer
approximation.
An Appendix at the end of this book sets out the calculations of the hydraulic
forces acting on a radial gate for water pressure on a closed gate, for a gate in the
__
ACTUAL
PRESSURE
ASSUMED
PRESSURE ~
ACTUAL
PRESSURE
PRESSURE
. PIVOT
/
/ /
/
L..
yASSUMED
L.__~
77777/7/77/77777777777~~7777
Types of gates
b
'~;:,
~ ""c:"'!
t ,
-:
:,.
';:J
11,.
PART
open position
conditions.
with drowned
Constructional
features
of gate
.'~.
,':
PLAN
arms
Gate arms are usually offset (Fig. 2.6). This reduces the bending moment on the
horizontal girders connecting the arms and, where necessary, permits the
trunnions to be recessed into the pier. In gates operating under drowned
discharge conditions in rivers it also prevents the trapping of debris between
the gate arms and the piers, and ice bridging between the gate arms and the sluice
.walls under severe winter conditions,
The lowest gate arm is usually positioned as low as possible for structural
reasons (Fig. 2.7).
Under drowned discharge conditions severe turbulence is set up in the
stilling basin and an unsteady roller occurs.f If the roller acts on the submerged
gate arms or on other structural members vibration is likely to occu.r.3ln this
case hydraulic considerations should override structural priorities so that
members are disposed in the most efficient manner (Fig. 2.8).
The hydrodynamic effects on the gate, such as wave action, are taken into
account in calculating the load on a gate. Hydrodynamic downpull forces (see
Chapter 9) have little structural effect on radial gates in open channels but must
be taken into consideration when calculating hoisting forces. They become
--
...
77777777777777
Figure 2. 7. Positioning of
arms oj {I radial gate
PREFERRED ARRANGEMENT
OF
GATE ARMS DUE TO HYDRAULIC
CONSIDERATIONS
Figttre2.8. Effectofreuerse
roller 011 tbe /OIUC,ft arm if a
radial gate under drowned
rli.fcbargc conditions
//7/7/
//777///
more significant when radial gates are used as culvert valves under high head
conditions (see Fig. 2.9).
Two-dimensional analyses of the strength of the weir plate assembly ignore
the additional strength and rigidity due to its curvature. To take this into
account, a three-dimensional analysis, such as finite elements, is required. This
is used mainly when gate vibration or the dynamic response to an earthquake is
in vestigated.
There are two main methods of constructing the skin plate assembly. The
skin plate is either stiffened vertically, as shown in Fig. 2.10(a), or horizontally
as shown in Fig. 2.10(b). The advantages and disadvantages of each method are
discussed in Chapter 5. Both methods of reinforcing the skin plate use main
structural members to tie the gate arms. Together with the gate arms they form
a portal which transmits the load to the trunnions. All members which are
welded to the skin plate are assumed to act compositely with the plate.
used
Types of gates
VERTICALLY
STIFFENED
HORIZONTALLY
STIFFENED
mater Ie/leicontrol
WIDE;
3.5m
HIGH
the displacer chambers and flow. to the river downstream of the gate. Due to
increased buoyancy of the displacers, the gate rises and discharge occurs under
the gate. Rise of the gate may cause a slight drop in upstream water level, which
will result in reduced flow over the weir and hence a lowering of the level in the
displacer chambers. The gate may close slightly as a consequence until balanced
conditions are achieved. The stability of the system is provided by head loss in
the inlet system and side-seal friction.
GAlE
____ ~
",SPlA:"
FRI( II Otl------------
PlV",
r- --
--==
SEAL
COUNfERBALANCE
WEIGH r
RI
__.\
\
/'
'(
\
ItIlH
wEln
OUlLE!
Types of gates
INTAKE
FROM
RIVER
minimise obstruction of the inlet due to flotsam. The pipe is usually made
large to enable rod ding out. The head loss, which should be 50 to 65 rnm, is
'provided by the setting of the inlet valve. Figure 2.13 shows a typical
arrangement.
the inlet weir is drowned and the gate operates according to the ratio Rl(R2 as
shown in Fig. 2.12. The gate will therefore lift more rapidly than the rise in flood
water and will cut out of the water to permit unobstructed flow under these
conditions.
If the initial difference between upstream and downstream level is 2 m or less,
the stage-discharge relationship must be known when the design is carried out,
so that the gate control system is arranged to avoid the downstream level taking
over control of the gate during some stage of the rising flood. This can also be
important during a falling flood when the downstream level falls more slowly
than the upstream level, resulting in a condition where the downstream water
level keeps the gate open and causes loss of retention level.
The displacers are designed so that their specific gravity is 1.05. Construction
takes the form of a stiffened box with watertight access covers for loading weights.
The total assembly forward of the pivot is out of balance to the extent that the
displacers are half immersed when the gate is in a steady condition. Equal forces
are then available for opening and closing the gate. With this arrangement the
mass of one displacer must be at least equal to the friction force of one side seal
multiplied by R1jR2 plus the pivot friction. A gate with displacers sized on this
basis would have no margin and would provide very coarse level control. With
displacers designed three times this size they will give good service.
The wide piers required to accommodate clisplacer chambers can be a
disad vantage in a sluice installs tion.
If self-aligning roller bearings are used for the pi vots the effort at the
displacers to overcome friction is negligible. For bronze bushed bearings it will
amount to 1-2.5 kN at the displacer.
It is normal practice to use displacers, but there are some gates where floats
have been used. Figure 2.14 shows a design where floats are held stable by the
link 'a' so that pivots 'b' and 'c' and the link form a parallel motion. In a 6 rn wide
gate, using Boats would result in an overall saving of 2.8 tonne.
To ensure consistent operation throughout the range of travel of the gate, the
centroid of the counterweight, the centre of the pivot bearing, the centre line of
the displacer suspension pivot and the centroid of the skin plate assembly should
be in one lioe.4 Departure from th.is is possible, provided the gate operation is
calculated separately for different openings.
Types of gates
Trunnions
Most gate manufacturers
use phosphor
bronze bushed bearings. Selflubricating bronze bearings have a starting coefficient of friction of 0.1-0.12,
which reduces to 0.07 during running. The gates in Figs 2.11 and 2.15
incorporate self-aligning roller bearings. Their coefficient of frictioo is 0.0018.
Using self-aligning roUer bearings eliminates uneven bearing pressure due to
deflection of the pivot pin in a bushed bearing. Self-lubricating bearings,
mentioned in Chapter 7 and illustrated in Fig. 7.16, significantly reduce
maintenance. They are available as plain bushes or self-aligning bearings.
Counterbalance
This can take the form of a reinforced concrete beam with pockets cast into the
upper section so that final balancing of the gate can be carried out. \X1here
DIAGRAMMATIC
LAYOUT
FLOAT CHAMBER
DOWNSTREAM
LEVEL
Downstream
Gates for downstream water level control use a float chamber downstream of
the trunnions (Fig. 2.15). These are only suitable for small gates, because the
discharge under the gate causes turbulent conditions. Larger gates have to be
severely damped, or displacer chambers in the piers must be constructed using
an intake positioned downstream of any turbulence. It is usual to baffle the
intake. Gates of this type are not subject to size limitations.
Causes ofinstabi/ity
and malfunction
of radial
automatic
gates
Malfunction
Such a program was run for the gate shown in Fig. 2.11 during the
preliminary design phase. The results demonstrated that the gate cut out of
the water too quickl y because the rapid rise in downstream level took over the
control of the gate. This caused an unstable condition because the upstream
level then dropped below the retention level. The program was rerun,
increasing the water level in the displacer chamber and the outlet from the
displacer, until stable conditions were obtained.
Types of gates
. ..
I
.
.
..~ I
+.:_ .._.<_:
SECTION
A-A
_"-'_
SECTION
~-.~A
.
-r
B-B
. ,!,. .
!!i
:,:
-~.-~A
Iii
PLAN
Control of the gate response to inflow into the float chamber is by a penstock
or a weir at the inlet to the pipe system and a valve at the outlet from the float
chamber. The tripod above the float chamber is for manual raising of the gate.
Gates wider than 10m have a float and counterweight at both sides. Quintela
et al. G list 37 gates of this type in Portugal. The gates appear to have been reliable
in service except for two gates where the hoist ropes failed due to corrosion. The
risk of malfunction due to .obstruction of the inlet to the control pipe and
jamming of the counterweight or of the float is mentioned in the paper. The
possibility of the hoist or the control ropes leaving the diverter pulleys must
be a further risk, also silt deposition in the chambers when the flood low carries
silt in suspension.
In South Africa, other automatic water operated gates have been developed
by Flowgate Projects of Randburg, South Africa. Their operation depends on
buoyancy tanks which are integral with the skin plate.
Operation of the automatic crest gate is shown in Fig. 2.17. It requires a
spillway weir with a vertical face on the upstream side to permit the gate to recess
under flood conditions. Level control achieved by this type of gate is coarse
compared with that of radial automatic gates. The inlet weir has to be protected
from wave action and debris. The face seal at the crest of the weir requires
BUOYANCY
TANK
ADJUSTABLE
OUTLET
VALVE
crest
Types of gates
INLET WEIR
FSL
SCOUR GATE
BUOYANCY
TANK
SCOUR
TUNNEL
SCREEN
CHAMBER
DRAIN PIPE
INFLOW
INTO WEIR
WATER LEVEL RISING
FSL
FLOW
accurate fabrication of the skin plate. Under overflow conditions, the nappe has
to be aerated (see also section on flap gates later in this chapter).
The automatic scour gate shown in Fig. 2.18 can provide scou ring and
maintain upstream water level.
Vertical-lift
gates
are:
scour
lifted, the counterbalance is reeved 2:1 so that it travels for only half the distance.
This results in an additional load on the superstructure of the order of2. 7 times
the mass of the gate, and requires a substantial support structure, adding to the
cost of gate installation. Most gates of this type that are used in open channels arc
of the fixed roller type (Fig. 2.19(a.
In gates in open channels, rollers are usually spaced out to take an equal load
of the hydrostatic forces acting on the gate. Roller alignment is critical, as
uneven contact of a roller can overload adjoining rollers. One method of
adjusting rollers to ensure equal contact on the roller path is shown in Fig. 7.13.
Downstream sealing of a gate is preferred because the water load compresses
the seals. Upstream sealing is required where a gateis located upstream in a shaft,
and access for inspection of the tunnel and the gate is via the shaft.
Fig. 2.19(b) is a diagram of a Stoney-roller gate. The roller trains, one on either
side within the gate slots, move at half the speed of the gate and are reeved 2:1 on
their suspension. Tn practice, the load distribution from the gate via the rollers to
the track is not even, and the permissible contact (Hertz) pressures for Stoney
rollers is half of the pressure for load roller wheels (see also Chapter 7). The rolling
load is transmitted from the rolling face on the gate directly to the track face and
Stoney-roller axles should theoretically be subject to only nominal rotational
friction. Tn fact this is not the case, since the rollers deform elastically imposing a
load on the axles or bushes, if they are fitted. In addition, inaccuracies in alignment
of the rollers can impose considerable additional loads. Breakdown of individual
rollers has occurred as a consequence. The slack in the tracking of a Stoney-roller
gate can cause gate vibration, especially under conditions of high velocity flow.
Stoney-roller gates have fallen into disfavour. There are, however, many gate
installations of this type which are being refurbished or will require
replacement.
Vertical-lift gates can also be fitted with overflow sections (see Fig. 2.20),
where limited overflow is required.
Where substantial overflow must be effected a hook-type gate is used (Fig.
2.21). This results in considerable complexity of construction, and high
accuracy of manufacture is required to maintain the seal between the two
sections.
Hook-type gates use either a single hoist so that the upper section is hoisted
first and when it reaches the full extent of travel it moves together with the lower
section, or two hoists. The latter arrangement is required for combined overand underflow or when the gate is to be used for underflow without moving the
upper leaf (sometimes required for aesthetic reasons). The hydraulic conditions
caused by combined over- and underflow can induce gate vibration and the
range of operation of such a gate may have to be restricted.
For long span vertical-lift gates where the skin plate structure is backed by
girders and where there is drowned discharge, similar considerations to those
mentioned earlier in this chapter concerning arms for radial gates apply. The
turbulent discharge conditions downstream of the gate, and an unsteady roller
- if it develops - can act on the structural members and cause local or general
gate vibrations. This is discussed further in Chapter 10.
Larger vertical-lift gates arc often manufactured
in sections with an
articulated joint between them. Figure 2.22 shows how this can be effected.
STONEY-ROLLER
KOTRI BARRAGE
PAKISTAN
1 B. 3m WIDE
6.3mHIGI-l
GATE
--x'
:,,04
~:
G 4"
,:
~:
.,:' "~
::-
(a)
1. DOWNSfREAM
2. UPSfREAM
SEALING.
. (al
(b)
SEALING.
FIXED
ROLLER
GAfE
Ibl
STONEY-ROLLER
GATE
The connecting bolts permit limited back and forth movement between
sections to allow the guide rollers of anyone section to centre. The connection
bolts are prestressed to IUnit deflection under load. Apart from the saving in site
welding that such designs permit, it is often possible to mount four wheels on
anyone section and omit adjustment of the guide rollers.
Figure 2.22 also shows the seal between the sections. This has to be jointed to
the side seals. This is more easily effected if the seal is on the downstream side of
the gate. An intermediate short section of a seal may be required if the seals are
not in the same plane.
An articulated vertical-lift gate has to make provision for transfer of shearing
forces from one section to the other so that racking forces do not shear the seal
between sections.
Transverse guidance of vertical-lift gates is provided by separate guide
rollers or slides.
HOIST
,-,\
i'----- -I
1
(j HOIST
FOR LOWER
Types of gates
LEAF
CURVED FACE TO
PREVENT
SEPARATION
SIDE SHIELDS TO
CURTAIL OVERFLOW.
VENTS NAPPE.
PROTECT HOIST CHAINS
AND GUIDE ROLL ER SLOTS
UPPER LEAF
LOWERED TO
EFFECT OVERFLOW
LOWER LEAF
RAISED TO
EFFECT UNDERSHOOT
...
- ; -:'.
-::--:-,-,--.---c,---.--~--c-_
Figllre2.21.
Hook-IJ'}Jegate
SEAL
PART OF
LOWER
SECTION
ONE OF TWO SETS
OF CONNE[ TO RS
OF GATE SECTIONS
SECTION A-A
VI EW ON ARROW'B'
Figure 2.22. Vertical-lift gatc
mansfactured in sections,dctailof
connectors
GAlE SECTIONS
TRANSVERSE
The gates are used to shut the intake to the turbine and can be operated under
balanced head or, iu emergency conditions, at an unbalanced head. Gates
capable of operating at differential heads of up to 30 m are still in use and have
proved to be very reliable. They were manufactured up to the 1970s.
p.._
,y
~
'\
"q1iT
, //
'"
-- 1 jl
t--
\
\
- .
I
I
E:~
J/
./
_'i'
_)1
,t-
~~
,'-
I=d_
J/
"\,
."
/_/
'1111"
"'-9 FE:
F IXED IN
BUI FREEE
__,
L
L.
L_
Flap gates
Bottom-hinged
flap gates
Bottom-hinged flap gates are used in tidal rivers to prevent the ingress of saline
waters or where special environmental
considerations
apply. They are
sometimes selected for reservoir spillways because they can be made to operate
~:I:"_~:.I~\~:~:
.....
,..;..
( a)
(b)
(e)
Types of gates
3 mm along both axes over the distance of gate movement. While this saves the
cost of embedded side plates, it exacerbates wear on the seals and increases
construction costs due to the requirements for accuracy and finish of the concrete.
Seal designs for bottom-hinged flap gates are illustrated in Chapter 7.
Where gates have to be fully retracted, the shape of the skin plate must be
formed so as to avoid low separation and subatmospheric pressure.
The arrangement shown in Fig. 2.26 can be a trap for debris. Where
operational reliability is paramount, it may be necessary to provide means of
clearing debris without dewatering the sluiceway or using a diver. This can be
done using jets of water, compressed air or by flushing with river water. No
information regarding practical experience in the use of compressed air
distribution systems for this application has been published. If river water is
used to flush out the recess, the design must avoid the creation of eddies and
dead pockets which will cause the transfer of debris from one part of the recess
to another without dislodging it. Debris will be trapped behind the overflow jet
and a bypass system, shown in Fig. 2.27, is necessary to clear flotsam. The gate is
elevated to stop overflow when flushing out.
Tilting gates can be manually or motor operated through screws, or actuated
by means of two hydraulic cylinders on either side of the gate or by a single
cylinder centrally positioned or on one side only. Small gates are more often
operated by ropes. The arrangement of the hoist machinery is then similar to
that shown in Fig. 6.1.
Rising screw-type gearing with twin lifting screws operated from a central
headstock gives the required large mechanical advantage and also provides a
self-locking feature which resists the water load tending to reverse drive the
gears. The use of exposed lifting screws either in the sluiceway or in a recess in
sluiceway walls is a potential source of malfunction due to contamination of the
threads by dust, insects and silt, especially when coated by an adhesive lubricant.
FLOW DIVIDERS
NAPPE
VENTING
Types of gates
DUCTS
II
II
II
II
II
II
--kJ1-'f-~--
---'--~-""-
"/
ROLLER
The resulting increase in friction of the screw threads has caused seizure of
screws, particularly when multistart threads have been used. Totally enclosed
screws operating in an oil bath are preferred.
Oil hydraulic operation by placing the ram under the gate makes it possible to
dispense with an overhead structure. Maintenance of the tam then requires
dewatering of the sluiceway, often considered a major drawback of this
configuration.
At the Baja sluices of the Dinorwic Pumped Storage Scheme in Wales,
planning requirements stipulated that no overhead structure should be
provided. The gates were therefore designed as a torque tube structure with
the pivot shaft extended into the hollow piers, where they were operated by a
hydraulic cylinder acting through a lever (see Fig. 2.28).
A speciaJ case of a bottom-hinged lap gate is the velocity-control structure in
the River Orwell at Ipswich.f 10 its elevated position, the gate acts as a
submerged weir reducing the cross-sectional area of the river, reducing the flow
and scouring action on the river bed.
Automatic tilting gates have been constructed with counterbalance above or
below the gate (see Fig. 2.29). The counterweight is arranged to balance the
overturning moment of the upstream waterload at normal retention level. With
a rise in level the gate becomes overtopped and the overturning moment is
increased. When this overcomes the resistance of the counterweight, the gate
opens. By careful proportioning and positioning of the counterweight system
and tbe pivot point these gates can be arranged to open and close in a series of
movements on a rising and falling upstream level. The degree of control is not
I'
Figllre2.28. BalaSllIicesoj
the Dinoru/icPum/Jed Storage
ScJJelJJe
,I
Types of gates
<,
J.v
II
II
1\
II
II
1\
.,
.,,'.:
accurate and a significant variation in level is required to effect full gate travel. If
on discharge the downstream level starts to rise, thus creating an overturning
moment tending to close the gate, it may be impossible to prevent the upstream
water level from rising.
The gate arrangement shown is inherently hydraulically unstable.
Disturbance may be set up due to wave motion in a reservoir, or a pulsating
surge or wave may be set up in the upstream reach of a long approach channel
of uniform section. Surging can be initiated by level drawdown immediately
upstream of the gate following a downward movement. The loss of water load
on the gate then causes a closing movement, and if the frequency of gate
oscillation coincides with that of the surge wave the gate movement is
accentuated and can become dangerous.
The only damping force present is the friction of the side seals, hence
hydraulic dashpots have to be added to the counterweight
system. A
control system which is more stable utilises the same principle as radial
automatic, gates by arranging the counterbalance weights so that they act
as displacers.
Venting
The nappe has to be vented to prevent gate vibration and nappe collapse. Flow
dividers are used to vent the nappe under moderate overflow conditions. The
design and spacing of flow dividers is critical.i' The d.ividers must project beyond
the gate lip and they must be wide enough to form an adequate opening in the
nappe for the admission of air. The flow of water over the gate lip expands but
as soon as it is no longer in contact with the divider it tends to close it up again.
Experimental work and prototype trials have been carried out to study nappe
oscillation and resulting vibrations.9-14
Model studies to Froude scale of
bottom-hinged
flap gates' incorporating
flow dividers are ineffective in
preventing self-excited nappe oscillations and resulting vibrarions.l" The
spacing of flow dividers is important. Pulpitel" gives some information on
flow dividers which can be applied in practice. Initially, flow dividers were
spaced at 2100 mm centres. Additional flow dividers were added between the
original ones. These projected 280 mrn beyond the skin plate and had a crest
width of 300 mm and a width over the tapering side sections of 450 mm. The
maximum overflow depth was 720 mm.
When the head of water above the gate lip is appreciable, flow dividers
become ineffectivel7 and additional venting through the sluiceway walls or
the piers has to be provided. A method for calculating the air demand of an
overflow jet is given in Chapter 9.
Top-hinged
flap gates
,I
Types of gates
Figure 2.30.
gate
Top-hillgedflap
conduit runs full, the eccentric hinge opens a gap which will cause flow over the
gate leaf in the open position. However, the eccentric binge arrangement
reduces the total mass of the flap compared with the weighted flap and therefore
provides an increased discharge for a given opening. This is due to the
relationship between the discharge and the mass of the flap. At a given gate
open.ing these are exponentially related.
Sealing is usually effected by face-to-face contact between the flap and body
of the gate. This requires exact positioning of the hinges. To ensure even contact
SOGREAH
Documentation
)!li?'i
':6:
Figllre2.J1.
Top-hillgedjJap
gate lIJith elastomeric le(1
THREADED
STUDS
between the sealing faces the pivot lugs are made adj ustable, One example of this
is shown in Fig. 2.32. The cushioning of the gate leaf in Fig. 2.33 is effected by
the movement of the projecting section of the flap in the seat. This acts like a
piston moving in a cylinder. The extent of cushioning is determined by the
clearance between the faces. The section in the seat in contact with the flap must
be tapered because the flap moves in an arc. If elastomeric seals are fitted they
should not be located on the face of the flap because this causes the flow to
separate. Seals should therefore be mounted on the frame. Under flow
conditions the flap rides the discharge jet. Under free discharge as well as some
conditions of drowned discharge a reverse roller forms at the lip of the flap (Fig.
2.34). This is similar to the discha rge conditions at underflow gates of the radial
and vertical-lift type. If the conduit flow is near full or under supercharged
conditions, transverse flap stiffeners at the lip of the flap can cause flow
reattachment which is likely to cause vibration. Transverse flap stiffeners sbould
be set back and the lip of the flap should be stiffened by webs leading from the
transverse stiffener to the leading edge.
When a series of flap gates are close to one anotber, such as in estuarial tidal
outfalls, the sideways discharge under the gate causes flow interference between
adjoining gates and results in hydraulic losses (Fig. 2.35). If the discharge through
the gates has to be maximised, the losses must be reduced by training walls. These
should extend as far as the length of the sector swept by the opening of tbe flap.
The theoretical treatment of the stage-discharge relationship of a flap gate by
Pethick and Harrison presupposes that there is no sideways discharge. It is
reproduced in Chapter 9.
Types of gates
SEAT
Wave action or flow reversaJ will cause rapid movement of the flap and can
result in severe slamming. This can be damped by a hydraulic cushion (Fig.
2.33), or by ao oil hydraulic damper. For a large gate it can be of the oleopneumatic kind, or a torsional damper can be introduced in the hinge assembly.
Another method of suspending top-hinged flap gates was used on the
lshmalia gates. This incorporated a cycloidal rocker-type bearing instead of
pivots or hinges and proved reliable in service.
'Fuse gates
Fuse gates consist of an alignment of elements standing on the crest of a dam on a
concrete sill with a toe abutment (Fig. 2.36). Underlow flood conditions they act
as a labyrinth weir (Fig. 2.36(b )). At higher floods, water flows into the well and
into the hollow base set over the sill (Fig. 2.36( c)). The uplift pressure on the base
and the change in the centre of gravity of the fuse gate causes the gate to overturn
CORRECT
::.::_=:._~.,"
"''\
(If
Hydraulic
II
gates and
valves
SPILLWAY
SEAL
BASE
SILL
ASSMBL Y
FUSE GATE
a ding
as a weir.
(d) Fuse
when the design flood level is reached (Fig. 2.36(d)). Fuse gates are made of steel,
concrete, or a combination of both. A seal is provided between adjoining fuse
gates. Their main application is to increase the reservoir capacity of existing
uncontrolled spillways, although they have also been fitted to new dams.
The discharge characteristics of fuse gates are similar to those of an ungated
spillway until the elements start to tilt. For large floods, the elements tilt
independently of each other, progressively following the rise in head water
level. Discontinuities arise as each element is activated. The reservoir retention
level, after the flood has subsided, is then the level of the concrete sill until the
fuse gates have been recovered and repositioned. When the elements overturn
they can be recovered just downstream of the dam and can sometimes be
reutilised, which may require reworking. If the velocity of flood discharge is
higher than 3 mis, the elements are carried away by the current.
Fuse gates can be set on a multiple level sill to refine flood routing.
Floating debris is said to have no effect on block stability and little impact on
the precise timing of tilting.
Existing dams, if fitted with fuse gates, have to be checked for the higher load
due to an effective increase in height and the transfer of the shearing forces
exerted at tbe sill.
For small and moderately large dams elements ranging from 0.5-2.5 m high
have been used and for larger scale projects fuse gates up to 6.5 m high have been
designed.
The main criteria for selection of fuse gates on an existing dam are:
o the additional loads imposed by fuse gates.
The main criteria for selection of fuse gates on all dams are:
o The required limits of control of the reservoir level. This may have to include
the reservoir level after tilting of fuse gates during a flood.
o The time and effort required to recover fuse gate sections after a flood.
Possible replacement of damaged gate sections or repair of reusable units.
o The frequency of flood events which cause fuse gates to tilt.
Types of gates
AIR
GATE RISING
FLOOD
CON TROL
AIR
GATE
GATE
GATE IN OPEN
LOl1ERING
IN MAINTENANCE
POSlfiON
POSITION
F igIJre2. 37. Thames Barrier and principle of operation of tbe Thames Barriergates
Aye/G)
(after
Types of gates
navigational depth of the opening (Fig. 2.37). To close the gates they are rotated
through approximately 90 until the flat skin plate is near to the vertical and the
curved skin plate is facing seawards. For maintenance, the gate is rotated
through a further 90 until it is fully inverted. A gate segment is supported at
its ends by two disc-shaped arms (Fig. 2.38).
The arms rotate on a shaft and bearing assembly (Fig. 2.39). The bearings are
self-aligning and are of the self-lubricating type described in Chapter 7.
Provision for supplementary lubrication is provided.
The main gate operating machinery is shown in Fig 2.40. It consists of two
hydraulic cylinders (1). They rotate a rocking beam (2) through crossheads and
links (3). The cylinders are disposed on either side of pivot shaft (4) of the
rocking beam. The cylinders develop a thrust of 15.21 MN at a pressure of
17.24 N/mm2. The rocking beam operates the disc shaped arm through the
connecting link (8).
To raise the gate, the upper cylinder pulls and the lower cylinder thrusts the
rocking beam causing it to rotate about its pivot, raising the gate link and so
rotating the gate. To rotate the gate through the link dead centre position a shift
and latch mechanism is provided. It is actuated by a hydraulic motor driving a lead
screw. Itis also used to lock the gate in the open, closed and maintenance positions.
SECTION THROUGH
~.:a:.::_
10
5
sz
-2375m
sz
-9'250m
RISING
..
0
SECTOR
.
GATE
_
10m
DOWNRIVER
sz -2830m
M.L.W.S.
LEVELS RELATE
TO ORONANCE
DATUM NEWLYN
GATE ARM
END
GATE SPAN
II
L PIER
GAlE
SIRU[fURE
GAlE
ARMI------lI-~
ARM
OMlllEO
SPHERI[AL
BEARIIlG
HA[ALLOY
WAIER
BARS
15mm OIA.BARS
SEAL
SECTION
Ii.
OF TRUNNION
Identical units are provided at each end of the gate. They can be used in an
emergency to raise a gate should the main machinery be out of action.
A high degree of redundancy is built into the system, specifically to the
power packs and gate raising machinery. The gates can be elevated into the
SEClION
THROUGH
METHOD OF OPERATING
ENLARGED
RI~lfIG
SECIOR
RISING
GAlE
SECTOR GATES
MECHANISM
Types of gates
defence position by the machinery on either side or by the two sets working
together, the fall-back being provided by the latch mechanisms. This also
applies to the operating controls.
Bottom-hinged
buoyant gates
The design of the Venice Barrier gates, which have so far not been built, was
evolved to avoid any piers within the navigation ways from the Adriatic Seainto
the Venice Lagoon. Cruise ships up to 30000 tonne displacement pass through
the Lido passage to the port of Venice and very large supertankers enter the
Lagoon through the Malamocco opening.
The gates recess into caissons in the navigation ways (Fig. 2.41). To raise
them, compressed air is admitted into the gates and water is expelled causing
the gates to rise into their operating position at an angle of 50" to the horizontal.
The gates can withstand a differential head between the waters of the Adriatic
and the Lagoon of up to 1.8 m. This is due to their buoyancy and mass. The four
barriers, each 400 m wide, are comprised of gates 20 m wide, moving
independently of one another (Fig. 2.42). There is a leakage flow between the
gates but the effect on the water level of the Lagoon is negligible because of its
large area (about 500 km\
Hydraulic
gates and
valves
20
40
60
================
BOrn
,;:.:.:.::'.:,'
LJ3v;.,..~~=S?===kk
F igtlre 2.43.
The gates can be transported by crane to a platform on the shore for servicing
and a replacement gate can be substituted (Fig. 2.43). Detachment and
reattachment of the hinges is by a series of hydraulic cylinders radially disposed
around the mandrel of the hinge (Fig. 2.44).
Air is admitted and exhausted via pipework in the service gallery of the caisson.
There are two independent, physically separate systems. The air is routed through
the detachable hinges. Detachment of the hinges automatically blocks the air
admission ports. In case silt is deposited on a mandrel of the hinge when a gate
has been removed, a flushing system is incorporated in the permanently attached
part of the hinge and is actuated prior to engagement of a replacement gate.
Sand and sedimentary matter will be deposited in the caissons during slack
tides and wave action. The caisson recess for the gate is in the form of a series of
troughs. A hydraulic transport and ejection system, consisting of evenly spaced
jets positioned above the crest of the troughs, clears about 80% of the material
which accumulates in the troughs. This is effected by supplying water under
pressure to the jets which then move the deposited material along the troughs
to the end, where jet pumps elevate and expel the silt and sand into the
..
21
navigatIon way.
The barriers and gates have been described by Lewin and Scotti.22
A large span vertical-lift gate forms the River Hull Tidal Surge Barrier (Fig.
2.45). When the gate is hoisted to its uppermost position it rotates and comes
to rest in the horizontal plane, reducing the overall height of the support
structure. This method of storing a gate in the fully open position is sometimes
adopted for vertical-lift lock gates. The upstream lock gate of the Kotri Barrage
on the River Indus in Pakistan is an example of this type of construction.
The flood surge protection gate at Barking Creek on the Thames,
downstream of the Thames Barrier, is of similar design (Fig. 2.46). The
navigation gate is 38.6m wide and 10.8 m high. The gate can be raised 39.4 ill
for ships to pass underneath. In the open position the gate is stored vertically.
The navigation gate is flanked on one side by two vertical-lift gates and on the
other side by one gate. The gate is counterbalanced and each end is suspended by
two Renold chains of165 mm pitch shown in Fig. 7.19. Each roller has a grease
nipple to lubricate the roller bush.
Two sets of hoist machinery are provided, located at high level in each tower.
A shaft interconnecting the hoists runs along the bridge tying the towers. Either
one of the hoists can be used to lift the gate but they cannot be used together.
In an emergency, the navigation and all the side gates can close under gravity.
Closure is then controlled either by hydraulic retarders or by hydraulically
operated disc brakes, mounted on the shaft which drives the chain hoist
sprocket. Figure 2.47 shows the arrangement of the overspeed disc brake at
one of the hoists of the navigation gate.
Types of gates
v,v,
0: ......
'w~
-.I<{
e::>
'"
......
,__
'"
CI:
6'"
CI:UJ
"';:;
......_.
:::Vl
>-
0::.
0>-
a:
3:
'"
>-<{
_J-
-'
-,0
<tVi
3:
UIll
'rr:::>
VI
VI
,__Vl
W
u
u
......0::
--'0
<{
<uLl-
'.~
t:
"
I'Q
~
!?;
c.-;
-.
'>i
~
h
~
~
>-l.,
>- I
c:
,
::;
i;;
.~
x."
>:c::
::>eo:
Vl
<(UJ
'"'"
0:..-'
...0
"l-
>'"
"-i
a:
0""
"'0
C>:_
W
I!)
::>
c,
Vl
J,
-'
'"
~
.~
>L.,
::>::>
,__'"
__
Types of gates
j_ NAVIGATIO'*
GENERAiGR
fCHAflNEl
!WllDING
I
I
I
I
,
,
IrL..",..oJ'U'1f'U"J
navigation opening leading to the dock is 30.498 m wide. The bed level at the
apron on the Thames side is at -11.85 m and the lood defence level is +7.2 m.
The structure consists of a bridge moving on a roller track (Fig. 2.48). The
overall length of the moving bridge is 72.5 m. A flap gate is hinged from the
bridge. To close the navigation opening, the bridge moves laterally so as to span
the lock entrance. The flap gate is then lowered to abut on a step in the
navigation way. The hydrostatic thrust due to a surge flood is resisted at the
bottom of the step and at the top it is transmitted via the gate binges to the
moving bridge. The bridge transfers the force to two thrust blocks on either
side of the lock entrance.
The bridge is surmounted by two hoist winches for lifting and lowering of
the gate. The winches are driven by low speed oil hydraulic motors. The bridge
IlRIOGf SIIW(JIIRE
is moved by rope sheaves which are also driven by oil hydraulic motors. The
sheaves utilise fixed haulage ropes. A stationary emergency winch can launch
the bridge but cannot retrieve it.
The gap between the gate and the underside of the bridge structure, where
the gate hinges are located, is sealed by hydraulic cylinder operated flaps.
Types of gates
affecting the sealing of the gate. The gate can be opened and shut against flow in
either direction. A disadvantage of this type oflock gate is the requirement for a
iarge recess in the river wall to accommodate the gate leaves when they are
opened.
Large sector lock gates safeguard a river on Rhode Island on the Atlantic
coast of the USA against surges due to hurricanes.
The largest example of a sector gate is the New \"X!aterwaystorm su rge barrier
in The Netherlands.2..1,24 It protects the Rotterdam area against floods from the
North Sea by closing off a 360 m wide waterway. The barrier consists of two
movable box shaped sector gates, each having an arc length of 203 m and a
height of 22 m. The radius at the sea side of the box is 240 m. Two trussed steel
arms support each of the buoyant boxes and transfer the water load to a ball and
socket joint (Fig. 2.50). The joints, of 10m diameter, allow the gates to be
loated into position. When water is admitted the gates settle on the sill 17 m
below datum.
In the open position the gates are parked in dry docks shaped in an arc. The
operation of closing and opening the gate leaves is by a 'locomobile'. It is fixed
and transfers traction to rails on top of the gate wall, thus the locornobile rides
on top of the gate sector, imparting horizontal movement. Figure 2.51 shows
the arrangement. The connection between the locomobile and the support on
the dock allows for the vertical motion which occurs when the gate is ballasted
and when it is buoyant. While the locornobile remains stationary, the gates
move horizontally.
The gate is required to lift to equalise a differential seaward head, or a higher
level on the river side. The initial design (Fig. 2.52(a)), was unstable under these
STEEL
WORK
II)
STEEL SECTOR GATE
I Z) SILL AND BOll0M
PROTECTION
13) PARKING
DRY DOCK
II.)
ABUTMENT
15) CONTROL CENJRE
16) BALL JOINT FOUNDATION
BLOCK
171 LOCOMOBILE
AND GUIDANCE
lOWER
TRUSS
CONCRETE
SOCKET
LUBRICATION
FILM
BALL OF
STEEL PLAiES
5m RADIUS
CO~lPAC1ED
SAND-FILL
[ROSS-SE[TlON
OF BALL
JOINT
conditions because the gates were strongly influenced by their own bottom side
geometry. Model studies arrived at a design which ensured stable operation
under both positive and negative heads (Fig. 2.52(b.
Other model tests investigated bed protection. Riprap bed protection was
required on both sides of the barrier in order to withstand hydraulic loads
during the closure operation (flood) and the opening operation (ebb).
Conventional vertical-lift gates were used at the Eastern Scheldt Storm
Surge Barrier in The Netherlands. The movable barrier comprises 63 gates
of 42.5 m width with height varying from 5 to 11.9 m and a depth of 6 m.
They are operated by a hydraulic cylinder at each end. The largest rams
have a stroke of 12 m and an internal diameter of nearly 2 m. If severe wave
action is encountered the cylinders are required to hold the gates rigid
under all hydraulic conditions, including waves impacting upwards and
downwards.
Types of gates
Bottom-hinged
flap gates
(b)
(ONT RACT
DESIGN
AMENDED
DESIGN
(after lepemP)
II
leaf can be lowered to provide overflow from the upstream reservoir and the
lower leaf can be raised for discharge under the gate when the ebb tide is below
the sill level and it is necessary to discharge silt accumulated at the gates.
Pointing gates
Pointing gates are vertically hinged double leaf flap gates. They operate
automatically without external power on movement of the tide. They arc used
on small ri vers to preven t the ingress of tidal water and provide protection
against storm surges. Older gates were constructed of hardwood. New gates
of this type have been constructed in steel. The gates open at low tide due to
river flow and close on the rising tide. Sealing is by face contact with the
masonry structure. The gates can slam on closure. Damage to masonry works
sometimes occurs as a result of violent closure. No attempt appears to have been
made to damp or to provide buffers on gates on the Somerset Levels in England.
10 existing installations, pointing gates are backed by vertical-lift gates which
provide a standby in case of gate breakdown or malfunction due to debris being
caught between the gate leaves, preventing complete closure.
SEA
BASIN
CABLE
Types of gates
DU[JS
MAXIMUM
MINIMUM
LEVEL
LEVEL
These gates are not suitable for low dams because of the deep excavation
required and the possibility of flooding of the float chamber due to downstream
water level. Some very large drum gates have been built, up to 40 m long and
9 m high.
Drum and sector gates have been superseded by radial gates at spillways
because the former are more complex to manufacture and therefore more cost! y.
The cost of civil engineering works associated with drum and sector gates is
significantly higher than with radial gates,
SERVICE ROAD
CABLE
M ACHIN~ f/Q!lt1
DUC1S--~rr+r~I+-
SEA
BASIN
MAXIMUM
LEVEL
MINIMUM
LEVEL
,---I--PANEL
PLACED
FOR llDAL CLOSURE
MAXIMUM LEVEL
WAVE ENERGY
DISSIPA10R
MINIMUM LEVEL
5
10
ISm.
t:=:::=1:::::;=!!==l
Scale
Hydraulic
gates and
valves
:
.' \ LOWERED
. '. :;'
POSITIO~y
: . :
',.:
'
.. :':',
.........
. .
..~
'.
'
..
.'
(al
(bl
DRUM GATE
SECTOR GATE
"
"e-
",.
Bear-trap gates
A bear-trap gate consists of two leaves, one hinged upstream, the other
downstream (see Fig. 2.56). Both leaves are sealed at their side and pivots and
are free to slide or roll relative to one another with a sliding seal at their juncture.
When the gate is lowered, the leaves come to rest in the horizontal position
with the upstream leaf on top of the downstream one. When the upstream water
level is admitted to chamber 'a' the gate can be raised.
The water pressure under the gate is controlled either by an adjustable weir or
by setting the inlet and outlet sluice valves in a control chamber in the sluiceway
abutment.
Bear-trap weirs have been used in the USA for log-sluicing operations, when
the skin plates are usually protected by hardwood skid timbers.
The accumulation of silt under a bear-trap weir set on the river bed has been a
source of trouble and various methods have been developed for the removal of
silt by sluicing.
The control system and the seals of a bear-trap gate are critical. There is a
recorded instance ofbreakdowo of a bear-trap weir due to vibration caused by
bad design of the hinge seal. This problem applies equally to bottom-hinged
flap gates.
LOWERED POSITION
...
...
':.'
,:'
',',
-.:
:'. :
.~
.
,.
"
Figttre2.56. Bear-trapgate
11
':
',.
:.
'
..
-,
,
.... ..
. ..
. .~
..
',
The calculations for each equilibrium condition of the gate have to be carried
out separately, taking into account the external water load, the differential head
required to raise tbe gate and the water level in the chamber to maintain the gate
in position. The gate can be arranged to operate automatically to maintain
upstream water level, although close control with variable tailwater levels is
difficult to achieve. Raising the gate by admission of water to chamber 'a'
requires effective side and sill seals. Bear-trap gates are now seldom constructed
and when used they are often raised by mechanical lifting gear travelling across
the weir.
Types of gates
HYDRAULIC
I---++---SERVOMOTOR
OUTLINE OF BULKHEAD
GATE FOR MAINTENANCE
OF OPERATING GATE
Slide gates are also used as intake gates. If the gate is only operated under
balanced conditions, the slide material has to be able to withstand the
hydrostatic forces with no head on the downstream side of the gate. The
frictional properties of the slide material in this application are of secondary
importance. Impregnated woven asbestos is sometimes used. If the slide gate
controls the intake flow, the slide-beating material is similar to that of the
control gates.
Intakes can be controlled by radial gates, as shown in Fig. 2.60. The
advantage of using radial gates is the absence of gate slots (which can cause
hydraulic problems in high velocity flow), guide rollers which have to operate
totally immersed, or slides. The gate is rigid with no slack in its movement and
operating forces are less than those required for vertical-Lift gates.
The disadvantages are the requirement for a chamber to retract the gate, the
fact that the gate cannot be withdrawn to the surface for maintenance and, in
some cases, the immersion of the operating cylinder or cylinders while the gateis
in the open position.
Cylinder gates (see Fig. 2.61), are used where the controlling gate must
operate in a shaft or intake tower. They are used as shut-off gates and for
regulating the intake. The gates are guided by rollers operating on tracks fixed
to the tower walls, and therefore have little mechanical friction to overcome
Types of gates
GATE MAINTENANCE
CHAMBER
The discharge from a slide gate is smooth and the only limitation is discharge
at very small openings when the flow does not spring clear of the gate lip and is
liable to produce cavitation damage at the bottom of the gate.
The gates are designed with tbe skin plate downstream and with open
stiffener girders on the upstream side. This causes flow circulation between
the girders which is not detrimental. An alternative design is box construction
with the space between the girder flanges filled in. Some gates, such as the
bottom outlet gate at the Victoria Dam in Sri Lanka, have been manufactured
from solid forged steel plate.
The bonnet is designed to withstand the full hydrostatic head without any
structural contribution from the embedding concrete.
The transverse deflection of the gate must be very low so that the slope at the
bearing faces does not cause uneven contact pressure at the slide faces. Keeping
the design contact pressure below the permissible bearing pressure of the slide
material will permit some variation in the imposed pressure.
Unless the deflection of the gate is very low, it may be desirable to crossradius the bearing to allow for gate deflection, in which case the contact pressure
(Hertz) calculations have to be carried out accordingly.
The slides can be of conventional bearing materials such as leaded bronze,
aluminium bronze or manganese bronze. Used on their own they require high
pressure grease lubrication.This has to be applied to the slide contact face within
the gate slot by pipes leading from grease nipples at the top of the bonnet to a
Types of gates
ACCESS SHAFT
FOR
GATE ERECTION AND
MAINTENANCE
PROVIDING
AIR SUPPLY
.... s
number of selected points on the slide face. Grease distribution grooves must be
incorporated in the slide face to distribute the grease. When applying grease,
effective distribution wiU occur only at grease outlets masked by the gate. There
is a danger that grease or lubricant on exposed slide faces can be washed away by
the recirculatory flow within the gate slots when the gate is in a partially or
wholly open position. Possible environmental contamination may have to be
considered.
The use of bearings with lubricant inserts eliminates the sliding seat greasing
pipes and ensures an even lubrication coverage. In this type of slide, tbe
lubricant is compressed into trepanned recesses in the bearing. The lubricant is
of a permanent, solid, thick film nature and is a compounded mixture of metals,
metallic oxides, minerals and other lubricating materials combined with a
lubricating binder. Graphites containing lubricants should not be used in
conjunction with stainless steel as they cause electrolytic action, which is
accelerated underwater.
GATE
HOIST
COUNTER BALANCE
III-t---+--GATE
IN THE
CLOSED POSIflON
GATE
SILL
When a gate is opened and discharges into an empty tunnel, an air demand is
created due to air entrainment in the air/water transition region. The calculation
of air demand is dealt with in Chapter 9.
Difficulties can be experienced at gate slots at high velocity flow. These are
discussed in greater detail in Chapter 9. Hydraulic problems in gate slots have
led to the development of jet-flow gates. The gates incorporate contraction
slopes on the conduit upstream from the gate slots to cause the flow to jump
the slots in order to avoid intermittent flow attachment (Fig. 2.64).
The jet-flow gate shown in Fig. 2.64 is of the United States Bureau of
Reclamation (USER) circular orifice type. Other types of jet-flow gate with
rectangular outlets have been developed. A rectangular gate seals flush at the
sill and the contraction section is omitted at the sill. The rectangular jet-flow
gate is appreciably cheaper to manufacture. Tbe cost advantage is to some
extent offset by the requirement for a transition section of conduit from
circular to rectangular and on the downstream side from rectangular to
circular.
Radial gates can be used as control gates in conduits, arranged as shown in
Fig. 2.65, or the gate may be located in a shaft (Fig. 10.16).
Types of gates
Figl/re2.63.
S'lidegate
JET
SPRING
POINT
AIR SUPPLY
.'
.;.:.....
.,;
-_
CONTRACTION
SECTION
As discussed earlier in this chapter, the main advantage of using a radial gate
is the absence of gate slots. The disadvantages are a substantially larger gate
chamber or shaft, and in many cases difficulty of access for initial assembly and
su bsequcnt maintenance.
A ring-follower gate (Fig. 2.66), is selected as a guard gate where the terminal
discharge is controlled by a valve. It removes the need for the upstream transition
section joining a circular conduit to a rectangular one, and a similar downstream
transition section to a circular cross-section. In the fully open position it provides
an unobstructed fluidway. The gate can therefore reduce hydraulic losses to outlet
works and result in economies in the transition sections.
The gate leaf retracts into the uppermost body, the bonnet section, when the
circular opening aligns with the fluidway to present an unobstructed flow
passage. To close the gate the circular opening is lowered into the bottom
section of the bonnet and the bulkhead port jon of the leaf blocks the f1uidway.
The lower bonnet has to be drained and designed for flushing of accumulated
sediment.
1/
"-.
TRANSITION
TO
SECTION
CIRCULAR
RECTANGULAR
TRANSITION
SECTION
Types of gates
SERVO-MD
TOR -------------J
-1BULK
HEAD
FOR SHUT-OFF
SLIDE
SECTION
-----1
---------1
SLIDE SECTION
IN POSITION
FOR GATE OPEN --------
SLIDE SECTION
IN POSITION
FOR GATE
SHUT --------
DRAINAGE
OF LOWER
34
The term stoplog derives from the time when it was general practice to isolate
a sluice installation with wooden beams. Stoplogs cannot be placed in flowing
water because they are liable to vibration during lowering and raising when
combined over- and underflow conditions occur. Stoplogs can be designed to
be guided by rollers, or more frequently by slides. Stoplogs incorporating guide
rollers are easier to place.
Maintenance gates and stop logs can be positioned by a rail-mounted gantry
crane or by a mobile crane. Emergency closure gates should only be placed by a
rail-mounted gantry crane. Hydraulic downpull forces (see Chapter 9) could
topple a mobile crane.
Figure 2.67 shows the essential features of a stoplog comprising lower seal
'a', upper seal contact plate 'b' and side seals 'c'. If the side seals are designed to
seal both upstream and downstream
it enables the gates to be tested
hydrostatically on commissioning before the reservoir is fully impounded. If
this is considered desirable the stoplogs must also be designed for reversal of
the hydrostatic thrust. Bypass valves 'd' are linked with the grappling beam
anchor points so that initial Eft movement of the grappling beam opens the
valves to equalise the water level upstream and downstream of the stoplog.
Landing sensing device 'e' on the stoplog is positioned on the sill beam or on
top of another section. The rod is displaced upwards and permits disengagement of the grappling beam. If the stoplog jams or meets an obstruction, release
of the grappling beam cannot be actuated or accidentally effected .
.A
ENLARGED
DETAIL
BYPASS VALVE
INOT SHOWN ON SEmON
OF
AA)-Bif}--
.:
----
ENLARGED
SECTION
END OF STOP LOG
.~
UPSTREAM
----
'b'
II
LANDING ----SENSING
PROBE
SECTION
A-A
SECTION
B-B
OF
Types of gates
IV
1//
~'"
113
I.
!/1~7
-e--
i- I-
f--l
'.'
.I
I;
-}-
'--..-
,_,...J
i~
-0:1.+ ~
~:
'\
UJ
- --!
>
_J
'\,
Vl
Vl
<(
a.
>a)
V-'
1---1/
It//
V
=
~
r-
~
~
v
1--
~\
1+-
>
:[1-
1---
<{
o ~;
K .._~;.i
8~
[/7
UJ
m
0
c::
o,
r=J
l!J
-,
Vl
Z
-,
UJ
til
'",-
l!J
CJ
Z
l/'/
-c
"<. -,
7
~ ~
OV
" 'L_
~_J
t=
r-
~f\..
1._r
!/v
.r+i
""LJr
_)/
_;;i--
--
Grappling beams
Figure 2.68 shows a typical grappling beam. It is designed to engage
automatically with a stoplog, and will automatically disengage once the
beam is in position and the landing sensing device is activated. By moving
lever 'a", the operator determines whether the grappling beam is in the
engage position (for recovering a stoplog) or the disengage position (for
placing a stoplog). Guiding in transverse and longitudinal directions is
effected by rollers.
Cranes
Emergency closure gates must be handled by a rail-mounted gantry crane,
whereas maintenance gate and stoplog handling is either carried out by a railmounted gantry crane or a mobile crane. Gantry cranes can be used to transport
maintenance gates and stoplogs from their storage area to the sluiceway whereas
a mobile crane cannot, as a rule, transport heavy gates or stoplogs. If this is the
case two small rail bogies can be provided.
Different functions are often carried out by the same crane. The crane in Fig.
2.57 can service the operating gate and place the maintenance gate. In other
installations it is used for servicing the operating gate and as a means of placing
stoplogs or an emergency closure gate. \X/hen the crane has a cantilever runway
with an auxiliary hoist, it can also raise a removable screen of the type shown in
Fig. 4.2. At freestanding intakes in a reservoir accessed by a bridge, the gantry
crane servicing the operating gate can also place the emergency closure gate, and
in some instances may incorporate screen raking machinery.
Creep speeds on all motions of a gantry crane should be provided to enable
accurate positioning. Creep speeds should be about 1/10 to 1/15 of normal
motion speeds.
Types of gates
Main application
1.
Radial gates
mororiscd
Sluice
installations
River control
Spillways
Barrages
2.
Radial
automatic
gates
Sluice
installations
River control
Radial gates
Spillways
3.
No outside source of
power required
Absence of machinery
float
operated
4.
5.
Advantages
Automatic
crest gates
Limited
application
spillways
Automatic
scour Gates
Limited
application at low
level outlets of
low height dams
at
No outside source of
power required
Absence of machinery
Low maintenance
No outside source of
power required
Absence of machinery
Low maintenance
Disadvantages
Wide piers to
accommodate
displacers
Counterbalance
visually intrusive
Can malfunction due
to incorrect design
Can malfunction due
to blockage of inlet or
control system
Wide piers to
accommodate
counterbalance and
floats
Can malfunction due
to rope system
Can malfunction due
to blockage of inlet or
control system
Requires special type
of spillway
Coarse level control
compared with 2.
Limited height of gate
compared with 1. and
2.
Limited height of
openll1g
Limited upstream
head
Free discharge onl y
Coarse control
Type
Main application
Advantages
Disadvantages
G.
Vertical-lift
gates
Sluice
installations
River control
Old installations:
0 barrages
o spillways
7.
Vertical-Uft
gates
hook type
River flow
control
Barrages where
ebb tide exposes
gates
8.
Free rolling
gates
Shut-off of
turbine intakes
9.
Rolling-weir
gates
River flow
control
Tidal barrages
Sluice
installations
River control
Tidal outlets
Very reliable
Capable of operating at
differential heads up to
30m
Can be manufactured
very wide
Complete separation of
saline and fresh water
Overflow to clear debris
No visually intrusive
overhead structu re
Can in some cases be
engineered to open
under gravity in an
emergency
No outside source of
power required
Automatic operation
Simple construction
Low maintenance
Type
12. Fuse gates
Main application
Advantages
Limited
application
spillways
Automatic operation
Operates as fixed weir
until it ti Its
Simple construction
Gates have to be
Easy addition to height
recovered after tilting
of a dam
and may require repair
Particularly applicable
. or replacement
in an unsophisticated
environment
Use may depend on
frequency of flood
events
Unobstructed
Requires piers
navigation passage
Permits some flow
Not visually intrusive
upstream
Can be raised for
Complex to fabricate
Complex machinery
maintenance and
inspection without
placing stoplogs
Unobstructed
Capable of
navigation passage
withstanding only
limited differential head
without piers
Gates move
No structure above
navigation passage bed independently under
wave action resulting in
level
Excellent from visual
leakage between gates
Requires detachable
considerations
hinges
Gates have to be
interchanged for
maintenance
High clearance and
High overhead
structure
large span can be
Gates are normally
achieved
counterbalanced
No underground
passages requi red
Large gates require
Gravity closure in an
lifting chains of the
type where each pin
emergency
Conventional hoist
can be lubricated
machinery
Can be maintained in
high level position
subject to safety
requirements
Lock required
No underground
Separate overhead
passages required
Gravity closure in an
structure
Gates arc normally
emergency
counterbalanced
Conventional hoist
Large gates require
machinery
lifting chains of the
Can be maintained
when elevated above
type where each pin
can be lubricated
waterway
to
Storm surge
barriers
14. Buoyant
gates
bottomhinged
Storm surge
barriers
Tidal barriers
15. Vertical-lift
gates for
navigation
channels
Storm surge
barriers
16. Vertical-lift
Storm surge
gates for nOI1- barriers
navigation
channels
Disadvantages
Types of gates
Main application
Advantages
Disadvantages
17. Mitregates
Storm surge
barriers
Unobstructed
navigation passage
Excellent from visual
considerations
18. Vertically
hinged sector
gates
Storm surge
barriers
Spillways
Unobstructed
navigation passage
Can be opened or dosed
against flow
Can be opened at
differential head
Can be constructed with
wider opening than
mitre gates
Excellent from visual
considerations
No outside source of
power required
Width of navigation
more limited than
some other gates
Cannot accept reverse
thrust
Opening has to be
effected when water
levels arc near! y equal
Heavy mitre thrust
Cannot close against
high flow
Requires wiele recess
in banks to
accommodate gates on
opening
20. Bear-trap
Water control in
logging rivers
Type
gates
No outside source of
power required
Clears debris
Provides unobstructed
flow
Complex gates
Requires extensive
civil engineering
works
Requires zero
downstream level
Control system critical
Can silt up
Not favoured
Seals cri tical
Control system critical
Can silt up
Rarely used
Control and
emergency
closure
Type
Main application
Advantages
2.
Vertical-lift
in take ga tes
rope
operated
Bulkhead gate
3.
Caterpillar or
coaster gates
Control and
emergency
closure
4.
Radial intake
gates
Control and
emergency
closure
Intake gate
5.
Cylinder
gates
Intake gate
o.
Slide gates
Control gates in
conduit
Back-up gate for
a control gate
Disadvantages
Types of gates
Cannot be used as a
control gate
Cannot be used as an
emergency closure
gate
.Requires balanced
head for operation
Guide slots required
Possible cavitation
problems
Requires bypass
system
Control gate for very
Wide gate slots
high heads
required
Caterpillar train
operates under water
Requires stem
connections between
servo-motor and gate
Cavitation problems
Slow operation to
raise to maintenance
position
Requites air admission
Very costly
Absence of gate slots
Requires chamber to
Requires no load rollers retract
or slides
High concentrated
load
Lintel seal critical
Requires dewatering
of tunnel to carry out
maintenance
Requires air admission
Capable of controlling
Low natural
intake flow and large
frequency of vibration
openings
clue to rope
suspension and low
friction
Possible vibration
problems
Large gates require
counterbalance to
reduce hoisting forces
Reliable control gate or Gate slots required
emergency closure gate Possible cavitation
Inherently damped due problems
to sliding friction
Requires bonnet for
withdrawal
Requires air admission
Type
Main application
Advantages
Disadvantages
J et-Ilow
Control gates in
conduit for high
head application
gates
8.
Radial gates
Control gates in
conduit
9.
Ringfollower
gates
Requires chamber to
retract
High concentrated
load
Lintel seal critical
Requires dewatering
of tunnel to carry out
maintenance
Requi res air admission
Large overall height
approximately three
times that of fluid way
Requires regular
flushing
Drain connection
must be provided
References
1. Rouse, H (1964): Engineering hydraulics, Proc. 4th Hydr. Conference, Iowa Institute
of Hydraulic Research, Jun 1949, John Wiley and SOIlS, inc.
2. Murphy, T E (1963): Model and prototype observation of gate oscillations, lOt!;
l.A .[-1. H.. COli gress, London, paper 3.1.
3. Lewin, J (1983): Vibration of Hydraulic Gates, [ourn.L. W.E.S., 37, 165-179.
4. Thorne, R B (1957): The design, fabrication and erection of radial automatic sluice
gates, Proc. T.C.E., 6th Feb, 126-133.
5. Lewin, J (1984): Radial automatic gates, Proc. 1sf lilt. Conference Channels and Channel
Control Strtlctllres, Southampton,
paper 1-195, editor Smith, K V I-l, Springer
Verlag.
6. Quintela, A C; Pinheiro, A N; Afonso, J R; Cordeiro, M S (2000): Gated spillways
and free flow spillways with long crests, Portuguese dams experience, 20th fCOLD
COIlj,1f"eSS, Beijing, Q.79-R.12,
Vol. TV, '171-189.
7. Townshend, PO (2000): Towards total acceptance of fully automated gates, Dams
2000, Proc. ojlhe Biennial Cotfereuce of the DDS, Bath, Jun., editor Tcdd, P, Thomas
Telford,81-94.
8. Randerson, R J (1979): A velociry control structure in the River Orwell, Ipswich,
[onm. [.W.E.S., 38, 135.
9. Petrikat, K (1958): Vibration tests on weirs, bottom outlet gates, lock gates, 1I7"tlter
Power, Feb., Mar., Apr. and May.
10. Naudaschcr, E (1965): Discussion on Nappe oscillation, Proc. A .S. C. B., [ourn.
Hydr. o; May.
11. Schwartz, HI (1964): Nappe oscillation, Proc. A .s.c.E., j01l17l. T-fJ,dr. Diu. HY6,
Nov., paper 4138.
12. Partenscky, H W; Swain, 1\ (1971): Theoretical study of flap gate oscillation, 14th
1.A .)-/. R. COligress, Paris, paper B26.
13. Krurnmet,
R (1965):
Swingungsverhalten
von
Verschlussorganen
irn
Stahlwasscrbau, For.rchllflginIJflgcfliellr/llesen, Bd. 31, No.5.
14. Falvey, H T (1979): Bureau of Reclamation experience with flow induced
vibrations, 19th l.A .H.R. Congress, Karlsruhe, paper C2.
15. Ogihara, K; Ueda, S (1979): Flap gate oscillation, 19th I.A.HR. Congress,
Karlsruhe, paper
16. Pulpitcl, L (1979): Some experiences with curing flap gate vibration, 19thl.A.H.R.
Congress, Karlsruhe, paper C12.
17. Nielson, F M; Pickert, E B (1979): Corps of Engineers experience with flow
induced vibrations, 19/h I.A. T-f. R. Congress, Karlsruhe, paper C3.
18. 1\yres, D (1983): The Thames Barrier: the background and basic engineering
requirements, in I. Mech. E. Seminar Proc. The Thames Barrier, 8th Jun.
19. Clark, P J; Tappin, R G (1977): Final design of Thames Barrier gate structures, in
Proc. l.e. E. Conference, Thames Barrier Design, 5th Oct, paper 7.
20. Fairweather, D M S; Kirton, R R H (1977): Operating machinery, in Proc. I.e. E.
Conference, Thames Barrier Design, 5th Oct, paper 8.
21. Hamilton, A J; Prosser, M J (1988): Vel/ice LagoonJlood protectiofl,0'dralflic
model oj
scouring D's/em, B.I-I.R.1\. report RR 2918.
22. Lewin, J; Scotti, A (1990): The flood prevention scheme of Venice: experimental
module, [onru.Lnst. (Pater E1Ivironmentai NJallagement, 4, 1, Feb.
23. Ieperen, A van (1994): Design of the New Waterway Storm Surge Barrier in The
Netherlands, J-i_ydropOlJJer
and Dams, May, pp.66~ 72.
24. Janssen, J P F M; Jorissen, R E; Iepercn, A van; Kouvenhoven, B J; Nederend, J
M; Pruijsscrs, A F; Ridder, H A J (1994): The Design and Construction of rhe New
Waterway Storm Surge Barrier in The Netherlands,
18th lCOLD
COligress,
Durham, C.15 pp.877~900.
25. Severn Barrage Committee (1981): Tidal pon/erfrom the Seuem Estllary, Vol I,Energy,
paper No. 46, HMSO.
26. Ball, J W (1959): Cavitation and vibration studies for a cylinder gate designed for
high heads, 8th l.A .H. R. Congress, Montreal, paper 9A.
27. Bixio, V; Cola, R; Garbin, C; Mariani, M (1985): On the hydraulic behaviour of a
cylinder gate in a vertical intake with radial symmetry openings, 2l1dTlIt. Couferenceon
the H.;'drolllics oj Flood and Flood Contral, Cambridge, paper D3.
en.
Types of gates
Valves
The flow in pipelines is controlled by valves. Tills chapter is mainly concerned with
large valves carrying out the function of terminal discharge and providing back-up
in circular conduits. The exceptions are pressure-reducing valves in pipelines (an
example of these is shown in Fig. 3.16), and the needle valve which can be used for
regulating flow in pipelines oras a terminal discharge valve. It is largely superseded
as a discharge valve. The main applications, advantages and disadvantages of the
various types of valve are summarised in a table at the end of this chapter.
Amongst terminal discharge valves, the hollow-cone valve predominates
because of its good energy dissipation
characteristics,
simplicity of
construction, lower cost and favourable coefficient of discharge. It is also least
prone to blockage.
Compared with the external control sleeve of the hollow-cone valve, hollowjet valves and needle valves with their internal moving parts are mechanically
more complex and therefore more difficult to service. The fluid passages are
more restricted and liable to blockage.
Trashracks and screens arc essential at intakes which serve conduits
containing valves, and may only be omitted when large hollow-cone valves
are used as these valves are less prone to blockage.
Sluice valves
Sluice valves are the most frequently used control devices in pipelines. In the
open position they provide an unobstructed fluid passage. In spite of their nonlinear flow control characteristics they are often used for control of low velocity
low. Because the valve blade is unsupported during raising and lowering, and
due to eddy shedding from the blade tip, they are only suitable for closure and
opening against flow at low velocities.
Butterfly valves
The butterfly valve is the most frequently used closure device in pressure
conduits because of its relativel y compact arrangement and simple construction.
In the open position the blade lies in the plane of flow of the fluid. Valves are
manufactured in sizes up to 4 m diameter and are able to withstand operating
heads up to 200 m.
There are two types of butterfly valve: the solid-disc valve, sometimes
referred to as lenticular (Fig. 3.1(a)), and the lattice-blade valve, also described
as a th rough-flow valve (Fig. 3.1 (b)). The latter offers the advantages of a stiffer
disc assembly and lower loss coefficient.'
Valve blades are normally manufactured in cast iron or carbon steel. Other
more corrosion resistant materials such as high nickel cast iron, stainless steel or
aluminium bronze are used as the material for the blade where the water carries
bed material or is aggressive to cast iron or steel. Valve bodies and blades ha ve
been coated for special applications with epoxy resin or ebonite.
Figure 3.2 shows different arrangements of seals for butterfly valves. Seal (a),
an all metal seal, is suitable only for low operating heads. At higher pressures an
elastomeric seal (b) of Neoprene or Nitrile is used, and the seal is pressll rised by
the upstream water head. A diaphragm seal is shown in (c). This seal is also
pressurised by the upstream water.
The seal seat for (a) and (b) is of stainless steel and is welded into the
housing. It is arranged flush, but for clarity is shown in the figure as
projecting.
(a)
(b)
Valves
In the closed position the disc of the versions shown in (a) and (b) is inclined
at approximately 80 to the conduit axis. This applies also to lattice-blade valves.
The latter can be provided with a second seal on the upstream side which can be
normally or hydraulically operated, allowing replacement of the downstream
seal while the conduit remains under pressure.
Butterfly valves are normally opened under balanced conditions and closed
against flow. They are not generally suitable for flow control, only as on/off
devices, because of flutter of the blades and eddy shedding from the blade tips.
Poor to opening of the valve, the pressure is balanced by means of a bypass pipe
which incorporates shut-off valves. At high heads, manual, electrical or oil
pressure actuated filling nozzles are used. A guard valve is arranged upstream
of the filling nozzle. Means of venting the conduit downstream of the valve
during the fiUing operation must be provided together with facilities for
draining.
Closure of valves controlling conduits or penstocks is usually by gravity (see
Fig. 3.3(a. DULing normal operation the lever arm for the falling weight is locked
in the valve open position. The locking mechanism is designed to release the arm
weight
by falling
(b) Closure
by double
acting
servo-motor
which is then able to rotate to the shut position. The release mechanism may be
triggered manually, electrically or indirectly by excessive velocity of flow in the
conduit or by loss of pressure. The valve is opened by the oil hydraulic servomotor which also controls the rate of closure. Shortly before the end of the closure
movement the discharge of oil from the annulus side of the cylinder is throttled to
ensure slow final closure of the valve. The closing time must be controlled so as to
minimise the water hammer in the penstock. This can also be achieved by
cushioning the last part of the movement of the piston. Ellis and Mualli have
analysed the closure characteristics of butterfly valves.
The servo-motor in Fig. 3.3(b) is double acting. Opening of the valve is
effected by oil or an emulsion supplied to the piston side of the cylinder from
Valves
a hydraulic pO\,ver pack, while the fluid on the annulus side is ported to
discharge. For closure of the valve, mains pressure is admitted from upstream
of the butterfly valve to the annulus side of the cylinder, while the oil is ported to
the tank of the power pack. The water used as a closing medium is filtered before
it reaches the control elements and the servo-motor.
In the event of emergency closure, cavitation will occur where the conditions
of initial operation with positive back pressure change, into a situation of
complete separation of flow. The time is usually short enough not to cause
any damage.
The loss coefficient for fully open butterfly valves is shown in Fig. 3.4. The
loss coefficient of a valve K; is defined as:
6.H
K ---
=r
v -
where 6.H
V
g
The loss coefficients for partially open valves are shown in Fig. 3.5.
Variations of more than 10% occur, particularly when the valve is nearly closed
and when the valve seating arrangement becomes very irnportant.f
Cavitation in valves
Cavitation is caused by the local pressure on the downstream side of a valve, by
the accelerated flow of the water as it contracts to pass through the valve
opening, and by the generation of turbulence. Eddies are formed in the intense
shear layer which surrounds the accelerated flow of 'water through the valve
2.0
>
./
1.0
""._ 0.8
z
UJ
LJ
u,
u,
/.
.....
0.4
0
LJ
0.2 ...
_J
0.1
r-- -
0.10
BEST
DESIGN
VALVES
./'"
Vl
Vl
TYPICAL
OF
COMMERCIAL
/'
0.6
/"
0.15
THICKNESS
0.25
0.20
RATIO
0,30
tid
.forfttily
1000
500
200
100
SO
I-
z
w
20
10 1=
~'\
;>
~
LL
LL
\.
"-
'\.
"\..
'\.
50
/
\.
'\ ,,\ /
<-,
t-
\..
_J
\.
Vl
Vl
0
20
\.
-, -,
!\.
10
"'"" ""
I'\.
05
"\.
02
0'
Figtlre3.5.
Loss coefficients
for partiallY open {mlle1flY
J
valves (after Miller)
"\.
\.
.~
<,
10 20
30 40
VALVE
SO 60
OPENING
70 80 90
IN DEGREES
opening. When the pressure inside the eddies, which is considerably less than
the fluid pressure in the penstock, approaches the vapour pressure, cavitation
bubbles gww from nuclei suspended in the water. As the ambient pressure
increases and the eddies degenerate due to viscous forces, the bubbles become
unstable and collapse. If this occurs next to a solid boundary, it creates noise,
vibration and in its intense form erosion damage. The cavitation index can be
used to express the level of cavitation as the ratio of forces suppressing or
preventing cavitation to the forces causing cavitation:
bd
bu
cavitation index a = -:--____:--:
btl - bd
+
b - bu
or--lm - bel
or
where bd
bit -
Valves
fjIJ
,}/2g
hIJ
btl
V
where
incipient cavitation
critical cavitation
a correction factor
velocity Vir from Fig. 3.6 for incipient cavitation or the
reference velocity Vcr from Fig. 3.6 for critical cav itation
T/ir
Vcr
C1
r-,
0
5
10
15
20
25
30
N/m
610
870
1230
1700
2330
3160
4230
Head of water: mm
62
89
125
174
238
323
433
1000 r---.,-----,---,---,--,----r--
....-y--y--,.--.----,
r
z
w
U
IL
lL
W
C)
Vl
Vl
C)
_J
(after JltJilie/)
10
11
12
1000
Valves
500
f----
-- -
--
200
100
50
>
:.:::
-J- --I
20
._
Z
OJ
10
w
u;
u.,
I--
I- -
;/
50
'0
w
0
w
Vl
Vl
~'_E
---E
'0
0 0
'0
E-~
" '"
~ " "
<0
'0
ci
1\
-0
'""
0
-0
20
0
..J
10
05
02
J 1
01
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
K
CdAv!(2gH)
where Q
Cd
A
g
H
CORRECTION
FAC TOR C,
IS FOUND FROM
FIG. 3.5
(3.1 )
discharge
discharge coefficient (approximately 0.85)
area of the valve based on the inside diameter of the valve
body
gravitational constant
net head at the valve entry
Vane failure
This has been attributed to a number of causes but the most likely one is
hydroelastic instability causing vibration normal to the chord of the valve and
twisting about the longitudinal axis. Destructive resonance occurs at a critical
velocity at which the flow couples the two forms of vibration in such a way as to
HYDRAULIC
OPERATING
CY"LINOERS
SLEEVE
VANES
SlPPORTING
CONE
SLIDING
~~
~~
SLEEVE
4-~~=====d~~SEAL
.I'--H-- ..
DISPERSING
CONE
VALVE
BODY
ALTERNATIVE
OPERATION
MECHANICAL
SLEEVE
BY ELECTRODRIVE
feed energy into the elastic system. Possible modes of vibration for a hollowcone valve are shown in Fig. 3.9.
6
Mercer has suggested a parametric value incorporating
a coefficient
depending on the ratio of shell-to-vane th.ickness and number of vanes. Valves
with a value less than 0.115 have operated successfully and valves with a value
greater than 0.130 have failed. Mercer's parametric value is:
...
Figllre 3.9. Possible uibration
modes of bollow-cona ualues
(after Meree/"6)
-VALVE
MODE 1
MODE 2
BODY
Table 3.2
Valves
Ts/Tv
C
5
1.00
2.35
6
0.50
1.98
6
0.90
2.40
6
1.00
2.48
6
1.20
2.53
Q/(CDTv)
(3.2)
J(Eg/e)
where
Q
C
D
Tv
E
g
e
/(
6
2.00
2.75
discharge
a dimensionless coefficient depending on K and N
valve diameter
vane thickness
Young's modulus
gravitational constant
mass per unit volume (of the material of the val ve)
ratio of shell thickness to vane thickness = Ts/Tv
number of vanes
(3.3)
Inserting the value of C in equation (3.3) shows that six-vane valves have 10%
higher frequencies than comparable four-vane valves, and that the thickness
of the shell relative to the vane does not have too great a bearing on the
frequency.
Nielson and Pickett7 reported a major vane failure of a 2740 mm diameter
hollow-cone valve. The failure was of the fatigue type. Mercer's parametric
value of the valve as originally constructed was 0.176.
Falvey'' cited severe vibration of two 2135 mm hollow-cone valves. The
observed 85 Hz frequency correlated well with estimates of its natural vibration
frequency based on the paper by Mercer. The valve opening in the prototype
had to be restricted to a maximum of 80%.
Shifting
attachment
As the hollow-cone val ve is opened the flow control may shift from the sleeve to
the valve body (see Fig. 3.10) and intermittent attachment and reattachment
may occur, resulting in severe vibration.f Under these conditions the opening
of the valve, that is the sleeve travel, has to be limited.
There have been a few instances of vibration of hollow-cone valves due to
excessive length of the sliding sleeve.
OPEN POSITiON
Deterioration of the seal between the valve body and the sliding sleeve can
result in leakage. In general this does not lead to vibration of the valve but if it
continues for a long time it can result in erosion.
The expanding cone shaped discharge pattern of the hollow cone is very
effective in aerating the water and dispersing the energy. Because of these
features, stilling basins are not normally used for tbe discharge of hollowcone valves. The usual angle of the cone is 45. Experimental investigations
involving valves with cones of different angles9 have been carried out.
Where hollow-cone valves are located in a tunnel or where the spray
from a widely dispersed jet is not acceptable, a hood is used to confine
and redirect the discharge (Fig. 3.11). In tunnels or conduits the hood
prevents erosion by the impinging jets and ensures that air is admitted from
upstream. Guidelines as to the optimum geometry of hoods are given by
Brighouse and Chang.Ill,11 The hood has to be arranged to minimise
splashback through the upstream opening and ribs are introduced on the
inside of the discharge section of the hood so that air is admitted to the
inside of the discharge jet.
Hollow-cone valves can be installed to clischarge into a stilling basin or
submerged as shown in Fig. 3.12. In the former case the main purpose of
the valve, to act as an energy dissipater, is limited because the jet IS
_I
Valves
VALVE DISCHARGING
VERTICALLY
INTO A STILLING
BASIN
SUBMERGED
HOLLOW-CONE
VALVE
H ollow-] et valves
Figure 3.13 illustrates the construction of a hollow-jet valve. Movement of
the cone controls the area of the discharge orifice. It is used as a terminal
discharge and control valve. The jet is compact and therefore entrains less
air than the hollow-cone val ve. It can be installed directly after a bend in the
pipework.
The hollow-jet valve is frequently installed so that it discharges at an angle of
30" into a stilling basin. The flow in the conduit past the movable cone and the
body results in a hollow jet, which initially maintains its shape and flares out
shortly before the point of impingement. A jet angle relative to the horizontal
loosens up the jet structure and reduces the intensity of impingement.
Inspection and servicing of the mechanical or the oil hydraulic actuator
requires removal of the valve in its entirety. The oil hydraulic cylinder of the
valve, Fig. 3.13(b), is operated by an external hydraulic power pack with
pipework which has to be routed through the jet.
EXTERNAL
OfliVE
a hollow-jet
actuator.
valve
valve
Valves
:... ~. ..
'~.'
DRAIN
AND FLUSH
FOR GATE
':;
-'. -.
LINE
BONNET
Figllre3.14. lnstallation of
bolknu-je: valves
Needle valves
Needle valves are used for regulating flow, either as terminal discharge valves or
for controlling high head flow in pipes. Their use in outlet works has been
supplanted in many applications by more economical and hydraulically efficient
valves, such as thehoUow-conevalve. Figure3.15 showsaninstallation and crosssection through an interior-differential
needle valve. The valve is closed by
admitting water pressure to chamber B and connecting chamber A to drain
through the spool valve located at the bottom of the needle valve. To open the
valve, water pressure is admitted tochamberAand
chamber Bisopened to drain.
To prcvent cavitation the discharge opening is arranged so that the
downstream cone anglc of the needle is slightly less than the downstream cone
angle of the body. A sharp flow separation point at the body seat is another
requirement if cavitation is to be avoided. The coefficient of discharge for use
in equation (3.1) is about 0.6 at full valve opening. This reduces to 0.4 at half
opening and 0.26 at quarter opening. Because of the low coefficient of discharge
at partial openings, the valve can dissipate energy when controlling high head
flow in pipes. Needle valves are manufactured in sizes up to 2 m and for working
heads up to 200 m.
Pressure-reducing valves
Figure 3.16 illustrates a pressure-reducing
valve. The energy dissipation is
effected by discharging some or aJl of the flow through the orifices of the
Valves
PERFORA rED
CY LINDER
IN PRESSURE REDUCING
POSITION.
This type of valve is suitable only as a regulating valve in closed pipe systems.
It is also used as a bypass valve. They are manufactured in sizes up to 1.5 m
diameter.
Sphere valves
Figure 3.17 shows a sphere valve, sometimes referred to as a rotary valve. The
section shows the lower half of the valve fully open and the upper half fully
closed.
Sphere valves have a dear bore and when fully open have a very low loss
coefficient. Resilient rubber seals a1:eused for valves working at pressures up
to 400 m head. Metal seals are used for higher heads. Valves are normally
supplied with an operating seal at the downstream end and with an additional
maintenance seal on the upstream end which is either operated manually or
hydraulically. Closure is droptight. The application of sphere valves is for
shut-off control on the pressure side of high head turbines and pumps. The
...
Valve in
Shut position
Valve in
Open position
Figtlre3.17.
Sphere vtJlve
usual arrangement is gravity closure of the valve and oil hydraulic piston
operation for opening. Alternative operation by servo-motor has been used
with opening effected by hydraulic oil acting on the piston side of the cylinder
and uncontrolled
water operating on the annulus side. On opening, the
hyd raulic oil on the piston side overcomes the force on the water side displacing
the uncontrolled water from the cylinder. When closure is initiated, or there is a
failure of the oil supply, the valve is closed by the uncontrolled water pressure.
Sphere valves are manufactured in sizes up to 3.5 m diameter for use at
working presslI res up to 500 m. Smaller size valves up to 2 m diameter are
available up to working pressures of 1000 m head.
1
Valves
.: ... ~"
.',
.. ~
,
I
I
,,'
,.
BUTIERFLY
VALVE
hOI/OIIl-COIIC
valve
r, .I.~.
HOLLOW-CONE
VALVE
I
I I
downstream of the butterfl y val ve. If the section of pipe upstream of the val ve
remains under pressure, admission of water under reservoir head is required
only to the downstream side. Apart from means of draining the sections of pipe,
provision for releasing air must be made. The closure characteristic of the
butterfly valve must be designed to minimise water hammer.
The matching of terminal discharge and guard valves may also be appropriate
for other valve combinations-apart
from butterfly and hollow-cone valves. It is
likely to be more critical when a hollow-cone valve is used for discharge because
of the low loss characteristic of a fully open hollow-cone valve.
Main application
Advantages
Disadvantages
1.
Sluice valves
Controlling low
at low velocities
Closure and
opening of flow
Low cost
Simple
Reliable
Unsupported valve
blade during raising
and lowering
Eddy shedding from
blade tip
2.
Butterfly valves
Relatively low
loss coefficient
Available in large
sizes
Capable of
working at high
heads
Closure by
gravity can be
arranged
Normally opened
under balanced
conditions
Possibility of blade
flutter
Possibility of eddy
shedding from blade
Very efficient
energy
dissipation
Simple
construction
Relatively low
cost
Can be operated
electromechanically or
by oil hydraulics
Good discharge
coefficient
Available in large
sizes
Least flow
obstruction of
any terminal
discharge valve
3.
Hollow-cone
valves
(Howell-Bunger
valves)
Terminal
discharge
tips
Type
Main application
Advantages
Disadvantages
4.
Hollow-jet
valves
Terminal
discharge
Dissipates energy
Can be arranged
to discharge into
a stilling basin at
an angle
5.
Needle valves
Terminal
discharge
Dissipates energy
Can be used as an
in-line pressure
reducing valve
6.
Pressurereducing valves
(perforated
cylinder type)
Pressure control
in closed pipes
Pressure control
Orifices in perforated
cyLinder can be
blocked by debris
Internal moving parts
Inspection and
servicing regUires
removal of valve
7.
Sphere valves
(rotary valves)
Shut-off control
in high pressure
conduits
Low loss
coefficient
Shuts droptight
Manufactured in
large sizes
Capable of
working at high
heads
Can be supplied
with a
maintenance seal
Valves
References
1. Bramham H T (1979): Developments in through flow butterfly valves. l,vater Power
and Dam Construction, Mar.
2. Ellis J; Mualla, W (1984): Dynamic behaviour of safety butterfly valves. [[7afer
PO/ve1' and Dam Construction, Apr., pp. 26-81
3. Miller D S (1978): lnternaifiow systems, B.H.R.A., Pluid Engineering.
4. Rahmeyer W J (,1981): Cavitation limits for valves, [ourn. A .IV.1,v.A., Nov, pp.
582-584.
5. Instrument
Society of America: Control valve capacity test procedure, ISA-S39.2,
Pittsburgh, Pa.
6. Mercer A G (1970): Vane failures of hollow-cone valves, I.A .H.R. Sympo.ritlnl,
Stockholm, paper G4.
7. Nielson F M; Pickett E B (1979): Corps of Engineers experiences with flowinduced vibrations, 19th I.A .H.R. COligress, Karlsruhe, paper C3.
8. Falvey H T (1979): Bureau of Reclamation
experience with flow-induced
vibrations, 19th1.A .1-:1.
R. COligress, Karlsruhe, paper C2.
9. Rao P V; Patel G G (1985): Hydraulic characteristics of cone valves with different
angles, Irrigation and Power, J ul., pp. 233-243.
10. Brighouse B A; Chang E (1982): Dcsigl1dataolldeflectorhoodsforhoIIOlv-col1cout/etvalves,
B.B.R.A., report 1939, Dec.
11. Brighouse B A; Chang E (1982): MOl1aSavlIl!ydroeicctric scheme, Fiji, Part 2, !I1odelstlldj'
ofI-IOlveli-BlIlIgcr /lO/vcfor the cOlltrolletlfilling outlet, B. H.R.A., report RR1818, MaL
1
Trashracks, screens and
debris
all
-,lL
i..:
/.
SCREEN
PART
FRONT
ELEVATION
,
SIDE
ELEVATION
Trashracks, screens
and debris
Failure of screens due to vibration of the screen bars has been recorded. '-3
Vibration depends on the natural frequency of the screen bars, the forcing
frequency and the potential development of resonance." Vibration occurs when
the two frequencies approach resonance.
The natural frequency of oscillation of screen bars in water is given by:
a
in=21f
where
J(
EIg
(171
in
natural frequency
a coefficient depending on the end fixity of the bar (bars
are normally welded to the supporting grid frame; a is
between 16 and 20 for bars 60-70 mrn deep having a
thickness-to-depth ratio of 5:1)
Young's modulus
moment of inertia of screen bar
gravitational constant
mass of screen bar
added mass of water; that is the mass of water vibrating
with the bar
length of bar between supports
E
I
g
!ll
m\v
L
\
+ lllwL3
. l?
where
= equivalent
17 = effective
171
or
Inw
= thickness
bid
of the bars
17
= '8 x d
Work done by Levin2 suggests that the value of the 17 term be limited to 0.55
times bar depth for a bar with a depth to thickness ratio 10, and to 1.0 times bar
depth for a bar with depth to thickness ratio 5. In practice, the effective spacing
between bars will be greater than the bar depth but the computed value of b should
be based on the suggested relationship, which will yield conservative results.
The forcing frequency due to vortex shedding at the downstream edge of the
screen bar is given by:
if
where
= (SV)ld
if
5
V
d
forcing frequency
Strouhal number
approach velocity
thickness of screen bar
The Strouhal number depends on the spacing between bars and the shape of
the bars. Levin 1,2 gives detailed information. For most design purposes the limit
value of the Strouhal number applies when the bar spacing to bar thickness
number is 5 or greater. For a bar fully rounded upstream and downstream the
limit value of the Strouhal number is 0.265, and for a screen bar with sharp
rectangular profile the number is 0.155.
Selecting the velocity is complicated by the difference between the flow
across the bars when the screen has been cleaned and when it is partially blinded
by debris. It is suggested that a range of values be used: at the lower end the net
velocity through the bars and at the upper end a value three times greater.
Although screens arc not normally operated so that debris is permitted to
accumulate to this extent, the local velocity of a partially blinded screen may
significantly exceed the average velocity. It is recommended that the natural
frequency of oscillation of the screen bars should cliffer by a factor of2-2.5 times
from the forcing frequency.
In the diagram of the velocity profile at La Plate Taille,3 the variation of flow
velocities about the mean value appears to lie between 0.5 and 1.8.
the
To satisfy the first criterion and take into account the possibility of a screen
becoming completely blocked, a bypass should be provided. Accommodating a
bypass may, however, present physical problems in urban river courses.
Screen instrumentation
Instrumentation
for submerged
inlets
It is a usual requi rernent that the head loss across the screen be measured to
indicate when the screen bas to be cleaned. At inlet depths of 30 m or more the
most suitable instrument is a bubbler device, because it can measure differential
head by means of a sensitive bridge.
Instrumentation
Screen raki ng
'A variety of screen raking machinery for fixed screens is available. Figure 4.3
illustrates two types.
Screens which are hoisted to the surface such as that shown shown in Fig. 4.1
have to be raked by hand using special purpose combs. An auxiliary crane is
provided at the screen's cleaning platform to handle logs or trees which have
been trapped by the screen bars.
Debris
In underflow gates, debris will not normally be discharged until a gate is
70-80% open. Under conditions of drowned discharge, debris becomes
trapped in the hydraulic jump which forms .in the stilling basin, and may
recirculate for an appreciable time. Floating oil cans or other metal containers
which repeatedly impact the submerged sections of gate arms or structural
stiffeners of the skinplate can be a noise nuisance at gate installations close
to dwellings. Floating debris can also cause damage to gate equipment and
to paintwork on gates.
At overflow gates or overflow sections of gates, debris and flotsam will be
discharged from upstream but can build up at flow breakers where these have
been provided to vent the nappe, or become trapped by the discharge roller
which forms downstream of the gate (see Fig. 2.27).
In bottom-hinged flap gates which recess into the river bed, floating timber
trapped upstream of the gate discharge roller can cause operational problems.
Trashracks,screens
and debris
l
At flood diversion channels which are controlled by gates, and which
incorporate a fixed weir alongside, a boom may be fixed across the flood channel
to divert floating debris to the weir.
A floating boom is often placed at the exit of a stilling basin to prevent debris
discharged over the weir from refluxing and entering the stilling basin when the
flood relief gates are shut. It can then remain trapped in the stilling basin when
discharge under the gates commences.
Floating booms for diverting debris are positioned across a waterway at an
angle of 30-45 to assist in driving flotsam towards the bank for clearance.
Slack must be provided in the stringer cable to allow for a rise in water level
during a flood. Most designs of floating booms are only partially successful.
In rivers which carry a large amount of flotsam during the flood season, an
appreciable load can accumulate on the structural stiffeners of gates. To prevent
this the rear of the gate skin plate is sometimes protected by wire mesh, but some
debris will still penetrate the mesh ~U1dbecomes more difficult to remove.
Another solution is to design horizontal stiffening members of the skin plate
assembly and horizontal tie members of gate arms as closed sections, adding a
fairing section where, due to the angle of the upper face of the stiffening
member, a ledge is still formed for debris to collect.
References
1. Levin, L (1967): Problcmcs dc Perte de Charge ct de Stabilite des Grilles de Prise
d'Eau, La HtliJie Blanche, 22, No.3, pp. 271~278.
2. Levin, L (1967): Etude Hydraulique des Grilles de Prise d'Eau, Proc. 7th Gen.
Meeting l.A. H.R., Lisbon, 1, p. Cll.
3. Vanbelli.ngen, R; Lejeune, A; Marchal, J; Poels, M; Salhoul, M (1%2): Vibration of
Screen at La Plate Taillc Hydro Storage Power Station in Belgium, Int. Conference on
,Ploul Induced Vibrations in FI1Iid Engineerillg, Reading, England, Sept., p. B2.
4. Sell, L E (1971): Hydroelectric Power Plant Trashrack Design, Proc.A .S.CE.,)
Pmuer Div., Vol. 97, Jan, No. POL
5. Magenis, S E (1988): Trash Screens in Urban A reas, I.W.E.M., River Engineering
Conference, Jan.
Trashracks, screens
and debris
l
Structural considerations
Design criteria
The design of gates can be analysed by conventional two-dimensional (2D)
structural analysis. This can also be applied to radial gates if the curvature of
the skin plate is ignored. If it is considered necessary to take curvature into
account, a finite element three-dimensional (3D) analysis is required. Existing
gate design standards require that designers consider the limit states at which
gates would become unfit for their intended use, by applying appropriate
factors for the ultimate limit state and the serviceability limit state.
The extent and magnitude of load factors applied is usually varied for
different design elements, operating conditions and the possible occurrence
of extreme events which cannot be reliably quantified. Examples are ship or
vessel impact, increases in weight due to entrained water or floating debris,
exceptional tide levels, jammed foreign bodies, increased loads due to
impeded movement caused by solid freezing, ice impact and possibly
irregular settlement and deformation of the foundation works. Some
designers or specification writers stipulate that a corrosion allowance be
added after calculations determining the sizes of members and plates have
been carried out. A reduction in working stresses or an increase in the load
factors is a more rational approach.
Detailed structural design standards for hydraulic gates are current in
Germany and the USA. The German standards are DIN 19704, Hydraulic Steel
Structures, Part: 1, Criteria for design and calculation, 19981 and DIN
Handbook 179, Water Control Structures 1, 1998.2 (At the time of preparation
of this edition, the German standards were not available in English. A previous
version of DIN 19704: Sept 1976 is available in an English translation.) The
USA standard is the US Army Corps of Engineers manual, Design of spillway
taintcr gates, January 2000.3 The designation 'tainter gate' is used in all US
references to radial gates. The design basis in this specification, while similar
to limit state design, is load and resistance factor design (LRFD) which can be
expressed mathematically as:
where '"Yi
load factors
that account
for variability
In
the
corresponding loads
nominal load effects
reliability factors
resistance factor tbat reflects the uncertainty
in
the
resistance for the particular limit state and, in a relative
sense, the consequence of attaining the limit state
nominal resistance
The load factors used in the Corps of Engineers Manual correspond to the
load factors used in BS5950: Part 1: 1990 Structural use of Steelwork in
Buildings, when these are interpreted to apply to the load conditions
encountered in hydraulic gates. Table 5.1 lists the load factors.
Calculations of hydrostatic loads on a radial gate for the water pressure on a
closed gate are set out in the Appendix to this book. Approximate calculations
are also shown for a gate in the open position when the downstream water level
is above the gate lip and the discharge is drowned. The third set of calculations is
for the gate load due to water pressure under overflow conditions.
Table 5. 1. Load 'factors Jor the design oj gates [after US Corps oj Engimers' Mantlal, Design oj Gate.r,
[anuary 2000)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Loading
Factor-y,
or,f
1.2
1.6
1.6
1.4
1.4
1.2
1.2
1.2
1.2
1.2
1.6
1.0
1.4
1.0
1.0
1.2
l
When considering the combination of loads under service conditions, the
following are deemed not to act simultaneously (numbering corresponds with
loading factors given in Table 5.1):
D
Structural
considerations
The maximum net hydrostatic load (4.) coinciding with either wave load
(15.), ice impact load (11.), or earthquake load (14.) is considered negligible.
The operating condition of maximum downward load of a hydraulic hoist
system (8.), wave load (15.) and ice impact load (11.) will not occur at the
same time.
The likelihood of opening 0[ closing a gate at the same time as an earthquake
occurs is considered negligible.
In the event of failure of one hoist, the likelihood of it coinciding with either
the maximum net hydrostatic load (4.), wave load (15.), ice impact load (11.)
or an earthquake load (14.) is considered negligible.
If the gate jams, the simultaneous occurrence of the following loads is
considered unlikely: the maximum net hydrostatic Load (4.), wave load
(15.), ice impact load (11.) or an earthquake load (14.).
VERTICALLY
STIFFENED
HORIZONTAllY
STIFFENED
the gap between the gate arms and the piers. Different arrangements of bracing
the gate arms are shown in Fig. 5.3.
When the skin plate assembly is stiffened by horizontal beams, as shown
in Fig. 5.1, the load on the skin plate is transmitted
by the horizontal
members to the two main vertical members which tie the gate arms in the
vertical plane.
When the skin plate is stiffened by curved vertical ribs, the load on the skin
plate is transmitted to the horizontal beams tying the gate arms and forming a
portal with the gate arms. The cu rved vertical ribs are usually of constant section
throughout. Spacing of the vertical ribs is based on the plate stress of the lowest,
Structural
considerations
most heavily loaded section. As a consequence, the upper section of the skin
plate assembly becomes more lightly stressed.
In horizontal reinforcement, the spacing of the stiffener beams can be varied
to equalise the panel stresses, and the beam sections can be selected to suit the
loading imposed on the adjoining plates.
Structurally, the horizontally reinforced skin plate assembly is more
economical. In radial gates, the reinforcing members of the skin plate constitute
about one-third of the total weight of the gate. The material saving will amount
to 10-15% of the weight of the skin plate assembly.
'I
Where the span of a radial gate is wide in relation to the height, the
horizontally reinforced skin plate is a more efficient choice. In such gates, an
arrangement of four arms per side can save weight and result in a torsionally stiff
structure. The four arms form a tapering box. Figure 5.4 shows a counterbalanced radial automatic gate designed on this basis.
One disadvantage of horizontally reinforcing the gate skin plate assembly of
river control gates is the potential for debris to accumulate on the beams. This
,i
-_J
can also occur on the lower horizontal tie member of the gate arms of vertically
stiffened skin plates.
\X/hen a gate operates with the downstream side partially submerged the sheared
flow in the stilling basin results in strong recirculation which impinges on the downstream side of the gate. As debris, particularly floating timber, is carried under flood
conditions and discharged under the gate, the recirculatory motion in the stilling
basin will deposit iton the horizontal beams. Thiscan becomeadif6cultmaintenance
problem. AttheTorrumbarry
Weiron theRiverMurray in Australia, aconsiderable
quantity of timber accumulates during the flood season and is discharged under the
gates. To prevent timber build-up, the horizontal skin plate stiffener beams were of
closed sections, designed so that lodgement of timber was prevented. Figure 5.5
shows a section of one of the six radial gates fabticated in this way.
At very large spillway gates, the skin plate is usually stiffened both
horizontally and vertically.
The potential for a gate to jam on one side when hoisted, or for failure of one
rope or multiple ropes on oneside ofa gate, has resultedindesigns which brace the
area between the gate arms to form a truss. This is implemented at thedownstream
side of the horizontal beams tying the gatearms.lt is regarded as standard practice
in the US Corps of Engineers' Manual 1110-2-2702. Figure 5.6 shows the load
diagram when a gate jams on one side (load factors have been omitted).
Structural
considerations
of the members depends on the panel stresses of the skin plate at the lowest, most
heavily loaded, part of the skin plate. This is continued to the crest of the gate,
resulting in relatively close spacing of stiffening members.
In all-welded construction T -sections are selected as stiffeners, and in riveted
construction rolled beams have to be used to provide a flange for riveting. To
Structural
considerations
~HOIS"NG
ROPE
k-1
o R
1 GATE
WEIGHT
W
-0
G_ GAlE
~GUIDE
ROLLERS
'GATE
JAMMED
RADIAL
GATE
GATE
ARMS
DOWNSTREAM
VERTICAL
TRUSS
FOR LOADING CONDITIONS
DUE TO
A JAMMED
GATE
weld the leg of a T-section, a minimum depth of 200 mm of the stiffener is
required. When repainting becomes necessary, the underside of the flanges is
difficult to access, particularly for paint preparation, and the quality of the finish
and its durability tends to suffer.
Horizontal stiffening beams of radial and vertical-lift gates which are welded
to the skin plate are also designed on the basis that they act compositel y with the
skin plate.
Composite construction
A form of stress analysis which is specific to gates and is not normally
encountered in onshore structures is the combination of panel and bending
stresses in stiffened plates subject to hydraulic pressure. The stiffening sections
Figllre5.6.
Load diagram and'
braeitl}!,to withslatulj"!llIni1Jg of
a radi,;/ gate
(a)
Figure 5. 7. N/etbodr of
reinforcing vertical-lift gales
(b)
welded to the skin plate of a gate (Fig. 5.8), act compositely with the skin plate to
form the top flange of a beam.
The composite beam, in bending, transfers the hydrostatic load to other
beams or support members. The skin plate is also subject to a panel stress at right
angles to the bending stress of the beam. These stresses have to be combined.
where q
t
pressure), and
"
-,'
Structural
considerations
..0
L.
F igJ/rc 5.9. Rectcltlglliar plate
witb all edgesfixed, IfIJiform
load over entire plate
and the maximum deflection at the centre of the plate is given by:
where E
Young's modulus.
For the case of a rectangular plate with three edges fixed and one edge free,
Table 5.2. Variables ill stress and deflectioll equations oj plates (all edges[ixed)
alb
1.0
1.2
1.4
1.6
1.8
2.0
00
fJI
0.3078
0.1386
0.0138
0.3834
0.1794
0.0188
0.4356
0.2094
0.0226
0.4680
0.2286
0.0251
0.4872
0.2406
0.0267
0.4974
0.2472
0.0277
0.5000
0.2500
0.0284
fJ2
n
uniform load over entire plate (Fig. 5.10 and Table 5.3).
At x = 0, Z = 0, the maximum stress is:
where R = the reaction force normal to the plate surface exerted by the
boundary support on the edge of the plate in Njrnm, The units of q
are N/mm2 and b is in mm.
Atx=
0, Z= 17:
Atx=
a/2,z=b
J-r7"""T"7"":r-r-T--r-r~V .
I
I ..
---;--1
of platrs
(Ibrre edgesfixed)
alb
0.25
0.50
0.75
1.0
1.5
2.0
3.0
/31
/32
/33
/1
0.020
0.016
0.031
0.114
0.125
0.081
0.066
0.126
0.230
0.248
0.173
0.148
0.286
0.341
0.371
0.321
0.259
0.511
0.457
0.510
0.727
0.484
1.073
0.673
0.859
1.226
0.605
1.568
0.845
1.212
2.105
0.519
1.982
1.012
1.627
/2
~:If
1111,11111C1IIIil,~lq
la)
LOAD
DIAGRAM
110 ME N T
Ib)
BENDING
(c)
MOMENT
DIAGRAM
IN
VI
Lin
1
0.984
0.938
0.867
0.783
Q.697
0.616
0.545
0.484
0.433
0.391
0.325
0.276
0.241
0.195
00
20
10
6.67
5
4
3.33
2.86
2.5
2.22
2
1.67
1.43
1.25
1
V2
1
0.861
0.753
0.660
0.580
0.512
0.453
0.404
0.363
0.324
0.295
0.250
0.215
0.190
0.155
A method and data for analysing curved skin plates and stiffener beams for
radial gates has been given by Wickert and Schmausser.6 The preferred 3D
analysis of a radial gate is by a finite element program.
The biaxial stresses represented by the panel stress acting at right angles to the
bending stress of the beam are combined so as to calculate the equivalent stress:
cr, and cry are the normal stresses in orthogonal directions, that is the panel st.ress
and the beam bending stress. They are substituted with their signs. T is the shear
stress, which is calculated as:
T
where T
S
I
d
A
= T [A
shear force
static moment about the centroid of the section of part of the
cross-section between the point concerned and the extreme
fibres
moment of inertia
web plate thickness
cross-sectional area of web plate.
References
1. OJ N 19704 (1998): I-l]drau/ic S tee! StrlfclJlres- Criteria for De.;ign and Calcuiation.
2. DIN Handbook 179 (1998): IF/ater Control Strtlctllres 1.
Structural
considerations
US Army Corps of Engineers (2000): Design of Spill Jill!)' Tainter Gelles, EMlll0-227023, Jan. 1.
4. Timoshenko,
S P; Woinowsky-Kricger,
S (1970): Theory of Plates and Shells, 2nd
edition, McGraw-HiU.
5. Roark, R J; Young, W C (1975): FOTnJIIlas [or Stress and StrailJ, 5th Edition,
McGraw-Hill.
6. Wickert, G; Schmausser, G V (1971): Slabhll{/sserbatl, Springer Verlag.
3.
Operating machinery
Gates may be operated by either electromechanical drives raising the gates by
ropes or chains, or by oil hydraulic cylinders. Screw jacks have been used, in
some installations, these are the preferred means of operating penstocks.
Electromechanical drives consist of electric motors driving hoisting drums
or chain sprockets through multistage reduction gear boxes. The gates close
under their own weight with the motors controlling speed of descent. Large
speed reductions are required from motors to the rope drums or the chain
sprockets. They can be in the range 1500:1 to 2200:1.
In most gate installations, that is at spillways and river control weirs, the
critical emergency operation is opening. This also applies to tunnel gates
controlling bottom outlets. For inlet gates for turbine passages it is the
opposite. They have to close in the event of turbine rejection or runaway
conditions.
A few radial gates have been constructed to open without power. To effect
opening under gravity the gate has to be counterbalanced so that closure motion
requires the drive effort, that is, the mass of the skin plate is overbalanced. The
gates are shown in Figs 2.11,2.14 and 2.15. The ropes for closu re from the hoist
drive are anchored at the counterweight
or along tile arms sustaining the
kentledge. The gear reduction of tbe hoist motor is kept low, while high
efficiency spur or helical spur gears are used to ensure tbat the gear train does
not become self-sustaining, which would prevent gravity lowering.
The same principle has been applied where gates are closed by oil hydraulic
cylinders. Tbe fail-safe operation is gate opening under gravity and power
closure by hydraulic cylinders. The spillway gates of tile Victoria Dam in Sri
l
Lanka were arranged in this manner, described in detail in Chapter 11.
Oil hydraulics applied to a hoisting cylinder, usually referred to as a servomotor, actuate gate closure as well as opening. Oil hydraulics permit the direct
application of large forces moving slowly, eliminating electric motors, brakes,
large multistage reduction gear boxes and hoisting drums.
A servo-motor can be used at each side of a gate and the cylinders can be
linked by a pipe ensuring that the same forces are exerted by both cylinders.
The elimination of transmission shafting and overhead hoisting machinery is
an advantage at locations where visual considerations are important, and where
the appearance of electromechanical machinery above the abutmeots or piers is
not acceptable.
Electromechanical drives
Single motor drives with line shafting driving rnultireduction gearboxes at each
end are a common method of layout. Apart from simplicity of control, it is
possible to mount two motors to duplicate the drive in the event of one motor
failing, and to arrange for manual winding of the motor extension shaft on
mains failure. Figure 6.1 shows two common layouts of single motor drives.
Squirrel-cage induction motors are invariably used in gate installations
because of their simplicity and minimal maintenance requirement. The hoist
capacity of the motor should be as close to the required load as possible, with
onlya small margin for an unexpected load combination. In the majority of cases
of hoist failure, due for instance to a racked gate, the force that caused the failure
of the hoist and sometimes the gate was most likely an oversized motor. The
motor is the easiest part of a power train to replace. It is also protected by
thermal overload devices which can be reset after stalling.
Motors should be able to start against fuJI load, the so-called 'hard start',
because of the considerable inertia of a gate system which has to be accelerated
to full speed. A standard squirrel-cage motor will develop about 150% of its fullload torque on starting, whereas 200(Xl is characteristic of a high torque motor.
Star-delta starting is not suitable for hoist motors, even with Wauchop 'no
break' winding. With star-delta winding, the starting torque is only 54% of
the motor full-load torque, and in order to accelerate a hoisting load, the motor
has to be oversized. In the event of an obstruction such a motor can wreck the
power train and the gate.
Operating
SECONDARY
OR HELICAL
REDUCER
machinery
SPUR
GEAR
ORUM
Figure 6. 1. A rrangcJnentJ of
single motor drives for radial
gates
Hydraulic
gates and
valves
PRIMARY
RE-
DRIVE.
SEl SYN
9jft~@mu{~lnllg"
ENLARGED SECtiON
X - X
HECIlAtliCAL SCREW
OPERATED RADIAL GATE. MOlORS
MACIlINES
3 PIlASE: ~PPlY
150%
- I-
- --
W
::l '--
-r-t
a:
a
t:)
{~
/ (./.:
III
Z
o
-. ~
x
z
Z/
n:
u
III
I? v
L~~~S~~~E!J
7Q,
1'-.
I~(
V
~:
--1/:
1-
,.L
/'
&
~
C ~~~~
V.
1\'('
1-
'i:
~
~
F-
Ie
;/
r+--
f\3! ~
-_-
-~
-__ - :?
'--
1--.
- l-
I--
_.
I--
- --
--
BASIC SElSYN
CONNECTIONS
--
I-
>-
s..s ~
;
W
SECONDAI1Y
WINDINGS
SYNCHRONISING lORQUE/DISPLACEMENT
Ar~GLE (IlARACTEnISlfCS FOFI A SElSYN
HECEIVER RUNNING AT 1200 n:v/min
of a pOJller S clEJ'11motor
drive
FUll
TORIlUE
mains failure, since the primary windings of the power Selsyns have to be
energised from the mains for synchronisation to be effective.
Alternative linked drives can be achieved by synchronous motors fed from a
variable-frequency supply or thyristor-controlled
DC motors with servocontrols. The former system also offers the benefit of speed variation, if
required. The starting characteristics are not as good as those obtained with
DC motors or high torque squirrel-cage AC motors.
Thyristor-controlled
DC motors with servo-controls can be up to 55 kW per
motor, or even larger.
With any thyristor control, great care should be taken with signal cables,
particularly those transmitting signals from solid-state devices. Signal cables
should be screened and should not run alongside motor feeder or control cables.
In some circumstances a clean supply should be considered.
Two hoisting ropes per side are frequently used. They are connected at the
gate anchorage point by a compensating beam to allow for differential rope
stretch. Articulation of the compensating beam is restricted so that, in the event
of the failure of one rope, hoisting can be continued with tbe other (Fig. 6.3).
Similar layouts are used when chains are the means of gatesuspension.ln
this
case, the hoist drums are replaced by chain sprockets.
HOIST
Operating
machinery
DRUI1
(b)
HOIST ROPE
BEHIND SKIN
PLATE
WHEN
SIDE
/0
t .. WIR[
R:; GAff
ROPE lfU$IOtl
RAOIUS
lRumuou
1.RUtlNIOU
PRESSURE
au
'W.
SKill
Pl A f (
FROM
(ABLE
R
E.
mGE
AEA(1I0"
HORE THAN
21 ,in
Ir)
TANGENT
The hoist ropes can be anchored to the gate either upstream of the skin plate,
Fig. 6.3(a), or downstream, Fig. 6.3(b). Upstream anchorage with the ropes in
contact with a wear plate welded to the skin plate has the disadvantage that
debris can become lodged between the ropes and the skin plate and that the
ropes are immersed during the majority of their working life. In spite of these
drawbacks this arrangement is sometimes used because the layout of the hoist
machinery shown in Fig. 6.1 is difficult to achieve or is incompatible with the
location of the gear box drive shaft upstream of the gate. However, it should be
a voided where possible.
Figure 6.4 shows the load imposed on the skin plate due to anchorage
upstream of the skin plate. A typical arrangement of a hoist for an intake gate
is shown in Fig. 6.5.
EJ
Operating
machinery
,.......
;-,r-
'1'-'
I
I
'I
"
"
II
""
II
""
"
""
..rh.
, II II
1
I I
I
I I,
I
,
,,
,,
,
,
1
Wi
r-J
,,
,
I
1
1
1
I
I
motion. For preliminary design of radial gates it is often assumed that the
cylinder will be at a 45 angle to the horizontal when the gate is closed, although
optimum angles may vary from that. If environmental or aesthetic considerations dictate that cylinders do not project into the skyline, different cylinder
arrangements can be engineered, but they willusuaUy be less efficient. Pistons
are either of high tensile steel, chrome plated or, for very long life, ceramic
coated.
Figure 6.7 shows the hydraulic circuit for a power pack for operating two
radial gates, each having two cylinders. Features of the ci rcuit are:
o Two motor and pump units where one acts as standby to the other.
o Automatic start-up and changeover from one pump motor unit to the other
in the event of one unit failing to run up to pressure. This is effected by valves
\Xl1 and \Xl2'
o For each gate, two directional control valves which select the raise, lower and
hold positions. One valve acts as a standby to the other. The changeover is
carried out manually.
o Alternative manual operation of the directional control valves in the event of
failure of the mains supply, when the gate is operated by a hand pump or by a
mobile plug-in power pack.
Figure 6.8 shows a possible circuit for the hydraulic cylinders. Together
Operating
lu
lU>
>-'
-'<{
~.-.-.-.. -.-.-.-.-.-.-.-.-.-.-.
~>
i
i
1
Nrn-tlfl\QC'OO
1
1
e-i
i'"'-
~I
i
~i
i/C_J'1
~l
1
!""'~'. ~
!
!
1
.
,
.
~i H-+t--_,..---.
.:
11
_.J
~I
.'"
__
i
i; i
e!--"~-i.._j
L._ .. _. __
._._._._._._._._._._._._._._
. ...L-_'
__
'_'
_J
-"
machinery
i'r___
'_'_'_'_'_'_'_':
.
I
.
I
I
I
.
I
1
1
1
al
~.
r".;;:-,'"
I"
1
i
1
1
1
"<,.
-,
;;'. I
-=~=f~
T
.
system diagram.
Features
of the circuit
Pressure equalisation pipe which balances the forces in the two cylinders.
Two pressure switches, one a standby to the other, to sense the loss of
pressure in the pressure equalisation pipe.
The circuit is arranged to arrest gate motion if there is an obstruction to
closure on the sill, even if the obstruction is off-centre:
The speed of operation is sensibly constant.
The cylinder-mounted manifolds ensure that the load is always positive in the
same port of the cylinder, the annulus. Thus the load holding facility within the
manifold is only on the positive load pressure port. Load holding is achieved by
the two pilot-operated check valves and a check valve.
A direct-operated relief valve is connected to tbe load holding port. It is set
above the normal operating pressure. In the event of an excessive external load,
this valve will relieve the resulting excess pressure and prevent damage to the
gate and the cylinder.
A shuttle valve is required in the manifold to provide a pilot signal which
operates the load holding pilot operated check valve when gate raising or
lowering is carried out. This allows free flow from each cylinder to equalise
the load pressure. It prevents torsional or racking loads in the gate by out-ofbalance cylinder forces.
Slide gates and roller gates, when controlling flow or opening and closing
under unbalanced hydraulic conditions, must be operated by a byd raulic servo,motor. Fluctuating and variable hydrodynamic forces act on the gate. To
prevent these from causing gate vibration, the piston and the annulus side of
the servo-motor are pressurised during opening and closing of the gate. In gates
subject to high heads of water and therefore high flow velocities, the closure
speed must be controlled to prevent hydraulic downpull forces from
accelerating the movement.
Operating gates are required to hold partial openings for a long time, and the
extent of the opening is often critical. Servo-motors are required to maintain
pressure in the piston and annulus side of the piston during partial opening,
although the hydraulic pumps are not in operation. This is effected by a
hydraulic accumulator. \X1hen the pressure in the accumulator falls below a
preset value, the pump or pumps are automatically started.
Leakage of oil occurs in hydraulic cylinders. Oil leakage past the piston seals
becomes progressively worse as seals become worn. Because of this, gates
operated by hydraulic cylinders in the open or partially open position will
gradually close. Gates are fitted with position indicators and these are utilised
to signal a predetermined gate movement, and to initiate a signal to restore gate
position.
Features which should be part of a reliable oil hydraulic system comprise:
(a) Inlet strainers for the tank.
(b) Suction and delivery filters with 5 J.l.mapertures.
(c) Two manually operated pumps.
Cd) Directional control valves which can be manually operated in addition to
Operating
machinery
Hoist speed
Hoist speeds of gates are conventionally 300 rum/min. In gates which control
water level, hoisting is either carried out in steps followed by a dwell period,
both of which are controlled by timers, or by set point control.
The final closure speed of servo-motors is usually decelerated. This can be
effected by using hydraulically cushioned cylinders, whose pistons have a
stepped crown which mates with a recess in the cylinder cap. If variable deli very
hydraulic pumps are used, the pump output is reduced when a limit switch is
actuated shortly before the end of the stroke of the piston rod is reached. In dual
pump operation, one pump is shut down by the limit switch.
For constant delivery by a hydraulic pump, the gate opening and closing
speed will vary with the cross-sectional area of the annulus and the piston side
of the servo-motor. This can be compensated by using two pumps, both
delivering only when the piston side is under full pressure. Two pumps also
ensure greater reliability. In the event of failure of one pump, it is accepted that
gate opening and closing speeds will vary. Figure 6.8 shows an alternative
arrangement of an oil hydraulic circuit for spillway gates, which ensures sensible
speeds for raising and lowering the gate when only one pump supplies the
power and the second pump is a standby.
Reference
1. Back FA A; Wilden D L (1988): Automatic flood routing at Victoria Dam,
Sri Lanka, Commission Internationale des Grandes Barrages, 16 If; Congress,
San Francisco, Q63, R52.
Seals
Seals are required to prevent loss of water. Jets emitted at inadequately sealed
sills, sides or soffits are a major source of gate vibration and gate noise. Under
severe winter conditions water leakage results in ice formation and can freeze a
gate to a pier. In addition, seal leakage in gates subjected to high head can cause
long-term damage to downstream concrete works. The selection of seals and the
design of their mountings is therefore important.
Since fluctuating pressure due to flow through gaps is a major cause of gate
vibration, aspects of seal arrangements are also discussed in Chapter 10.
Seals at old gates were either ofleather or staunching bars. They were used to
prevent water leakage at piers or abutments.
Block seals of timber or iignumvitae which were fitted to older installations are
not suitable, as explained in Chapter 10. Modern seals are extruded or moulded
from an elastomer. The usual materials are natural' rubber or polychloroprene,
known commercially as Neoprene. The elastomers are compounded to produce
the necessary properties such as tensile strength, tear resistance, low water
absorption, compression set and ultraviolet resistance, and they contain
antioxidants. Seals are normally specified to have a Shore A hardness of 65 with
a tolerance of 5. High head gates often have seals of greater hardness. At
hardness significantly lower than 65, the coefficient of friction between a seal
and a stainless steel sliding surface increases. The friction coefficient is also
affected by the surface finish of the seal contact face.
Approximate values are:
Shore A 55 hardness: coefficient of friction 0.8
Shore A 65 hardness: coefficient of friction 0.7
Shore A 80 hardness: coefficient of friction 0.6
For PTFE covered seals the coefficient of friction reduces to 0.1.
Side and top seals rely on water pressure to aid sealing. The gap between a
gate skin plate and the side rubbing strip must be sufficiently wide to permit
inaccuracies of tracking and deflection under load, including dimensional
variations to the seal contact faces. The seals must be able to accommodate
these variations. If the gap is made too wide, the seal may not develop
adequate contact pressure or may extrude through the gap under the
hydrostatic head.
Seals should be preset, that is the stem should be placed under deflection, but
not under compression, because the force required to compress a seal bulb is
very much greater than that required to deflect the stem of a seal. If friction is
to be minimised, a preset of 3-5 mm is advisable. Less will be adequate for
effective sealing but is unlikely to be sufficient to allow for dimensional
variations of the seal contact plates and the gate structure.
Bottom seals rely on the weight of the gate to provide the contact pressure for
sealing. Figure 7.1 shows typical seal shapes.
Bulb seals in the shape
of a musical
note
These are more frequently used in their solid, rather than hollow, bulb
form. The solid bulb seal has a reduced contact area because it deforms less
than the hollow bulb type under water pressure and is less liable to
compression set.
ruLe
SEALS
i , SlIAPED
(j\1rT\
I'-!_)
DOuaLE ejLEI
+1
SEALS
PI FE COVERED SEALS
SEALS
r,AlE
SKill
PLATE
.Sf.IUNG
I'IASIIER
'"
STAINLESS
~ ~*=I
!;;
~
;;;
SiEEL
fERRULE
~rLb-~~;:-----SLUICEWAY
WALL
(a)
GASKET
SEAliNG
WASIIER
SIAINLESS
StEEL
fEnnULE
TO ENSURE
EVEN (LAflPltlG---+:A'~
Sf AlItlG
WA SliER
(b)
Figure 7.2(a) shows the commonest ,trrangement of a bulb side seal. Distance
c must allow an adeguate area for the upstream water pressure to force the seal
into contact with the stainless steel contact plate embedded in the pier face. The
seal clamping plate must not be too close to the bulb to permit the seal to flex
under pressure, distance b. Mounting the side seal on an angle section permits
site adjustment. Clamping of the seal mounting angle and the skin plate should
be on the downstream side of the skin plate to permit the sill seal to be extended
to a junction with the side seal.
L-shaped seals
This type of seal (Fig. 7.2(b)) can be more effective than the bulb seal, but can
only be used for movement in one plane, that is in radial or vertical-lift gates and
not in bottom-hinged flap gates. The L-shaped seal is inherently more flexible
and can blow through or 'fold under' if clearances become excessive.
Sill seals
SilJ seals (Fig. 7.3) are formed by a rectangular elastorneric section. They should
be mounted as far as possible downstream of the gate. Bulb seals, elastomeric or
timber blocks should not be used as sill seals.
To ensure even clamping by seal retaining strips, the bolts passing through
bulb and L-shaped seals should incorporate ferrules, as shown in Fig. 7.2.
Bolt holes through the skin plate for seal clamping are a frequent source of
leakage. This is avoided by fitting nylon washers or washers faced with an
elastomer under the nuts.
Splices in seals should be hot vulcanised, whether in the factory or on site,
and corners should be moulded. Special moulds are available for the junction of
different seals.
When designing a gate sealing system it is desirable to arrange sill and side
seals to be in the same plane. This applies also to the uppermost or lintel seal for
gates in conduit. It simplifies the sealing of the junction between the seals. If
this cannot be achieved a block seal must be introduced to bridge the gap, but
these are difficult to engineer successfully (95;', of seal leakages occur at corner
seals).
Rubber has a tendency to contact bond when kept under high compression
for a long time. Where gates are infrequently operated and are under high
hydrostatic head, it is advisable to uprate the coefficient of friction by 2()%.
1
Detail design aspects
UPSTRf AM
DOWNS TREAM
3mm
SILL
DEFLEcrIO'~
OF SEIIL
BEAM
vertical-lift gate
radial gate
submergible radial gate
bottom-hinged flap gate
mitre gate
drum gate
double leaf vertical-lift gate
vertical-lift gate with flap
bear-trap gate
cylinder gate
vertical-lift tunnel gate
culvert valve (reverse tainter gate)
~I~;?~>~
(a)
(b)
(c)
(d)
(g)
(h)
(i)
(k)
to
(e)
(I)
(In )
req1liring soeziing
FLUME WALL
SE ALiNG
--------i
WASHER
(a)
1
Detail design aspects
(el
area with the sill beam. Other seal profiles have been used but are incorrect for
hydraulic reasons. This is more extensively discussed in Chapter 10 where it is
pointed out that to prevent gate vibration there must be a sharp point of flow
separation.
The seal is set to deflect 2-3 mm when the gate seats on the sill. The sill seal
should be located downstream of the weir plate; upstream it can cause flow
separation. This favou rs the arrangement of the side seal downstream of the skin
plate, as shown on the lower section ofFig.7.5(a). The seals are then in tbe same
plane and a leakage path at the gate corners can be eliminated.
The sill seal can be used to take up only Limited variation between the weir
plate and the sill. Excessive projection of the seal below the weir plate causes it to
deflect and leakage to occur. The angle at the sill should not be more acute than
45, otherwise lip deflection will result in sealing difficulties.
Inpractice, local leakage can occu r clue to dimensional variations and the seal
clamping arrangement shown in Fig. 7.6(b) has been employed. The upper
adjustment screws are used to increase the pressure on the lip of the seal, forcing
it into contact with the sill.
Sealing vertical-lift
gates
Side sealing
Fig. 7.7.
Sill.realil1g
Sealing bottom-hinged
Side sealing Figure 7.9 shows different arrangements of side seals. Bottomhinged flap gates are frequently used at tidal barrages because they can prevent
the flow of estuaria.l salt water into the fresh water river course. In this
application they may have to resist pressure in either direction. The seal
arrangements of Fig. 7.9(a) and (b) will effect this.
,.
~ ',' i,
,~
"
'II'
t-----+-I--t- ....
rUl }:,,)
...
:,:, ;.
UPSTREAM
- '\
~~.
-..',
OOWNsrllWI
~f--ll-
nOWNSTflEAM
IJPSTI1EAt-1
Sf ALING
SEALING
+ 0.86
where f.L = coefficient of friction and p" = applied pressure N/mm2. For PTFE
faced seals f.L is reasonably constant at a value of 0.1 throughout the range. The
size of the area in contact has no influence on the coefficient of friction.
UPSIREAI'l
-(
r
DOWN 5 J REA H
The seal friction equals twice the length of the seal (on one side of the gate) X
average pressure (Pa) X coefficient of friction (f-L) X width of the seal in contact
with the seal plate.
Seal contact plates should have a machined face and a surface finish of )1.6
or better.
Side seals are designed to deflect 3 mrn on assembly to ensure that they
remain in contact with their contact plates due to dimensional variations and
thermal contraction. For calculating hoisting forces due to seal friction, the
force to deflect a seal adds to the friction force.1
If calculations for gate vibration are carried out (see Chapter 10), the seal
friction is a damping force. It would be prudent to omit the additional force
due to seal deflection.
SIDE SlAUNCHING
a= RESISTING
DIRECT Ion
OVERflOW
SKIN
PLA I E
(bl
Figure 7.11 shows a side guide roller. Smaller gates of aspect area up to 40 m are
sufficiently rigid not to require side guide rollers. However, in the event of rope
Hydraulic
gates and
valves
or GAl!
HOVEH[NI
.'
-\
GAl friVOl
([Hill[
LUll
Of IIHIGf Sf An IIHi
-(_I
,
Figure 7.10. Hinge sealfor
bol totJl-bingedf/ap grzteJ
Sill
SEAL
TO WITHSTAND
~(
.l-).-I /'
"-1--
UPSTREAH
HEAO ONLY
failure or jamming ofa gate due to other causes, the provision of guide rollers
would be justified. Larger gates are fitted with two guide rollers per side.
gates
In
",
.!--_i
.,
:'1
'.'
\.
(J
and (J 2
Operational
condition
Contact surfaces
Rolling motions
between nonhardened contact
surfaces
Rail-wheel
(JI
Roller-wheel or
roller-axle (J2
Gates (not
frequently
operated)
l.85 (Jll
1.6
(Jll
cylindrical
cylindrical
2.00 (JB
The two materials in contact with one another have different ultimate tensile
strengths and aB is the lesser of the two values. For hardened contact faces, the
stresses may be increased according to the hardness of the material.
The above values apply to rolling components temporarily immersed in
water. For rollers permanently immersed in water and temporarily exposed to
heavy water flows, the Hettl stresses should be reduced:
for 0 to 300 revolutions under load per year:
above 300 to 2000 revolutions under load per year:
above 2000 to 20,000 revolutions under load per year:
above 20,000 revolutions under load per year:
by
by
by
by
10%
15'%
30%
40%
For spherical revolving surfaces (crowned rolling face) with a diameter ratio
of 2=15:1, the permissible Hertz2 pressure between wheel and rail may be
increased by 50%. The diameter ratio is twice the radius of the crown of the load
roller divided by the diameter of the load roller.
The values given in Table 7.1 should be halved for Stoney rollers, due to the
uneven load distribution which occurs in Stoney-roller trains. This does not
apply to caterpillar rollers because they are not set in a fixed train.
These criteria lead to very high stresses, particularly if manganese or nickel
manganese steels are used. USA and UK practice is to use lower Hertz pressures
(0.7- o.s of these values).
Bearings are either bronze alloy bushes or bushes with lubricant inserts as
shown in Fig. 7.16, used because of their established performance in underwater
conditions. This applies whether the bearing is sealed or not, because the seal is
liable to break down after 10-15 years due to ageing or wear. Graphite
containing lubricants should not be used in conjunction with stainless steel as
this causes electrolytic action which is accelerated underwater.
If a plain bush is used the roller is often crowned to ensure that it will
centralise and that the pressure distribution is symmetrical. On vertical-lift gates
crowning of the load rollers compensates for deflection of the gate, which
would otherwise cause excessive contact pressure on one side of the roller.
With self-aligning bearings of the angular contact type, the rollers have
parallel faces and the articulation of the bearing ensures centring of the roller.
The design of the sealing system and the back-up to prevent ingress of water
through any bolted face is critical. This is achieved by '0' rings. For maximum
reliability it is necessary to ensure that seals are lubricated. Some gate
manufacturers provide separate grease passages to each seal.
-10000 k N
300
The difference in the coefficient offriction between a bronze bush and a roller
bearing (0.1 as compared with 0.0018) makes an appreciable difference to the
lifting force required, as the following example illustrates (see Fig. 7.12).
With a bronze bush:
H,
10000
Where
adjustment
important
The roller
accurately
position.
10000
0.0018
300/600 = 9 kN
Trunnion assembly
The trunnion assembly consists of the bearing housing or pedestal which is
bolted to the trunnion beam, which is anchored to the pier. A trunnion spindle
or axle and a bushing or bearing complete the assembly.
The gate load is transmitted either via the trunnion beam to the pier nosing
or, in some smaller gates, directly into a recess in the sluice wall (Fig. 7.14).
Spherical bearings can take the form of roller bearings (Figs 7.13 and 7.15), or
slide bearings formed by an inner or outer sphere (Fig. 7.16).
GATE
~_
ECCr:ll1nICIIY
10
N'ln~1 AI.IGrrl.IEIII
Spherical bearings are more expensive than bushed bearings because of their
size and additional parts. However, they compensate for a degree of misalignment
of the gate arms, construction tolerances and thermal changes in dimension. Such
misalignments must be very small because they can result in problems with the
tracking of a radial gate. When bushed bearings are used, even a slight
misalignment of the gate arms in the horizontal plane causes non-uniform pressure
distribution on the trunnion shaft and the bush. The coefficient of friction of a
conventional lubricated bronze bearing material varies between 0.2 for starting
conditions to 0.08-0.1 for permissible bearing pressures of 200-300 bar, although
in practice designers tend to limit bearing pressures to 70-80% of the permissible
values. For self-lubricating bronze or alloy bearings the starting coefficient of
friction is stated to be 0.1, and under running conditions 0.08.
Self-aligning double row roller bearings, such as the ones shown in Figs 7.13
and 7.15, have a coefficient offriction of 0.0018, although a higher value is often
assumed for design purposes.
The friction at the interface of the bearing and shaft in Fig. 7.16, or the two
faces moving relative to one another in a self-aligning bearing, causes a bending
moment on the gate arms. From this point of view, roller bearings are the
preferred choice. On the other hand, roller bearings require maintenance, which
is not the case for self-lubricating bearings.
Figure 7.17 shows a trunnion assembly with a spherical self-aligning, selflubricating bearing. The lubrication is provided by solid lubricant inserts, as
shown in Fig. 7.16.
The trunnion shaft is of austenitic stainless steel, the same as the material of
the bearing's inner ring. A bronze sleeve is interposed between the trunnion
shaft and the inner ring to prevent galling of the stainless steels, which could
prevent disassembly. The axial centre holes in the trunnion shaft permit
application of hydraulic pressure should it become necessary to disassemble
the bearing during the lifetime of the gate.
ONE BEARING
FREE 10 HOVE
AXIALLY OPPOSI IE
B[ARING
10 BE !'IXED
-PIFE
FABfll (
IORIENTED
FIBRES I
SO LlO
LlJlllll(ANI
~I
I----OUTLINE
Of
TRUNNION BEAM
tendons is typically 10-15 m. The ducts are formed using either galvanised steel
tube or polyethylene, and are arranged to prevent moisture entering the duct.
Prestressing cables for this application are encapsulated in oil to permit
restressing after relaxation.
Limit switches
Limit switches are one of the most vulnerable components of the gate hoist.
Problems range from jcing over to corrosion of contacts, failure to actuate,
Hydraulic
gates and
valves
breaking of limit switch arms and loss of calibration. Since limit switches
control overhoisting, failure can have catastrophic consequences (see Chapter
12).
Rotary (geared) limit switches are frequently used, especially in tunnel gate
installations where the gate has to be lowered a long distance. Preferably they
should act only as a back-up to position limit switches. Rotary limit switches are
difficult to calibrate due to stretch of wire ropes and thermal changes. Use of
pre-stretched ropes avoids the need to adjust switches due to the initial set of
ropes. The effect of thermal changes can be compensated by setting the gate in
the closed position in winter and in the fully open position in summer.
Because of the importance and vulnerability of limit switches they should be
provided with a back-up switch, and the electrical circuit should be arranged so
that the function of the limit switches can be tested at the gate control cubicle.
One arrangement is to make the operating limit switch resetting and the backup limit switch non-resetting to alert the operator that the primary switch has
failed. However, this can inhibit gate operation at a critical time. It also provides
no indication that the back-up limit switch has failed, due perhaps to icing or
corrosion, while the operating limit switch is still functional.
Electrical testing of a limit switch will only reveal that the electrical circuit is
functioning. Failure is often caused by a spring fracture, breaking of the arm or
loss of the roller follower which will not show up on an electrical test.
Ropes
The usual practice is to provide two ropes per side for hoisting and to attach
them to the gate by means of a load compensating arm. The arm is arranged so
that in the event that one rope fails, hoisting can continue with the other.
Fibre core ropes provide increased flexibility compared with steel core ropes;
however, the strength of fibre core ropes is less than that of steel core ropes of
equivalent size. Increased flexibility permits a smaller diameter rope drum, but
this advantage is cancelled out if strength requirements result in the selection of
a greater diameter rope. At gates, sections of hoist wire ropes are either
immersed for long periods in water or, at best, subjected to frequent splashing.
The core becomes a zone of moisture accumulation and decomposition. Fibre
core wire ropes should not be used on hydraulic gates.
Stainless steel wire ropes have outstanding corrosion resistance. For larger
sizes of rope (greater than 15 mm) there is an appreciable cost difference. If
replacement of ropes is difficult and likely to create operational problems, this
becomes a factor in the selection of stainless steel ropes. Another consideration
is fatigue life, which is lower for stainless steel ropes. If the gates are frequently
operated or if they pass over sheaves, creating additional bending stresses, high
tensile steel wire ropes should be considered.
7
Spearman recorded electrolytic corrosion of stainless steel wire ropes on
spillway gates where the ropes were located upstream of the gate skin plate. It was
not mentioned whether the contact area between the ropes and the skin plate was
lined with stainless steel. In the absence of lining, the chafing between the ropes and
the skin plate causes removal of paint on the skin plate and early corrosion.
Galvanised wire ropes can provide a reasonable level of corrosion protection.
Some high tensile strength galvanised wire ropes have a lower breaking load
than un gal vanised ropes. The reduction can be in the region of 3-5% for similar
size ropes. This is due to a reduction in the diameter of the rope strands to
compensate for the added galvanising zinc. The reduction in breaking load docs
not apply to all ropes.
Ropes are subject to elongation due to settlement of the wires in the strands
and the strands in the rope. When using geared limit switches of the type that
measure a gate's distance of travel by the rotation of a hoisting drum, the limit
switches have to be reset after the initial constructional extension of the rope has
occurred. In most cases it is advantageous to use ropes which have been
prestressed.
6
The coefficient of linear expansion of wire ropes is 12.5 X 10- per "C, the
same as for all steels. Withlong falls, thermal elongation and shortening of ropes
62.0
68.7
49.0
59.0
42.0
Chains
Two types of chain are used for hoisting gates.
In the roller chain shown in Fig. 7.19(a), known as a 'Galle' chain, the link
pins rotate in the chain links and the link pins slide in the sprocket teeth during
hoisting.
In Fig. 7.19(b) the link pin carries a bush. With this type of chain there is no
sliding of the pin relative to the teeth of the sprocket, rather movement occurs
between the bush and the pin. The chain absorbs less power and can be supplied
with grease nipples in each pin so that all rotating faces can be lubricated. The
chain in Fig. 7.19(a) is usually lubricated by drip feed.
,.-:-
-~
I_
==5
-=:=..
,.r;::=J_
?
'c:;::;;;:::;J'
(bl
(a)
ELEVATION
PLAN
PI TCH
165mm.
300 mm
ROLLER
01 AMETER
I DOmin
120 mm
INSIDE
WIDTH
10lmm
146mrn
BREAKING
LOAD
2B6 TONNES
1.57TONNE
(el
Difficulties have been experienced with chains of type (a) due to high friction
at the pin face which bears on the link. This has resulted in chains failing to fully
articulate. This is less likely to happen with chains of types (b) and (c). However,
both types have suffered from corrosion, particularly where chains are used
upstream of the skill plate of a radial gate and remain submerged in water for
long periods. Service lubrication of the joints and pins of both chain types is
only partially successful and will not prevent corrosion.
Corroded chain links will develop kinks at the joints. Even under load, some
chains have not straightened out. Kinks change the effective length of a chain
and the gate will not be raised evenly.
(0' cOIII'lcSj' of
References
1. Semperit: Gate seals cata/oglle, 3rd edition, Semperit Technische Producte, \l(Iien,
Austria.
2. DIN 19704 (1976): Hydraulic steel structures: criteriafor design and calculation.
3. Bureau of Reclamation (1996): Forensic report, spillw{!}gate 3failure Folsom Dam, US
Bureau of Reclamation, Mid-Pacific Regional Office, Sacramento, California, Nov.
4. Koltuniuk, R; Todd, R (1996): Determination of trunnion friction coefficient from tests on
reinforced spilllJJC!} radial gate, Folsom Dam, Ca., Bureau of Reclamation, Technical
Service Center, Denver, Colorado, Jun.
5. US Army Corps of Engineers (1993): Design ofbydrauiic steel strnctures, EM 1110-22105, Mar.
6. US Army Corps of Engineers (2000): Design of spillway tainter gates, EM 1110-2-2702, Jan.
7. Spearman, P C (1967): Design and development of radial spillway gates in New
Zealand, New Zealand Engineering, Feb.
8. Briden Ropes: Steel tuire ropes and fittings, publication
1304. Bridon Ropes,
Doncaster, South Y orkshi re.
Embedded parts
Sill beams, side-seal contact and roller faces on radial gates, gate guide roller and
sliding paths for vertical-lift gates and tunnel lining sections of high head gates
have to be embedded in concrete. They must be rigidly secured and accurately
aligned. The practice is to provide cut-outs in the primary concrete and means of
fixing alignment screws. The embedded parts are then accurately set up and
secondary concrete is cast around them.
To illustrate the sequence of erection, an example of an embedded sill beam is
shown in Fig. 8.1.
The pads (1) for the adjusting studs are cast into the primary concrete. The
adjusting studs (2) are then welded to the pads. This is followed by positioning
the sill beam (3), aligning and levelling by adjusting the nuts on the studs. The
final operation is to cast the secondary concrete. Adjusting studs should not be
less than 15 mm in diameter. They should not be assumed to tie in the primary
and secondary stage concrete; separate reinforcement should be provided to
carry out this function. Dovetailing the first stage blockouts on the sides is
advantageous.
Best practice is to machine the top flange of the sill beam and to line it with a
stainless steel sill-seal contact plate. The plate is either welded to the beam, or in
some cases screwed to the beam so that it can be renewed. If this practice is
adopted, insulation against electrolytic corrosion between the carbon steel and
the stainless steel is advisable.
~.~:
~
..' .~~
'.~' .
..
': ..
:.
)
o
...
. "::
"'"
,'
.. . .;
,..
'
, I
I
PRIMARY
CONCRETE
o
Figure 8.1. Embedded parts for
sill seals
PRIMARY
CONCRETE
'b_::
~NDARY
\
~~~~RETE
. ..,'
"
.>.
\ .:
.,
.. " p
..~ .
.>,
, ~\
c'
,.1
--:
'.'
o .' ._
Figure 8.2 is an example of the embedded side-seal contact face for a radial
gate and the rolJer path for the gate side guide rollers. (1) is a rail for the gate
transverse guide roller, (2) is the roller path and (3) is the side-seal contact plate.
Figure 8.3 is an illustration of the gate slot of a high-head vertical-lift roller
gate. (1) is a rail for the gate transverse guide roller, (2) is the roller path and (3) is
the side-seal contact plate.
The method of providing a rigid fixing and alignment of the embedded parts
of a lintel seal for a high-head slide gate is shown in Fig. 8.4.
It is highly desirable for all faces in contact with water to be of stainless steel.
The corrosion resistance of stainless steels depends on the alloying content of
chromium and nickel. Austenitic stainless steels with a chromium content of
15% or greater and nickel of 10% or greater have the best corrosion resistance
of the three groups of stainless steel (see Chapter 15). Unprotected low carbon
steels should not be closer than 75 mm to a water face.
Steel linings in gate slots or tunnel inverts can be repainted when stoplogs or
bulkhead gates have been positioned and the section has been dewatered. This
excludes steel linings for stoplog slots, which cannot be refurbished throughout
the existence of the structure unless the reservoir or river reach is drained down.
In practice the application of stainless steel to faces in contact with water is often
confined to seal contact and sliding faces.
The design criteria for thrust faces of embedded parts, sill beam and slide or
roller paths are empirical. The distribution ofload is assumed to be effected by
the lower flange of the beams (Fig. 8.5(b. The dimensions of the distribution
cross-section ate given in DIN 19704.1 In a gate slot, the minimum distance
from the outer edge of the concrete should, as a rule, be not less than 150 mm.
The design of the beam is conventionally based Oll that of a beam on an elastic
foundation with the modulus of concrete baving a value of C 200 N/mm3.
The usual checks apply, such as the compressive stress of the concrete below
Embedded parts
..~
I
I
w
l-
UJ
c:
U
I
I
,I
"
W
I-
UJ
c:
u
0
u
>-
c:
>-
<t
c:
<r
0
Z
0
u
0:
c,
UJ
Vl
.. --_
'
SECOIIOARY
CONCRETE
F-==-
I
I
L- -
-I------t-.....J
..
: .:
:' .0.
':.~:..:
:.
"",,\
,'"
" \J
..
:'
.,,~
DOWNSTnEAH
//GJ\~
/
/
WIDTH
or BEARING
(ON(RE TE
!_ -
L-b lb,J
2"
WIDTH
OF BEARING
(ON (RE fE
(a)
\~-- l
(b)
h~b2-b1
the transmission area and the shear stress in the beam. Some designers advocate a
corrosion allowance for all embedded parts of carbon steel. Tfthe usual design
practice for hydraulic equipment is followed (that is, derating the permissible
working stresses) this can be accommodated within such an allowance.
At high head gates the jet emitted under the gate at small openings can
cavitate and become attached to the bottom of the conduit. It is therefore
common practice to line the invert for 2-4 m downstream of the sill, and to line
upstream between one-half and two-thirds of the downstream length.
In the situation where a hydraulic jump downstream of a gate is contained
within a concrete tunnel, considerable erosion to the invert can occur due to
recirculation of debris within the jLlmp.2 Under such conditions it may be
necessary to extend the invert lining to protect the concrete.
References
1.
2.
DIN 19704 (1976): Hydraulic steel structures; criteria for desigll and calculation.
Lewin, J; Whiting, J R (1986): Gates and valves in reservoir low level outlets;
learning from experience, BNCOLD/11l7ES
Conference 0/1 Reservoirs, Edinburgh,
Sept., p.77.
Embedded
parts
[!]
Hydraulic considerations
pertaining to gates
The first section in this chapter outlines the basic data required to determine the
stage-discharge characteristics of gates. The discharge coefficients for radial
gates used in the equations are not directly comparable because the definition
of energy head varies. In some cases it may be the head upstream of the gate, in
others the head to the middle of the gate opening, and in drowned discharge
both the true energy head, the difference between upstream and downstream
water levels, and the downstream head enter the equation. Little information
has been published on the stage-discharge relationship of top-hinged flap gates.
Available data have been included in this chapter.
The next section deals with hydraulic downpull forces on vertical-lift gates.
This hydrodynamic effect is usually ignored when designing gates in open
channels, because under these conditions it is low and is absorbed by the margin
of hoisting force provided in gate installations. It becomes important for high
head gates. Because of the number of variables involved in determining
hydraulic downpuJl, calculations must be considered approximate. All research
in this field has been carried out on models representing vertical-lift gates,
although hydraulic downpull forces also act on radial gates.
Later sections draw attention to instability in a reach of a watercourse which
can be caused by the operation of a gate when there is limited ponded up water.
Problems can also arise when there is a change from 3D to 20 flow. This is the
condition of flood flow from a reservoir into a sluiceway. This type of problem
can be resolved by a physical model study. The occurrence of reflux at a
multigate installation and observed flow oscillations are described, as well as
the hysteresis effect of gate discharge during raising and lowering.
The final section begins with a discussion of vorticity at intakes. The
introduction of air into a conduit can cause severe pressure fluctuations at
control gates. Awareness that free vortices can occur should prompt
reconsideration of the design of an intake.
Cavitation and erosion are factors whenever flow velocities of the order of
13-15 mls are reached or exceeded. Cavitation can affect gate slots and the invert
of tunnels. Information relevant to gates is included in this section and IS
complementary to cavitation in valves which was discussed in Chapter 3.
Go
1.17
g
H
discharge
coefficient of discharge
gate opening (denoted b in Fig. 9.5)
gate width
gravitational constant
upstream water head
The variables affecting the discharge characteristics of a radial gate are shown
in Fig. 9.1.
An example of the coefficient of discharge map for free and submerged low
2
under a radial gate based on Metzler is shown in Fig. 9.2.
7
Buyalski used similar maps to derive discharge algorithms which can be programmed into a computer for automatic control of gates or for calculating
Hydraulic
considerations
pertaining to gates
---~-~a
r"
-..,."...,.
~--.t-.--~
0.8
b_
a0.7
1-
FREE D1S~iGE:-'
0.6
--:
//
V
_...-
L
,/
0.5
Cd
..---
"
SUBMERGED
DISCHARGE~
----r tffi
~
=0.3.;
0.4
/
I-- r/
[\- ~
-~/
H--
I // v
/
VI v
'fy].:
Yl= 0.4
a"
v- 7
; bo
!a'~
II
/ '~
0.5
/ /
a~V
V, :~
b-o
/~
/;, VI
f
t7
/'
0.4
0.2\
0.1)
0.3
//j
0.2
./'
0.6
.>
nr::
I~' .
fl; /
rj I
Y3 =0.2
f;:0.3
I~
/ 1~410.3 V6.2 ~
~O.1
/ /
0.3
0.7
b
a 1=0.2
I--- lZl!_.4
f-
a
~V l!=o
totY*
./
~.1
---\ ~
lA
_.
>--
0.6
0.2
r/Y~ :=0.8
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
YI
1'/ a = 1.5
1.0
1.1
that the experimental data show that different gate lip designs (even a minor
modification) can result in a -7% to +12% difference in the coefficient of
discharge Cu' The different seal configurations investigated were the rectangular
hard rubber section, a seal of musical note shape and no seal, that is, metal edge
contact. A seal of musical note shape should not be used as a lip seal as it can lead to
gate vibration.
The US Corps of Engineers' Hydraulic Design CriterialO include graphs for
free discharge for ratios a] R_ = 0.1, 0.5 and 0.9, where a is the height of the gate
pivot above the sluiceway floor and R is the radius of curvature of the skin plate.
The graphs arc based on Toch,ll Metzlel and Geotilini.12 However, the
geometry of many gates is outside the range of these ratios. The charts are useful
because they incorporate an ancillary graph to give adjustment factors when the
gate sill is raised above the floor of the channel.
The chart produced by the US Corps of Engineers':' for the coefficient of
submerged discharge is independent of the aIR ratio and is plotted for the ratio
of raised sill downstream submergence over gate opening. It appears that the
height of the sill above the approach bed is not an important factor if)
submerged flow controlled by gates. One of the graphs (sheet 320-8) is
reproduced in Fig. 9.3.
40.--- -,---
-,
-t-r-r-r-t-
30t---------------,--
3r---
---
DISCHARGE
COEHICIEN
r (
For radial gates on spillway crests, the discharge through a partially open gate
can be computed using the same basic orifice equation:
Q = OJ y'(2gH)
where C
A
g
H
(9.2)
coefficient of discharge
area of opening
gravitational constant
head to the centre of opening
Hydraulic
considerations
pertai ning to gates
Hydraulic
gates and
valves
110
I
II
-,
10 5
I~W
SUGGESI(D
FOR X/Hd
SUGGESTED
DESIGN
CURVE FOR X/Hcl.: 00
10 0
95
90
B5
80
\l
Hd
I,
V1
W
DE51G N CURVE
: 01 to 03
I~"\.
,"~
~0.
f-Y/
l:J
w
/'~Hd
lld
(ENTRE
UHE WEIR
(REST
CD..
70
U:CGoBrIgiT
WHERE.
Go = Nur
B : GATE
65
GATE
~llOrH
a PEHING
60
55
F igllre 9.4. US Corps oj
Ellgineers' chart [or the
coefficient oJ discharge jar radial
gales 01/ spil/illtl).crests jor gate
lip angle jl'01l1 50' to 110'
50
0.55 0.60
0.65
0.70
DISCHARGE
0.75
0.80
COEFFICIENT
0.85
0.90
0'"
Hydraulic
considerations
pertaining to gates
0'6
05
04
Cd
0'3
0'2
0,1
0
0
10
12
14
16
Yl/b
and the separation of the water stream from, or its adhesion to, the gate surface.
The hydrodynamic pressure is of a pulsing character due to random velocity
pulsations,18,19 the instability of the flow separation point from the gate, and
water level oscillations induced by wave motion or a change in flow conditions.
Distribution
of mean and fluctuating pressures for different operating
conditions is usually determined by a model study.2o
An analytical method of determining the mean value of local hydrodynamic
pressure and a mathematical model of pressure pulsation for a non-submerged
bottom-hinged
gate with circular curvature has been given by Rogala and
Winter?1
Reference 21 derives an equation to determine the hydrodynamic pressme at
an arbitrarily chosen point on an overflow hinged gate which is neither
submerged nor fully aerated. The pressure at a point depends on the geometric
parameters of the gate, its position and the flow discharge. The equation is:
pgR
where
Ho-Y - 0.85
= --
p
p
g
R
1-1
exp( 117')
(9.3)
ENERGY
HEAD
\ RADIUS
lf7
W
W
Stage-discharge
q
L
was expressed
as a plot of three
upstream head
distance from hinge to the centre of gravity of the flap
mass per unit width of flap gate
gravitational constant
clischarge per unit width of flap gate
depth of flap
density of fluid
2
For a given gate, the mass parameter 2(311l!PL is constant and so the stageclischarge relationship is read from the ordinate in Fig. 9.7 and bl I: values on the
appropriate vertical line. For design purposes the mass parameter can be
~Il"""-C-..!
h
Hydraulic
considerations
pertaining to gates
S:>
~~~::;~m~,''l'b'
i"l!ll~lzy,D,;j~qb~i~a
07
.:t'fq,"]ll~![1;'."iHii
--,---,--_..,-_----,
r----,r---,---r-
l!
l
0- 2
0-1
o-~
0-7
08
09
10mlp l'
is unity at the point where the bjL curve intersects the vertical axis,
22
The figures show that throughout the drowned flow regime for a constant
upstream water level, discharge reduces very rapidly as tailwater increases;
b - do is the effective head across the gate and, as expected, q is proportional to
where b
do
(c) Empirical
upstream head
tailwater head.
stage--discharge
relationship
Tests carried out on circular flap valves at the State University ofIowa, USA to
establish loss of head through flap valves derived the following empirical
formula:
_ 4V
L -
where L
V
g
d
2
X
exp
(-1.15V)
./
vd
(9.7)
It is assumed that this formula applies only to free discharge, although this was
not stated.
"7
HINGE~
.....-
~
I
h
"
Hydraulic
considerations
pertaining to gates
I
I
mg
---
:;
...........
o-~
0-]
!L
t7gl1 0-2
110'0/
Pnrume
l er
o-t
dd
030
0
0-]
06
01
0-8
0-1
08
2pmlpLI
0'5
o-~
0- ]
~J
Porameter
O' 2
L
0'79
0-' -
_.'--
0'
0]
0-4
06
7pm/pl'
F igllre 9_ 8_ RectallgularjlaJi
APPflOXIHAI
vtt or u v
OISI
muuuou
y
Figure 9.9. Diagram of ttmnel gate
under slIlmierged'[kn conditions
23
(after 'Naudasdser et ',II. )
(9.S)
where F
aT
an
B
d
p
V
downpull force
downpull coefficient at the top of the gate
downpull coefficient at the bottom of the gate
width of gate
depth of gate
density of water
velocity in the contracted section of the jet
The geometric parameters which affect downpull at the bottom of the gate
are:
(i)
(ii)
(iii)
(iv)
(v)
the
the
the
the
the
the
Hydraulic
considerations
pertaining to gates
e.s
0..4
- 1.0.
0..3
0..8
0..2
- 0..6
II)
ts
'"
L.J
ts
0..1
0..0.
Cd
FREE
DISCH,E/
.
- 0..4
~
Cd
0..2
SUBMERGEIJ
DISCHARGE
-0..1
0..0.
0..2
0.4
0..6
V'Vo
0..8
1.0.
0.6 -
0.2
O=-----"I,,----''---L---'--JJ
0.2
0.4
0.6
O.B
Y/ Yo
'0
08 ~
\~
\\,
\
06
04
aB 0'2
I
-,
- --_
-,
...._-
i'--.
" -,
.....
<,
F~1!,1I,.e9.12. Dependence of
bottmn dO/of/p,,1I coefficient 011
gate<l!,coJl1et~)'fora ratio 0/
conduit heigh! to depth of tbe
gate,yo/ d = 6 ,dte/" Tbang
and Nnudascher'" )
-02
-0,'
'0
08
_'-----
..-e\t IdV
.
= 06"
_t
..... _
.....
,/
rid 1=0.4
I
\ ._.
~
v")Yd
-0 2
e/d '=081
=4
.>..41/d
-- ....
=6L
~ hid' =8
t...::::::
<,
....
....
'
___
.....
e = '30
as 0.2
-0'4
_f-
l\
04
Il
06
"
r-,
<,
",
-rId = 0
---r/d=O'4
.............
h d =12
>-
Hydraulic
considerations
pertaining to gates
FIXED WEIR
RADIAL
AUmMATIC
GATE
Figttrc9.14.
1V'(//Jc(/ctiol1duc
limited j)(mded-lip water
If)
chr;/lgcfrOIll three-dimensional
to nuo-dimensiona! jlOII!
by the tail water. 27 It caused bouncing of radial gates due to surges of flow which
moved back into the gate bay and struck the bottom girder of the gate. The fluid
flow was strong enough to lift the gates, causing the bouncing phenomenon.
Extension of the piers would have reduced the load on the gates. Flow
oscillation has also been noted in other model studies.28
Hydraulic considerations
conduit
pertaining to gates in
Vorticity at intakes
When air is introduced into a conduit, severe pressure fluctuations can occur at a
control gate due to the build-up of stagnated air under high pressure at the
conduit crown upstream of the gate. The formation of free vortices at intakes
must therefore be avoided.
Two factors principally determine the formation of vorticity at an intake:
submergence and circulation of the ar.eroach low. Circulation is the primary
parameter influencing subrnergence.f ,-0 Gulliver etal.31 have suggested that, as
a first approximation prior to a model study, two design parameters should be
D1/2
Hydraulic
considerations
pertaining to gates
(9.9)
where Cis a constant of suggested value 0.3 for symmetrical approach flow. The
units of Care sm -1/2. The formula was derived from observations of 29
prototype intakes.
Intake screens will, by streaming the flow, permit lower submergence of
inlets before the onset of formation of free vortices.33,34
Aotivortex devices have been used, such as lon~ approach channel walls over
the intake, or the hooded inlet developed by Song 5 and Blaisdell and Donelly.36
o
Fr
{gO or fgs
-v
Figure 9. 16. Vortex
[ormation at a submerged
intake
Gate slots
Vertical-lift gates of the roller or slide type require recessed slots in abutments or
piers for the movement of the gate guide rollers or slides. The flow of water
across the slots causes flow separation at the upstream edge of the slot and
reattachment on the downstream side. Eddies are set up within the slots and
vortices are formed. Under conditions of high velocity flow cavitation can
occur within gate slots.
Flow conditions due to gate slots are influenced by the upstream and
downstream edge shape and the cavity depth to length ratio (Fig. 9.17).
Radiusing the upstream edge increases the flow into the cavity hence
increasing energy dissipation, so this should be avoided. A radius on the
downstream edge reduces energy dissipation. A single, stable vortex forms in
cavities where the d/IV ratio is close to unity. This results in low losses. Between
d/N} ratios of 0.2 and 0.8 circulation is unstable, with periodic disturbances
influencing the main flow. Loss coefficients for sharp edged gate slots have a
minimum value of about 0.01 with d/JV ratios of 0.5; this rises to 0.03 for
d / IJ} = 2.5. (The loss coefficient is defined as the ratio of head loss V2/2g where
V is the mean velocity.)
The flow past the gate slot results in a reduction in pressure on the conduit
wall immediately downstream (rom the slot. Cavitation can occur within the
slot or downstream from the slot when high velocity flow occurs and there is
insufficient pressure in the region of the slot.
Cavitation in valves was discussed and varying intensities of cavitation were
differentiated in Chapter 3.
Incipient cavitation is the onset of the phenomenon and usually occurs
intermittently over a restricted area. Noise is slight and there is no damage
except at isolated local conditions, such as a step.
The next stage is critical cavitation where noise and vibration are acceptable
and damage will occur only after long periods of operation. This is usually
adopted as a design criterion in gate and valve installations.
PERIODIC VORTEX
FLOW DEFLECTED
SHEDDING
---INTO
CAVlfY
-STABLE
--
VORrEX
Hydraulic
considerations
pertaining to gates
I-~
HIGH LOSS
HEDIUH
LOSS
lOW lOSS
A further stage is incipient damage when pitting occurs after short periods of
operation and is accompanied by a high noise leveL
Choking cavitation occurs when the outlet pressure is lowered to vapour
pressure. At this stage the flow is unaffected by the downstream pressure, and
flow and pressure loss relationsh.ips no longer apply. Close to choking, noise,
vibration and damage due to pitting are at a maximum.
Specialist literature should be consulted for super cavitation, the stage
beyond choking cavitation .
. Cavitation of gate slots was investigated by BaU41 and Galperin.42 May43
reviews cavitation in hydraulic structures and deals extensively with cavitation
due to gate slots.
The cavitation parameter as of a slot is given by:
(9.10)
where hi
hv
V
head
vapour head
flow velocity
gravitational constant
(9.11)
where !Cis is the value of incipient cavitation at the upstream or the downstream
edge of the gate slot.
In the graph of C3, 8 is the thickness of the boundary layer which can be
calculated from the boundary layer eguation for smooth turbulent flow.43 Since
gate slots on high head gates are long (Ill) in relation to the boundary layer
thickness (8), using a value for C3 of 1.4 will result in safe designs.
The results are applicable to a fully open gate and when flow is approximately
two-dimensional. The latter condition may not apply to an intake gate.
Galperin'12 also gives data for vertical-lift gates which are partially open.
Typical values of asi for gates discharging under submerged conditions can vary
between 1.0ata gate opening of35%and 2.5 at 90% open. For free discharge the
range is 0.3-1.0.
IB
1 6 I--
y/
U/S
r--
: r>
"
(ORHE~
/_;-- Kts
[2
1'4
f- I-/
I 2 110
(ORNER
~.
t--
t--
r-,
......
l- I--
I- -
I'--.
0'2
,-
I-
I-- 1_
002
(ORNER
U/S
I-f--- -1--1-
01,-
(ORNER
r- 1- -
r-,
06
---~I-
t-
U/S
V'
(zO'8
I-
---j-
~ ~
~i'-
0'04
t::--"
0-06
0'08
0'10
U'W
.r
;-~h=rT-
'6
TI
,4
--
1-
!:==:=
"0
( 3 O'B
06
0",
~_1L__j
ze
7
IL-,-
- - ~
--
r--
05
KI'--
- -
o 2
6
10
12
II,
W/o
o =THI(KNESS
OF BOUNDARY
LAYER
Hydraulic
considerations
pertaining to gates
(9.12)
where Hd
C
H"
Figure 9.17 illustrated flow patterns in the plane of the slot. In addition,
forced vorticity occurs in the vertical direction resulting in a complex 3D low
when the gate is in the open or partially open position. The pressure coefficients
0.3
REff
~m::E ~mtJllE'
"d:CH,
0.2
f---
---
-,
--
,
\
,lJ
._
0.1 --
lh
0
lJ
UJ
0:
!a
\
\
Z
UJ
0
IT:
0.0
0:
0...
--
--
---
--
-- ~~I-t--I-
--
---
- 0.1 --
.,,_
---
.,. "
___
It
--
--- --
no
suns
WITII UT
OFFSET
--..!...=
___
--- --
-1.0
\ ~
,
--
--
\
\
, I
\I.LI
".
--
1.0
.1 'W
Z.O
3.0
R201W
1, U TAPER
Figure 9.19. Pressure
coefficie1llS forgate slots 1]Ji/1i
and witbou! downstream offsets
EJ
given in Fig. 9.19 therefore show only the improvement which can be achieved
by providing a downstream offset. There arc likely to be significant resultant
hydraulic forces on the gate rail tending to lift it from its mounting, due to high
stagnation pressures which can develop near the downstream edge of the slot
where the gate rail is fastened.
4'1
Ball showed tbat deflectors at the upstream edges of slots produce an
ejector action which lowers pressures at the slot far below the reference pressure
and will induce cavitation. A very large deflector which causes a heavy
contraction can be used successfully, and is the basis of the design of jet-flow
gates (Fig. 2.64).
Some of the conclusions in the paper by BaJI41can be used as a guide for the
design of gate slots. Offset corners of slots and a variable rate of convergence are
most desirable from hydraulic considerations. Arcs used in this design should
have radii in tbe range of about 100-250 times tbe offset of the downstream
corner. Ellipses can also be used with excellent results.
The upstream corners of the gate slots should not be rounded or notched as
both are detrimental to pressure distribution.
The widening of slots permits more expansion of the jet into the slot, tending
to increase the contraction
at the downstream corner. However, pressure
conditions are acceptable for a wide range of slot width-to-depth
ratios in
designs using offset corners with converging walls. This is particularly true
for the 24:1 convergence and the long radius curved convergence.
Sharp downstream corners of gate slots should always be offset away from
the flow. The offset of the downstream corner of a gate slot should be small
and related to the slot width. Within reasonable limits, this offset is not
critical.
Abrupt offsets into the flow and irregularities in flow surfaces are particularly
troublesome. Offsets of less than 3 mm will cause damage. It is extremely
important to provide smooth continuous surfaces downstream from gates
operating under high heads.
Gate conduits
The investi~ation of the bottom outlet of the San Roque Dam in the
Philippines"
demonstrated severe turbulent flow separation upstream of the
control gate installation, of the type illustrated in Fig. 9.20. This was of a
periodic nature causing peak pressur.e surges. The geometry of the approach
section of the tunnel and the transition to the conduits containing the gates
affected the pressures in the gate chamber.
The paper also drew attention to the desirability of the conduit's crosssectional area at the point of gate discharge being less than that of the approach
tunnel, to avoid subatmospheric pressures which could limit the opening of the
control gate.
Where two or more gates are to be installed in parallel it is necessary to
consider effects due to the confluence of the jets downstream and potential
combinations of the jets downstream and of asymmetrical flow. Problems can
result from flow separation, unstable flow, excessive bulking, oblique flow and
cross waves.
Hydraulic
considerations
pertaining to gates
"'e
~ 30 f-t---j--,..-i=-='F=!-11--1
~
CROSS
- -.
SItAPE
SOLID
SHAPE
FtATCHED:
TUHN'El
I AH EHDE 0 ).
BLACK: TUtillEl
(OAIGltlAL I.
'"
~ 15r-t--~_-r-r-r'_'_--~~~-4
V1
TURBULENCE
DUE TO ORIGINAL
TUNNEL
SHAPE
Koch,"" in the model study of the bottom outlet at the Randenigala Project in
Sri Lanka, found that a downstream length of 8 m was insufficient for the
dividing wall. With velocities up to 43.2 mis, flow was separating from the
curved face of the dividing wall. In order to guard against low pressures which
were likely to resultin cavitation, it was necessary to extend and taper the wall by
35 m and incorporate facilities for air entrainment.
The bottom outlet of the Mrica Hydroelectric Project on the Island of Java,
Indonesia45 has twin conduits, each housing a control and emergency closure
gate of the slide type. The dividing wall extends 8.8 ill downstream of the
control gate's sill with no physical flow separation beyond the wall. The
maximum jet velocity is 33.55 m/s.
Air demand
Flow under a gate
When a vertical-lift gate in a conduit is opened and the downstream section of
the conduit contains no water, a demand for air arises due to entrainment of air
in the issuing jet.
The total air demand consists of two different parts: air entrainment in the
water flow as bubbles or larger air pockets in the air/water transition region, and
air flowing above the transition zone because of the drag of the flowing mixture.
At initial gate openings, the issuing jet is accompanied by spray which entrains a
high proportion of air.
To explain differences in air demand, flow has been classified.49 The total air
demand for free surface flow in conduits is not normally at a maximum when
gates are fully open. Often two maxima exist; one for very small gate openings,
when spray flow occurs at 4-8% of gate opening, and a second one usually larger
than the first when the gate opening is between 40-70%.
If a hydraulic jump occurs further air entrainment will take place. Kalinske
and Robertson50 expressed this in terms of the ratio of air flow to water flow. Air
entrainment
without
jumps has been investigated
by a number of
researchers.49,51
The suggested design assumptionf
is:
(J
0.03(Fr - 1) 1.06
where (J
Fr
V
g
)'
(9.13)
The contraction coefficient for a gate with a 45 lip is 0.8. The above formula
results in significantly more conservative volumes of air than those arising from
the in vestigation of Kalinske and Robertson.5o
The air admission pipes should be designed for velocities of not more than
40 mls to prevent excessive pressure loss due to flow resistance in the ducts as
well as entrance and exit air flow losses. These cause subatmospheric pressure
conditions in the water conduit.
Air flow losses can be calculated from data in the CIBSE Guide. 53
where QA
Q
a
air demand
flow over the gate
a coefficient depending on height of fall of the nappe, h4,
depth of the nappe, h3, and the Froude number Fr of the
nappe
where III is the velocity of the approach flow to the gate and hl is the energy head
above the lip of the gate.
The Froude number of the nappe is:
Hydraulic
considerations
pertaining to gates
150
v
J
100
II
50
Fr=3-10
20
-, /
/~
FR\=13-15
lO ~
05
02
0,
Figllre.9.21. Coefficient of air
demand jar an ouerflou: gate
0'
02
II
05
1-0
II
5 . 10
a
The ducts should be arranged so that the air supply in any position of
overflow of the gate is not blocked by the downstream water level, and that
the air is admitted under the gate. In order to achieve this, the outlets of the
air supply ducts are staggered as shown in Fig. 2.27. It may even be necessary
to stagger the termination of the air supply pipes relative to one another in
opposite sluiceway walls. It is usual practice to screen the outlets of the vent
ducts. The screens must be set back from the face of the sluiceway so that
they do not damage tbe side seal of tbe gate when it moves over the duct
outlets.
References
1. Rouse, 1-1 (1949) editor: Engineering hydraulics, Proc. 4th Hydr. Conference, Iowa
Institute of Hydraulic Research, John Wiley, p. 540.
2. Metzler, 0 E (1948): A model study of tainter gate operation, MS thesis, State
University of Iowa, in Prot. 4th Hydr. Conference, Iowa Institute of Hydraulic
Research, editor Rouse, H.
3. Lewin, J (1980): Hydraulic gates,jollrn. 1. W.E.S., 34, No.3, p. 237.
4. Chow, V T (1959): Opmcbarme/lJ)'drolllics,
McGraw-Hill.
5. Franke, P G; Valentin, F (1969): The determination of discharge below gates in case
of variable tailwatcr conditions, [ourn. Hydr. Re.r., 7, NQ. 4.
6. Young, L R; Fellerrnan, L (1971): Too11lcsluices calibration tests, B.H.R.A., report RR
1105, Jul.
7. Buyalski, C P ("1983): Canal radial gate discharge, algorithms and their use, Pros.
SpeciafitJ' Conf. on A donnees in Irrigation and Drainage: SlIT7Jiving external pressures,
Jackson, USA, July, editors Borelli, J; Hasfurther, VR; Burman, RD, New York,
USA, A.S.C.E., pp. 538-545.
8. Lewin, J (1983): Vibration of hydraulic gates, [ourn. [.IP.E.S., 37, 165-182.
Hydraulic
considerations
pertaining to gates
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
Ables, J I-I (1979): Vortex preble: (II iutal:e Loiuer SI Antho,!), Falls Lock and Dam,
Mississippi River, Milllicapoli.r, Minnesota, US Army Engineer Waterways Experiment
Station, USA, technical report HL-79-9, May.
Ziegler, E R (1976): Hydraulic made! vortex slt/(!)' Grand Coulee third powerplaut,
Engineering Research Centre, Bureau of Reclamation, Denver, Colorado, USA,
Feb.
Song, C C S (1974): Hydraulic mode! tests for MO)'field POIPer Plant, University of
Minnesota, St Anthony Falls Hydraulic Laboratory, project report No. 148, Apr.
Blaisdell, F W; Donnelly, C A (1958): Hydranlics of closcd condnit spill/pO)'s:part X, the
hood inlet, Agricultural Research Service, St Anthony Fails Hydraulic Laboratory,
technical paper 20, Series B.
Wagner, W E (1967): Glen Canyon Dam diversion tunnel outlets, Proc. A SCE.,
[ourn. I-:f)'dr. ot, 93, I-IY6, Nov, pp.113-134.
Ball, J W (1963): Construction finishes and high-velocity flow, Proc. A .S.CE.,
[ourn. Constr. Div., 89, (C02).
Dickson, R S; Murley, K 1\ (1983): Dartmouth
Dam low level outlet aeration
ramps, A ncold j\llagazil1e.
Anastassi, G (1983): Besondere Aspekte der Gestaltung von Grundablassen
in
Stollen (Design of high-pressure tunnel outlets), Wa.rsCrlvirtschaft, 73, 12.
Bail, J \YJ (1959): Hydraulic characteristics of gate slots, Proc.A .5. C E.,foum. Hydr.
Div., 85, HY10, 81-114.
Galperin, R (1971): l-IydrauJic structures operation under cavitation conditions,
14tb lA.H
R. COl/gress, Paris, Vol. 5, pp.45--48.
May, R W P (1987): Cauitatiouinbydraulic
structures: occnrrence and preuention, Hydraulic
Research, Wallingford, repoft SR79.
Koch, H J (1982): Schubstrahlzusamrnenfuhrung
bei einern Grundablass mit
Nemencinanderliegenden
Segmentschutzen
(Confluence of two jets created by
two parallel segment gates of a bottom outlet), If?assBn/lirtschaft, 72, 3.
Bruce, B A; Crow, D A (1984): III/rica Hydroelectric Project: &ydralllic mode! sttldy of tbe
culvert control structure, B.H. R. A., report RR2325.
Nielson, F M; Pickett, E B (1979): Corps of Engineers experiences with flowinduced vibrations, 19IhI.A.H.R.
COligress, Karlsruhe, paper C3.
Petrikat, K (1979): Seal vibration, 19th fA .1-],R. COligress, Karlsruhe, paper C14.
Naudascher, E (1972): Enrwurfskriterien
fur Schwingungssicbers
Talsperrenverschliisse (Design criteria for avoiding vibration
of high head gates),
If?'assenuirtscbaft 62, 112.
Sharma, H R (1973): A ir demandforbigb bead gatedconduits, University ofTrondheim,
Oct.
Kalinske, P; Robertson,
R A (1943): Closed conduit flow: Symposium on
entrainment of air in flowing water, A .S.CE., Transactions, paper 2205.
Wunderlicb,
W (1961): Beitrag Zlir T3eliifltlJ/g des A bflusses ill Tiefausldssen
(Commentary on air demand in cond u it gates), Tcchnische 1-1ochsch ulc, Karlsru he.
US Army Corps of Engineers: A ir demand, reglilated outlet works, Hydraulic Design
Criteria, Sbeet 050-1.
Chartered Institute of Building Services Engineers: Guide, C4-48 and C4-49, Figure
C4.3 Air flow in round ducts, Figure C4.4 Air flow in rectangular ducts.
Gate vibration
Gate vibration, when it occurs, can be a serious problem. It can result in
structural damage or restrict operation at certain gate openings. 1n some cases
vibration of a gate will occur under specific hydraulic conditions which may
only become manifest years after commissioning of the installation. Even when
these conditions have been identified it may not be easy to reproduce them so
that they can be investigated. Apparent steady-state conditions may be subject
to a minor hydraulic disturbance which overcomes the damping forces acting
on the gate and initiates an unstable motion, giving rise to oscillations of
increasing amplitude.
This chapter is intended as an introduction to the subject and offers some
guidance on design features which will prevent vibration.
Many gates incorporate elements which arc likely to result in vibration but
c?ntinue to operate satisfactorily. One possibility is that disturbing forces are of
low magnitude and are damped out; this can be the case with small gates in river
courses. Also, since gates are designed for long return period events, it may be
tbat the conditions which could cause vibration have not yet occurred. It does
not follow that gates of similar design will be equally satisfactory at a higher
head or scaled up in size.
Most of the research papers on gates deal with vibration problems. Vibration
is perhaps the most frequent cause of malfunction of gates.
(a) Extraneously
(!)'
1112 + c- + Jg
where
dt
III
y
I
c
k
.F
= F(t)
(10.1)
mass
displacement
time
damping (viscous)
spring rigidity
impressed force
(10.2)
where
Cw
kw
F
(c + cw) > 0
As a first approximation
c>O
Figllre 10.1.
systeu:
Thellibralil1g
(10.3)
if added mass damping is neglected:
(10.4)
Gate vibration
(c+cw)<O
(10.5)
(10.6)
wbere cz; = natural frequency of oscillation of the gate in water and the damping
ratio ( is:
(=
2(7/1
+ mw)J(k/(m +
(10.7)
<1
!IIw))
(10.S)
pD2L
where p
D
fluid density
gate depth or characteristic
immersion)
spanwisc width of gate
body
dimension
(gate
Jr
=-
27f
J((k + kw)/(m +
171w))
(10.9)
TR=---...".
1-
(10.10)
(!//r)2
Excitation frequencies
Two possible sources of disturbing frequencies are the vortex trail shed from
the bottom edge of a partly open gate and the pressure waves that travel
upstream in a conduit to the reservoir and are reflected back. The frequency
of a vortex trail in the case of flow induced vibration can be defined by the
Strouhal number:
S =f{L/V)
where
f
L
excitation frequency
a representative length of the flow geometry (in the case of a
tunnel gate it is the width, or twice the projection of the gate
into the conduit)
a representative flow velocity at the gate
1/ =
where He
J(2gHe)
f = J(2gHc)
7L
(10.11)
The vortex trail will spring from the upstream edge of a flat bottomed gate,
causing pressure pulsations at the bottom of the gate. Where the gate has a 450
lip or a larger angle, the vortex trail springs from the downstream edge,
eliminating bottom pulsations. The use of flat bottomed vertical-lift gates to
control low is strongly discouraged.
Abelev4,s has presented a number of studies to establish the dominant
Strouhal numbers S and the excitation coefficient C\ The S values for
horizontal excitation of a culvert gate are shown in Fig. 10.2.
For a flat bottomed gate, S numbers for vertical excitation were given by
Naudaschei and are reproduced in Fig. 10.3.
The Strouhal numbers of vertical-lift gates for culverts and of the flat
bottomed type in Figs 10.2 and 10.3 are within a range 0.4-3.0 and 0.180.30 respectively. Martin et al.' reported a Strouhal number of around 0.33
for a fixed vertical gate model with extended lip. In a laboratory study by
Kanne8 on gates elastically suspended in the vertical direction, values of
0.3-0.5 were measured for a gate opening of three times the gate thickness.
There is little information available on Strouhal numbers for radial gates
either in model or prototype scales. However, at least two investigations have
been carried out. The radial gates at the Barkley Dam on the Cumberland
River, Kentucky, were subject to severe vibration.9,1o The gates were
15.2 m high and 16.8 m wide. Vibration occurred within a range of gate
opening of 0.75-2.75 m. The vibration freguency was 30 Hz. At the
Torrumbarry Weir on the River Murray in Victoria, Australia, radial gates
6.25 m high and 11 m wide vibrated11 when the gates were between 3-4 m
open. This occurred at a vibration frequency of 10Hz. In both cases the lips
of the gates were submerged by high tailwater levels when high frequency
vibration occurred. The Strouhal number for the Barkley and Torrumbarry
prototype gates was about 0.1.
The removal of sill seals from the Barldey gates during field testing largely
eliminated the excitation.f At Torrumbarry, spoilers were fitted to the bottom
of the gates to break up the regularity of eddy shedding.ll This was successful in
preventing gate vibration.
Gate vibration
4.0
3.0
S = fh/Vs
"
'" "
r-,
2.0
......._
to
o
<,
r--
0.2
0.4
--
0.6
1777777777777/
O.B
1.0
f = EXCITATION
DUE
FREUUENCY
TO FLOW
0'4
f-/
03
-02
f--
[7
17'
1--
--
f--
-- --
V
-- --
-- --
:-
f
Figllre 10.3. Strouba! number
for uerttcal excitation of aflat
/Jott()11ler(g(J~(after
Nnndasclm"}
01
---~
02
0-3
0,4 0-5
= [XC IT ATION
DUE TO FLO'"
FREQUENCY
06
O/b
However, it remains unclear why high flexural vibration of radial gates with
high taiJwater levels is a relatively rare phenomenon, because the lip geometry of
the gates at Barkley and Torrumbarry is typical of many radial gates.
A list of Strouhal numbers for various gate configurations is required to
enable quantitative anaJyses of transmission ratios to be carried out for most
gates.
The frequency of a reflected pressure wave is given by:
.r = V p/4L
where
VP
(10.12)
velocity of the pressure wave (the vaJue of V p ranges from
1400 mls for a relatively inelastic conduit to 1000 mls for a
relatively elastic pipe)
length of conduit upstream from the gate
.fr
=-
21f
where E
.r
V((gE)/(12sO'))
(10.13)
Gate vibration
Added mass
Figure 10.4 illustrates the simplest case of added mass. In this case, the added
mass is the total mass of water above the piston; oscillation of the mass of the
piston 111 forces the mass of water above to move with the same velocity. \X1hena
submerged body oscillates with small amplitude in a stagnant fluid, some of the
fluid will oscillate in phase with the vibration of the body, but the further away
the fluid is from the body the smaller its velocity compared with the body. The
added mass 1llw has the dimension of a virtual volume of fluid.
Computational methods have been established for determining 7Jl",. They
assume stagnant fluid conditions with potential flow induced by a body in
harmonic vibration. The investigations
were carried out by Wendel,12
Zieokiewicz and Natb 13 as well as Derunz and Geers. 14
Added mass coefficients em (see equation (10.8 have been established
.
'15-'l!l
experimentally.
The experimental work of Hardwick15 and Thang17 showed that the
amplitude of vibration has little effect on em, but Thang found an appreciable
effect at frequencies between 9-12 Hz and some effect at vibration frequencies
greater than 23 I-Iz.
A gate stiffened by girders is analagous to the condition in Fig. 10.4. The
added mass and em will be considerably greater than that of a closed type gate.
The design of the gate bottom also has an appreciable effect on em, as shown
in Fig. 10.9.
K=RIGIDITY
OFTHE
GATE SUSPENSION
Figllre 10.4. Added mass
1 5
k-::'_
1 4
em
1 3
1 2
11
1 0
/'
Hydraulic
gates and
valves
EFFECT
\VFREQUENCY
.i->: .~.-
_.
--- --
----
--~'T
.-
----
___.. f
cf
Gate opening
E = Submergence
degree
W = Circular
vibration
frequency
= 2n:f.
v = Kinematic viscosity
f :: Vibration frequency
Yo = Vibration amplitude
(Hz)
h-cf
0
inflow
010 = 0.67
E = 586
7
wiTH
(ONDU
-- -- ---
v--:- .--t::-- .
BOTiH
IT PARTS\
V:iX '
I ./
/I /
WITHOUT
CONDUIT
I"-WITH01UT
t5....--mr
__.J
1/'
,_--
.CONDUIT
UPSTREAI1
PART
~/
I-
jl/
~V
Gate vibration
1----
1/' -~
. 5
J-_.--__ __
-----
->
/'
./
"'./
10
12
14
16
E=h-cf
-0Cmr:: Reduced added mass coefficien t
( without 75% contained
water mass
-without
open stiffening
girders)
Other dynamic forces which can cause gate vibration are wave action,
cavitation, two-phase flow and water column separation.
Kolkman3 has suggested that vibration due to unsteady flow probably arises
via a mechanism involving a fluctuation discharge coefficient, induced by the
added mass flow of the vibrating gate. These conditions cannot be analysed
theoretically and are described later in this chapter.
3-0
2'5
\ -,
h-O
E= -0- =13
I'--.
20
\
1'5
"
-0
E=-O-
--- -
= 585
- - - -
',0
h
0'5
011
Vibration
This is probably the most frequent cause of gate vibration. The mechanism of
self-excitation due to seal leakage has been explained by Petrikatl9 and Lewin,2o
although explanation of the hydrodynamic effect differs in the two papers.
Pctrikat mentions a case of vibration caused by the top seal for a Im\7 level
Br-----,-----,-----,-----r-----,----,,----,
both conduit
parts
~L
E =h(/
: 15
.... ...........
-----. --or......
Z ....
._-_ _~
without
conduil
fronl
pari
- - -
3~--~~--~----~-----L----~----~
__ ~
5
10
F igllre 10.9. Effect of gate bottom
added "MSS coefficient Cm (after
Tballg'1)
011
Gate vibration
vertical-lift gate at the Bharani Dam, and Kiummet21 discusses a similar case of
vibration of a radial gate at a bottom outlet due to leakage of tbe top seal. Other
examples of vibration caused by seal leakage have been given by Kolkman3 and
Mitch ell. 22
Sill seals
Sill seals should be of rectangular shape and should be moulded in a moderately
hard elastomer (Shore A hardness 65) for gates in open channels and medium
head gates. For high head gates a Shore hardness of 75 is more appropriate.
Bottom seals can be metal to metal in order to overcome seal induced
.
192324
b ut penect
C
. hi' sue 1 an arrangement 1S difElCU It to
V1ib ratJon,
"
se ali
WIt
2
effect with large and heavy gates and sometimes also with smaller ones. to
Under no circumstances should elastomeric seals project more than 5 mm below
the faceplate of a gate, and in high head gates the projection should be no more
than is required to effect a seal, about 3 mm. Wide block seals of timber or other
materials are not sujtable,24,25 because they can shift the point of flow
attachment.f"
Musical note shape seals have been used as sill seals,19,23 as
evidenced from comparatively recent model investigations,
despite their
ill!?
UNSUITABLE
Sill
SEAL ARRANGEMENT
CORRECT ARRANGEMENT
GATE IS REQUIRED
A flOW
OF Sill
SEAL
TO CLOSE AGAINST
OF 200M3 IS
(a)
UNSATISFACTORY
Sill
SEAlAAAANGEMENT
RADIAL
SUITABLE
FORA
ARRANGEMENTS
Silt
OF
SEAL
GATE
(bl
of suitabl
unsuitability for this purpose. Diaphragm seals on bottom-hinged gates are also
vulnerable to vibration. A seal of this type failed on a bear-trap weir and was
replaced by a sliding seal.27 Figure 10.10 shows some arrangements of seals
which are unsuitable, alongside the correct configuration.
Slight vibration initiated by leakage from the sill seal can usually be identified
by vibration of skin panels and small amplitude ripples upstream of the gate.
Severe vibration can cause high amplitude movement of a gate and may be
attended by loud noise.
Side seals
Leakage past the side seals of gates in open channels rarely causes vibration of a
gate as a single unit, but can initiate flexing of local structural members.f"
sometimes severely.21. 25 Tt mostly results in seal flutter, which can be very noisy.
Using two seals one after the other to suppress the jet from a leaking primary seal
is an unacceptable solution, because vibration can still be initiated by the
primary seal. The junction between sill and lintel seals and side seals often
presents design problems. Although special moulds are available for transition
sections, these require rigid attachment. It is difficult to assess the incidence of
vibration due to leakage at corners; it is probably high. The likelihood of corner
leakage can be reduced by arranging side and horizontal seals in the same plane.
Lintel seals
Vibration of medium and high bead gates due to flow past or impinging on
Iintel seals or protruding lips has been recorded by Petrikatl9,29 and Krummet. 21
Based on these and similar cases, centre bulb seals should be used in preference
to musical note shape seals (see Fig. 10.11). Flow past the openin created at the
lintel once the seal is no longer in contact can induce vibration.2
Inemergency closure gates and draft tube inlet gates, when gates are used for
initial filling of the tunnel, it is highly desirable if not essential for the seal to
remain in contact with embedded parts of the lintel structure for the degree of
opening required to fill the tunnel. Since leakage through narrow gaps leads to
vibration,2lll1y design of an upstream sealing gate which aims at a rapid increase
of the lintel seal gap after opening, will lead to a cantilever mounting of thcseal2
and be subject to excitation due to impingement of the flow.
stiffening members
omrc
GAl E
(a)
Gate vibration
110t,
of
OP E.IUUO
( b)
because a free shear layer lies close to the bottom of the gate\ Figure 10.12( c) is
the configuration of the bottom section of a diversion tunnel gate which would
lead to flow reattachment problems because the structural stiffening member is
placed too low.
60rrOl1
SECiiON
OF A RADIAL
AUTOMATIC
GATE 'NHICH SUFFERED
SEVERE VIBRATION PROBLEMS
BIlU CE H AL 2GJ
(b)
(a)
:":'~'.'
r:
0/ snsuitabie
UNSUITABLE
DESIGN OF
TlfE BorTOM SECTION OF A DIVERSION
TUNNEL GATE. THE GATE IS REQUIRED
TO ClOSE AGAINST
A FLOW OF 150m3/ s.
structural
(e)
Gate vibration
FLOW
FLOW
jump. This is likely to cause problems where structural members are located low
on the skin plate.
32
~-777777777777
Siable
Il.ollocl~nl
77777777777
lnler mlt te nt
R..altochmenl
7777777777
No Reollochnnent
(0)
SEPARATED
FLOW
-f""__-~,
EIITRAPPED
- FLUID
(b)
POSSIBLE SI-AR
LAYER DEFLECTION
OF
ENTRAPPED FLUID
Gate vibration
Two-phase flow
\Vhere air can be introduced into a conduit, severe pressure fluctuations can
occur at the control gate due to the build-up of stagnated air under high pressure
at the conduit crown upstream of the gate. The air, which is uniformly
distributed at the head race tunnel, accumulates and forms air pockets due to
the relatively low velocity of flow in the conduit and the long d.istance upstream
from the gate. The air.pockets stagnate at the upstream side of the skin plate until
they are partially drawn under the gate. When the pressurised air is released, it
reaches atmospheric value almost instantaneously, with explosive force.
A particularly severe problem of this type was noted in a model study by
Rouve and Traut,37 as shown in Fig. 10.16. Discussing the paper, Maurice Kenn
of Imperial College, London pointed out that air entraining water flows are
notoriously difficult to model, except perhaps when tested with full scale
velocities. Because of scaling problems, pressure fluctuations in prototypes
may prove less severe than those suggested by model tests.
Nielson and Pickett23 have recorded severe vibration of a reverse radial gate
which was attributed to the collapse of large vapour cavities near the gate. The
gate acted as a control valvefor a high lift lock with a maximum differential head
of28.1 m. This type of problem can only be solved by venting upstream as well
as downstream of the gate.
Singh et a/.38 have reported the dislocation of a bulkhead on a tower type
intake due to air compression. Air entraining vortices had formed at the intake
under some lower reservoir levels, and the subsequent operation of the
emergency gate 200 rn downstream of the portal caused surges which increased
the pressure on the trapped air, driving it up the intake in an air/water spout
which dislocated the bulkhead.
.
' -.
Gate vibration
~'
I'
'.'
_.
,.,
"".
..
,.
'
..'
.....,
"','
-t,:." .
AIR EIJTRAIIlE[)
.
_.:..,_,...-=---. /
BY TtT
. .
" ".......
GATE OPEN
30'/.
~ 2
co
L_~~-r~--~~~-r~~~r-~or-
50
10
VJITHOUT
TlHE
1 0
IN SECONDS
AIR
Willi
AIR
IN HODEL TEST (SCALE 1:25) SEVERE PERIODIC PRESSURE fLUCTUA TrOUS WERE OBSERVED
UPSTREAH AND DOWNSTREAH OF TilE GATE. TilE GATE VIBRATED AT A HODEL FREQUENCY
Of O.2-0.71Iz. PRESSURE fLUCTUATIONS OCCURRED AT GATE OPENINGS OF 2o-4o~ AT
TWO-PIIASE fLOW COUDITIONS ONLY.
FZ!!,lfre10.16.
RUBBER (NEOPRENE I
BOLSTER PRELOADING
GUIDE WHEELS
References
1. Naudascher, E (1979): On identification and preliminary assessment of sources of
flow induced vibration, 19th fA .HK Congress, Karlsruhe, paper Ct.
2. Kolkman, P A (1984): Vibration of hydraulic structures, in Dcoelopments in bydraulic
cllginccring2, editor Novak, P, Elsevier Applied Science Publishers.
3. Kolkrnan, P A (1979): Deuelopoeent oJ vibration free gate des~gll, Delft Hydraulics
Laboratory, publ. 219.
4. Abelev, AS (1979): Investigation of the total pulsating hydrodynamic load acting
on bottom outlet sliding gates and its scale modelling, 8tb 1.A .1-1.1\.. Congress,
Montreal, paper 10A1.
5. Abelev, A S (1963): Pulsations of hydrodynamic loads acting on bottom gates of
hydraulic structures and their calculating methods, 10tbf.A .H.R Congress, London,
paper 3.21.
6. Naudaschcr, E (1964): Hydrodynamische
und Hydro-elastiche
Beanspruchung
von Tiefschutzen, DCI'J'/ablbrJl/, Nos 7 and 9.
7. Martin, W \'(/; Naudascher,
E; Pradrnanabharn,
M (1975): Fluid dynamic
excitation, involving flow instability, Proc.A.S.C.E.,jol/l'II.
F[ydr. Div, 101, HY6,
Jun.
8. Kanne, S (1989): Vibration oJ a vertical-lift gate n/ith variable bottom geometry (in
German), Diploma thesis, University of Karlsruhe, Germany, referred to in
204
1
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Gate vibration
33.
34.
35.
36.
37.
38.
Control objectives
In rivers
The operation of gates in rivers is designed to maintain the upstream water level
for navigation or for water abstraction and to pass flood flow. Gates may be
used to lower the water level for construction
purposes such as bank
consolidation or for channel improvement works; such tasks are carried out
under the control of an operator. Automatic control is mainly focused on
upstream level control. \Xlhere flood warning systems are in operation the water
level in a reach or in a reservoir may be lowered in anticipation of a flood.
In reservoi rs
Safety of the dam must be ensured by preventing the reservoir level from rising
to within a few metres of the crest, as overtopping could destroy fill dams or
other types of dam.
Hydraulic
gates and
valves
SLOPE OF OUTFLOW
CURVE SHOULD
PREFERABLY
BE
LESS
THAN INFLOW
CURVE
INFLOW
FLOOD
HYDROGRAPH
TIME-
INFLOW
FLOOD
HYDROGRAPH
- - --
.....
"
,
SPILLWAY
OUTFLOW
\
\
of
flood propagatioll
208
1
rate
TIME-
,,
of motorised
operation
This takes the form of increasing the outflow in steps. At its simplest level gate
openings may be determined by the operator based on experience in responding
to a signal indicating river or reservoir level.
To reduce reliance on operator judgement, curves may be used which show
in 30-60 min intervals the increase in gate opening as a function of river or
reservoir level and rate of change of level during the interval.3 The charts can
be designed so that, for a gi ven rate of change of level, a specified discharge rate
is not exceeded in oreler to limit downstream flooding.
Remote control
of motorised
operation
This may be located remote from the spillway, barrage or weir, or may be carried
out from a centre controlling a number of dams. In all cases alternative or
standby operation of the gates is provided at the spillway or barrage.
Automatic methods
Cascade controls
Cascade control is mostly applied to the control of reservoirs and will therefore
be discussed in that context. The distance between the retention level and the
maximum reservoir level is divided into a series of steps, each corresponding to
a gate opening. On reaching a specific water level, the gate hoist motion is
started and the gates open in seguencc to their predetermined height, controlled
by limit switches. A frequent refinement is to provide alternative limit switches,
permitting greater gate openings if one of several spillway gates is out of
operation due to maintenance or malfunction.
Cascade control is generally used in conjunction with power actuation of
gates, either by electric motor driven winches or by oil hydraulic cylinders,
but there arc exceptions.
At the Victoria Dam in Sri Lanka counterbalanced radial spillway gates were
devised," which can open under gravity. The control system is of the cascade
type but ca.n operate mechanically without external power, with electrical
controls as a standby.
Eight gates, each 12.5 m wide, open in pairs in four stages at 0.7,2.5,4.7 and
9.35 m. Opening is actuated by floats, which over a rise of upstream water level
of O.64m operate oil hydraulic poppet valves. These direct the oil from the
piston side of the hoist cyLinders to the tank, permitting the gates to open under
gravity. When the gates have reached their appropriate opening step, an
actuator on the gate closes another poppet valve which stops the flow of oil
and locks the cylinder in position. Other floats are connected over pulleys to
electric limit switches which energise the relays for solenoid operation of oil
hydraulic, di rectional control valves. This provides a standby system for
opening the gates. Closure of the gates is by oil hydraulic cylinders, supplied
by electric motor driven, oil hydraulic power packs.
Level control
Electromechanical
This is actuated by a predetermined rise in water level above
the retention level, which initiates opening of the gates in steps and in seguence.
A water level control band is set. When the upper limit of the control band is
reached, the opening motion is started and continues in steps until the level falls
below the upper limit, when the motion stops. Closure of gates commences
when the lower limit of the control band is reached. The raising and lowering
motions are interrupted by a dwell period to prevent hunting.
An ultimate upper water level limit switch initiates an alarm signal, and some
control systems are designed so that the gate hoist dwell period is cancelled
during the time when the uppermost level is reached or exceeded.
Computer controlled (jeedback control rystenl' ) Th is moves the gates in tu rn to a set
point after determining the desired outflow with reference to the measured
inflow. Control instructions arc issued by a computer which is programmed
with the strategy for maintaining upstream water level.
In proportional
integral derivative (PlO) control, the value at which
upstream water level is to be maintained is compared with the value transmitted
by water level sensors. The difference between the two signals, the error, is
computed at fixed intervals and is used to operate the proportion
and
integration algorithm. The proportional
term causes an immediate and
longer-term corrective action, and as long as the error persists the integral term
will increase or decrease continuously so as to open or close the spillway gates.
Feedback control systems" compare the actual value of a variable with its
desired value and take the necessary corrective action. In gate control the
variable is usually water level, although the rate of change of water level may
be used as an additional control parameter. It is characteristic of a good
feedback-based closed-loop system that it maintains the desired level, the set
point, and corrects for any variations with a minimum of oscillation. The gate
should respond so that a small error results in a small opening and a larger error
in a larger gate opening. The actual as compared with the desired level should be
closely tracked by the system.
A block diagram of a closed-loop system is shown in Fig. 11.3
Changes
Measured
value
r---~-I
Set
Point
in water level
Level
measurement
--..--
Water level
- --.
Comparator
.. _- .... ---
Output
I
I
Controller
-~-
I
I
I
I
I
Corrective
device
__
... 1
Error signal
Criteria
of a successful
a closed-loop system
(a) How well the system reduces the error signal to zero or almost zero.
(I;) The final difference between the measured value and the set point, called the
'offset' in control terminology.
(c) The speed with which a system responds or restores agreement (in gate
operation speed of response is not an important factor).
The system should be stable, that is, free from large and violent oscillations.
(~
Modes
of control
(~
Proportional control
In proportional control, the output is proportional to the error signal. When
there is an increase in water level due to increased inflow, the controUer will
actuate raising of the gate to compensate for the increase in flow. Since gate
opening is proportional
to the error signal the new opening can only be
maintained if there is a permanent error, thus proportional systems tend to have
a permanent error, the offset.
OFF SE T
l_
S~,O!i!.T_T
CHANGE
VALUE
I
I
I
st r
POINTI
GI\TE MOVEMENT
IN I[GnAL
INII IA L
SET POINT
Figure t t. Proportional
plu.r inte,gral control
HEASUnEO
AUION
-----TIME
Proportional
(PID) control
The derivative term measures the rate of change and causes the system to react
more rapidly (Fig. 11.5). PID conttol is most frequently used in gate position
control.
Application
Symbols:
ofPID control
X1
Xz
= error signal = XI - X2
T; = sampling interval
Vp = output of proportional term algoridun
Vi = output of integral term algorithm
VI: = Vp+ Vi
q = nominal demanded flow rate
= gate angle, which is corrected In order to be proportional to
gate opening
Tgs = interval between gate error signals
KJ = proportional constant
K2 = integral constant
Xc
I
I
PROPORTIONAL
INTEGRAL
PLUS DERIVATIVE
(ONTROL
SET POIN T
CHANGE"-
r - -I
I
I
SET POINII
GATE OPENING
I
I
INITIAL
SE T POIN I
EJ
-----TII1E
PROPORTIONAL
CONTROL
11
The value x, of the upstream water level to be mainta.ined, is compared with the
value X2 transmitted by a water level sensor. The sensor may be a float, an
electrode or an ultrasonic device, pressure transducer or bubbler device. The
difference between the two signals Xc = X'-X2
is computed every T; seconds
and is fed into the Pl algorithm. The output V,of this algorithm is
VI' = K,xc
Vi = VI'
+ Kzxc T,
proportional
term
integral term
An upper and lower limit are placed on Vi (maximum and zero flow,
respectively).
The difference Xc between the retention level and the actual water level
requires the gate to rise and discharge the increased inflow into the reach or
reservoir, and a level difference Xc will correspond to a specific gate opening.
The gate angle is not directly related to gate opening, and the coefficient of
discharge for flow under a gate varies with gate opening. The computer must
therefore be programmed to convert the required flow rate into a set of
demanded gate angles using a polynomial fit. This may require inbuilt logical
hysteresis. The computer determines the required outflow rate from the signal
Vt and converts it to the new gate angle e.
For free discharge conditions, the upstream water level and gate angle are
input to the control program. For gates in barrages where the discharge can be
submerged, the downstream water level is also measured and transmitted to the
computer to enable the required outflow to be calculated and converted to
demanded gate angle. From gate control considerations where upstream level
has to be maintained, knowledge of the exact relationship between gate angle,
i.e. gate opening, and discharge under or over a gate is not essential. It is only
required if the data are logged to obtain a record of the flood hydrograph.
In multigate installations, each sluiceway will normally behave as if it is
independent of the adjoining sluiceways, provided the approacb to the gates is
sensibly straight. Thus the computer can be programmed to calculate the gate
opening
depending 011 how many gates are operational.
The value of the demanded gate angles will now be held for the rest of the
current sampling period Ts. The demand angles are compared with the
measured gate angles and generate gate error signals every Tgs seconds. Tgsillust
be considerably shorter than Ts, because the dynamics of the gate loop arc very
much faster than the response of the reservoir or river reach.
If any of the error signals exceed a threshold, the gate hoist motors arc started
up and close or open the gates to adjust flow. This adjusts the upstream water
level until the difference between the actual and required levels is eliminated.
e,
During a flood, the inflow will rise and subsequently fall. Any difference
between the demanded level and the measured level will
o cause an immediate correcti ve action due to the proportional term
o cause a longer-term corrective action to eliminate any difference between
and Xl (the error).
Xl
If there is any difference between Xl and Xl, the integral term will increase or
decrease continuously causing the gates to open or close. Hence there is no
requirement within the control system to have an accurate relationship between
actual and demanded outflow rates. If there is inconsistency between the two,
the integral action will compensate for it. Ultimately, after a change of inflow to
the reservoir or the reach, the flow over or under the gates will balance to
eliminate any difference between Xl and Xl'
To facilitate design of the control system and the parameters T" Tgs, KI and
](1, the variables mentioned in the following paragraphs must be fixed:
[I
o The maximum allowable change in water level above retention for a given
percentage change in the inflow and the time at which this maximum should
occur (Fig. 11.6).
o The time for the transient to settle wi thin a specified proportion of the peak
(Fig. 11.6).
o The maximum permitted variation in level, shown in Fig. 11.6. The time at
which it should occur together with a time at which this should decay to a
given tolerance band, shown by lines a and c in Fig. 11.7.
In practice this is sometimes approached differently. When a gate is operated
by an electromechanical hoist, it is advisable to limit the number of motor starts
per hour to ensure long life. If, for instance, this is fixed at four starts per hour, a
check is required that gate opening is compatible with the rate of change of
upstream level at the steepest part of the design hydrograph, and that the change
in water level above retention during the rest period of the motor is acceptable.
This should be the case when all gates in a multigate sluice installation are in
operation. It may be critical when one gate out of a two- or three-gate sluice is
MAXIHUM
VARIATION
IN LEVEL
RETENTION
LEVEL
TlME-
MAXII1UH
LEVEL
HEW DEMAND
(RE TENT ION)
LEVEL
ORIGINAL
RETENTION
LEVEL
TIME out of action due to maintenance or a defect. The gate controller can be
programmed to override any restriction on the frequency of operation when
one or several gates are not available, or when the increase in upstream water
level exceeds a critical value.
\X1hengates are operated by oil hydraulic servo-motors, frequency of motor
starting is less important because motors driving oil hydraulic pumps are started
with the pump offloaded.
Any gate installation must be operated so as to minimise surges upstream and
downstream. These could affect navigation, fishermen and other river users,
and could also be important if there is a sluice installation under automatic
control in the downstream reacb. Where the possibility of surges exists,
transients in upstream level measurement must be filtered out. The widely
adopted gate hoisting rate of 300 mm per minute does not prevent surges in
rivers unless the hoisting time is limited.
For a given change in flow rate through the sluices, or a given change in
demanded level, there will be a time limit for the system to respond, depending
on the flow discharge due to gate opening and the upstream head, or the
difference between upstream and downstream head. The speed of response is
also influenced by the rate of inflow. To a first approximation, the rate of fall
in level is given by:
rate of fall in level
outflow
9;! --------------
The speed of response can be made higher for small signals by increasing KI
and Kz. Making Kzlarger makes the system faster but less damped, i.e. more and
more oscillatory with progressively larger overshoots and undershoots.
A usual and conservative limit for a maximum outflow qz is:
level tolerance.
then
K1
it is advisable to
T ~ 15021s
Therefore it is inadvisable to use an intersample interval T; greater than
15021/20s = 750s.
Other control panuneters can be used apart from upstream level, such as
constant downstream level which is common for irrigation channels.
Telemetry
Automatically operated sluices and spillway gates are in most cases remotely
supervised by telemetry at a central monitoring station. Usually the station
supervises several installations, sometimes carrying out different functions. Key
information is displayed on a VDU and recorded on an event printer. For a sluice
gate installation, in addition to essential data such as upstream and downstream
water levels and gate positions, other data will be transmitted such as:
o availability of mains suppl y
o availability of transformers
o condition and availability of standby generating plant
o gate availability
o aggregated faults in anyone motor circuit or gate equipment
o valve position (if valves are part of the installation).
If the electrical switchgear and standby generators
building, additional data may include:
l
I
(
I
HOIST
GEARBOX
-------'
COMPRESSED
AIR S JOIlAGE
VESSELS
AT 40 BAR
COMPRESSOR
P.S.
3.
SYSJ EM
START
SIGNAL
SYSTEM
PRESSURE
AT 6 BAR
2.
A 111
COMPR E SSOR /
COMPRESSORS
i-'l
I
I
....
_...1
A III
MOTOR
",'"
EXHAUSJ
1. ON/OFF
VALVES
THE SYSTEM
IS
TO ISOLATE
STORAGE
REIlUIRED
TO OPERATE.
2. PRESSURE
CONTROL
TO AIR HOTOR.
l.
PRESSURE
STORAGE
SWIHH
PRESSURE
VALVE
ro
TO MAINTAIN
flOIVAfE
IN VESSELS.
FllOM
AIR
MOTORS
CORRO
COHPRESSORS
UNJIL
PRESSURE
ro
11AINfAIN
CHARGER
BATTERIES
DC tIOTDRS
BATlEIlY
(HARGER
D(
BAT TERIES
A ( ~IOroR
Gates requiring
Alternative means
of control
Inall automatic control systems, means are provided to revert to manual control
by an operator pressing 'open' and 'close' buttons if the automatic controls fail
to operate. These also serve to actuate the gates for testing, servicing and
maintenance. The practice of testing spillway gates varies at different dams
and with different operators.l"
Central computer operation for a cascade of dams or barrages is backed up by
10
a local computer at each barrage, for example on the River Rh6ne. At another
project the spillway gate control computer can, in the event of failure, transfer
control functions to a second computer at the power station. At Kotmale Dam8
an electromechanical level control system takes over if the computer control
fails to maintain maximum retention level.
Computer controUed systems are almost always designed to be self-checking.
At Kotmale8 an independent checking and warning system coexists with that
incorporated in the computer.
A difficulty arises in training operators to manually control spillway gates
which are normally operated automatically. Their experience of dealing with
flood events is limited and intermittent. Where a computer is provided such
training could be by simulation, although at present such applications appear
to have been confined to testing control strategies (Kotmale Dam) rather than
operator training.
Instrumentation
The two main parameters which must be measured in order to control gates and
valves are water level, which may include both upstream and downstream water
levels, and gate or valve opening. In some cases direct flow measurement is also
required, although this is more frequently deduced from calibration curves of
water level and gate or valve opening.
discharge for most valves is not constant throughout the range of valve
openings and a rating curve is therefore required. However, in small closed
conduits direct measurement of Bow is usually possible.
VVaterlevel
measurement
instruments
Electrodes
Electrodes operate as on/off devices. A number of electrodes can be used to
provide cascacle control where a gate or gates open in steps depending on water
level. Each step corresponds to the setting of a limit switch.
For level control two electrodes are used, set apart by the dead band, the
upper and the lower limit of control. The upper electrode initiates raising of
the gate in steps when it becomes submerged. De-energising of the electrode
causes the hoist motion to stop. When the water level exposes the lower
electrode, closure of the gate is started in steps.
The upper electrode is usually duplicated, one acting as a standby to the
other. An additional electrode is sometimes provided at higher level to warn
when a danger point has been reached.
Electrodes can sometimes be energised by dripping water, and can be
protected against such inadvertent activation by a sheath of nylon or Pvc.
Level gauges
of the
The float is linked by a tape to the instrument which displays level on a circular
dial, usually with two hands, like a clock. The float is steadied by guide wires.
The measuring tape actuates a sprocket wheel which operates the hands through
reduction gearing. Float level gauges are available with electrical analogue or
digitally measured value transmission, and with a series of contacts for
signalling high, low or intermediate levels. The measuring range can be up to
30 m. While the accuracy of the mechanical reading is about 2 mrn the
electrical transmission, accuracy and bias depends on the range of level the
instrument has to cover.
Pressure transmitters
Pressure transmitters arc used for depth measurement. They are encapsulated
integrated silicon strain gauge bridges. They are available for a range of
pressures from zero up to 500 bar and even higher. For water level
measurement, vented gauges are used with a conventional 4-20 rnA range.
Linearity and hysteresis are obtainable to O.l 'Yo. For good accuracy, it is
advantageous to select a pressure transmitter which only just exceeds the
required range of water level. Most pressure transmitters have a high overload
capacity. For high reliability, three pressure transmitters are arranged on a
'voting' basis.
proportional to the depth of the water at that point. This pressure is transmitted
via the pneumatic tube and a service unit to the receiver, which applies a force
proportional to the value to be measured to the balance beam system. When the
equilibrium of the balance is disturbed by a change in the measured value, the
displaced beam triggers a control contact. The servo-motor is actuated and
moves a travelling mass until the equilibrium of the balance is re-established.
The servo-motor and the travelling mass are connected via gearing to a digital
display counter and analogue or digital switching and transmission units.
Bubbler devices can be accurate to O.25%.
Devices are arranged so that a blocked measuring nozzle can be cleaned by
the application of full compressor pressure. Operationally, bubbler devices
require more frequent checking and maintenance than float level gauges or
pressure transmitters. They are frequently employed to measure head loss across
a screen, or where a wide range of water level has to be measured accurately. In
practice bubbler devices are often found to be non-operational due to lack of
maintenance or incorrect setting by inexperienced staff.
References
1. Lewin, J (1985): The control
on hydra/die
aspects
0)' flood
and flood
Engineering.
2. Lewin, J; Denham, H (1983): An adaptive control system for flood routing
through a reservoir, 1sf Int. Conference 011 I!Jdrolllic aspects offlood and flood control,
London, B.l-I.R.A.,Fluid Engineering.
3. Anon (1976): Flood control by reservoirs, Chapter 6, Spillway operation, Section 6,
Considerations for spillwayoperation, Hydrologic Engineering Centre, USArmy
Corps of Engineers, Feb.
4. Evans, T E; Halifax,P J; Floyd, D S (1983): A real time computer operated model
developed for the River Medway Flood Storage Scheme, 1.rt lilt. Conference 011
h),drat/lic aspects off/ood and flood control, London, 13.H.R.A., FI uid Engineeri ng.
5. Back, P A A; Wilden, 0 L (1988): Automatic flood routing at Victoria Dam, Sri
Lanka, Commission lntemationale des Grands Barrages, t otb Congress, San Francisco,
Q63, R52.
s.
7.
8.
9.
10.
of
Dam safety has been examined and an extensive technical literature exists on the
subject. Statistics of dam failures have been collected and analysed. 1
Corresponding
investigations into the hazard and reliability of reservoir
appurtenances are more recent. There is a greater awareness that the integrity
of a dam installation includes tbe reliability of gates controlling flood release and
the facility to empty a reservoir if a fault develops.
In an analysis of causes of embankment incidents and failures, according to
USCOLD (USNRC, 1983),22% of240 dams experienced malfunction of gates.
Since the publication of this analysis a few catastrophic events have been
recorded involving spillway gates and bottom outlets, and a number which
demonstrated risk.
The most serious risk is posed by common cause failures, that is failures
which affect the operation of a total system. These may consist of failure of the
mains supply and back-up system, failure of central control systems, fire,
explosion, an aircraft crash or ship collision in the case of barriers, or a natural
disaster such as an earthquake.
._
AXIS
OF ROTATION
Wflltl'tATE
WINCH
AXIS
'tARriNG;
r.1m
mo ..
".1m
1.Jm
PArH
OF COLLAPSE
eUCKLING
Corrosion on the steel trunnion pins had increased trunnion friction over
time. Collapse occurred when a strut brace in one of the radial arms sheared at
its connection (discussed in Chapter 7).
The possibility of failure had existed for some time and could have been
predicted, however, this was the first actual case. Was it recognised that the
design of the lubrication system and the choice of materials for the bearing
system were vulnerable? Was this not modified because of the number of similar
installations which appeared to have operated satisfactorily?
In 1992 a spillway gate malfunctioned at the Tarbela Dam, Pakistan,8 when it
became stuck during a lowering operation. It collapsed, breaking two hoist ropes,
damaging the gate and the weir. The gate was 28.6 m high and 15.2 m wide. Over a
, w-
.....
long period, the clearance between the side-sealing plates on the piers (the seal
contact plates) and the clamping bar securing the rubber seal on the gate had
deteriorated. The cause of the dimensional change was not reliably established.
An instance of failure of a dam due to inability to open the spillway gates
. ')
occurre d iill S pain.
In 1988 at the Seton Dam in Canada 10 the wire lifting ropes for the 8.1 m wide
by 10.3 m high radial gate broke during a scheduled operation, allowing the gate
- which was about half open - to fall on the sill, causing some structural damage.
The pin connections of the ropes to the gates had seized, causing the wire rope to
bend acutely: first one side snapped, allowing the gate to twist and jam, then the
second snapped. After repair, the same gate was damaged again in 1989 when
local frost expansion packed shut the operating contacts on the hoist motor. It
wound up the gate to beyond its stops until the motor fuse was blown, but not
before structural damage was caused. (Why did the overload protection fail to
come into operation?)
The Canadian Terzaghi Darn10 has vertical-lift gates 7.6 m wide by 10.7 m
high moved by screw-stem hoists. In 1994 one was damaged when the
downward power drive was not stopped in time and the gate was forced on
the sill. During a dewatering test in 1995, one of the gates jammed in the fully
open position. This was due to congealed lubricant and wind-blown dirt on the
exposed long screws.
A similar event occurred at Bray Weir on the Thames, where the open screws
operating double leaf vertical-lift gates jammed due to congealed lubricant,
dead insects and wind blown dust. In this case the problem was aggravated by
a high lead angle of the screws due to a four-start thread. This increased the
screw friction.
A spillway gate of a Swedish dam collapsed3 due to debris accumulation.
Also in Sweden, a serious breakdown occurred during the remote control of a
sector gate3 due to the gate passing the upper limit switch. The bolts on the gate
bearings sheared, causing the gate to break loose and to move down the
spillway.
At the Jackson Meadows Dam in California, the trunnions of the three radial
spillway gates became displaced.11 The dam, completed in 1995, has gates 9.1 m
wide by 4.6 m high. The displacement varied under load and temperature
conditions and was the result of failure to post-tension the trunnion assembly
anchor bolts. Part of the displacement was due to elastic deflection of the
trunnion assembly.
An example of serious gate vibration occurred at one of the spillway gates at
Dundreggan near Loch Ness in Scotland.12 Vibration at low gate opening
caused numerous fatigue cracks at stress concentrations. The cause was flow
reattachment at the gate lip. The gate had been installed during the 1950s and
had previously been operated at larger gate openings.
It should be a matter of course, on changing operational procedures, that the
possible technical conseguences are investigated; also that gate vibration should
be reported as soon as it is noticed, even if it is confined to a limited range.
At a large multigate sluice installation, self-exciting wave oscillations
occurred in the upstream basin when six openings discharged while four others
were closed by gates. 13
Some operational failures of gates due to severe winter conditions are
mentioned in Chapter 13 on ice formation, and while two examples of gate
vibration are included in the selection above, Chapter 10 on gate vibration is
more representative of the problem as a whole.
[not comprebensiue]
Oil ",ajor
failares
of sjJilllJlq)'galc
installations
Failure
Type of failure
Country
Malfunction of gates
(fatalities)19
Power suppl/o
Power supply'"
Power supply (fatalitiesj"
Vibration of outlet gate21
Power failure19,22
Russia
Structural failures of
spillway gates (5 failures)
Hoist failures
Controls
Seals leakage
Romania
India
Spain
India
Romania
Trunnion friction 7
Ice loading/brittle
fracture23
Vibration4
Vibration 12
Debris accumulation3
USA
Damaged hoist
Hoist failure/motor stalled
Hoist rope failurefjammed
gate
Hoist chain failure
Hoist chain failure
Limit switch failure
Hoist rope sheave seizure
Hoist rope failure
Hoist failure
Hoist rope failure/gate
jammed
Hoist failure/gate jammed
Brake failure/runaway gate
Brake failurcjrunaway gate
Hungary
USA
South Africa
Spain
France
Ice formation
Icc formation
Ice formation
Canada
Sweden
Norway
Russia
Japan
UK
Sweden
Spain
Canada
Sweden
Canada
Canada
Australia
Pakistan
Canada
UK
Sweden
Slovenia
dam and had completely blocked the outlet. The problem became dangerous
following a major storm in 1993.
A number of bottom outlets are never, or rarely, exercised. A survey of
reservoir appurtenances at dams in Indonesia identified some bottom outlets
which had not been operated since impounding of the reservoirs. These are
not isolated cases. Similar situations were noted in Sweden.3 Seals under
high pressure are subject to contact welding over time. Gates and bottom
outlets which have not been regularly moved may be difficult or impossible
to raise.
In 1974, the diversion tunnel of the Tarbela Dam on the River Indus in
Pakistan collapsed during the construction phase due to cavitation damage.
The cause was the sticking of a control gate in a partially open position. This
was one of the most destructive failures of a tunnel gate. rs
Table 12.2. Stlll1mao, oj published information (IIOt comprebensiue] 011majorfaiiures 0/ reservoir bottom
outlets
Failure
Type of failure
Country
Silting of inlet
Silting of inlet
Cavitation
Structural collapse of gate
Structural collapse of gate
Serious vibration (9 cases)
Cavitation of discharge valve
Cavitation of discharge valve
Cavitation of discharge valves
Spain
Romania
Pakistan
Romania
Romania
Romania
New Zealand
Australia
Turkey
When the discharge from a tunnel gate does not result in supercritical flow,
the possibility of creating highly sheared flow is present and the Tarbela
incident illustrated its destructive power.
A comprehensi ve survey of the operation of bottom outlets at 50 large dams
was carried out in Romania.16 While it may not be representative of experience
in other countries, significant deterioration,
incidents and failures were
recorded. Damage had occurred at 38 gate installations. 60% of the incidents
and failures were due to vibration problems, including two structural failures
which occurred after 8 and 20 years' operation. Four instances of intake
clogging made the bottom outlets unavailable and nine vibration problems
were classified as 'serious'.
Frequent operational
o
o
o
o
problems or deficiencies
Table 12.3. Faull JrequCIICJ' pel' 10year.rb,J1 type o/gatc (after Lagerhol!II, 1966
Type of gate
Number of gates
survey
In
Radial gates
Secto r ga res
Vertical-lift gates,
roller type
Vertical-lift gates,
sliding type
Needles
Stoplogs
234
1
Fault frequency
per 10 years
Fault frequency
% per gate per
year
362
107
590
235
125
770
6.5
10.0
13.1
2418
73
0.3
944
433
11
0.1
44
1.0
o
o
o
o
o
o
o
o
o
o
o
Norway; however, such problems must exist in other parts of the world
subject to severe winter weather (extreme environmental factors)
seal leakage, which can cause gate vibration and in winter can result in
freezing of gates (see also sections on detail design aspects, extreme
environmental factors and gate vibration)
failure of heating systems
failure of hoisting systems - ropes, winding screws, wedging of gates, motor
overloads, brake failures, chains
gate vibration
discharge valve cavitation
cavitation at the bottom outlet of high head dams
silting of the intake of bottom outlets
lack of regular exercise of bottom outlets (also mentioned by Lagerholm ')
floating debris in extreme floods
electrical cable fractures
control system malfunction
instrumentation.
236
1
Even a relatively small gate opening will cause a surge wave to travel up a
reservoir or a reach of river. On reflection it can affect the water level recorder
and in an automatic control system the reflected wave can initiate opening of a
gate. When the ponded-up water is limited it can cause severe instability, as
described in Chapter 9.
Some methods of gate operation can be a danger to river users downstream of
a dam or barrage and may even cause local flooding. Fatalities have occurred due
to excessive continuous opening of gates at the onset of a flood.
24
Vandalism can result in similar problems. Shaw and Halun
record that
vandals gained access in J ul y 1998 to the Windsor Dam near Ladysmith in
South Africa. Willie there is security at the dam it is not permanently staffed.
The vandals penetrated to the control panel and raised the 12 m wide by 8 m
deep main sluice gate by 500 mm. A large quantity of water was released into
the River Klip at a time when flow is usually at its lowest. Six children playing in
the river downstream of the dam were caught unawares by the sudden flow rise
and were washed away. Two children were rescued, but four died.
The determination of reliability must include potential shortcomings due to
design, operation, maintenance, operator training, inspection and supervision.
Common cause failures predominate, affecting the whole installation.
A ship collision occurred at the Thames Barrier in October 1997. The vessel
Sand kite collided with one of the piers and partially sank (Fig. 12.5). It
discharged its cargo of sand and gravel onto one of the rising sector gates;
fortunately there was no significant structural damage. If the gate had been
rendered temporarily inoperable, it could oat have been closed when a tidal
surge was expected. The potential for such an accident exists at barriers with
intermediate piers.
Maintenance can be deficient, of a variable standard at different stations, or
completely neglected because there is no maintenance budget or authority to
order spares, as was the case at hydropower plants in a tropical country. Stockkeeping of spare parts for gates and auxiliary equipment is absent in many cases.
Many reservoir control structures are 30--40 years old; spare parts are no longer
obtainable even for mass produced articles. A reliability assessment should
include a parallel contingency plan covering how replacement or substitution
of potentially vulnerable parts or components can be carried out.
Marking and documentation of spillway gates and auxiliary systems is often
indistinct or defective.
In a reliability assessment, factoring the age and condition of the plant is
difficult and depends on judgement. To reduce the uncertainty, recording of
comprehensive and systematic data from maintenance and failures should be
practised. This can provide information for an analysis of wear, degradation
due to ageing of plant or machines, and incipient failures where a large number
of similar components are used. Limit switches, oil hydraulic valves, seals and
electrical relays are examples of devices which can be present in tens or hundreds
in gate control and hoist systems.
In a high-level assessment of failure probabilities human factors are Likely to
dominate, such as availability of operating staff, even ts wh ich preven t staff from
reaching their control station, misreading of instruments, incorrect action,
failure of communication and failure to follow laid down practice.
A reliability assessment considers all the factors which can cause
unavailability of safety critical hydraulic structures. It can also be expanded to
quantify the contribution of each event or action to failure of the top event,
which is the safe passage of a flood or, in the case of a tidal barrier, prevention
pf the flood.
The operational and physical condition of gate systems often varies during
the lifetime of the installation. This can apply to spillway gates, and even more
to tidal barriers. Developments and encroachments on the river downstream of
a dam can affect the consequences of operational procedures at the spillway.
At tidal defence barriers, the rise in water level due to global warming will
cause higher and more frequent storm surges. These will have wide ranging
consequences.
Risks which were of a low order when a structure was first commissioned can
become more significant. For instance, at the time of carrying out a risk
assessment of the Thames Barrier in 1988,25 the probability of an aircraft crash
affecting any of the barriers was considered very small and therefore did not
warrant further consideration. Since then, the London City Airport has started
operation. Statistics indicate that the areas for peak probability of aircraft crash
now include some of the barriers on the Thames.
Spillway gate and barrier control systems are sometimes upgraded. For
example, at the spillway gates at Dundreggan 12 automatic controls were
introduced; at the Thames Barrier the original relay-based control system was
replaced by a programmable logic control (PLC) system because of problems
with the previous installation. Changes can also comprise alterations in working
practice, manning, training and the retirement of experienced operating staff.
A problem of this kind affecting spillway gate operating personnel in rural
24
South Africa has been reported by Shaw and Hakin.
As a consequence of the
increasing influence of HlV and AIDS, negati ve population growth rates across
southern Africa are causing loss of key trained personnel in ru ral areas where
they are difficult to replace, and where the provision of adequately skilled
back-up personnel is simply not possible.
Reliability of mechanical equipment can be affected by the consequences of
alkali aggregate reaction of the civiJ engineering works. Notable examples are
the Kariba Dam and Owen Falls Dam. After symptomatic repairs, these dams
are now monitored through management programmes. On a smaller scale, in a
rural environment in Africa, repair before the problem reaches a critical point is
. 24
not, al ways rea Irsttc.
Reliability assessments must, therefore, be regarded as a time dependent
overview of the causes of unavailability of gated structures. They should be
updated at intervals
to take into acount management,
operational,
environmental, technical and hydraulic changes.
Reliability
indices
The probabilistic reliability derived from a fault tree analysis can be expressed as
failure per demand (in the case of a spillway gate, this is the opening of the gate).
For the Thames Barrier, this was determined at 1.55 X 10~4 per gate per
demand?" Expressed differently, there is a chance that a single gate will fail to
close on one in 560 closure demands, and that two of the ten gates will fail to
close at one full closure in approximately 6000 closure demands. This reliability
assessment was carried out before control of the Thames Barrier was upgraded
by replacing the relay systems with programmable logic controllers (PLCs).
In the hazard and reliability study of the Flood Prevention Scheme for the
City of Venice, failure was defined as flooding of Venice more than 280 mm
above Venice datum. The design resulted in 1 event in 800 years.32
For the New Waterway Storm Surge Barrier in The Netherlands.f" the
derived reliability targets were:
o probability of not closing due to human or technical errors less than 1O~3 on
demand
o probability of collapse less than 1O~6 in any year
o probability of not opening due to human or technical errors less than 1O~4 on
demand.
For the Seven Mile Dam in British Columbia, reliability analysis35 resulted in:
o probability of failure of spillway gates to open due to environmental
9.68 X 1O~6
hazards
Hydraulic
gates and
valves
structural and mechanical plant which have some similarity. The available data
will probably be of low statistical validity. The selection of a failure probability
for each item of a fault tree branch will therefore involve asignificant element of
engineering judgement. Such judgement, whether exercised by an individual or
collectively, depends on experience.
Problems encountered with bottom outlets often stem from the interaction of
structural and mechanical aspects with hydrodynamics. Assignment of failure
probabilities to fault tree branches when investigating bottom outlets is therefore
even more dependent on judgement and knowledge of theory and practice.
Some technical papers record failure events which resulted in serious hazards.
This would not have been highlighted in a conventionally constructed fault tree
and would have resulted in a low failure probability. Integrating experience and
knowledge on a wide scale may identify areas where hazards resulting from a
rare combination of factors would result in a different construction of a fault
tree, or simply indicate the need for remedial action.
12421
installation will comprise a number of gates, and failure of a single gate, or even
several gates, would permit the discharge of a flood less than the probable
maximum or the design flood. Since interest is in a complete failure of the
spillway capacity, the analysis is likely to be dominated by events with the
potential to affect all of a set of multiple gates. Random independent failures
of multiple components are clearly possible but the probability of occurrence
is likely to be of a much lower order than common cause events, that is, an event
which affects the total installation - for instance, failure of the mains supply.
While multiple independent failures have not been shown in the fault tree, they
would be included in a real analysis.
The increase in inflow to a reservoir during a flood varies depending on the
geography of the catchment area feeding the river or rivers discharging into the
reservoir. Snowmelt can usually be predicted well in advance, whereas a steepsided reservoir in a mountainous area would exhibit a very rapid increase in the
rate of inflow, which could be as short as a few hours. A fault tree would usually
omit any failure mode which can be rectified within the period between the
onset of a flood and the time when multiple gates are required to open to
maintain the reservoir level.
The impression created by the fault tree shown here might be that the events
causing failure are fairly obvious. In practice, it would be developed and
expanded to ensure that all events are considered, not just structural, mechanical
and electrical ones. The reliability of an installation depends as much on human
factors, such as the actions of operators, the organisation, the standard of
maintenance, training and communication as on other events. External events,
, particularly when they can cause a common cause failure, would be included,
such as a lightning strike, an earthquake, exceptional wave action due to a
landslip into the reservoir or a storm of rare intensity.
The structure of the tree also allows for specific areas where redundancy
provides higher reliability, or where combinations of failures could cause
unexpected results.
The gates of the fault tree, which link the progression from a lower to a
higher event, indicate when a failure can be caused by either of several different
occurrences, so-called 'OR' gates (such as gates 3, 4, 5 and 20). The 'AND' gates
(7 and 10) require that all the immediate lower events occur to cause failure and
the next higher event.
\X1hena fault tree has been fully developed, it may raise concerns which result
in design changes. Using data on the reliability of components, the fault tree top
events can be quantified. This provides a relative ranking of the identified
failure modes, as well as an estimate of the absolute failure probability. Most
useful is the insight gained into potential vulnerabilities of the system. Even
an approximate quantification can support concerns which are raised by good
engineering judgement and experience.
References
1. (COLD (1995): Dam Failures: A Statistica! A nalysis, Bulletin 99, International
Cornission on Large Darns, Paris.
2. USNCR (1983): Safery oj existing dams: eualuatiou and improvcnICII/, US National
Research Council, National Academy Press, Washington DC, USA.
3. Lagerholm, S (1996): Safety and Reliability of Spillway Gates, lCOLD Jjll1po.ritl1ll,
repair andllpgradillg of dams, Stock holm, Jun.
4. Yano, K (1968): 011 tbc evcntofthegatede.rtructioll Ofthe IIPachi Dam, Disaster Prevention
Research Institute, Annals of Kyoto University, Japan, II-B, 1-17.
5. Ishii, N; Irnachi, K; Hirose, A (1968): Instability of elastically suspended taintergate system caused by surface waves on the reservoir of a dam,A "I. Joc. Mech, Eng.,
Fluids Eng. Diu., [oint applied mechanics, fluids engineering and bioengineering conference,
New Haven, Conn, Jun., paper No. 77-FE-25.
6. Ishii, N; lmachi, K; Hirose, A (1979): Dynamic instability of minter gates, 19th
f.A .H.R. Congress, Karlsruhe, Paper C9.
7. Bureau of Reclamation (1996): Forensic report oj spilbvqy gate 3failure, Folsom Dam,
Bureau of Reclamation, Mid-Pacific Regional Office, Sacramento, Cal, USA,
Nov.
8. Khan, K A; Siddique, N A (1994): Malfunction of a spillway gate at Tarbela after
27 years of normal operation, ICOLD,
16tbCongress, Durban, Q71, R27, pp. 411428.
9. Water World (1982): Overtopped Spallisb Dam collapses as spillwqy gates stqy shut ; 5, No.
11, p. 8.
10. Watson, M A (1997): Spillway gates: Will they open safely, [COLD, 19th Congress,
Florence.
11. McManus, R A (1999): Measurement of tainter gate trunnion displacements,
Jackson Meadows Darn, California, ASCE
Conference Hydro's Future, Las Vegas,
Nevada, Sep.
12. Noble, M; Lewin, J (2000): Three cases of gate vibration, Isritisb Dalll Society, ttt
Conference, Bath, Jun., in Proceedings, editor Tedd, P. Thomas Telford.
13. Kolkman, P A (1984): Vibration of hydraulic structures, in Developments in bydraulic
cllgillecrillg2, editor P Novak, Elsevier, p. 46.
14. Romeo, R (1996): Drawdown of the Barasona Reservoir, report by the author at the
Symposium on repair and upgrading of dams, Stockholm, Jun. Hydropouer ant!
Dams, 1996, Issue 5, pp. 65-74.
15. Kenn, M J; Garrod, A D (1981): Cavitation damage and the Tarbcla TUIUlel
collapse of'1974, ProclCE, Part 1, Vol. 70, Feb., pp. 65-89.
16. lonescu, S e/ al. (1994): Damage and remedial work during operation of several
bottom outlets, lCOLD,
16tb Congress, Durban, Q71, R7, pp. 79-90.
17. Rajar, R; Rryzanowski, 1\. (1994): Self-induced opening of spillway gates on the
Mavcice Dam - Slovenia, ICOLD,
16tb Congress, Durban Q71, R8, pp. 97-112.
18. Narayana Murty, T V S (1979): Failure of the Machhu-Il Dam, Indian [ournal o]
Pou/er and River Valli!)' Development, Mar., 54-67.
19. Reynolds, P; Hindley, M (1994): Double dam flood failures, ~IPater Pomer and Dam
Construction, 46, No.9, p. 2.
20. Utillas, J L; Garno, 1\.; Soriano, 1\. (1992): Reconstruction of the Tous Dam, LIPaler
Power and Dam Constrnaion, 44, No.9, pp. 55-65.
21. Sagar, B T A; Tullis, J P (1970): Problems with recent high-head gate installations,
Proc. ojill/emaliollall!),dralflic
research symposium, Stockholm, Sweden, paper Fl.
22. Diacon, 1\.; Sternatiu, D; Mircca, N (1992): An analysis of tbe Belci Dam failure,
IIPatcr Power and Dall1 Constructian, 44, No.9, pp. 67-72.
12441
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Ice formation
Ice
Criteria for the design of gates under ice conditions in Northern Europe are
given in DIN 19704,1 where different empirical rules apply to inland and
estuarial conditions.
Hydraulic
gates and
valves
by debris (timber, ice and other foreign objects) or lateral loading due to
thermal expansion of ice sheets. This is a distributed load of 73 kN/m acting
in the downstream direction along the width of the gate at the upstream water
level.
Starosolsky'' gives extensive information on ice formation and provides
4
some data which can be used to arrive at a rational basis of design. Otsub0
and johansson'' give examples of the effect of severe ice formatioo on gates
and some means that have been employed to mitigate it.
Under winter conditions of light frost, the side seals of gates can freeze to
their contact face. Ao attempt to operate a gate under these conditions could
result in tearing of the seal.
Where there is an operational risk of side seals freezing to their contact face,
either because moisture is trapped between the seal and its contact face or
because there is leakage past the seal, the side staunching has to be heated when
air frost occurs. Lintel seals will also have to be protected against freezing.
Heating of side-seal contact plates for radial gates under frost conditions is
extended throughout the length of travel of the side seals. It is effected by
electrical resistance cables, usually of the mineral insulated, stainless steel
sheathed kind. Figure 13.1 shows such an arrangement. Heating cables have
to be insulated to prevent undue conduction of heat to the flume walls. An
alternative method involves the circulation of hot oil.
Heating provision was made at the Tees Barrage in the UK where it was
introduced for some of the bottom-hinged
barrage gates in their closed
position. The seal contact plates of the large vertical-lift gate of the Barking
Barrier on the Thames also incorporate provision for heating.
SIDE SEAL
(0 N fACT PL ATE
GA TE
SKINPL A TE
248
1
HEA TING
ELEMENTS
INSULATION
Ice formation
To prevent ice from forming immediately upstream at the top of gates, air
bubbler nozzles can be embedded upstream of a gate. They provide a constant
emission of air supplied by compressors through nozzles set in the concrete
invert. This causes relatively warm water at the bottom to circulate upwards,
preventing ice from forming at the watcr surface. If the nozzles are not close
enough to the bottom of the curved gate skin plate, ice may form at the bottom
sill. Tn addition, leakage past the sill seal can create ice which bonds the bottom
of gates to the sill beam.
Air bubbler systems are discussed in the US Army Corps of Engineers
9
10
Manual,EMll10-8-1
(FR) and Haynesetal.
Other problems which have occurred during severe winter conditions have
inclucled:
o Absence of a spillway gate building for machinery and control equipment,
aUowing hoist machinery to be covered in ice and snow.
o An uninsulated spillway gate building, which caused oxidation of electrical
contacts resulting in failure.
o Attachments of the hydraulic cylinder operating a bottom-hinged flap gate
locate~ in a ~it which had not been drained, causing the piston clevis to freeze
In position.
o Structural failure of a spillway gate occurred on the River Svir in Russia, due
11
to ice loading and brittle fracture.
o At radial gates with parallel arms, ice has formed between the arms and the
sluice walls, forming a bond.
References
1. DIN 19704 (1976): /-I)'drtll/lic steel structnres: criteria [or design ami calculation.
2. US Army Corps of Engineers (2000): Engil/eer Manna] 1110-2-2702, Design of
spillway minter gates, Dept. of the A rrny, Washington DC, 1st Jan.
3. Starosolsky, 0 (1985): DelleloplllCllts ill b),draufic engineerillg - 3, editor Novak, P,
Elsevier Science Publishers, pp. 175-219.
4. Otsubo, K (1959): lee problems of gates at hydro-electric plant in northern districts
of Japan, 8thLA.H.R.
COligress, Montreal, Vol. TIl, pp. 2-S1-1 to 2-S1-2.
5. Johansson, H (1959): Ice problems relating to dam gates, 8tb T.A.f-f.R. COligress,
Montreal, Vol. Ill, pp. 27-S1-1 to 27-S1-3.
6. Watson, M A (1997): Spillway gates: will they open safely, [COLD 19th Congress,
florence, Q74, R4.
7. Beckerman, R W; Wagner, P A (1999): Solutions to spillway tainter gate problems,
A .5. C.E. Conference,'D'dro's [uture, Las Vegas, Nevada, Sept.
8. Lagerholm, S (1996): Safety and reliability of spillway gates, ICOLD Sj'll'PO.rilllll,
Repair alit! Upgradillg oj Dams, Stockholm, Jun.
9. US Army Corps of Engineers: Engineer manna! EM 1110-8-1 (FR)) luinterllalligtllioll
011 inlandiuatenoays,
Departmen. of the Army, Washington DC.
"10. I-Iaynes, F D; Hachnel, R; Clark, C; Zabilansky, L (1997): Ice control techniques for
corps projects, Technical Report RJ:!.MR-HY -14, US Army Engineer Waterways
Experiment Station, Vicksburg, IvIS.
11. Freishisr, A R; Rozina, L D; Rakhmanova, A L (1976): Operating experience gained
with flat hydraulic gates under winter conditions (English translation), Cidrotekbnicbeskoe .llroitaf SIIlO,USSR (No.4 pp. 348-358), Plenum Publishing, New York.
250
1
The evaluation of the IvIDE and the OBE are outside the scope of this book.
Guidance can be obtained from references 4,5 and 6, and for the United Kingdom
from references 1, 5 and 7. Charles et a/.1 give information on other regions.
When considering a spillway gate installation, the ground acceleration at the
toe of a dam during an earthquake has to be reassessed at the crest of the dam.
The shock can be considerably amplified over and above that at the dam base,
and the spillway gates will be subject to this amplified acceleration.
j_---------
WATER
Earthqual<e effects on
gates
SURFACE
~Py=
.1
huwH
HYDRODYNAMIC
DUE TO ADDED
ADDED
MASS
DUE TO WESTERGAARD
~--------,
TOTAL PRESSURE
P;
PRESSURE
MASS
HYDRODYNAMIC
PRESSURE ADDED
MASS
ON SPILLWAY
HYDROSTATIC
PRESSURE
GATE
= -awJ(J-fy)
8
where Py
a
}I!
I-I
Y
The assumptions
face cannot apply. However, the Westergaard formula is extensively used when
seismic evaluation of gate installations is carried out and is considered a
sufficient approximation.
In 1999 the lzmit earthquake in Turkey, close to the Yuvacik Dam, caused
very high wave motion in the reservoir. At the time, the reservoir level was
below the spillway weir and the gates were not subjected to hydrostatic and
added mass loading, so the combined effect of the additional impact of wave
motion was not experienced.
Kolkman 12 has developed a relatively simple method for approximating the
hydrodynamic mass for hydraulic gates of various configurations. It is based on
2D flow, without wave radiation, and can be implemented using a spreadsheet
for solving the potential low problem. It was developed for investigating gate
vibration but its use in computing hydrodynamic mass, due to seismic motion,
is equally valid.
Methods of analysis
Equivalent-static or pseudo-static method
The equivalent-static
method represents the seismic forces as an equivalentstatic force equal to the effective mass of the structure multiplied by the peak
acceleration of the input response sjectrum, thus allowing for dynamic
amplification of the input acceleration.
Computation involves the hydrostatic forces on the gate due to the reservoir
water level, plus the load imposed by the added mass of water, multiplied by the
peak acceleration of the input response spectrum <:. Additional forces are the
amplification exerted at the trunnions, the gate sill and the side guide rollers in
the case of a radial gate.
The vertical acceleration is often assumed to be 0.6 of the horizontal
acceleration. This ratio is likely to increase within the vicinity of the epicentre.
It is considered excessive to assume that peak vertical and peak horizontal
accelerations will be in phase, so a phase separation should be assumed. In radial
gates the amplification of forces at the sill beam and trunnions, due to vertical
acceleration, need not be added to the horizontal effects.
Structural damping for steel of 3% and 5% is recommended in connection
with the OBE and the MDE, respectively." Friction damping will occur at the
side seals and the trunnions of a.radial gate. For side seals this can be estimated
from the data given in Chapter 7.
An equivalent-static analysis is a 2D model and does not represent important
3D effects in the gate. It is advisable to confine equivalent-static
analyses to
preliminary assessments. Dynamic finite element 3D analyses are recommended
for gates with complex structural arrangements, for which 2D modelling would
be inadequate.f
ICOLD (1999)4 suggests that a dynamic analysis should be considered if the
structure has a fundamental frequency of less than 33 Hz. This would apply to
most gated structures. However, the response spectra at 5% damping show
natural frequencies for hard sites between about 5-12 Hz. For soft sites, the
values are between 3-8 Hz. 13 Most of the energy inan earthquake ground motion
Allowable stresses
US manuals and codes permit a 33% increase in allowable stresses when
combining earthquake loading with the normal dead and live loads. ICOLD
(1999)4 suggests that allowable stress values up to 90% of the material strength
may be appropriate because earthquakes impose an extreme loading. Since
permanent distortion due to yielding Of exceeding the proof stress could make
a gate inoperable, the material strength should, in this instance, be the yield
point stress or an appropriate proof stress. lCOLD (1999/ does not justify the
suggested allowable stress values.
Earthqual<e effects on
gates
will act as a spring due to their relatively low modulus of elasticity (see Chapter
7). The integrity of the overhead structure is at risk, and by extension that of the
gate. The practice of partially counterbalancing radial gates was used on some
older structures to reduce the hoisting effort. If these arc assessed under
earthquake conditions the possible failure of the counterbalance ropes should
be investigated, since fracture may prevent raising of the gates.
In most cases, hoisting gantries and their access ladders will require a
dynamic assessment. This should also include the anchorage of hoist motors,
gear boxes and hoist dri ve supports, as well as the vibration oflong transmission
shafts and their resulting stresses.
Lighting columns on elevated gantries are vulnerable during an earthquake
and may cause consequential damage on failure. Area lighting and emergency
lighting can also cause damage. Lights on gantries should be located at platform
level.
Spillway gate installations include a bridge lin king the piers and abutments.
This should also be part of a seismic assessment.
Overhead hoisting machinery is either mounted on the bridge, or sometimes
on a separate structure spanning the piers. The integrity of the structures, and
especially their anchorages, as well as the bolted machinery connections to the
structure, must be considered under earthquake conditions.
Many spillway gate installations include a gantry crane for handling
stoplogs. These are liable to topple over or jump rails during an earthquake.
To prevent this, rail keep plates are fitted. During the 1987 earthquake at the
Matahina Dam on the North Island of New Zealand, many tail keep plates,
fitted to prevent toppling of the transformers, sheared. The dynamic effect of
a swaying gantry crane and any measures designed to prevent toppling over
must be fully considered.
Systems analysis
A systems analysis20 to determine the impact of failed components on the ability
of safety critical spillway gates and bottom outlets to withstand an earthquake
without major consequence would include the following, in addition to the
issues already discussed:
o The analysis will be dominated by events with the potential to affect all of a
set of multiple gates. Random independent failure of multiple components is
clearly possible but its probability of occurrence is likely to be of a much
lower order than common cause events. Multiple independent failures have
to be included. The potential for equipment to be out of action because of
maintenance is frequently an important contributor to the failure of intended
redundancy.
Earthqual<e effects on
gates
Earthquake effects on
gates
..... ~
e ,
'
...':
"
.
,
:I.~,'~:"
.'.' ,
, .
~ . \:.
"
....
.: ,.
EJ
An alternative design in which the piers and the sluiceway floor are
monolithic may prevent jamming of a gate. An arrangement of this kind was
considered for the Torrumbarry Weir on the River Murray in Australia.
Electrical cables and oil hydraulic pipes which cross the movement joint
between two sections of a pier have to permit movement.
Rope sheaves and J:ope drums require substantial rope guards to prevent
hoist ropes from jumping. grooves. In vertical-lift roller or slide gates,
horizontal accelerating forces will be applied through the transverse guide
slippers or guide wheels. Figure 14.3 shows an arrangement of spring-loaded
transverse guide wheels to absorb shock. This arrangement is also used on
stoplogs and draft tube gates when they are placed under balanced pressure, in
order to locate them close to the sealing faces for good initial sealing.
RUBBER
260
1
BUFFERS
Control buildings
Control buildings must be designed to withstand earthquakes. Suspended light
fittings should not be used and emergency light fittings should be selfcontained. Oil hydraulic pipework passing through walls should be rubber
sleeved. Electrical trunking and any switchgear which is wall mounted should
be rigidly secured. Holding down bolts of electrical control cabinets and oil
hydraulic power packs must be substantial and, where they pass through sheet
metal, the area must be reinforced and the load distributed. Fire-fighting
equipment should be available.
Transformer oil tanks are vulnerable to seismic shocks, as are the day tanks of
standby generating plants. Rail mounted transformers have toppled in
earthquakes even when substantial rail keep plates were fitted.
FAn_URE
OFSPlLLWAY
GATES
OPERATOR
FAILUIWTO
RECOVER
ERRONEOUS
OPERATION
OF BOITOM
OUILET
FAILURE
OFIlOITOM
OUlLET
DAMrAILURE
rI
SEiSMIC_
EVENT
I
NO
DAM rAILUREFROM
LOWER LEVEL
OK
UNNECESSARY
DOWNSTREAM
DAMAGE
OK
EJ
full spillway flow within one hour of earthquake'. In practice this may always be
the requirement for the more severe categories of seismic event because of the
difficulty of determining the real extent of dam damage. However, the potential
for downstream damage due to operating the spillways to their full extent will
be an important factor in deciding what action to recommend.
Operatorfailure to recover While the spillway gates may fail to respond in the
manner intended because of control or other failures, operators may be able to
recover the situation in time by various planned or ad boc actions. The extent to
which this is possible will depend on the time available, but also on other factors
such as whether the dam is normally manned, whether the operators are
practised in fault finding and recovery and whether advice is available on a
communication link which is still operating.
Unintended operation of bottom outlet
Operator action or equipment malfunction
may lead to either the spillway gate or the bottom outlet opening when it is not
required. The potential for downstream damage as a result of such opening
means these events must be considered in the analysis. For the present example
it will be assumed that operators will follow clear operation procedures and will
not initiate opening of either spillway gates or bottom outlets unnecessarily. In
the case of equipment malfunction it will be similarly assumed that the operators
would quickly recognise any unintended operation of the spillway gates because
these are immediately visible to them. However, it is less clear that they would
see and recognise control indication of bottom outlet initiation or that the
outfall of the bottom outlet would be visible. This event is therefore retained
within the event tree.
Failure of botto: ouiiet Depending on the extent of damage to the dam, it may
be necessary to open the bottom outlet to provide continued lowering of the
reservoir water level. Failure to open could lead to dam failure despite successful
operation of the spillway gates, although the lowering of water levels resulting
from that operation could mitigate the consequences of any subsequent dam
failure.
Event sequence c011sequences The anal ysis is driven and limited by consideration of
events that concern the dam owners. This example is limited to those events
with the potential to cause fatalities. In practice, the owner may be interested
in a wider range of consequences such as damage to generating capacity etc.
Balanced against this interest is the greater complexity that would be required
in the event trees and the fault trees.
References
1. Charles, J A; Abbiss,
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Earthquake effects on
gates
Materials
Most gates are fabricated in low carbon structural steels. Except in high head
gates, the higher tensile stress grades are rarely employed because deflection
often becomes the critical design parameter so that sealing faces do not open
under load. In gates which are subject to ice formation, low temperature
structural steels are used to prevent brittle fracture.
In fluidways and valves where high velocity flow is experienced, and where
there is a risk of cavitation, stainless steel is selected. The same material is used
for linings downstream of gates in high head tunnels where the boundary layers
have not developed sufficiently to protect the walls from high velocity flows.
. Seal contact faces and guide roller paths are also constructed in stainless steel.
The degree of corrosion resistance of stainless steels is determined by the total
amount of nickel and chromium components. Thus high nickel and chromium
austenitic steels are generally the best choice.
Operational experience of the use of austenitic stainless steels for shafts, pins
and other parts suggests that the chromium nickel molybdenum steels
(European Norm 316 group) give better performance than the unstabilised
austenitic chromium nickel steels (European Norm 304 group).
Stainless steels in contact with one another are subject to galling, the chafing
process which causes seizure of two parts moving relative to one another. Two
stainless steel components which are physically assembled in contact with one
another, and which may have to be disassembled or replaced at some stage,
require a sleeve of a different material such as phosphor or aluminium bronze
to be interposed. Figure 7.17 shows a stainless steel self-aligning trunnion
bearing mounted on a stainless steel shaft with an interposed bronze sleeve.
Operationally there is no movement between the parts, but the absence of the
sleeve might make it impossible to disassemble them. The same figure shows
passages in the shaft to enable the use of high pressure oil hydraulic fluid to ease
separation when replacement is required.
Galling can also occur at stainless steel bolts and nuts. At a number of gate
installations where austenitic stainless steel nuts and bolts have been used,
selecting bolts of the unstabilised steel grade and nuts of chromium nickel
molybdenum steel has averted the problem.
Stainless steels are often welded to low carbon structural steels. To prevent
carbon migration at the welds, stabilised austenitic chromium nickel steels must
be used. The steel is stabilised by the addition of titanium,
Under some
conditions, unstabilised stainless steels can be welded to low carbon steels using
special welding rods with molybdenum and/or titanium additions.
Stainless steel clad carbon steels are used to achieve cost reductions. The layer
of stainless steel is rolled on top of the carbon steel to produce an integral plate.
It is desirable that the cladding be at least 1 mm thick. Welding of a clad steel to a
carbon steel is possible, but reguires a special technique. The protection of cut
faces and bolt holes in clad steels presents problems. The cost ad vantage of clad
steel does not always justify its use.
A limited number of cases of electrolytic underwater corrosion have been
reported due to the proximity of stainless and carbon steels but, in general,
operating experience in inland waters has been favourable.
Large cathode (stainless steel) to anode (carbon steel) area ratios should be
avoided. Surfaces of both metals should be painted. If only the anode metal is
painted and there is a small defect in the coating, the cathode to anode area ratio
will be very large and rapid corrosion will occur.
In saline waters, nickel-copper alloy (Monel metal) is used, and also in fresh
water where parts are in contact with brass or bronze. Nickel-copper alloy is
available in plate and sheet form, rod and bar, and is more expensive than
stainless steel. It is closer to the brass and bronze alloys in the galvan.ic series
and its corrosion resistance, especially in saline waters, is superior to that of
stainless steels.
Guide rollers, shafts and pins which are permanently or occasionally
submerged in water, are selected in high chromiumferritic stainless steel, in free
machining austenitic chromium-nickel
steel or in nickel-copper alloy. Their
use in connection with bearing materials of leaded bronze is common.
Martensitic stainless steels can be suitable for some applications but should be
used with caution for heavily loaded rotating parts, as some have a relatively low
fatigue limit.
Bolts and nuts, particularly where they may have to be removed for
maintenance or replacement, should be in stainless steel or MoneL Bronze or
brass bolting is not suitable.
The usual range of engineering materials from ductile iron castings to alloy
steel castings, from medium carbon to high tensile alloy steels, is used. Their
selection and application are comparable to that found in general engineering
practice, except that factors of safety are generally higher. Grey iron castings arc
not, as a rule, used for stress carrying components in gate installations because
they are liable to brittle fracture.
Certain areas of gates are more susceptible to corrosion than others. It is most
likely to occur at crevices, conjunctions of dissimilar metals, areas where debris
or mud can accumulate, and particularly where water can pond up. Other areas
prone to corrosion, due to the difficulty of applying an adequate protective
coating, are sharp corners, locations of bolts, drain holes and areas which are
difficult to access. Erosion occurs at radial gates where hoisting ropes are
located upstream of the skin plate assembly. To avoid damage to the protective
coating of the skin plate, a stainless steel plate is welded along the contact face.
The water line upstream of radial and vertical-lift gates in free surface flow is
another area vulnerable to corrosion.
The use of closed sections as main structural members, and also as stiffener
members of skin plate assemblies, offers advantages such as reduced paint areas,
reduction of pockets or surfaces where water or silt can accumulate, structural
efficiency and appe-arance. Figure 5. 7(b) illustrates the problem of ensuring that
the underside of the flanges of the Tvscction stiffener beams of a skin plate are
adequately painted in the first instance, and that the areas can be prepared for
repainting.
To prevent moisture penetration and condensation before the sections are
sealed by welding, box section members or other enclosed sections are filled
with dry inert gas or a vapour phase inhibitor. Bolt holes in enclosed sections
should be avoided or, where essential, a structural sleeve should be provided to
maintain the moisture protection of the enclosed areas. Welds must be
continuous; 100% testing to avoid discontinuities and pinholes is advisable.
The selection of the most appropriate paint system is difficult. There are a
l
number of books ,2 and papers3,4 which give guidance, as well as British
5
Standard 5493. The sequence of selection in the BS is given by a table, but to
an engineer who is not a specialist in the subject, it does not assist in choosing
from many combinations of acceptable systems.
A frequent practice is to blast clean to Swedish Standard Sai [z and to apply a
wash coat and repeat applications of epoxy coal tar to 350 flm minimum dry film
thickness. When applying the epoxy coal tar coating under normal ambient
temperatures and normal curing, a polyamide cured epoxy is used. Under low
ambient temperatures, when quick curing is necessary, an isocyanate cured
epoxy coal tar is applied.
Epoxy coal tar paints are available in a restricted range of colours. This can
present a difficulty when painting spillway gates in the tropics. It is desirable to
reduce the solar heat gain on the downstream side of gates by using a white or
aluminium colour coating. The differential expansion of a spillway gate due to
the heating of one face by the sun, and a lower temperature due to reservoi r
water on the upstream side, can result in leakage at the sill seal.
Until some years ago, metal sprayed and thick paint coatings were widely
used on bridges. There is no literature on tbe effectiveness and durability of this
protective treatment for gates.
The large radial automatic gate at the Pulteney sluices in the City of Bath was
zinc sprayed to 250!Lm without subsequent painting. The protection was
effective for approximately 25 years. The difficulty of ensuring adequate and
consistent thickness of a sprayed coat at corners and sharp re-entrant angles
militates against repetition of this treatment.
Materials and
protection
Most gates are assembled with some bolted connections. The mating surfaces
must be protected from corrosion, even if high strength friction grip bolts are
used. The bolts have to be derated due to contact with a painted face. Joints which
have been coated and are assembled with friction grip bolts have also to be derated.
The CIRlA67 and Transport and Road Research Laborarory'' reports offer
guidance. Some gate designers favour seal welding of bolted connections.
High head gates often incorporate enclosed sections. The delta configuration
of the lip of a slide or roller gate is an example. It is possible for oxygen and
moisture to penetrate such enclosed sections. Moisture penetrates by differences
in vapour pressure, and as a result of changes in temperature it can condense and
cause local corrosion. Bolt holes in enclosed sections should be avoided, and
welds, which must be continuous, should be tested to avoid discontinuities
and pinholes.
Paint selection, preparation for painting, tbe testing of painted surfaces and
repair of damaged coatings constitute a specialist subject. The preceding notes
mention some factors pertaining to hydraulic structures and provide references
to literature for further reading.
Cathodic protection
Cathodic protection is the technique of reducing the corrosion rate of an
immersed metallic structure by making the steady-state or corrosion electrical
potential of the metal more electronegative. The thermodynamic considerations
have been dealt with by Shrier,9 Pearson io and others.
\"XThentwo dissimilar metals are electrically connected and immersed in an
electrolyte, which can be fresh or salt water, a current flows through the
electrolyte and the metal so that anions enter the solution from the anode, and
at the same time electrons move from the anode to the cathode via the metallic
connection. The level of corrosion protection depends on the amount of current
flowing, which in turn depends on the electromagnetic force (e.m.f.) and
various ohmic and non-ohmic resistances in the circuit.
Thee.m.f. may be provided by a metal which is more electronegative than the
metal to be protected (sacrificial protection) or by an external e.m.f. and an
auxiliary anode (impressed current protection).
Cathodic protection of gates in sea water or estuarial locations is mainly by
the use of sacrificial anodes of zinc, 11 magnesium or aluminium. Impressed
current cathodic protection is used when corrosion conditions are severe and
where inspection and remedial work during tbe lifetime of the structure are
impossible or impractical.
Cathodic protection is not effective in the splash zone of a gate, and in a tidal
application will afford only reduced protection in the upper tidal zones.
The current density required for steel for adequate cathodic protection in
moving fresh water is 55-65 mAjm2. Instilling basins where the water is highly
turbulent and can contain dissolved oxygen, the range is 55-165 mAjm2.ln sea
water the current density for cathodic protection
is within the range
55-300 mAjm2, whereas in highly polluted estuarine water 600-2000 mA/m2
may be required.
Magnesi um is probabl y the most widely used sacrificial anode material, as the
high current yield ensures maximum current distribution. The addition of
aluminium to magnesium reduces self-corrosion, but minor alloying elements
such as copper, nickel and iron can significantly increase this tendency and
counteract the efficiency of magnesium as a sacrificial anode. Alloying elements
are therefore controlled within limits in magnesium anodes.
Current output is related to the composition of the anodes, surface area and
shape, while the working life is dependent on the ratio of surface area to weight
together with the current demand of water at the gate location.
Although the principles of cathodic protection are essentially simple, its
practical application to the protection of steel structures, such as gates,
immersed in water appears to be more of an art than a science.
Cathodic protection applied to a structure, particularly when applied only to
elements of a structure, can present a danger to adjacent unprotected structures
or parts.
A further application for cathodic protection is the prevention of cavitation
damage. It requires high current densities in order that the hydrogen freely
evol ved from the protected metal can act as a gas cushion between the collapsing
vapour cavities and the metal surfaces. This makes it impractical to protect more
than a limited surface area. \X/here cavitation cannot be avoided, such as in the
area downstream of a high head tunnel gate, it is more economical to provide a
replacement liner.
BS 7361: Part 1 Cathodic protection is a detailed guide for the design of
sacrificial and impressed current protective systems.
Paint coatings used on gates subject to cathodic protection must be
compatible. This should be checked with the paint manufacturer.
References
1. Hudson, J C (1940): The corrosion of iron and steel, Chapman and Hall.
2. Evans, L lR (1960): Thecono.rionandoxidatioliriflllctrtlJ,
Edward Arnold Ltd, Chapter
13.
3. ClRIA (1982): Painti1Jgsteeilvork, editor Haigh,] P.
4. I-IMSO (1971): Report of tbe Committee 011 Corrosion and Protection.
5. British Standard 5493: 1977, Code of practice for protectiue coating rif iron and steel
structures against corrosion. (Note: this standard has been proposed for obsolescence
and has been partially replaced by BS EN ISO 12944 Parts] to 8.)
6. CIRIA (1969): Protection of steel fa:yillg surfaces, editor Day, K J, interim research
report.
7. CIRIA (1980): Desigllguiriallce notes forfrictiongrip
bolted connections, ed itor Cheal, B D,
technical note 98.
8. Black, W; Moss, D S (1968): I-ligb strength friction grip bolts - slij) factors aud protected
/cD/illg surfaces, Transport and Road Research Laboratory, report LR 153.
9. Shrier, L L (1963): Corrosion, section 11, Cathodic protection, George Newness Ltd.
10. Pearson, J M (1955): Fundamentals
of cathodic protection,
in Section VIl,
corrosion protection, The corrosion brJlldbook, editor Uhlig, I-I H, John Wiley and
Sons Inc.j'Chaprnan and Hall Ltd.
11. Day, K J ('1977): Protective treatment, in Proc. i.c.e. Conference Thames Barrier
desigll, 5 Oct., paper 16.
Materials and
protection
Model studies
Numerous model studies of gates and gate installations have been carried out.
Many relate to a particular project and were undertaken to obtain specific
numerical results. These have Limited validity and only show certain
relationships of observed hydraulic parameters within the range of the
experiments undertaken. Some studies have explored problems which are not
specific to an individual installation; extrapolation from data gathered during
empirical investigations has led to the formulation of general guidelines for
other similar cases of flow. They can therefore supplement experience and
enhance understanding of fluid dynamic behaviour.
v, =
V/,j(gd)
where V
flow velocity
gravitational constant
depth/length
g
d
Velocity
Vm
V;,/So.s
Hydraulic
gates and
valves
where the suffixes m denotes model and p denotes prototype conditions
For viscous forces it can be shown by dimensional analysis that the Reynolds
number R, in the model and prototype should be the same. Both the Reynolds
and the Froude numbers for a model and prototype cannot be made equal. Any
difference in the Reynolds number is not of great import as long as both model
and prototype have high values (above 100000) and similar roughness to
diameter ratios. Under these conditions the head loss is a common function of
the square of the velocity in both model and prototype. If the reduced Reynolds
number of the model approaches the point of transition of turbulent to laminar
flow, the laminar flow could occur in the model but turbulent flow would occur
in the prototype. This must be avoided, consequently a minimum operable
Reynolds number has to be chosen. Increasing the velocity to improve
Reynolds number correlation is a technique often used to test for safety margins
that have been eroded by these non-scale effects. A full discussion of the theory
of similarity has been given by Novak and Cabelka, 1
models to Froude
Hydraulic models of gates and gate installations can be divided into three
categories, although the division is arbitrary and categories overlap to some
extent.
A 2D model which can be constructed in a flume is intended for the study of
gate characteristics, and to verify the design of an associated stilling basin or a
weir crest. The objectives of such a model may range from determining the
discharge characteristics of a radial gate, or the hydraulic downpull forces acting
on a vertical-lift gate in a tunnel, to studying the interaction between an
operating gate and a guard gate in a conduit.
An example, Fig. 16.1, is the model study of the spillway gates of the
2
Kotmale Dam in Sri Lanka. The objectives were to determine the gate
Model studies
____ 6___ __
12741
UK the rising sector gates of the Thames Barrier were the subject of
hydrodynamic load and vibration studies. 10,11 Studies to compare model and
P 13
.
prototype results. -, have been undertaken. These, together with the success
of designs which have been tested using models of hydtoelastic similitude,
justify confidence in vibration models.
Some actual or potential vibration problems can be investigated by
constructing a model only to Froude scale. This requires observation of the flow
conditions, coupled with the experience to judge whether these are likely to
cause gate vibration. A similar model strategy can be employed to study possible
conditions of flow separation and reattachment at bottom sections of gates.
Cavitation can be an important cause of dynamic load as well as causing
significant loss of material. Gates have to be designed to be cavitation free, or
to be subjected only temporarily to a low degree of cavitation. To satisfy the
theoretical requirement, model research must be such that model and prototype
vapour pressures occur at equivalent locations. The relationship between the
pressure at any other location in the flow and the pressure in the critical location
2
is given by P V , where p is the density of water and Vis the reference velocity.
Therefore the criterion is that the Thoma number:
(p -
Pv-.pour)
pV2
is correctly reproduced, where p = water pressure and Pvapour = vapour
pressure of air.
This cannot easily be done and requires a special test facility, 14 such as the one
developed by the Snowy Mountains Engineering Corporation in its Fluid
Mechanics Laborarory.!" Small-scale models have been used successfully at
Imperia] College, London, to indicate likely patterns of cavitation and even of
..
. lor
chi tee ements 0f a Iarge structure ."15, 16
cavrtauon
erosion
References
P; Cibelka, J (1981): Models in 0,dmulic engineering - pbysicalJJrincijJles and
design applications, Pitman Advanced Publishing Program.
Milan, D; Habraken, P (1984): Kotmale, report 011 spillJvaJ' radial gates, model tests,
General Technical Services, Lyon (unpublished).
Palmer, M H (1979): NJ,draulic !nodelstudy of the River JVledll'ayjlood relief schemo control
structure, B. H. R.A., repo.rt RR 1572.
Bruce, B A; Crow, D A (1984): Mrica Hydroelectric Project: bydraulic model sttltjy oj tbe
draiodown fit/vert control strnctnre, B.H. R.A., report RR 2325, Nov.
Abelev, AS (1959): Investigations of the total pulsating hydrodynamic load acting
00 bottom
outlet sliding gates and its scale modelling, 8th .I.A .H.R. Congress,
Montreal, paper A 10.
Abelev, J\ S (1963): Pulsations of hydrodynamic loads acting on bottom gates of
hydraulic structures and their calculation methods, 10th l.A.H.R.
Congress,
London.
Kolkman, P A (1976): Flow-induced gate vibrations, Delft Hydraulics Laboratory,
publication 164.
1. Novak,
2.
3.
4.
5.
6.
7.
Model studies
8. Haszpra,
(1979): !vJ()defling~ydroelaslicvi/;ralio/lS,
Pitman Publishing.
9. Kolkrnan, P A (1959): Vibration tests in a model of a weir with elastic similarity on
Froude scale, Rthl.A.I-1.R.
Callgress, Montreal, paper A29.
10. Crow, D A; King, R; Prosser, H J (1977): Hydraulic model studies of the rising
sector gate; hydrodynamic loads and vibration studies, Int. Conference on Thames
Barrier Design, London, Oct.
11. Hardwick, J D (1977): Hydraulic model studies of the rising sector gate conducted
at Imperial College, Tilt. Conference OJ/ Thames Barrier Design, London, Oct., I.e. E.,
London, 1978.
12. Gclccdst, M; Kolkrnan, P A (1963): Comparison of measurements on the prototype
and the elastically similar model of the Hagestein Weir, 10th I.A.l-f.R.
Congress,
London, paper 3.21.
13. Geleedst, M; Kolkrnan, P A (1965): Comparative vibration measurements on the
prototype and the elastically similar model of the Hagestcin Weir under flow
conditions, 11tb 1.A.H.K Congress, Leningrad, paper 4.7.
14 Lcslcighter, E J; Harrison, R D (1981): Development of a cavitation test facility,
Conference, Tllstitliliono/Eligineers,
Australia, Canberra, Mar.
15. Kenn, M J; Garrod, A D (1981).' Cauitatio dOli/age and tbe Tarhela Tunnel collapse of
1974, Proc.Lnstn Civ. Engrs, Part 1,70, Feb: Discussion Proc, Proc. 1m/II Ci. Eltgrs,
Part 1, 1982, 70, Nov.
16. Keno, M J (1983): Cavitation and cavitation damage in concrete structures, Proc.
6th fill. Conference all erosionby liqnid and solid iff/paci, Cambridge, Sept.
Envi ronmental
considerations
Environmental considerations for spillway gate installations differ from those
which apply at gated river control structures. The former are part of a
prominent structure, a dam and the reservoir formed by the dam. The
environmental impact involves the whole system, including changes to the river
feeding the reservoir and the watercourse receiving the discharge from the
spillway. These wider aspects are outside the scope of this book.
Some hydraulic effects due to river and estuary control structures also have
an environmental effect and are briefly outlined in this chapter.
In the UK, rivers were controlled initially for commercial navigation but are
now more frequently treated as an amenity. The position is different in Central
.and Eastern Europe, where river transport plays an important role.
considerations
Navigation
This requires the inclusion of one or more locks. \Vhen locks are located
alongside tbe control structure, extended piers are required upstream and
downstream of the lock to prevent cross-flow and to provide qu.iescent water
conditions when river craft enter and leave a lock. At some structures, mooring
basins are provided for boats and vessels waiting to enter a lock. These
structures have an environmental impact which has to be balanced against
navigation requirements.
Flood management
and maintenance
of river
levels
Fish passes
At many rivers, fish passes for migratory fish are an important part of the
control structure. Criteria for the the design of fish passes is a specialist subject
and have been given by Beach. ) At hydroelectric stations in rivers in the Scottish
Highlands, fish ladders and fish lifts are used. Since the low in fish passes is
continuous, even if all gates at a weir are closed because of low river flow, fish
passes are frequently located at or near the centre of a river barrage or weir.
Migrating fish are attracted to falling water or the turbulent conditions in a
stilling basin due to the discharge of gates. As a result, priority of operation
during low river flow is often given to those gates adjoining a fish pass.
Avoidance
of erosion of the
river bed
Avoidance of river bed erosion due to discharge under or over gates requires
dissipation of the energy of the discharge by a stilling basin, and sometimes
armouring of the river bed downstream of the stilling basin.
278
1
are associated with high river flows or flood events. Thus debris accumulation
upstream of gates will occur during most of the year when river flow is low. To
reduce or eliminate this discharge, gate overflow is required. The bottomhinged flap gate (see Figs 2.25-2.27), is effective in this respect.
Radial gates with overflow section (Fig. 2.1 (b) and vertical-lift gates with
overflow section (Fig. 2.20), will discharge upstream floating debris at low
river lows. Submergible radial gates (Fig. 2.1 (c, operate in the same way.
Accumulation
of floating debris and logs of wood on open structural
members stiffen.ing the skin plate of a gate can be avoided by adopting the
construction used for the radial gates of the Torrumbarry Weir on the River
Murray in Australia mentioned in Chapter 5.
Not all debris will clear in a stilling basin. It becomes trapped by recirculation
due to sheared low, and by the induced hydraulic jump caused by kicker plates
or energy dissipating blocks at the end of tbe stilling basin. The bypass system
illustrated in Fig. 2.27 can assist in clearing floating debris from a stilling basin
when gates are shut. When tbe span of a gate exceeds 3-4 m it will leave dead
areas where debris will remain. Floating oil or beer cans in a stilling basin can
generate noise by repeated impact on the downstream side of gates.
Since most rivers are controlled by a succession of weirs, clearance of debris
at a gate structure results ill its accumulation at the next downstream weir. The
removal of debris from rivers aided by debris booms has met with only limited
success.
Environmental
considerations
iY-;;;'"
gates built early in the 20tb century are often considered visually pleasing and
well integrated with the environment, perhaps because masonry does not have
the same industrial associations as steel.
Figures 17.2-17.4 show three different types of gate of the same aspect area
which have been designed to provide overflow to clear debris and to minimise
the visual impact of the sluice structure.
Vandalism is an important
consideration
in designing river control
structures and protecting existing installations. This requires high fences.
280
1
,.
.ll
Sensitive solutions which effectively impede access to river barrages but do not
clash with environmental considerations arc difficult to achieve.
Barriers
Closing and opening
I
'I
I
of barriers
Closing of barriers can cause surge waves upriver, because on closing a barrier
the natural flow of the river is stopped. If tidal flow upriver is arrested, a surge
wave downriver can be initiated. Similarly, on opening a barrier against a slight
differential head a surge wave is started. The time taken to close or open a barrier
is another critical element. Operational management of a barrier must take into
account the effect on shipping, pleasure craft and, in some cases, people who are
fishing. The method of barrier closure may have to be investigated for different
surges to determine the reflected wave when the barrier is closed.
due to a barrier
Bed protection
Bed protection adjoining the barrier may have to be investigated to take into
account the possibility of failure of a gate to close, causing a scouring action.
Navigation
and shipping
Barrages
Siltation
Siltation can be a problem at barrages. Tidal rivers carry sediment whicb may be
flushed out by the tides. When fluvial transport rates are a consideration, a
morphological model study may be necessary at the design stage.3 Bottomhinged flap gates, which permit only overflow, are selected at barrages to
prevent ingress of saline water. If the low tide recedes from the barrage during
part of the tidal cycle, hook-type vertical-lift gates (Fig. 2.21) enable flushing of
silt to be carried out from the ponded-up watercourse or an upstream bay. This
is effected by raising the lower leaf of the gate, causing discharge under the gate.
\'(fave action may have to be considered at some barrages.
Groundwater
effects at barrages
Water quality
The main function of a barrage is to improve the amenity of the area. This
cannot be achieved if the quality of the impounded water is not acceptable.
The first requirement is to reroute discharges into the river upstream of the
barrage. If saline water is allowed to penetrate the barrage, it can fonn a stagnant
layer at bed level and can create anaerobic conditions commonly associated with
unpleasant smells.
Under prolonged dry weather conditions and low river flows, dissolved
oxygen levels may be depleted, requiring oxygen injection.
Creating a fresh water reach of river or a freshwater lake can provide
conditions for algal growth during some summer conditions.f
Fish migration
Fish passes are incorporated into barrages." Usually migrating fish move from
sea water through brackish water to fresh water. At barrages, the migratory fish
move di rectly from sea water to fresh water and uice versa. What effect this has on
the fish population is not yet fully established.
It is clear that stratification due to presence of some saline water and the
resulting anaerobic conditions would be harmful to fish and fauna.
References
1. Beach, M (-((1984): Fisb pass design> criteria/or the design ami approlial riffish passes and
other structures tafacilitaie tbe passage o/migratolJlfish in riuers. Lowestoft: MAFF, Fish.
Res. Tech. Rep (MAFF Direct Fish. Res Lowestoft), No. 78.
2. rVfcCallum, I R (1996): Navigation aspects of barrage design, Jnt. Confermcebarrages,
mgincering design and enuironmeutal impacts, Cardiff, Sept. in Barrages, editors Burt, N;
Watts, J, John \X/iley and SODS,1996, pp. 465-478.
3. Han, Z C; Shou, W B; Shao, Y Q (1996): Comparison of siltation between
prediction and field data in downstream of tidal barrage, lilt. Conference on barrages
pp.129-137.
4. 1-1 R Wallingford (1994): Tees Barrier IT7eir: imestigation ofoperatingrttles, Report 2943,
Jan.
5. Reynolds, C S (1994): The threat of algal blooms in proposed estuarine barrages,
models, predictions, risks, lilt. Conference 011 barrages, pp. 83-89.
6. Gough, P J (1994): Potential impact of estuarine barrages on migratory fish in
England and WaJes, lnt. Conference 011 barrages, pp. 73-81.
Envi ronmental
considerations
Access
Spillway gate installations and river control structures require stoplogs to
permit temporary closure of the fluidway for emergency situations, gate
maintenance or repair. The slots for placing stoplogs or bulkheads should
provide for adequate working space and the erection of scaffolding between
the gate and the closure structure. At most spillway gates only upstream
stoplogs are required, unlike river control gates where the downstream water
level is frequently above the sill and a second set of stoplogs may be necessary.
At large gates where the discharge under a gate can be at high velocity,
stoplog or bulkhead slots can cause disturbance of the approach flow, such as
eddies and vorticity. Where this is a factor, stoplog slots can be provided with
withdrawable masking plates to produce a smooth face at the pier or abutment.
Gates require inspection and access to parts which have to be serviced or
replaced. On large radial gates, a walkway along the uppermost gate arms
should be provided together with access ladders for inspection for corrosion
of structural members, rope or chain anchorages and seals. It is important that
welds for attachments and brackets supporting access ladders and handrailing
are not located so as to cause stress concentrations at the gate structure.
Gate inspection and side-seal replacement is often carried out with radial
gates in the elevated position. For safety reasons, radial gates should be dogged
GATE
oII-
dogg:rlgmon()l
,---1
Inspection
In practice, gate structures rarely need to be inspected unless there are signs of
distress or significant corrosion. The recommended procedure is to develop an
inspection plan for each different design of gate. For radial gates this is set out in
the US Corps of Engineers specific requirements for inspection of hydraulic
steel structures.' It is not economical to carry out a detailed inspection of all
the gates of a spillway gate installation or a barrage. It should be limited to
critical areas, that is, the elements which are fracture critical and whose failure
would cause collapse or render the gate inoperative.
Fracture critical locations, or areas susceptible to weld-related cracking, may
include trunnion weldments, trunnion beams, gate arms (particularly at their
attachments to the skin plate assembly), and welds attaching lifting brackets.
Intersecting welds are also sometimes subject to stress concentration.
Visual examination is the first method used to inspect all critical elements. If
cracks are suspected, non-destructive test methods should be used, such as dye
penetrant, magnetic particle or ultrasonics.
Inspection for corrosion should be part of the regular maintenance
procedure. Before there is a noticeable breakdown of the paint system, 10Gti
corrosion may have started at crevices, tbe junction of dissimilar metals, seal
clamping plates, location of bolts, drain holes, sharp corners or edges.
2
Guidelines to quantifying corrosion damage are given in Greimann at al.
3
Kumar and Odeh provide information on the corrosive behaviour of stainless
steels under conditions experienced at gates.
o
o
o
o
prevention of rust
resistant to water washout
must be soft to be pumped into clearance of the load zone
mineral-base oil must be used
must have antiwear, antiscuff properties; the recommended grease must
contain soluble antiwear additives, such as sulphur and phosphorous
compounds but no molybdenum disulphide or polytetraOuoroethylene
must have good adherence properties, that is, contain the chemical additives
iso-butylene or polyethelene
non-corrodible towards bronze, low copper corrosion properties
long stable life
will not separate in storage.
Maintenance and
operation of gate
installations
References
1. US Army Corps of Engineers: .Resporlsibililjl Jor hydraulic structures, Specfication
ERll12-02-8157 .
2. G rcimann, L; Stecker, J; Rens, K (1990): Manage1llent systen: [or mitre lock gate.f,
technical report REMR-OM-08,
US Army Engineer Waterways Experiment
Station, Vicksburg, MS.
3. Kumar, A; Odeh, A A (1989): Mecbanical properties and corrosion bebauiours of stainless
steel Jar locks, dams and 0'droelectric plant applications, technical report REMR-EM-6,
US Army Engineer Waterways Experiment Station, Vicksburg, MS.
4. Bureau of Reclamation (1996): Forensic report on spillllJtry gate 3 failure, Folsom 00111,
Bureau of Reclamation, Mid-Pacific Region, Sacramento, California, Nov.
5. Noble, M; Lewin, J (2000): Three cases of gate vibration, Proc. of the Biennial
Conference oj the /3ritish Dan; .)ocielj', Bath, Jun., in Dams 2000, Thomas Telford,
editor Teclcl, P, pp. 95-108.
Appendix
Calculation of hydrostatic
load on radial gates
Gate loads due to water pressure
w
a
e
b
a
g
gate width
height of gate
centre of pressure of hydrostatic thrust on the gate skin plate
upstream water level
downstream water level
height of trunnion
gate opening
horizontal hydrostatic thrust on tbe gate skin plate
vertical hydrostatic thrust on the gate skin plate
resultant of horizontal and vertical hydrostatic thrust passing
through the trunnions
angle of Tit with respect to the horizontal
gravitational constant
Linear units are in metres, forces are in kN.
Note: The vertical force is the water displaced by the gate. The horizontal
and the vertical component of hydrostatic thrust pass through tbe centroid
of the area in question, the resultant in turn passing through the axis of the
trunnions.
The ratio dl R (Fig. A.l) should be equal to or less than 0.707 to ensure that
the sill seal is at an angle of at least 45" to the sill beam. Angles less than 45 can
cause leakage at the sill, which can result in gate vibration.
The force diagrams are valid for gates in the closed position. The forces
are lower when there is discharge under the gate. A more precise
determination
of at least the horizontal component
under discharge
conditions requires an estimation of the discharge characteristics of the gate,
see Hun ter Rouse. 1
"0
.Q
<.9
This is the most severe loading condition. In gates in river installations wben
there is always a downstream water level above the sill it can occur due to
inadvertent drainage of the downstream reach or testing of the gate behind
stoplogs.
TH
e
al
a1
a2
z,
Z2
= 4 (ligW')
kN
.0
,.
_________
\
FigureA .2. Forces due to iuater pressure
290
1
Oil
a ciosedgate
ZI
.1
Z3
Calculation of
hydrostatic load on
radial gates
Z2 - Zl
area 1
A1 = (1/2)(Z3b)
distance AB chord length = 2Rsin(a2/2)
area of triangle ABC = (1/2)R cos(az/2) X 2R sin(a2/2)
Z
= R cos(a2/2)sin(az/2)
area of sector ABC = (1fR2az)/360
area 2
~h = [(1fR2a2)/360][RZcos(a2/2)sin(az/2)]
Tv
vertical hydrostatic thrust = (A 1 + A 2)gL~ kN
Tit
resultant hydrostatic thrust = -J(T~J+ T~) kN
a
tan-I (Tv /TH)
TH2
ez
TH2
2. (b - Go)
~ Wg[(b - Go)2
2
(e7 - Go)TH2
.-
= ~(b
2
- (c -
- Go)2Wg X
Wg [(b 2TH2
2. (c - GotJ~g]
rV"g -
Go)3 _
3
Go)2]
(/1 - 3 Go)
(c - Go)3]
3
_ ~ (c _ Go)zU7g X
2
(c -
Go)
3
+Go
.R sin a4
R sin a6
Z2 - ZI
-J[R2 - (d - c)2]
Z4 - ZI
FiglfrcA.3.
Forces due to
011 a ,gate ill the
open position wit]: drowned
discharge
tuaterpressure
area 1
AI
= (23 + 25) (b - c)
2
cos-1
((d - c)/R)
cos
_I
[Cd - c) - (b - c)] =
R
cos
_I (-
d - b)
R
--
7rR2as
A2 =--360
2
as.
R cos 2
as
SIO
_I
TV2
TH2
Lf::7g
"
C3
-r-[bf - (bl
area 3
./h = 23 -
6TH3
I__I~I
~.
?
(/;1 -
a) 1
a)
~
'"
.;,rn
-.-
_I
bflf;7g
kN
__ c3_1,f;7_g
6
1,f;7g
= J(T~3 + T~3)
kN
TV3
6 [H"
+ Az + A3)gU;7
-TI-13
_ (bl - a)3Wg
_,-
[b1
(b1
a) - c 1
_ITV4
--
TH4
The calculations are an overestimate which ignores the velocity head due to flow
. over the gate.
To refine calculations b2
where bl
/J
b1
(~)
energy head
O.571F J(g)
o
1,f;7(b -
a)
(b - a)3/2
Calculation of
hydrostatic load on
radial gates
i
i
V2
hv
2g
hv
0.564[1
+ 0.150~
a)] TV J(g)
[(b - a)
+ 0.001f/2
Submergible gate
o Use as condition 3. The forces on the gate below the sill are due to the
downstream level and balance.
References
1. Rouse, H (1950): Engineeri11g b),draulics, Chapter Vlll, John Wiley and Sons, New
York.
2. Ackers, P; White, W R; Perkins, J A; Harrison A J M (1978): Weirs and flumes for
flow measurement, John Wiley and Sons, p. 47.
D
Index
access, gate installations, 285-286
added mass, gate vibration, "19"1-194
air demand, conduits, 180-182,201
air supply pipes, submerged outlet gates,
51
articulated vertical lift gates, 20
automatic control methods, 209-216
automatic control systems, choosing,
2"17-218
automatic crest and scour gates, 16-17,63
automatic tilting gates, 27, 29
Barking Creek Barrier, 41,43
Barkley Dam, Kentucky, 189-"190
barrier and barrage gates, 35-48
bottom-hinged buoyant gates, 39-41
bottom-hinged flap gates, 47
caisson gates, 48
environmental considerations, 281-282
flap gates, 41-44
hook-type double leaf gates, 47-48, 64
large span vertical-lift gates, 41
lock gates, 44-47
pointing gates, 48
rising sector gates, 35-39
tidal power barrages, 48, 64
battery powered standby systems, 220-221
bear-trap gates, 50-51, 66
bed protection, 46
Bondi scheme, 30-31
bottom-hinged buoyant gates, 39-41,65
bottom-hinged flap gates, 23-30, 64
debris, 26-27, 97-99
disadvantage, 25
gravity standby systems, 222
nappe oscillation, 29-30
oil hydraulic operation, 120
seals, 134-137, 138
storm surge protection, 47
venting, 29-30
versions, 25
bottom outlet tunnel gares, earthquakes,
256-257
Index
pointing gates, 48
PRA sec probabilistic risk analysis
pressure methods, water level
measurement, 224-225
pressure-reducing valves, 85-87,81, 91
pressure waves, gate vibration, 190
probabilistic risk analysis (PRA), 239
programmable logic control (PLC), 218,
237
proportional control, 211
proportional integral derivative (1'10)
control,210-216
proportional plus derivative (PD) control,
211-212
proportional plus integral (PI) control,
211-212
protection, materials and, 265-269
proximity switches, 147
pseudo-static analysis, 254-255
radial automatic gates, 9-15,63
radial gates, 4-15, 56-58
advantages, 4
arms arrangement, 1
computer program, 14-'15
control weir, 1'1
counterbalance, 13-14
as cuIvert val vc, 8
debris, 107,108
disadvantages, 4
discharge characteristics, 158-162
displaccrsjdisplacer chambers, 11-13
distribution of pressure head, 6
downstream water level control, 13, 14
electromechanical drives, 116-120
embedded parts, 151-155
float operated, 63
floats, 12
gate arms, 103-107
hoist rope attachments, 119-120
hydraulic forces, 5-6
hydrostatic load calculation, 289-294
inlet system, 10-11
instability causes, 14
jamming, 107, 109
malfunction causes, 14
motorised, 63
oil hydraulic operation, 120-126
piers, dividing, 14
toller faces, 151-155
seals, 132-134
side guide rollers, 137-140
side seal contacts, 151-155,259
stiffening, 103-107
structural design, 103-107
trunnions, 13, 103-104, 142, 144-145,
146
1.
Index
Second Edition
JACK LEWIN
This book written by the leading expert in the field, is the definitive reference work on gates and valves. Based on
Professor Jack Lewin's extensive practical experience, this fully revised edition includes new sections covering
hazard and reliability of gates and valves, earthquake effects on gates and operating machinery, environmental
impact and aesthetics, and maintenance. Previous sections on the trunnions of radial gates, seals, ice formation, gate
operation and structural design have been extended. Llkewise, the chapter on types of gates has been expanded to
include new types of water-operated automatic gates, rolling weir gates, fuse gates and barrier gates and their
details.
This essential resource describes the principal options available to engineers and designers and outlines the main
advantages, disadvantages and potential problems associated with each type of gate and value .
Professor Jack Lewin is visiting professor of hydraulics at City University and has worldwide experience of the design
and project management of hydropower schemes and flood prevention. Professor Lewin has lectured throughout the
world on this subject and has been an extemal examiner for the Open University. He has twice won the Chartered
Institution'of Water and Environmental Management's Herbert Lapworth Medal for an outstanding contribution to the
art and science of water engineering and has also won the Marshall Nixon Prize.
SECTION
THROUGH
1"'----
11
r-;
..
j
1"
... - ... --
-_ .. _-
..
ISBN D-7277-299D-X
I780727
1111111729903
I" I
......
'"I Thomas Telford
RISING
SECTOR
GATE
1[b.