HEAD LOSS CHARACTERISTICS OF FLAP

GATES AT THE ENDS OF DRAIN PIPES

J. A. Replogle, B. T. Wahlin
ABSTRACT. Flap gates are commonly used at the end of pipe drains and pump outlets to prevent backflows of water and entry
of small animals. Flap gates are relatively inexpensive, with low maintenance costs, but can trap debris in their hinge systems.
Many texts refer to studies performed on flap gates at the University of Iowa in 1936, which may be limited in value because
they are for specific lightweight gates. More recent studies in England have attempted to generalize the characteristics
for pinhinged flap gates and place the Iowa studies into a broader perspective. The flapgate backpressure effect on a gate
with free outfall appears to be small. However, under certain submerged conditions, a flap gate will increase the upstream
backwater levels. This may be critical in sewerage situations and some surface land drainage cases. Recently, at least one
manufacturer sells a rubbercoated steel gate cover with a flexure hinge made of the same rubber material. While it reduces
the opportunity for trash to catch, such as may happen on a pinnedhinge type of pivot, this rubber hinge arrangement
essentially becomes a springloaded gate with the force of closure due to both the weight of the gate and the elastic properties
of the hinge. Users have questioned whether this arrangement introduces significant backpressure. We therefore tested a
rubberhinged flap gate to verify whether these gates fit into the general pattern of the limited previous studies on pinhinged
gates, which is to exhibit a continuous decrease in backpressure with increasing flow rate, and hence gate opening. The design
information for pinhinged flap gates is also updated to make it more readily available for design uses. The rubber hinge
resulted in a slight deviation towards more head loss, approximately 3 mm (0.1 in.) at larger gate openings compared to about
1 mm (0.025 in.) for pinhinged gates, which is attributed to the flexure strength of the rubber hinge. Thus, for freeflow outlet
applications, flap gates of either the pinnedhinge or the flexure style add small head losses that amount to about 1% to 2%
of the pipe diameter. These studies and the review of previous work allow users to evaluate whether flap gates of either the
pinned hinge design or the rubberized flexure design can cause detrimental backpressure on a drainage system under free
outfall situations.
Keywords. Drain outlets, Drainage, Drainage design, Flap gates.

lap gates are commonly used at the ends of pipe

drains and pump outlets to prevent backflows of
water and entry of small animals. Large sizes are
frequently found in tidal areas to reduce inflows
during high tides and permit outflows during low tides. Other
uses are to prevent flood flows from an upstream storm from
backing into lowlands during the passing of the flood flow.
Under these conditions, the gate closes under the influence of
its own weight and the hydrostatic pressure from the
downstream side. When the water levels on the downstream
side recede, the gate reopens and flow can again drain to the
lowered receiving waters. These installations are relatively
inexpensive, and maintenance costs are low. Malfunctions
can occur when debris lodges in the gate opening or in the
pinned hinges that are common to many types of flap gates,
requiring regular inspections.

Article was submitted for review in December 2002; approved for

publication by the Soil & Water Division of ASAE in April 2003.
The authors are John A. Replogle, ASAE Member Engineer,
Research Hydraulic Engineer, U.S. Water Conservation Laboratory,
Phoenix, Arizona; and Brian T. Wahlin, WEST Consultants, Inc., Tempe,
Arizona, formerly Civil Engineer, U.S. Water Conservation Laboratory,
Phoenix, Arizona. Corresponding author: John A. Replogle, U.S. Water
Conservation Laboratory, 4331 E. Broadway, Phoenix, AZ 85040; phone:
6024371702; fax: 6024375291; email: jreplogle@uswcl.ars.ag.gov.

When properly maintained, flap gates can minimize

backflow into drainage systems for most combinations of
drainage flows and receiving stream water levels, even
though the gate is expected to introduce some constriction on
the flow that translates into an increase in upstream head. The
effect on free outfall situations should be small (SCS, 1973).
However, under certain conditions, when the receiving water
level has receded to the point of being slightly lower than the
previously protected lowland, a flap gate in the system will
maintain a slightly increased upstream backwater level. This
may be critical in sewerage situations and some surface land
drainage cases. Just how much this increase will be at various
levels of submergence, and at levels including free outfall,
has been studied in only a few instances, and guidance is not
well disseminated.
At least one manufacturer (PlastiFab) offers a rubbercoated steel gate cover with the hinge made of the same
rubber material, making a total gate thickness ranging from
1.5 to 3 cm (0.75 to 1.2 in.) depending on the basic pipe
diameter. Besides reducing the opportunity of trash to catch
on a pinned hinge type of pivot, this rubber hinge arrangement acts as a springloaded gate, with the force of closure
due to both the weight of the gate and the elastic properties
of the hinge. While the existing information appears to
support the idea that only small backpressure effects are
generated from any flap gate (SCS, 1973; Burrows and
Emmonds, 1988), users continued to question whether this

Transactions of the ASAE

Vol. 46(4): 10771084

2003 American Society of Agricultural Engineers ISSN 00012351

1077

was still true with the elasticloaded closure. We therefore

tested a flap gate on a pipe that was 20.3 cm (8 in.) in diameter
to verify whether these gates fit into the general pattern of the
limited previous studies.

BACKGROUND

Information on flapgate characteristics is not abundant.

Many texts refer to flap gates or tide flaps, including the
classic book on sanitary engineering by Babbitt (1922 and
subsequent revisions, including Babbitt and Baumann, 1958)
and a more recent book by Linsley and Franzini (1979).
Pethick and Harrison (1981) presented a theoretical treatment of rectangular flap gates. In 1936, the Hydraulic
Laboratory of the (then) State University of Iowa conducted
a series of tests to determine the head lost by water
discharging through ArmcoCalco flap gates (currently
offered by Hydro Gate as model 10C flap gates). The gates,
which were 46, 61, and 76 cm (18, 24, and 30 in.) in diameter,
were supplied from commercial stock (Armco, 1978).
Burrows and Emmonds (1988) pointed out that the Iowa
tests may be limited in value because they were for specific
lightweight gates, and that Armco suggested that heavy
gates may cause more head loss than their reported values of
15 mm (0.6 in.) for a 0.305 m (12 in.) flap gate and 110 mm
(4.3 in.) for a 2.134 m (84 in.) flap gate. The terms light and
heavy were not well defined. These values were extrapolated and interpolated by the Soil Conservation Service
(SCS) (now the Natural Resources Conservation Service) in
their internal Engineering Handbook Series as design
guidance. Portions of that handbook, which include this
guidance, are reproduced and quoted in SCS (1973, Section 16, Chapter 7, pp. 726). Burrows and Emmonds (1988)
cautioned that the Iowa tests applied only to the gates tested
and may not be relevant for general application.
Armco (1978) presented some quantitative headloss
information on gates that it manufactured, again based
primarily on the Iowa studies. More recently, Burrows et al.
(1997) reported on a British study with the objective to
estimate flow rates based on flap gate opening. The analysis
presented fit the data in a general sense, but values differed
on the order of 20% to 30%. The most consistent data were
for free discharge and no downstream submergence. The
study also did not report the backpressure effects directly or
the head losses that were experienced. Head losses for
submerged gates were reported for two model gates by
Burrows and Emmonds (1988).
Most or all of these flap gates have been the simple
hingepin type with the hinge pin in the plane of the pipe end
and gateface contact plane. Burrows et al. (1997) consider
the effects of the hinge being offset slightly upstream from
this contact plane in their study of discharge ratings. Such an
offset hinge mechanism is used in a weightedgate system to
control water levels in upstream irrigation canals (Raemy and
Hager, 1998; Burt et al., 2001).

THEORETICAL CONSIDERATIONS

The flow situation from a flap gate is complicated, so a

detailed theoretical approach is difficult. A dimensional
analysis approach was presented by Burrows and Emmonds
(1988), but it introduces more variables than equations and

1078

HL

Hinge

Tailwater
Z

Total energy line

Flap gate attached

Total energy line

Flap gate removed
v / 2g
2

LH

Flap
gate

Average velocity
V=Q/A

Pipe
Gate
weight
force
(W)
Figure 1. Definition sketch (modified from Burrows and Emmonds, 1988).

cannot be directly applied. Burrows and Emmonds also

conducted laboratory tests on two models with 80 mm
(3.15 in.) and 135 mm (5.32 in.) gates of the configuration
shown in figure 1, where most of the variables used herein are
defined.
Burrows and Emmonds (1988) collected their data under
slightly submerged conditions, and the data showed an
increase in head loss due to the flap gates. Interestingly, the
smaller flap gate exhibited about 30% more head loss than the
larger gate. They attribute this increase to the relative length
of the gate pivot arms (LH /D), which is the distance from the
gate center to the hingepin center (LH ) divided by the gate
diameter (D) (fig. 1). Because the hinge arm is nearly the
same length for both gate sizes, the relative length (LH /D)
varies from 0.71 for the larger gate to 0.85 for the smaller
gate, or about 14% increase in LH /D from the large to the
small gate. Thus, the change in head loss (30%) seems to be
about twice the change in LH /D.

QUALITATIVE CONSIDERATIONS OF THE

PROBLEM

For a flap gate, the backpressure effect at low heads in the

pipe should be noticeable because the pressure component is
used to force open the gate. Even relatively heavy gates will
swing under relatively small pressures. For example, an iron
gate about 1.25 cm (0.5 in.) thick with a mass of 15 kg (33 lb.)
covering the end of a 38 cm (15 in.) pipe at an end angle of
1 horizontal to 10 vertical causes only a 15 N (3.3 lbf)
component force as a capping pressure. The flap gate would
start to open at an average overpressure of 0.390 kPa
(0.0566 psi), or 39 mm (1.6 in.) of water head, which is
readily supplied by a partly filled horizontal pipe. If,
however, a large flow exists and the pipe velocities become
large, then the jet forces support the gate by the deflection of
the jet, and the backpressure from the gate becomes even less
noticeable. Thus, for a freeoutfall flap gate at its higher
range of flow rates, the energy to open the gate is borrowed
from the downstream deflected jet, and little or no upstream
backpressure is detectable. This qualitatively explains the
behavior shown in figure 2, which was reconstructed for SI
units from SCS (1973) for light flap gates with free outfall.
As mentioned, the term light flap gate is not well defined
but was used to refer to the Iowa selection of a commercial

TRANSACTIONS OF THE ASAE

0.2

0.4

Head loss (in.)

0.6
0.8
1.0

1.2

1.4
100
50

D = 122 cm (48 in.)

D = 107 cm (42 in.)

1000

D = 91 cm (36 in.)
D = 76 cm (30 in.)

10

D = 61 cm (24 in.)

D = 51 cm (20 in.)
D = 46cm (18 in.)
D = 41 cm (16 in.)
D = 36 cm (14 in.)
D = 30 cm (12 in.)

100

1
0.5

D = 25 cm (10 in.)
D = 20 cm (8 in.)

0.2

10
0

10

15
20
25
Head Loss (mm)

30

35

Flow through flap gate (1000 gpm)

Flow through flap gate (L/s)

10000

40

Figure 2. Head loss for light flap gates with free outfall (after SCS, 1973).

flap gate designed to withstand a backpressure head of 3 m

(10 ft) of water. Note in figure 2 that as the flow rate for a
given pipe size increases, the head loss decreases.
An equation that approximates figure 2 is:
HL
1 gD5
=
D 176 Q2

(1)

where
g = acceleration due to gravity (9.81 m/s2, or 32.16 ft/s2)
HL = head loss (m, or ft)
D = pipe diameter (m, or ft)
Q = discharge rate (m3/s, or cfs).
Equation 1 was not derived directly from theory but rather
from curves fit to the diagrams presented in the SCS
reference. This equation does not account for changes in gate
weight. However, it does illustrate the small backpressure
that can be expected from the addition of a flap gate to the end
of a drainage pipe with a free outfall.
Burrows and Emmonds (1988) indicated that the added
head loss from a flap gate was reflected in the angle of the
gate opening regardless of how it is achieved, by gate weight,
spring, or manually applied force. This result was for
submerged gates, where the submergence was to the
elevation of the pipe top, Z/D = 1 (fig. 1).

LABORATORY VERIFICATION

Laboratory studies were conducted to verify the above

qualitative inferences. The flap gate shown in figure 3 was
attached to the outlet end of a 20 cm (8 in.), Schedule 40
plastic pipe. The flap angle was about 15 from vertical. The
flap gate opening was slightly larger and elliptical, 23 cm
(9 in.) wide by 24 cm (9.5 in.) high. The pipe had piezometer
taps at 30.5 cm (1 ft) intervals, with extra taps at 15.25 cm
(6 in.) and 45.7 cm (18 in.) from the flap gate. The flow rate
varied from about 35 to 85 L/s (1.25 to 3 cfs), and flow
velocities ranged from 0.9 m/s (3 ft/s) to about 2 m/s (7 ft/s)
in the flap gate outlet section.
Because the previous tests were for light flap gates
mounted vertically with pintype hinges, we attempted to

Vol. 46(4): 10771084

Figure 3. Flap gate with rubber flexure hinge.

verify whether the elastic hinge and the mounting at about

15 from vertical for the test gate behaved similarly.
Expectations were that the sloping face would affect the
initial low flow values because the weight component of the
gate would inhibit opening. As noted, the gate would need to
be very heavy to overcome the full pipe static head before
flow begins.
The changes in backpressure readings were small, typically only 1 to 3 mm (0.025 to 0.1 in.), and challenged the limits
of the standard vertical manometer system that was readily
available. However, a large number of readings tend to
reduce random error, so simple averages of readings from the
20 manometers were used. Slopeintercept comparisons did
not fare better than the simple averages. Both methods
produced similar data variation.
The pressure increase as read from the piezometers was
noted for the flap gate in action and for the flap gate lifted out
of the flow. This difference was interpreted as the increase in
backpressure head due to the gate system being in place, and
hence the additional head loss. Extra weights, either one or
two, were fastened to the gate in the form of circular plates
of steel 23 cm (9 in.) in diameter, each weighing 2 kg

1079

(4.46 lb). The rubberized gate weighed about 2.8 kg (6.2 lb),
including some small brackets used to hold the extra weights.
These plates and the gate produced an 18 N (4 lbf) horizontal
force on the end of the pipe, which was sloped back toward
the flow at about 15 from vertical (fig. 3). As the gate swings
upward, increasing horizontal force is needed, but eventually
the dynamics of the flow, namely the impact of the flowing
water and not the hydrostatic forces alone, sustains the gate.
This combination of forces varies with the angle of the gate
opening. The flexure of the rubber hinge causes an added
force, again varying with the opening, of about 5.5 to 10 N
(1.25 to 2.25 lbf), which further complicates the system.

DISCUSSION OF RESULTS

The starting horizontal force needed to open the gate was

about 18 N (4 lbf), plus that needed to overcome the flexure
force, which was near zero in the closed position. This
opening force at zero flow must come from the static pressure
distributed over a flap gate area of 387 cm2 (60 in.2). This
force is readily provided by the static head of the horizontal
pipe, when filled, because the gate then has 20 cm (8 in.) of
head against it with hydrostatic distribution, offering a
potential static force of about 45 N (10 lbf).
In figure 4, the angle of the opening produced similarly
small head losses, for the gate alone and with the addition of
one added weight, shown in the figure as a dimensionless
ratio to the velocity head, hv = V2/2g. Adding the second
weight produced a significant increase in additional head loss
(fig. 4). The date groups in figure 4 are fitted with power
curves to compare with the Burrows and Emmonds (1988)
curve. The freeoutfall flap gate appears to have a threshold
weight below which the backpressure changes slowly or not
at all.
This increase in head loss is interpreted as exceeding the
threshold weight for the behavior of a socalled light flap
gate for this pipe diameter, and it indicates that the gate
weight when two weights were added exceeded the ability of
the hydrostatic pressure in a nearly nonflowing pipe to open
the gate. However, we noted above that 45 N (10 lbf) are
available and that only 18 N (4 lbf) are required to initially
move the gate.
With the gate weight alone and with a high flow velocity,
making the ratio of HL /hv low, the head loss approaches zero.

Conversely, when the gate weight is large and the velocity

head is low, backpressures increased significantly.
An explanation offered for why the head loss is greatest for
low flow rates is that the hydrostatic pressure is used to hold
the gate open until the velocity head can accomplish the task
in freeoutfall gates. The velocity head in these tests varied
from 5 to 24 cm (0.2 to 0.8 ft). Thus, the velocity head value
approaches that of the flap gate diameter. It appears that when
the velocity head is of a magnitude that approaches or
exceeds the pipe (gate) diameter, the additional head loss
transmitted upstream approaches zero.
Also shown in figure 4 is an approximated average of
Burrows and Emmonds (1988) results for submerged gates.
Here, the gap angle is wider, probably due to the buoyancy
forces on their light model gates, and the head loss values are
greater, probably due to submerged conditions. There is no
threshold static force due to the pipe being full and at rest, as
with the freeoutfall situations tested, because the tail water
equalizes the pressure on both sides of the flap gate. The static
head forces that are needed to open the gate, even with
reduced force due to gate buoyancy, do not develop until
there is a head increase between the upstream waters and the
receiving waters. Being submerged, the velocity head energy
cannot be borrowed from the downstream trajectory, and
the turbulent flow losses around the gate flap are manifested
as detectable backpressure in the upstream pipe system. The
tradeoff between static pressure and velocity head, while it
should still occur, affects the upstream backwater, hence the
submerged flow curve does not approach zero (fig. 4). Rather,
the indication is that the head loss (hL ) can vary from about
one velocityhead (HL = V 2/2g) in the pipe at low flows to less
than onetenth velocityhead at high flows. No head loss
comparisons with gate angle or submerged flow were
available for the SCS curves.
Burrows and Emmonds (1988) presented dimensionless
plots of their results in terms of the dimensionless grouping
WD2/Q2 vs. angle of gate (fig. 5), where W is the gate
weight, D is the pipe diameter, r is the fluid density, and Q
is the discharge rate. All of the data, regardless of gate weight,
fell approximately on the same curve for the flap gate we
studied, which had an LH /D ratio of 0.5. The Burrows and
Emmonds (1988) curves are for flap gates with extended
hinge arms of LH /D = 0.71 and 0.85 (see fig. 1). The
progression is not continuous among the three curves. One

1.2
1.0

HL/ h v

0.8

Added two weights plus

rubber flexure force
(free outfall)

0.6
0.4

Added one weight plus

rubber flexure force
(free outfall)

0.2

Burrows and Emmonds

(1988) submerged gate

Gate weight plus

rubber flexure force
(free outfall)

0.0
0

10

20

30
40
50
Angle of gate opening (degrees)

60

70

Figure 4. Head loss (HL ) relative to velocity head (hv ) as a function of flap gate opening and gate weight.

1080

TRANSACTIONS OF THE ASAE

3
Gate plus
two weights

2.5

Burrows and Emmonds (1988)

L H /D = 0.85 (submerged)

Gate plus
one weight

o Gate only

WD

/r Q

Burrows and Emmonds (1988)

L H /D = 0.71 (submerged)

1.5
1

Flexure and weights

L H /D = 0.5 (free outfall)

0.5
0
0

10

20

30

40

50

60

70

80

Angle of gate opening (degrees)

Figure 5. Gate weight (W), diameter (D), and discharge rate (Q) related to gate angle.

would expect that for a given flow rate, the longer arm would
produce a smaller gate angle to achieve a required orifice
opening. This indeed is indicated in figure 5 for the
submerged flap gates of Burrows and Emmonds (1988), but
the shorter arm of our test gate does not follow this pattern.
Discharging around the gate into free air instead of under
water is one possible explanation. Adding to that is the
expectation that buoyancy forces of the submerged gates
would increase the angle over that of unsubmerged conditions.
The angle of the gate opening as a function of the
dimensionless grouping of figure 5 might be expected
because the angle defines a quasiorifice opening, and for
orifice flow, the discharge rate is directly proportional to the
area of the orifice opening but only to the square root of
differential head.
In mathematical form, if a flap gate follows the behavior
of an orifice, then the discharge (Q) is a function of the orifice
area (A) and the differential head (h), or:
Q = CA 2gh

(2)

where C is a constant. If we take the log of both sides of the

equation and the derivative of the equation, we obtain:
dQ dA 1 dh
=
+
Q
A 2 h

(3)

Thus, increases in Q can be accommodated by combined

increases in A and h. The value of A is variable because the
gate opens in response to increases in h, but changes in h are
half as effective as changes in A. Moreover, A increases
further from the impact of the velocity head (hv = V2/2g)
caused by the increasing flow (Q). This means that the area
(A) contributes more toward accommodating increases in Q
than does h. In addition, h is gradually replaced by the
velocity head (hv ) as the primary driver in opening the gate,
which reduces the value of h. Making the gate heavier can be
tolerated up to a threshold value, and then the flap gate angle
for that particular flow rate changes when the impact force
from hv requires additional help from h (figs. 4 and 5).
Figure 6, in dimensionless format, hopefully would have
unified flap gates in terms of gate weight force, diameter,
flow rate, and head loss. It obviously does not. The gate

Vol. 46(4): 10771084

weight (W) is the vertical force with no component forces

calculated directly. The pivot arm length does not change for
our gates when extra weights are added. Again, the
smallerweight condition and noaddedweight condition
behave similarly. The second extra weight appears to indicate
that backpressure can be transmitted upstream from freeoutfall flap gates.
Figure 7, with discharge (Q) plotted against additional
head loss (HL ), indicates that the additional head loss is about
the same regardless of the flow rate, despite data variation for
the weighted gate runs. This differs from the results for
pinhinged gates shown in figure 2 (replotted here as
pinhinged gate), which show decreased head loss with
increasing discharge. The tested rubberized gate is assumed
to be of lighter weight than the pinhinged gates of figure 2.
The test gate points would be expected to be to the left of and
parallel to the pinhinged gate curve, had the gate been
pinhinged itself. That most of the points appear to the right
of, and do not parallel, the pinhinged gate curve is attributed
to the changing flexure force of the rubber hinge. The wider
the angle of opening caused by the discharge rate, the more
back force is generated by the flexure. This tends to maintain
the additional loss to a more or less constant value throughout
the discharge range, and it does not appear to parallel the
pinhinged gate curve until higher flow rates. Again, the idea
of exceeding a threshold weight is supported by the wide
spacing between the plotted points for one weight added and
two weights added.
Figure 7 also shows that added weight increases backpressure, or additional head loss (HL ). The 18 N (4 lbf) force
needed to initially unseat the heaviest test configuration
requires a static head of about 5.6 cm (2.2 in.) of water depth
on the gate face. However, recall that 45 N (10 lbf) are
available as the pipe becomes full. Thus, the gate opens
before the pipe is full, and velocity head conversion starts.
Likewise, for the smaller weight, 12.7 N (2.86 lbf), or 4 cm
(1.6 in.) of head are needed before the gate starts to open.
Neither of these values is reflected in any direct way in
figure 7, and no consistent pattern with gate weight is readily
apparent. However, figure 7 supports the Armco (1978)
suggestion that heavy gates may cause more head loss than
their reported values. It also suggests that when the pipe must
be nearly full to open the gate, then backpressure could

1081

Test gate: 2.80 kg plus flecture

W = 3.43 to 3.68 kg
D = 22 cm

SCS gates: assumed 1 cm steel,

dia. = 5 cm larger than pipe size of
W = 4 kg, D = 20 cm, free outfall

Test gate: 4.83 kg plus flecture

W = 5.41 to 5.56 kg
D = 22 cm

WD 2 /Q

1
Burrows and Emmonds
(1988) submerged gate

Test gate: 6.85 kg

plus flecture
W = 7.4 1 to 7.57 kg
D = 22 cm

0
0.0

0.1

0.2

HL

0.3

0.4

0.5

V 2 / 2g
Figure 6. Dimensionless plots of flow parameters for several flap gates.
100
Rubber flexure
plus two weights

Rubber flexure

80

Q (L/s)

60
Rubber flexure
plus one weight

40

From eq. 1: D = 20 cm
pinhinged gate

20
0
0

H L (cm)
Figure 7. Comparison of pinhinged gates to gate with flexure hinge and with added weight.

increase beyond the value for light flap gates. This idea of
requiring a full pipe to start the opening of the gate could
serve as the separation between heavy and light gates.
However, the opening force is also a function of the slope
angle of the flap gate mounting, which complicates the
definition.
OTHER CONSIDERATIONS
In an agricultural drainage system in which the drain pipes
usually flow partly full, the effects of adding a light flap
gate such as the one tested are nearly undetectable. In
submerged situations, an increase in upstream backwater can
be expected, but the effect should be small, reflected more as
an increase in time to drain the final increment of water than
on the final drainage level, because the gates seldom seat
tightly without significant backpressure from downstream.
Flap gates can be installed so that they remain underwater, as
recommended by Armco (1978), to slow corrosion that
repeated wetting and exposure to air can cause. In addition,
floating debris is more of a problem for gates that alternate
between free and submerged flow. Debris clogging is of less
concern on freely discharging gates.

1082

Another consideration is the extrapolation of these results

to other pipe sizes. Standard hydraulic modeling laws
indicate that the results will reliably scale to length ratios of
at least 10, if the length dimensions of all parts are similar.
EXAMPLE APPLICATION
Situation
A lowlying, level, 4 ha (10 acre) farm field is surface
drained through a levy berm with a 20 m (65 ft) length of
20 cm (8 in.) smooth plastic pipe into a creek with a
nonflooding water surface only 50 cm (20 in.) below the
field surface (fig. 8). The pipe bottom outlet is only 0.25 m
(10 in.)above the stream surface. The inlet is depressed at the
field end with a flaredcone inlet structure that reduces inlet
losses, so that the hydraulic grade line is affected primarily
by the change in velocity head. The pipe is installed
horizontally through the berm. Frequent flows from upstream
can temporarily inundate the outlet to 1.5 m (5 ft) deep for up
to 12 hours and cause flooding in the cropped field, even
when there are no local rains. It is suggested that a flap gate
be attached to the pipe outlet.

TRANSACTIONS OF THE ASAE

20 cm
(8 in.)
25 cm
(10 in.)

Field

Stream

Figure 8. Example drainage situation.

Required
Design (select) a standard, light, hinged flap gate for
this pipe outlet. Comment on the increase in backpressure
that is calculated. How might this affect the field to be
protected?

Thus, the rubberized flap gate can be used in this example

with advantage. Protecting the area from back flooding
leaves only local rainfall. Even a relatively heavy rain of
10 cm (4 in.) should be drained within a day after the stream
flow recedes.

Discussion and Solution

Standard hydraulic calculations, using 1.5 m (5 ft) of head
on smooth pipe, indicate that the pipe could carry approximately 113 L/s (4 cfs, or 4 acreinches per hour) at the
maximum stream depth over the pipe outlet of 1.5 m (5 ft).
Under preflapgate conditions, the field could receive about
36 acreinches backwards through the pipe, or 9 cm (3.6 in.)
on the field. This would later drain at a slower rate when the
receiving stream lowered.
Using standard pipe flow calculations again, the drainage
flow would only have about 30 cm (1 ft) of head instead of
1.5 m (5 ft) and would flow at only about 51 L/s (1.8 cfs). This
would more than double the time that was needed to inundate
the field. However, depending on the subsoil water table,
much of this flooded water could be expected to infiltrate in
less than a day. Thus, the drainback time may approach, or
be less than, the inundation time of 12 hours. Local rains
would aggravate this on the field. A flap gate would limit the
drainage and field inundation to local rainfall amounts minus
infiltration.

Submerged Flow
The drainage starts when the flap gate is slightly
submerged and the field water level is still higher than the
stream surface. In this case, the flow does not start until the
gate buoyancy weight is balanced by the slight overpressure
from the fieldlevel water. The net force moment on the
2.8 kg (6.2 lb) rubberized gate (ignoring buoyancy) mounted
at 15 must be 27.5 N (6.2 lbf) acting about the hinge at 2.73
cm (1.075 in.), or 75 Ncm (6.66 lbfin.) must be countered
by a uniform overpressure (P) in the pipe acting at the center
distance of 10 cm (4 in.) from the hinge (the hydrostatic
forces and hydrostatic distributions inside and outside the
gate cancel), or a moment of P 10 cm (4 in.), or P = 7.4 N
(1.67 lbf) distributed over an area of 325 cm2 (50.2 in.2), or
an overpressure of 2.3 cm (0.92 in.) head differential before
the gate cracks open and a small flow begins.
The assumed heavier gate of figure 6, weighing 4 kg
(8.8 lbs), would increase the backpressure to about 3 cm
(1.2 in.) of water head. Thus, the increased backpressure at
opening for either the light pinhinged gate or the
rubberhinged gate would be on the order of 2.5 cm (1 in.) of
water head and would decrease with gate opening for either
gate.

Standard Flap Gate Design Selection

From figure 2 or equation 1, D = 0.2 m (0.66 ft or 8 in.),
Q = 0.051 m3/s (1.8 cfs, or 808 gpm), and g = 9.81 m/s2
(32.2 ft/s2) yields 0.0014 m = 1.4 mm (0.0047 ft = 0.056 in.)
head loss. This is judged to be insignificant, and the gate can
be added with negligible effect on the drainage time.
Because the gate is frequently submerged, the hinge could
be subjected to trash entanglements. Thus, a rubberized flap
gate might be indicated to avoid trash accumulation.
Estimate the backpressure that can be expected in the pipe
and the influence this rubberized flap gate might have on the
draining times from the field.
From figure 7, the backpressure increase from a rubberized flexure hinge is about 3 mm (0.1 in.) of water head at
about 50 L/s (800 gpm) flow rate. This is nearly double that
for the pinhinged gate but is still low enough not to
materially affect drainage rates.

Vol. 46(4): 10771084

CONCLUSIONS

Rubberhinged flap gates added to the end of a drain pipe

do not create significant flow restrictions in the usual
operating ranges for horizontal pipes discharging freely.
When discharging into air, the influence is about 3 mm
(0.1 in.) of additional head loss for an operating range of
discharges from 35 to 85 L/s (1.25 to 3.0 cfs) in a 20 cm (8 in.)
pipe.
Head loss in pinhinged gates is inversely proportional to
discharge rate. Rubberflexure hinges tend to cause a nearly
constant head loss regardless of flow rate. However, the
resulting backpressure is still small, on the order of 1% of the
pipe diameter.

1083

The additional head loss due to adding a flap gate to a

freely discharging pipe is minimal because the energy to open
the gate at higher flow rates is taken from the trajectory
energy of flow, essentially after the flow has exited the pipe,
and hence little or no backpressure is detected.
Previous studies by others indicate that submerged gates
cause greater head loss than freely discharging gates, which
may be a factor in lowland drainage situations. This loss is
only roughly characterized, and may be on the order of 0.1 to
1 times the velocity head (hv ) for high to low flow rates,
respectively.
The angle of the flap gate opening is a function of
WD2/Q2. The angle of opening for a given discharge and
flap gate is expected to behave as an orifice, meaning that the
change in the area of the opening is directly proportional to
the discharge rate but only to the square root of head change.
Because the static pressure is directly proportional to head,
and the velocity forces are proportional to the square of the
velocity, it appears that a change in area is more readily
accomplished than a change in differential head, and thus the
head becomes inversely proportional to discharge.
These studies and the review of previous work allow users
to evaluate whether flap gates, of either the pinhinged or
rubberhinged design, can cause detrimental backpressure
on a drain outlet system under free outfall and submerged
outlet situations.

1084

REFERENCES

Armco. 1978. Handbook of Drainage and Construction Products.

Armco Drainage & Metal Products, Inc. (Out of print, summary
of Iowa results currently available on web
site:www.hydrogate.com).
Babbitt, H. E. 1922. Sewerage and Sewage Treatment. 7th ed,
1953. New York, N.Y.: John Wiley and Sons.
Babbitt, H. E., and R. Baumann. 1958. Sewerage and Sewage
Treatment. 8th ed. New York, N.Y.: John Wiley and Sons.
Burrows, R., G. A. Ockleston, and K. H. M. Ali. 1997. Flow
estimation from flapgate monitoring. J. Chartered Institution of
Water and Environmental Management (U.K.) 11(5): 346356.
Burrows, R., and J. Emmonds. 1988. Energy head implications of
the installation of circular flap gates on drainage outfalls. J.
Hydraulic Research 26(2): 131142.
Burt, C. M., R. Angold, M. Lehmkuhl, and S. Styles. 2001. Flap
gate design for automatic upstream water level control. J.
Irrigation and Drainage Eng. 127(2): 8491.
Discussion/closure in 128(4): 264265.
Linsley, R. K., and J. B. Franzini. 1979. Water Resources
Engineering. 3rd ed. New York, N.Y.: McGrawHill.
Pethick, R. W., and A. J. M. Harrison. 1981. The theoretical
treatment of the hydraulics of rectangular flap gates. Subject
B(C), Paper No. 12 in Proc. IAHR XIX Congress, 247254.
Madrid, Spain: International Association of Hydraulic
Engineering and Research.
Raemy, F., and W. H. Hager. 1998. Hydraulic level control by
hinged flap gate. Proc. Inst. Civil Eng. Water Maritime and
Energy 130: 95103.
SCS. 1973. Drainage of Agricultural Lands. Published by Water
Information Center, Inc. (Reformatted and reprinted from
Section 16, Drainage of agricultural land, in the National
Engineering Handbook issued in 1971 by the USDA Soil
Conservation Service.)

TRANSACTIONS OF THE ASAE