9

HYDRAULIC STRUCTURES

9.1 List the principal advantages and disadvantages of gravity, arch and butress dams, and give an
indication of where each type may be used. What type of spillway are typically used with these dams?

GRAVITY DAMS

Advantages
1. Earth-rock or roller compacted concrete dams are relatively simple to construct and, in the correct location,
economical as well.
2. Can be constructed from whatever material is available locally (eg rock, clay, waste from industry - such as
waste products from china clay extraction in Devon and Cornwall) or from concrete.
3. Gravity concrete dams can be built almost anywhere, provided the foundation is strong enough and the
materials are available.
Disadvantages
1. Overtopping of earth dams is dangerous, so they must be designed to ensure this does not happen.
2. Earth dams are generally restricted to less than 150 m in height.
3. Conventional concrete dams are relatively expensive, and the large volumes of concrete required can cause
problems with heat generation and cracking during construction.
Where used
1. Rock or earth fill dams are suited to wide shallow valleys. Their wide base means they can be used where the
foundation material is relatively weak.
2. Concrete gravity dams can be constructed almost anywhere, but when over 20 m high require a relatively
strong foundation.
Spillway types
1. Earth/rock dams must not be overtoped and so usually do not have a spillway on the crest. Therefore chute
spillways or shaft spillways are commonly used.
2. There are no such problems with concrete dams, so almost any type of spillway is possible, including
overflow, chute and shaft spillways. Some dams may also have crest gates. Sometimes the topography of the site
may lend itself to a particular spillway type.

ARCH DAMS

Advantages
1. Use perhaps only 20% of the concrete required for a gravity dam.
2. May be the only option where a very high dam is required, provided the location is suitable, of course.
Disadvantages
1. Technically complex, being thin, curved structures. Some have a double curvature (ie in plan and section).
2. Sudden failure is possible if things go wrong (eg foundation failure).
Where used
1. Where there is a steep sided, narrow valley or gorge.
2. Where the rock foundation is strong enough to withstand the high stresses imparted by the arching action.
Spillway types
1. Almost any type of spillway is possible, including overflow, chute and shaft spillways. Sometimes the
topography of the site may lend itself to a particular spillway type.

BUTRESS DAMS

Advantages
1. Use only about 60% of the concrete required for a gravity dam.
2. Their reduced weight means they can be built on weaker foundations.
Disadvantages
1. May not be much cheaper than a concrete gravity dam if the formwork is complex etc.
2. Depending upon base width and height, may still require a relatively strong foundation.
Where used
1. Almost anywhere, but relatively narrow valleys may be desirable.
2. Where the foundations material is relatively strong and can carry the weight.

3rd ed, Understanding Hydraulics 57 © Les Hamill 2001, 2011

Spillway types
1. Almost any type of spillway is possible, including overflow, chute and shaft spillways. Some dams may
incorporate crest gates. Sometimes the topography of the site may lend itself to a particular spillway type.

9.2 Look at Tables 9.3 and 12.14, which list some notable dam incidents and spillway design standards,
and the associated text. What message does this give to engineers and politicians with respect to dam
safety?

Don’t make the mistake of thinking that a 1 in 10 000 year flood is unlikely to occur, or will occur in 10 000
years’ time.
Don’t underestimate the design flood and its probability of occurring in the very near future. Equation (12.11)
and Example 12.10 show that there is a very real risk of failure. This is often not understood by some engineers,
never mind politicians or the public!
Don’t underestimate the size of spillway required. Around 30% of dam failures may be attributable to inadequate
spillways. Some dams are over 100 years old. How do they cope with any increased flood magnitude and
frequency arising from global warming? New dams may still be around in 200 years. How can be they be
designed to allow for this?
Don’t underestimate the need for good management and routine maintenance. Of the dam failures that occurred
after 1900, about half were due to overtopping. The possibility of overtopping can be reduced if the reservoir
level is kept low (to provide additional storage and freeboard) during the wet season. About 21% of these
overtopping failures resulted from problems with the operation of spillway gates. Therfore, it may be wise to
design spillways that operate automatically without moving mechanical parts.
Don’t gamble with public safety and people’s lives. There has already been too many failures.
Arguably, there should be well established evacuation procedures in case things do go wrong; the public should
understand these procedures and know where to go to reach safety.

9.3 A 4.0 m wide vertical sluice gate is positioned in a horizontal, rectangular channel of the same width.
The gate operates freely and must pass a discharge of 15.0 m3/s without the upstream head exceeding a
value of 3.5 m. (a) Use equation (9.2) and Fig 9.13 to determine the height at which the gate should be set
to give an upstream depth of 3.5 m. (b) What is the approximate depth of the jet at the vena contracta?
(c) Assuming an energy head loss through the gate of 0.05V22/2g and that α1 = α2 = 1.05, check the answer
from part (b) using the energy equation. (d) If the normal depth in the channel is 1.95 m, confirm that the
gate is actually discharging freely.

(a) First iteration: assume C = 0.54 in equation (9.2), which is:

Q = C aO(2gH1)1/2
15.00 = 0.54 aO (19.62 × 3.50)1/2
Thus aO = 3.35 m2
Y = 3.35/4.00 = 0.84 m.
This gives H1 /Y = 3.50/0.84 = 4.2 and from Fig 9.13 C = 0.54 as assumed above, so no further iterations are
needed.
(b) It is often assumed that the coefficient of contraction ( CC) is about 0.6 so this would give the approximate
depth of the jet at the vena contracta as H2 = 0.6 × 0.84 = 0.50 m.
(c) As a check, using the energy equation:
H1 + α1V12/2g = H2 + α2V22/2g + 0.05V22/2g
Assume initially that H2 = 0.500 m, as in (b) above then:
V2 = 15.00/(0.50 × 4.00) = 7.500 m/s and α2V22/2g = 1.05 × 7.502/19.62 = 3.010 m.
Energy head loss = 0.05V22/2g = 0.05 × 7.5002/19.62 = 0.143 m.
Now V1 = 15.00/(3.50 × 4.00) = 1.071 m/s and α1V12/2g = 1.05 × 1.072/19.62 = 0.061 m.
Substituting the above values into the energy equation gives:
3.500 + 0.061 = 0.500 + 3.010 + 0.143
3.561 ≠ 3.653
The energy heads upstream and downstream are not equal. A trial and error solution is needed. To reduce the
velocity head downstream, try H2 = 0.508 m.
H1 = 3.500 m and α1V12/2g = 0.061 m, as before.
V2 = 15.00/(0.508 × 4.00) = 7.382 m/s and α2V22/2g = 1.05 × 7.3822/19.62 = 2.916 m.
Energy head loss = 0.05V22/2g = 0.05 × 7.3822/19.62 = 0.139 m.
Substituting these values into the energy equation gives:
3.500 + 0.061 = 0.508 + 2.916 + 0.139
3.561 ≈ 3.563

3rd ed, Understanding Hydraulics 58 © Les Hamill 2001, 2011

This is near enough, so H2 ≈ 0.508 m, roughly the same as in part (b). A small change to CC , α or the assumed
energy head loss would eliminate the discrepancy.
(c) From equation (8.32), the critical depth in the channel is:
DC = (Q 2/gB 2)1/3 = (15.02/9.81 × 4.02)1/3 = 1.13 m
Thus H2 < DC so the jet is supercritical and so likely to be discharging free.
Also, if DN = 1.95 m then VN = 15.00/(1.95 × 4.00) = 1.92 m/s and FN = 1.92/(9.81 × 1.95)1/2 = 0.44 so from
equation (8.36) the depth at which a hydraulic jump will be initiated (D1) is:
D1 = (DN/2)[(1 + 8FN2)1/2 − 1] = (1.95/2)[(1 + 8 × 0.442)1/2 − 1] = 0.58 m.
Since the depth at the vena contracta is H2 = 0.50 m < 0.58 m a jump will form some distance downstream of the
vena contracta, so the gate will operate freely.

Had the jet been submerged then the above solution would not be valid: a new value of C corresponding to
submerged flow would have to be obtained from Fig 9.13 and an iterative solution obtained.

9.4 A vertical underflow sluice gate 5.5 m wide discharges 18.2 m3/s into a channel of the same width. The
gate is set 0.90 m above the bed (Y), the downstream depth at the gate is 1.3 m and the normal depth in
the channel is DN = 2.4 m. (a) Confirm that the gate is operating in the submerged condition. (b) Use
equation (9.2) and Fig 9.13 to determine the upstream depth.

(a) Assume DN = 2.40 m = the depth after a hydraulic jump.

VN = 18.2/(2.40 × 5.5) = 1.38 m/s so FN = V2 /(gD2)1/2 = 1.38/(9.81 × 2.40)1/2 = 0.28
The sequent depth to DN is D1 where from equation (8.36)
D1 = (DN /2)[(1 + 8FN2)1/2 − 1]
= (2.40/2)[(1 + 8 × 0.282)1/2 − 1]
= 0.33 m
Thus a jump can form only if the initial depth (D1) is 0.33 m (or less). Since the depth downstream of the gate is
1.3 m a jump is clearly not possible and the gate is submerged. This is probably quite obvious since the critical
depth in the channel is DC = (Q 2/gB 2)1/3 = (18.22/9.81 × 5.52)1/3 = 1.04 m, so the flow downstream of the gate is
above critical depth.
(b) The velocity of the flow underneath the gate is VO = 18.2/(0.90 × 5.5) = 3.68 m/s and FO = VO/(gY)1/2 =
3.68/(9.81 × 0.90)1/2 = 1.24. The value of DN/Y = 2.4/0.90 = 2.67, so Fig 9.13 gives C = 0.45 and so:
Q = C aO (2gH1)1/2 (9.2)
18.2 = 0.45 (0.90 × 5.5) (19.62 × H1)1/2
H1 = 3.40 m

9.5 A bridge has 9 round nosed masonry piers each 2.5 m thick. They are equally spaced in a river
channel that is 127 m wide. The normal depth in the channel is 2.1 m when the discharge is 530 m3/s.
(a) Use Yarnell’s equation to calculate the afflux caused by the piers. (b) Check the answer in (a) using
the d’Aubuisson equation. (c) If the piers have a length to width ratio of 4, what is the potential scour
depth at the pier if the flow hits the piers at an angle of 0°° and 30°°?

(a) For round nosed piers Table 9.6 gives KY = 0.90, DN = 2.1 m, VN = 530/(2.1 × 127) = 1.987 m/s, FN =
1.987/(9.81 × 2.1)1/2 = 0.438, the total width of the piers = 22.5 m, the channel width B = 127 m, so the total
width of the bridge openings b = (127 − 22.5) = 104.5 m, and the channel contraction ratio
m = (1 − b/B) = (1 − 104.5/127) = 0.177
H1* = KY DN FN2 (KY + 5FN2 − 0.6)(m + 15m4)
= 0.90 × 2.1 × 0.4382 (0.90 + 5 × 0.4382 − 0.6)(0.177 + 15 × 0.1774)
= 0.363 × 1.259 × 0.192
= 0.088 m
(b) A trial and error solution is necessary. Assuming H1* = 0.088 m as above then D1 = 2.188 m and
V1 = 530/(2.188 × 127) = 1.907 m/s. With m = 0.177, interpolating the values in Table 9.6 gives KA = 1.10 so:
Q = KA b DN (2gH1* + V12)1/2
530 = 1.10 × 104.5 × 2.1 (19.62H1* + 1.9072)1/2
2.196 = (19.62H1* + 3.637)1/2
4.822 = 19.62H1* + 3.637
H1* = 0.060 m
Now V1 = 530/(2.160 × 127) = 1.932 m/s and V12 = 3.733 so:
4.822 = 19.62H1* + 3.733
H1* = 0.056 m

3rd ed, Understanding Hydraulics 59 © Les Hamill 2001, 2011

With V1 = 530/(2.156 × 127) = 1.936 m/s and V12 = 3.748 then:
4.822 = 19.62H1* + 3.748
H1* = 0.055 m
Thus from parts (a) and (b) it would appear that the afflux H1* is about 0.055 m to 0.088 m.
(c) Very approximately, from Table 9.4 the potential scour depth for round nosed piers is about 1.5bP = 1.5 × 2.5
= 3.75 m. With a length to width ratio of 4, a 30° angle of attack would double the scour depth so that it becomes
7.50 m.

9.6 The design discharge for a culvert is 13.8 m3/s when the normal depth in the channel is 1.9 m. The
river channel has a bankful width of 4.2 m, a slope of 1 in 300 and a Manning’s n of about 0.040 s/m1/3.
The maximum permissible upstream depth is 2.2 m (ie 2.5 m allowing a 0.3 m freeboard). The culvert will
have a length of 35 m. Design a single barrel, rectangular culvert that will operate with outlet control,
following the general procedure and example given in the text. The site is environmentally sensitive so
allow for the invert being 0.15 m below existing bed level, thus enabling a layer of stones and gravel to be
provided on the channel bottom.

Maximum permissible upstream depth, H1 = 2.2 m.

Make the slope of the culvert barrel SB = the natural slope of the stream SO = 1 in 300.
Length of culvert = 35 m.
Say the width of the culvert barrel = the bankfull width BS = 4.20 m.
Assume the tailwater depth = the channel’s normal depth = 1.90 m.
Assume the depth of flow in the culvert barrel, DB = 1.90 m.
Velocity of flow in the barrel, VB = 13.80/(4.20 × 1.90) = 1.73 m/s (0.75 < 1.73 < 2.00 m/s so OK).
Make culvert height Y = tailwater depth + culvert freeboard + 0.15 m = 1.90 + 0.60 + 0.15 = 2.65 m.
Use a rectangular concrete box section 4.20 m wide × 2.70 m high.
From equation (8.32), the critical depth in the culvert is:
DC = (Q 2/gB 2)1/3 = (13.802/9.81 × 4.22)1/3 = 1.03 m.
Check tailwater level: 0.5 (DC + Y) = 0.5(1.03 + 2.55) = 1.79 m, which is less than 1.90 m, so use 1.90 m. Here Y
was taken as 2.70 − 0.15 = 2.55 m to allow for the stone layer on the bottom.
Again, to allow for the bed of gravel and stones, say the composite bed/concrete roughness of the barrel is nB =
0.040 s/m1/3 (probably too high, but on the safe side).
RB = AB/PB = (1.90 × 4.20) / (4.20 + 2 × 1.90) = 1.00 m.
Estimate the headwater level from equation (9.15):
HWL = H2 + 1.5VB2/2g + VB2nB2L/RB4/3
= 1.90 + (1.5 × 1.732/19.62) + (1.732 × 0.0402 × 35/1.004/3)
= 1.90 + 0.23 + 0.17
= 2.30 m
From equation (9.17), upstream depth, H1 = HWL − SBL
= 2.30 − (1/300) × 35 = 2.18 m. (< 2.20 m so OK)
Submergence of the entrance will not occur until H1 > 1.1Y = 1.1 × 2.55 = 2.81 m. Since H1 = 2.18 m
submergence is impossible or unlikely. As in Example 9.5, a check calculation could be conducted to see at what
discharge the culvert entrance would submerge, but it is not really necessary in this case. If it was not for the
restriction on upstream depth there would be an argument for reducing the size of the barrel.

9.7 (a) With respect to a broad crested weir, what is meant by the ‘submergence ratio’ and why is it
important? (b) Water flows along a rectangular channel in which there is a broad crested weir with a
horizontal crest. The channel is 9.0 m wide, and upstream of the weir the depth of flow measured from the
channel bed is 1.1 m when the discharge is 8.24 m3/s. Ignoring any loss of energy, what is the minimum
height of the weir that will allow it to function with critical depth on the crest? (c) If a broad crested weir
has a coefficient of discharge, C, of 1.65, and if it completely spans a 17.4 m wide rectangular channel,
what would be the head over the weir when the discharge is 6.8 m3/s.

(a) The submergence ratio of a weir is defined as the ratio of the head of water above the crest on the
downstream side of the weir (HD) to the upstream depth (H1), that is HD /H1. It is important because if the
submergence ratio is exceeded (by HD becoming too large) the critical flow on the crest will be ‘drowned out’.
Under these conditions the discharge equation (such as equation 9.19 and 9.20) will no longer be applicable,
since they are derived by assuming critical flow on the weir crest.
(b) Using the same notation as Fig 9.27 in the text:
V1 = Q/A1 = 8.24/(9.0 × 1.1) = 0.832 m/s

3rd ed, Understanding Hydraulics 60 © Les Hamill 2001, 2011

For a rectangular channel DC = (Q 2/gB 2)1/3
DC = (8.242/9.81 × 9.02)1/3
DC = 0.440 m
VC = (gDC)1/2 = (9.81 × 0.440)1/2
VC = 2.079 m/s
With reference to Fig 9.27, ignoring energy losses the height of the weir, p, is obtained from equation 9.21:
V12/2g + D1 = VC2/2g + DC + p
(0.8322/19.62) + 1.1 = (2.0792/19.62) + 0.440 + p
0.035 + 1.1 = 0.220 + 0.440 + p
p = 0.475 m
Thus at this discharge the minimum height of the weir for critical depth on the crest is 0.475 m.
(c) Equation 9.20 is: Q = C b H13/2 where C is the weir coefficient and b the length of the crest, which in this case
equals the width of the channel (17.4 m). If Q = 6.8 m3/s and C = 1.65 then:
6.8 = 1.65 × 17.4 × H13/2
H13/2 = 0.237
H1 = 0.383 m
With reference to Fig 9.26, the head over the weir is 0.383 m when the discharge is 6.8 m3/s.

9.8 (a) List the advantages and disadvantages of a throated flume compared to a broad crested weir when
used to measure the discharge in an open channel. (b) A flat bed throated flume is to be constructed at a
position in a rectangular open channel where the normal depth is 0.55 m, the channel width is 7.5 m, the
channel slope is 1/250 and the channel has a Manning n value of 0.035 s/m1/3. Assuming that the flume has
a typical coefficient of discharge of 1.65, what is the maximum throat width that will still induce critical
flow in the flume?

(a) Advantages compared to a broad crested weir.

(i) Causes a relatively small obstacle to flow, so
(ii) causes a relatively small loss of energy, and hence
(iii) causes a relatively small afflux.
(iv) The streamlined shape of a flume does not trap floating debris or silt.
(v) Easily constructed and robust.
(vi) Requires little maintenance.
(vii) Has a good (relatively high) modular limit (submergence ratio).
Disadvantages compared to a broad crested weir.
(i) A flume is not very suitable for wide river channels or large discharges.
(ii) It may be difficult to design a flume to cover a range of flows, whereas it is relatively easy to design
a compound weir, which is why they are quite common.
(iii) It may be difficult to design a flume so that critical depth occurs in the throat with a standing wave in
the diverging portion.
(iv) A flume may be more difficult to construct because of its shape in plan (but at the same time (v)
above is also true!).
(b) Starting with the normal depth, DN , of 0.55 m which occurs in the unconstricted rectangular channel.
A = 0.55 × 7.5 = 4.125 m2
P = 7.5 + 2 × 0.55 = 8.6 m
R = A/P = 4.125/8.6 = 0.480 m
The discharge can be calculated from the Manning equation with n = 0.035 s/m1/3 and SO = 1/250 or 0.004.
Q = (A/n) R2/3 SO1/2
Q = (4.125/0.035) × 0.4802/3 × 0.0041/2
Q = 117.857 × 0.613 × 0.0632
Q = 4.569 m3/s
The flume will submerge (and cease to be accurate) at the modular limit when DD/D1 = 0.75, where DD is the
downstream depth and D1 is the upstream depth. Assuming that the downstream depth DD is the normal depth
(because the upstream depth will be increased by the afflux) then DD = 0.55 m. Thus the corresponding
minimum depth upstream is DD/0.75 therefore:
D1 = 0.55/0.75 = 0.733 m
This is the value of the upstream depth measured above the flat bed of the flume as used in the discharge
equation (equation 9.22), which is:
Q = C bC D13/2 where from above Q = 4.569 m3/s and C is given as 1.65.
4.569 = 1.65 × bC × 0.7333/2
4.569 = 1.035 × bC

3rd ed, Understanding Hydraulics 61 © Les Hamill 2001, 2011

bC = 4.412 m
The maximum width that will induce critical flow in the throat is 4.412 m (the channel is 7.5 m wide).

3rd ed, Understanding Hydraulics 62 © Les Hamill 2001, 2011