BIE 5300/6300 Assignment #9

Open-Channel Transition Design

23 Nov 04 (due 2 Dec 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

Given:

• The design flow rate is 20.0 m3/s

• The upstream trapezoidal section side slopes of m = 1.25
• The upstream trapezoidal section bed width is b = 3.40 m
• The downstream rectangular section has b = 2.75 m
• The bed slope of the upstream trapezoidal section is 0.000220 m/m
• The bed slope of the downstream rectangular flume is 0.00332 m/m
• The upstream and downstream channels are concrete-lined
• The length of the transition will be L = 10.0 m
• Uniform flow will prevail upstream & downstream of the transition

Required:

1. For uniform flow, do you expect subcritical flow both in the trapezoidal and
rectangular sections?
2. Specify the rate of change of bed width with distance, using a 3rd-degree
polynomial, through the transition.
3. Specify the rate of change of side slope with distance through the transition.
4. Determine the elevation of the bed (invert) of the transition versus distance along
the transition so that the energy line has a constant slope through the transition,
matching the upstream and downstream uniform flow depths.
5. Make sure the water surface through the transition has a constant, uniform slope.
6. What is the total bed elevation change across the transition?
7. Show your results graphically, with a side view and a plan view of the transition.
 BIE 5300/6300 Assignment #10
Drop Spillway Design
2 Dec 04 (due 7 Dec 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

Given:

• The design flow rate for an earthen canal is 120.0 cfs.

• There is a 7.5-ft drop in canal bed elevation at a location in the canal.
• The average base width of the canal is 10 ft, and the average inverse side slope
is 2:1, for a trapezoidal cross section.
• The canal longitudinal bed slope is 0.000125 ft/ft upstream of the drop location,
and downstream of the drop location it is 0.000129 ft/ft.
• The average water temperature in the canal is 65°F.
• Assume a Manning roughness of 0.018 for the earthen canal.
• Assume uniform flow conditions at the design flow rate in the downstream
channel.
• Forty percent of the canal flow rate is delivered to agricultural water users, and
the rest goes to municipal water users.

Required:

1. Design a drop spillway for the given conditions at the location of the 7.5-ft drop.
2. Use English units for the design.
3. Use the design procedure given in the lecture notes, but iterate to make the
basin area (b x L) as small as feasible at the design flow rate; however, if
possible, do not make the basin width, b, greater than the average base width of
the earthen canal.
4. Add 10% to the upstream normal depth for freeboard, determining the height of
the headwall at the upstream sides of the stilling basin.
5. Produce side view and plan view technical drawings of the drop spillway,
indicating the dimensions of the energy dissipation structure.
A Design Solution:

I. Uniform-Flow Depths

• From the ACA program, the following uniform-flow depths were found:

• Upstream normal depth: 3.91 ft

• Downstream normal depth: 3.88 ft

both of which are for the design flow rate of 120 cfs.

II. Minimum Stilling Basin Area

• The stilling basin width is limited to a maximum of b = 10.0 ft in this problem.

• Start at this limit and decrease b incrementally to find the minimum area (L x b).
• Set up calculations in a spreadsheet, using the equations from the lecture notes.

b hc yend ht 2.15hc Eq. 14 ydrop xf xt xs xa xb xc L Lxb

(ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (sq ft)
10.0 1.648 0.659 4.540 3.542 OK -8.159 7.555 5.232 8.013 7.784 1.318 2.883 11.985 119.850
9.5 1.705 0.682 4.563 3.665 OK -8.182 7.702 5.337 8.167 7.935 1.364 2.983 12.282 116.680
9.0 1.767 0.707 4.588 3.800 OK -8.207 7.861 5.450 8.334 8.098 1.414 3.093 12.605 113.442
8.5 1.836 0.734 4.615 3.948 OK -8.234 8.034 5.572 8.515 8.274 1.469 3.213 12.956 110.130
8.0 1.912 0.765 4.646 4.110 OK -8.265 8.223 5.706 8.712 8.467 1.529 3.346 13.342 106.739
7.5 1.996 0.798 4.679 4.291 OK -8.298 8.429 5.854 8.928 8.679 1.597 3.493 13.768 103.260
7.0 2.090 0.836 4.717 4.493 OK -8.336 8.657 6.016 9.166 8.912 1.672 3.657 14.241 99.685
6.5 2.196 0.878 4.759 4.721 OK -8.378 8.911 6.198 9.431 9.171 1.757 3.842 14.770 96.004
6.0 2.316 0.926 4.807 4.979 invalid -8.426 9.195 6.401 9.728 9.462 1.853 4.053 15.367 92.205
5.5 2.454 0.982 4.863 5.277 invalid -8.482 9.518 6.633 10.064 9.791 1.963 4.295 16.049 88.272
5.0 2.615 1.046 4.927 5.623 invalid -8.546 9.887 6.898 10.450 10.168 2.092 4.577 16.838 84.188

• It is noted that b ≤ 6.0 ft violates Eq. 14.

• In the above table, b = 6.5 ft gives the lowest basin area (96 ft2).
• More precisely, b ≈ 6.41 ft gives the minimum value without violating Eq. 14:

b hc yend ht 2.15hc Eq. 14 ydrop xf xt xs xa xb xc L Lxb

(ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (ft) (sq ft)
6.50 2.196 0.878 4.759 4.721 OK -8.378 8.911 6.198 9.431 9.171 1.757 3.842 14.770 96.004
6.49 2.198 0.879 4.760 4.725 OK -8.379 8.916 6.201 9.437 9.177 1.758 3.846 14.781 95.930
6.48 2.200 0.880 4.761 4.730 OK -8.380 8.922 6.205 9.442 9.182 1.760 3.850 14.792 95.855
6.47 2.202 0.881 4.762 4.735 OK -8.381 8.927 6.209 9.448 9.188 1.762 3.854 14.804 95.780
6.46 2.205 0.882 4.763 4.740 OK -8.382 8.932 6.213 9.454 9.193 1.764 3.858 14.815 95.705
6.45 2.207 0.883 4.764 4.745 OK -8.383 8.938 6.217 9.459 9.199 1.766 3.862 14.826 95.630
6.44 2.209 0.884 4.765 4.750 OK -8.384 8.943 6.221 9.465 9.204 1.767 3.866 14.838 95.555
6.43 2.212 0.885 4.766 4.755 OK -8.385 8.949 6.225 9.471 9.210 1.769 3.870 14.849 95.480
6.42 2.214 0.886 4.767 4.760 OK -8.386 8.954 6.229 9.476 9.215 1.771 3.874 14.861 95.405
6.41 2.216 0.886 4.767 4.765 OK -8.386 8.960 6.232 9.482 9.221 1.773 3.878 14.872 95.330
6.40 2.218 0.887 4.768 4.770 invalid -8.387 8.965 6.236 9.488 9.226 1.775 3.882 14.883 95.254
 • Choose a stilling basin width of b = 6.5 ft (rounding up to the nearest half foot).
• Then,

• hc = 2.20 ft
• xa = 9.17 ft
• xb = 1.76 ft
• xc = 3.84 ft
• L = 14.77 ft
• ht = 4.76 ft
• yend = 0.88 ft
• ydrop = -8.38 ft

• Some of the above values could be rounded up, but in this design they will
remain as calculated (other dimensions will be rounded, as shown below).

III. Headwall & Wingwalls

• Adding 10% to the upstream normal depth (as specified), the headwall height
should be 1.1(3.91) = 4.30 ft above the origin, which is at the crest height.
• There will need to be a converging section at the stilling basin inlet because the
stilling basin width is less than the upstream channel width. This is given a 45-
degree convergence, as shown in the plan view drawing (see below).
• According to design procedures, the wingwall height at the end sill is to be 0.85hc
= 0.85(2.20) = 1.87 ft above the tail water surface.
• The side walls should slope linearly from the headwall to the beginning of the
wingwalls over the length, L, of the stilling basin.
• Also according to design procedures, the wingwalls splay out at 45 degrees, and
the tops slope downward at 45 degrees.
• Extending the wingwalls to intersect with the base of the downstream channel
side slopes, the length of each wingwall will be:

2
⎛ 10.0 − 6.5 ⎞
L wing = 2⎜ ⎟ = 2.47 ft
⎝ 2 ⎠

• Round this up to 2.50 ft.

• Riprap and or other measures to help prevent erosion at and just downstream of
the wingwalls will also be necessary to complete this design.

IV. Floor Block Sizing & Spacing

• This design will follow the guidelines in which the floor blocks have the same
width and length, which is equal to:
 Lblocks = 0.5hc = 0.5(2.20) = 1.10 ft

• Floor block height will be:

yblocks = 0.8hc = 0.8(2.20) = 1.76 ft

which in this design will be rounded down to 1.75 ft.

• Number of floor blocks for 50% occupation of the stilling basin width:

6.5
N= = 2.95
2(1.10)

which in this design will be rounded up to three blocks.

• Equal block spacing across the stilling basin width gives:

6.50 − 3(1.10)
block spacing = = 0.80 ft
4

• Finally, the proportion of the stilling basin width occupied by the three floor blocks
will be:

⎛ 3(1.10) ⎞
occupied width = 100 ⎜ ⎟ = 50.8%
⎝ 6.50 ⎠

which is in the recommended range of 50 – 60%.

V. Footings & Other Details

• Footing depth: 2.00 ft.

• All concrete work to be steel reinforced.
• Concrete floor and wall thicknesses: 5 inches.
• No longitudinal sills required.

VI. Drawings

• Side and plan views: see below.

 2.20 ft 14.77 ft 1.77 ft

upstream 4.30 ft

1.87 ft

10.93 ft 3.84 ft
downstream

2.00 ft
1.10 ft

Side View

1.10 ft
10.00 ft

6.50 ft

upstream 0.80 ft downstream

Plan View
 BIE 5300/6300 Assignment #11
Siphon Spillway Design
7 Dec 04 (due 10 Dec 04)

Show your calculations in an organized, neat format. Indicate any

assumptions and or relevant comments.

Given:

• An overflow spillway needs to be designed for a location along a canal.

• The concrete-lined canal has a maximum flow rate of 4.0 m3/s.
• The concrete lining depth is 2.3 m, and the depth to the top of the earthen berms
is 2.5 m.
• The canal bed elevation at this location is 239 m above mean sea level (msl).
• The maximum water surface elevation in the open drainage channel on the downhill
side of the canal at this location is 238.5 m.
• The canal cross section is trapezoidal in shape.
• The canal bed width is 1.65 m, and the inverse side slopes are 1.5.
• The longitudinal bed slope is 0.000284 m/m.

Required:

1. Use metric units for the design.

2. Use the Chezy equation with C = 60.
3. Design a siphon spillway for the given conditions.
4. Assume a Cd value of 0.67 when applying the orifice equation for the siphon
spillway capacity.
5. Refer to section 4-14 to 4-16 of the USBR book, and to the lecture notes.
6. Determine the following:

a. What is the depth of flow in the canal at the maximum discharge? Take this as
FSL (full supply level).
b. What siphon spillway crest elevation (referenced to msl) do you recommend?
c. What is the estimated available head, H, across the siphon spillway?
d. Is H < hatm, where hatm is mean atmospheric pressure head?
e. What are the barrel dimensions (b and D), given a rectangular barrel section
(according to USBR guidelines, D ≥ 2.0 ft)?
f. What are RCL, RC and RS?
g. What is the estimated full-pipe unit discharge, q, in the siphon spillway?
h. What is the estimated maximum unit discharge through the siphon spillway
(“vortex” equation)? Make sure this is more than the discharge calculated from
the “orifice equation,” otherwise the design is not acceptable.
i. What is the minimum required vent (siphon breaker) pipe inside diameter?
j. What is the required height of the outlet deflector sill (1.5D)?
 k. What is the required height of the outlet ceiling, h2?
l. What is the hydraulic seal in the canal, above the top of the opening to the
siphon spillway, on the downhill canal back at Qmax? Is it greater than the
minimum values of 1.5hv +0.5 ft, or 1.0 ft (whichever is greater)?
m. Create a side-view drawing of your design, with Fig. 4-17 as a model.

Solution:

a) Normal depth in the canal

By iteration, the normal depth is found to be hn = 1.289 m at 4.0 m3/s, with C = 60.
This is taken to be the full supply level (FSL).

b) Elevation of siphon spillway crest

By USBR design guidelines, the crest elevation is 0.2 ft (0.061 m) above FSL.
Then, the crest elevation is: 239.000 + 1.289 + 0.061 = 240.350 m above msl.

c) Available head

The available head, H, is measured from the crest elevation to the downstream
water surface elevation (see Fig. 4-17). Taking the maximum downstream water
surface elevation: H = 240.350 – 238.500 = 1.850 m.

d) Is H < hatm?

The mean atmospheric pressure head at an elevation of 239 m is estimated as:

hatm = 10.3 − 0.00105(239.00) = 10.05 m

Then, H < hatm, which is as required for siphonic operation.

e) Barrel dimensions

Try the minimum recommended barrel height, D = 2.0 ft (0.610 m). The width, b,
will be determined below.

f) RCL, RC and RS

Use the recommended ratio RCL/D = 2.0. Then, RCL = 2.0(0.610) = 1.22 m. RC =
RCL – ½D = 0.915 m. And, RS = RCL + ½D = 1.525 m.

g) Full pipe unit discharge

Use the orifice equation with Cd = 0.67:

 q = CdD 2gH = 0.67(0.610) 2(9.81)(1.850) = 2.46 m2 /s

h) Maximum unit discharge

This is based on the “vortex” equation:

⎛R ⎞
qmax = RC 2g(0.7h) ln ⎜ S ⎟
⎝ RC ⎠
⎛ 1.525 ⎞
= 0.915 2(9.81)(0.7)(10.05) ln ⎜ ⎟
⎝ 0.915 ⎠
= 5.49 m2 /s

Thus, qmax > q, as required. Then, the required barrel width is:

Q 4.00
b= = = 1.63 m
q 2.46

Finally, the area of the barrel is: A = bD = (1.63)(0.61) = 0.992 m2.

i) Siphon breaker diameter

This is taken as 1/24th of the barrel area, or 0.992/24 = 0.0413 m2. For a circular
pipe cross section, this give an inside diameter of 0.229 m. Round up to the
nearest available steel pipe size.

j) Outlet deflector sill height

This height is 1.5D = 1.5(0.61) = 0.915 m.

k) Outlet ceiling height

This is equal to:

h2 = 1.5D + Ecritical + 0.3048

Critical depth for the design discharge (4.0 m3/s) is:

(Q / b) ( 4.00 /1.63 )
2 2

hc =3 =3 = 0.850 m
g 9.81
 The velocity at critical depth would be:

Q 4.00
Vc = = = 2.89 m/s
bhc (1.63)(0.850)

And, the velocity head at critical flow is:

Vc2 (2.89)2
= = 0.425 m
2g 2(9.81)

Finally,

h2 = 1.5(0.61) + 0.425 + 0.850 + 0.3048 = 2.49 m

l) Hydraulic seal at inlet

Using the minimum dimensions, the inlet along the side slope of the canal is b x 2D,
or 1.63 wide by 1.22 high (in meters). This gives a total inlet area of (1.63)(1.22) =
1.99 m2. Then, the maximum inlet velocity head is:

Q2 (4.00)2
hv = = = 0.206 m
2gA 2 2(9.81)(1.99)2

Minimum hydraulic seal is defined as:

1.5hv + 0.152 = 1.5(0.206) + 0.152 = 0.461 m

If the inlet begins at the base of the canal side slope (see Fig. 4-17), this hydraulic
seal corresponds to a vertical depth of:

0.461 sin ⎡⎣ tan−1 (1/1.5 )⎤⎦ = 0.256 m

The depth of water above the inlet is, then: 1.289 – 0.256 = 1.033 m, which exceeds
the hydraulic seal requirement of 0.461 m. Thus, the hydraulic seal is sufficient.

m) Side-view drawing

See the drawing below.

 15o

FSL
1.289 m
ft

1.85 m
1.5:1 2
0.
30o
m high
0.61 m
22

0.915 m
water
1.

2.49 m
1.2
2
sill

m
1.63 m
60o 1:1

0.61 m
1.525 m 3.100 m
ll all
Barrel
ff wa ff w
to uto
cu c
Lecture 2
Flumes for Open-Channel Flow Measurement

“Superb accuracy in water measurement, Jessica thought.”

Dune, F. Herbert (1965)
I. Procedure for Installing a Parshall Flume to Ensure Free Flow

• If possible, you will want to specify the installation of a Parshall flume such
that it operates under free-flow conditions throughout the required flow range
• To do this, you need to specify the minimum elevation of the upstream floor of
the flume
• Follow these simple steps to obtain a free-flow in a Parshall flume, up to a
specified maximum discharge:

1. Determine the maximum flow rate (discharge) to be measured

2. Locate the high water line on the canal bank where the flume is to be
installed, or otherwise determine the maximum depth of flow on the
upstream side
3. Select a standard flume size and calculate hu from the free-flow equation
corresponding to the maximum discharge capacity of the canal
4. Place the floor of the flume at a depth not exceeding the transition
submergence, St, multiplied by hu below the high water line

• In general, the floor of the flume should be placed as high in the canal as
grade and other conditions permit, but not so high that upstream free board is
compromised.
• The downstream water surface elevation will be unaffected by the installation
of the flume (at least for the same flow rate)
• As an example, a 0.61-m Parshall flume is shown in the figure below
• The transition submergence, St, for the 0.61-m flume is 66% (see table)
• The maximum discharge in the canal is given as 0.75 m3/s, which for free-
flow conditions must have an upstream depth of (see Eq. 3): hu =
(0.75/1.429)1/1.55 = 0.66 m
• With the transition submergence of 0.66, this gives a depth to the flume
floor of 0.66(0.660 m) = 0.436 m from the downstream water surface
• Therefore, the flume crest (elevation of hu tap) should be set no lower than
0.436 m below the normal maximum water surface for this design flow
rate, otherwise the regime will be submerged flow
• However, if the normal depth for this flow rate were less than 0.436 m, you
would place the floor of the flume on the bottom of the channel and still
have free flow conditions
• The approximate head loss across the structure at the maximum flow rate
will be the difference between 0.660 and 0.436 m, or 0.224 m
• This same procedure can be applied to other types of open-channel
measurement flumes
BIE 5300/6300 Lectures 19 Gary P. Merkley
II. Non-Standard Parshall Flume Calibrations

• Some Parshall flumes were incorrectly constructed or were intentionally built with
a non-standard size
• Others have settled over time such that the flume is out of level either cross-wise
or longitudinally (in the direction of flow), or both
• Some flumes have the taps for measuring hu and hd at the wrong locations (too
high or too low, or too far upstream or downstream)
• Some flumes have moss, weeds, sediment or other debris that cause the
calibration to shift from that given for standard conditions
• Several researchers have worked independently to develop calibration
adjustments for many of the unfortunate anomalies that have befallen many
Parshall flumes in the field, but a general calibration for non-standard flumes
requires 3-D modeling
• There are calibration corrections for out-of-level installations and for low-flow
conditions

III. Hysteresis Effects in Parshall Flumes

• There have been reports by some researchers that hysteresis effects have been
observed in the laboratory under submerged-flow conditions in Parshall flumes
• The effect is to have two different flow rates for the same submergence, S, value,
depending on whether the downstream depth is rising or falling
• There is no evidence of this hysteresis effect in Cutthroat flumes, which are
discussed below

IV. Software

• You can use the ACA program to develop calibration tables for Parshall,
Cutthroat, and trapezoidal flumes
• Download ACA from:
http://www.engineering.usu.edu/bie/faculty/merkley/Software.htm
• You can also download the WinFlume program from:
http://www.usbr.gov/pmts/hydraulics_lab/winflume/index.html

Gary P. Merkley 20 BIE 5300/6300 Lectures

V. Submerged-Flow, Constant Flow Rate

• Suppose you have a constant flow rate through a Parshall flume

• How will hu change for different hd values under submerged-flow conditions?
• This situation could occur in a laboratory flume, or in the field where a
downstream gate is incrementally closed, raising the depth downstream of the
Parshall flume, but with a constant upstream inflow
• The graph below is for steady-state flow conditions with a 0.914-m Parshall flume
• Note that hu is always greater than hd (otherwise the flow would move upstream,
or there would be no flow)

BIE 5300/6300 Lectures 21 Gary P. Merkley

 Parshall Flume (W = 0.914 m)

1.0

0.9 Submerged flow conditions.

3
Constant flow rate: Qs = 1.00 m /s.

0.8
hu

0.7

0.6

0.5

0.4
0.4 0.5 0.6 0.7 0.8 0.9 1.0
hd

hd hu Q S Regime
(m) (m) (m3/s)
0.15 0.714 0.999 0.210 free
0.20 0.664 0.999 0.301 free
0.25 0.634 0.999 0.394 free
0.30 0.619 1.000 0.485 free
0.35 0.615 1.002 0.569 free
0.40 0.619 1.000 0.646 free
0.45 0.631 1.000 0.713 submerged
0.50 0.650 1.001 0.769 submerged
0.55 0.674 1.000 0.816 submerged
0.60 0.703 1.000 0.853 submerged
0.65 0.736 1.000 0.883 submerged
0.70 0.772 0.999 0.907 submerged
0.75 0.811 1.001 0.925 submerged
0.80 0.852 1.004 0.939 submerged
0.85 0.894 1.000 0.951 submerged
0.90 0.938 1.002 0.959 submerged

Gary P. Merkley 22 BIE 5300/6300 Lectures

VI. Cutthroat Flumes

• The Cutthroat flume was developed at USU

from 1966-1990
• A Cutthroat flume is a rectangular open-
channel constriction with a flat bottom and
zero length in the throat section (earlier
versions did have a throat section)
• Because the flume has a throat section of zero
length, the flume was given the name
“Cutthroat” by the developers (Skogerboe, et
A Cutthroat flume
al. 1967)
• The floor of the flume is level (as opposed to a Parshall flume), which has the
following advantages:

1. ease of construction − the flume can be readily placed

inside a concrete-lined channel
2. the flume can be placed on the channel bed
• The Cutthroat flume was developed to operate satisfactorily under both free-flow
and submerged-flow conditions
• Unlike Parshall flumes, all Cutthroat flumes have the same dimensional ratios
• It has been shown by experiment that downstream flow depths measured in the
diverging outlet section give more accurate submerged-flow calibration curves
than those measured in the throat section of a Parshall flume
• The centers of the taps for the US and DS head measurements are both located
½-inch above the floor of the flume, and the tap diameters should be ¼-inch

Cutthroat Flume Sizes

• The dimensions of a Cutthroat flume are identified by the flume width and length
(W x L, e.g. 4” x 3.0’)
• The flume lengths of 1.5, 3.0, 4.5, 6.0, 7.5, 9.0 ft are sufficient for most
applications
• The most common ratios of W/L are 1/9, 2/9, 3/9, and 4/9
• The recommended ratio of hu/L is equal to or less than 0.33

Free-flow equation

• For Cutthroat flumes the free-flow equation takes the same general form as for
Parshall flumes, and other channel “constrictions”:

nf
Qf = Cf W (hu ) (1)

where Qf is the free-flow discharge; W is the throat width; Cf is the free-flow

coefficient; and nf is the free-flow exponent

BIE 5300/6300 Lectures 23 Gary P. Merkley

 • That is, almost any non-orifice constriction in an open channel can be calibrated
using Eq. 1, given free-flow conditions
• The depth, hu, is measured from the upstream tap location (½-inch above the
flume floor)

3 6
1 1
B = W + L/4.5

B = W + L/4.5
W Flow
Inlet Outlet
Section Section

Lu = 2L/9 Ld = 5L/9
Lin = L/3 Lout = 2L/3
Piezometer
Tap for hu
Top View Piezometer
Tap for hd

L
Side View

• For any given flume size, the flume wall height, H, is equal to hu for Qmax,
according to the above equation, although a slightly larger H-value can be used
to prevent the occurrence of overflow
• So, solve the above free-flow equation for hu, and apply the appropriate Qmax
value from the table below; the minimum H-value is equal to the calculated hu

Gary P. Merkley 24 BIE 5300/6300 Lectures

Submerged-flow equation

• For Cutthroat flumes the submerged-flow equation also takes the same general
form as for Parshall flumes, and other channel constrictions:

Cs W(hu − hd )nf
Qs = ns
(2)
[ −(log10 S)]

where Cs = submerged-flow coefficient; W is the throat width; and S = hd/hu

• Equation 2 differs from the submerged-flow equation given previously for

Parshall flumes in that the C2 term is omitted
• The coefficients Cf and Cs are functions of flume length, L, and throat width, W
• The generalized free-flow and submerged-flow coefficients and exponents for
standard-sized Cutthroat flumes can be taken from the following table (metric
units: for Q in m3/s and head (depth) in m, and using a base 10 logarithm in Eq.
2)
• Almost any non-orifice constriction in an open channel can be calibrated using
Eq. 2, given submerged-flow conditions

Cutthroat Flume Calibration Parameters for metric units

(depth and W in m and flow rate in m3/s)

Cf nf St Cs ns Discharge (m3/s)
W (m) L (m)
min max
0.051 0.457 5.673 1.98 0.553 3.894 1.45 0.0001 0.007
0.102 0.457 5.675 1.97 0.651 3.191 1.58 0.0002 0.014
0.152 0.457 5.639 1.95 0.734 2.634 1.67 0.0004 0.022
0.203 0.457 5.579 1.94 0.798 2.241 1.73 0.0005 0.030
0.102 0.914 3.483 1.84 0.580 2.337 1.38 0.0002 0.040
0.203 0.914 3.486 1.83 0.674 1.952 1.49 0.0005 0.081
0.305 0.914 3.459 1.81 0.754 1.636 1.57 0.0008 0.123
0.406 0.914 3.427 1.80 0.815 1.411 1.64 0.0011 0.165
0.152 1.372 2.702 1.72 0.614 1.752 1.34 0.0005 0.107
0.305 1.372 2.704 1.71 0.708 1.469 1.49 0.0010 0.217
0.457 1.372 2.684 1.69 0.788 1.238 1.50 0.0015 0.326
0.610 1.372 2.658 1.68 0.849 1.070 1.54 0.0021 0.436
0.203 1.829 2.351 1.66 0.629 1.506 1.30 0.0007 0.210
0.406 1.829 2.353 1.64 0.723 1.269 1.39 0.0014 0.424
0.610 1.829 2.335 1.63 0.801 1.077 1.45 0.0023 0.636
0.813 1.829 2.315 1.61 0.862 0.934 1.50 0.0031 0.846
0.254 2.286 2.147 1.61 0.641 1.363 1.28 0.0009 0.352
0.508 2.286 2.148 1.60 0.735 1.152 1.37 0.0019 0.707
0.762 2.286 2.131 1.58 0.811 0.982 1.42 0.0031 1.056
1.016 2.286 2.111 1.57 0.873 0.850 1.47 0.0043 1.400
0.305 2.743 2.030 1.58 0.651 1.279 1.27 0.0012 0.537
0.610 2.743 2.031 1.57 0.743 1.085 1.35 0.0025 1.076
0.914 2.743 2.024 1.55 0.820 0.929 1.40 0.0039 1.611
1.219 2.743 2.000 1.54 0.882 0.804 1.44 0.0055 2.124

BIE 5300/6300 Lectures 25 Gary P. Merkley

 • Note that nf approaches 1.5 for larger W values, but never gets down to 1.5
• As for the Parshall flume data given previously, the submerged-flow calibration is
for base 10 logarithms
• Note that the coefficient conversion to English units is as follows:

(0.3048)1+nf
Cf (English) = 3
Cf (metric) (3)
(0.3048)

• The next table shows the calibration parameters for English units

Cutthroat Flume Calibration Parameters for English units

(depth and W in ft and flow rate in cfs)

Discharge (cfs)
W (ft) L (ft) Cf nf St Cs ns
min max
0.167 1.50 5.796 1.98 0.553 3.978 1.45 0.004 0.24
0.333 1.50 5.895 1.97 0.651 3.315 1.58 0.008 0.50
0.500 1.50 5.956 1.95 0.734 2.782 1.67 0.013 0.77
0.667 1.50 5.999 1.94 0.798 2.409 1.73 0.018 1.04
0.333 3.00 4.212 1.84 0.580 2.826 1.38 0.009 1.40
0.667 3.00 4.287 1.83 0.674 2.400 1.49 0.018 2.86
1.000 3.00 4.330 1.81 0.754 2.048 1.57 0.029 4.33
1.333 3.00 4.361 1.80 0.815 1.796 1.64 0.040 5.82
0.500 4.50 3.764 1.72 0.614 2.440 1.34 0.016 3.78
1.000 4.50 3.830 1.71 0.708 2.081 1.49 0.034 7.65
1.500 4.50 3.869 1.69 0.788 1.785 1.50 0.053 11.5
2.000 4.50 3.897 1.68 0.849 1.569 1.54 0.074 15.4
0.667 6.00 3.534 1.66 0.629 2.264 1.30 0.024 7.43
1.333 6.00 3.596 1.64 0.723 1.940 1.39 0.050 15.0
2.000 6.00 3.633 1.63 0.801 1.676 1.45 0.080 22.5
2.667 6.00 3.662 1.61 0.862 1.478 1.50 0.111 29.9
0.833 7.50 3.400 1.61 0.641 2.159 1.28 0.032 12.4
1.667 7.50 3.459 1.60 0.735 1.855 1.37 0.068 25.0
2.500 7.50 3.494 1.58 0.811 1.610 1.42 0.108 37.3
3.333 7.50 3.519 1.57 0.873 1.417 1.47 0.151 49.4
1.000 9.00 3.340 1.58 0.651 2.104 1.27 0.042 19.0
2.000 9.00 3.398 1.57 0.743 1.815 1.35 0.088 38.0
3.000 9.00 3.442 1.55 0.820 1.580 1.40 0.139 56.9
4.000 9.00 3.458 1.54 0.882 1.390 1.44 0.194 75.0

Unified Discharge Calibrations

• Skogerboe also developed “unified discharge” calibrations for Cutthroat flumes,

such that it is not necessary to select from the above standard flume sizes
• A regression analysis on the graphical results from Skogerboe yields these five
calibration parameter equations:

Cf = 6.5851L−0.3310 W 1.025 (4)

Gary P. Merkley 26 BIE 5300/6300 Lectures

 nf = 2.0936L−0.1225 − 0.128(W / L) (5)

ns = 2.003(W / L)0.1318 L−0.07044(W / L)−0.07131 (6)

−0.3555
St = 0.9653(W / L)0.2760 L0.04322(W / L) (7)

ns
Cf ( − log10 St )
Cs = nf
(8)
(1 − St )
• Note that Eqs. 4-8 are for English units (L and W in ft; Q in cfs)
• The maximum percent difference in the Cutthroat flume calibration parameters is
less than 2%, comparing the results of Eqs. 4-8 with the calibration parameters
for the 24 standard Cutthroat flume sizes

VII. Trapezoidal Flumes

• Trapezoidal flumes are often used for small flows, such as for individual furrows
in surface irrigation evaluations
• The typical standard calibrated flume is composed of five sections: approach,
converging, throat, diverging, and exit
• However, the approach and exit sections are not necessary part of the flume
itself

• Ideally, trapezoidal flumes can measure discharge with an accuracy of ±5%

under free-flow conditions
• But the attainment of this level of accuracy depends on proper installation,
accurate stage measurement, and adherence to specified tolerances in the
construction of the throat section
• Discharge measurement errors are approximately 1.5 to 2.5 times the error in the
stage reading for correctly installed flumes with variations in throat geometry from
rectangular to triangular sections
BIE 5300/6300 Lectures 27 Gary P. Merkley
 F

Gary P. Merkley
S
W P
F U
R S
W
Plan View

28
F Throat End View

A B C B D
F

P
Side View End View

BIE 5300/6300 Lectures

 • In the following table with seven trapezoidal flume sizes, the first two flumes are
V-notch (zero base width in the throat, and the last five have trapezoidal throat
cross sections

Flume Dimensions (inches)

Description
Number A B C D E F G P R S U W
1 o
Large 60 -V 7.00 6.90 7.00 3.00 7.00 1.00 6.75 2.00 1.50 4.00 3.50 0.00
2 o
Small 60 -V 5.00 6.05 5.00 2.00 4.25 1.00 4.00 2.00 1.00 2.38 2.50 0.00
3 o
2”-60 WSC 8.00 6.41 8.50 3.00 8.50 1.00 13.50 4.90 1.50 6.00 4.30 2.00
4 o
2”-45 WSC 8.00 8.38 8.50 3.00 8.50 1.00 10.60 4.90 1.50 10.60 4.30 2.00
5 o
2”-30 WSC 8.00 8.38 8.50 3.00 8.50 1.00 10.00 4.90 1.50 17.30 4.30 2.00
6 o
4”-60 WSC 9.00 9.81 10.00 3.00 10.00 1.00 13.90 8.00 1.50 8.00 5.00 4.00
7 o
2”-30 CSU 10.00 10.00 10.00 3.00 10.80 1.00 9.70 10.00 1.50 16.80 5.00 2.00
Note: All dimensions are in inches. WSC are Washington State Univ Calibrations, while CSU are
Colorado State Univ Calibrations (adapted from Robinson & Chamberlain 1960)

• Trapezoidal flume calibrations are for free-flow regimes only (although it would
be possible to generate submerged-flow calibrations from laboratory data)
• The following equation is used for free-flow calibration

nft
Qf = Cft (hu ) (9)

where the calibration parameters for the above seven flume sizes are given in the
table below:

Flume Qmax
Cft nft
Number (cfs)
1 1.55 2.58 0.35
2 1.55 2.58 0.09
3 1.99 2.04 2.53
4 3.32 2.18 2.53
5 5.92 2.28 3.91
6 2.63 1.83 3.44
7 4.80 2.26 2.97
Note: for h u in ft and Q in cfs

V-Notch Flumes

• When the throat base width of a trapezoidal flume is zero (W = 0, usually for the
smaller sizes), these are called “V-notch flumes”
• Similar to the V-notch weir, it is most commonly used for measuring water with a
small head due to a more rapid change of head with change in discharge
• Flume numbers 1 and 2 above are V-notch flumes because they have W = 0

BIE 5300/6300 Lectures 29 Gary P. Merkley

VIII. Flume Calibration Procedure

• Sometimes it is necessary to develop site-specific calibrations in the field or in

the laboratory
• For example, you might need to develop a custom calibration for a “hybrid” flume,
or a flume that was constructed to nonstandard dimensions
• To calibrate based on field data for flow measurement, it is desired to find flow
rating conditions for both free-flow and submerged-flow
• To analyze and solve for the value of the unknown parameters in the flow rating
equation the following procedure applies:

1. Transform the exponential equation into a linear equation using

logarithms
2. The slope and intersection of this line can be obtained by fitting the
transformed data using linear regression, or graphically with log-log
paper
3. Finally, back-calculate to solve for the required unknown values

The linear equation is:

Y = a + bX (10)

The transformed flume equations are:

Free-flow:

log(Q f ) = log ( Cf W ) + nf log(hu ) (11)

So, applying Eq. 10 with measured pairs of Qf and hu, “a” is log Cf and “b” is nf

Submerged-flow:

⎡ Qs ⎤
log ⎢ n ⎥ = log(Cs W) − ns log [ −(logS)] (12)
⎢⎣ (hu − hd ) f ⎥⎦

Again, applying Eq. 10 with measured pairs of Qs and hu and hd, “a” is log Cs and
“b” is ns

• Straight lines can be plotted to show the relationship between log hu and log Qf
for a free-flow rating, and between log (hu-hd) and log QS with several degrees of
submergence for a submerged-flow rating
• If this is done using field or laboratory data, any base logarithm can be used, but
the base must be specified
• Multiple linear regression can also be used to determine Cs, nf, and ns for
submerged flow data only − this is discussed further in a later lecture

Gary P. Merkley 30 BIE 5300/6300 Lectures

IX. Sample Flume Calibrations

Free Flow

• Laboratory data for free-flow conditions in a flume are shown in the following
table
• Free-flow conditions were determined for these data because a hydraulic jump
was seen downstream of the throat section, indicating supercritical flow in the
vicinity of the throat

Q (cfs) hu (ft)
4.746 1.087
3.978 0.985
3.978 0.985
2.737 0.799
2.737 0.798
2.211 0.707
1.434 0.533
1.019 0.436
1.019 0.436
1.019 0.436
1.019 0.436
0.678 0.337

• Take the logarithm of Q and of hu, then perform a linear regression (see Eqs. 10
and 11)
• The linear regression gives an R2 value of 0.999 for the following calibration
equation:

Qf = 4.04h1.66
u (13)

where Qf is in cfs; and hu is in ft

• We could modify Eq. 13 to fit the form of Eq. 6, but for a custom flume calibration
it is convenient to just include the throat width, W, in the coefficient, as shown in
Eq. 13
• Note that the coefficient and exponent values in Eq. 13 have been rounded to
three significant digits each – never show more precision than you can justify

Submerged Flow

• Data were then collected under submerged-flow conditions in the same flume
• The existence of submerged flow in the flume was verified by noting that there is
not downstream hydraulic jump, and that any slight change in downstream depth
produces a change in the upstream depth, for a constant flow rate
• Note that a constant flow rate for varying depths can usually only be obtained in
a hydraulics laboratory, or in the field where there is an upstream pump, with an
unsubmerged outlet, delivering water to the channel
BIE 5300/6300 Lectures 31 Gary P. Merkley
 • Groups of (essentially) constant flow rate data were taken, varying a downstream
gate to change the submergence values, as shown in the table below

Q (cfs) hu (ft) hd (ft)

3.978 0.988 0.639
3.978 1.003 0.753
3.978 1.012 0.785
3.978 1.017 0.825
3.978 1.024 0.852
3.978 1.035 0.872
3.978 1.043 0.898
3.978 1.055 0.933
3.978 1.066 0.952
3.978 1.080 0.975
3.978 1.100 1.002
3.978 1.124 1.045
2.737 0.800 0.560
2.736 0.801 0.581
2.734 0.805 0.623
2.734 0.812 0.659
2.733 0.803 0.609
2.733 0.808 0.642
2.733 0.818 0.683
2.733 0.827 0.714
2.733 0.840 0.743
2.733 0.858 0.785
2.733 0.880 0.823
2.733 0.916 0.876
2.733 0.972 0.943
1.019 0.437 0.388
1.019 0.441 0.403
1.010 0.445 0.418
1.008 0.461 0.434
1.006 0.483 0.462
1.006 0.520 0.506

• In this case, we will use nf in the submerged-flow equation (see Eq. 12), where nf
= 1.66, as determined above
• Perform a linear regression for ln[Q/(hu – hd)1.66] and ln[-log10S], as shown in Eq.
12, giving an R2 of 0.998 for

1.66
1.93 (hu − hd )
Qs = 1.45
(14)
( − log10 S )
where Qs is in cfs; and hu and hd are in ft

• You should verify the above results in a spreadsheet application

Gary P. Merkley 32 BIE 5300/6300 Lectures

References & Bibliography

Abt, S.R., Florentin, C. B., Genovez, A., and B.C. Ruth. 1995. Settlement and submergence
adjustments for Parshall flume. ASCE J. Irrig. and Drain. Engrg. 121(5).
Abt, S., R. Genovez, A., and C.B. Florentin. 1994. Correction for settlement in submerged Parshall
flumes. ASCE J. Irrig. and Drain. Engrg. 120(3).
Ackers, P., White, W. R., Perkins, J.A., and A.J.M. Harrison. 1978. Weirs and flumes for flow
measurement. John Wiley and Sons, New York, N.Y.
Genovez, A., Abt, S., Florentin, B., and A. Garton. 1993. Correction for settlement of Parshall flume.
J. Irrigation and Drainage Engineering. Vol. 119, No. 6. ASCE.
Kraatz D.B. and Mahajan I.K. 1975. Small hydraulic structures. Food and Agriculture Organization of
the United Nations, Rome, Italy.
Parshall, R.L. 1950. Measuring water in irrigation channels with Parshall flumes and small weirs. U.S.
Department of Agriculture, SCS Circular No. 843.
Parshall R.L. 1953. Parshall flumes of large size. U.S. Department of Agriculture, SCS and
Agricultural Experiment Station, Colorado State University, Bulletin 426-A.
Robinson, A.R. 1957. Parshall measuring flumes of small sizes. Agricultural Experiment Station,
Colorado State University, Technical Bulletin 61.
Robinson A. R. and A.R. Chamberlain. 1960. Trapezoidal flumes for open-channel flow
measurement. ASAE Transactions, vol.3, No.2. Trans. of American Society of Agricultural
Engineers, St. Joseph, Michigan.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965a. Submerged Parshall flumes
of small size. Report PR-WR6-1. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965c. Measuring water with
Parshall flumes. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G. V., Hyatt, M. L., Anderson, R. K., and K.O. Eggleston. 1967a. Design and calibration
of submerged open channel flow measurement structures, Part3: Cutthroat flumes. Utah Water
Research Laboratory, Logan, Utah.
Skogerboe, G.V., Hyatt, M.L. and K.O. Eggleston 1967b. Design and calibration of submerged open
channel flow measuring structures, Part1: Submerged flow. Utah Water Research Laboratory.
Logan, Utah.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965b. Submergence in a two-foot
Parshall flume. Report PR-WR6-2. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G. V., Hyatt, M. L., England, J. D., and J. R. Johnson. 1967c. Design and calibration of
submerged open-channel flow measuring structures Part2: Parshall flumes. Utah Water
Research Laboratory. Logan, Utah.
Working Group on Small Hydraulic Structures. 1978. Discharge Measurement Structures, 2nd ed.
International Institute for Land Reclamation and Improvement/ILRI, Wageningen, Netherlands.

BIE 5300/6300 Lectures 33 Gary P. Merkley

 Wright J.S. and B. Taheri. 1991. Correction to Parshall flume calibrations at low discharges. ASCE J.
Irrig. and Drain. Engrg.117(5).
Wright J.S., Tullis, B.P., and T.M. Tamara. 1994. Recalibration of Parshall flumes at low discharges.
J. Irrigation and Drainage Engineering, vol.120, No 2, ASCE.

Gary P. Merkley 34 BIE 5300/6300 Lectures

 BIE 5300/6300
Fall Semester 2004

Irrigation Conveyance & Control:

Flow Measurement & Structure Design
Lecture Notes

Biological & Irrigation Engineering Department

Utah State University, Logan, Utah
Gary P. Merkley ii BIE 5300/6300 Lectures
 Preface
These lecture notes were prepared by Gary P. Merkley of the Biological &
Irrigation Engineering Department at USU for use in the BIE 5300/6300 courses. The
material contained in these lecture notes is the intellectual property right of G.P.
Merkley, except where otherwise stated.

Many thanks are extended to USU engineering students, past and present,
whose numerous suggestions and corrections have been incorporated into these
lecture notes.

Copyright Notice

This material has been duplicated by special permission of the copyright holder.
It is not to be duplicated or used for purposes other than learning resource support for
Utah State University. Any violation of this agreement is punishable under existing
copyright laws.

© Copyright 2000, 2001, 2002, 2003, 2004

BIE 5300/6300 Lectures iii Gary P. Merkley

Gary P. Merkley iv BIE 5300/6300 Lectures
 Contents
Page

1 Course Introduction ........................................................................................... 1

2 Flumes for Open-Channel Flow Measurement ................................................ 19
3 Current Metering in Open Channels ................................................................ 35
4 Current Metering in Open Channels ................................................................ 45
5 Field Exercise #1 for Open-Channel Flow Measurement ................................ 57
6 Field Exercise #2 for Open-Channel Flow Measurement ................................ 59
7 Weirs for Flow Measurement ........................................................................... 61
8 Weirs for Flow Measurement ........................................................................... 73
9 Broad Crested Weirs ....................................................................................... 87
10 Broad Crested Weirs ....................................................................................... 95
11 Calibration of Canal Gates............................................................................. 107
12 Calibration of Canal Gates............................................................................. 123
13 Flow Measurement in Pipes........................................................................... 141
14 Flow Measurement in Pipes........................................................................... 153
15 Canal Design Basics...................................................................................... 161
16 Channel Cross Sections ................................................................................ 171
17 Design of Earthen Canals .............................................................................. 191
18 Sample Earthen Canal Designs..................................................................... 209
19 Canal Linings ................................................................................................. 221
20 Inverted Siphons............................................................................................ 229
21 Culvert Design & Analysis.............................................................................. 237
22 Sample Culvert Designs ................................................................................ 243
23 Culvert Hydraulic Behavior ............................................................................ 249
24 Flumes & Channel Transitions....................................................................... 257
25 Design Example for a Channel Transition...................................................... 265
26 Energy Dissipation Structures........................................................................ 275
27 Protective Structures ..................................................................................... 293
28 Safety Considerations.................................................................................... 301

These lecture notes are formatting for printing on both sides of the page, with odd-
numbered pages on the front. Each lecture begins on an odd-numbered page, so
some even-numbered pages are blank.

BIE 5300/6300 Lectures v Gary P. Merkley

Gary P. Merkley vi BIE 5300/6300 Lectures
Units, Constants and Conversions
28.35 g/oz
15.85 gpm/lps (= 60/3.785)
7.481 gallons/ft3
448.86 gpm/cfs (= 7.481*60)
3.7854 litres/gallon

6.89 kPa/psi
1 cb = 1 kPa
10 mb/kPa, or 100 kPa/bar
2.308 ft/psi, or 9.81 kPa/m (head of water)
14.7 psi = 101.3 kPa = 10.34 m (head of water) = 1,013 mbar = 1 atm
62.4 lbs/ft3, or 1000 kg/m3 (max density of pure water at 4°C)
0.1333 kPa/mmHg

1 ppm ≈ 1 mg/liter (usually)

1 mmho/cm = 1 dS/m = 550 to 800 mg/liter

0.7457 kW/HP
1 langley = 1 cal/cm2
0.0419 MJ/m2 per cal/cm2

0.3048 m/ft
1.609 km/mile
2.471 acre/ha
43,560 ft2/acre
1,233 m3/acre-ft

57.2958 degrees/radian
π ≈ 3.14159265358979323846
e ≈ 2.71828182845904523536

ºC = (ºF – 32)/1.8
ºF = 1.8(ºC) + 32

Ratio of weight to mass at sea level and 45° latitude: g = 9.80665 m/s2

BIE 5300/6300 Lectures vii Gary P. Merkley

Gary P. Merkley viii BIE 5300/6300 Lectures
Lecture 1
Course Introduction
“What is never measured is never managed”
Rep. Stephen Urquhart (2003)
I. Textbook and Other Materials

• The main references are Design of Small Canal Structures, USBR; and Water
Measurement Manual, USBR. At least one copy of each will be on reserve in the
library.
• Some material will also be referred to from Irrigation Fundamentals, by
Hargreaves & Merkley, as well as from other books and sources
• BIE 5300/6300 lecture notes by G.P. Merkley are required

II. Homework

• All work must be organized and neat

• There will be some computer programming and or spreadsheet exercises
• Submitting work late:10% reduction per day, starting after class

III. Tests

• One mid-term exam

• Final exam is not comprehensive

IV. Subject Areas

• Flow measurement
• Open channels
• Full pipe flow

• Design of conveyance infrastructure

• Canals, flumes, chutes
• Canal linings
• Siphons
• Culverts
• Energy dissipation structures

V. Why Measure Water?

Flow measurement is a key element in:

1. Water Management. Without knowing flow rates it is usually difficult to

quantify deliveries to water users, and in this case the evaluation of water
management practices is only vague
2. Water Quality Analysis. This relates to concentrations, rate of movement,
direction of movement, and dispersion of contaminants, and other issues
BIE 5300/6300 Lectures 1 Gary P. Merkley
 3. Water Rights and Water Law. This includes volumetric delivery allotments,
groundwater pumping, and excess water (e.g. irrigation runoff), among others

• Good quality, fresh water, is becoming more and

more scarce as people exploit water resources
more aggressively, and as the world population
increases
• This increases the importance of water
measurement
• It is unlikely that the regional and global situations
on water availability and water quality will improve
in the foreseeable future

When a resource is measured, it is implied that it

has significant value; when not measured, the
implication is of little or no value Nilometer, Cairo, Egypt

VI. Some Fundamental Flow Measurement Concepts

• Most flow measurement devices and techniques are based on the measurement
of head (depth or pressure) or velocity
• One exception to this is the salt dilution method (described below)
• Here, the term “flow rate” refers to volumetric rate, or volume per unit time
• Thus, we apply mathematical relationships between head and discharge, or take
products of velocity and cross-sectional area

• Strictly speaking, all open-channel and most pipe flow measurement techniques
cause head loss

“The inability to make accurate measurements is not necessarily

because of instrumentation deficiencies, but is a fundamental
property of the physical world – you cannot measure something
without changing it” (paraphrased) W.K. Heisenberg (1901-76)

• However, some methods incur negligible losses (e.g. ultrasonic)

• It is usually desirable to have only a small head loss because this loss typically
translates into an increased upstream flow depth in subcritical open-channel flow
• In open-channel flow measurement, devices can operate under free flow and
submerged flow regimes
• In free flow, we are concerned with the upstream head because critical flow
occurs in the vicinity of the flow measurement device. As long as this is true,
changes in downstream depth will not affect discharge at that location.
• In submerged flow, we are concerned with a head differential across the flow
measurement device.
• In this class, the terms flow rate and discharge will be used interchangeably.

Gary P. Merkley 2 BIE 5300/6300 Lectures

VII. Flow Measurement Accuracy

• Perhaps the most accurate method for measuring flow rate is by timing the filling
of a container of known volume
• However, this is often not practical for large flow rates
• Typical flow measurement accuracies are from ±2% to ±20% of the true
discharge, but this range can be much greater
• Measurements of head, velocity, and area are subject to errors for a variety of
reasons:

1. Approach Conditions

• High approach velocity

• Approach velocity not perpendicular to measurement device

2. Turbulence and Eddies

• Rough water surface

• Swirling flow near or at measurement location

3. Equipment Problems

• Staff gauges, current meters, floats,

etc., in disrepair
• Shifting calibrations on pressure
transducers and other meters
• Poor installation (non-horizontal
crest, wrong dimensions, etc.)

4. Measurement Location

• Local measurements (uniform flow

assumption?)
• Stream gauging stations (need
steady flow)

5. Human Errors

• Misreading water levels, etc.

• Misuse of equipment, or improper application of equipment

VIII. Simple Flow Measurement Methods

• The following are considered to be special methods, because they are mostly
simple and approximate, and because they are not usually the preferred methods
for flow measurement in open channels

BIE 5300/6300 Lectures 3 Gary P. Merkley

 • Preferred methods are through the use of calibrated structures (weirs, flumes,
orifices, and others), and current metering

1. Measurement by Observation

• In this method one must rely on experience to estimate the

discharge in an open channel, simply by observing the flow in
the channel and mentally comparing it to similar channels from
which the flow rate was measured and known
• This method is usually not very accurate, especially for large
flows, but some very experienced hydrographers can (with
some luck) arrive at a close estimate

2. Measurement by Floats

• The average flow velocity in an open channel can be estimated by measuring

the speed of a floating object on the surface of the water
• This can be done by marking uniform distances along the channel and using
a watch to measure the elapsed time from a starting location to respective
downstream locations
• However, in practice, usually only a single distance (say 10 m) is used

float

wa
te
r su
r fa
ce
10 m

• It is a good idea to have more than one measurement point so that the
velocity can be averaged over a reach, and to lessen the chance of an error
• Then, a graph can be made of float travel distance versus time, with the slope
equal to the surface velocity of the water
• Select a location in which the channel is fairly straight, not much change in
cross-section, smooth water surface, and no abrupt changes in bed elevation
or longitudinal slope
• Note that wind can affect the velocity of the float, changing the relationship
between surface velocity and average flow velocity

Gary P. Merkley 4 BIE 5300/6300 Lectures

 • Care should be taken to obtain measurements with the float moving near the
center of the surface width of flow, not bumping into the channel sides, and
not sinking

• The float speed will be higher than the average flow velocity in the channel,
unless perhaps the float travels near one of the channel banks or is
obstructed by vegetation
• You can estimate the average velocity in the channel by reducing the float
speed by some fraction
• The following table is from the U.S. Bureau of Reclamation
• It gives coefficients to multiply by the measured float velocity, as a function of
average depth, to obtain the approximate average flow velocity in the channel
Average Depth
(ft) (m) Coefficient
1 0.30 0.66
2 0.61 0.68
3 0.91 0.70
4 1.22 0.72
5 1.52 0.74
6 1.83 0.76
9 2.74 0.77
12 3.66 0.78
15 4.57 0.79
>20 >6.10 0.80

• To obtain average depth, divide the cross-sectional area by the top width of
the water surface (do not use an area-weighted average of subsection
depths)
• The coefficients in the above table only give approximate results; you can
typically expect errors of 10 to 20% in the flow rate
• What happens to the above coefficient values when the average water depth
is below 1 ft (or 0.3 m)?
• Some hydrographers have used partially submerged wooden sticks which are
designed to approximate the mean flow velocity, precluding the need for
coefficients as in the above table
• One end of the stick is weighted so that is sinks further

• The stick will give the correct velocity only for a small range of water depths
• The float method is not precise because the relationship between float speed
and true average flow velocity is not well known in general
• Other methods should be used if an accurate measurement is desired
BIE 5300/6300 Lectures 5 Gary P. Merkley
 Sample calculation:

The float method is applied in a rectangular channel with a base width of 0.94 m
and a uniform water depth of 0.45 m. Ten float travel times are recorded over a
distance of 5.49 m (18 ft), with calculated surface velocities:

Trial Time (s) V (m/s)

1 7.33 0.749
2 6.54 0.839
3 7.39 0.743
4 7.05 0.779
5 6.97 0.788
6 6.83 0.804
7 7.27 0.755
8 6.87 0.799
9 7.11 0.772
10 6.86 0.800
Avg: 7.02 0.783

Alternatively, the average surface velocity can be taken as 5.49/7.02 = 0.782

m/s, which is very close to the average velocity of 0.783 m/s from the above table
(as is usually the case). The surface velocity coefficient can be taken as 0.67,
interpolating in the previous table of USBR data, and the cross-sectional area of
the channel is (0.45)(0.94) = 0.42 m2. Then, the estimated flow rate is:

(0.67)(0.783 m/s)(0.42 m2) = 0.22 m3/s

3. Dye Method

• The dye method, or “color-velocity method”, can be used to

measure the flow velocity, similar to the float method
• However, in this method a slug of dye is injected into the
stream, and the time for the slug of dye to reach a downstream location is
measured
• This time can be taken as the average of the time for the first portion of the
dye to reach the downstream location, and the time for the last portion of the
dye to reach that location (the dye will disperse and elongate as it moves
downstream).
• The test section should not be too long, otherwise the dye will have dispersed
too much and it is difficult to visually detect the color differential in the water
• Usually, it is appropriate to use a test section of approximately 3 m
• Dyes used in this type of measurement should be nontoxic so as not to
pollute the water
• Food coloring can be used, as can other colored chemicals, such as
“flouricine”
Gary P. Merkley 6 BIE 5300/6300 Lectures
 4. Salt Dilution

• In this method, an aqueous solution of known salt concentration, C1, is poured

into the stream at a constant rate, q
• The completely mixed solution is measured at a downstream location,
providing the concentration C2
• The downstream location should be at least 5 m (perhaps up to 10 or 15 m)
from the point of injection, otherwise incomplete mixing may result in large
errors; that is, you might measure a highly-concentrated slug of water, or you
might miss the slug altogether, if you try to measure too close to the point of
injection
• After measuring the existing salt concentration in the flow (without adding the
concentrated solution), C0, the stream discharge, Q, can be calculated as,

∆t [QC0 + qC1] = ∆t [(Q + q)C2 ] (1)

or,
⎛ C − C2 ⎞
Q = q⎜ 1 ⎟ (2)
⎝ C2 − C0 ⎠

• The above two equations represent a mass balance, where: ∆t can be in s

(doesn’t really matter, because it cancels out); C can be in mg per liter; and q
and Q can be in lps
• This method is not used to measure velocity, but total volumetric flow rate
• Concentrations are normally expressed as mmho/cm, or dS/m

5. Uniform Flow

• In this method the channel bed slope, average cross-section, and average
depth are measured
• A roughness value is estimated, and the Manning or Chezy equation is
applied to calculate the discharge
• This method is valid only for steady uniform flow, and is severely limited by an
inability to accurately estimate the roughness value
• And because it is only valid for steady uniform flow, it cannot be
applied in general since these flow conditions are often not found
in open channels
• Ideally, both bed slope and water surface slope are measured to
verify whether the flow is uniform or not
• The discharge can be estimated by giving a range of probable flow
rates for maximum and minimum roughness values (also
estimated), based on the channel appearance and size
• The roughness can be estimated by experience, or by consulting
hydraulics handbooks which provide tables and figures, or photographs

BIE 5300/6300 Lectures 7 Gary P. Merkley

 6. Pitot Tube

• A simple pitot tube can be

positioned into the flow to measure
the velocity head
• One end of the tube is pointed into
the flow, and the other end is
pointed up vertically out of the water
− both ends are open
• The submerged end of the tube is positioned to be essentially parallel to the
flow
• Solving for velocity in the equation for velocity head:

V = 2gh (3)

• This method is best applied for higher flow velocities because it is difficult to
read the head differential at low velocities, in which large errors in the
estimation of velocity can result

IX. Introduction to Flumes

• Measurement flumes are open-channel

devices with a specially-shaped,
partially-constricted throat section
• The flume geometry is often designed
to cause the flow regime to pass
through critical depth, providing a
means for determining the rate of flow
from a single (upstream) water depth measurement − this is the advantage of a
free-flow regime
• When the water surface exceeds specified limits, submerged-flow conditions
occur and two water depth measurements are required (upstream and
downstream)
• In channels with small longitudinal bed slopes, it may be desirable to install a
flume to operate under conditions of submerged-flow rather than free flow in
order to:

1. reduce energy losses

2. place the flume on the channel bed to minimize the increase
in upstream water surface elevation

(the above two reasons are essentially the same thing)

• Many different flow measurement flumes have been designed and tested, but
only a few are commonly found in practice today

Gary P. Merkley 8 BIE 5300/6300 Lectures

X. Flume Classifications

There are two principal classes of flumes: short-throated and long-throated

Short-throated flumes:

• Critical flow conditions occur in regions of curvilinear flow (assuming the

regime is free flow)
• These include flumes with side contractions and bottom contractions, and
some type of transition section
• In general, laboratory calibrations are required to obtain flow coefficients
for rating
• Under favorable operating conditions the discharge can be determined
with an accuracy of ±2 to ±5% for free flow

Long-throated flumes:

• Critical flow conditions are created in a region of parallel flow in the control
section, again, assuming free flow conditions
• These linear-stream flow conditions are much better theoretically defined;
thus, rating relations can be reasonably well predicted
• Generally, flows larger than 10 lps can be measured with an error of less
than ±2% in an appropriately dimensioned flume
• Broad-crested weirs are an example of long-throated flumes

XI. Advantages and Disadvantages of Flumes

Advantages

1. capable of operation with relatively small head loss, and a high transition
submergence value (compared to sharp-crested weirs)
2. capable of measuring a wide range of free-flow discharges with relatively
high tail-water depths, using a single water depth measurement
3. capable of measuring discharge under submerged flow conditions using
two water depth measurements
4. both sediment and floating debris tend to pass through the structure
5. no need for a deep and wide upstream pool to reduce the velocity of
approach

Disadvantages

1. usually more expensive to construct than weirs

2. must be constructed carefully and accurately for satisfactory
performance
3. cannot be used as flow control structures (compared to adjustable weirs,
orifice gates, and other structures)

BIE 5300/6300 Lectures 9 Gary P. Merkley

 4. often need to use standard design dimensions, unless you want to
develop your own calibration curve

XII. Free, Submerged, and Transitional Flow

• When critical flow occurs the flow rate through the flume is uniquely
related to the upstream depth, hu
• That is, the free flow discharge can be obtained with only a single water
depth measurement

Qf = f(hu ) (4)

• When the tail-water depth is increased such that the flume operates under
submerged-flow conditions, both upstream, hu, and downstream, hd, depth
measurements are required.
• Let S be the submergence ratio, or S = hd/hu. Then, Qs is a function of the
head differential, (hu - hd), and S

Qs = f(hu − hd,S) (5)

• The value of submergence which marks the change from free flow to submerged
flow, and vice versa, is referred to as the transition submergence, St.
• At this condition the discharge given by the free-flow equation is exactly
the same as that given by submerged-flow equation

XIII. Parshall Flumes

• The Parshall flume is perhaps the most commonly used open-channel flow-
measuring device in irrigation systems in the U.S. and elsewhere
• It was developed at Colorado State University by Ralph Parshall from 1915-1922
• Some characteristics of this flume design are:

1. This flume has specially designed converging, throat and diverging

sections
2. It has been designed to measure flow from 0.01 to 3,000 cfs (1 lps to 85
m3/s), or more
3. Under typical conditions, free-flow accuracy is ±5% of the true discharge
4. Under favorable conditions (calm upstream water surface, precise flume
construction, level upstream flume floor) free-flow accuracy can be ±2%
of the true discharge
5. The head loss across a comparably-sized sharp-crested weir under free-
flow conditions is roughly four times that of a Parshall flume operating
under free-flow conditions
6. It is usually designed to operate under free-flow conditions

Gary P. Merkley 10 BIE 5300/6300 Lectures

 7. Size selection is based on the flume width which best fits the channel
dimensions and hydraulic properties
8. As a general rule, the width of the throat of a Parshall flume should be
about one-third to one-half the width of the upstream water surface in
the channel at the design discharge and at normal depth

• The general forms of the free-flow and submerged-flow equations for flumes,
including the Parshall flume, are:

Free Flow

nf
Qf = Cf W (hu ) (6)

Submerged Flow

Cs W(hu − hd )nf
Qs = (7)
[ −(log10 S + C2 )]ns

where nf and ns are the free-flow and submerged-flow exponents, respectively;

and W is the throat width.

• It is strongly recommended that you use the same units for W and depth (hu and
hd) in Eqs. 6 and 7 (i.e. don’t put W in inches and hu in feet)
• Below are two views of a Parshall flume
• Note that both hu and hd are measured from the upstream floor elevation, that is,
from a common datum
• This is in spite of the fact that the downstream tap is supposed to be located at
an elevation equal to H – Y below the upstream floor, as shown in the figure
below

• The diverging outlet section of the flume is not required when the structure is
placed at a drop in bed elevation, whereby it would always operate under free-
flow conditions
• The USBR (1974) discusses “modified Parshall flumes” which fit a particular
canal profile
• The following table gives dimensions (A-H, K, X & Y) and discharge ranges for
the 23 standard Parshall flume sizes (see the following figure showing the
dimensional parameters) in metric units

BIE 5300/6300 Lectures 11 Gary P. Merkley

Gary P. Merkley 12 BIE 5300/6300 Lectures
Parshall Flume Dimensions in metric units (see the above figure)

Dimensions (m) Q (m3/s)

W (m) A B C D E F G H K X Y min max
0.025 0.167 0.093 0.363 0.356 0.076 0.203 0.152 0.029 0.019 0.008 0.013 0.00028 0.0057
0.051 0.214 0.135 0.414 0.406 0.114 0.254 0.203 0.043 0.022 0.016 0.025 0.00057 0.011
0.076 0.259 0.178 0.467 0.457 0.152 0.305 0.381 0.057 0.025 0.025 0.038 0.00085 0.017
0.152 0.394 0.394 0.621 0.610 0.305 0.610 0.457 0.114 0.076 0.051 0.076 0.00142 0.082
0.229 0.575 0.381 0.879 0.864 0.305 0.457 0.610 0.114 0.076 0.051 0.076 0.00283 0.144
0.305 0.845 0.610 1.372 1.343 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0113 0.453
0.457 1.026 0.762 1.448 1.419 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0142 0.680
0.610 1.207 0.914 1.524 1.495 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0198 0.934
0.762 1.391 1.067 1.632 1.600 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0227 1.16
0.914 1.572 1.219 1.676 1.645 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0283 1.42
1.219 1.937 1.524 1.829 1.794 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0368 1.93
1.524 2.302 1.829 1.981 1.943 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0623 2.44
1.829 2.667 2.134 2.134 2.092 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.0736 2.94
2.134 3.032 2.438 2.286 2.242 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.116 3.43
2.438 3.397 2.743 2.438 2.391 0.610 0.914 0.914 0.229 0.076 0.051 0.076 0.130 3.96
3.048 4.756 3.658 4.350 4.267 0.914 1.829 1.219 0.343 0.152 0.305 0.229 0.170 5.66
3.658 5.607 4.470 4.972 4.877 0.914 2.438 1.524 0.343 0.152 0.305 0.229 0.227 9.91
4.572 7.620 5.588 7.772 7.620 1.219 3.048 1.829 0.457 0.229 0.305 0.229 0.227 17.0
6.096 9.144 7.315 7.772 7.620 1.829 3.658 2.134 0.686 0.305 0.305 0.229 0.283 28.3
7.620 10.668 8.941 7.772 7.620 1.829 3.962 2.134 0.686 0.305 0.305 0.229 0.425 34.0
9.144 12.313 10.566 8.084 7.925 1.829 4.267 2.134 0.686 0.305 0.305 0.229 0.425 42.5
12.192 15.481 13.818 8.395 8.230 1.829 4.877 2.134 0.686 0.305 0.305 0.229 0.566 56.6
15.240 18.529 17.272 8.395 8.230 1.829 6.096 2.134 0.686 0.305 0.305 0.229 0.708 85.0

• It is noted that Parshall flumes were developed using English units, but these
days we often prefer metric units
• Anyway, many of the dimensions in English units were not even “round”
numbers, often being specified to the 32nd of an inch
• The next table shows Parshall flume dimensions for the same 23 standard sizes,
but in feet, rounded to the thousandth of a foot, with discharge ranges in cubic
feet per second

BIE 5300/6300 Lectures 13 Gary P. Merkley

Parshall Flume Dimensions in English units (see the above figure)
Dimensions (ft) Q (cfs)
W (ft) A B C D E F G H K X Y min max
0.083 0.549 0.305 1.190 1.167 0.250 0.667 0.500 0.094 0.063 0.026 0.042 0.01 0.2
0.167 0.701 0.443 1.359 1.333 0.375 0.833 0.667 0.141 0.073 0.052 0.083 0.02 0.4
0.250 0.849 0.583 1.531 1.500 0.500 1.000 1.250 0.188 0.083 0.083 0.125 0.03 0.6
0.500 1.292 1.292 2.036 2.000 1.000 2.000 1.500 0.375 0.250 0.167 0.250 0.05 2.9
0.750 1.885 1.250 2.885 2.833 1.000 1.500 2.000 0.375 0.250 0.167 0.250 0.10 5.1
1.000 2.771 2.000 4.500 4.406 2.000 3.000 3.000 0.750 0.250 0.167 0.250 0.40 16.0
1.500 3.365 2.500 4.750 4.656 2.000 3.000 3.000 0.750 0.250 0.167 0.250 0.50 24.0
2.000 3.958 3.000 5.000 4.906 2.000 3.000 3.000 0.750 0.250 0.167 0.250 0.70 33.0
2.500 4.563 3.500 5.354 5.250 2.000 3.000 3.000 0.750 0.250 0.167 0.250 0.80 41.0
3.000 5.156 4.000 5.500 5.396 2.000 3.000 3.000 0.750 0.250 0.167 0.250 1.0 50.0
4.000 6.354 5.000 6.000 5.885 2.000 3.000 3.000 0.750 0.250 0.167 0.250 1.3 68.0
5.000 7.552 6.000 6.500 6.375 2.000 3.000 3.000 0.750 0.250 0.167 0.250 2.2 86.0
6.000 8.750 7.000 7.000 6.865 2.000 3.000 3.000 0.750 0.250 0.167 0.250 2.6 104
7.000 9.948 8.000 7.500 7.354 2.000 3.000 3.000 0.750 0.250 0.167 0.250 4.1 121
8.000 11.146 9.000 8.000 7.844 2.000 3.000 3.000 0.750 0.250 0.167 0.250 4.6 140
10.000 15.604 12.000 14.271 14.000 3.000 6.000 4.000 1.125 0.500 1.000 0.750 6.0 200
12.000 18.396 14.667 16.313 16.000 3.000 8.000 5.000 1.125 0.500 1.000 0.750 8.0 350
15.000 25.000 18.333 25.500 25.000 4.000 10.000 6.000 1.500 0.750 1.000 0.750 8.0 600
20.000 30.000 24.000 25.500 25.000 6.000 12.000 7.000 2.250 1.000 1.000 0.750 10 1000
25.000 35.000 29.333 25.500 25.000 6.000 13.000 7.000 2.250 1.000 1.000 0.750 15 1200
30.000 40.396 34.667 26.521 26.000 6.000 14.000 7.000 2.250 1.000 1.000 0.750 15 1500
40.000 50.792 45.333 27.542 27.000 6.000 16.000 7.000 2.250 1.000 1.000 0.750 20 2000
50.000 60.792 56.667 27.542 27.000 6.000 20.000 7.000 2.250 1.000 1.000 0.750 25 3000

• The minimum flow rate values represent the limits of the validity of the free-flow
rating equation
• For submerged flow conditions, a minimum flow rate also applies because if it is
very low, the difference between hu and hd will be virtually indistinguishable
(perhaps 1 mm or less)

• The next table gives calibration parameters (Cf, Cs, nf, ns) and transition
submergence (St) for standard Parshall flume sizes (metric units)
• Use Eq. (3) or (4) to get flow rate in m3/s, where depths are in metres
• The C2 value in Eq. (4) is equal to about 0.0044 (dimensionless) for all of the
standard Parshall flume sizes
• Standard sizes were developed in English units, so the throat width values show
below are “odd” numbers, but the ft-inch equivalents are given in parentheses
• Note that St is transition submergence − the value tends to increase with the size
of the flume, up to a maximum of about 0.80
• Be aware that the St values in the table below are for the maximum flow rate; for
other flow rates it is different
• Also note that the values in the table below are for a base 10 logarithm in Eq. (4)
• In practice, under extreme submerged-flow conditions, the head differential, hu-
hd, can be less than 1 mm and no measurement is possible with the flume

Gary P. Merkley 14 BIE 5300/6300 Lectures

 Parshall Flume Calibration Parameters for metric
units (depth and W in m, flow rate in m3/s)

Throat
Width (m) Cf Cs nf ns St
0.025 (1”) 2.38 2.10 1.55 1.000 0.56
0.051 (2”) 2.38 2.15 1.55 1.000 0.61
0.076 (3”) 2.32 2.14 1.55 1.000 0.64
0.152 (6”) 2.50 2.02 1.58 1.080 0.55
0.229 (9”) 2.34 1.91 1.53 1.060 0.63
0.305 (12”) 2.26 1.76 1.52 1.080 0.62
0.457 (18”) 2.32 1.71 1.54 1.115 0.64
0.610 (24”) 2.34 1.74 1.55 1.140 0.66

Metric Units
0.762 (30”) 2.36 1.70 1.56 1.150 0.67
0.914 (3’) 2.37 1.70 1.56 1.160 0.68
1.219 (4’) 2.40 1.66 1.57 1.185 0.70
1.524 (5’) 2.43 1.65 1.58 1.205 0.72
1.829 (6’) 2.46 1.62 1.59 1.230 0.74
2.134 (7’) 2.49 1.61 1.60 1.250 0.76
2.438 (8’) 2.49 1.59 1.60 1.260 0.78
3.048 (10’) 2.47 1.52 1.59 1.275 0.80
3.658 (12’) 2.43 1.50 1.59 1.275 0.80
4.572 (15’) 2.40 1.48 1.59 1.275 0.80
6.096 (20’) 2.37 1.46 1.59 1.275 0.80
7.620 (25’) 2.35 1.45 1.59 1.275 0.80
9.144 (30’) 2.33 1.44 1.59 1.275 0.80
12.192 (40’) 2.32 1.43 1.59 1.275 0.80
15.240 (50’) 2.31 1.42 1.59 1.275 0.80

• The following table gives calibration parameters (Cf, Cs, nf, ns) and transition
submergence (St) for standard Parshall flume sizes in English units

BIE 5300/6300 Lectures 15 Gary P. Merkley

 Parshall Flume Calibration Parameters for English
units (depth and W in ft, flow rate in cfs)

Throat
Width Cf Cs nf ns St
1 inches 4.06 3.59 1.550 1.000 0.56
2 inches 4.06 3.67 1.550 1.000 0.61
3 inches 3.97 3.66 1.550 1.000 0.64
6 inches 4.12 3.32 1.580 1.080 0.55
9 inches 4.09 3.35 1.530 1.060 0.63
12 inches 4.00 3.11 1.520 1.080 0.62
18 inches 4.00 2.95 1.540 1.115 0.64
24 inches 4.00 2.97 1.550 1.140 0.66
30 inches 4.00 2.89 1.555 1.150 0.67

English Units
3 feet 4.00 2.87 1.560 1.160 0.68
4 feet 4.00 2.78 1.570 1.185 0.70
5 feet 4.00 2.71 1.580 1.205 0.72
6 feet 4.00 2.64 1.590 1.230 0.74
7 feet 4.00 2.59 1.600 1.250 0.76
8 feet 4.00 2.55 1.600 1.260 0.78
10 feet 4.01 2.48 1.590 1.275 0.80
12 feet 3.96 2.45 1.590 1.275 0.80
15 feet 3.90 2.41 1.590 1.275 0.80
20 feet 3.85 2.38 1.590 1.275 0.80
25 feet 3.82 2.36 1.590 1.275 0.80
30 feet 3.80 2.34 1.590 1.275 0.80
40 feet 3.77 2.33 1.590 1.275 0.80
50 feet 3.75 2.32 1.590 1.275 0.80

• It is seen that nf, ns, and St are dimensionless, but Cf & Cs depend on the
units
• Also, the submerged-flow coefficient, Cs, is for a base-10 logarithm
• Note that, for 1 ft ≤ W ≤ 8 ft, the nf value can be approximated as:

nf ≈ 1.522 W 0.026 (8)

where W is in ft

• For W > 8 ft, nf is constant, at a value of 1.59

• For W > 8 ft, ns and St also remain constant at 1.275 and 0.80, respectively

Gary P. Merkley 16 BIE 5300/6300 Lectures

References & Bibliography

Abt, S.R., Florentin, C.B., Genovez, A., and Ruth, B.C. 1995. Settlement and Submergence
Adjustments for Parshall Flume. ASCE J. Irrig. and Drain. Engrg., 121(5):317-321.
Blaisdell, F.W. 1994. Results of Parshall Flume Tests. ASCE J. Irrig. and Drain. Engrg., 120(2):278-
291.
Brater, E.F., and King, H.W. 1976. Handbook of Hydraulics. McGraw-Hill Book Co., New York, N.Y.
Parshall, R.L. 1945. Improving the Distribution of Water to Farmers by Use of the Parshall Measuring
Flume. USDA Soil Conservation Service, in cooperation with the Colorado Agric. Exp. Station,
Colorado State Univ., Fort Collins, CO.
Parshall, R.L. 1953. Parshall Flumes of Large Size. USDA Soil Conservation Service, in cooperation
with the Colorado Agric. Exp. Station, Colorado State Univ., Fort Collins, CO.
Skogerboe, G.V., Hyatt, M.L., and Eggleston, K.O. 1967. Design & Calibration of Submerged Open
Channel Flow Measurement Structures, Part 1: Submerged Flow. Utah Water Research
Laboratory, Utah State Univ., Logan, UT.
Skogerboe, G.V., Hyatt, M.L., and Eggleston, K.O. 1967. Design & Calibration of Submerged Open
Channel Flow Measurement Structures, Part 2: Parshall Flumes. Utah Water Research
Laboratory, Utah State Univ., Logan, UT.
USBR. 1997. Water Measurement Manual. U.S. Bureau of Reclamation, Denver, CO. (also available
from Water Resources Publications, LLC, http://www.wrpllc.com/)
Wright, S.J., and Taheri, B. 1991. Correction to Parshall Flume Calibrations at Low Discharges.
ASCE J. Irrig. and Drain. Engrg., 117(5):800-804.

BIE 5300/6300 Lectures 17 Gary P. Merkley

Gary P. Merkley 18 BIE 5300/6300 Lectures
Lecture 2
Flumes for Open-Channel Flow Measurement

“Superb accuracy in water measurement, Jessica thought.”

Dune, F. Herbert (1965)
I. Procedure for Installing a Parshall Flume to Ensure Free Flow

• If possible, you will want to specify the installation of a Parshall flume such
that it operates under free-flow conditions throughout the required flow range
• To do this, you need to specify the minimum elevation of the upstream floor of
the flume
• Follow these simple steps to obtain a free-flow in a Parshall flume, up to a
specified maximum discharge:

1. Determine the maximum flow rate (discharge) to be measured

2. Locate the high water line on the canal bank where the flume is to be
installed, or otherwise determine the maximum depth of flow on the
upstream side
3. Select a standard flume size and calculate hu from the free-flow equation
corresponding to the maximum discharge capacity of the canal
4. Place the floor of the flume at a depth not exceeding the transition
submergence, St, multiplied by hu below the high water line

• In general, the floor of the flume should be placed as high in the canal as
grade and other conditions permit, but not so high that upstream free board is
compromised.
• The downstream water surface elevation will be unaffected by the installation
of the flume (at least for the same flow rate)
• As an example, a 0.61-m Parshall flume is shown in the figure below
• The transition submergence, St, for the 0.61-m flume is 66% (see table)
• The maximum discharge in the canal is given as 0.75 m3/s, which for free-
flow conditions must have an upstream depth of (see Eq. 3): hu =
(0.75/1.429)1/1.55 = 0.66 m
• With the transition submergence of 0.66, this gives a depth to the flume
floor of 0.66(0.660 m) = 0.436 m from the downstream water surface
• Therefore, the flume crest (elevation of hu tap) should be set no lower than
0.436 m below the normal maximum water surface for this design flow
rate, otherwise the regime will be submerged flow
• However, if the normal depth for this flow rate were less than 0.436 m, you
would place the floor of the flume on the bottom of the channel and still
have free flow conditions
• The approximate head loss across the structure at the maximum flow rate
will be the difference between 0.660 and 0.436 m, or 0.224 m
• This same procedure can be applied to other types of open-channel
measurement flumes
BIE 5300/6300 Lectures 19 Gary P. Merkley
II. Non-Standard Parshall Flume Calibrations

• Some Parshall flumes were incorrectly constructed or were intentionally built with
a non-standard size
• Others have settled over time such that the flume is out of level either cross-wise
or longitudinally (in the direction of flow), or both
• Some flumes have the taps for measuring hu and hd at the wrong locations (too
high or too low, or too far upstream or downstream)
• Some flumes have moss, weeds, sediment or other debris that cause the
calibration to shift from that given for standard conditions
• Several researchers have worked independently to develop calibration
adjustments for many of the unfortunate anomalies that have befallen many
Parshall flumes in the field, but a general calibration for non-standard flumes
requires 3-D modeling
• There are calibration corrections for out-of-level installations and for low-flow
conditions

III. Hysteresis Effects in Parshall Flumes

• There have been reports by some researchers that hysteresis effects have been
observed in the laboratory under submerged-flow conditions in Parshall flumes
• The effect is to have two different flow rates for the same submergence, S, value,
depending on whether the downstream depth is rising or falling
• There is no evidence of this hysteresis effect in Cutthroat flumes, which are
discussed below

IV. Software

• You can use the ACA program to develop calibration tables for Parshall,
Cutthroat, and trapezoidal flumes
• Download ACA from:
http://www.engineering.usu.edu/bie/faculty/merkley/Software.htm
• You can also download the WinFlume program from:
http://www.usbr.gov/pmts/hydraulics_lab/winflume/index.html

Gary P. Merkley 20 BIE 5300/6300 Lectures

V. Submerged-Flow, Constant Flow Rate

• Suppose you have a constant flow rate through a Parshall flume

• How will hu change for different hd values under submerged-flow conditions?
• This situation could occur in a laboratory flume, or in the field where a
downstream gate is incrementally closed, raising the depth downstream of the
Parshall flume, but with a constant upstream inflow
• The graph below is for steady-state flow conditions with a 0.914-m Parshall flume
• Note that hu is always greater than hd (otherwise the flow would move upstream,
or there would be no flow)

BIE 5300/6300 Lectures 21 Gary P. Merkley

 Parshall Flume (W = 0.914 m)

1.0

0.9 Submerged flow conditions.

3
Constant flow rate: Qs = 1.00 m /s.

0.8
hu

0.7

0.6

0.5

0.4
0.4 0.5 0.6 0.7 0.8 0.9 1.0
hd

hd hu Q S Regime
(m) (m) (m3/s)
0.15 0.714 0.999 0.210 free
0.20 0.664 0.999 0.301 free
0.25 0.634 0.999 0.394 free
0.30 0.619 1.000 0.485 free
0.35 0.615 1.002 0.569 free
0.40 0.619 1.000 0.646 free
0.45 0.631 1.000 0.713 submerged
0.50 0.650 1.001 0.769 submerged
0.55 0.674 1.000 0.816 submerged
0.60 0.703 1.000 0.853 submerged
0.65 0.736 1.000 0.883 submerged
0.70 0.772 0.999 0.907 submerged
0.75 0.811 1.001 0.925 submerged
0.80 0.852 1.004 0.939 submerged
0.85 0.894 1.000 0.951 submerged
0.90 0.938 1.002 0.959 submerged

Gary P. Merkley 22 BIE 5300/6300 Lectures

VI. Cutthroat Flumes

• The Cutthroat flume was developed at USU

from 1966-1990
• A Cutthroat flume is a rectangular open-
channel constriction with a flat bottom and
zero length in the throat section (earlier
versions did have a throat section)
• Because the flume has a throat section of zero
length, the flume was given the name
“Cutthroat” by the developers (Skogerboe, et
A Cutthroat flume
al. 1967)
• The floor of the flume is level (as opposed to a Parshall flume), which has the
following advantages:

1. ease of construction − the flume can be readily placed

inside a concrete-lined channel
2. the flume can be placed on the channel bed
• The Cutthroat flume was developed to operate satisfactorily under both free-flow
and submerged-flow conditions
• Unlike Parshall flumes, all Cutthroat flumes have the same dimensional ratios
• It has been shown by experiment that downstream flow depths measured in the
diverging outlet section give more accurate submerged-flow calibration curves
than those measured in the throat section of a Parshall flume
• The centers of the taps for the US and DS head measurements are both located
½-inch above the floor of the flume, and the tap diameters should be ¼-inch

Cutthroat Flume Sizes

• The dimensions of a Cutthroat flume are identified by the flume width and length
(W x L, e.g. 4” x 3.0’)
• The flume lengths of 1.5, 3.0, 4.5, 6.0, 7.5, 9.0 ft are sufficient for most
applications
• The most common ratios of W/L are 1/9, 2/9, 3/9, and 4/9
• The recommended ratio of hu/L is equal to or less than 0.33

Free-flow equation

• For Cutthroat flumes the free-flow equation takes the same general form as for
Parshall flumes, and other channel “constrictions”:

nf
Qf = Cf W (hu ) (1)

where Qf is the free-flow discharge; W is the throat width; Cf is the free-flow

coefficient; and nf is the free-flow exponent

BIE 5300/6300 Lectures 23 Gary P. Merkley

 • That is, almost any non-orifice constriction in an open channel can be calibrated
using Eq. 1, given free-flow conditions
• The depth, hu, is measured from the upstream tap location (½-inch above the
flume floor)

3 6
1 1
B = W + L/4.5

B = W + L/4.5
W Flow
Inlet Outlet
Section Section

Lu = 2L/9 Ld = 5L/9
Lin = L/3 Lout = 2L/3
Piezometer
Tap for hu
Top View Piezometer
Tap for hd

L
Side View

• For any given flume size, the flume wall height, H, is equal to hu for Qmax,
according to the above equation, although a slightly larger H-value can be used
to prevent the occurrence of overflow
• So, solve the above free-flow equation for hu, and apply the appropriate Qmax
value from the table below; the minimum H-value is equal to the calculated hu

Gary P. Merkley 24 BIE 5300/6300 Lectures

Submerged-flow equation

• For Cutthroat flumes the submerged-flow equation also takes the same general
form as for Parshall flumes, and other channel constrictions:

Cs W(hu − hd )nf
Qs = ns
(2)
[ −(log10 S)]

where Cs = submerged-flow coefficient; W is the throat width; and S = hd/hu

• Equation 2 differs from the submerged-flow equation given previously for

Parshall flumes in that the C2 term is omitted
• The coefficients Cf and Cs are functions of flume length, L, and throat width, W
• The generalized free-flow and submerged-flow coefficients and exponents for
standard-sized Cutthroat flumes can be taken from the following table (metric
units: for Q in m3/s and head (depth) in m, and using a base 10 logarithm in Eq.
2)
• Almost any non-orifice constriction in an open channel can be calibrated using
Eq. 2, given submerged-flow conditions

Cutthroat Flume Calibration Parameters for metric units

(depth and W in m and flow rate in m3/s)

Cf nf St Cs ns Discharge (m3/s)
W (m) L (m)
min max
0.051 0.457 5.673 1.98 0.553 3.894 1.45 0.0001 0.007
0.102 0.457 5.675 1.97 0.651 3.191 1.58 0.0002 0.014
0.152 0.457 5.639 1.95 0.734 2.634 1.67 0.0004 0.022
0.203 0.457 5.579 1.94 0.798 2.241 1.73 0.0005 0.030
0.102 0.914 3.483 1.84 0.580 2.337 1.38 0.0002 0.040
0.203 0.914 3.486 1.83 0.674 1.952 1.49 0.0005 0.081
0.305 0.914 3.459 1.81 0.754 1.636 1.57 0.0008 0.123
0.406 0.914 3.427 1.80 0.815 1.411 1.64 0.0011 0.165
0.152 1.372 2.702 1.72 0.614 1.752 1.34 0.0005 0.107
0.305 1.372 2.704 1.71 0.708 1.469 1.49 0.0010 0.217
0.457 1.372 2.684 1.69 0.788 1.238 1.50 0.0015 0.326
0.610 1.372 2.658 1.68 0.849 1.070 1.54 0.0021 0.436
0.203 1.829 2.351 1.66 0.629 1.506 1.30 0.0007 0.210
0.406 1.829 2.353 1.64 0.723 1.269 1.39 0.0014 0.424
0.610 1.829 2.335 1.63 0.801 1.077 1.45 0.0023 0.636
0.813 1.829 2.315 1.61 0.862 0.934 1.50 0.0031 0.846
0.254 2.286 2.147 1.61 0.641 1.363 1.28 0.0009 0.352
0.508 2.286 2.148 1.60 0.735 1.152 1.37 0.0019 0.707
0.762 2.286 2.131 1.58 0.811 0.982 1.42 0.0031 1.056
1.016 2.286 2.111 1.57 0.873 0.850 1.47 0.0043 1.400
0.305 2.743 2.030 1.58 0.651 1.279 1.27 0.0012 0.537
0.610 2.743 2.031 1.57 0.743 1.085 1.35 0.0025 1.076
0.914 2.743 2.024 1.55 0.820 0.929 1.40 0.0039 1.611
1.219 2.743 2.000 1.54 0.882 0.804 1.44 0.0055 2.124

BIE 5300/6300 Lectures 25 Gary P. Merkley

 • Note that nf approaches 1.5 for larger W values, but never gets down to 1.5
• As for the Parshall flume data given previously, the submerged-flow calibration is
for base 10 logarithms
• Note that the coefficient conversion to English units is as follows:

(0.3048)1+nf
Cf (English) = 3
Cf (metric) (3)
(0.3048)

• The next table shows the calibration parameters for English units

Cutthroat Flume Calibration Parameters for English units

(depth and W in ft and flow rate in cfs)

Discharge (cfs)
W (ft) L (ft) Cf nf St Cs ns
min max
0.167 1.50 5.796 1.98 0.553 3.978 1.45 0.004 0.24
0.333 1.50 5.895 1.97 0.651 3.315 1.58 0.008 0.50
0.500 1.50 5.956 1.95 0.734 2.782 1.67 0.013 0.77
0.667 1.50 5.999 1.94 0.798 2.409 1.73 0.018 1.04
0.333 3.00 4.212 1.84 0.580 2.826 1.38 0.009 1.40
0.667 3.00 4.287 1.83 0.674 2.400 1.49 0.018 2.86
1.000 3.00 4.330 1.81 0.754 2.048 1.57 0.029 4.33
1.333 3.00 4.361 1.80 0.815 1.796 1.64 0.040 5.82
0.500 4.50 3.764 1.72 0.614 2.440 1.34 0.016 3.78
1.000 4.50 3.830 1.71 0.708 2.081 1.49 0.034 7.65
1.500 4.50 3.869 1.69 0.788 1.785 1.50 0.053 11.5
2.000 4.50 3.897 1.68 0.849 1.569 1.54 0.074 15.4
0.667 6.00 3.534 1.66 0.629 2.264 1.30 0.024 7.43
1.333 6.00 3.596 1.64 0.723 1.940 1.39 0.050 15.0
2.000 6.00 3.633 1.63 0.801 1.676 1.45 0.080 22.5
2.667 6.00 3.662 1.61 0.862 1.478 1.50 0.111 29.9
0.833 7.50 3.400 1.61 0.641 2.159 1.28 0.032 12.4
1.667 7.50 3.459 1.60 0.735 1.855 1.37 0.068 25.0
2.500 7.50 3.494 1.58 0.811 1.610 1.42 0.108 37.3
3.333 7.50 3.519 1.57 0.873 1.417 1.47 0.151 49.4
1.000 9.00 3.340 1.58 0.651 2.104 1.27 0.042 19.0
2.000 9.00 3.398 1.57 0.743 1.815 1.35 0.088 38.0
3.000 9.00 3.442 1.55 0.820 1.580 1.40 0.139 56.9
4.000 9.00 3.458 1.54 0.882 1.390 1.44 0.194 75.0

Unified Discharge Calibrations

• Skogerboe also developed “unified discharge” calibrations for Cutthroat flumes,

such that it is not necessary to select from the above standard flume sizes
• A regression analysis on the graphical results from Skogerboe yields these five
calibration parameter equations:

Cf = 6.5851L−0.3310 W 1.025 (4)

Gary P. Merkley 26 BIE 5300/6300 Lectures

 nf = 2.0936L−0.1225 − 0.128(W / L) (5)

ns = 2.003(W / L)0.1318 L−0.07044(W / L)−0.07131 (6)

−0.3555
St = 0.9653(W / L)0.2760 L0.04322(W / L) (7)

ns
Cf ( − log10 St )
Cs = nf
(8)
(1 − St )
• Note that Eqs. 4-8 are for English units (L and W in ft; Q in cfs)
• The maximum percent difference in the Cutthroat flume calibration parameters is
less than 2%, comparing the results of Eqs. 4-8 with the calibration parameters
for the 24 standard Cutthroat flume sizes

VII. Trapezoidal Flumes

• Trapezoidal flumes are often used for small flows, such as for individual furrows
in surface irrigation evaluations
• The typical standard calibrated flume is composed of five sections: approach,
converging, throat, diverging, and exit
• However, the approach and exit sections are not necessary part of the flume
itself

• Ideally, trapezoidal flumes can measure discharge with an accuracy of ±5%

under free-flow conditions
• But the attainment of this level of accuracy depends on proper installation,
accurate stage measurement, and adherence to specified tolerances in the
construction of the throat section
• Discharge measurement errors are approximately 1.5 to 2.5 times the error in the
stage reading for correctly installed flumes with variations in throat geometry from
rectangular to triangular sections
BIE 5300/6300 Lectures 27 Gary P. Merkley
 F

Gary P. Merkley
S
W P
F U
R S
W
Plan View

28
F Throat End View

A B C B D
F

P
Side View End View

BIE 5300/6300 Lectures

 • In the following table with seven trapezoidal flume sizes, the first two flumes are
V-notch (zero base width in the throat, and the last five have trapezoidal throat
cross sections

Flume Dimensions (inches)

Description
Number A B C D E F G P R S U W
1 o
Large 60 -V 7.00 6.90 7.00 3.00 7.00 1.00 6.75 2.00 1.50 4.00 3.50 0.00
2 o
Small 60 -V 5.00 6.05 5.00 2.00 4.25 1.00 4.00 2.00 1.00 2.38 2.50 0.00
3 o
2”-60 WSC 8.00 6.41 8.50 3.00 8.50 1.00 13.50 4.90 1.50 6.00 4.30 2.00
4 o
2”-45 WSC 8.00 8.38 8.50 3.00 8.50 1.00 10.60 4.90 1.50 10.60 4.30 2.00
5 o
2”-30 WSC 8.00 8.38 8.50 3.00 8.50 1.00 10.00 4.90 1.50 17.30 4.30 2.00
6 o
4”-60 WSC 9.00 9.81 10.00 3.00 10.00 1.00 13.90 8.00 1.50 8.00 5.00 4.00
7 o
2”-30 CSU 10.00 10.00 10.00 3.00 10.80 1.00 9.70 10.00 1.50 16.80 5.00 2.00
Note: All dimensions are in inches. WSC are Washington State Univ Calibrations, while CSU are
Colorado State Univ Calibrations (adapted from Robinson & Chamberlain 1960)

• Trapezoidal flume calibrations are for free-flow regimes only (although it would
be possible to generate submerged-flow calibrations from laboratory data)
• The following equation is used for free-flow calibration

nft
Qf = Cft (hu ) (9)

where the calibration parameters for the above seven flume sizes are given in the
table below:

Flume Qmax
Cft nft
Number (cfs)
1 1.55 2.58 0.35
2 1.55 2.58 0.09
3 1.99 2.04 2.53
4 3.32 2.18 2.53
5 5.92 2.28 3.91
6 2.63 1.83 3.44
7 4.80 2.26 2.97
Note: for h u in ft and Q in cfs

V-Notch Flumes

• When the throat base width of a trapezoidal flume is zero (W = 0, usually for the
smaller sizes), these are called “V-notch flumes”
• Similar to the V-notch weir, it is most commonly used for measuring water with a
small head due to a more rapid change of head with change in discharge
• Flume numbers 1 and 2 above are V-notch flumes because they have W = 0

BIE 5300/6300 Lectures 29 Gary P. Merkley

VIII. Flume Calibration Procedure

• Sometimes it is necessary to develop site-specific calibrations in the field or in

the laboratory
• For example, you might need to develop a custom calibration for a “hybrid” flume,
or a flume that was constructed to nonstandard dimensions
• To calibrate based on field data for flow measurement, it is desired to find flow
rating conditions for both free-flow and submerged-flow
• To analyze and solve for the value of the unknown parameters in the flow rating
equation the following procedure applies:

1. Transform the exponential equation into a linear equation using

logarithms
2. The slope and intersection of this line can be obtained by fitting the
transformed data using linear regression, or graphically with log-log
paper
3. Finally, back-calculate to solve for the required unknown values

The linear equation is:

Y = a + bX (10)

The transformed flume equations are:

Free-flow:

log(Q f ) = log ( Cf W ) + nf log(hu ) (11)

So, applying Eq. 10 with measured pairs of Qf and hu, “a” is log Cf and “b” is nf

Submerged-flow:

⎡ Qs ⎤
log ⎢ n ⎥ = log(Cs W) − ns log [ −(logS)] (12)
⎢⎣ (hu − hd ) f ⎥⎦

Again, applying Eq. 10 with measured pairs of Qs and hu and hd, “a” is log Cs and
“b” is ns

• Straight lines can be plotted to show the relationship between log hu and log Qf
for a free-flow rating, and between log (hu-hd) and log QS with several degrees of
submergence for a submerged-flow rating
• If this is done using field or laboratory data, any base logarithm can be used, but
the base must be specified
• Multiple linear regression can also be used to determine Cs, nf, and ns for
submerged flow data only − this is discussed further in a later lecture

Gary P. Merkley 30 BIE 5300/6300 Lectures

IX. Sample Flume Calibrations

Free Flow

• Laboratory data for free-flow conditions in a flume are shown in the following
table
• Free-flow conditions were determined for these data because a hydraulic jump
was seen downstream of the throat section, indicating supercritical flow in the
vicinity of the throat

Q (cfs) hu (ft)
4.746 1.087
3.978 0.985
3.978 0.985
2.737 0.799
2.737 0.798
2.211 0.707
1.434 0.533
1.019 0.436
1.019 0.436
1.019 0.436
1.019 0.436
0.678 0.337

• Take the logarithm of Q and of hu, then perform a linear regression (see Eqs. 10
and 11)
• The linear regression gives an R2 value of 0.999 for the following calibration
equation:

Qf = 4.04h1.66
u (13)

where Qf is in cfs; and hu is in ft

• We could modify Eq. 13 to fit the form of Eq. 6, but for a custom flume calibration
it is convenient to just include the throat width, W, in the coefficient, as shown in
Eq. 13
• Note that the coefficient and exponent values in Eq. 13 have been rounded to
three significant digits each – never show more precision than you can justify

Submerged Flow

• Data were then collected under submerged-flow conditions in the same flume
• The existence of submerged flow in the flume was verified by noting that there is
not downstream hydraulic jump, and that any slight change in downstream depth
produces a change in the upstream depth, for a constant flow rate
• Note that a constant flow rate for varying depths can usually only be obtained in
a hydraulics laboratory, or in the field where there is an upstream pump, with an
unsubmerged outlet, delivering water to the channel
BIE 5300/6300 Lectures 31 Gary P. Merkley
 • Groups of (essentially) constant flow rate data were taken, varying a downstream
gate to change the submergence values, as shown in the table below

Q (cfs) hu (ft) hd (ft)

3.978 0.988 0.639
3.978 1.003 0.753
3.978 1.012 0.785
3.978 1.017 0.825
3.978 1.024 0.852
3.978 1.035 0.872
3.978 1.043 0.898
3.978 1.055 0.933
3.978 1.066 0.952
3.978 1.080 0.975
3.978 1.100 1.002
3.978 1.124 1.045
2.737 0.800 0.560
2.736 0.801 0.581
2.734 0.805 0.623
2.734 0.812 0.659
2.733 0.803 0.609
2.733 0.808 0.642
2.733 0.818 0.683
2.733 0.827 0.714
2.733 0.840 0.743
2.733 0.858 0.785
2.733 0.880 0.823
2.733 0.916 0.876
2.733 0.972 0.943
1.019 0.437 0.388
1.019 0.441 0.403
1.010 0.445 0.418
1.008 0.461 0.434
1.006 0.483 0.462
1.006 0.520 0.506

• In this case, we will use nf in the submerged-flow equation (see Eq. 12), where nf
= 1.66, as determined above
• Perform a linear regression for ln[Q/(hu – hd)1.66] and ln[-log10S], as shown in Eq.
12, giving an R2 of 0.998 for

1.66
1.93 (hu − hd )
Qs = 1.45
(14)
( − log10 S )
where Qs is in cfs; and hu and hd are in ft

• You should verify the above results in a spreadsheet application

Gary P. Merkley 32 BIE 5300/6300 Lectures

References & Bibliography

Abt, S.R., Florentin, C. B., Genovez, A., and B.C. Ruth. 1995. Settlement and submergence
adjustments for Parshall flume. ASCE J. Irrig. and Drain. Engrg. 121(5).
Abt, S., R. Genovez, A., and C.B. Florentin. 1994. Correction for settlement in submerged Parshall
flumes. ASCE J. Irrig. and Drain. Engrg. 120(3).
Ackers, P., White, W. R., Perkins, J.A., and A.J.M. Harrison. 1978. Weirs and flumes for flow
measurement. John Wiley and Sons, New York, N.Y.
Genovez, A., Abt, S., Florentin, B., and A. Garton. 1993. Correction for settlement of Parshall flume.
J. Irrigation and Drainage Engineering. Vol. 119, No. 6. ASCE.
Kraatz D.B. and Mahajan I.K. 1975. Small hydraulic structures. Food and Agriculture Organization of
the United Nations, Rome, Italy.
Parshall, R.L. 1950. Measuring water in irrigation channels with Parshall flumes and small weirs. U.S.
Department of Agriculture, SCS Circular No. 843.
Parshall R.L. 1953. Parshall flumes of large size. U.S. Department of Agriculture, SCS and
Agricultural Experiment Station, Colorado State University, Bulletin 426-A.
Robinson, A.R. 1957. Parshall measuring flumes of small sizes. Agricultural Experiment Station,
Colorado State University, Technical Bulletin 61.
Robinson A. R. and A.R. Chamberlain. 1960. Trapezoidal flumes for open-channel flow
measurement. ASAE Transactions, vol.3, No.2. Trans. of American Society of Agricultural
Engineers, St. Joseph, Michigan.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965a. Submerged Parshall flumes
of small size. Report PR-WR6-1. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965c. Measuring water with
Parshall flumes. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G. V., Hyatt, M. L., Anderson, R. K., and K.O. Eggleston. 1967a. Design and calibration
of submerged open channel flow measurement structures, Part3: Cutthroat flumes. Utah Water
Research Laboratory, Logan, Utah.
Skogerboe, G.V., Hyatt, M.L. and K.O. Eggleston 1967b. Design and calibration of submerged open
channel flow measuring structures, Part1: Submerged flow. Utah Water Research Laboratory.
Logan, Utah.
Skogerboe, G.V., Hyatt, M. L., England, J.D., and J. R. Johnson. 1965b. Submergence in a two-foot
Parshall flume. Report PR-WR6-2. Utah Water Research Laboratory, Logan, Utah.
Skogerboe, G. V., Hyatt, M. L., England, J. D., and J. R. Johnson. 1967c. Design and calibration of
submerged open-channel flow measuring structures Part2: Parshall flumes. Utah Water
Research Laboratory. Logan, Utah.
Working Group on Small Hydraulic Structures. 1978. Discharge Measurement Structures, 2nd ed.
International Institute for Land Reclamation and Improvement/ILRI, Wageningen, Netherlands.

BIE 5300/6300 Lectures 33 Gary P. Merkley

 Wright J.S. and B. Taheri. 1991. Correction to Parshall flume calibrations at low discharges. ASCE J.
Irrig. and Drain. Engrg.117(5).
Wright J.S., Tullis, B.P., and T.M. Tamara. 1994. Recalibration of Parshall flumes at low discharges.
J. Irrigation and Drainage Engineering, vol.120, No 2, ASCE.

Gary P. Merkley 34 BIE 5300/6300 Lectures

Lecture 3
Current Metering
“Six hours the waters run in, and six hours they run
out, and the reason is this: when there is higher water
in the sea than in the river, they run in until the river
gets to be highest, and then it runs out again”
The Last of the Mohicans, J.F. Cooper (1826)
I. Introduction

• Current metering in open channels is both

a science and an art
• One cannot learn how to do current
metering well by only reading a book
• This is a velocity-area flow measurement
method for open-channel flow
• A current meter is used to measure the
velocity at several points in a cross section
• The velocities are multiplied by respective
subsection areas to obtain flow rates

II. Types of Current Meters

• There are many companies that

manufacture good quality current meters, Current metering in an irrigation canal
and there are many types of current
meters, including mechanical and electromagnetic versions
• Current meters with a rotating unit that senses the water velocity are either vertical-
shaft or horizontal-shaft types
• The vertical-axis current meter has a rotating cup with a bearing system that is
simpler in design, more rugged, and easier to service and maintain than horizontal-
shaft (axis) current meters
• Because of the bearing system, the vertical-shaft meters will operate at lower
velocities than horizontal-axis current meters
• The bearings are well protected from silty
water, the adjustment is usually less Price AA
Current Meter
sensitive, and the calibration at lower
velocities is more stable
• Two of the commonly used vertical-axis
current meters are the Price Type A (or
AA) current meter and the Price Pygmy
current meter, the latter intended for use
with shallow flow depths and relatively low
velocities (less than 0.5 fps, or 0.15 m/s)

BIE 5300/6300 Lectures 35 Gary P. Merkley

 • The commonly-used Price AA current meter can measure velocities up to about 8
fps (2.4 m/s)
• None of the Price current meters can accurately measure velocities less than about
0.2 fps (0.07 m/s)

Price Type AA Current Meter

• The horizontal-shaft current meters use a propeller

• These horizontal-axis rotors disturb the flow less than vertical-axis cup rotors
because of axial symmetry in the flow direction
• Also, the horizontal-shaft current meters are less sensitive to vertical velocity
components in the channel

Price Pygmy Current Meter

• Because of its shape, the horizontal-axis current meter is less susceptible to

becoming fouled by small debris and vegetative material moving with the water
• Some common horizontal-axis current meters are the Ott (German), the Neyrpic
(France) and the Hoff (USA)

Gary P. Merkley 36 BIE 5300/6300 Lectures

 Ott Current Meters

• Some recent models have proven to be both accurate and durable when used in
irrigation channels
• Electromagnetic current meters are available that contain a sensor with the point
velocity displayed digitally
• Some earlier models manifested considerable electronic noise under turbulent flow
conditions (even the latest models still have problems if near steel-reinforced
concrete infrastructure, such as bridge piers)
• Present models yield more stable velocity readings, with averaging algorithms
• However, recent lab tests have shown that the Price current meters are more
accurate than at least two types of electromagnetic meter throughout a range of
velocities, and significantly more accurate at low velocities (J.M. Fulford 2002)

• Mechanical current meters automatically integrate and average velocities when

rotations per specific time interval are counted (e.g. count the rotations of the meter
in a 30-s or 60-s interval)
• The photograph below shows a digital display from an Ott current meter. The
number on the left is the elapsed time in seconds (to tenths of a second), and the
number on the right is the number of revolutions of the propeller

BIE 5300/6300 Lectures 37 Gary P. Merkley

 • With a traditional current meter, you use an electronic counter as shown above,
earphones that emit a beep for each rotation of the propeller, or other device, using
a stopwatch if necessary
• Then you divide the number of revolutions by the elapsed seconds to get a value in
revolutions per second
• The longer the duration (elapsed time) per measurement, the greater the integration
effect as the propeller speeds up and slows down with fluctuations in the flow
(unless the flow is perfectly stable at the location)
• The revolutions per second are directly proportional to the velocity of the flow,
according to the calibration of the instrument (see below)

III. Care of the Equipment

• Accuracy in velocity measurements can only be expected when

the equipment is properly assembled, adjusted, and maintained
• The current meter should be treated as a delicate instrument
that needs meticulous care and protective custody, both when
being used and when being transported
• The current meter necessarily receives a certain amount of hard
usage that may result in damage, such as a broken pivot, chipped bearing, or bent
shaft that will result in the current meter giving velocity readings that are lower than
actual velocities
• Measurements near bridge piers and abutments, water depth readings taken at
cross-sections having irregular bed profiles with the current meter attached to the
measuring line, and floating debris, represent the greatest hazards to the equipment
(Corbett, et al. 1943)
• Damage to current meter equipment during transport is generally due to careless
packing or negligence
• A standard case is provided by all manufacturers of current meter equipment, which
should be used before and after taking discharge measurements
• The equipment case should always be used when transporting the current meter,
even when the distance is relatively short
• Transport of assembled equipment from one location to another is one of the most
common sources of damage

Gary P. Merkley 38 BIE 5300/6300 Lectures

IV. Spin Test

• A “spin test” can be performed on a current meter to determine whether it is

spinning freely or not; this is done while the current meter is out of the water
• For a Price-type current meter, the USBR (1981) recommends putting the shaft in a
vertical position and giving the cups a quick turn by hand
• Ideally, the cups should spin for at least 3 minutes, but if it is only about 1½ minutes
the current meter can still be used, provided the velocity is not very low
• The cups should come to a smooth and gradual stop

V. Current Meter Ratings

• Usually, a current meter is

calibrated in a towing tank, which is
a small, straight open channel with
stagnant water
• The current meter is attached to a
carriage that travels on rails
(tracks) placed on the top of the
towing tank
• Then, a series of trials are
conducted wherein the current
meter is towed at different constant
velocities
• For each trial, the constant velocity
of the carriage is recorded, as well
as the revolutions per second (rev/s) of the current meter
• These data are plotted on rectangular coordinates to verify that a straight-line
relationship exists; then, the equation is determined by regression analysis
• The table below is an example of a velocity rating based on the rating equation for a
current meter:
Velocity (m/s) = 0.665 (rev/s) + 0.009 (1)

BIE 5300/6300 Lectures 39 Gary P. Merkley

 Sample Velocity Rating for a Current Meter, with Velocity in m/s
Time REVOLUTIONS
(seconds) 5 10 15 20 25 30 40 50 60 80 100

40 0.092 0.175 0.258 0.342 0.425 0.508 0.674 0.840 1.007 1.339 1.672

41 0.090 0.171 0.252 0.333 0.415 0.496 0.658 0.820 0.982 1.307 1.631

42 0.088 0.167 0.247 0.326 0.405 0.484 0.642 0.801 0.959 1.276 1.592

43 0.086 0.164 0.241 0.318 0.396 0.473 0.628 0.782 0.937 1.246 1.556

44 0.085 0.160 0.236 0.311 0.387 0.462 0.614 0.765 0.916 1.218 1.520

45 0.083 0.157 0.231 0.305 0.378 0.452 0.600 0.748 0.896 1.191 1.487

46 0.081 0.154 0.226 0.298 0.370 0.443 0.587 0.732 0.876 1.166 1.455

47 0.080 0.151 0.221 0.292 0.363 0.434 0.575 0.716 0.858 1.141 1.424

48 0.078 0.148 0.217 0.286 0.355 0.425 0.563 0.702 0.840 1.117 1.394

49 0.077 0.145 0.213 0.280 0.348 0.416 0.552 0.688 0.823 1.095 1.366

50 0.076 0.142 0.209 0.275 0.342 0.408 0.541 0.674 0.807 1.073 1.339

51 0.074 0.139 0.205 0.270 0.335 0.400 0.531 0.661 0.791 1.052 1.313

52 0.073 0.137 0.201 0.265 0.329 0.393 0.521 0.648 0.776 1.032 1.288

53 0.072 0.135 0.197 0.260 0.323 0.385 0.511 0.636 0.762 1.013 1.264

54 0.071 0.132 0.194 0.255 0.317 0.378 0.502 0.625 0.748 0.994 1.241

55 0.070 0.130 0.190 0.251 0.311 0.372 0.493 0.614 0.735 0.976 1.218

56 0.068 0.128 0.187 0.247 0.306 0.365 0.484 0.603 0.722 0.959 1.197

57 0.067 0.126 0.184 0.242 0.301 0.359 0.476 0.592 0.709 0.942 1.176

58 0.066 0.124 0.181 0.238 0.296 0.353 0.468 0.582 0.697 0.926 1.156

59 0.065 0.122 0.178 0.234 0.291 0.347 0.460 0.573 0.685 0.911 1.136

60 0.064 0.120 0.175 0.231 0.286 0.342 0.452 0.563 0.674 0.896 1.117

• Of course, if you have a calculator or

spreadsheet software, it may be
preferable to use the equation directly
rather than interpolating in a table, which
is based on the equation anyway
• The calibration equation is always linear,
where the constant term is the threshold
flow velocity at which the current meter
just begins to rotate; thus, current
meters have limits on the velocities that
can be measured
• Laboratory nozzles with uniform velocity
distributions at a circular cross section
are sometimes used (instead of a towing
tank) to calibrate current meters, as in
the Utah Water Research Lab

Gary P. Merkley 40 BIE 5300/6300 Lectures

VI. Methods Of Employing Current Meters

Wading

• The wading method involves

having the hydrographer stand in
the water holding a wading rod
with the current meter attached
to the rod
• The wading rod is graduated so
that the water depth can be
measured. The rod has a metal
foot pad which sets on the Wading method in an open channel
channel bed
• The current meter can be placed at any height on the wading rod and is readily
adjusted to another height by the hydrographer while standing in the water
• A tag line is stretched from one bank to the other, which can be a cloth or metal
tape
• This tag line is placed perpendicular to the flow direction
• The zero length on the tag line does not have to correspond with the edge of the
water on one of the banks
• This tag line is used to define the location of the wading rod each time that a current
meter measurement is made (recheck measurements each time, and check units)
• The wading rod is held at the tag line
• The hydrographer stands sideways to the flow direction, facing toward one of the
banks
• The hydrographer stands 5-10 cm downstream from the tag line and approximately
50 cm to one side of the wading rod
• During the measurement, the rod needs to be held in a vertical position and the
current meter must be parallel with the flow direction
• An assistant can signal to the hydrographer whether or not the rod is vertical in
relation to the flow direction
• If the flow velocity at the bank is not zero, then this velocity should be estimated as
a percentage of the velocity at the nearest measuring point (vertical)
• Thus, the nearest measuring point
should be as close to the bank as
possible in order to minimize the
error in the calculated discharge for
the section adjacent to the bank

Bridge

• Many of the larger irrigation

channels have bridges at various
locations, such as headworks and
cross regulators, but they may not

BIE 5300/6300 Lectures 41 Gary P. Merkley

 be located at an appropriate
section for current meter
measurements
• However, culverts often
prove to be very good
locations, with current meter
measurements usually being
made on the downstream
end of the culvert where
parallel streamlines are
more likely to occur
• Bridges often have piers,
which tend to collect debris
on the upstream face, that
should be removed prior to
undertaking current meter measurements
• Either a hand line or a reel assembly may be used from a bridge
• In either case, a weight is placed at the bottom of the line, which sets on the channel
bed in order that the line does not move as a result of the water flow
• The current meter is then placed at whatever location is required for each
measurement
• For a hand line assembly, the weight is lowered from the bridge to the channel bed
and the reading on the graduated hand line is recorded; then, the weight is lifted
until it is setting on the water surface and the difference in the two readings on the
hand line is recorded as the water depth
• Afterwards, the current meter is placed at the appropriate location on the hand line
in order to make the velocity measurement
• If a weight heavier than 10-15 kg is required in order to have a stable, nearly
vertical, cable line, then a crane-and-reel assembly is used
• The reel is mounted on a crane designed to clear the handrail of the bridge and to
guide the meter cable line beyond any interference with bridge members
• The crane is attached to a movable base for convenience in transferring the
equipment from one measuring point (vertical) to another

Cableway

• For very wide canals, or rivers, with water depths exceeding 150 cm, a cable is
placed above the water with vertical supports on each bank that are heavily
anchored for stability
• The cable supports a car (box) that travels underneath the cable using pulleys. This
car carries the hydrographer and the current meter equipment
• The cable has markers so that the location across the channel is known
• A hand line or a cable reel assembly is used depending on the size of the weight
that must be used

Gary P. Merkley 42 BIE 5300/6300 Lectures

 Boat

• For some very wide

channels, such as those
often encountered in the
Indian subcontinent (and
many other places), the
installation of a cableway
is a significant expense
• Consequently, a boat is
commonly employed
instead of the cableway
• Some friends on the banks should help hold the boat in place with ropes while the
velocity measurements are taken
• Either a hand line or a cable reel assembly is used in this case
• This method is not as convenient as the wading method, and it takes longer to
make measurements, but it is sometimes the best alternative

References & Bibliography

Fulford, J.M. 2002. Comparison of Price Meters to Marsh-McBirney and Swoffer Meters. WRD
Instrument News, March.

BIE 5300/6300 Lectures 43 Gary P. Merkley

Gary P. Merkley 44 BIE 5300/6300 Lectures
Lecture 4
Current Metering

I. Velocity Measurement Techniques

Vertical Velocity Method

• The most complete method for establishing the mean velocity at a vertical section is
to take a series of current meter velocity measurements at various depths in the
vertical
• Often, the current meter is placed below the water surface at one-tenth of the water
depth and a velocity measurement is made, then the current meter is placed at two-
tenths of the water depth; this procedure is continued until the velocity has finally
been measured at nine-tenths below the water surface
• Of particular importance are the velocity measurements at relative water depths of
0.2, 0.6 and 0.8 because they are used in the simpler methods
• When the above field procedure has been completed for a number of verticals in the
cross section, the data are plotted
• The relative water depth, which varies from zero at the water surface to unity at the
channel bed, is plotted on the ordinate starting with zero at the top of the ordinate
scale and unity at the bottom of the ordinate scale
• Velocity is plotted on the abscissa
• A smooth curve can be fitted on the data points for each vertical, from which the
mean velocity for the vertical can be determined
• Also, the relative water depth(s) corresponding with the mean velocity on the
velocity profile can be compared between each vertical
• Because the field procedure and data analysis for this method
are time consuming, simpler methods are commonly used
• Some of the more common methods are described in the
following sections
• However, the vertical velocity method provides an opportunity to determine whether
or not the simpler procedures are valid, or if some adjustments are required

Two-Point Method

• The most common methodology for establishing the mean

velocity in a vertical is the Two-Point Method
• Based on many decades of experience, a current meter
measurement is made at two relative water depths -- 0.2
and 0.8
• The average of the two measurements is taken as the mean
velocity in the vertical

V0.2 + V0.8
V= (1)
2

BIE 5300/6300 Lectures 45 Gary P. Merkley

 • In some field cases the velocity profile is distorted
• For example, measurements taken downstream from a structure may have very
high velocities near the water surface that can be visually observed, or near the
channel bed which can be sensed by the hydrographer when using the wading
method
• If there is any suspicion that an unusual velocity profile might exist in the cross
section, the vertical velocity method can be used to establish a procedure for
determining the mean velocity in a vertical for that cross section

Six-Tenths Method

• For shallow water depths, say less than 75 cm, the Six-
Tenths Method is used
• However, shallow is a relative term that is dependent on the
type (size) of current meter being used
• A single current meter measurement is taken at a relative
water depth of 0.6 below the water surface and the resulting
velocity is used as the mean velocity in the vertical
• In irrigation canals, this method is commonly used at the first vertical from each
bank, while the two points method is used at all of the other verticals in the cross-
section
• Frequently, the first vertical from each bank has a low velocity so the discharge in
each section adjacent to the left and right (looking downstream) banks represents a
very small portion of the total discharge in the cross-section
• In situations where shallow flow depths exist across most of the cross section, and
the six-tenths method must be used because of the type of current meter that is
available, then it is likely there will be considerable error in the velocity
measurement, perhaps more than ten percent

Three-Point Method

• This is a combination of the previous two methods

• The mean velocity is calculated as the sum of the results from the previous two
methods, divided by two:

1 ⎛ V0.2 + V0.8 ⎞
V= ⎜ + V0.6 ⎟ (2)
2 ⎝ 2 ⎠
Integration Method

• In this approach, experienced hydrographers can slowly lower and raise the current
meter two or three times along a vertical line in the stream
• The resulting “integrated” velocity along the vertical is then used to determine the
flow rate in a cross-section
• This method is subject to large errors, however, and should only be used for quick
checks
Gary P. Merkley 46 BIE 5300/6300 Lectures
II. Velocity at Vertical Walls

• Vertical walls are frequently encountered in irrigation systems

• Usually, this occurs in rectangular channels lined with concrete or
brick-and-mortar
• Even earthen canals will likely have some structures with a
rectangular cross section
• In some cases, there may be a vertical retaining wall along only one
side of the canal to stabilize the embankment

• In such cases, visual observation will usually disclose that the velocity very near the
vertical wall is significantly greater than zero
• These data are given in the table below
• The mathematical relationship between the parameters is:

Vw Vw / VD
= (3)
Vx Vx / VD

and,
⎛ V / V ⎞ 0.65 Vx
Vw = Vx ⎜ w D ⎟ = (4)
⎝ Vx / VD ⎠ Vx / VD
where,

D= depth of flow measured at the vertical wall;

x= horizontal distance from the wall toward the center of the channel (x
equals zero at the wall);
VD = mean velocity in the vertical at x = D;
Vx = mean velocity measured in the vertical at a horizontal distance x from
the wall (x ≤ D);
Vw = calculated mean velocity in the vertical at the wall, where x = 0;
Vx / VD = relative mean velocity in the vertical at a horizontal distance x from the
wall;
Vw / VD = relative mean velocity in the vertical at the wall, where x = 0

BIE 5300/6300 Lectures 47 Gary P. Merkley

 • The following table gives some values for the relationship between x/D and Vx/VD:

x/D Vx / VD
0.0 0.650
0.1 0.825
0.2 0.884
0.3 0.916
0.4 0.936
0.5 0.952
0.6 0.964
0.7 0.975
0.8 0.984
0.9 0.993
1.0 1.000

• When applying this procedure in a spreadsheet or other computer application, you

can use the following equation to accurately define the same relationship:

Vx 0.65 + 10.52 ( x / D )
= (5)
VD 1 + 10.676 ( x / D ) − 0.51431( x / D )2

• The figure below is a graphical representation of Eq. 5

1.00

0.95

0.90

0.85
Vx/VD

0.80

0.75

0.70

0.65

0.60
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
x/D

Gary P. Merkley 48 BIE 5300/6300 Lectures

 • The ratio Vx / VD is obtained from the above table after having measured Vx at a
horizontal distance x from the wall
• The accuracy of the estimated mean velocity at the wall will be enhanced by
measuring the mean velocity in a vertical located as close to the vertical wall as the
current meter equipment will allow
• Thus, if a current meter measurement could be made at a distance D/4 from the
wall, then the estimated mean velocity at the vertical wall would be the mean
velocity measured at D/4 from the wall multiplied by the ratio 0.65/0.90, where the
0.90 value is obtained from interpolating in the table
• In this example, the relative horizontal distance from the wall is x/D = (D/4)/D = 0.25
• Note that if x/D = 0, Vx / VD = 0.65, giving Vw = Vx (which is logical because
the measurement is at the wall)
• Note also that x should be less than D
• Special current meters exist for measuring velocities very close to vertical walls,
but they are expensive and not very common

III. Selection of Measuring Cross Section

• The most commonly used criterion in

selecting a channel cross section for
current meter measurements is that it be
located in a straight channel
• Cross sections having large eddies and
excessive turbulence are to be avoided
• A cross section with stagnant water near
one of the banks should be avoided
• Avoid cross sections where the flow depths
are shallow (except near the banks) and
the flow velocities are too low
• Rantz (1982) recommends that the flow
depths should exceed 15 cm and the flow velocities should exceed 15 cm/s
• It is preferable to select a cross section with little or no aquatic growth that can
cause problems with the rotation of the current meter – but this is true for
electromagnetic current meters too, although to a lesser extent, because vegetation
in the canal tends to cause velocity fluctuations
• A cross-section is preferred where the channel bed is not highly irregular so that the
area of the cross section can be accurately determined
• An irregular channel bed will affect the velocity profiles

IV. Subdivision of a Cross Section into Verticals

• The current meter is used to measure the mean velocity of each vertical in the cross
section
• In addition, the spacing of the verticals is used in determining the cross-sectional
area of each section, where a section is defined as the cross-sectional area of flow
between two verticals.
BIE 5300/6300 Lectures 49 Gary P. Merkley
 • In natural channels, the measuring cross-section should be subdivided into twenty
(20) or more verticals for a relatively smooth channel bed; but for lined canals,
twenty verticals is usually excessive and unnecessary
• For an irregular channel bed, more verticals are needed, not only to better define
the cross-sectional area of flow, but also because an irregular bed causes more
variation in the velocity distribution
• Verticals do not need to be spaced closer than 0.3 m across the width of the
channel (Corbett et al. 1943), but for small canals the verticals can be closer than
that, especially when there are only 3 or 4 verticals across the section
• For concrete-lined trapezoidal cross-section canals of small and medium size, it is
typical to take verticals at the mid-points of the side slopes on each side, and at the
two vertices where the side slopes meet the canal bottom, then dividing the base
width into 3 to 5 equally-spaced verticals

• An example earthen canal cross-section is illustrated in the figure below

• The most important verticals for defining the cross-sectional area of flow are shown
in this figure (for this example)
• The data from the above figure will be used below in sample calculations

V. Measuring Water Depths

• The water depth must be known at each vertical in order to calculate the cross-
sectional area of flow for two sections, one on each side of the vertical
• Accurately determining the flow areas is just as important as accurate velocity
measurements
• The greatest sources of error in measuring the depth of water are:

1. an irregular channel bed

2. a channel bed that is soft so that a weight or a rod sinks into the
material, indicating a water depth greater than actually exists
3. human error (simple mistakes… not paying attention to detail?)

• Another source of error: water “piles up” on the upstream edge of the rod and is
lower on the downstream edge, requiring the hydrographer to sight across the rod,
looking both upstream and downstream to get a reading
Gary P. Merkley 50 BIE 5300/6300 Lectures
VI. Recording of Data

• There are various formats for recording current metering data, and various
computational procedures (all of which are similar)
• These days, it is usually convenient to transfer the data to a spreadsheet application
and do the computations therein

Date: _________________ Channel: ______________________________ Station: ________________________

distance velocity
observed

revolu-
flow

mean
depth

depth

depth

width
tions
from

area
time
start at point mean mean rate
point in vertical in section

Hydrographers: ____________________________________________________________________
No. _____ of _____ page(s) Computations __________________________________________
Checked by: ________________________________________________________________________

VII. Computational Procedure

• The computational procedure for an example current meter discharge measurement

is given in the table below (see the figure above also)
• The major verticals had readings of 0.82, 1.23, 2.22, 3.70 and 4.50 m along the tag
line
• Intermediate verticals were selected as listed in the table
• The water surface is contained between 0.27 and 4.77 m along the tag line
• The first cross-section is contained between 0.27 and 0.82 m along the tag line
• The velocity at the bank is roughly estimated to be 10% of the mean velocity in the
vertical at 0.82 m along the tag line (the velocity at the bank is often listed as zero)

• Because of the shallow water depth at 0.82 m, the six-tenths method was used in
making the current meter measurement, which resulted in a velocity of 0.208 m/s
• The discharge in this cross section is less than 0.5% of the total discharge

BIE 5300/6300 Lectures 51 Gary P. Merkley

 • For the last cross-section, which contains a vertical wall, a set of current meter
measurements were made at 4.50 m along the tag line, with the mean velocity in
the vertical being 0.553 m/s
• The distance, x, from the vertical wall was 4.77 - 4.50 = 0.27 m
• The depth of water, D, at the vertical wall was 0.92 m
• Thus, x/D is 0.27 /0.92 = 0.29

Gary P. Merkley 52 BIE 5300/6300 Lectures

 • The relative mean
velocity in the vertical is
0.91, whereas the
relative mean velocity at
the wall is 0.65
• The total flow rate in the
cross section is
estimated to be 1.831
m3/s (see the following
table), but this can be
rounded to 1.83 m3/s
because it is almost
certain that the accuracy
is less than four
significant digits

distance velocity (m/s) mean flow

from depth depth time mean in mean in depth width area rate
edge (m) (m) fraction revolutions (s) at point vertical section (m) (m) (m2) (m3/s)
0.27 0.00 -- -- -- 10% 0.0208
0.1144 0.135 0.55 0.0743 0.008
0.82 0.27 0.6 20 67 0.208 0.2080
0.2245 0.480 0.41 0.1968 0.044
1.23 0.69 0.2 20 54 0.255 0.2410
0.8 20 61 0.227 0.2580 0.775 0.32 0.2480 0.064
1.55 0.86 0.2 25 58 0.296 0.2750
0.8 25 68 0.254 0.3020 0.905 0.35 0.3168 0.096
1.90 0.95 0.2 25 50 0.342 0.3290
0.8 30 65 0.316 0.3583 0.965 0.32 0.3088 0.111
2.22 0.98 0.2 30 49 0.416 0.3875
0.8 30 57 0.359 0.4310 1.000 0.28 0.2800 0.121
2.50 1.02 0.2 40 53 0.511 0.4745
0.8 40 62 0.438 0.4903 1.050 0.30 0.3150 0.154
2.80 1.08 0.2 40 49 0.552 0.5060
0.8 40 59 0.46 0.5310 1.105 0.30 0.3315 0.176
3.10 1.13 0.2 40 43 0.628 0.5560
0.8 40 56 0.484 0.5683 1.155 0.30 0.3465 0.197
3.40 1.18 0.2 50 52 0.648 0.5805
0.8 50 66 0.513 0.6008 1.200 0.30 0.3600 0.216
3.70 1.22 0.2 50 49 0.688 0.6210
0.8 50 61 0.554 0.6120 1.175 0.30 0.3525 0.216
4.00 1.13 0.2 50 51 0.661 0.6030
0.8 50 62 0.545 0.5893 1.105 0.25 0.2763 0.163
4.25 1.08 0.2 50 55 0.614 0.5755
0.8 50 63 0.537 0.5643 1.025 0.25 0.2563 0.145
4.50 0.97 0.2 40 47 0.575 0.5530
0.8 40 51 0.531 0.4740 0.945 0.27 0.2552 0.121
4.77 0.92 At vertical wall: 0.553 (0.65/0.91) = 0.3950 Totals: 4.50 3.9178 1.831

BIE 5300/6300 Lectures 53 Gary P. Merkley

 • The two tables below give sample current metering data and flow rate calculations
(one in Spanish, the other in English)
Aforo con Molinete en Sistema de Riego Chacuey
Ubicación: Inicio del canal principal Método: Tres Puntos
15 agosto 01 08h30 Sección: Rectangular, de concreto

Distancia desde Profundidad Fracción de Velocidad (m/s) Profundidad Ancho Area Caudal
el inicio Profundidad del punto promedio de promedio de Promedio
(m) (m) los puntos subsección (m) (m) (m2) (m3/s)
derecha pared vertical: (0.93)(0.65)/0.893=
0.00 0.88 0.68 0.68

0.80 0.88 0.20 0.176 0.141

0.20 0.88 0.2 1.00
0.6 0.92 0.93
0.8 0.87 0.93 0.88 0.20 0.176 0.163
0.40 0.88 0.2 1.01
0.6 0.92 0.92
0.8 0.84 0.93 0.88 0.20 0.176 0.163
0.60 0.88 0.2 0.99
0.6 0.94 0.93
0.8 0.84 0.94 0.88 0.20 0.176 0.165
0.80 0.88 0.2 1.01
0.6 0.95 0.94
0.8 0.86 0.90 0.88 0.20 0.176 0.158
1.00 0.88 0.2 0.89
0.6 0.87 0.85
0.8 0.76 0.73 0.88 0.18 0.158 0.116
1.18 0.88 0.62 0.62
izquierda pared vertical: (0.85)(0.65)/0.886=
Totales: 1.18 1.038 0.905

Logan & Northern Canal Company

Location: Upstream of Parshall flume Method: Two-point & six-tenths
09 Aug 01 10h30 Section: Earthen

Distance Depth Depth Velocity (m/s) Average Width Area Flow

from edge Fraction point avg of avg of Depth Rate
(m) (m) points subsection (m) (m) (m2) (m3/s)

0.000 0.000 0.02 0.02

0.10 0.200 0.61 0.122 0.012
0.610 0.400 0.6 0.18 0.18
0.23 0.488 0.61 0.297 0.067
1.219 0.575 0.2 0.29 0.27
0.8 0.25 0.30 0.638 0.61 0.389 0.115
1.829 0.700 0.2 0.35 0.32
0.8 0.30 0.31 0.755 0.61 0.460 0.143
2.438 0.810 0.2 0.35 0.30
0.8 0.25 0.30 0.825 0.61 0.503 0.149
3.048 0.840 0.2 0.32 0.29
0.8 0.27 0.28 0.815 0.61 0.497 0.138
3.658 0.790 0.2 0.29 0.26
0.8 0.23 0.23 0.743 0.61 0.453 0.103
4.267 0.695 0.2 0.20 0.19
0.8 0.18 0.18 0.598 0.61 0.364 0.065
4.877 0.500 0.6 0.16 0.16
0.14 0.378 0.61 0.230 0.031
5.486 0.255 0.6 0.11 0.11
0.06 0.128 0.61 0.078 0.005
6.096 0.000 0.01 0.01

Totals: 6.10 3.392 0.829

Gary P. Merkley 54 BIE 5300/6300 Lectures

References & Bibliography

Sontek. 2003. 6837 Nancy Ridge Dr., Suite A, San Diego, CA. www.sontek.com
USBR. 1997. Water Measurement Manual. U.S. Government Printing Office, Denver, CO.

BIE 5300/6300 Lectures 55 Gary P. Merkley

Gary P. Merkley 56 BIE 5300/6300 Lectures
Lecture 5
Field Exercise for Flume Calibration

I. Introduction

• In this field exercise, we will check the dimensions of a flow measurement flume,
also applying the:

1. Observation method;
2. Float method;
3. Uniform flow method; and,
4. Dye method.

• You will write this up as a homework exercise with the following sections:

1. Date, location, participant names;

2. Introduction (describe what was done);
3. Procedure;
4. Data analysis;
5. Summary & conclusions; and,
6. References and or bibliography.

• Include a few digital photographs in the report (we will bring a camera)
• You may turn the report in by groups, if desired, but everyone in the group must
contribute significantly to the work

II. Field Activities (Procedure)

• Dress appropriately for field work

• Guess (observation) the flow rate by looking at the channel and or flume
• Use the “float method” to estimate the flow rate
• Use the “dye method” to estimate the flow rate
• Take several dimensional and elevational measurements on the flume, including
the water surface elevations
• Notice if the flume is operating under free or submerged conditions
• Elevational measurements should include at least five points on the upstream
floor of the flume, and the top of the flume walls
• Measure the channel bed elevations downstream of the flume at 15 – 20 ft
distance intervals for a total distance of at least 300 ft
• Measure data to define the channel cross section, downstream of the flume, at
two locations (try to get representative locations)
• Estimate the Manning roughness value
• Write down any notes or observations which might be relevant to the work and
the flow measurement results

BIE 5300/6300 Lectures 57 Gary P. Merkley

Gary P. Merkley 58 BIE 5300/6300 Lectures
Lecture 6
Field Exercise for Current Metering

I. Introduction

• In this field exercise, we will do current metering in a canal, at the same locations
as the flume from the previous field work
• We will use three current meters:

1. A Price current meter;

2. A Marsh-McBirney electromagnetic current meter; and,
3. An Ott-type current meter.

• The results of the current metering will be compared to the known calibration of a
measurement flume
• You will write this up as a homework exercise with the following sections:

1. Date, location, participant names;

2. Introduction (describe what was done);
3. Procedure;
4. Data analysis;
5. Summary & conclusions; and,
6. References and or bibliography.

• Include a few digital photographs in the report (we will bring a camera)
• You may turn the report in by groups, if desired, but everyone in the group must
contribute significantly to the work

II. Field Activities (Procedure)

• Dress appropriately for field work

• Use the current meters to measure the flow rate
• Measure the upstream & downstream depths at the flume, including the water
surface elevations

BIE 5300/6300 Lectures 59 Gary P. Merkley

Gary P. Merkley 60 BIE 5300/6300 Lectures
Lecture 7
Weirs for Flow Measurement

I. Introduction

• Weirs are overflow structures built Suppressed rectangular weir

across open channels to measure
the volumetric rate of water flow
• The crest of a measurement weir is
usually perpendicular to the
direction of flow
• If this is not the case, special
calibrations must be made to
develop a stage discharge
relationship
• Oblique and “duckbill” weirs are
sometimes used to provide nearly
constant upstream water depth, but
they can be calibrated as measurement devices

Duckbill weir

• Some general terms pertaining to weirs are:

notch....... the opening which water flows through

crest....... the edge which water flows over
nappe....... the overflowing sheet of water
length....... the “width” of the weir notch

BIE 5300/6300 Lectures 61 Gary P. Merkley

II. Advantages and Disadvantages of Weirs

Advantages

1. Capable of accurately measuring a wide range of flows

2. Tends to provide more accurate discharge ratings than flumes and orifices
3. Easy to construct
4. Can be used in combination with turnout and division structures
5. Can be both portable and adjustable
6. Most floating debris tends to pass over the structure

Disadvantages

1. Relatively large head required, particularly for free flow conditions. This
precludes the practical use of weirs for flow measurement in flat areas.
2. The upstream pool must be maintained clean of sediment and kept free of
weeds and trash, otherwise the calibration will shift and the measurement
accuracy will be compromised

III. Types of Weirs

• Weirs are identified by the shape of their opening, or notch

• The edge of the opening can be either sharp- or broad-crested

(1) Sharp-crested weir

• A weir with a sharp upstream corner, or edge, such that the water
springs clear of the crest
• Those most frequently used are sharp-crested rectangular, trapezoidal,
Cipoletti, and triangular or 90° V-notch weirs
• According to the USBR, the weir plate thickness at the crest edges
should be from 0.03 to 0.08 inches
• The weir plate may be beveled at the crest edges to achieve the
necessary thickness

(2) Broad-crested weir

• A weir that has a horizontal or nearly horizontal crest sufficiently long in

the direction of flow so that the nappe will be supported and hydrostatic
pressures will be fully developed for at least a short distance
• Broad-crested weirs will be covered in detail later in the course
• Some weirs are not sharp- nor broad-crested, but they can be
calibrated for flow measurement

Gary P. Merkley 62 BIE 5300/6300 Lectures

Weirs may also be designed as suppressed or contracted

(1) Suppressed weir

• A rectangular weir whose notch (opening) sides are coincident with the
sides of the approach channel, also rectangular, which extend
unchanged downstream from the weir
• It is the lateral flow contraction that is “suppressed”

(2) Contracted weir

• The sides and crest of a weir are far away from the sides and bottom
of the approach channel
• The nappe will fully contract laterally at the ends and vertically at the
crest of the weir
• Also called an “unsuppressed” weir
• Calibration is slightly more complex than for a suppressed weir

IV. Types of Flow

(1) Free flow

• Also called “modular” flow, is a condition in which the nappe discharges

into the air
• This exists when the downstream water surface is lower than the lowest
point of the weir crest elevation
• Aeration is automatic in a contracted weir
• In a suppressed weir the sides of the structure may prevent air from
circulating under the nappe, so the underside of the nappe should be
vented (if used for flow measurement)
• If not vented, the air beneath the nappe may be exhausted, causing a
reduction of pressure beneath the nappe, with a corresponding increase in
discharge for a given head

(2) Submerged flow

• Also referred to as “non-modular” flow, is a condition in which the

discharge is partially under water, where changes in the downstream
depth will affect the flow rate
• A condition which indicates the change from free-flow to submerged-flow
is called transition submergence, where submergence is defined as the
ratio of downstream to upstream specific energy (Ed/Eu)

• For practical application of weirs as flow measurement devices, it is preferable

that they operate under free-flow conditions so that only the upstream depth
need be measured to arrive at a discharge value

BIE 5300/6300 Lectures 63 Gary P. Merkley

 • The calibration of free-flow weirs is more accurate that the calibration of
submerged-flow weirs

V. Approach Velocity and Gauge Location

• Large errors in flow measurement can occur because of poor flow conditions,
high-velocity and turbulence in the area just upstream of weir
• In general, the approaching flow should be the same as the flow in a long,
straight channel of the same size
• The upstream section of channel is sometimes called the “weir pool”
• For best flow measurement accuracy, the velocity of approach to a weir should
be less than 0.5 fps, or about 0.15 m/s
• This value is approximately obtained by dividing the maximum discharge by the
product of channel width and water depth (for a rectangular channel section),
which measured at the upstream point 4 to 6 times the weir head
• This point is the preferred staff gauge location upstream of the weir
• A tranquil flow condition should extend upstream from the weir a distance of 15
to 20 times the head on the weir
• The weir pool can be a wide channel section just upstream, thereby obtaining a
sufficiently low approach velocity
• Never place a weir in an open-channel reach with supercritical flow; a hydraulic
jump will form upstream and the water surface at the weir will not be tranquil

You can install a weir in a supercritical channel and a hydraulic

jump will occur upstream of the weir, but there will be too much
turbulence (unless the sill is very high). Always check the
upstream Froude number in weir designs.

VI. Guidelines for Designing & Operating Weirs

1. The weir should be set at the lower end of a long pool sufficiently wide and deep
to give an even, smooth flow
2. The centerline of the weir notch should be parallel to the direction of the flow
3. The face of the weir should be vertical, not leaning upstream nor downstream
4. The crest of the weir should be level, so the water passing over it will be of the
same depth at all points along the crest (does not apply to V-notch weirs, but the
centerline of the V-notch opening should be vertical)
5. The upstream edge should be sharp so that the nappe touches the crest only at
the leading (upstream) edge
6. Ideally, though not always practical, the height of the crest above the bottom of
the pool, P, should be at least three times the depth of water flowing over the
weir crest (check this condition for the maximum flow rate) – note that some
calibrations do not have this restriction, as described below
7. The sides of the pool should be at a distance from the sides of the crest not less
than twice the depth of the water passing over the crest (for unsuppressed
rectangular weirs):

Gary P. Merkley 64 BIE 5300/6300 Lectures

 ⎛B−L⎞
⎜ 2 ⎟ > 2hu (1)
⎝ ⎠

8. For accurate measurements the depth over the crest should be no more than
one-third the length of the crest
9. The depth of water over the crest should be no less than two inches (50 mm), as
it is difficult to obtain sufficiently accurate depth readings with smaller depths
10. The crest should be placed high enough so water will fall freely below the weir,
leaving an air space under the over-falling sheet of water. If the water below the
weir rises above the crest, this free fall is not possible, and the weir is then
operating under submerged-flow conditions.
11. To prevent erosion by the falling and swirling water, the channel downstream
from the weir should be protected by loose rock or by other material
12. You can assume that the discharge measurement accuracy of a sharp-crested
weir under free-flow conditions is within ±2% under the best field conditions
13. Don’t design a weir in which the minimum measurable flow rate is less than 2%
of the maximum flow rate, because you will not be able to accurately measure
such small flows.

BIE 5300/6300 Lectures 65 Gary P. Merkley

 • Note that it is not always possible to achieve the above guidelines when using
sharp-crested weirs for flow measurement in open channels
• But some things can be compensated for, such as an approach velocity which is
greater than 0.5 fps (0.15 m/s), as described below
• Also, the P > 3hu restriction is not always necessary (e.g. the Ce graphs below
have hu/P up to a value of 2.4)
• As the ratio of P/hu decreases, the calculated flow rate over the weir is
increasingly underestimated
• Never let P < hu unless you are prepared to develop a custom calibration

VII. Derivation of the Free-Flow Weir Equations

• An equation for accurately describing the head-discharge relationship over a weir

under free-flow conditions cannot be derived purely from theoretical
considerations assuming one-dimensional flow
• Theoretical calibrations can be derived based on 3-D flow analysis and a few
assumptions, but so far this can only be done with models using numerical
approximations
• In terms of one-dimensional flow, the Bernoulli equation can be written from a
point upstream of the weir to the crest location, as follows:

Vu2 V2
ht = h + = Cht + hL + v (2)
2g 2g

• Solving for the mean flow velocity at the vena contracta, Vv,

Gary P. Merkley 66 BIE 5300/6300 Lectures

 Vv = 2g ht (1 − C) − hL (3)
• Taking the liberty to combine some terms,

Vv ≈ C' 2ght (4)

• From continuity, Q = AvVv, and expressing the area of the vena contracta in
terms of the weir opening, Av = CcA, where Cc is the contraction coefficient,

Q = Cc A C' 2ght (5)

• Letting Cd = Cc C' 2g ,
Q = Cd A h t (6)

• For a horizontal-crested rectangular weir, A = hL. Therefore,

Q = Cd Lh ht ≈ CdLh3 / 2 (7)

• For a V-notch weir, A = h2 tan(θ/2), and,

⎛θ⎞ ⎛θ⎞
Q = Cd tan ⎜ ⎟ h2 ht ≈ Cd tan ⎜ ⎟ h5 / 2 (8)
⎝2⎠ ⎝2⎠

θ
• Letting Cdv = Cd tan ,
2

Q = Cdvh5 / 2 (9)

• For field calibrations it is useful to apply Eq. 7 for rectangular weirs and Eq. 9 for
triangular weirs
• These coefficients will include the effects of approach velocity, nappe shape, weir
opening contraction, and head loss
• Note that Eqs. 7 and 9 are dimensionally correct for either cfs or m3/s, given the
above definition for Cd
• Note also that Eq. 9 is of the same form as the free-flow calibration equation for
nonorifice open-channel constrictions
• The general form of Eq. 9 can be used to calibrate most weirs, regardless of
whether they are sharp-crested or not, when both the coefficient and the
exponent on the “h” term are taken to be calibration parameters (based on field
or lab data)

BIE 5300/6300 Lectures 67 Gary P. Merkley

VIII. Sharp-Crested Rectangular Weirs

• A convenient method of including the variation in the velocity of approach and the
contraction of the water jet over the weir is to relate Cd to the ratio hu/P, where P
is the vertical distance from the upstream channel bed to the weir crest
• A larger discharge for a given hu would be passed when hu/P is large
• In other words, when hu/P is large, the influence of the vertical component is
relatively small, and there is less contraction
• This is done through a coefficient called “Ce”

Kindsvater and Carter (1957) weir equation, for Q in cfs:

Q = CeL ehe3 / 2 (10)

Le = L + KL (11)

he = hu + KH (12)

where Le = the effective weir length

L = the measured weir length
he = the effective head
hu = the measured head above the weir crest (ft)
Ce = the effective discharge coefficient
KH = a small correction to the measured head (ft)

For weirs with L/B = 1 (suppressed weirs)

(a) According to the Kindsvater and Carter tests:

hu
Ce = 3.22 + 0.40 (13)
P
for KH = 0.003 ft and KL = -0.003 ft, with Q in cfs and head in feet.

(b) According to the Bazin (1886) tests:

hu
Ce = 3.25 + 0.445 (14)
P
for KH = 0.012 ft and KL = 0, with Q in cfs and head in feet.

(c) According to the Schroder and Turner (1904-1920) tests:

hu
Ce = 3.21 + 0.45 (15)
P
Gary P. Merkley 68 BIE 5300/6300 Lectures
 for KH = 0.004 ft and KL = 0, with Q in cfs and head in feet.

(d) According to USBR tests:

hu
Ce = 3.22 + 0.44 (16)
P
for KH = 0.003 ft and KL = 0, with Q in cfs and head in feet.

• It is seen that Eqs. 13 through 16 will give very similar results

• Note also that KL is either zero or very small, and often negligible
• You can see that some of the above relationships were developed 100 years ago

For weirs with L/B < 1 (unsuppressed weirs)

• Equations 10 - 12 still apply in this case

• The contraction effect is to decrease the magnitude of the coefficient, Ce
• The relationship of Ce to the constriction ratio L/B can be found in figures (see
below) presented by Kindsvater and Carter (1957)
• The KH values remain the same (but multiply the respective KH values in Eqs. 13
- 16 by 0.3048 to use meters instead of feet)
• KL values can also be determined graphically (see below)

Sharp-crested, rectangular weirs, English units:

0.015 0.015 4.2 4.2

4.0 4.0
0.010 0.010 hu in ft, Q in cfs
3.8 3.8
1.0
=
L/B 0.9
KL (ft) 0.005 0.005 Ce 3.6 L/B
= 3.6
= 0.8
L/B
3.4 3.4
6
0.000 0.000 L/B = 0.
L/B = 0.4
3.2 L/B = 0.2 3.2

-0.005 -0.005 3.0 3.0

0 0.2 0.4 0.6 0.8 1.0 0 0.4 0.8 1.2 1.6 2.0 2.4
L/B hu /P

BIE 5300/6300 Lectures 69 Gary P. Merkley

 Sharp-crested, unsuppressed, rectangular weirs, metric units:
0.005 0.005

0.004 0.004

0.003 0.003

0.002 0.002
KL (m)
0.001 0.001

0.000 0.000

-0.001 -0.001

-0.002 -0.002
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
L/B

Note: suppression occurs at L/B = 1

Sharp-crested, unsuppressed, rectangular weirs, metric units:

2.30 2.30

2.25 2.25

2.20 2.20

2.15 hu in m, Q in m3/s 2.15

2.10 2.10

2.05 1.0 2.05

=
L/B
0 .9
Ce 2.00 L/B
=
2.00

1.95 = 0.8 1.95

L/B

1.90 1.90
0.7
L /B =
1.85 .6 1.85
L/B = 0

1.80 L/B = 0.5 1.80

L/B = 0.4
1.75 L/B = 0.3 1.75
L/B = 0.2

1.70 1.70
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
hu/P

Gary P. Merkley 70 BIE 5300/6300 Lectures

 • Observe that the abscissa scale in the above graph for Ce goes up to a maximum
of hu/P = 2.4, which exceeds the recommended maximum of 0.333, as discussed
previously in this lecture
• Nevertheless, the above calibration procedure allows for hu/P > 0.333

“B” for Rectangular Weirs in Non-rectangular Sections

• Note that rectangular-notch weirs in non-rectangular channel sections are always

unsuppressed
• When applying the above calibrations to rectangular weirs in non-rectangular
channel sections, let B equal the width of the upstream cross-section at the
elevation of the weir crest

Equations Instead of Graphs

• It may be more convenient to approximate the above graphical solutions for KL

and Ce by equations when applying the relationships on a computer or calculator
• A rational function fits the Ce lines in the above graph (in metric units):

⎛h ⎞
Ce = αce ⎜ u ⎟ + βce (17)
⎝P⎠
where Ce is for Q in m3/s, and,

⎛L⎞
βce = 1.724 + 0.04789 ⎜ ⎟ (18)
⎝B⎠
and,
⎛L⎞
−0.00470432 + 0.030365 ⎜ ⎟
αce = ⎝B⎠ (19)
2
⎛L⎞ ⎛L⎞
1 − 1.76542 ⎜ ⎟ + 0.879917 ⎜ ⎟
⎝B⎠ ⎝B⎠

• A combination of a straight line and a polynomial approximates the KL curve, for

KL in meters:

For 0 ≤ L/B ≤ 0.35:

⎛L⎞
KL = 0.002298 + 0.00048 ⎜ ⎟ (20)
⎝B⎠

For 0.35 < L/B ≤ 1.00:

BIE 5300/6300 Lectures 71 Gary P. Merkley

 4 3 2
⎛L⎞ ⎛L⎞ ⎛L⎞
KL = −0.10609 ⎜ ⎟ + 0.1922 ⎜ ⎟ − 0.11417 ⎜ ⎟
⎝B⎠ ⎝B⎠ ⎝B⎠ (21)
⎛L⎞
+0.028182 ⎜ ⎟ − 0.00006
⎝B⎠

where KL is in meters

References & Bibliography

Kindsvater and Carter (1957)

Gary P. Merkley 72 BIE 5300/6300 Lectures

Lecture 8
Weirs for Flow Measurement

I. Cipoletti Weirs

• The trapezoidal weir that is most

often used is the so-called Cipoletti
weir, which was reported in ASCE
Transactions in 1894
• This is a fully contracted weir in which
the notch ends (sides) are not
vertical, as they are for a rectangular
weir
• The effects of end contraction are
compensated for by this trapezoidal
notch shape, meaning that mathematical corrections for end contraction are
unnecessary, and the equation is simpler
• The side slopes of the notch are designed to correct for end contraction (as
manifested in a rectangular weir), splayed out at angle of 14° with the vertical, or
nearly 1 horizontal to 4 vertical (tan 14° ≈ 0.2493, not 0.25 exactly)
• Some researchers have claimed than the side slopes should be greater than 1:4
in order to eliminate the effects of end contraction
• The sloping sides provides the advantage of having a stable discharge coefficient
and true relationship of:

Q = CL h3 / 2 (1)

The discharge equation by Addison (1949) is:

⎛ ⎛2⎞ ⎞
Q = ⎜ 0.63 ⎜ ⎟ 2g ⎟ L h3 / 2 = Ccip L h3 / 2 (2)
⎝ ⎝3⎠ ⎠
where L is the weir length (equal
to the width of the bottom of the
crest, as shown above); and h is
the upstream head, measured
from the bottom (horizontal part)
of the weir crest

• The units for L & h are feet for Q

in cfs, with Ccip = 3.37
• The units for L & h are m for Q in
m3/s, with Ccip = 1.86

BIE 5300/6300 Lectures 73 Gary P. Merkley

 • Eq. 2 is of the same form as a rectangular sharp-crested weir
• Eq. 2 (right-most side) is simpler than that for unsuppressed rectangular and
triangular sharp-crested weirs because the coefficient is a simple constant (i.e.
no calibration curves are needed)

X. V-Notch Weirs

• Triangular, or V-notch, weirs are

among the most accurate open
channel constrictions for measuring
discharge
• For relatively small flows, the notch
of a rectangular weir must be very
narrow so that H is not too small
(otherwise the nappe clings to the
A small V-notch weir (for furrows)
downstream side of the plate)
• Recall that the minimum hu value for a rectangular weir is about 2 inches (50
mm)
• But with a narrow rectangular notch, the weir cannot measure large flows without
correspondingly high upstream heads
• The discharge of a V-notch weir increases more rapidly with head than in the
case of a horizontal crested weir (rectangular or trapezoidal), so for the same
maximum capacity, it can measure much smaller discharges, compared to a
rectangular weir

• A simplified V-notch equation is:

Q = Ch5/2 (3)

Gary P. Merkley 74 BIE 5300/6300 Lectures

 • Differentiating Eq. 3 with respect to h,

dQ 5 3 / 2
= Ch (4)
dh 2
• Dividing Eq. 4 by Eq. 3 and rearranging,

dQ 5 dh
= (5)
Q 2 h
• It is seen that the variation of discharge is around 2.5 times the change in head
for a V-notch weir
• Thus, it can accurately measure the discharge, even for relatively small flows
with a small head: h is not too small for small Q values, but you still must be able
to measure the head, h, accurately
• A rectangular weir can accurately measure small flow rates only if the length, L,
is sufficiently small, because there is a minimum depth value relative to the crest;
but small values of L also restrict the maximum measurable flow rate
• The general equation for triangular weirs is:

hu
⎛θ⎞
Q = Cd 2 2g tan ⎜ ⎟
⎝2⎠
∫ (hu − hx ) hx dh (6)
0

because,
dA = 2x dh (7)
x
= tan(θ / 2) (8)
hu − hx

dQ = Cd 2gh dA (9)

BIE 5300/6300 Lectures 75 Gary P. Merkley

 • Integrating Eq. 6:

8 ⎛θ⎞
Q = Cd 2g tan ⎜ ⎟ hu2.5 (10)
15 ⎝2⎠

• For a given angle, θ, and assuming a constant value of Cd, Eq. 10 can be
reduced to Eq. 3 by clumping constant terms into a single coefficient

• A modified form of the above equation was proposed by Shen (1981):

8 ⎛θ⎞
Q= 2g Ce tan ⎜ ⎟ he5 / 2 (11)
15 ⎝2⎠

where,
he = hu + K h (12)

• Q is in cfs for hu in ft, or Q is in m3/s for hu in m

• The Kh and Ce values can be obtained from the two figures below
• Note that Ce is dimensionless and that the units of Eq. 11 are L3T-1 (e.g. cfs,
m3/s, etc.)

Sharp-crested triangular (V-notch weirs):

• Shen (ibid) produced the following calibration curves based on hydraulic

laboratory measurements with sharp-crested V-notch weirs

• The curves in the two figures below can be closely approximated by the following
equations:

Kh ≅ 0.001⎡⎣ θ (1.395 θ − 4.296 ) + 4.135⎤⎦ (13)

for Kh in meters; and,

Ce ≅ θ ( 0.02286 θ − 0.05734 ) + 0.6115 (14)

for θ in radians
• Of course, you multiply a value in degrees by π/180 to obtain radians
• Some installations have an insertable metallic V-notch weir that can be placed in
slots at the entrance to a Parshall flume to measure low flow rates during some
months of the year

Gary P. Merkley 76 BIE 5300/6300 Lectures

 3
0.100

Kh (inches)
2 0.075
Kh (mm)
0.050
1
0.025

0
20 40 60 80 100
Notch Angle, θ (degrees)

0.60

0.59
Ce
0.58

0.57
20 40 60 80 100
Notch Angle, θ (degrees)

XI. Sutro Weir

• Sutro weirs have a varying cross-sectional shape with depth

• This weir design is intended to provide high flow measurement accuracy for both
small and large flow rates
• A Sutro weir has a flow rate that is linearly proportional to h (for free flow)
• A generalized weir equation can be written as:

Q f = k + αhβ (15)

where k = 0 for the V-notch and rectangular weirs, but not for the Sutro; and β is
as defined below:

BIE 5300/6300 Lectures 77 Gary P. Merkley

 • The Sutro weir functions like a rectangular weir for h ≤ d
• This type of weir is designed for flow measurement under free-flow conditions
• It is not commonly found in practice

XII. Submerged Flow over Weirs

Single Curve

• Villamonte (1947) presented the following from his laboratory results:

0.385
⎛ ⎛ h ⎞nf ⎞
Qs = Qf ⎜ 1 − ⎜ d ⎟ ⎟ = K sQ f (16)
⎜ ⎝ hu ⎠ ⎟
⎝ ⎠
• For hd ≤ 0, Ks = 1.0 and the flow is free
• For hd > hu, there will be backflow across the weir
• For hu = hd, the value of Qs becomes zero (this is logical)
• The value of Qf is calculated from a free-flow weir equation
• The exponent, nf, is that which corresponds to the free-flow equation (usually, nf
= 1.5, or nf = 2.5)
• The figure below shows that in applying Eq. 16, hu & hd are measured from the
sill elevation

• Eq. 16 is approximately correct, but may give errors of more than 10% in the
calculated flow rate, especially for values of hd/hu near unity

Gary P. Merkley 78 BIE 5300/6300 Lectures

 Multiple Curves

• Scoresby (1997) expanded on this approach, making laboratory measurements

which could be used to generate a family of curves to define the submerged-flow
coefficient, Ks
• The following is based on an analysis of the laboratory data collected by
Scoresby (ibid). The flow rate through a weir is defined as:

Q = K sCf LHnuf (17)

where Q is the flow rate; Cf and nf are calibration parameters for free-flow
conditions; L is the “length” of the crest; Hu is the total upstream hydraulic head
with respect to the crest elevation; and Ks is a coefficient for submerged flow, as
defined above. As before, the coefficient Ks is equal to 1.0 (unity) for free flow
and is less than 1.0 for submerged flow. Thus,

K s ≤ 1.0 (18)

• Below is a figure defining some of the terms:

2
V
2g EL
HG L

hu Hu

hd

flow
weir

5hu
• The coefficient Ks can be defined by a family of curves based on the value of
Hu/P and hd/Hu
• Each curve can be approximated by a combination of an exponential function
and a parabola
• The straight line that separates the exponential and parabolic functions in the
graph is defined herein as:

BIE 5300/6300 Lectures 79 Gary P. Merkley

 ⎛h ⎞
Ks = A ⎜ d ⎟ + B (19)
⎝ Hu ⎠

• The exponential function is:

β
⎛ h ⎞
Ks = α ⎜1− d ⎟ (20)
⎝ Hu ⎠

• The parabola is:

2
⎛h ⎞ ⎛h ⎞
Ks = a ⎜ d ⎟ + b ⎜ d ⎟ + c (21)
⎝ Hu ⎠ ⎝ Hu ⎠

• Below the straight line (Eq. 19) the function from Eq. 20 is applied
• And, Eq. 21 is applied above the straight line
• In Eq. 19, let A = 0.2 y B = 0.8 (other values could be used, according to
judgment and data analysis)
• In any case, A+B should be equal to 1.0 so that the line passes through the point
(1.0, 1,0) in the graph (see below).
1.0

0.9

0.8

0.7

0.6
0.2 < Hu/P < 2.0
0.20
K 0.5
0.40
0.60
0.4 0.80
1.00
0.3 1.20
1.40
1.60
0.2 1.80
2.00
0.1

0.0
-0.20 0.00 0.20 0.40 0.60 0.80 1.00
hd/Hu

• This curve is defined by Eq. 20, but the values of α and β depend on the value of
Hu/P
Gary P. Merkley 80 BIE 5300/6300 Lectures
 • The functions are based on a separate analysis of the laboratory results from
Scoresby (ibid) and are the following:

α = 0.24
FG H IJ + 0.76
HPK
t
(22)

FH I
β = 0.014 G J + 0.23
HPK
t
(23)

• The point at which the two parts of the curves join is calculated in the following:

F hI F hI
AG J + B = αG1 − J
β

(24)
HH Kt H HK t

• Defining a function F, equal to zero,

β
⎛h ⎞ ⎛ h ⎞
F = A ⎜ d ⎟ + B − α ⎜1− d ⎟ = 0 (25)
⎝ Hu ⎠ ⎝ Hu ⎠
β−1
∂F ⎛ h ⎞
= A + αβ ⎜ 1 − d ⎟ (26)
⎛h ⎞ ⎝ Hu ⎠
∂⎜ d ⎟
⎝ Hu ⎠

• With Eqs. 25 and 26, a numerical method can be applied to determine the value
of hd/Hu
• Then, the value of Ks can be determined as follows:

⎛h ⎞
Ks = A ⎜ d ⎟ + B (27)
⎝ Hu ⎠

• The resulting values of hd/Hu and Ks define the point at which the two parts of the
curves join together on the graph

XIII. Overshot Gates

• So-called “overshot gates” (also known as “leaf gates,” “Obermeyers,”

“Langeman,” and other names) are weirs with a hinged base and an adjustable
angle setting (see the side-view figure below)
• Steel cables on either side of the gate leaf are attached to a shaft above and
upstream of the gate, and the shaft rotates by electric motor to change the
setting
• At large values of the angle setting the gate behaves like a weir, and at lower
BIE 5300/6300 Lectures 81 Gary P. Merkley
 angles it approximates a free overfall (but this distinction is blurred when it is
recognized that these two conditions can be calibrated using the same basic
equation form)
• These gates are manufactured by the Armtec company (Canada), Rubicon
(Australia), and others, and are easily automated
• The figure below shows an overshot gate operating under free-flow conditions

hu
L
P
θ
hinge
• The calibration equations presented below for overshot gates are based on the
data and analysis reported by Wahlin & Replogle (1996)
• The representation of overshot gates herein is limited to rectangular gate leafs in
rectangular channel cross sections, whereby the specified leaf width is assumed
to be the width of the cross section, at least in the immediate vicinity of the gate;
this means that weir end contractions are suppressed

• The equation for both free and submerged flow is:

2 2g
Q = K s Ca Ce Gwh1.5
e (28)
3
where Q is the discharge; θ is the angle of the opening (10° ≤ θ ≤ 65°), measured
from the horizontal on the downstream side; Gw is the width of the gate leaf; and
he is the effective head

• The effective head is defined as: he = hu + KH, where KH is equal to 0.001 m, or

0.0033 ft
• KH is insignificant in most cases
• For θ = 90°, use the previously-given equations for rectangular weirs
• For he in m, Q is in m3/s; for he in ft, Q is in cfs
• The calibration may have significant error for opening angles outside of the
specified range
• The coefficient Ce is a function of θ and can be approximated as:

Gary P. Merkley 82 BIE 5300/6300 Lectures

 ⎛h ⎞
Ce = 0.075 ⎜ u ⎟ + 0.602 (29)
⎝P⎠
where P is the height of the gate sill with respect to the gate hinge elevation (m
or ft)

• The value of P can be calculated directly based on the angle of the gate opening
and the length of the gate leaf (P = L sinθ, where L is the length of the gate)
• The coefficient Ca is a function of the angle setting, θ, and can be adequately
described by a parabola:

Ca = 1.0333 + 0.003848 θ − 0.000045 θ2 (30)

where θ is in degrees

• The submerged-flow coefficient, Ks, is taken as defined by Villamonte (1947), but

with custom calibration parameters for the overshot gate type.

C2
⎡ ⎛ h ⎞1.5 ⎤
K s = C1 ⎢1 − ⎜ d ⎟ ⎥ (31)
⎢⎣ ⎝ hu ⎠ ⎥⎦

where,
C1 = 1.0666 − 0.00111 θ for θ < 60
(32)
C1 = 1.0 for θ ≥ 60

and,
C2 = 0.1525 + 0.006077 θ − 0.000045 θ2 (33)

in which θ is in degrees

• The submerged-flow coefficient, Ks, is set equal to 1.0 when hd ≤ 0

• See the figure below for an example of an overshot gate with submerged flow

BIE 5300/6300 Lectures 83 Gary P. Merkley

 hu hd

P
θ

XIV. Oblique and Duckbill Weirs

• What about using oblique or duckbill weirs for flow measurement?

• The problem is that with large L values, the hu measurement is difficult because
small ∆h values translate into large ∆Q
• Thus, the hu measurement must be extremely accurate to obtain accurate
discharge estimations

flow

duckbill weir oblique weir inverse duckbill weir

XV. Approach Velocity

• The issue of approach velocity was raised above, but there is another standard
way to compensate for this
• The reason this is important is that all of the above calibrations are based on
zero (or negligible) approach velocity, but in practice the approach velocity may
be significant
• To approximately compensate for approach velocity, one approach (ha ha!)
method is to add the upstream velocity head to the head term in the weir
equation
• For example, instead of this…

nf
Qf = Cf ( hu ) (34)

…use this (where V is the mean approach velocity, Q/A):

Gary P. Merkley 84 BIE 5300/6300 Lectures

 nf
⎛ V2 ⎞
Qf = Cf ⎜ hu + ⎟⎟ (35)
⎜ 2g
⎝ ⎠
or,
nf
⎛ Q2f ⎞
Qf = Cf ⎜ hu + ⎟ (36)
⎜ 2gA 2 ⎟⎠
⎝

which means it is an iterative solution for Qf, which tends to complicate matters a
lot, because the function is not always well-behaved

• For known hu and A, and known Cf and nf, the solution to Eq. 36 may have
multiple roots; that is, multiple values of Qf may satisfy the equation (e.g. there
may be two values of Qf that are very near each other, and both positive)
• There may also be no solution (!*%&!#@^*) to the equation
• Conclusion: it is a logical way to account for approach velocity, but it can be
difficult to apply

XVI. Effects of Siltation

• One of the possible flow measurement errors is the effect of siltation upstream of
the weir
• This often occurs in a canal that carries a medium to high sediment load
• Some weirs have underflow gates which can be manually opened from time to
time, flushing out the sediment upstream of the weir
• The effect is that the discharge flowing over the weir can be increased due to a
higher upstream “apron”, thus producing less flow contraction
• The approximate percent increase in discharge caused by silting in front of a
rectangular weir is given below:

Percent Increase in Discharge

X/W
P/W 0 0.5 1.0 1.5 2.0 2.5
0.00 10% 13% 15% 16% 16%
0.25 5% 8% 10% 10% 10%
zero

0.50 3% 4% 5% 6% 6%
0.75 1% 2% 2% 3% 3%
1.00 zero

BIE 5300/6300 Lectures 85 Gary P. Merkley

 hu

P
W

X
• W is the value of P when there is no sediment deposition upstream of the weir
• X is the horizontal distance over which the sediment has been deposited
upstream of the weir – if X is very large, use the top of the sediment for
determining P, and do not make the discharge correction from the previous table
• The reason for the increase in discharge is that there is a change in flow lines
upstream of the weir
• When the channel upstream of the weir becomes silted, the flow lines tend to
straighten out and the discharge is higher for any given value of hu

References & Bibliography

Addison. 1949.
Kindsvater and Carter. 1957.
Flinn, A.D., and C.W.D. Dyer. 1894. The Cipoletti trapezoidal weir. Trans. ASCE, Vol. 32.
Scoresby, P. 1997. Unpublished M.S. thesis, Utah State Univ., Logan, UT.
Wahlin T., and J. Replogle. 1996.

Gary P. Merkley 86 BIE 5300/6300 Lectures

Lecture 9
Broad Crested Weirs

I. Introduction

• The broad-crested weir is an open-channel flow measurement device which

combines hydraulic characteristics of both weirs and flumes
• Sometimes the name “ramp flume” is used in referring to broad-crested weirs
• As with related open-channel measurement devices, the broad-crested weir has an
upstream converging section, a throat section, and a downstream diverging section
• The broad-crested weir can be calibrated for submerged flow conditions; however,
it is desirable to design this device such that it will operate under free-flow
conditions for the entire range of discharges under which it is intended to function

• When operating under free-flow conditions, critical flow will occur over the crest
(sill), and the discharge is uniquely related to the upstream flow depth – in this
case, the downstream conditions do not affect the calibration
• The broad-crested weir can be calibrated in the field or laboratory; however, a
major advantage of the structure is that it can be accurately calibrated based on
theoretical equations without the need for independent laboratory measurements

• The flow depth upstream of the measurement structure must always be higher
than it would be in the absence of the structure because there is always some
head loss
• Downstream of the structure the depth will not be affected; so, the required head
loss is manifested (in one way) as an increase in the upstream depth

BIE 5300/6300 Lectures 87 Gary P. Merkley

 hu
hc hd
zu zd

hu 2zu - 3zu L
II. Transition Submergence

• The typical transition submergence ranges for modular flow are:

Parshall flume ....................... 58 to 80%

Cutthroat flume ..................... 55 to 88%
Broad-crested weir................ 70 to 95%

• This means that the broad-crested weir can usually function as a free-flow
measurement device with less increase in the upstream water depth, which can be
a significant advantage

III. Advantages and Disadvantages

Advantages

1. the design and construction of the structure is simple, thus it can be relatively
inexpensive to install
2. a theoretical calibration based on post-construction dimensions can be obtained,
and the accuracy of the calibration is such that the discharge error is less that two
percent (this is assuming correct design and installation of the structure)
3. as with other open-channel flow measurement structures operating under free-flow
(modular) conditions, a staff gauge which is marked in discharge units can be
placed upstream; this allows a direct reading of the discharge without the need for
tables, curves, or calculators
4. the head loss across the structure is usually small, and it can be installed in
channels with flat slopes without greatly affecting existing upstream flow depths
5. floating debris tends to pass over and through the structure without clogging

Disadvantages

1. for water supplies with sediment, there will be deposition upstream of the structure
2. the upstream water depth will be somewhat higher than it was without the structure
3. farmers and other water users tend to oppose the installation of this structure
because they believe that it significantly reduces the channel flow capacity.
Although this is a false perception for a correctly designed broad-crested weir, it
Gary P. Merkley 88 BIE 5300/6300 Lectures
 does represent an important disadvantage compared to some other flow
measurement devices

IV. Site Selection

• The channel upstream of the broad-crested weir should be fairly straight and of
uniform cross-section
• The flow regime in the upstream section should be well into the subcritical range so
that the water surface is stable and smooth (Fr2 < 0.20, if possible). For this reason
it is best to avoid locating the structure just downstream of a canal gate or turnout,
for example, because the water surface is often not stable enough for an accurate
staff gauge reading
• The use of a stilling well and float assembly (or other water level sensing device) to
measure water level can partially compensate for fluctuating water levels, although
it involves additional cost
• Preferably, there are no gates or channel constrictions downstream of the structure
which would cause non-modular flow
• In fact, it is desirable to locate the structure just upstream of an elevation drop if
possible
• The presence of adjustable gates downstream complicates the design even more
than for fixed constrictions because the depth will depend on both discharge and
gate setting
• Other factors involved in the site selection are the stability of the channel bed and
side slopes in the upstream direction (in the case of earthen canals), and the
accessibility for measurement readings and maintenance
• If the upstream channel is not stable, the calibration may change significantly, and
sediment can accumulate rapidly at the structure, also affecting the calibration

V. Design Considerations

• One of the important advantages of the broad-crested weir is that it can be

accurately calibrated according to theoretical and empirical relationships
• This means that it is not necessary to install "standard" structure sizes and rely on
laboratory calibration data
• The ability to calibrate the structure using equations instead of measurements is
based on the existence of parallel streamlines in the control section over the crest
• In many other open-channel flow measurement devices the streamlines are not
straight and parallel in the control section, and although a theoretical calibration
would be possible, it requires complex hydraulic modeling
• On the other hand, theoretical calibration of the broad-crested weir is relatively
simple
• The broad-crested weir should be located and dimensioned so that the flow is
modular over the full operating range of the device
• If there is a significant drop in the channel bed immediately downstream of the
structure, then the height of the crest may not be important in achieving critical
depth

BIE 5300/6300 Lectures 89 Gary P. Merkley

 • However, the relative dimensions of the structure are important to obtain
"favorable" flow conditions over the crest, that is, flow conditions which conform to
the inherent assumptions for accurate theoretical calibration
• Thus, the height and length of the crest are important dimensions with relation to
the upstream flow depth
• In any case, adequate design of the structure dimensions is essentially a process
of trial-and-error, and therefore can be greatly facilitated through use of a calculator
or computer program

Sill Height

• One of the most important design parameters is the height of the sill above the
upstream channel bed
• This height should be sufficient to provide modular flow for the entire range of
discharges that the broad-crested weir is intended to measure; however, it should
not be higher than necessary because this would cause undue increases in the
upstream water level after installation
• Thus, a design objective is to determine the minimum crest height for which
modular flow can be obtained, and not to exceed this minimum height
• Excessively tall broad-crested weirs are not a problem in terms of water
measurement or calibration, they are only troublesome with respect to
unnecessarily raising the upstream water level
• The lower limit on sill height is based on the Froude number in the upstream
channel section (Fr2 < 0.20)

Upstream and Downstream Ramps

• The converging upstream ramp should have a slope of between 2:1 and 3:1 (H:V).
If flatter, the ramp is longer than necessary and there will be additional hydraulic
losses which detract from the calibration accuracy
• If the ramp is steeper than 2:1, unnecessary turbulence may be created in the
converging section, also causing addition head loss
• The diverging ramp at the downstream end of the crest should have a slope of
between 4:1 and 6:1 (H:V), or should be truncated (non-existent). The 6:1 ratio is
preferred in any case, and this same ratio is used in the diverging sections of other
flow measurement devices, in both open-channel and pipe flow, to minimize head
losses from turbulence
• If the 6:1 ratio causes an excessively long downstream ramp, then the length
should be abruptly truncated (see the figure below), not rounded off

DS ramp
flow truncated

Gary P. Merkley 90 BIE 5300/6300 Lectures

 • Many broad-crested weirs do not have a downstream ramp − the structure is
terminated with a vertical wall just downstream of the throat section. In many
cases, the energy that could be "recovered" by the inclusion of a downstream ramp
is not enough to justify the additional expense. Also, the benefits of a downstream
ramp are more significant in large broad-crested weirs (more than 1 m high)
• Most of the head loss across the structure occurs due to turbulence in the
diverging section, and in many cases the losses in the converging and throat
sections may be neglected in calibration calculations

hu
hc
zu

hu 2zu - 3zu L zd
hd

Lateral Flow Contraction

• The side slope in the throat section of the broad-crested weir is usually the same
as that in the upstream section, but it does not need to be the same
• In very wide and earthen channels it is common practice to reduce the width of the
throat section and design for a zero side slope (i.e. a rectangular section)
• When the side slope is reduced it is usually because the vertical flow contraction
obtained by the crest height is insufficient to induce modular flow conditions.
Therefore, in some cases lateral flow contraction is also required

Ratio of hu/L

• The ratio of upstream head to crest length is limited by a maximum of

approximately 0.75, and a minimum of approximately 0.075
• The lower limit is imposed to help maintain a reasonably small ratio of head loss to
total upstream head (relative to the crest elevation)
• The upper limit is meant to avoid a non-hydrostatic pressure distribution on the
crest
• The calibration procedure is valid only for horizontal, parallel streamlines in the
control section in the throat. When the ratio is between these limits, the theoretical
calibration should be accurate to within two percent of the actual discharge
• The ratio of upstream flow depth (referenced from the sill elevation) to the throat
length is approximately 0.5 for the average discharge over a correctly-designed
broad-crested weir
• The figure below shows dimensions for a sample BCW design
BIE 5300/6300 Lectures 91 Gary P. Merkley
VI. Modular Limit

• The modular limit is defined as the

ratio of specific energies in the
downstream and upstream sections
(Ed/Eu) at which transitional flow
exists
• The velocity heads in the upstream
and downstream sections will
normally be small compared to the
flow depths, so this ratio may be
approximated by the transition
submergence, St = hd/hu for Qf = Qs
• When the ratio is greater than the
modular limit, the structure is
submerged and the flow is non-
modular

Gary P. Merkley 92 BIE 5300/6300 Lectures

 • Under non-modular conditions the theoretical calibration is invalid since it assumes
that critical flow occurs somewhere over the crest in the throat section
• Field calibration of the structure for submerged-flow conditions is possible, but the
results will be less accurate and the structure's use as a measurement device will
be less convenient

Determining Downstream Depth

• For existing channels with a straight section downstream of the broad-crested weir,
and without hydraulic controls such as sluice gates, the value of hd can be
determined according to normal flow conditions. That is, for a given discharge, the
value of hd can be calculated using the Manning or Chezy equations
• In the case of a downstream control which causes a backwater effect at the broad-
crested weir, the issue becomes complicated since the actual submergence ratio
across the structure depends not only on the discharge, but also on the control
setting (which creates an M1 profile upstream toward the broad-crested weir). This
is a common situation because the water surface profile in most irrigation channel
reaches is affected by downstream flow control structures
• For this reason, it is preferable to have a drop in elevation immediately
downstream of the broad-crested weir, or to have a straight canal section without
any nearby control structures in the downstream direction

Energy Balance and Losses

• Given an upstream depth and its associated discharge (according to the

calibration), the value of downstream specific energy, Ed, for the modular limit is
calculated by subtracting estimated head losses from the upstream specific
energy, Eu
• These head losses include friction and turbulence across the broad-crested weir,
and equations exist to approximate the respective values according to cross-
sectional geometry, expansion ratios, and roughness coefficients (Bos, et al. 1984)
• Computer programs for developing the theoretical calibration contain these
equations; however, it is worth noting that the losses due to wall friction are very
minimal, and accurate estimation of roughness coefficients is not necessary for
calibration
• The majority of the losses occur due to the sudden expansion downstream of the
throat section, and these losses are either estimated or calculated (using empirical
relationships) depending on the downstream ramp dimensions

Calculating the Modular Limit

• The modular limit will vary according to discharge for a given installation, and its
calculation can be summarized as follows:

1. Given an upstream depth, channel cross-section, and weir calibration,

calculate the discharge (assuming modular flow), and add depth plus
velocity head to produce Eu
BIE 5300/6300 Lectures 93 Gary P. Merkley
 2. Calculate head losses due to wall friction in the converging and throat
sections, then estimate the expansion loss downstream of the throat, and
add these two values
3. Subtract the combined losses from Eu, giving the value of Ed. The modular
limit is, then, Ed/Eu
4. Determine the downstream depth for the given discharge. This can be done
using the Manning equation for uniform flow, or by another equation when a
backwater profile exists from a downstream control
5. Compare the calculated downstream water level with the value of hd, which
is equal to Ed minus the corresponding downstream velocity head. If the
calculated water level is less than or equal to hd, the flow will be modular
because the actual head loss is greater than that required for modular flow

Gary P. Merkley 94 BIE 5300/6300 Lectures

Lecture 10
Broad Crested Weirs

I. Calibration by Energy Balance

• The complete calibration of a broad-crested weir includes the calculation of head

losses across the structure. However, the calibration can be made assuming no
losses in the converging and throat sections, and the resulting values will usually
be very close to those obtained by the complete theoretical calibration
• The procedure which is presented below is useful to illustrate the hydraulic
principles which govern the broad-crested weir characteristics, and to check the
calibration of an existing structure in the field with a programmable calculator
• The simplified calibration approach does not include the calculation of the modular
limit; however, this is an important consideration in the design and operation of a
broad-crested weir because the structure is usually intended to operate under
modular flow conditions

(1) The specific energy of the flow upstream of the broad-crested weir can be set
equal to the specific energy over the crest (or sill) of the structure. The energy
balance can be expressed mathematically as follows:

Vu2 V2
hu + zu + = hc + zu + c (1)
2g 2g

where hu is the upstream flow depth, referenced from the sill elevation; Vu is the
average velocity in the upstream section, based on a depth of (hu + zu); zu is the
height of the sill above the upstream bed; hc is the depth over the crest where
critical flow is assumed to occur; and Vc is the average velocity in the critical flow
section over the crest.

Note that the zu term cancels from Eq. 1.

Recognizing that Q = VA, where Q is the volumetric flow rate (discharge) and A is
the area of the flow cross-section,

Q2 Q2
hc − hu + − =0 (2)
2g A c2 2g A u2

which can be reduced to,

Q2 ⎛ 1 1 ⎞
hc − hu + ⎜⎜ 2 − 2 ⎟⎟ = 0 (3)
2g ⎝ A c A u ⎠

BIE 5300/6300 Lectures 95 Gary P. Merkley

 (2) Critical flow over the crest can be defined by the Froude number, which is equal
to unity for critical flow. Thus, the square of the discharge over the crest can be
defined as follows:

g A 3c
Q2 = (4)
Tc

where Tc is the width of the water surface over the crest. This last equation for Q2
can be combined with the equation for energy balance to produce the following:

A 3c ⎛ 1 1 ⎞
hc − hu + ⎜⎜ 2 − 2 ⎟⎟ = 0 (5)
2 Tc ⎝ A c Au ⎠

This last equation can be solved by trial-and-error, or by any other iterative method,
knowing hu, zu, and the geometry of the upstream and throat cross-sections. The
geometry of the sections defines the relationship between hc and Ac, and between
hu and Au (important: if you look carefully at the above equations, you will see that
Au must be calculated based on a depth of hu + zu). The solution to Eq. 5 gives the
value of hc.

(3) The final step is to calculate the discharge corresponding to the value of Ac,
which is calculated directly from hc. This is done using the following form of the
Froude number equation:

g A 3c
Q= (6)
Tc

This process is repeated for various values of the upstream flow depth, and in the
end a table of values for upstream depth and discharge will have been obtained.
From this table a staff gauge can be constructed. This simple calibration assumes
that the downstream flow level is not so high that non-modular flow exists across
the structure.

• See the computer program listing on the following two pages

• In the design of broad-crested weirs it is often necessary to consider other factors
which limit the allowable dimensions, and which restrict the flow conditions for
which the calibration is accurate
• Complete details on broad-crested weir design, construction, calibration, and
application can be found in the book "Flow Measurement Flumes for Open
Channel Systems", 1984, by M.G. Bos, J.A. Replogle, y A.J. Clemmens

Gary P. Merkley 96 BIE 5300/6300 Lectures

 //-----------------------------------------------------------------------------
// Broad crested weir calibration for free flow by energy balance equation.
// Written in Object Pascal (Delphi 6) by Gary Merkley. September 2004.
//-----------------------------------------------------------------------------
unit BCWmain;

interface

uses
Windows, Messages, SysUtils, Classes, Graphics, Controls, Forms, Dialogs,
StdCtrls, Buttons;

type
TWmain = class(TForm)
btnStart: TBitBtn;
procedure btnStartClick(Sender: TObject);
private
function NewtonRaphson(hu:double):double;
function EnergyFunction(hc:double):double;
function Area(h,b,m:double):double;
function TopWidth(h,b,m:double):double;
end;

var
Wmain: TWmain;

implementation
{$R *.DFM}

const
g = 9.810; // weight/mass (m/s2)
bu = 2.000; // base width upstream (m)
mu = 1.250; // side slope upstream (H:V)
zu = 1.600; // upstream sill height (m)
bc = 6.000; // base width at control section (m)
mc = 1.250; // side slope at control section (H:V)
L = 1.500; // sill length (m)

var
hu,Au,hc: double;

function TWmain.NewtonRaphson(hu:double):double;
//-----------------------------------------------------------------------------
// Newton-Raphson method to solve for critical depth. Returns flow rate.
//-----------------------------------------------------------------------------
var
i,iter: integer;
dhc,F,Fdhc,change,Ac,Tc: double;
begin
result:=0.0;

for i:=1 to 9 do begin

hc:=0.1*i*hu;

for iter:=1 to 50 do begin

dhc:=0.0001*hc;
F:=EnergyFunction(hc);
Fdhc:=EnergyFunction(hc+dhc);
change:=Fdhc-F;
if abs(change) < 1.0E-12 then break;
change:=dhc*F/change;
hc:=hc-change;

if (abs(change) < 0.001) and (hc >= 0.001) then begin

Ac:=Area(hc,bc,mc);
Tc:=TopWidth(hc,bc,mc);
result:=sqrt(g*Ac*Ac*Ac/Tc);
BIE 5300/6300 Lectures 97 Gary P. Merkley
 Exit;
end;
end;
end;
end;

function TWmain.EnergyFunction(hc:double):double;
//-----------------------------------------------------------------------------
// Energy balance function (specific energy), equal to zero.
//-----------------------------------------------------------------------------
var
Ac,Tc: double;
begin
Ac:=Area(hc,bc,mc);
Tc:=TopWidth(hc,bc,mc);
result:=hc-hu+0.5*Ac*Ac*Ac*(1.0/(Ac*Ac)-1.0/(Au*Au))/Tc;
end;

function TWmain.Area(h,b,m:double):double;
//-----------------------------------------------------------------------------
// Calculates cross-section area for symmetrical trapezoidal shape.
//-----------------------------------------------------------------------------
begin
result:=h*(b+m*h);
end;

function TWmain.TopWidth(h,b,m:double):double;
//-----------------------------------------------------------------------------
// Calculates top width of flow for symmetrical trapezoidal shape.
//-----------------------------------------------------------------------------
begin
result:=b+2.0*m*h;
end;

procedure TWmain.btnStartClick(Sender: TObject);

//-----------------------------------------------------------------------------
// Entry point for calculations (user clicked the Start button).
//-----------------------------------------------------------------------------
var
i: integer;
F: TextFile;
strg: string;
Q,humin,humax: double;
begin
AssignFile(F,'BCWenergy.txt');
Rewrite(F);
Writeln(F,' hc (m) hu (m) Q (m3/s)');

humin:=0.075*L;
humax:=0.75*L;

hu:=humin;

for i:=0 to 100 do begin

Au:=Area(hu+zu,bu,mu);
Q:=NewtonRaphson(hu);
strg:=Format('%12.3f%12.3f%12.3f',[hc,hu,Q]);
Writeln(F,strg);

hu:=hu+0.03;
if hu > humax then break;
end;

CloseFile(F);
end;

end.

Gary P. Merkley 98 BIE 5300/6300 Lectures

II. Calculation of Head Loss

Throat (Control Section)

Head loss in the throat (where the critical flow control section is assumed to be
located) can be estimated according to some elements from boundary layer theory.
The equation is (Schlichting 1960):

CF L Vc 2
(hf )throat = (7)
2gR

where L is the length of the sill; Vc is the average velocity in the throat section; and R is
the hydraulic radius of the throat section. The values of V and R can be taken for
critical depth in the throat section. The drag coefficient, CF, is estimated by assuming
the sill acts as a thin flat plat with both laminar and turbulent flow, as shown in the
figure below (after Bos, Replogle and Clemmens 1984).

The drag coefficient is calculated by assuming all turbulent flow, subtracting the
turbulent flow portion over the length Lm, then adding the laminar flow portion for the
length Lm. Note that CF is dimensionless.

⎛m⎞ ⎛m⎞
CF = CT,L − ⎜ ⎟ CT,m + ⎜ ⎟ CL,m (8)
⎝L⎠ ⎝L⎠

where CL,m is the thin-layer laminar flow coefficient over the distance m, which
begins upstream of the weir crest:

1.328
CL,m = (9)
(Re )m
When (Re)L < (Re)m, the flow is laminar over the entire crest and CF = CL,L, where
CL,L is defined by Eq. 9.

The CT,L and CT,m coefficients are calculated by iteration from:

BIE 5300/6300 Lectures 99 Gary P. Merkley

 ⎡ ⎛ 1 k ⎞ ⎤
0.544 CT,x = CT,x ⎢5.61 CT,x − ln ⎜ + ⎟ − 0.638 ⎥ (10)
⎢⎣ ⎜ Re x CT,x 4.84x CT,x ⎟ ⎥⎦
⎝ ⎠

where x is equal to L or m, for CT,L and CT,m, respectively; Re is the Reynolds number;
and k is the absolute roughness height. All values are in m and m3/s. Below are some
sample values for the roughness, k.

Material Roughness, k
and Condition (mm)
Glass 0.001 to 0.01
Smooth Metal 0.02 to 0.1
Rough Metal 0.1 to 1.0
Wood 0.2 to 1.0
Smooth Concrete 0.1 to 2.0
Rough Concrete 0.5 to 5.0

The Reynolds number can be calculated as follows:

Vx
(R e ) x = (11)
ν

where ν is the kinematic viscosity (a function of temperature); x is equal to m or L; and

V is the average velocity in the throat section for critical flow. If (Re)L is less than (Re)m,
the boundary layer over the sill is laminar only, and CF = CL,m.

The value of m can be estimated as:

ν⎛ L⎞
m= ⎜ 350,000 + ⎟ (12)
V⎝ k⎠

where the units are (m2/s)/(m/s) = m

Diverging Section

The head loss in the downstream diverging section is estimated as:

2
ξ ( Vc − Vd )
(hf )ds = (13)
2g

where (hf)ds is the head loss in the diverging section (m); Vc is the average velocity in
the control section, at critical depth (m/s); and Vd is the average velocity in the
downstream section (m/s), using hd referenced to the downstream channel bed
elevation (not the sill crest).
Gary P. Merkley 100 BIE 5300/6300 Lectures
 The coefficient ξ is defined as:

log10 ⎡114.6 tan−1 ( zd / Ld ) ⎤ − 0.165

ξ= ⎣ ⎦ (14)
1.742
where Ld is the length of the downstream ramp; and zd is the height of the downstream
ramp, also equal to the difference in elevation between the sill and the downstream
bed elevation. Note that the recommended value for the zd/Ld ratio is 1/6.
If the downstream ramp is not used, Ld equals zero. In this case, assume ξ = 1.2.

Converging Section

After having calculated the CF drag coefficient for the throat section, and assuming the
same roughness value for the channel and structure from the gauge location to the
control section, the upstream losses can be estimated. These are the losses from the
gauge location to the beginning of the sill.

For the section from the gauge to the beginning of the upstream ramp, the head loss is
estimated as:

⎛ CF Lgauge ⎞ Vu2
(hf )gauge =⎜ ⎟ (15)
⎝ R u ⎠ 2g

where Lgauge is the distance from the gauge to the beginning of the upstream ramp; Vu
is the average velocity in the upstream section (at the gauge); and Ru is the hydraulic
radius at the gauge. All values are in metric units (m and m/s).

For the upstream ramp, the same equation can be used, but the hydraulic radius
changes along the ramp. Therefore,

1 ⎛ CF Lu ⎞ ⎛ Vu2 Vr2 ⎞
(hf )us = ⎜ ⎟⎜ + ⎟ (16)
2 ⎝ 2g ⎠ ⎜⎝ Ru Rr ⎟⎠

where the values of Vr and Rr, at the entrance to the throat section (top of the ramp),
are estimated by calculating the depth at this location:

hr = hc + 0.625 (hu − hc ) (17)

BIE 5300/6300 Lectures 101 Gary P. Merkley

III. Photographs of BCW Construction

Gary P. Merkley 102 BIE 5300/6300 Lectures

BIE 5300/6300 Lectures 103 Gary P. Merkley
Gary P. Merkley 104 BIE 5300/6300 Lectures
References & Bibliography

BIE 5300/6300 Lectures 105 Gary P. Merkley

Gary P. Merkley 106 BIE 5300/6300 Lectures
Lecture 11
Calibration of Canal Gates

I. Suitability of Gates for Flow Measurement

Advantages

• relatively low head loss

• often already exists as a control device
• sediment passes through easily

Disadvantages

• usually not as accurate as weirs

• floating debris tends to accumulate
• calibration can be complex for all flow conditions

II. Orifice and Non-Orifice Flow

• Canal gates can operate under orifice flow conditions and as

channel constrictions
• In general, either condition can occur under free or
submerged (modular or non-modular) regimes
• Orifice flow occurs when the upstream depth is sufficient to
“seal” the opening − in other words, the bottom of the gate is
lower than the upstream water surface elevation
• The difference between free and submerged flow for a gate operating as an orifice
is that for free flow the downstream water surface elevation is less than CcGo,
where Cc is the contraction coefficient and Go is the vertical gate opening,
referenced from the bottom of the gate opening (other criteria can be derived from
momentum principles)
• The distinguishing difference between free and submerged flow in a channel
constriction is the occurrence of critical velocity in the vicinity of the constriction
(usually a very short distance upstream of the narrowest portion of the
constriction)
• This lecture focuses on the calibration of gates under non-orifice flow conditions

BIE 5300/6300 Lectures 107 Gary P. Merkley

III. Rating Open Channel Constrictions

• Whenever doing structure

calibrations, examine the
data carefully and try to
identify erroneous values or
mistakes
• Check the procedures used
in the field or laboratory
because sometimes the
people who take the
measurements are not
paying attention and make
errors
• Never blindly enter data
into a spreadsheet or other
computer program and accept the results at “face value” because you may get
incorrect results and not even realize it
• Always graph the data and the calibration results; don’t perform regression and
other data analysis techniques without looking at a graphical representation of
the data and comparison to the results

Free Flow

• Gates perform hydraulically as open channel constrictions (non-orifice flow) when

the gate is raised to the point that it does not touch the water surface
• The general form of the free-flow equation is:

n
Qf = Cf hu f (1)

where the subscript f denotes free flow; Qf is the free-flow discharge; Cf is the
free-flow coefficient; and nf is the free-flow exponent

• The value of Cf increases as the size of the constriction increases, but the
relationship is usually not linear
• The value of nf is primarily dependent upon the geometry of the constriction with
the theoretical values being 3/2 for a rectangular constriction and 5/2 for a
triangular constriction

Sample Free-Flow Constriction Calibration

• Sample field data for developing the discharge rating for a rectangular open-
channel constriction are listed in the table below
• The discharge rate in the constriction was determined by taking current meter
readings at an upstream location, and again at a downstream location

Gary P. Merkley 108 BIE 5300/6300 Lectures

 • This is a good practice because the upstream and downstream flow depths are
often significantly different, so that the variation in the measured discharge
between the two locations is indicative of the accuracy of the current meter
equipment and the methodology used by the field staff

Date Discharge Water Surface Elevation

(m3/s) in Stilling Well (m)
21 Jun 86 0.628 409.610
21 Jun 86 1.012 409.935
21 Jun 86 1.798 410.508
21 Jun 86 2.409 410.899
Note: The listed discharge is the average discharge
measured with a current meter at a location 23 m upstream
of the constriction, and at another location 108 m
downstream.

• The free-flow equation for the flow depths measured below the benchmark (at
411.201 m) is:
1.55
Qf = 0.74 ( hu ) x (2)

Col. 1 Col. 2 Col. 3 Col. 4 Col. 5

Discharge Water Surface (hu)sw Tape Measurement (hu)x
(m3/s) Elevation (m) (m) (m) (m)
0.628 409.610 0.918 1.604 0.905
1.009 409.935 1.243 1.294 1.215
1.797 410.508 1.816 0.734 1.775
2.412 410.899 2.207 0.358 2.151
Notes: The third column values equal the values in column 2 minus the floor
elevation of 408.692 m. The values in column 5 equal the benchmark elevation of
411.201 m minus the floor elevation of 408.692 m, minus the values in column 4.

• The “Tape Measurement” in the above table is for the vertical distance from the
benchmark down to the water surface
• If a regression analysis is performed with the free-flow data using the theoretical
value of nf = 3/2,

1.5
Qf = 0.73 ( hu )sw (3)
or,
1.5
Qf = 0.75 ( hu ) x (4)

• The error in the discharge resulting from using nf = 3/2 varies from -1.91% to
+2.87%.

BIE 5300/6300 Lectures 109 Gary P. Merkley

 Submerged Flow

• The general form of the submerged-flow equation is:

nf
Cs ( hu − hd )
Qs = ns
(5)
( − logS )
where the subscript s denotes submerged flow, so that Qs is the submerged-flow
discharge, Cs is the submerged-flow coefficient, and ns is the submerged-flow
exponent

• Base 10 logarithms have usually been used with Eq. 5, but other bases could be
used, so the base should be specified when providing calibration values
• Note that the free-flow exponent, nf, is used with the term hu - hd
• Consequently, nf is determined from the free-flow rating, while Cs and ns must be
evaluated using submerged-flow data
• The theoretical variation in ns is between 1.0 and 1.5
• Note that the logarithm term in the denominator of Eq. 5 can be estimated by
taking the first two or three terms of an infinite series (but this is not usually
necessary):

1 2 1 3 1 4
loge (1 + x) = x − x + x − x +… (6)
2 3 4

Graphical Solution for Submerged-Flow Calibration

• The graphical solution used to be performed by hand on log-log paper before

PCs and programmable calculators became widely available
• This is essentially how the form of the submerged-flow equation was derived in
the 1960’s
• This solution technique assumes that the free-flow exponent, nf, is known from a
prior free-flow calibration at the same structure
• So, assuming you already know nf, and you have data for submerged flow
conditions, you can plot Qs versus (hu – hd), as shown in the figure below where
there are five measured points

Gary P. Merkley 110 BIE 5300/6300 Lectures

 • The above graph has a log-log scale
• The slope of the parallel lines is equal to nf, as shown above (measured using a
linear, not log, scale), and each line passes through one of the plotted data
points
• The five values of Qs at each of the horizontal dashed lines are for ∆h = 1.0
• Get another sheet of log-log paper and plot the five Q∆h = 1.0 values versus -
log10S, as shown in the figure below

• Draw a straight line through the five data points and extend this line to –log10S =
1.0 (as shown above)
• The value on the ordinate at –log10S = 1.0 is Cs
• The slope of the line is –ns (measured with a linear scale)
• You might not prefer to use this method unless you have no computer

BIE 5300/6300 Lectures 111 Gary P. Merkley

 Submerged-Flow Calibration by Multiple Regression

• Multiple regression analysis can be used to arrive directly at all three values (Cs,
nf, and ns) without free-flow data
• This can be done by taking the logarithm of Eq. 5 as follows:

logQs = logCs + nf log ( hu − hd ) − ns log ( − logS ) (7)

• Equation 7 is linear with respect to the unknowns Cs, nf, and ns

• Such a procedure may be necessary when a constriction is to be calibrated in the
field and only operates under submerged-flow conditions
• You can also compare the calculated nf values for the free- and submerged-flow
calibrations (they should be nearly the same)
• Spreadsheet applications and other computer software can be used to perform
multiple least-squares regression conveniently
• This method can give a good fit to field or laboratory data, but it tends to
complicate the calculation of transition submergence, which is discussed below

Sample Submerged-Flow Constriction Calibration

• In this example, a nearly constant discharge was diverted into the irrigation
channel and a check structure with gates located 120 m downstream was used
to incrementally increase the flow depths
• Each time that the gates were changed, it took 2-3 hours for the water surface
elevations upstream to stabilize
• Thus, it took one day to collect the data for a single flow rate
• The data listed in the table below were collected in two consecutive days

Discharge Tape Measurement Tape Measurement

Date
(m3/s) from U/S Benchmark (m) from D/S Benchmark (m)
22 Jun 86 0.813 1.448 1.675
22 Jun 86 0.823 1.434 1.605
22 Jun 86 0.825 1.418 1.548
22 Jun 86 0.824 1.390 1.479
22 Jun 86 0.793 1.335 1.376
23 Jun 86 1.427 0.983 1.302
23 Jun 86 1.436 0.966 1.197
23 Jun 86 1.418 0.945 1.100
23 Jun 86 1.377 0.914 1.009
23 Jun 86 1.241 0.871 0.910

Gary P. Merkley 112 BIE 5300/6300 Lectures

 Qs (hu)x (hd)x
S -log10S Q∆h=1
(m3/s) (m) (m)
0.813 1.061 0.832 0.784 0.1057 7.986
0.823 1.075 0.902 0.839 0.0762 12.486
0.825 1.091 0.959 0.879 0.0560 19.036
0.824 1.119 1.028 0.919 0.0367 33.839
0.793 1.174 1.131 0.963 0.0164 104.087
1.427 1.526 1.205 0.790 0.1024 8.305
1.436 1.543 1.310 0.849 0.0711 13.733
1.418 1.564 1.407 0.900 0.0458 25.005
1.377 1.595 1.498 0.939 0.0273 51.220
1.241 1.638 1.597 0.975 0.0110 175.371

• A logarithmic plot of the submerged-flow data can be made

• Each data point can have a line drawn at a slope of nf = 1.55 (from the prior free-
flow data analysis), which can be extended to where it intercepts the abscissa at
hu - hd = 1.0
• Then, the corresponding value of discharge can be read on the ordinate, which is
listed as Q∆h=1.0 in the above table
• The value of Q∆h=1.0 can also be determined analytically because a straight line
on logarithmic paper is an exponential function having the simple form:

nf
Cs (hu − hd )
Qs
=
( − logS )ns = (hu − hd )
nf
(8)
n
Q∆h=1 Cs (1) f
( − logS )ns
then,
nf
Qs = Q∆h=1.0 (hu − hd ) (9)

or,

Qs
Q∆h=1.0 = nf
(10)
(hu − hd )
where Q∆h=1.0 has a different value for each value of the submergence, S

• Using the term Q∆h=1.0 implies that hu - hd = 1.0 (by definition); thus, Eq. 10
reduces to:

BIE 5300/6300 Lectures 113 Gary P. Merkley

 nf
Cs (1.0 ) −ns
Q∆h=1.0 = nf
= Cs ( − logS ) (11)
(hu − hd )
• Again, this is a power function where Q∆h=1.0 can be plotted against (-log S) on
logarithmic paper to yield a straight-line relationship
• Note that the straight line in such a plot would have a negative slope (-ns) and
that Cs is the value of Q∆h=1.0 when (-log S) is equal to unity
• For the example data, the submerged-flow equation is:

1.55
0.367 ( hu − hd )
Qs = 1.37
(12)
( − logS )
Transition Submergence

• By setting the free-flow discharge equation equal to the submerged-flow

discharge equation, the transition submergence, St, can be determined
• Consider this:

Cs (1 − St ) f hnuf
n
Cf hnuf = (13)
( − logSt ) ns

then,

f ( St ) = Cf ( − logSt )
ns
− Cs (1 − St )
nf
=0 (14)

• In our example, we have:

1.37 1.55
0.74 ( − logS ) = 0.367 (1 − S ) (15)

or,
1.55
0.367 ( hu − hd )
0.74h1.55
u = (16)
( − logS )1.37
and,
1.37 1.55
0.74 ( − logS ) = 0.367 (1 − S ) (17)

• The value of S in this relationship is St provided the coefficients and exponents

have been accurately determined
• Again, small errors will dramatically affect the determination of St
Gary P. Merkley 114 BIE 5300/6300 Lectures
 1.37 1.55
0.74 ( − logSt ) = 0.367 (1 − St ) (18)

• Equation 18 can be solved to determine the value of St, which in this case is 0.82
• Thus, free flow exists when S < 0.82 and submerged flow exists when the
submergence is greater than 82%
• The table below gives the submergence values for different values of Qs/Qf for
the sample constriction rating

S Qs/Qf S Qs/Qf
0.82 1.000 0.91 0.9455
0.83 0.9968 0.92 0.9325
0.84 0.9939 0.93 0.9170
0.85 0.9902 0.94 0.8984
0.86 0.9856 0.95 0.8757
0.87 0.9801 0.96 0.8472
0.88 0.9735 0.97 0.8101
0.89 0.9657 0.98 0.7584
0.90 0.9564 - -

1.55
Qs 0.367 ( hu − hd ) 1
= = (19)
Qf ( − logS )1.37 0.74h1.55
u

which is also equal to:

1.55
Qs 0.496 (1 − S )
= (20)
Qf ( − logS )1.37
• For example, if hu and hd are measured and found to be 1.430 and 1.337, the first
step would be to compute the submergence, S,

1.337
S= = 0.935 (21)
1.430

• Thus, for this condition submerged flow exists in the example open-channel
constriction
• In practice, there may only be a “trivial” solution for transition submergence, in
which St = 1.0. In these cases, the value of Cs can be slightly lowered to obtain
another mathematical root to the equation. This is a “tweaking” procedure.
• Note that it is almost always expected that 0.50 < St < 0.92. If you come up with
a value outside of this range, you should be suspicious that the data and or the
analysis might have errors.

BIE 5300/6300 Lectures 115 Gary P. Merkley

IV. Constant-Head Orifices

• A constant-head orifice, or
CHO, is a double orifice
gate, usually installed at
the entrance to a lateral or
tertiary canal
• This is a design promoted
for years by the USBR,
and it can be found in
irrigation canals in many
countries
• The idea is that you set
the downstream gate as a
meter, and set the
upstream gate as
necessary to have a
constant water level in the
mid-gate pool
• It is kind of like the double
doors in the engineering building at USU: they are designed to act as buffers
whereby the warm air doesn’t escape so easily when people enter an exit the
building
• But, in practice, CHOs are seldom used as intended; instead, one of the gates is
left wide open and the other is used for regulation (this is a waste of materials
because one gate isn’t used at all)
• Note the missing wheel in the upstream gate, in the above figure
• Few people know what CHOs are for, and even when they do, it is often
considered inconvenient or impractical to operate both gates
• But these gates can be calibrated, just as with any other gate

V. General Hydraulic Characteristics of Gates for Orifice Flow

• It is safe to assume that the exponent on the head for orifice flow (free and
submerged) is 0.50, so it is not necessary to treat it as an empirically-determined
calibration parameter
• The basic relationship for orifice flow can be derived from the Bernoulli equation
• For orifice flow, a theoretical contraction coefficient, Cc, of 0.611 is equal to π/(π+2),
derived from hydrodynamics for vertical flow through an infinitely long slot
• Field-measured discharge coefficients for orifice flow through gates normally range
from 0.65 to about 0.9 − there is often a significant approach velocity
• Radial gates can be field calibrated using the same equation forms as vertical
sluice gates, although special equations have been developed for them

Gary P. Merkley 116 BIE 5300/6300 Lectures

VI. How do You Know if it is Free Flow?

• If the water level on the upstream side of the gate is above the bottom of the
gate, then the flow regime is probably that of an orifice
• In this case, the momentum function (from open-channel hydraulics) can be used
to determine whether the flow is free or submerged
• In some cases it will be obviously free flow or obviously submerged flow, but the
following computational procedure is one way to distinguish between free orifice
and submerged orifice flow
• In the figure below, h1 is the depth just upstream of the gate, h2 is the depth just
downstream of the gate (depth at the vena-contracta section) and h3 is the depth
at a section in the downstream, a short distance away from the gate

• If the value of the momentum function corresponding to h2 is greater than that

corresponding to h3, free flow will occur; otherwise, it is submerged
• The momentum function is:

Q2
M = A hc + (22)
gA

where A is the cross-sectional area and hc is the depth to the centroid of the area
from the water surface

• The following table shows the values of A and Ahc for three different channel
sections

BIE 5300/6300 Lectures 117 Gary P. Merkley

 Section A Ahc

Rectangular bh bh2
2
⎡ ⎛ m1 + m2 ⎞ ⎤ ⎡ b ⎛ m1 + m2 ⎞ ⎤ 2
Trapezoidal ⎢b + ⎜ 2 ⎟ h⎥ h ⎢2 + ⎜ 6 ⎟ h⎥ h
⎣ ⎝ ⎠ ⎦ ⎣ ⎝ ⎠ ⎦

D2 D3 ⎡ sin3 ( θ / 2 ) cos ( θ / 2 )( θ − sin θ ) ⎤

Circular ( θ − sin θ ) 4
⎢
3
−
4
⎥
8 ⎢⎣ ⎥⎦

• For rectangular and trapezoidal sections, b is the base width

• For trapezoidal sections, m1 and m2 are the inverse side slopes (zero for vertical
sides), which are equal for symmetrical sections

−1 ⎛ 2h ⎞
• For circular sections, θ is defined as: θ = 2cos ⎜1 − ⎟
⎝ D⎠

where D is the inside diameter of the circular section

• There are alternate forms of the equations for circular sections, but which yield
the same calculation results.

• The depth h2 can be determined in the either of the following two ways:

1. h2 = Cc Go where Cc is the contraction coefficient and Go is the

vertical gate opening
2. Equate the specific energies at sections 1 and 2, where
Q2
E =h+ , then solve for h2
2gA 2

• Calculate M2 and M3 using the depths h2 and h3, respectively

• If M2 ≤ M3, the flow is submerged; otherwise the flow is free

• Another (simpler) criteria for the threshold between free- and submerged-orifice
flow is:

CcGo = hd (23)

where Cc is the contraction coefficient (≈ 0.61)

Gary P. Merkley 118 BIE 5300/6300 Lectures

 • But this is not the preferred way to make the distinction and is not as accurate as
the momentum-function approach

VII. How do You Know if it is Orifice Flow?

• The threshold between orifice and nonorifice flow can be defined as:

Cohu = Go (24)

where Co is an empirically-determined coefficient (0.80 ≤ Co ≤ 0.95)

• Note that if Co = 1.0, then when hu = Go, the water surface is on the verge of
going below the bottom of the gate (or vice-versa), when the regime would
clearly be nonorifice
• However, in moving from orifice to nonorifice flow, the transition would begin
before this point, and that is why Co must be less than 1.0
• It seems that more research is needed to better defined the value of Co

• In practice, the flow can move from any regime to any other at an underflow
(gate) structure:

Free Submerged

Orifice FO SO

Non-orifice FN SN

VII. Orifice Ratings for Canal Gates

• For free-flow conditions through an orifice, the discharge equation is:

Qf = CdCv A 2ghu (25)

where Cd is the dimensionless discharge coefficient; Cv is the dimensionless

velocity head coefficient; A is the area of the orifice opening, g is the ratio of
weight to mass; and hu is measured from the centroid of the orifice to the
upstream water level

BIE 5300/6300 Lectures 119 Gary P. Merkley

 • The velocity head coefficient, Cv, approaches unity as the approach velocity to
the orifice decreases to zero
• In irrigation systems, Cv can usually be assumed to be unity since most irrigation
channels have very flat gradients and the flow velocities are low
• The upstream depth, hu, can also be measured from the bottom of the orifice
opening if the downstream depth is taken to be about 0.61 times the vertical
orifice opening
• Otherwise, it is assumed that the downstream depth is equal to one-half the
opening, and hu is effectively measured from the area centroid of the opening
• The choice will affect the value of the discharge coefficient

• If hu is measured from the upstream canal bed, Eq. 4 becomes

Qf = CdCv A 2g(hu − CgGo ) (26)

where Cg is usually either 0.5 or 0.61, as explained above.

• If the downstream water level is also above the top of the opening, submerged
conditions exist and the discharge equation becomes:

Qs = CdCv A 2g ( hu − hd ) (27)

where hu - hd is the difference in water surface elevations upstream and

downstream of the submerged orifice.

• An orifice can be used as an accurate flow measuring device in an irrigation

system
• If the orifice structure has not been previously rated in the laboratory, then it can
easily be rated in the field
• As mentioned above, the hydraulic head term, (hu - CgGo) or (hu - hd), can be
relied upon to have the exponent ½, which means that a single field rating
measurement could provide an accurate determination of the coefficient of
discharge, Cd
• However, the use of a single rating measurement implies the assumption of a
constant Cd value, which is not the case in general
• Adjustments to the basic orifice equations for free- and submerged-flow are often
made to more accurately represent the structure rating as a function of flow
depths and gate openings
• The following sections present some alternative equation forms for taking into
account the variability in the discharge coefficient under different operating
conditions

Gary P. Merkley 120 BIE 5300/6300 Lectures

VIII. Variation in the Discharge Coefficient

• Henry (1950) made numerous laboratory measurements to determine the

discharge coefficient for free and submerged flow through an orifice gate
• The figure below shows an approximate representation of his data

BIE 5300/6300 Lectures 121 Gary P. Merkley

Gary P. Merkley 122 BIE 5300/6300 Lectures
Lecture 12
Calibration of Canal Gates

I. Free-Flow Rectangular Gate Structures

• For a rectangular gate having a gate opening, Go, and a gate width, W, the free-
flow discharge equation can be obtained from Eq. 26 of the previous lecture,
assuming that the dimensionless velocity head coefficient is equal to unity:

(
Qf = CdGoGw 2g hu − CgGo ) (1)

where Go is the vertical gate opening; Gw is the gate width; GoGw is the area, A,
of the gate opening; and, Cg is between 0.5 and 0.61

• The upstream flow depth, hu, can be measured anywhere upstream of the gate,
including the upstream face of the gate
• The value of hu will vary only a small amount depending on the upstream location
chosen for measuring hu
• Consequently, the value of the coefficient of discharge, Cd, will also vary
according to the location selected for measuring hu

• One of the most difficult tasks in calibrating a gate structure is obtaining a highly
accurate measurement of the gate opening, Go
• For gates having a threaded rod that rises as the gate opening is increased, the
gate opening is read from the top of the hand-wheel to the top of the rod with the
gate closed, and when set to some opening, Go
• This very likely represents a measurement of gate opening from where the gate
is totally seated, rather than a measurement from the gate lip; therefore, the
measured value of Go from the thread rod will usually be greater than the true
gate opening, unless special precautions are taken to calibrate the thread rod

• Also, when the gate lip is set at the same elevation as the gate sill, there will
undoubtedly be some flow or leakage through the gate
• This implies that the datum for measuring the gate opening is below the gate sill
• In fact, there is often leakage from a gate even when it is totally seated because
of inadequate maintenance
• An example problem will be used to illustrate the procedure for determining an
appropriate zero datum for the gate opening

Sample Calibration

• Calibration data (listed in the table below) were collected for a rectangular gate
structure
• The data reduction is listed in the next table, where the coefficient of discharge,
Cd, was calculated from Eq. 27
BIE 5300/6300 Lectures 123 Gary P. Merkley
 Discharge, Qf Gate Opening, Go Upstream Benchmark
(m3/s) (m) Tape Measurement (m)
0.0646 0.010 0.124
0.0708 0.020 1.264
0.0742 0.030 1.587
0.0755 0.040 1.720
0.0763 0.050 1.787
0.0767 0.060 1.825

Qf Go hu
Cd
(m3/s) (m) (m)
0.0646 0.010 1.838 0.756
0.0708 0.020 0.698 0.677
0.0742 0.030 0.375 0.654
0.0755 0.040 0.242 0.635
0.0763 0.050 0.175 0.625
0.0767 0.060 0.137 0.620
Note: The discharge coefficient, Cd, was calculated
using the following equation:

⎛ G ⎞
Qf = CdGoGw 2g ⎜ hu − o ⎟
⎝ 2 ⎠

• A rectangular coordinate plot of Cd versus the gate opening, Go, is shown in the
figure below

0.76

0.74
Coefficient of discharge, Cd

0.72

0.70

0.68

0.66

0.64

0.62
0.01 0.02 0.03 0.04 0.05 0.06
Gate opening, Go (m)

Gary P. Merkley 124 BIE 5300/6300 Lectures

 • Notice that the value of Cd continues to decrease with larger gate openings
• One way to determine if a constant value of Cd can be derived is to rewrite Eq. 5
in the following format (Skogerboe and Merkley 1996):

⎛ G + ∆Go ⎞
Qf = Cd ( Go + ∆Go ) Gw 2g ⎜ (hu )∆G − o ⎟ (2)
⎝ o 2 ⎠

where ∆Go is a measure of the zero datum level below the gate sill, and

(hu )∆Go = hu + ∆Go (3)

• Assuming values of ∆Go equal to 1 mm, 2 mm, 3 mm, etc., the computations for
determining Cd can be made from Eq. 3
• The results for ∆Go equal to 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, 6 mm, 7 mm, 8
mm and 12 mm (gate seated) are listed in the table below
• The best results are obtained for ∆Go of 3 mm − the results are plotted in the
figure below, which shows that Cd varies from 0.582 to 0.593 with the average
value of Cd being 0.587
• For this particular gate structure, the discharge normally varies between 200 and
300 lps, and the gate opening is normally operated between 40-60 mm, so that a
constant value of Cd = 0.587 can be used when the zero datum for Go and hu is
taken as 3 mm below the gate sill
• Another alternative would be to use a constant value of Cd = 0.575 for ∆Go = 4
mm and Go greater than 30 mm

BIE 5300/6300 Lectures 125 Gary P. Merkley

Gary P. Merkley 126 BIE 5300/6300 Lectures
BIE 5300/6300 Lectures 127 Gary P. Merkley
II. Submerged-Flow Rectangular Gate Structures

• Assuming that the dimensionless velocity head coefficient in Eq. 27 is unity, the
submerged-flow discharge equation for a rectangular gate having an opening,
Go, and a width, W, becomes:

Qs = CdGoGw 2g (hu − hd ) (4)

where GoGw is the area, A, of the orifice

• Field calibration data for a rectangular gate structure operating under

submerged-flow conditions are listed in the table below
• Note that for this type of slide gate, the gate opening can be measured both on
the left side, (Go)L, and the right side, (Go)R, because the gate lip is not always
horizontal
• The calculations are shown in the second table below

Discharge, Qs Gate Opening Benchmark Tape Measurements

(Go)left (Go)right Upstream Downstream
(m3/s) (m) (m) (m) (m)
0.079 0.101 0.103 0.095 0.273
0.095 0.123 0.119 0.099 0.283
0.111 0.139 0.139 0.102 0.296
0.126 0.161 0.163 0.105 0.290
0.141 0.180 0.178 0.108 0.301
0.155 0.199 0.197 0.110 0.301

Qs Go hu hd h u - hd
Cd
(m3/s) (m) (m) (m) (m)
0.079 0.102 0.823 0.643 0.180 0.676
0.095 0.121 0.819 0.633 0.187 0.674
0.111 0.139 0.816 0.620 0.196 0.668
0.126 0.162 0.813 0.626 0.187 0.666
0.141 0.179 0.810 0.615 0.195 0.660
0.155 0.198 0.808 0.615 0.193 0.659

• As in the case of the free-flow orifice calibration in the previous section, a trial-
and-error approach can be used to determine a more precise zero datum for the
gate opening
• In this case, the submerged flow equation would be rewritten as:

Qs = Cd ( Go + ∆Go ) Gw 2g (hu − hd ) (5)

Gary P. Merkley 128 BIE 5300/6300 Lectures

 where ∆Go is the vertical distance from the gate sill down to the zero datum level,
as previously defined in Eq. 2

Qs Go h u - hd Cd

(m3/s) (m) (m) ∆Go = 4 mm ∆Go = 6 mm ∆Go = 8 mm

0.079 0.102 0.1801 0.650 0.638 0.626

0.095 0.121 0.1865 0.651 0.641 0.631

0.111 0.139 0.1960 0.649 0.640 0.631

0.126 0.162 0.1869 0.650 0.642 0.635

0.141 0.179 0.1949 0.646 0.639 0.632

0.155 0.198 0.1931 0.646 0.640 0.634

Note: The discharge coefficient, Cd, was calculated from Eq. 32:

• As before, the criteria for determining ∆Go is to obtain a nearly constant value of
the discharge coefficient, Cd
• The above table has the example computational results for determining the
discharge coefficient, Cd, according to adjusted gate openings, Go, under
submerged flow conditions

III. Calibrating Medium- and Large-Size Gate Structures

• A different form of the submerged-flow rating equation has been used with
excellent results on many different orifice-type structures in medium and large
canals
• The differences in the equation involve consideration of the gate opening and the
downstream depth as influential factors in the determination of the discharge
coefficient
• The equation is as follows:

Qs = Cs hs Gw 2g (hu − hd ) (6)

and,
β
⎛G ⎞
Cs = α ⎜ o ⎟ (7)
⎝ hs ⎠

where hs is the downstream depth referenced to the bottom of the gate opening,
α and β are empirically-fitted parameters, and all other terms are as described
previously

BIE 5300/6300 Lectures 129 Gary P. Merkley

 hu G
o hs
hd

• Note that Cs is a dimensionless number

• The value of the exponent, β, is usually very close to unity
• In fact, for β equal to unity the equation reverts to that of a constant value of Cs
equal to α (the hs term cancels)

• The next table shows some example field calibration data for a large canal gate
operating under submerged-flow conditions
• The solution to the example calibration is: α = 0.796, and β = 1.031
• This particular data set indicates an excellent fit to Eqs. 6 and 7, and it is typical
of other large gate structures operating under submerged-flow conditions

Data Discharge Go ∆h hs
Go/hs Cs
Set (m3/s) (m) (m) (m)
1 8.38 0.60 3.57 2.205 0.272 0.206
2 9.08 0.70 3.00 2.010 0.348 0.268
3 5.20 0.38 3.31 1.750 0.217 0.168
4 4.27 0.30 3.41 1.895 0.158 0.125
5 5.45 0.40 3.43 2.025 0.198 0.149
6 12.15 0.95 2.63 2.300 0.413 0.334
7 5.49 0.38 3.72 1.905 0.199 0.153
8 13.52 1.10 2.44 2.405 0.457 0.369
9 14.39 1.00 3.84 2.370 0.422 0.318
10 16.14 1.13 3.79 2.570 0.440 0.331
11 6.98 0.50 3.70 1.980 0.253 0.188
12 11.36 0.58 7.64 2.310 0.251 0.183
13 7.90 0.42 6.76 2.195 0.191 0.142
14 7.15 0.38 6.86 2.110 0.180 0.133
15 7.49 0.51 3.98 2.090 0.244 0.184
16 10.48 0.70 3.92 2.045 0.342 0.266
17 12.41 0.85 3.76 2.205 0.385 0.298
18 8.26 0.55 3.91 2.065 0.266 0.208
Note: the data are for two identical gates in parallel, both having the
same opening for each data set, with a combined opening width of 2.20
m.

Gary P. Merkley 130 BIE 5300/6300 Lectures

 • A similar equation can be used for free-flow through a large gate structure, with
the upstream depth, hu, replacing the term hs, and with (hu - Go/2) replacing (hu -
hd)
• As previously mentioned, the free-flow equation can be calibrated using (hu -
0.61Go) instead of (hu - Go/2)
• The following figure shows a graph of the 18 data points; the straight line is the
regression results which gives α and β

-0.6

-0.8

alpha = 0.796
-1.0
beta = 1.031
-1.2

-1.4
ln(Cs)

-1.6

-1.8

-2.0

-2.2

-2.4
-2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6
ln(Go/hs)

IV. Radial Gate Orifice-Flow Calibrations

• This structure type includes radial (or “Tainter”) gates as calibrated by the USBR
(Buyalski 1983) for free and submerged orifice-flow conditions
• The calibration of the gates follows the specifications in the USBR “REC-ERC-
83-9” technical publication, which gives calibration equations for free and
submerged orifice flow, and corrections for the type of gate lip seal
• The calibration requires no field measurements other than gate dimensions, but
you can add another coefficient to the equations for free flow and orifice flow in
an attempt to accommodate calibration data, if available
• Three gate lip seal designs (see figure below) are included in the calibrations:

BIE 5300/6300 Lectures 131 Gary P. Merkley

 1. Hard-rubber bar;
2. Music note; and,
3. Sharp edge

• The gate lip seal is the bottom of the gate leaf, which rests on the bottom of the
channel when the gate is closed
• The discharge coefficients need no adjustment for the hard-rubber bar gate lip,
which is the most common among USBR radial gate designs, but do have
correction factors for the other two lip seal types
• These are given below for free and submerged orifice flow

Gr

hu
Go P hd

• The gate radius divided by the pinion height should be within the range 1.2 ≤ Gr/P
≤ 1.7
• The upstream water depth divided by the pinion height should be less than or
equal to 1.6 (hu/P ≤ 1.6)
• If these and other limits are observed, the accuracy of the calculated flow rate
from Buyalski’s equations should be within 1% of the true flow rate

Gary P. Merkley 132 BIE 5300/6300 Lectures

 Free Orifice Flow Free, or modular orifice flow is assumed to prevail when the
downstream momentum function corresponding to CcGo, where Go is the vertical
gate opening, is less than or equal to the momentum function value using the
downstream depth. Under these conditions the rating equation is:

Ff = Qf - CfcdaGoGw 2ghu = 0 (8)

where Cfcda is the free-flow discharge coefficient; Go is the vertical gate

opening (m or ft); Gw is the width of the gate opening (m or ft); and hu is the
upstream water depth (m or ft); Cfcda is dimensionless

• Cfcda is determined according to a series of conic equations as defined by

Buyalski (1983) from an analysis of over 2,000 data points
• The equations are lengthy, but are easily applied in a computer program

Eccentricity

⎛ ⎛ Gr ⎞ ⎞
2
AFE = 0.00212 ⎜ 1.0+31.2 ⎜ − 1.6 ⎟ ⎟ + 0.901
⎜ ⎝ P ⎠ ⎟⎠
⎝ (9)

⎛ ⎛ Gr ⎞ ⎞
2
BFE = 0.00212 ⎜ 1.0+187.7 ⎜ − 1.635 ⎟ ⎟ − 0.079
⎜ ⎝ P ⎠ ⎟⎠
⎝ (10)

⎛G ⎞
FE = AFE − BFE ⎜ o ⎟
⎝ P ⎠ (11)

where Gr is the gate radius (m or ft); and P is the pinion height (m or ft)

Directrix

⎛ ⎛ Gr ⎞ ⎞
2
AFD = 0.788 − 0.04 ⎜ 1.0+89.2 ⎜ − 1.619 ⎟ ⎟ (12)
⎜ ⎝ P ⎠ ⎟⎠
⎝

BIE 5300/6300 Lectures 133 Gary P. Merkley

 ⎛G ⎞
BFD = 0.0534 ⎜ r ⎟ + 0.0457 (13)
⎝ P ⎠

⎛ ⎛ Go ⎞ ⎞
2
FD = 0.472 − BFD ⎜ 1.0- ⎜ − AFD ⎟ ⎟ (14)
⎜ ⎝ P ⎠ ⎟⎠
⎝

Focal Distances

⎛G ⎞ Go
FX1 = 1.94 ⎜ o ⎟ − 0.377 ≤ 0.277
⎝ P ⎠ P
(15)
⎛G ⎞ Go
FX1 = 0.18 ⎜ o ⎟ + 0.111 > 0.277
⎝ P ⎠ P

⎛G ⎞
FY1 =0.309 − 0.192 ⎜ o ⎟ (16)
⎝ P ⎠
and,
hu
FXV = − FX1 (17)
P

The correction on Cfcda for the “music note” gate lip seal design is:

⎛G ⎞
Ccorrect = 0.125 ⎜ o ⎟ + 0.91 (music note) (18)
⎝ P ⎠

The correction on Cfcda for the “sharp edge” gate lip seal design is:

⎛G ⎞
Ccorrect = 0.11⎜ o ⎟ + 0.935 (sharp edge) (19)
⎝ P ⎠

For the hard-rubber bar gate lip seal design, Ccorrect = 1.0. The preceding corrections
on Cfcda for the “music note” and “sharp edge” gate lip seal designs were chosen
from the linear options proposed by Buyalski (ibid).

Gary P. Merkley 134 BIE 5300/6300 Lectures

 Finally,

⎛ ⎞
Cfcda = Ccorrect ⎜ FE2 (FD + FXV ) − FXV 2 + FY1 ⎟
2
(20)
⎝ ⎠

Radial gates in parallel at a check structure

Submerged Orifice Flow The submerged orifice rating equation is:

Fs = Qs - CscdaGoGw 2ghu = 0 (21)

where Cscda is the submerged-flow discharge coefficient; and all other terms
are as previously defined; both Cscda and Cds are dimensionless

• Note that the square-root term does not include the downstream depth, hd,
but it is included in the lengthy definition of Cscda
• As in the free-flow case, Cscda is determined according to a series of conic
equations:

Directrix
−1
⎛ ⎛G ⎞ ⎞
ADA = ⎜ 11.98 ⎜ r ⎟ − 26.7 ⎟ (22)
⎝ ⎝ P ⎠ ⎠

BIE 5300/6300 Lectures 135 Gary P. Merkley

 ⎛ P⎞
ADB = 0.62 − 0.276 ⎜ ⎟ (23)
⎝ Gr ⎠

−1
⎛ ⎛G ⎞ ⎞
AD = ⎜ ADA ⎜ o ⎟ + ADB ⎟ (24)
⎝ ⎝ P ⎠ ⎠

⎛G ⎞
BDA = 0.025 ⎜ r ⎟ − 2.711 (25)
⎝ P ⎠

⎛G ⎞
BDB = 0.071 − 0.033 ⎜ r ⎟ (26)
⎝ P ⎠

⎛G ⎞
BD = BDA ⎜ o ⎟ + BDB (27)
⎝ P ⎠

⎛h ⎞
DR = AD ⎜ d ⎟ + BD (28)
⎝P⎠

D = DR −1.429 (29)

Eccentricity
⎛G ⎞
AEA = 0.06 − 0.019 ⎜ r ⎟
⎝P ⎠ (30)

⎛G ⎞
AEB = 0.996 + 0.0052 ⎜ r ⎟
⎝P ⎠ (31)

−1
⎛ ⎛G ⎞ ⎞
AE = ⎜ AEA ⎜ o ⎟ + AEB ⎟
⎝ ⎝ P ⎠ ⎠ (32)

Gary P. Merkley 136 BIE 5300/6300 Lectures

 ⎛G ⎞
BEK = 0.32 − 0.293 ⎜ r ⎟
⎝P ⎠ (33)

⎛ ⎛ Go ⎞ ⎞
2
BE = BEK + 0.255 ⎜ 1.0 + 1.429 ⎜ − 0.44 ⎟ ⎟ (34)
⎜ ⎝ P ⎠ ⎟⎠
⎝

ER = AE (D ) + BE
(35)

⎛ ER ⎞
E = ln ⎜ ⎟
⎝ D ⎠ (36)
Vector V1
E (D )
V1 = (37)
1.0 + E

Focal Distance
⎛P⎞
AFA = 0.038 − 0.158 ⎜ ⎟ (38)
⎝ Gr ⎠

⎛G ⎞
AFB = 0.29 − 0.115 ⎜ r ⎟ (39)
⎝ P ⎠

⎛G ⎞
AF = AFA ⎜ o ⎟ + AFB (40)
⎝ P ⎠

⎛P ⎞
BFA = 0.0445 ⎜ ⎟ − 0.321 (41)
⎝ Gr ⎠

BIE 5300/6300 Lectures 137 Gary P. Merkley

 ⎛P⎞
BFB = 0.155 − 0.092 ⎜ ⎟ (42)
⎝ Gr ⎠

⎛ P ⎞
BF = BFA ⎜ ⎟ + BFB (43)
⎝ Go ⎠

AF (hd )
FY = BF − (44)
P

• If FY ≤ 0, then let FY = 0 and FX = 0. Otherwise, retain the calculated value

of FY and,

FX = V12 + FY 2 − V1 (for FY > 0) (45)

hu h
VX = − V1 − d − FX (46)
P P

The correction on Cscda for the “music note” gate lip seal design is:

⎛G ⎞
Ccorrect = 0.39 ⎜ o ⎟ + 0.85 (music note) (47)
⎝ P ⎠

The correction on Cscda for the “sharp edge” gate lip seal design is:

⎛G ⎞
Ccorrect = 0.11⎜ o ⎟ + 0.9 (sharp edge) (48)
⎝ P ⎠

• For the hard-rubber bar gate lip seal design, Ccorrect = 1.0
• The preceding corrections on Cscda for the “music note” and “sharp edge” gate
lip seal designs were chosen from the linear options proposed by Buyalski
(ibid)

Finally,

Gary P. Merkley 138 BIE 5300/6300 Lectures

 ⎛ ⎞
Cscda = Ccorrect ⎜ E2 (D + VX ) − VX 2 + FY ⎟
2
(49)
⎝ ⎠

References & Bibliography

Buyalski, C.P. 1983. Discharge algorithms for canal radial gates. Technical Report REC-ERC-83-9.
U.S. Bureau of Reclamation, Denver, CO. 232 pp.
Brater, E.F., and H.W. King. 1976. Handbook of hydraulics. 6th edition. McGraw-Hill Book Company,
New York, N.Y. 583 pp.
Buyalski, C.P. 1983. Discharge algorithms for canal radial gates. Technical Report REC-ERC-83-9.
U.S. Bureau of Reclamation, Denver, CO. 232 pp.
Chow, V.T. 1959. Open-channel hydraulics. McGraw-Hill Book Company, New York, N.Y. 680 pp.
Daugherty, R.L., and J. B. Franzini. 1977. Fluid mechanics with engineering applications. 7th edition.
McGraw-Hill Book Company, New York, N.Y. 564 pp.
Davis, C.V. and K.E. Sorensen (eds.). 1969. Handbook of applied hydraulics. McGraw-Hill Book
Company, New York, N.Y.
French, R.H. 1985. Open-channel hydraulics. McGraw-Hill Book Company, New York, N.Y. 705 pp.
Hu, W.W. 1973. Hydraulic elements for USBR standard horseshoe tunnel. J. of the Transportation
Engrg. Div., ASCE, 99(4): 973-980.
Hu, W.W. 1980. Water surface profile for horseshoe tunnel. Transportation Engrg. Journal, ASCE,
106(2): 133-139.
Press, W.H., S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery. 1992. Numerical recipes in C: the
art of scientific computing. 2nd Ed. Cambridge Univ. Press, Cambridge, U.K. 994 pp.
Shen, J. 1981. Discharge characteristics of triangular-notch thin-plate weirs. Water Supply Paper
1617-B. U.S. Geological Survey.
Skogerboe, G.V., M.L. Hyatt, R.K. Anderson, and K.O. Eggleston. 1967. Cutthroat flumes. Utah Water
Research Laboratory Report WG31-4. 37 pp.
Skogerboe, G.V., L. Ren, and D. Yang. 1993. Cutthroat flume discharge ratings, size selection and
installation. Int’l Irrig. Center Report, Utah State Univ., Logan, UT. 110 pp.
Uni-Bell Plastic Pipe Association. 1977. Handbook of PVC pipe: design and construction. Uni-Bell
Plastic Pipe Association, Dallas, TX.
Villamonte, J.R. 1947. Submerged-weir discharge studies. Engrg. News Record, p. 866.

BIE 5300/6300 Lectures 139 Gary P. Merkley

Gary P. Merkley 140 BIE 5300/6300 Lectures
Lecture 13
Flow Measurement in Pipes

I. Introduction

• There are a wide variety of methods for measuring discharge and velocity in
pipes, or closed conduits
• Many of these methods can provide very accurate measurements
• Others give only rough estimates
• But, in general, it is easier to obtain a given measurement accuracy in pipes
when compared to measurement in open channels
• Some of the devices used are very expensive and are more suited to industrial
and municipal systems than for agricultural irrigation systems

II. Pitot Tubes

• The pitot tube is named after Henri Pitot who used a bent glass tube to measure
velocities in a river in France in the 1700s
• The pitot tube can be used not only for measuring flow velocity in open channels
(such as canals and rivers), but in closed conduits as well
• There are several variations of pitot tubes for measuring flow velocity, and many
of these are commercially available

• Pitot tubes can be very simple devices with no moving parts

• More sophisticated versions can provide greater accuracy (e.g. differential head
meters that separate the static pressure head from the velocity head)
• The pitot static tube, shown in the figure below, is one variation of the device
which allows the static head (P/γ) and dynamic (total) head (P/γ + V2/2g) to be
separately measured

BIE 5300/6300 Lectures 141 Gary P. Merkley

 • The static head equals the depth if open-channel flow
• Calibrations are required because the velocity profile can change with the flow
rate, and because measurement(s) are only a sampling of the velocities in the
pipe

• The measurement from a pitot tube can be accurate to ±1% of the true velocity,
even if the submerged end of the tube is up to ±15% out of alignment from the
flow direction
• The velocity reading from a pitot tube must be multiplied by cross-sectional area
to obtain the flow rate (it is a velocity-area method)
• Pitot tubes tend to become clogged unless the water in the pipe is very clean
• Also, pitot tubes may be impractical if there is a large head, unless a manometer
is used with a dense liquid like mercury

III. Differential Producers

• This is a class of flow measurement devices for full pipe flow

• “Differential producers” cause a pressure differential which can be measured and
correlated to velocity and or flow rate in the pipe
• Examples of differential producers:

• Venturis
• Nozzles
• Orifices

• Measured ∆P at a differential producer depends on:

• Flow rate
• Fluid properties
• Element geometry

IV. Venturi Meters

• The principle of this flow measurement device was first documented by J.B.
Venturi in 1797 in Italy
• Venturi meters have only a small head loss, no moving parts, and do not clog
easily

Gary P. Merkley 142 BIE 5300/6300 Lectures

 • The principle under which these
devices operate is that some
pressure head is converted to
velocity head when the cross-
sectional area of flow decreases
(Bernoulli equation)
• Thus, the head differential can be
measured between the upstream
section and the throat section to
give an estimation of flow velocity,
and this can be multiplied by flow area to arrive at a discharge value
• The converging section is usually about 21º, and the diverging section is usually
from 5 to 7º

Head
loss
∆h

21º 5º - 7º

Flow
D1 D2 D1

• A form of the calibration equation is:

2g ∆h(sg − 1)
Q = C A2 (1)
4
1− β

where C is a dimensionless coefficient from approximately 0.935 (small throat

velocity and diameter) to 0.988 (large throat velocity and diameter); β is the ratio
of D2/D1; D1 and D2 are the inside diameters at the upstream and throat sections,
respectively; A2 is the area of the throat section; ∆h is the head differential; and
“sg” is the specific gravity of the manometer liquid

• The discharge coefficient, C, is a constant value for given venturi dimensions

• Note that if D2 = D1, then β = 1, and Q is undefined; if D0 > D1, you get the square
root of a negative number (but neither condition applies to a venturi)
• The coefficient, C, must be adjusted to accommodate variations in water
temperature

BIE 5300/6300 Lectures 143 Gary P. Merkley

 • The value of β is usually between 0.25 and 0.50, but may be as high as 0.75
• Venturi meters have been made out of steel, iron, concrete, wood, plastic, brass,
bronze, and other materials
• Most modern venturi meters of small size are made from plastic (doesn’t corrode)
• Many commercial venturi meters have patented features
• The upstream converging section usually has an angle of about 21° from the pipe
axis, and the diverging section usually has an angle of 5° to 7° (1:6 divergence,
as for the DS ramp of a BCW, is about 9.5°)

• Straightening vanes may be required upstream of the venturi to prevent swirling

flow, which can significantly affect the calibration
• It is generally recommended that there should be a distance of at least 10D1 of
straight pipe upstream of the venturi
• The head loss across a venturi meter is usually between 10 and 20% of ∆h
• This percentage decreases for larger venturis and as the flow rate increases
• Venturi discharge measurement error is often within ±0.5% to ±1% of the true
flow rate value

V. Flow Nozzles

• Flow nozzles operate on the same principle as venturi meters, but the head loss
tends to be much greater due to the absence of a downstream diverging section
• There is an upstream converging section, like a venturi, but there is no
downstream diverging section to reduce energy loss
Gary P. Merkley 144 BIE 5300/6300 Lectures
 • Flow nozzles can be less expensive than venturi meters, and can provide
comparable accuracy
• The same equation as for venturi meters is used for flow nozzles
• The head differential across the nozzle can be measured using a manometer or
some kind of differential pressure gauge
• The upstream tap should be within ½D1 to D1 upstream of the entrance to the
nozzle
• The downstream tap should be approximately at the outlet of the nozzle (see the
figure below)

Head
loss
HGL

Flow D1
D2

A Flow Nozzle in a Pipe

• The space between the nozzle and the pipe walls can be filled in to reduce the
head loss through the nozzle, as seen in the following figure

Head
loss
HGL

Flow D1
D2

A “Solid” Flow Nozzle in a Pipe

BIE 5300/6300 Lectures 145 Gary P. Merkley
VI. Orifice Meters

• These devices use a thin plate with an orifice, smaller than the pipe ID, to create
a pressure differential
• The orifice opening is usually circular, but can be other shapes:

• Square
• Oval
• Triangular
• Others

• The pressure differential can be measured, as in venturi and nozzle meters, and
the same equation as for venturi meters can be used
• However, the discharge coefficient is different for orifice meters
• It is easy to make and install an orifice meter in a pipeline – easier than a nozzle

• Orifice meters can give accurate measurements of Q, and they are simple and
inexpensive to build
• But, orifice meters cause a higher head loss than either the venturi or flow nozzle
meters
• As with venturi meters and flow nozzles, orifice meters can provide values within
±1% (or better) of the true discharge
• As with venturi meters, there should be a straight section of pipe no less than 10
diameters upstream
• Some engineers have used eccentric orifices to allow passage of sediments –
the orifice is located at the bottom of a horizontal pipe, not in the center of the
pipe cross section

• The orifice opening can be “sharp” (beveled) for better accuracy

• But don’t use a beveled orifice opening if you are going to use it to measure flow
in both directions
• These are the beveling dimensions:

upstream downstream

½D1 30 to 45 deg
0.005D1
to 0.02D1 ½D2
CL
Gary P. Merkley 146 BIE 5300/6300 Lectures
 • The upstream head is usually measured one pipe diameter upstream of the thin
plate, and the downstream head is measured at a variable distance from the
plate
• Standard calibrations are available, providing C values from which the discharge
can be calculated for a given ∆h value
• In the following, the coefficient for an orifice plate is called “K”, not “C”
• The coefficient values depend on the ratio of the diameters and on the Reynold’s
number of approach; they can be presented in tabular or graphical formats

Head loss
∆h

vena contracta
Flow D1
D2

0.5D1
D1
An Orifice Meter in a Pipe

• In the figure below, the Reynold’s number of approach is calculated for the pipe
section upstream of the orifice plate (diameter D1, and the mean velocity in D1)
• Note also that pipe flow is seldom laminar, so the curved parts of the figure are
not of great interest
• An equation for use with the curves for K:

⎡⎛ P ⎞ ⎛P ⎞⎤
Q = KA 2 2g ⎢⎜ u + zu ⎟ − ⎜ d + zd ⎟ ⎥ (2)
⎣⎝ γ ⎠ ⎝ γ ⎠⎦

• The above equation is the same form as for canal gates operating as orifices
• The ratio β is embedded in the K term
• Note that zu equals zd for a horizontal pipe (they are measured relative to an
arbitrary elevation datum)
• Note that Pu/γ is the same as hu (same for Pd/γ and hd)
• Also, you can let ∆h = hu - hd

BIE 5300/6300 Lectures 147 Gary P. Merkley

 • Pd is often measured at a distance of about ½D1 downstream of the orifice plate,
but the measurement is not too sensitive to the location, within a certain range
(say ¼ D1 to D1 downstream)
• The following graph shows the K value for an orifice meter as a function of the
ratio of diameters when the Reynold’s number of approach is high enough that
the K value no longer depends on Re

Orifice Meter Coefficient for High Reynold's Number

0.70

0.69

0.68

0.67

0.66

0.65
K

0.64

0.63

0.62

0.61

0.60
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70
D2/D1

Gary P. Merkley 148 BIE 5300/6300 Lectures

Orifice Plate Calibrations

• A perhaps better way to calibrate sharp-edged orifice plates in pipes is based on

the following equations
• Flow rate can be calculated through the orifice using the following equation:

2g∆h(sg − 1)
Q = Cd A 2 (3)
1− β 4

where Cd is a dimensionless orifice discharge coefficient, as defined below; A2 is

the cross-sectional area of the orifice plate opening; g is the ratio of weight to
mass; ∆h is the change in piezometric head across the orifice; and, β is a
dimensionless ratio of the orifice and pipe diameters:

D2
β= (4)
D1

where D2 is the diameter of the circular orifice opening; and, D1 is the inside
diameter of the upstream pipe

• In Eq. 3, “sg” is the specific gravity of the manometer fluid, and the constant “1”
represents the specific gravity of pure water
• The specific gravity of the manometer liquid must be greater than 1.0
• Thus, if a manometer is used to measure the head differential across the orifice
plate, the term “∆h(sg - 1)” represents the head in depth (e.g. m or ft) of water
• If both ends of the manometer were open to the atmosphere, and there’s no
water in the manometer, then you will see ∆h = 0
• But if both ends of the manometer are open to the atmosphere, and you pour
some water in one end, you’ll see ∆h > 0, thus the need for the “(sg – 1)” term
• Note that the specific gravity of water can be slightly different than 1.000 when
the water is not pure, or when the water temperature is not exactly 5°C
• See the figure below
• Note also that the manometer liquid must not be water soluble!

BIE 5300/6300 Lectures 149 Gary P. Merkley

 flow

Head of water =
∆h(sg - 1)

sg = 1 ∆h

sg > 1

• The inside pipe diameter, D1, is defined as:

D1 = ⎡⎣1 + αp ( T°C − 20 )⎤⎦ (D1 )meas (5)

in which T°C is the water temperature in °C; (D1)meas is the measured inside pipe
diameter; and αp is the coefficient of linear thermal expansion of the pipe material
(1/°C)

• The coefficient of linear thermal expansion is the ratio of the change in length per
degree Celsius to the length at 0°C
• See the following table for linear thermal expansion values

Gary P. Merkley 150 BIE 5300/6300 Lectures

 Coefficient of Linear
Material
Thermal Expansion (1/°C)
Cast iron 0.0000110
Steel 0.0000120
Tin 0.0000125

Metal
Copper 0.0000176
Brass 0.0000188
Aluminum 0.0000230
Zinc 0.0000325
PVC 0.0000540
Plastic

ABS 0.0000990
PE 0.0001440
Glass 0.0000081
Other

Wood 0.0000110
Concrete 0.0000060 – 0.0000130

• For the range 0 to 100 °C, the following two equations can be applied for the
density and kinematic viscosity of water
• The density of pure water:

ρ = 1.4102(10)−5 T 3 − 0.005627(10)−5 T 2 + 0.004176(10)−6 T + 1,000.2 (6)

where ρ is in kg/m3; and T is in °C

• The kinematic viscosity of pure water:

1
ν= (7)
83.9192 T + 20,707.5 T + 551,173
2

where ν is in m2/s; and T is in °C

• Similarly, the orifice diameter is corrected for thermal expansion as follows:

D2 = ⎡⎣1 + αop ( T°C − 20 )⎤⎦ (D2 )meas (8)

where αop is the coefficient of linear thermal expansion of the orifice plate
material (1/°C); and (D2)meas is the measured orifice diameter

• Note that the water temperature must be substantially different than 20°C for the
thermal expansion corrections to be significant

• The coefficient of discharge is defined by Miller (1996) for a circular pipe and
orifice plate in which the upstream tap is located at a distance D1 from the plate,
and the downstream tap is at a distance ½D1:
BIE 5300/6300 Lectures 151 Gary P. Merkley
 Cd = 0.5959 + 0.0312β2.1 − 0.184β8
0.039β 4 91.71β2.5 (9)
+ − 0.0158β +
3

1 − β4 R0e.75

in which Re is the Reynolds number.

• Similar Cd equations exist for other orifice plate configurations, and for venturis
• The Cd expression for venturis is much simpler than that for orifice plates
• The Reynold’s number is a function of the flow rate, so the solution is iterative
• The calculated value of Cd is typically very near to 0.6, so if this is taken as the
initial value, usually only one or two iterations are needed:

1. Specify T, ∆h, αp, and αop

2. Calculate or specify ρ and ν
3. Calculate D1 and D2
4. Calculate β = D1/D2
5. Let Cd = 0.60
6. Calculate Q
7. Calculate Re
8. Calculate Cd

• Repeat steps 6 - 8 until Q converges to the desired precision

References & Bibliography

Miller, R.W. 1996. Flow measurement engineering handbook. 3rd Ed. McGraw-Hill Book Co., New
York, N.Y.
USBR. 1996. Flow measurement manual. Water Resources Publications, LLC. Highlands Ranch,
CO.

Gary P. Merkley 152 BIE 5300/6300 Lectures

Lecture 14
Flow Measurement in Pipes

I. Elbow Meters

• An elbow in a pipe can be used as a flow measuring

device much in the same way as a venturi or orifice plate
• The head differential across the elbow (from inside to
outside) is measured, and according to a calibration the
discharge can be estimated
• The taps are usually located in the center of the elbow (e.g. at
a 45° angle for a 90° elbow), but can be at other locations
toward the upstream side of the elbow
• Some companies manufacture elbow meters for flow
measurement, but almost any pipe elbow can be calibrated
• Elbow meters are not as potentially accurate as venturi,
nozzle, and orifice meters
• Typical accuracy is about ±4% of Q
• One advantage of elbow meters is that there
need not be any additional head loss in the
piping system as a result of flow
measurement
• The graph below is a sample calibration
curve for a 1½-inch elbow meter in a USU
hydraulic lab where the head differential
(inside to outside tap) is measured in inches
of mercury, using a manometer (data are from Dr. L.S. Willardson)

Calibration for a 1.5-inch Elbow Meter

4.0

3.5
Head differential (inches of Hg)

3.0

2.5

2.0

1.5

1.0

0.5

0.0
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Discharge (cfs)

BIE 5300/6300 Lectures 153 Gary P. Merkley

II. Variable Area Meters

• These are vertical cylinders with a uniformly expanding

cross-section in the upward direction
• A float inside the cylinder stabilizes at a certain elevation
depending on the flow rate through the cylinder

• Note that the outside walls are usually transparent to allow direct readings by eye

III. Horizontal Trajectory Method

• From physics, an accelerating object will travel a distance x in time t according to

the following equation (based on Newton’s 2nd law):

a t2
x = vot + (1)
2
where x is the distance; vo is the initial velocity at time 0; t is the elapsed time;
and a is the acceleration

• Flow emanating from a horizontal pipe will fall a height y over a distance x
• The horizontal component (x-direction) has almost no acceleration, and the
vertical component (y-direction) has an initial velocity of zero
• The vertical acceleration is equal to the ratio of weight to mass, or g = 9.81 m/s2
(32.2 ft/s2)
• Therefore,
gt 2
x = v ot and, y = (2)
2

Gary P. Merkley 154 BIE 5300/6300 Lectures

 • Then by getting rid of t, knowing that Q = VA, and the equation for the area of a
circle, the flow rate is calculated as follows:

π D2 x
Q= (3)
2y
4
g

where D is the inside diameter of the circular pipe

• This equation is approximately correct if x and y are measured to the center of

mass of the discharge trajectory
• Errors occur because in practice it is difficult to measure exactly to the center,
and because of possible wind and other turbulent effects
• Also, the pipe might not be exactly horizontal (although a correction could take
this into account, according to the same analysis given above)
• Tables of coefficient values derived from experiments allow x and y to be
measured from the top of the trajectory
• However, measurements can be difficult because the flow is often very turbulent
at a distance from the pipe end

• The previous equation can be simplified as:

x
Q = 3.151CD2 (4)
y

where C is a coefficient to adjust the calculated discharge value when the ratios
of x/D or y/D are smaller than 8 and 5, respectively (otherwise, C equals unity)

• Equation 4 is valid for x, y and D in ft, and Q in cfs

• The following table is for C values for use with Eq. 4

BIE 5300/6300 Lectures 155 Gary P. Merkley

 y/D x/D
1.00 1.50 2.00 2.50 3.00 4.00 5.00 8.00
0.5 1.44 1.28 1.18 1.13 1.10 1.06 1.03 1.00
1.0 1.37 1.24 1.17 1.12 1.09 1.06 1.03 1.00
2.0 1.11 1.09 1.08 1.07 1.05 1.03 1.00
3.0 1.04 1.04 1.04 1.04 1.03 1.00
4.0 1.01 1.01 1.02 1.03 1.02 1.00
5.0 0.97 0.99 1.00 1.01 1.01 1.00

• This method can also be used for pipes flowing partially full (i.e. A < πD2/4), and
experimental data are available to assist in the estimation of discharge for these
conditions

IV. California Pipe Method

• This is the horizontal pipe method for partially-full pipes

• It is somewhat analogous to the calibration for a weir or
free overfall
• The following equation is in English units:

1.88
⎛ a⎞
Q = 8.69 ⎜ 1 − ⎟ D2.48 (5)
⎝ D⎠

where a and D are defined in the figure below (ft); and Q is discharge in cfs

• The ratio a/D is limited to: a/D > 0.45

• This method was published in the 1920’s
• Measurement accuracy is only ±10%, at best
• The pipe must be exactly horizontal (level), with circular cross section
• The pipe must discharge freely into the air, unsubmerged

Gary P. Merkley 156 BIE 5300/6300 Lectures

V. Vertical Trajectory Method

• As with pipes discharging horizontally into the air, there is a method to measure
the flow rate from vertical pipes
• This is accomplished by assuming a translation of velocity head into the
measurable height of a column of water above the top of the pipe
• Thus, to estimate the flow rate from pipes discharging vertically into the air it is
only necessary to measure the:

1. inside diameter of the pipe, D; and,

2. the height of the jet, H, above the pipe

• This is a nice idea on “paper,” but in practice, it can be difficult to measure the
height of the column of water because of sloshing, surging, and splashing
• Also, the act of measuring the height of the column can significantly alter the
measured value
• The table below gives flow rate values in gpm for several pipe diameters in
inches

Jet Pipe Diameter (inch)

Height 2 3 4 5 6 8 10 12
(inch) (gpm) (gpm) (gpm) (gpm) (gpm) (gpm) (gpm) (gpm)
2 28 57 86 115 150 215 285 355
2½ 31 69 108 150 205 290 385 480
3 34 78 128 183 250 367 490 610
3½ 37 86 145 210 293 440 610 755
4 40 92 160 235 330 510 725 910
4½ 42 98 173 257 365 570 845 1060
5 45 104 184 275 395 630 940 1200
6 50 115 205 306 445 730 1125 1500
7 54 125 223 336 485 820 1275 1730
8 58 134 239 360 520 890 1420 1950
9 62 143 254 383 550 955 1550 2140
10 66 152 268 405 585 1015 1650 2280
12 72 167 295 450 650 1120 1830 2550
14 78 182 320 485 705 1220 2000 2800
16 83 195 345 520 755 1300 2140 3000
18 89 208 367 555 800 1400 2280
20 94 220 388 590 850 1480 2420
25 107 248 440 665 960 1670 2720
30 117 275 485 740 1050 1870 3000
35 127 300 525 800 1150 2020
40 137 320 565 865 1230 2160
From Utah Engineering Experiment Station Bulletin 5, June 1955.
“Jet Height” (first column) is the height from the top of the pipe to the top of the jet.

BIE 5300/6300 Lectures 157 Gary P. Merkley

VI. Vortex Shedding Meters

• The vortex shedding meter can be accurate to within ±½% to ±1% of the true
discharge
• The basic principal is that an object placed in the flow will cause turbulence and
vortices in the downstream direction, and the rate of fluctuation of the vortices
can be measured by detecting pressure variations just downstream

• This rate increases with increasing velocity, and it can be used to give an
estimate of the discharge
• This requires calibration for a particular pipe material, pipe size, element shape
and size, fluid type, and temperature
• It is essentially a velocity-area flow measurement method, but it is calibrated to
give discharge directly

Gary P. Merkley 158 BIE 5300/6300 Lectures

 • Vortex shedding meters are commercially available
and are used with a variety of fluids, not only with
water, and can operate well over a large pressure
range and high flow velocities
• Size selection for these meters is important in order
to avoid cavitation
• There need not be any moving parts in the meter
• This meter can be used for velocities up to
approximately 50 m/s, or 180 kph
• The response of the device is linear for Reynolds
numbers of 10,000 or more
• Errors can result from pipe vibration due to external
machinery or other causes, or when the velocity is
too low; however, some sophisticated devices have
been developed and tested to correct for such errors

VII. Ultrasonic Meters

1. Doppler

• An emitted pressure wave reflects off a deflector plate

• Difference between transmitted and reflected frequencies correlates to
flow velocity
• Liquid does not have to be clean – in fact, it may not work well if the liquid
is “too clean” because it needs particles to reflect the signal

2. Transit-time

• Also called “time-of-flight”

• The liquid should be fairly clean with this method
• Devices generates high-frequency (≈1 MHz) pressure wave(s)
• Time to reach an opposing wall (inside the pipe) depends on:

a) Flow velocity
b) Beam orientation (angle)
c) Speed of sound through the liquid medium

• Upstream straightening vanes may be needed to avoid swirling flow

• May have a single or multiple transmitted sound beams

VIII. Other Measurement Devices

• Collins meters
• Commercial propeller flow meters
• Electromagnetic flow meters
• Volumetric tank

BIE 5300/6300 Lectures 159 Gary P. Merkley

 • Weight tank

References & Bibliography

Brater, E.F. and H.W. King. 1976. Handbook of hydraulics. McGraw-Hill.
Daugherty, R.L. and J.B. Franzini. 1977. Fluid mechanics with engineering applications. McGraw-Hill.
Ginesi, D. 1987. Putting new technology to work in flow measurement. Chilton’s I&CS 60:2:25-28.
Greve, F.W. 1928. Measurement of pipe flow by the coordinate method. Purdue Engrg. Experiment
Station Bulletin #32.
Israelson, O.W. and V.E. Hansen. 1962. Irrigation principles and practices. John Wiley, 3rd Ed., pp.
140-145.
King, L.G. 1974. Drainage laboratory manual. BIE Dept., USU (BIE 605 course notes).
Ledoux, J.W. 1927. Venturi tube characteristics. Trans. ASCE, vol. 91.
Lucas, G.P. and J.T. Turner. 1985. Influence of cylinder geometry on the quality of its vortex
shedding signal. Proc. Int’l Conference on Flow Measurement (FLOWMECO 1985), Univ. of
Melbourne, Australia, pp. 81-89.
Miller, R.W. 1996. Flow measurement engineering handbook. 3rd Ed. McGraw-Hill Book Co., New
York, N.Y.
Sovik, R.E. 1985. Flow measurement - some new considerations. Mech. Engrg., May, 107(5):48-52.
Tily, P. 1986. Practical options for on-line flow measurement. Process Engrg., London, 67(5):85-93.
U.S. Bureau of Reclamation. 1981. Water measurement manual. 2nd Ed., Denver, CO.

Gary P. Merkley 160 BIE 5300/6300 Lectures

Lecture 15
Canal Design Basics

I. Canal Design Factors

• There are a number of criteria to consider when designing canals

• Below is a list of main criteria, not necessarily in order of importance:

1. Flow rate capacity requirements (demand)

2. Expected flow rate entering the canal (supply)
3. Construction cost
4. Safety considerations
5. Hydraulic operational characteristics
6. Water management needs
7. Maintenance requirements
8. Environmental conservation
9. Need for emergency spill structures
10. Cross-channel surface drainage needs
11. Need for drainage directed into the canal
12. Right-of-way (easements) along the canal path
13. Secondary uses (clothes washing, swimming, others)
14. Aesthetics

• Historically, flow rate capacity and construction cost have been the dominant
design criteria, but it is better to take into account all of the above factors before
finalizing a design
• This is not to say that you necessarily have to dwell on an issue like aesthetics,
for example
BIE 5300/6300 Lectures 161 Gary P. Merkley
 • However, issues such as dynamic operation, maintenance requirements and
need for spillways have often been given only cursory attention during the design
phase, requiring subsequent post-construction modifications to the infrastructure
• Water management and operational needs are very similar
• Secondary uses can
include things like
navigation (large canals),
clothes washing, other
domestic uses, aquatic
production, bathing, and
many others
• Remember that every
design has both common
and unique (site-specific)
features, compared to
other canals

II. Capacity-Based Design

• This is an important consideration because a canal must have sufficient capacity,

but not “too much”
• Construction and maintenance costs increase significantly with larger canals
• Actual required delivery system capacity depends on:

1. size of the irrigated area

2. cropping patterns (crop types, planting & rotation schedules)
3. climatological conditions
4. conveyance efficiencies
5. on-farm efficiencies
6. availability & exploitation of other water sources (conjunctive use)
7. type of delivery schedule (continuous, rotation, on-demand)
8. non-agricultural water needs

• It is often recommendable to allow for a safety factor by increasing capacities by

10% to 20% in case crops change, an expansion in irrigated area occurs,
conveyance losses increase, and other possible factors
• The magnitude of design safety factors is very subjective and debatable
• Capacity requirements can change with different crop types, different total area,
different planting schedules, and different efficiencies due to maintenance and
rehabilitation (or lack thereof)
• On-demand delivery schedules require higher capacities because the combined
requirements will tend to peak higher at some point during each growing season
(on-demand delivery schemes can fail when there is not enough water or not
enough conveyance capacity)
• Administrative losses can be significant, especially if the delivery schedule is very
flexible (need extra water running through the system to buffer sudden turnout
deliveries, else spill the excess)
Gary P. Merkley 162 BIE 5300/6300 Lectures
 • The required design flow rate capacity is usually known from independent
calculations
• For example, irrigation project canal capacities are based on peak crop
evapotranspiration requirements and net irrigated area
• A typical main canal capacity is approximately 1 lps per irrigated hectare
• Irrigation canal capacity may also be partially based on non-irrigation
requirements, such as municipal supply, industry, fishery & wildlife conservation,
and others
• Of course, the capacity of the canals will depend on location, whereby the
capacity requirements tend to decrease in the downstream direction due to
deliveries in upstream reaches
• But, if the capacity for each reach is known based on crop water and other
requirements, and one or more canal layouts have been identified, the design
problem becomes one of cross-sectional shape and size, and longitudinal bed
slope

• An important point in capacity-based designs is that most canal designs are

“static”, based only on the hydraulic ability to carry up to a specified maximum
flow rate
• The problem with this is that many designs did not consider the “dynamics” of
canal operation, nor the type of delivery schedules envisioned
• This oversight has caused many operational difficulties and has limited the
operational flexibility of many systems, sometimes severely
• The dynamics of canal operation can be taken into account through design-
phase modeling, either with physical models or mathematical models

• In earthen canals, and for canals in general, the most efficient cross section is a
secondary consideration to erodibility, maintenance, safety, and convenience
• The ratio of flow depth, h, to canal bottom width, b, usually varies from 1:2 to 1:4
for small canals, and from 1:4 to 1:8 in large canals
• Freeboard can be designed into the canal size at ¼ of the maximum water depth
plus one foot (maximum of 6 ft)
• Less freeboard is required if the canal is carefully controlled during operation
• Top width of the bank should allow for a vehicle to pass on one side; the other
side can be more narrow

III. System Layout Considerations

• A primary concern in the layout of the system is that it serves the purpose of
conveying and distributing water to key locations in the area of service
• Another concern is that the excavation and earthen fill volumes not be excessive
• When large volumes of excavation and or fill are required, the construction costs
can increase tremendously
• In fill areas, compaction of the soil material is very important, to avoid settlement
problems and possible structural failure

BIE 5300/6300 Lectures 163 Gary P. Merkley

 • In reaches constructed over fill, the seepage
losses tend to be high, even if the canal is
lined Central Arizona Project
• For these reasons, canals are often designed canal (USBR)
to follow the existing topography for the
design bed slope, which often means routing
the canals indirectly so that earth moving
work can be minimized, or at least held to an
acceptable level
• The selection of longitudinal bed slope should
also take into account the existing slopes of
the terrain, so as to minimize deviations in
canal routing
• Curves in canals should not be too sharp;
following are some recommended limits:

Channel Capacity Minimum Curve

(m3/s) Radius (m)
< 15 300
15-30 600
30 -90 1,000
> 90 1,500

• In bends, the radius of curvature should usually be between 3 and 7 times the
top width of flow at maximum design discharge (larger radius for larger canals)

IV. Designing for Maximum Discharge and Uniform Flow

• For a known design discharge, known longitudinal bed slope, and selected cross-
sectional shape, the Manning or Chezy equation can be solved for the required
depth
• Or, for a known design discharge, known longitudinal bed slope, and specified
maximum depth, the Manning equation can be solved for the required base width
of a rectangular section
• In general, the equation can be solved for any “unknown”, where all other
parameters are specified
• You can also go to the field and measure everything but roughness under
steady, uniform flow conditions, then calculate the value of n
• Avoid critical flow at or near design discharge (unstable water surface)

V. Manning Equation

• The Manning equation has been used to size canals all over the world
• It is an empirical equation for approximating uniform flow conditions in open
channels
• A roughness factor, n, is used in the equation

Gary P. Merkley 164 BIE 5300/6300 Lectures

 • This factor is dependent on the type and condition of the canal lining
• But in reality, the factor also depends on the Reynold’s number, NR (that is, the
size and shape of the cross section, not just the roughness of the lining material)
• In practice, it is often erroneously assumed that n is independent of NR

• In Metric units:
1
Q= AR2 / 3 So (1)
n
where Q is in m3/s; A is cross-section flow area (m2); R is hydraulic radius (m),
equal to A divided by wetted perimeter; and So is the longitudinal bed slope
(dimensionless)

• In English units, a coefficient must be added to the equation:

1
≈ 1.49 (2)
(0.3048 m / ft)1/ 3

• In English units:
1.49
Q= AR2 / 3 So (3)
n
where Q is in cfs; A is in ft2; and R is in (ft)

• An alternative to the Manning equation is the Chezy equation

VI. Chezy Equation

• The Chezy equation is an alternative to the Manning equation, and can be

applied as described above
• It is also an empirical equation for approximating uniform flow conditions in open
channels, but it has more of a theoretical basis
• The Chezy equation has a diagram analogous to the Moody diagram for the
Darcy-Weisbach equation (pipe flow head loss) that takes the Reynold’s number
into account, which makes it technically more attractive than the Manning
equation
• Another advantage is that the Chezy equation can be applied successfully on
steeper slopes than the Manning equation

Q = CA RSo (4)

where Q is in m3/s; A is cross-section flow area (m2); R is hydraulic radius (m),

equal to A divided by wetted perimeter; and So is the longitudinal bed slope
(dimensionless)
BIE 5300/6300 Lectures 165 Gary P. Merkley
VII. Chezy C Value

• The units of C are m½/s

• Note that the numerical value of C increases for smoother surfaces, which is
opposite to the behavior of Manning’s n
• The relationship between C and Manning’s n is (for m and m3/s):

R1/ 6
C= (5)
n
• The relationship between C and the Darcy-Weisbach f is:

8g
C≈ (6)
f
• Thus, C can be defined as a function of relative roughness (ε/R) and Reynold’s
number, and the resulting graph looks much like the Moody diagram, vertically
inverted
• Reynold’s number can be defined like this:

4RV
NR = (7)
ν

where R is the hydraulic radius (m), A/Wp; V is the mean flow velocity in a cross
section (m/s); and ν is the kinematic viscosity of water (m2/s)

• For a full circle:

A πr 2 r
R= = = (8)
Wp 2πr 2

whereby 4R = D (diameter), so use 4R in general for non-circular sections

• Kinematic viscosity is a function of water temperature

Water Temperature (°C) Kinematic Viscosity (m2/s)

0 0.000001785
5 0.000001519
10 0.000001306
15 0.000001139
20 0.000001003
25 0.000000893
30 0.000000800
40 0.000000658
50 0.000000553
60 0.000000474

Gary P. Merkley 166 BIE 5300/6300 Lectures

BIE 5300/6300 Lectures 167 Gary P. Merkley
 • For laminar flow (NR < 2000) and units of m and m3/s:

C = 1.107 NR (9)

which is analogous to the Blasius equation (Darcy-Weisbach f)

• For turbulent smooth flow (NR > 2000 & ε ≈ 0) and units of m and m3/s:

⎛ 0.28C ⎞
C = −17.7 log10 ⎜ ⎟ (10)
⎝ NR ⎠

• For turbulent transitional flow (NR > 2000 & ε > 0) and units of m and m3/s:

⎛ ε / R 0.28C ⎞
C = −17.7 log10 ⎜ + ⎟ (11)
⎝ 12 NR ⎠

• For turbulent rough flow (NR > 20,000 & ε > 0), where C is no longer a function of
NR, and units of m and m3/s:

⎛ 12 ⎞
C = 17.7 log10 ⎜ ⎟ (12)
⎝ ε /R ⎠

which gives the flat (horizontal) lines for fully turbulent flow

• To determine the threshold between turbulent transition and turbulent rough flow
for a given ε/R ratio, first determine C from Eq. 11, then calculate NR as:

75C
NR = (13)
ε /R

• Other equations exist to define the C value as a function of NR and relative

roughness, and these can be found in hydraulics textbooks & handbooks
• For R in ft, A in ft2, and Q in cfs, multiply C by:

0.3048
= 2.006 (14)
0.3048

• That is, in English units:

Q = 2.006 C A RSo (15)

Gary P. Merkley 168 BIE 5300/6300 Lectures

 where Q is in cfs; A is in ft2; and R is in ft

• Note that for all but laminar flow, you must iterate to solve for C
• This can be done quickly and easily in a computer program, and the results can
be presented as in the graph above

VIII. Chezy Epsilon Values

• Epsilon (roughness height) values depend on channel lining material type &
condition:

Material & Condition ε (m)

Very smooth and essentially seamless concrete 0.0003
Smooth concrete with joints between panels 0.0005
Rough concrete surfaces 0.0012
Very rough concrete surfaces 0.004 to 0.005
Gunite with a smooth finish 0.0005 to 0.0015
Untreated gunite 0.003 to 0.010

References & Bibliography

Davis, C.V. and K.E. Sorensen (eds.). 1969. Handbook of applied hydraulics. McGraw-Hill Book
Company, New York, N.Y.
Labye, Y., M.A, Olsen, A. Galand, and N. Tsiourtis. 1988. Design and optimization of irrigation
distribution networks. FAO Irrigation and Drainage Paper 44, Rome, Italy. 247 pp.
USBR. 1974. Design of small canal structures. U.S. Government Printing Office, Washington, D.C.
435 pp.

BIE 5300/6300 Lectures 169 Gary P. Merkley

Gary P. Merkley 170 BIE 5300/6300 Lectures
Lecture 16
Channel Cross Sections

I. Channel Cross Section Parameters

• Common cross-sectional shapes are rectangular, trapezoidal and circular

• Geometrically, rectangular cross-sections are just special cases of trapezoidal
sections
• Circular cross sections are hydraulically more “efficient” than other cross-
sectional shapes, but they are only used for small channel sizes
• Circular cross sections are usually made of precast concrete mixes, and elevated
above the ground
• Tunnels designed for open-channel flow are sometimes built with special cross
sections (e.g. “horseshoe” sections)
• The standard horseshoe cross section has a semicircular top portion, and an
intersection of three larger circles for the lower portion – it can be considered a
modification of a circular section
• A semi-circular channel cross section is the best shape for an open-channel,
including open-channel flow in tunnels, but the horseshoe shape has been used
in dozens of tunnels to allow for greater floor width, thereby facilitating the
passage of equipment through the tunnel

Standard horseshoe cross section (bold curves): diameter of upper section

is half of the diameter of the three larger circles

BIE 5300/6300 Lectures 171 Gary P. Merkley

 • Trapezoidal cross sections can be symmetrical or non-symmetrical

• For trapezoidal cross sections, the inverse side slope (H:V) is usually
between zero and 2.0
• Common inverse side slopes are zero (rectangular section), 1.0 and 1.5
• There are tradeoffs between low and high values of side slope:

1. canals with low inverse side slopes occupy less land area
2. high inverse side slopes are more stable and may require
less maintenance
3. high inverse side slopes are safer, if animals or people could
fall into the canal, because it is easier to climb out
4. rectangular cross sections can be simpler to build, when lined
with concrete (especially for small cross sections)
5. it may be easier to build and install structures and transitions
for rectangular sections
6. medium-range side slopes correspond to greater hydraulic
efficiency

• Compound sections are not uncommon

• For example, a combination of a trapezoidal and rectangular section:

Gary P. Merkley 172 BIE 5300/6300 Lectures

II. Freeboard Recommendations

• Freeboard means the extra depth of a canal section, above the water surface for
100% flow rate capacity, usually for uniform-flow conditions
• A freeboard value should be added to the maximum expected depth to allow for:

1. deviations between design and construction (these are

ubiquitous, only varying in magnitude from place to place)
2. post-construction, non-uniform land settlement
3. operational flexibility (including operator mistakes)
4. accommodate transient flow conditions
5. provide a more conservative design (in terms of flow
capacity)
6. increase in hydraulic roughness due to lining deterioration,
weed growth, and for other reasons
7. wind loading
8. other reasons

• Thus, with freeboard, under maximum flow conditions (full supply level, FSL),
canal overflow is not impending
• If the canal starts to overflow, enormous erosive damage can occur in just a few
minutes
• According to Murphy’s Law, these things usually happen about 3:00 am, when no
one is around. Then, everyone finds out at about 6:30 am after it has been
spilling for hours.
• Many reaches of canal in many countries are routinely operated with virtually no
freeboard, and disasters often occur
BIE 5300/6300 Lectures 173 Gary P. Merkley
 A canal which is overtopping the banks and spilling water.

A canal with impending spillage (zero freeboard).

Gary P. Merkley 174 BIE 5300/6300 Lectures

 • Traditional wisdom says that the minimum design freeboard for “small canals” is
1 ft (0.3048 m), so this would be the minimum for most any canal, but for very
small canals it would certainly be excessive
• For canals with flow rates up to 3,000 cfs (85 m3/s), the freeboard should be up
to 4 ft (1.2 m), and for intermediate flow rates, the freeboard should be 1 ft plus
25% of the maximum expected water depth
• However, the design freeboard can extend above the lined portion of a concrete-
lined canal, and it often does (the berm of a lined canal is almost always higher
than the top of the lining)
• Also, considerable judgment is required to know what the required freeboard
might actually be, including knowledge of the operational modes in the canal
(these will help determine the need for freeboard)
• Special analysis should go into the determination of required freeboard for canals
with capacities exceeding 3,000 cfs, but such analysis can also be included for
any size canal
• The analysis should also include economic criteria, because on large and or long
canals, a difference of a few centimeters in freeboard is likely to mean a
difference of millions of dollars in construction costs
• Nevertheless, the USBR published data on freeboard guidelines for up to 20,000
cfs (560 m3/s) capacities, and these are found in the plots below
• To put things in perspective, note that very few irrigation canals exceed a
capacity of 100 m3/s; most main canals have a capacity of less than 20 m3/s

BIE 5300/6300 Lectures 175 Gary P. Merkley

III. Trapezoidal Cross Section

Symmetrical section:

A = h ( b + mh )
(16)

T = b + 2mh (17)

Wp = b + 2h m2 + 1
(18)
Gary P. Merkley 176 BIE 5300/6300 Lectures
 w = h m2 + 1 (19)

T −b
m=
4w 2 − (T − b)2 (20)

h2 ⎛ 2hm ⎞
hc = ⎜b + ⎟ (21)
2A ⎝ 3 ⎠

h2 ⎛ 4hm ⎞
h= ⎜b + ⎟ (22)
2A ⎝ 3 ⎠

where h is the depth from the bottom (or “invert”) of the cross section up to the
centroid of the cross-sectional area; and hc is the depth from the water surface
down to the area centroid:

hc = h − h (23)

IV. Nonsymmetrical Trapezoidal Cross Section

A = h [b + 0.5(m1 + m2 )h]
(24)

T = b + h ( m1 + m2 )
(25)

Wp = b + h ⎛⎜ m12 + 1 + m22 + 1 ⎞⎟
⎝ ⎠ (26)

w = h m12 + 1 or h m22 + 1
(27)
BIE 5300/6300 Lectures 177 Gary P. Merkley
 h2 ⎡ h ⎤
hc = ⎢b + ( m1 + m2 ) ⎥ (28)
2A ⎣ 3 ⎦

h2 ⎡ 2h ⎤
h= ⎢
2A ⎣
b + ( m1 + m2 ) ⎥ (29)
3 ⎦

V. Circular Cross Section

T
β

hc
h
h

• In the following, angle β is in radians

⎛ 2h ⎞
β = 2cos −1 ⎜ 1 − ⎟ (30)
⎝ D⎠

D2
A= (β − sinβ ) (31)
8
or,
⎡ h−r ⎞ π⎤
A = (h − r) 2hr − h2 + r 2 ⎢sin−1 ⎛⎜ ⎟+ (32)
⎣ ⎝ r ⎠ 2 ⎥⎦

Gary P. Merkley 178 BIE 5300/6300 Lectures

 where r = D/2
2
⎛β⎞
T = D sin ⎜ ⎟ = D 1 − ⎛⎜ 1 −
2h ⎞
⎟ (33)
⎝2⎠ ⎝ D⎠

βD
Wp =
2 (34)

D⎛ β⎞
h= ⎜ 1 − cos ⎟
2⎝ 2⎠ (35)

D 2
h= − (hD − h2 )3 / 2 (36)
2 3A

hc = h − h (37)

Circular Channel with D = 1.0

7.0 0.50

beta 0.45
6.0 Top Width

Hydraulic radius & depth to centroid

Area 0.40

Wetted Perimeter
5.0
0.35
Hydraulic Radius
Depth to Centroid
Beta, T, A, Wp

0.30
4.0

0.25

3.0
0.20

0.15
2.0

0.10

1.0
0.05

0.0 0.00
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Depth

Nondimensional curves of circular cross-section geometry.

BIE 5300/6300 Lectures 179 Gary P. Merkley

VI. Standard Horseshoe Cross Section

r
r/2 h3

h2
h1

• The following is for a “standard” horseshoe cross section

• Divide the depth into three segments
• Note that h1 + h2 + h3 = r (see the above figure)
• Determine h1 by solving for the intersection of two of the circles

⎡ ⎛ 1 + 7 ⎞⎤
h1 = r ⎢1 − ⎜⎜ ⎟⎟ ⎥ (38)
⎢⎣ ⎝ 4 ⎠ ⎥⎦

then,
r
h2 = − h1
2 (39)

and,
r
h3 =
2 (40)

• In the following, all angles are in radians

Gary P. Merkley 180 BIE 5300/6300 Lectures
Top width (at water surface):

• For 0 < h ≤ h1:

2
T = 2r 1 − ⎛⎜ 1 − ⎞⎟
h
(41)
⎝ r⎠

• For h1 < h ≤ r/2:

⎛ 2 ⎞ ⎛ 2
r⎞
⎛ r ⎞ r ⎛ r ⎞
T = ⎜ r − ⎜h − ⎟ − ⎟ − ⎜ − r − ⎜h − ⎟ + ⎟
2 2
(42)
⎜ ⎝ 2⎠ 2⎟ ⎜ ⎝ 2⎠ 2⎟
⎝ ⎠ ⎝ ⎠

or,
2
⎛ r⎞
T = 2 r − ⎜h − ⎟ − r
2

⎝ 2⎠ (43)

• For r/2 < h < r:

2
T = r 1 − ⎜⎛ 1 −
2h ⎞
⎟ (44)
⎝ r ⎠

and T = 0 when h = 0 or h = r

Cross-sectional area:

• For 0 ≤ h ≤ h1:

⎡ h−r ⎞ π⎤
A = (h − r) h ( 2r − h ) + r 2 ⎢sin−1 ⎛⎜ ⎟+ (45)
⎣ ⎝ r ⎠ 2 ⎥⎦

• For h1 < h ≤ r/2:

⎛ 1 ⎞
A = r 2 ⎜ α 2 − α1 − [cot (α1 ) − cot (α 2 )] ⎟ − A a + A b + A1 (46)
⎝ 4 ⎠

where A1 is the cross-sectional area corresponding to h = h1; cot(α1) is the

cotangent of α1, equal to 1/tan(α1); and,

BIE 5300/6300 Lectures 181 Gary P. Merkley

 r 2 − h22
ϕ1 = (47)
h2

2
⎛r ⎞
r − ⎜ − h⎟
2

⎝2 ⎠
ϕ2 = (48)
r
−h
2

α1 = tan−1 ( ϕ1 )
(49)

α 2 = tan−1 ( ϕ2 )
(50)

2
1 ⎛⎜ 2 ⎛ r r ⎞⎟
2
⎞
Aa = r − ⎜ − h⎟ − (51)
ϕ2 ⎜ ⎝2 ⎠ 2⎟
⎝ ⎠

and,
2
1⎛ r⎞
Ab = ⎜ r 2 − h22 − ⎟ (52)
ϕ1 ⎝ 2⎠

• Note that h2 = r/2-h1

• Note that α1 and Ab are constants for a given value of r
• Note that α2 = π/2 and Aa = 0 when h = r/2

• Another way to calculate this (h1 < h ≤ r/2) area is by integration:

y2 y2 ⎛ r 2⎞
A = 2∫ x dy + A1 = 2 ∫ 2
⎜ − + r − y ⎟ dy + A1 (53)
y1 y1 ⎝ 2 ⎠

which yields the following expression:

y
⎡ 2 2 2 −1 ⎛ y ⎞ ⎤
2
A = ⎢ −r y + y r − y + r sin ⎜ ⎟ ⎥ + A1 (54)
⎣ ⎝ r ⎠ ⎦ y1

where y2 and y1 are the integration limits:

Gary P. Merkley 182 BIE 5300/6300 Lectures

 r
y2 = h − (55)
2

r C1
y1 = (56)
2
where,
⎛ 1+ 7 ⎞
C1 = 1 − ⎜⎜ ⎟⎟ (57)
⎝ 2 ⎠

and,
C1 ⎛⎜ C12 ⎞
⎟ − sin−1 ⎛⎜ C1 ⎞⎟
C2 = 1− 1− (58)
2 ⎜ 4 ⎟ ⎝ 2 ⎠
⎝ ⎠

Finally, applying the integration limits:

⎛ 2⎞
2⎡ −1 ⎛ 2h − r ⎞ ⎤
⎛ r ⎞⎜ 2 ⎛ r⎞ ⎟
A = r ⎢C2 + sin ⎜ ⎟ ⎥ − ⎜ h − ⎟ ⎜ r − r − ⎜ h − ⎟ ⎟ + A1 (59)
⎣ ⎝ 2r ⎠⎦ ⎝ 2⎠ ⎝ 2⎠
⎝ ⎠

where A1 is the area corresponding to h = h1 (Eq. 45)

• Equation 59 is preferred over Eq. 46 because it is simpler and yields the same
result for h1 < h ≤ r/2

• For r/2 < h ≤ r:

⎛ r⎞ r2 2h − r ⎞
A = ⎜ h − ⎟ h ( r − h ) + sin−1 ⎛⎜ ⎟ + A2 (60)
⎝ 2⎠ 4 ⎝ r ⎠

where A2 is the area corresponding to h = r/2 (Eq. 59)

Depth to area centroid:

• For 0 ≤ h ≤ h1:

r3
h=
A
⎡π −1 ⎛ h ⎞⎤
⎢ 2 − sin ⎜ 1 − r ⎟ ⎥ −
⎣ ⎝ ⎠⎦
h(2r − h)
3A
(
hr − 2h2 + 3r 2 ) (61)

BIE 5300/6300 Lectures 183 Gary P. Merkley

 where A is as calculated by Eq. 45; and h is the depth measured from the area
centroid to the bottom of the cross section

• For h1 < h ≤ r/2, the moment of area with respect to x is:

Mx = ∫ ( yx ) dy = ∫ y ⎛⎜ −r + 2 r 2 − y 2 ⎞⎟ dy
⎝ ⎠
Mx = −r ∫ y dy + 2∫ y r 2 − y 2 dy (62)
y2
⎡ ry 2 2 2 3/2⎤
Mx = ⎢ −
⎢⎣ 2 3
− r − y2( ) ⎥
⎥⎦ y
1

where y1 and y2 are integration limits, exactly as defined above for cross-
sectional area. Applying the integration limits:

2 3/2
r⎛ r⎞ 2⎡ 2⎤
Mx = r C3 − ⎜ h − ⎟ − ⎢r 2 − ⎜⎛ h − r ⎟⎞ ⎥
3
(63)
2⎝ 2⎠ 3⎣ ⎝ 2⎠ ⎦

where,
3/2
C12 2 ⎛ C12 ⎞
C3 = + ⎜1 − ⎟ (64)
8 3 ⎜⎝ 4 ⎟⎠

where C3 is a constant; and C1 is as defined in Eq. 57

• The value of Mx will be negative because it is calculated based on coordinate

origins at h = r/2, so the depth to centroid for a given depth, h, must be shifted
upward by the amount r/2:

r Mx
hx = + (65)
2 Ax

which will be a positive value, with Ax being the cross-sectional area

corresponding to the same integration limits, y1 and y2:

⎛ 2⎞
2⎡ −1 ⎛ 2h − r ⎞ ⎤
⎛ r ⎞⎜ 2 ⎛ r⎞ ⎟
A x = r ⎢C2 + sin ⎜ ⎟⎥ − ⎜ h − ⎟ ⎜ r − r − ⎜ h − ⎟ ⎟ (66)
⎣ ⎝ 2r ⎠ ⎦ ⎝ 2⎠ ⎝ 2⎠
⎝ ⎠

and C2 is also as previously defined

Gary P. Merkley 184 BIE 5300/6300 Lectures
 • The composite value of h must account for the calculations up to h = h1, so for
depths from h1 to r/2, the following area-weighted relationship is used to obtain
the exact depth to the area centroid:

⎛r M ⎞
A x ⎜ + x ⎟ + A1h1
h= ⎝ 2 Ax ⎠ (67)
A x + A1

where A1 and h 1 are the values corresponding to h = h1 (Eqs. 45 and 61)

• For r/2 < h ≤ r, the moment of area with respect to x is:

r3 2
− [h(r − h)]
3/2
Mx = (68)
12 3

• The cross-sectional area from r/2 up to some h value is:

2
r ⎛ 2h − r ⎞
A x = ⎜⎛ h − ⎟⎞ h(r − h) + sin−1 ⎜
r
⎟ (69)
⎝ 2⎠ 4 ⎝ r ⎠

which is Eq. 60 minus the A2 term

• The composite value of h must account for the calculations up to h = r/2, so for
depths from r/2 to r, the following area-weighted relationship is used to obtain the
exact depth to the area centroid:

⎛r M ⎞
A x ⎜ + x ⎟ + A 2h2
h= ⎝ 2 Ax ⎠ (70)
A x + A2

where A2 and h 2 are the values corresponding to h = r/2 (Eqs. 59 and 67)

Wetted perimeter:

• For 0 ≤ h ≤ h1:

⎛ h⎞
Wp = 2r cos−1 ⎜ 1 − ⎟ (71)
⎝ r⎠

• For h1 < h ≤ r/2:

BIE 5300/6300 Lectures 185 Gary P. Merkley

 ⎡ ⎛ r − 2h ⎞ −1 ⎛ C1 ⎞ ⎤
Wp = 2r ⎢cos−1 ⎜ ⎟ − cos ⎜ − ⎟ ⎥ + Wp1 (72)
⎣ ⎝ 2r ⎠ ⎝ 2 ⎠⎦

where C1 is as defined in Eq. 57; and Wp1 is the wetted perimeter corresponding
to h = h1 (Eq. 71)

• Note that the term “cos-1(-C1/2)” is a constant, based on h1

• For r/2 < h ≤ r:

⎡ ⎛ 2h ⎞ π ⎤
Wp = r ⎢cos−1 ⎜ 1 − ⎟ − 2 ⎥ + Wp 2 (73)
⎣ ⎝ r ⎠ ⎦

where Wp2 is the wetted perimeter corresponding to h = r/2 (Eq. 72)

Standard Horseshoe Cross Section

1.0 5.0

0.9 4.5

0.8 4.0
Top width
0.7 3.5
Area
T/r, A/r , and hbar/r

Centroid depth
0.6 3.0
Wetted perimeter

Wp/r
0.5 2.5
2

0.4 2.0

0.3 1.5

0.2 1.0

0.1 0.5

0.0 0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Depth, h/r

Nondimensional geometric values in a standard horseshoe cross section.

• The increase in area with the standard horseshoe cross section (compared to a
circular section with a diameter of r) is only about 5.6% for a full section

Gary P. Merkley 186 BIE 5300/6300 Lectures

VII. “Efficient” Canal Sections

• Sometimes it is useful to apply an “efficient” cross section to maximize

channel capacity for a given bed slope and roughness
• However, other considerations such as side slope stability, safety, and lining
material may be more important
• Comparisons between the most efficient section and other sections show that
relative changes in the section often do not affect the capacity significantly –
capacity is much more sensitive to changes in roughness and bed slope

VIII. Most Efficient Trapezoidal Canal Section

• How to calculate the most efficient trapezoidal cross section?

• Minimize the wetted perimeter with respect to cross-sectional area of flow (or
depth), or maximize the hydraulic radius (R = A/Wp)
• Express the wetted perimeter as a function of A, m, and h, where h is depth
• Keep area, A, as a constant, otherwise you will get Wp = 0 for the most
efficient section
• Differentiate Wp with respect to depth, h, and set it equal to zero
• For a symmetrical trapezoidal cross section:

A = h(b + mh) (1)

Wp = b + 2h m2 + 1 (2)

1. Write the wetted perimeter in terms of A, h, and m (get rid of b by combining

Eqs. 1 and 2):
A
b= − mh (3)
h

A
Wp = − mh + 2h m2 + 1 (4)
h
2. Differentiate Wp with respect to h (A and m constant) and equate to zero (to
minimize Wp for a given area):

∂Wp −A
= 2
− m + 2 m2 + 1 = 0 (5)
∂h h

BIE 5300/6300 Lectures 187 Gary P. Merkley

 3. Solve Eq. 5 for A

A = h2 ⎛⎜ 2 m2 + 1 − m ⎞⎟ (6)
⎝ ⎠

4. For R = A/Wp, use Eq. 6 to obtain

h2 ⎛⎜ 2 m2 + 1 − m ⎞⎟
R= ⎝ ⎠ (7)
b + 2h m2 + 1
5. Now, manipulate Eq. 7

bh + 2h2 m2 + 1 − bh − mh2
R= (8)
2
b + 2h m + 1

h ⎛⎜ b + 2h m2 + 1 ⎞⎟ − h ( b + mh )
R= ⎝ ⎠ (9)
b + 2h m2 + 1

hWp − A
R= = h−R (10)
Wp

6. Therefore, h = 2R, or,

h
R= (11)
2
You could also directly manipulate Eq. 6 to get the same result: 2A = hWp

7. For the most efficient rectangular section,

A bh h
R= = = (12)
Wp b + 2h 2

which results in b = 2h (bed width twice the maximum flow depth).

8. For the most efficient trapezoidal section we will get Wp = T + b, where T is the
top width of flow (b+2mh), which for a symmetrical trapezoid means that the
length of each side slope (for depth h) is T/2. It also means that b = T/2, and this
corresponds to half of a regular six-sided polygon, or a hexagon. The interior
angle of a hexagon is 120 degrees, so m = 1/tan(60°) = 0.577.

Gary P. Merkley 188 BIE 5300/6300 Lectures

IX. Parabolic Canal Section

• Suppose you have a parabolic channel section…

• Define half of a symmetrical parabolic section as:

h = Kx 2 (13)

• The cross-sectional area of flow (for half of the section) is:

h h h 2h3 / 2
A = ∫ x dh = ∫ dh = (14)
0 0 K 3 K

• The wetted perimeter (again, half of the section) is:

2
∑
2
Wp = lim ⎡⎣ f ( x + ∆x ) − f ( x ) ⎤⎦ + ( ∆x ) (15)
∆x →0

or,
1/ 2
⎛ 2 ⎞
Wp =
∫ ⎜ ⎛ df ⎞ + 1⎟
⎜ ⎜⎝ dx ⎟⎠
⎝
⎟
⎠
dx (16)

where f = Kx2. This derivation can be described graphically as follows:

BIE 5300/6300 Lectures 189 Gary P. Merkley

 where the curve is broken up (discretized) into successive linear segments...

2 2
∆s ≈ ⎡⎣ f ( x + ∆x ) − f ( x ) ⎤⎦ + ( ∆x ) (17)

• For y = f(x) = Kx2,

2
⎛ df ⎞ 2 2
⎜ ⎟ = 4K x
⎝ dx ⎠ (18)

• Then,
h/K
Wp = ∫ 4K 2 x 2 + 1 dx (19)
0

• After integration (using integration tables), the wetted perimeter for half of the
parabolic section is:

h⎛h 1 ⎞ 1 ⎡ ⎛ h h 1 ⎞⎤
Wp = K + + ln ⎢ 2K ⎜ + + ⎟⎟ ⎥ (20)
K ⎜⎝ K 4K 2 ⎟⎠ 4K ⎢⎣ ⎜⎝ K K 4K 2 ⎠ ⎥⎦

which of course is a function of both K (curvature) and depth (h)

• An analysis of the hydraulic radius for such a parabolic section shows that the
hydraulic radius decreases monotonically as K increases from an infinitesimally
small value, so there is no “most efficient” value of K
• Chow (1959) has some equations (exact and approximate) for various channel
section shapes, including the parabola defined in this case

References & Bibliography

Davis, C.V. and K.E. Sorensen (eds.). 1969. Handbook of applied hydraulics. McGraw-Hill Book
Company, New York, N.Y.
Hu, W.W. 1973. Hydraulic elements for USBR standard horseshoe tunnel. J. of the Transportation
Engrg. Div., ASCE, 99(4): 973-980.
Hu, W.W. 1980. Water surface profile for horseshoe tunnel. Transportation Engrg. Journal, ASCE,
106(2): 133-139.
Labye, Y., M.A, Olsen, A. Galand, and N. Tsiourtis. 1988. Design and optimization of irrigation
distribution networks. FAO Irrigation and Drainage Paper 44, Rome, Italy. 247 pp.
USBR. 1963. Linings for irrigation canals. U.S. Government Printing Office, Washington, D.C. 149 pp.

Gary P. Merkley 190 BIE 5300/6300 Lectures

Lecture 17
Design of Earthen Canals

I. General

• Much of this information applies in general

to both earthen and lined canals
• Attempt to balance cuts and fills to avoid
waste material and or the need for “borrow
pits” along the canal
• It is expensive to move earth long
distances, and or to move it in large
volumes
• Many large canals zigzag across the terrain
to accommodate natural slopes; this makes
the canal longer than it may need to be, but A borrow pit
earthwork is less
• Canals may also follow the contours along hilly or mountainous terrain
• Of course, canal routing must also consider the location of water delivery points
• In hilly and mountainous terrain, canals generally follow contour gradients equal
to the design bed slope of the canal
• Adjustments can be made by applying geometrical equations, but usually a lot of
hand calculations and trial-and-error are required
• As previously discussed, it is generally best to follow the natural contour of the
land such that the longitudinal bed slope
is acceptable
• Most large- and medium-size irrigation
canals have longitudinal slopes from
0.00005 to 0.001 m/m
• A typical design value is 0.000125 m/m,
but in mountainous areas the slope may
be as high as 0.001 m/m: elevation
change is more than enough
• With larger bed slopes the problems of
sedimentation can be lessened An earthen channel
• In the technical literature, it is possible to
find many papers and articles on canal design, including application of
mathematical optimization techniques (e.g. FAO Irrig & Drain Paper #44), some
of which are many years old
• The design of new canals is not as predominant as it once was

II. Earthen Canal Design Criteria

• Design cross sections are usually trapezoidal

• Field measurements of many older canals will also show that this is the range of
averaged side slopes, even though they don’t appear to be trapezoidal in shape
BIE 5300/6300 Lectures 191 Gary P. Merkley
 • When canals are built on hillsides, a berm on the uphill side should be
constructed to help prevent sloughing and landslides, which could block the canal
and cause considerable damage if the canal is breached

III. Earth Canal Design: Velocity Limitations

• In designing earthen canals it is necessary to consider erodibility of the banks

and bed -- this is an “empirical” exercise, and experience by the designer is
valuable
• Below are four methods applied to the design of earthen channels
• The first three of these are entirely empirical
• All of these methods apply to open channels with erodible boundaries in alluvial
soils carrying sediment in the water

1. Kennedy Formula
2. Lacey Method
3. Maximum Velocity Method
4. Tractive-Force Method

1. Kennedy Formula

• Originally developed by British on a canal system in Pakistan

• Previously in wide use, but not used very much today

( )
C2
Vo = C1 havg (1)

where Vo is the velocity (fps); and havg is the mean water depth (ft)

• The resulting velocity is supposed to be “just right”, so that neither erosion

nor sediment deposition will occur in the channel
• The coefficient (C1) and exponent (C2) can be adjusted for specific
conditions, preferably based on field measurements
• C1 is mostly a function of the characteristics of the earthen material in the
channel
• C2 is dependent on the silt load of the water
• Below are values for the coefficient and exponent of the Kennedy formula:

Table 1. Calibration values for the Kennedy formula.

C1 Material
0.56 extremely fine soil
0.84 fine, light sandy soil
0.92 coarse, light sandy soil
1.01 sandy, loamy silt
1.09 coarse silt or hard silt debris
Gary P. Merkley 192 BIE 5300/6300 Lectures
 C2 Sediment Load
0.64 water containing very fine silt
0.50 clear water

Kennedy Formula
(clear water: C2 = 0.50)
1.2
C1 = 0.56
C1 = 0.84
1.0
C1 = 0.92
C1 = 1.01
C1 = 1.09
0.8
Velocity, V o (m/s)

0.6

0.4

0.2

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Depth, db (m)

Figure 1. Velocity values versus water depth for the Kennedy formula with
clear water.

2. Lacey Method

• Developed by G. Lacey in the early part of the 20th century based on data from
India, Pakistan, Egypt and elsewhere
• Supports the “Lindley Regime Concept”, in which Lindley wrote:

“when an artificial channel is used to convey silty water, both bed and
banks scour or fill, changing depth, gradient and width, until a state of
balance is attained at which the channel is said to be in regime”

• There are four relationships in the Lacey method

• All four must be satisfied to achieve “regime” conditions

BIE 5300/6300 Lectures 193 Gary P. Merkley

 1. Velocity
V = 1.17 fR (2)
2. Wetted Perimeter
Wp = 2.67 Q (3)
3. Hydraulic Radius
R = 0.47 3 Q / f (4)
4. Bed Slope
f2/3
S = 0.000547 (5)
Q1/ 6
where,
f = 1.76 dm (6)

and, dm is the mean diameter of the bed and side slope materials (mm); V is the
mean velocity over the cross-section (fps); Wp is the wetted perimeter (ft); R is
the hydraulic radius (ft); S is the longitudinal bed slope (ft/ft); and Q is discharge
(cfs)

• The above relationships can be algebraically manipulated to derive other

dependent relationships that may be convenient for some applications
• For example, solve for S in terms of discharge
• Or, solve for dm as a function of R and V
• Here are two variations of the equations:

V = 0.00124 d11/
m
12
/S (7)
and,
V = 0.881 Q1/ 6d1/
m
12
(8)

• A weakness in the above method is that it considers particle size, dm, but
not cohesion & adhesion

Lacey General Slope Formula:

1.346 0.75
V= R S (9)
Na

where Na is a roughness factor, defined as:

Na = 0.0225f 0.25 ≅ 0.9nR0.083 (10)

where n is the Manning roughness factor

Gary P. Merkley 194 BIE 5300/6300 Lectures

 • This is for uniform flow conditions
• Applies to both regime and non-regime conditions
• Appears similar to the Manning equation, but according to Lacey it is more
representative of flow in alluvial channels

3. Maximum Velocity Method

• This method gives the maximum permissible mean velocity based on the type of
bed material and silt load of the water
• It is basically a compilation of field data, experience, and judgment
• Does not consider the depth of flow, which is generally regarded as an important
factor in determining velocity limits

Table 2. Maximum permissible velocities recommended by Fortier and Scobey

Velocity (fps)
Clear Water with
Material water colloidal silt
Fine sand, colloidal 1.5 2.5
Sandy loam, non-colloidal 1.75 2.5
Silt loam, non-colloidal 2 3
Alluvial silt, non-colloidal 2 3.5
Firm loam soil 2.5 3.5
Volcanic ash 2.5 3.5
Stiff clay, highly colloidal 3.75 5
Alluvial silt, colloidal 3.75 5
Shales and hard "pans" 6 6
Fine gravel 2.5 5
Coarse gravel 4 6
Cobble and shingle 5 5.5

Table 3. USBR data on permissible velocities for non-cohesive soils

Particle Mean velocity

Material diameter (mm) (fps)
Silt 0.005-0.05 0.49
Fine sand 0.05-0.25 0.66
Medium sand 0.25 0.98
Coarse sand 1.00-2.50 1.80
Fine gravel 2.50-5.00 2.13
Medium gravel 5.00 2.62
Coarse gravel 10.00-15.00 3.28
Fine pebbles 15.00-20.00 3.94
Medium pebbles 25.00 4.59
Coarse pebbles 40.00-75.00 5.91
Large pebbles 75.00-200.00 7.87-12.80

BIE 5300/6300 Lectures 195 Gary P. Merkley

IV. Introduction to the Tractive Force Method

• This method is to prevent scouring, not sediment deposition

• This is another design methodology for earthen channels, but it is not 100%
empirical, unlike the previously discussed methods
• It is most applicable to the design of earthen channels with erodible boundaries
(wetted perimeter) carrying clear water, and earthen channels in which the
material forming the boundaries is much coarser than the transported sediment
• The tractive force is that which is exerted on soil particles on the wetted
perimeter of an earthen channel by the water flowing in the channel
• The “tractive force” is actually a shear stress multiplied by an area upon which
the stress acts
• A component of the force of gravity on the side slope material is added to the
analysis, whereby gravity will tend to cause soil particles to roll or slide down
toward the channel invert (bed, or bottom)
• The design methodology treats the bed of the channel separately from the side
slopes
• The key criterion is whether the tractive + gravity forces are less than the “critical”
tractive force of the materials along the wetted perimeter of the channel
• If this is true, the channel should not experience scouring (erosion) from the flow
of water within
• Thus, the critical tractive force is the threshold value at which scouring would be
expected to begin
• This earthen canal design approach is for the prevention of scouring, but not for
the prevention of sediment deposition
• The design methodology is for trapezoidal or rectangular cross sections
• This methodology was developed by the USBR

V. Forces on Bed Particles

• The friction force (resisting particle movement) is:

Ws tanθ (11)

where θ is the angle of repose of the bed material and Ws is the weight of a soil
particle

Gary P. Merkley 196 BIE 5300/6300 Lectures

 • Use the angle of repose for wet (not dry) material
• θ will be larger for most wet materials
• Note that “tan θ” is the angle of repose represented as a slope

Angle of Repose for Non-Cohesive Earthen Material

42

40 Very angular
Moderately angular
38 Slightly angular
Slightly rounded
36
Angle of repose (degrees)

Moderately rounded
34 Very rounded

32

30

28

26

24

22

20
1 10 100
Soil particle size (mm)

Figure 2. Angle of repose (degrees from horizontal), θ, for non-cohesive

earthen materials (adapted from USBR Hyd Lab Report Hyd-366).

• The shear force on a bed particle is:

aTbed (12)

where “a” is the effective particle area and Tbed (lbs/ft2 or N/m2) is the shear
stress exerted on the particle by the flow of water in the channel

• When particle movement is impending on the channel bed, expressions 1 and 2

are equal, and:

Ws tan θ = aTbed (13)

or,

Ws tan θ
Tbed = (14)
a

BIE 5300/6300 Lectures 197 Gary P. Merkley

VI. Forces on Side-Slope Particles

• The component of gravity down the side slope is:

Ws sinφ (15)

where φ is the angle of the side slope, as defined in the figure below

Figure 3. Force components on a soil particle along the side slope of an

earthen channel.

• If the inverse side slope is m, then:

⎛ 1⎞
φ = tan−1 ⎜ ⎟ (16)
⎝m⎠

• The force on the side slope particles in the direction of water flow is:

aTside (17)

where Tside is the shear stress (lbs/ft2 or N/m2) exerted on the side slope particle
by the flow of water in the channel

• Note: multiply lbs/ft2 by 47.9 to convert to N/m2

• Combining Eqs. 15 & 17, the resultant force on the side slope particles is
downward and toward the direction of water flow, with the following magnitude:

Ws2 sin2 φ + a2Ts2ide (18)

• The resistance to particle movement on the side slopes is due to the orthogonal
component of Eq. 15, Wscosφ, as shown in the above figure, multiplied by the
coefficient of friction, tan θ
• Thus, when particle movement is impending on the side slopes:

Gary P. Merkley 198 BIE 5300/6300 Lectures

 Ws cos φ tan θ = Ws2 sin2 φ + a2Tside
2
(19)

• Solving Eq. 19 for Tside:

Ws
Tside = cos2 φ tan2 θ − sin2 φ (20)
a

• Applying trigonometric identities and simplifying:

Ws tan2 φ
Tside = cos φ tan θ 1 − (21)
a tan2 θ

or,

Ws sin2 φ
Tside = tan θ 1 − (22)
a sin2 θ

VII. Tractive Force Ratio

• As defined in Eq. 14, Tbed is the critical shear on bed particles

• As defined in Eqs. 20-22, Tside is the critical shear on side slope particles
• The tractive force ratio, K, is defined as:

Tside
K= (23)
Tbed

where Tside and Tbed are the critical (threshold) values defined in Eqs. 4 & 9-11

• Then:

sin2 φ tan2 φ
K = 1− = cos φ 1 − (24)
sin2 θ tan2 θ

VIII. Design Procedure

• The design procedure is based on calculations of maximum depth of flow, h

• Separate values are calculated for the channel bed and the side slopes,
respectively

BIE 5300/6300 Lectures 199 Gary P. Merkley

 • It is necessary to choose values for inverse side slope, m, and bed width, b to
calculate maximum allowable depth in this procedure
• Limits on side slope will be found according to the angle of repose and the
maximum allowable channel width
• Limits on bed width can be set by specifying allowable ranges on the ratio of b/h,
where b is the channel base width and h is the flow depth
• Thus, the procedure involves some trial and error

Step 0

• Specify the desired maximum discharge in the channel

• Identify the soil characteristics (particle size gradation, cohesion)
• Determine the angle of repose of the soil material, θ
• Determine the longitudinal bed slope, So, of the channel

Step 1

• Determine the critical shear stress, Tc (N/m2 or lbs/ft2), based on the type
of material and particle size from Fig. 3 or 4 (note: 47.90 N/m2 per lbs/ft2)
• Fig. 3 is for cohesive material; Fig. 4 is for non-cohesive material
• Limit φ according to θ (let φ ≤ θ)

Step 2

• Choose a value for b

• Choose a value for m

Step 3

• Calculate φ from Eq. 16

• Calculate K from Eq. 24
• Determine the max shear stress fraction (dimensionless), Kbed, for the
channel bed, based on the b/h ratio and Fig. 6
• Determine the max shear stress fraction (dimensionless), Kside, for the
channel side slopes, based on the b/h ratio and Fig. 7

Gary P. Merkley 200 BIE 5300/6300 Lectures

 Permissible Tc for Cohesive Material
100

Lean clay
Clay
Heavy clay
Sandy clay
T c (N/m2)

10

1
0.1 1.0 10.0
Void Ratio

Figure 4. Permissible value of critical shear stress, Tc, in N/m2, for cohesive
earthen material (adapted from USBR Hyd Lab Report Hyd-352).

About Figure 4: The “void ratio” is the ratio of volume of pores to volume of solids.
Note that it is greater than 1.0 when there is more void space than that occupied
by solids. The void ratio for soils is usually between 0.3 and 2.0.

BIE 5300/6300 Lectures 201 Gary P. Merkley

 Permissible Tc for Non-Cohesive Material
100

Clear water
Low content of fine sediment
High content of fine sediment
Coarse, non-cohesive material
Tc (N/m2)

10

size for which gradation gives 25%

of the material being larger in size

1
0.1 1.0 10.0 100.0

Average particle diameter (mm)

Figure 5. Permissible value of critical shear stress, Tc, in N/m2, for non-
cohesive earthen material (adapted from USBR Hyd Lab Report Hyd-352).

• The three curves at the left side of Fig. 5 are for the average particle diameter
• The straight line at the upper right of Fig. 5 is not for the “average particle
diameter,” but for the particle size at which 25% of the material is larger in size
• This implies that a gradation (sieve) analysis has been performed on the earthen
material

particle gradation

75% 25%

smallest largest

Gary P. Merkley 202 BIE 5300/6300 Lectures

 • The three curves at the left side of Fig. 5 (d ≤ 5 mm) can be approximated as
follows:

Clear water:

Tc = 0.0759d3 − 0.269 d2 + 0.947 d + 1.08 (25)

Low sediment:

Tc = 0.0756 d3 − 0.241d2 + 0.872d + 2.26 (26)

High sediment:

Tc = −0.0321d3 + 0.458 d2 + 0.190d + 3.83 (27)

where Tc is in N/m2; and d is in mm

• The portion of Fig 5. corresponding to “coarse material” (d > 5 mm) is

approximated as:

Coarse material:

Tc = 2.17d0.75 (28)

• Equations 25-28 are for diameter, d, in mm; and Tc in N/m2

• Equations 25-28 give Tc values within ±1% of the USBR-published data
• Note that Eq. 28 is exponential, which is required for a straight-line plot with log-
log scales

BIE 5300/6300 Lectures 203 Gary P. Merkley

Figure 6. Kbed values as a function of the b/h ratio.
Notes: This figure was made using data from USBR Hydraulic Lab Report Hyd-366. The
ordinate values are for maximum shear stress divided by γhSo, where γ = ρg, h is water depth,
and So is longitudinal bed slope. Both the ordinate & abscissa values are dimensionless.

Gary P. Merkley 204 BIE 5300/6300 Lectures

Figure 7. Kside values as a function of the b/h ratio.
Notes: This figure was made using data from USBR Hydraulic Lab Report Hyd-366. The
ordinate values are for maximum shear stress divided by γhSo, where γ = ρg, h is water depth,
and So is longitudinal bed slope. Both the ordinate and abscissa values are dimensionless.

BIE 5300/6300 Lectures 205 Gary P. Merkley

 • Regression analysis can be performed on the plotted data for Kbed & Kside
• This is useful to allow interpolations that can be programmed, instead of reading
values off the curves by eye
• The following regression results give sufficient accuracy for the max shear stress
fractions:

0.153
⎛b⎞
Kbed ≅ 0.792 ⎜ ⎟ for 1 ≤ b / h ≤ 4
⎝h⎠ (29)
⎛b⎞
Kbed ≅ 0.00543 ⎜ ⎟ + 0.947 for 4 ≤ b / h ≤ 10
⎝h⎠

for trapezoidal cross sections; and,

AB + C (b / h)
D

K side ≅ (30)
B + (b / h)
D

where,
A = −0.0592 (m ) + 0.347 (m ) + 0.193
2
(31)

−0.000311(m)
7.23
B = 2.30 − 1.56e (32)

−0.00143 (m )
5.63
C = 1.14 − 0.395e (33)

−3.29
−35.2 (m )
D = 1.58 − 3.06e (34)

for 1 ≤ m ≤ 3, and where e is the base of natural logarithms

• Equations 29 give Kbed to within ±1% of the values from the USBR data for 1 ≤
b/h ≤ 10
• Equations 30-34 give Kside to within ±2% of the values from the USBR data for 1
≤ m ≤ 3 (where the graphed values for m = 3 are extrapolated from the lower m
values)

• The figure below is adapted from the USBR, defining the inverse side slope, and
bed width
• The figure below also indicates locations of measured maximum tractive force on
the side slopes, Kside, and the bed, Kbed
• These latter two are proportional to the ordinate values of the above two graphs
(Figs. 6 & 7)

Gary P. Merkley 206 BIE 5300/6300 Lectures

 Step 4

• Calculate the maximum depth based on Kbed:

KTc
hmax = (35)
KbedγSo

• Recall that K is a function of φ and θ (Eq. 13)

• Calculate the maximum depth based on Kside:

KTc
hmax = (36)
K side γSo

where γ is 62.4 lbs/ft3, or 9,810 N/m3

• Note that K, Kbed, Kside, and So are all dimensionless; and Tc/γ gives units
of length (ft or m), which is what is expected for h
• The smaller of the two hmax values from the above equations is applied to
the design (i.e. the “worst case” scenario)

Step 5

• Take the smaller of the two depth, h, values from Eqs. 35 & 36
• Use the Manning or Chezy equations to calculate the flow rate
• If the flow rate is sufficiently close to the desired maximum discharge
value, the design process is finished
• If the flow rate is not the desired value, change the side slope, m, and or
bed width, b, checking the m and b/h limits you may have set initially
• Return to Step 3 and repeat calculations
• There are other ways to attack the problem, but it’s almost always iterative

BIE 5300/6300 Lectures 207 Gary P. Merkley

 • For a “very wide” earthen channel, the channel sides become negligible and the
critical tractive force on the channel bed can be taken as:

Tc ≅ γhSo (37)

• Then, if So is known, h can be calculated

IX. Definition of Symbols

a effective particle area (m2 or ft2)

b channel base width (m or ft)
h depth of water (m or ft)
hmax maximum depth of water (m or ft)
K tractive force ratio (function of φ and θ)
Kbed maximum shear stress fraction (bed)
Kside maximum shear stress fraction (side slopes)
m inverse side slope
So longitudinal bed slope
Tbed shear stress exerted on a bed soil particle (N/m2 or lbs/ft2)
Tc critical shear stress (N/m2 or lbs/ft2)
Tside shear stress exerted on a side slope soil particle (N/m2 or lbs/ft2)
Ws weight of a soil particle (N or lbs)
φ inverse side slope angle
γ weight of water per unit volume (N/m3 or lbs/ft3)
θ angle of repose (wet soil material)

References & Bibliography

Carter, A.C. 1953. Critical tractive forces on channel side slopes. Hydraulic Laboratory Report No.
HYD-366. U.S. Bureau of Reclamation, Denver, CO.
Chow, V.T. 1959. Open-channel hydraulics. McGraw-Hill Book Co., Inc., New York, NY.
Davis, C.S. 1969. Handbook of applied hydraulics (2nd ed.). McGraw-Hill Book Co., Inc., New York,
NY.
Labye, Y., M.A. Olson, A. Galand, and N. Tsiourtis. 1988. Design and optimization of irrigation
distribution networks. FAO Irrigation and Drainage Paper 44, United Nations, Rome, Italy. 247
pp.
Lane, E.W. 1950. Critical tractive forces on channel side slopes. Hydraulic Laboratory Report No.
HYD-295. U.S. Bureau of Reclamation, Denver, CO.
Lane, E.W. 1952. Progress report on results of studies on design of stable channels. Hydraulic
Laboratory Report No. HYD-352. U.S. Bureau of Reclamation, Denver, CO.
Smerdon, E.T. and R.P. Beasley. 1961. Critical tractive forces in cohesive soils. J. of Agric. Engrg.,
American Soc. of Agric. Engineers, pp. 26-29.

Gary P. Merkley 208 BIE 5300/6300 Lectures

Lecture 18
Sample Earthen Channel Designs

I. Example “A”: Design Procedure for an Earthen Canal

• Design an earthen canal section in an alluvial soil such that the wetted
boundaries do not become eroded
• The canal will follow the natural terrain at an estimated So = 0.000275 ft/ft with a
preliminary design side slope of 1.5:1.0 (h:v)
• The bed material has been determined to be a non-cohesive “coarse light sand”
with an average particle diameter of 10 mm, 25% of which is larger than 15 mm
• Thus, 15 mm will be used to determine Tc in Fig. 5
• Tests have shown that the angle of repose for the bed material is approximately
34°, measured from the horizontal
• For the Manning equation, use a roughness value of 0.030

• The design discharge (Qmax) is 650 cfs

• The source of water is such that there will be a low content of fine sediment
• The canal can be assumed to be straight, even though there will be bends at
several locations

• Design the section using a trapezoidal shape with a bed width to depth ratio, b/h,
of between 1.0 and 5.0
• The design should also be such that the canal bed and sides do not erode
• Adjust the side slope if necessary, but keep it within the range 0.5:1 to 2.0:1 (h:v)
• Note that φ < θ must be true to allow for a stable side slope

1. Design the canal using the tractive force method

2. Compare the results for the case in which it is assumed that the channel is
very wide (i.e. critical tractive “force” = γhSo)
3. Compare with results from the Kennedy formula
4. Compare with results from the Lacey method
5. Compare with results from the maximum velocity method using Table 2, and
again using Table 3

Solution to Example “A” Design Problem:

1. Tractive Force Method

Critical Tractive Force

• The critical tractive force can be estimated from Figure 5 (see above)
• The material is non-cohesive, and 25% of the particles are larger than 15 mm
• This gives Tc ≈ 16.3 N/m2 (0.34 lbs/ft2) for the 15 mm abscissa value

BIE 5300/6300 Lectures 209 Gary P. Merkley

 Angle of Repose

• The angle of repose, θ, is given as 34°

• Then, the ratio of Tside to Tbed is:

T sin2 φ
K = side = 1− 2
= 1 − 3.2sin2 φ (1)
Tbed sin θ

• Design requirements for this example call for a side slope between 0.5 & 2.0
• Actually, the range is restricted to 1.5 to 2.0 because the preliminary design
side slope of 1.5:1 corresponds to an angle φ = 33.7°
• This is less than the angle of repose, θ = 34°, but it is very close
• Make a table of K values:

Table 1. K values for different side slopes

m φ K
1.5 33.7° 0.122
1.6 32.0° 0.318
1.7 30.5° 0.419
1.8 29.1° 0.493
1.9 27.8° 0.551
2.0 26.6° 0.599

Maximum Shear Stress Fractions

• From Figure 7 (see above), the maximum shear stress fraction for sides, Kside,
is approximately equal to 0.74 in the range 1.0 < (b/h) < 5.0, and for side
slopes from 1.5 to 2.0
• Take Kside as a constant for this problem: Kside ≈ 0.74
• The maximum shear stress fraction on the channel bed, Kbed, will fall on the
curve for trapezoidal sections, and will vary from 0.79 to 0.97 within the
acceptable range1.0 < (b/h) < 5.0
• Make a table of Kbed values according to b/h ratio (from Figure 6):

Table 2. Values of Kbed for different b/h ratios

b/h Kbed
1 0.79
2 0.90
3 0.94
4 0.96
5 0.97

Gary P. Merkley 210 BIE 5300/6300 Lectures

 Manning Equation

• The Manning roughness, n, is given as 0.030. The longitudinal bed slope is

given as 0.000275 ft/ft
• The side slope can be any value between 1.5 and 2.0
• Construct a table of depths (normal depths) for the maximum design
discharge of 650 cfs, then make another table showing bed width to depth
ratios:
Table 3. Flow depths (ft) for 650 cfs

inverse side slope, m

b (ft)
1.5 1.6 1.7 1.8 1.9 2.0
10 10.11 9.92 9.74 9.58 9.43 9.29
15 9.01 8.87 8.74 8.62 8.51 8.41
20 8.10 8.00 7.91 7.83 7.75 7.67
25 7.36 7.29 7.23 7.16 7.10 7.04
30 6.75 6.70 6.65 6.60 6.56 6.52
35 6.25 6.21 6.17 6.14 6.10 6.07

Table 4. Bed width to depth ratios for 650 cfs

inverse side slope, m

b (ft)
1.5 1.6 1.7 1.8 1.9 2.0
10 0.989 1.008 1.027 1.044 1.060 1.076
15 1.665 1.691 1.716 1.740 1.763 1.784
20 2.469 2.743 2.766 2.554 2.581 2.608
25 3.397 3.429 3.458 3.492 3.521 3.551
30 4.444 4.478 4.511 4.545 4.573 4.601
35 5.600 5.636 5.672 5.700 5.738 5.766
Note: bold values fall outside the acceptable range of 1.0 < (b/h) < 5.0.

• In Table 4 it is seen that the bed width must be less than 35 ft, otherwise the
required b/h ratio will be greater than 5.0
• Also, it can be seen that bed widths less than 10 ft will have problems
because the b = 10 and m = 1.5 combination gives b/h < 1.0

Allowable Depth for Tractive Force: Side Slopes

• Taking Kside from Figure 7, the maximum allowable depth according to the
tractive force method for side slopes is:

TcK 0.34 1 − 3.2 sin2 φ

hmax = = = 26.8 1 − 3.2 sin2 φ (2)
K side γSo ( 0.74 )( 62.4 )( 0.000275 )

where the unit weight of water is taken as γ = 62.4 lbs/ft3

BIE 5300/6300 Lectures 211 Gary P. Merkley

 • Make a table of hmax values (essentially independent of b/h) for different
values of the angle φ:

Table 5. Maximum depth values for different side slopes

m φ hmax
1.5 33.7° 3.3
1.6 32.0° 8.5
1.7 30.5° 11.2
1.8 29.1° 13.2
1.9 27.8° 14.8
2.0 26.6° 16.0

• Now the design possibilities will narrow further

• Compare depths calculated by the Manning equation for 650 cfs with the
maximum allowable depths by tractive force method, according to side slope
traction (combine Tables 3 & 5):

Table 6. Ratio of Manning depths to hmax for 650 cfs

inverse side slope, m

b (ft)
1.5 1.6 1.7 1.8 1.9 2.0
10 3.06 1.17 0.87 0.73 0.64 0.58
15 2.73 1.04 0.78 0.65 0.58 0.53
20 2.45 0.94 0.71 0.59 0.52 0.48
25 2.23 0.86 0.65 0.54 0.48 0.44
30 2.05 0.79 0.59 0.50 0.44 0.41
Note: bold values are out of range (uniform flow depths are too high).

• From the above table it is seen that the side slope must now be between 1.6
and 2.0, otherwise the required flow depths will exceed the limit imposed by
the tractive force method for side slopes

Allowable Depth for Tractive Force: Channel Bed

Again, taking Kbed from Figure 6, the maximum allowable depth according to the
tractive force method for the channel bed is:

KTc 0.34 1 − 3.2sin2 φ 19.8

hmax = = = 1 − 3.2sin2 φ (3)
Kbed γSo Kbed ( 62.4 )( 0.000275 ) Kbed

where Kbed is a function of the b/h ratio

Gary P. Merkley 212 BIE 5300/6300 Lectures

 Table 7. Maximum allowable depths (ft) according to bed criterion

side slope, m
b (ft)
1.5 1.6 1.7 1.8 1.9 2.0
10 n/a 8.00 10.50 12.40 13.80 15.00
15 n/a 7.20 9.70 11.20 12.50 13.60
20 n/a 6.80 8.90 10.60 11.90 12.90
25 n/a 6.60 8.70 10.30 11.50 12.50
30 n/a 6.50 8.60 10.10 11.30 12.30
Note: bold depth values are not acceptable.

• Comparing with Table 3 from Manning equation, most of the depths

required for 650 cfs at m = 1.6 are higher than allowed by tractive force
method (bed criterion)
• However, the combination of m= 1.6 and b = 30 falls within acceptable
limits

Final Tractive Force Design

• The allowable depths according to the bed criterion are all less than the
allowable depths (for the same m values) from the side slope criterion
• Therefore, use the bed criterion as the basis for the design
• The permissible values for m = 2.0 are all much higher than those from the
Manning equation
• Permissible values for m = 1.9, 1.8, and 1.7 are also higher than those
from the Manning equation
• For m = 1.6, only the b = 30 ft bed width is within limits (less than that
required by the Manning equation for 650 cfs)

• Make a judgment decision based on economics, convenience of

construction, area occupied by the channel (channel width), safety
considerations, and other factors

• Reject the 30-ft bed width; it will be wider than necessary

• Due to lack of other information, recommend the 20-ft bed width and 1.7
side slope option
• At this point it would not be useful to consider other intermediate values of
b and m
• For b = 20 ft and m = 1.7, the depth will be about 7.91 ft (allowable is 8.9 ft
from Table 4), and the mean flow velocity at 650 cfs is: V = 2.5 fps

2. Tractive Force Method: Assume “very wide channel”

In this case, K is equal to 1.0, and,

BIE 5300/6300 Lectures 213 Gary P. Merkley

 Tc 0.34
hmax = = = 19.8 ft (4)
γS (62.4)(0.000275)

• Thus, the depth would have to be less than 19.8 ft

• This is a much less conservative value than that obtained above (OK)

3. Kennedy Formula

• For a “coarse light sand”, C = 0.92

• For fine sediment in water, m = 0.64. Then,

0.64
Vo = CDm = 0.92 ( 7.91) = 3.5 fps (5)

• This is more than the tractive force design velocity of 2.5 fps. (OK)

4. Lacey Method

• Mean diameter of bed material, dm = 10 mm

• The hydraulic radius at 650 cfs from the tractive force design is R =
A/Wp = 265/51 = 5.2 ft
• Then,
f = 1.76 dm = 5.57 (6)

V = 1.17 fR = 1.17 ( 5.57 )( 5.2) = 6.3 fps (7)

• This is also more than the tractive force design velocity of 2.5 fps. (OK)

5. Maximum Velocity Method

• From Table 1, the maximum permissible velocity for “coarse light sand”, say
“fine gravel”, is 5.0 fps for water transporting colloidal silt
• For the same material and clear water, the maximum is 2.5 fps -- a large
difference based on a fairly subjective determination
• Also, from the given information of this example problem, an exact match the
materials listed in Table 1 is not possible
• From Table 2, for “coarse sand”, the maximum permissible velocity is 1.8 fps

Gary P. Merkley 214 BIE 5300/6300 Lectures

II. Example “B”: Design Procedure for an Earthen Canal

• Design an earthen canal section using the tractive force method such that the
bed and side slopes are stable
• The design flow rate is 90 m3/s and the water is clear
• The earthen material is non-cohesive fine sand with average particle size of 0.5
mm and an angle of repose of 27°
• Assume that the inverse side slope is fixed at m = 3.0 for this design. Use a
Manning’s n of 0.02
• Determine the minimum bed width, b
• Determine the maximum longitudinal bed slope, So
• Recommend a freeboard value for the design discharge
• Make a sketch to scale of the channel cross section and the water surface at the
design discharge

Solution to Example “B” Design Problem:

1. Convert to English units

• 90 m3/s = 3,178 cfs

2. Check angle of repose

• θ = 27°, or 0.471 rad (angle of repose)

• φ = tan-1 (1/m) = tan-1 (1/3.0) = 0.322 rad (18.4°)
• φ < θ, so the side slopes are potentially stable

3. Tractive force ratio, K

sin2 φ sin2 (0.322)

K = 1− = 1− = 0.718
sin2 θ sin2 (0.471)

4. Critical shear stress, Tc

• From Figure 5, for non-cohesive material, with an average particle size of 0.5
mm and clear water, Tc ≈ 0.03 lb/ft2

5. Max shear stress fractions

• Arbitrarily limit b/h to a minimum of 1 and maximum of 9 (note that b/h < 1 is
usually not reasonable or feasible for an earthen channel)
• This is a very wide range of b/h values anyway
• The table below shows Kbed and Kside for 1 ≤ b/h ≤ 9
• The Kside values are extrapolated from the curves for 1 ≤ m ≤ 2 because in our
case, m = 3
BIE 5300/6300 Lectures 215 Gary P. Merkley
 b/h Kbed Kside
1 0.79 0.78
2 0.90 0.81
3 0.94 0.83
4 0.97 0.83
5 0.98 0.83
6 0.98 0.84
7 0.99 0.84
8 0.99 0.84
9 1.00 0.84

6. Uniform flow depths

• Calculate uniform flow depths for various b and So values, using n = 0.02 and
the Manning equation
• Note that b will have to be fairly large because it is an earthen channel and
the design discharge is rather large itself
• The uniform flow calculations are done in a computer program and the results
are given in the table below

Uniform flow depths (in ft) for varying b & So (Manning; n = 0.02)
Longitudinal bed slope, So
b (ft)
0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 23.750 20.379 18.609 17.437 16.573 15.894 15.340 14.874 14.473 14.122
40 22.479 19.161 17.426 16.278 15.434 14.773 14.233 13.780 13.390 13.050
50 21.317 18.063 16.367 15.250 14.429 13.787 13.264 12.825 12.448 12.119
60 20.256 17.074 15.423 14.337 13.541 12.920 12.415 11.991 11.628 11.312
70 19.289 16.184 14.579 13.527 12.757 12.158 11.671 11.263 10.913 10.609
80 18.406 15.382 13.825 12.806 12.063 11.485 11.016 10.623 10.287 9.995
90 17.602 14.659 13.149 12.164 11.446 10.889 10.437 10.059 9.737 9.456
100 16.867 14.005 12.542 11.589 10.896 10.359 9.923 9.560 9.250 8.980
110 16.194 13.413 11.994 11.073 10.404 9.885 9.465 9.115 8.816 8.557
120 15.578 12.874 11.499 10.607 9.960 9.459 9.054 8.717 8.429 8.179

7. Base width to depth ratio, b/h

• Calculate b/h values for each of the uniform-flow depths in the above table
• Values in the following table are shown in bold where greater than the
specified maximum of 9

Gary P. Merkley 216 BIE 5300/6300 Lectures

Bed width to depth ratios (b/h) for varying b & So
Longitudinal bed slope, So
b (ft) 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 1.263 1.472 1.612 1.720 1.810 1.888 1.956 2.017 2.073 2.124
40 1.779 2.088 2.295 2.457 2.592 2.708 2.810 2.903 2.987 3.065
50 2.346 2.768 3.055 3.279 3.465 3.627 3.770 3.899 4.017 4.126
60 2.962 3.514 3.890 4.185 4.431 4.644 4.833 5.004 5.160 5.304
70 3.629 4.325 4.801 5.175 5.487 5.758 5.998 6.215 6.414 6.598
80 4.346 5.201 5.787 6.247 6.632 6.966 7.262 7.531 7.777 8.004
90 5.113 6.140 6.845 7.399 7.863 8.265 8.623 8.947 9.243 9.518
100 5.929 7.140 7.973 8.629 9.178 9.653 10.078 10.460 10.811 11.136
110 6.793 8.201 9.171 9.934 10.573 11.128 11.622 12.068 12.477 12.855
120 7.703 9.321 10.436 11.313 12.048 12.686 13.254 13.766 14.237 14.672

• The following two tables are interpolated Kbed and Kside values
• Values shown in bold (following two tables) are greater than 1.0 and
considered to be infeasible

Interpolated Kbed values

Longitudinal bed slope, So
b (ft) 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 0.821 0.840 0.852 0.861 0.867 0.873 0.878 0.882 0.885 0.889
40 0.865 0.886 0.899 0.909 0.916 0.922 0.928 0.932 0.936 0.940
50 0.902 0.926 0.940 0.950 0.958 0.965 0.970 0.975 0.969 0.969
60 0.935 0.960 0.975 0.970 0.971 0.972 0.973 0.974 0.975 0.976
70 0.965 0.970 0.973 0.975 0.977 0.978 0.980 0.981 0.982 0.983
80 0.971 0.975 0.978 0.981 0.983 0.985 0.986 0.988 0.989 0.990
90 0.975 0.980 0.984 0.987 0.990 0.992 0.994 0.996 0.997 0.999
100 0.979 0.986 0.990 0.994 0.997 0.999 1.002 1.004 1.006 1.007
110 0.984 0.992 0.997 1.001 1.004 1.007 1.010 1.013 1.015 1.017
120 0.989 0.998 1.004 1.008 1.012 1.016 1.019 1.022 1.024 1.027

Interpolated Kside values

Longitudinal bed slope, So
b (ft) 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 0.775 0.775 0.776 0.776 0.777 0.777 0.777 0.777 0.777 0.778
40 0.777 0.777 0.778 0.778 0.779 0.779 0.779 0.779 0.780 0.780
50 0.778 0.779 0.780 0.780 0.780 0.781 0.781 0.781 0.828 0.828
60 0.779 0.780 0.781 0.828 0.828 0.828 0.828 0.828 0.828 0.828
70 0.781 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828
80 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828
90 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828 0.828
100 0.828 0.828 0.828 0.828 0.828 0.828 0.829 0.829 0.829 0.829
110 0.828 0.828 0.828 0.829 0.829 0.829 0.829 0.829 0.829 0.829
120 0.828 0.828 0.829 0.829 0.829 0.829 0.829 0.829 0.829 0.829

8. Ratio of max depth to uniform flow depth

• Using the interpolated max shear stress fractions, apply the following
equations to calculate maximum water depth:

BIE 5300/6300 Lectures 217 Gary P. Merkley

 KTc
hmax =
KbedγSo

and,
KTc
hmax =
K side γSo

where the smaller of the two values is taken for the design

• Bold values in the following two tables have a uniform flow depth which
exceeds the calculated maximum depth, and are removed from consideration

Ratio of max depth (based on Kbed) to uniform-flow depth

Longitudinal bed slope, So
b (ft) 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 1.771 1.008 0.726 0.575 0.480 0.415 0.366 0.329 0.299 0.275
40 1.775 1.016 0.734 0.583 0.488 0.422 0.373 0.336 0.306 0.281
50 1.795 1.032 0.748 0.596 0.500 0.433 0.383 0.345 0.318 0.294
60 1.822 1.053 0.765 0.621 0.525 0.458 0.408 0.369 0.338 0.313
70 1.855 1.099 0.811 0.654 0.554 0.484 0.431 0.391 0.358 0.331
80 1.932 1.151 0.851 0.687 0.582 0.509 0.454 0.411 0.377 0.349
90 2.012 1.201 0.889 0.719 0.609 0.533 0.475 0.431 0.395 0.366
100 2.090 1.250 0.926 0.749 0.636 0.556 0.496 0.450 0.412 0.382
110 2.167 1.298 0.962 0.779 0.661 0.578 0.516 0.468 0.429 0.397
120 2.241 1.344 0.997 0.807 0.685 0.599 0.535 0.484 0.444 0.411

Ratio of max depth (based on Kside) to uniform-flow depth

Longitudinal bed slope, So
b (ft) 0.00001 0.00002 0.00003 0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.00010
30 1.876 1.092 0.797 0.637 0.536 0.466 0.414 0.373 0.341 0.314
40 1.977 1.159 0.849 0.681 0.574 0.500 0.445 0.402 0.367 0.339
50 2.081 1.226 0.902 0.725 0.613 0.535 0.476 0.431 0.372 0.344
60 2.186 1.295 0.955 0.727 0.615 0.537 0.479 0.434 0.398 0.368
70 2.292 1.287 0.953 0.770 0.653 0.571 0.510 0.462 0.424 0.393
80 2.264 1.354 1.005 0.813 0.691 0.605 0.540 0.490 0.450 0.417
90 2.367 1.421 1.056 0.856 0.728 0.638 0.570 0.518 0.475 0.441
100 2.470 1.488 1.107 0.899 0.765 0.670 0.600 0.545 0.500 0.464
110 2.573 1.553 1.158 0.941 0.801 0.702 0.629 0.571 0.525 0.487
120 2.675 1.618 1.208 0.982 0.837 0.734 0.657 0.597 0.549 0.509

• It is clear that the bed slope must be less than So = 0.00003 ft/ft
• This is a very small slope
• But there is a large range of possible bed widths
• The lower b values will result in great depths, and the higher b values will take
up a wide “swath” of land. Most feasible will probably be a compromise
between these extremes, perhaps 60 < b < 90 ft.
• To complete one design possibility, recommend b = 80 ft & So = 0.00002 ft/ft
• This corresponds to a uniform flow depth of h = 15.4 ft (see above)

Gary P. Merkley 218 BIE 5300/6300 Lectures

 9. Freeboard

• Using the freeboard curves from the previous lecture, with Q = 3,178 cfs, the
height of the bank above the water surface should be about 4.5 ft
• Then, the depth of the channel is 15.4 + 4.5 = 19.9 ft. Round up to 20 ft.

10. Cross-section sketch

20.0 ft 15.4 ft 1
3
80.0 ft
References & Bibliography

Carter, A.C. 1953. Critical tractive forces on channel side slopes. Hydraulic Laboratory Report No.
HYD-366. U.S. Bureau of Reclamation, Denver, CO.
Chow, V.T. 1959. Open-channel hydraulics. McGraw-Hill Book Co., Inc., New York, NY.
Davis, C.S. 1969. Handbook of applied hydraulics (2nd ed.). McGraw-Hill Book Co., Inc., New York,
NY.
Labye, Y., M.A. Olson, A. Galand, and N. Tsiourtis. 1988. Design and optimization of irrigation
distribution networks. FAO Irrigation and Drainage Paper 44, United Nations, Rome, Italy. 247
pp.
Lane, E.W. 1950. Critical tractive forces on channel side slopes. Hydraulic Laboratory Report No.
HYD-295. U.S. Bureau of Reclamation, Denver, CO.
Lane, E.W. 1952. Progress report on results of studies on design of stable channels. Hydraulic
Laboratory Report No. HYD-352. U.S. Bureau of Reclamation, Denver, CO.
Smerdon, E.T. and R.P. Beasley. 1961. Critical tractive forces in cohesive soils. J. of Agric. Engrg.,
American Soc. of Agric. Engineers, pp. 26-29.

BIE 5300/6300 Lectures 219 Gary P. Merkley

Gary P. Merkley 220 BIE 5300/6300 Lectures
Lecture 19
Canal Linings

I. Reasons for Canal Lining Installing plastic canal

lining (courtesy R.W. Hill)
1. To save water (reduce
seepage)
2. To stabilize channel bed
and banks (reduce
erosion)
3. To avoid piping through
and under channel
banks
4. To decrease hydraulic
roughness (flow
resistance)
5. To promote movement, rather than deposition, of sediments
6. To avoid waterlogging of adjacent land
7. To control weed growth
8. To decrease maintenance costs and facilitate cleaning
9. To reduce excavation costs (when extant material is unsuitable)
10. To reduce movement of contaminated groundwater plumes

• The most common and (usually) most important reason is to reduce seepage
losses (and this may be for a variety of reasons)
• The assumption that lining will solve seepage problems is often unfounded,
simply because poor maintenance practices (especially with concrete linings) will
allow cracking and panel failures, and tears and punctures in flexible membranes
• Seepage losses from canals can be beneficial in that it helps recharge aquifers
and makes water accessible to possibly larger areas through groundwater
pumping. The extent of aquifers is more continuous than that of canals and
canal turnouts. But, pumping ($energy$) is usually necessary with groundwater,
unless perhaps you are downhill and there is an artesian condition (this is the
case in some places).
• “Administrative losses” and over-deliveries can add up to a greater volume of
water than seepage in many cases (that means that canal lining is not always the
most promising approach to saving water in the distribution system)

• Sometimes, only the bottom of a canal is lined when most of the seepage has
been found to be in the vertical direction
• It may be advisable to perform soil compaction testing under concrete linings to
determine if steps need to be taken to avoid subsequent settlement of the canal
• Lining to decrease maintenance costs can backfire (costs may actually increase)
• Concrete pipe is an alternative to lined canals, but for large capacities the pipes
tend to cost more
• Many billions of dollars have been spent world-wide during the past several
decades to line thousands of miles of canals
BIE 5300/6300 Lectures 221 Gary P. Merkley
II. Some Types of Lining and Costs

Type Typical Costs

1. Soil

• Lime
• Bentonite clay
• “High-swell” Bentonite & coarse clay or other “bridging material”
• Geosynthetic clay liner (“Bentomat”)
• Soil mixed with portland cement
• Thin compacted earth (6 - 12 inches)
• Thick compacted earth (12 - 36 inches)

2. Fly Ash ............................................................................................... $3.00/yd2

3. Masonry (stone, rock, brick)
4. Concrete (portland cement)

• Nonreinforced concrete ........................................................... $5.00/yd2

• Reinforced concrete (with steel)
• Gunite, a.k.a. shotcrete, a.k.a. cement mortar (hand
or pneumatically applied; w/o steel reinforcement)................ $12.00/yd2
• Gunite, a.k.a. shotcrete, a.k.a. cement mortar (hand
or pneumatically applied; w/ steel reinforcement).................. $15.00/yd2

5. Plastic

• Polyvinyl Chloride (PVC) ......................................................... $5.00/yd2

• Oil Resistant PVC
• Chlorinated Polyethylene (PE)
• Low Density Polyethylene........................................................ $4.00/yd2
• High Density Polyethylene..................................................... $10.00/yd2
• Polyurethane foam with or without coatings

6. Asphalt (bituminous)

• Sprayed (“blown”) asphalt

• Asphaltic Concrete .................................................................. $4.00/yd2

7. Synthetic Rubber

• Butyl Rubber............................................................................ $8.00/yd2

• Neoprene Rubber
• Shotcrete over geosynthetic .................................................. $37.00/yd2
• Concrete over geosynthetic ................................................... $26.00/yd2

Gary P. Merkley 222 BIE 5300/6300 Lectures

III. Comments on Different Lining Materials

• The USBR has had a Pneumatic application of shotcrete

long-standing research
program on canal lining
materials and
installation techniques
(began in 1946, but
essentially discontinued
in recent years)
• There are many
publications with
laboratory and field
data, design guidelines
and standards, and
other relevant
information (but you have to dig it all up because it doesn’t come in one book)
• Many technical articles can be found in the journals on canal lining materials,
construction methods, and experience with different types of linings

Earthen Linings

• Earthen linings usually require significant over-excavation, and transport of

suitable material (in large volumes) from another site
• Many earthen linings are 2 - 3 ft thick; “thin” linings are 6 - 12 inches thick
• Clay linings can crack after only a few cycles of wetting and drying, causing
increased seepage loss. Bentonite clay swells considerably when wet, but
cracks may not completely seal after the canal has been dried, then filled with
water again.
• Bentonite is a special kind of clay, usually made of up decomposed volcanic ash,
and containing a high percentage of colloidal particles (less than 0.0001 cm in
diameter)
• High-swell Bentonite may swell 8 to 12 times in volume when wetted; other types
may swell less than 8 times in volume
• Bentonite disperses well when mixed with soft water, but may flocculate (clump
up) when mixed with hard water. Flocculation can be avoided by adding one or
more dispersing agents (e.g. tetrasodium pyrophosphate, sodium
tripolyphosphate, sodium hexametaphosphate). Low-swell Bentonite tends to
flocculate easier.
• Repeated drying-wetting cycles can cause loss of lining density, loss of stability,
and progressive deterioration of the lining
• Other than Bentonite, clay linings may be of montmorillonite, or montmorillonite -
chlorite
• Some clay linings have been treated with lime to stabilize the material. The
addition of lime to expansive soils (e.g. Bentonite) improves workability and
increases structural strength

BIE 5300/6300 Lectures 223 Gary P. Merkley

 Portland Concrete

• Small concrete-lined canals are

usually non-reinforced Rubber strip at panel joint
• Steel reinforcement (rebar or steel
mesh) is also not commonly used on
large canals anymore unless there
are compelling structural reasons
• The elimination of steel reinforcement
from concrete canal linings saves
about 10 to 15% of the total cost
(USBR 1963)
• During the past several years it has
become popular to install concrete
linings in small canals at the same
time as final excavation and finishing, often using a laser to control the alignment
and longitudinal slope
• Some “underwater” concrete lining operations have been performed in recent
years on full canals (so as not to disrupt delivery operations)
• Careful shaping, or finishing, of the native soil is an important step in the
preparation for concrete lining simply because it can greatly reduce the required
volume of concrete (significantly lowering the cost)

Manual concrete lining of a canal reach (every other panel is

poured first to facilitate formwork)

Gary P. Merkley 224 BIE 5300/6300 Lectures

 • Reinforced concrete can contain rebar and or wire mesh. Reinforcement is
usually for structural reasons, but also to control cracking of the lining
• Concrete panel joints may have rubber strips to prevent seepage
• Weep holes or flap valves are often installed in cut sections of a concrete-lined
canal to relieve back pressures which can cause failure of the lining
• Flap valves may be installed both in side slopes and in the canal bed
• Some concrete-lined canals have (measured) high seepage loss rates,
particularly in “fill” sections of canal, and in soils with high permeability (usually
sandy soils) -- but, seepage rates are rarely measured; they are “assumed”
based on tables in books
• British researchers report that their investigations show that if 0.01% of the area
of a concrete canal lining is cracked (0.01% are cracks), the average seepage
rate may be the same as that of an unlined canal
• Soil mixed with Portland cement, especially sandy soil, can be an acceptable
cost-saving approach to canal lining

IV. Concrete Lining Thickness

• Lining thickness is often chosen in a somewhat arbitrary manner, but based on

experience and judgment, and based on the performance of existing linings on
other canals
• Thinner linings may crack, but this does not have to be a problem if the cracks
are sealed during routine maintenance (not all concrete-lined canals enjoy
routine maintenance)
• Concrete lined channels often have high seepage loss rates due to cracks and
unsealed panel joints
• Grooves are often specified to control the location and extent of cracking, which
can be expected even under the best conditions
• The selection of lining thickness is an economic balance between cost and
durability (canals perceived to be very important will have more conservative
designs -- municipal supplies, for example)
• The USBR has suggested the following guidelines:

Lining Type Thickness (inch) Discharge (cfs)

Unreinforced concrete 2.00 0-200
2.50 200-500
3.00 500-1,500
3.50 1,500-3,500
4.00 > 3,500
Asphaltic concrete 2.00 0-200
3.25 200-1,500
4.00 > 1,500
Reinforced concrete 3.50 0-500
4.00 500-2,000
4.50 > 2,000

BIE 5300/6300 Lectures 225 Gary P. Merkley

 Lining Type Thickness (inch) Discharge (cfs)
Gunite (shotcrete) 1.25 0-100
1.50 100-200
1.75 200-400
2.00 > 400

Plastic and Rubber

• Plastic linings are also referred to as “geomembranes” or “flexible membrane”

linings
• Plastic canal linings have been in use for approximately 40 years
• Plastic and rubber linings are covered with soil, soil and rock, bricks, concrete, or
other material for

1. protection
• ozone “attack” and UV radiation
• puncture due to maintenance machinery
and animal feet, etc.
• vandalism

2. anchoring
• flotation of the lining (high water table)
• resist gravity force along side slope
• wind loading

• Plastic linings are typically 10 to 20 mil (0.010 to 0.020 inches, or 0.25 to 0.5 mm)
-- thicker membranes are usually recommendable because of increased
durability, and because the overall installation costs only increase by about 15%
for a doubling in thickness
• The USBR previously used 10 mil plastic linings, but later changed most
specifications to 20 mil linings
• Plastic linings of as low as 8 mil (PE), and up to 100 mil have been used in
canals and retention ponds
• Low density polyethylene (LDPE) is made of nearly the same material as
common trash bags (such as “Hefty” and “Glad” brands), but these trash bags
have a thickness of only 1.5 - 2 mils
• Plastic canal linings are manufactured in rolls, 5 to 7 ft in width, then seamed
together in a factory or shop to create sheets or panels of up to 100 ft (or more)
in width
• Rubber membrane linings can have a thickness ranging from 20 to 60 mil
• Flexible plastic and synthetic rubber linings are susceptible to damage
(punctures, tears) both during and after installation
• Flatter than normal side slopes (say 3:1) are sometimes preferred with plastic
linings to help prevent the possible migration of the lining down the slope, and to
help prevent uncovering of the lining by downward movement of soil
Gary P. Merkley 226 BIE 5300/6300 Lectures
 • Correctly installed plastic and synthetic rubber linings are completely impervious,
provided they have not been damaged, and provided that the flow level in the
channel does not exceed the height of the lining
• Plastic liners will “age” and lose plasticizer, causing a loss of flexibility and
greater potential for damage. Increased plasticizer during fabrication has been
shown to be effective in this regard

plas-ti-ciz-er (plas'tuh sie zuhr) n. a group of substances that are used in plastics to
impart viscosity, flexibility, softness, or other properties to the finished product

• Some canals in central Utah have had plastic linings for more than 30 years, and
most of it is still in good condition (measured seepage is essentially zero in the
lined sections, but some evidence of puncture/tearing has been found)
• Plastic lining material is sometimes used to retrofit existing concrete-lined canals
after the concrete lining canal fails and or continued maintenance is considered
infeasible
Preparing a canal section for buried
membrane lining (courtesy R.W. Hill)

• In the former Soviet Union, thin PE lining has been placed under precast slabs of
concrete lining in some canals
• In India, some canals have been lined with plastic (PE) on the bottom, and bricks
or tiles on the side slopes
• Polyethylene (PE) is the lowest cost geomembrane material, PVC is next lowest.
Some newer materials such as polyolefin are more expensive

Exposed and Buried Membranes

• Exposed membrane linings have been tried, but tend to deteriorate quickly for
various reasons
• Exposed membrane linings have recently been installed in some full (operating)
canals

BIE 5300/6300 Lectures 227 Gary P. Merkley

 • Buried membrane lining should have a cover layer of soil of approximately 1/12th
of the water depth, plus 10 inches
• Some vegetation can penetrate these types of linings (asphaltic too), so
sometimes soil sterilant is applied to the soil on the banks and bed before lining

Fly Ash

• Fly ash is a fine dust particulate material (roughly the size of silt) produced by
coal-burning power plants, usually in the form of glassy spheres
• Fly ash contains mostly SiO2 (silicon dioxide), Al2O3 (aluminum oxide), and Fe2O3
(iron oxide)
• Fly ash is often mixed with soil to form canal linings, the mixture being more
dense and less permeable than soil alone
• Fly ash is sometimes mixed with both soil and portland cement

V. References & Bibliography

ASAE. 1994. Standards. Amer. Soc. Agric. Engr., St. Joseph, MI.
Davis, C.V. and K.E. Sorensen (eds.). 1969. Handbook of applied hydraulics. McGraw-Hill Book
Company, New York, N.Y.
Frobel, R.K. 2004. EPDM rubber lining system chosen to save valuable irrigation water. Proc. of the
USCID conference, October 13-15, Salt Lake City, UT.
USBR. 1968. Buried asphalt membrane canal lining. USBR research report No. 12, Denver Federal
Center, Denver, CO.
USBR. 1963. Linings for irrigation canals. USBR technical report, Denver, CO.
USBR. 1984. Performance of plastic canal linings. USBR technical report REC-ERC-84-1, Denver
Federal Center, Denver, CO.
USBR. 1971. Synthetic rubber canal lining. USBR technical report REC-ERC-71-22, Denver Federal
Center, Denver, CO.
USBR. 1986. Tests for soil-fly ash mixtures for soil stabilization and canal lining. USBR technical
report REC-ERC-86-9, Denver Federal Center, Denver, CO.
USBR. 1994. Water operation and maintenance. USBR technical bulletin No. 170, Denver Federal
Center, Denver, CO.
www.geo-synthetics.com

Gary P. Merkley 228 BIE 5300/6300 Lectures

 Lecture 20
Inverted Siphons
Portions of the following were adapted from the USBR
publication “Design of Small Canal Structures” (1978)

I. Introduction

• Siphons, or inverted siphons, are used to convey water across a natural

depression, under a road, or under a canal
• Siphons are usually made of circular concrete pipe or PVC, connecting two canal
reaches in series
• Some siphons have rectangular cross-sections
• Siphons may have a straight lateral alignment, or may have changes in direction

Cebadas Inverted Siphon (Riobamba, Ecuador)

Entrance
1000
Exit
S li
990

980

970

960

950
Relative Elevation (m)

940

930

920 Road to Cebadas

910

900

890

880

870
Cebadas River
860
0 200 400 600 800 1000
Distance (m)

• When going across a depression, the siphon should be completely buried,

usually with a minimum of about 1 m of cover
• Siphons are like culverts, but instead of sloping down from inlet to outlet, they
slope down, then back up to the outlet
• Also, siphons usually have only one pipe (not two or three in parallel, as with
many culverts)
• Some siphons have multiple pipes in parallel; for example, when the original flow
capacity is to be increased
• In general, siphons are longer than culverts
BIE 5300/6300 Lectures 229 Gary P. Merkley
 • Open-channel flumes are alternatives to siphons, but may be more expensive to
build and/or to maintain, and may not be aesthetically acceptable
• Siphons can be very dangerous (the time to travel from inlet to outlet is usually
longer than a someone can hold their breath)
• Siphons can be very problematic with sediment-laden water because sediment
may tend to deposit at the low point(s)
• With large siphons, periodic cleaning is possible, but may be impractical with
smaller siphons
• Also, cleaning usually means a significant interruption in water delivery service
• Sometimes gravel and rock can enter the siphon
• However, trash racks and/or screens should always be provided at the inlet

II. Structural Components of Siphons

1. Pipe

• Most siphons are built out of pre-cast Siphon installation (USBR)

concrete pipe (PCP), reinforced with steel
in larger diameters
• Concrete pipe head class can go up to
200 ft (86 psi) or more
• Pre-stressed concrete cylinder pipe has
steel wire wrapped around the pipe with a
mortar coating
• Pre-stressed concrete pipe is typically
used on large siphons (over 10 ft in
diameter), with lengths of 20 ft, and
placed in trench with a special vehicle
• Pre-stressed concrete pipe was
previously considered the most
economical option for large siphons, but
experience has shown that the wire may
corrode in only 15-18 years of service
(e.g. Salt River project in Arizona)
• Older siphons have been built out of
asbestos-concrete (AC) mixtures, but this has been discontinued in the USA due
to health risks from exposure to asbestos
• Siphons can also be built with plastic or steel pipe, including other less common
possibilities (even wood)
• USBR siphon designs always have a single pipe, or barrel, but this is not a
design restriction in general, and you can find siphons with two or more pipes in
parallel
• USBR siphon designs always have circular pipe cross-sections

• A siphon in the Narmada canal in India was recently built to carry 40,000 cfs
(1,100 m3/s) across a depression; it has multiple rectangular conduits in parallel
Gary P. Merkley 230 BIE 5300/6300 Lectures
 • The Central Arizona Project (CAP) has four large siphons with 21-ft diameter pre-
stressed concrete pipe; some of these have already been replaced because of
corrosion and subsequent structural failure

2. Transitions
Concrete pipe repairs on
• Transitions for siphons are the an inverted siphon
inlet and outlet structures
• Most siphons have inlet and
outlet structures to reduce head
loss, prevent erosion and piping,
and maintain submergence
(“hydraulic seal”)
• It can be very hazardous to omit
inlet and outlet structures
because these locations are
often at steep embankments that
would erode very quickly in the
event of a breach or overflow
• An emergency spillway is
sometimes located just upstream
of a siphon inlet
• Transitions in smaller siphons may be of the same design at the inlet and outlet,
and standard designs can be used to reduce costs
• With larger siphons, it may be desirable to do a “site-specific” transition design,
possibly with different designs for the inlet and outlet

3. Gates and Checks

• Gates and checks can be installed:

(a) at the entrance of a siphon to control the upstream water level

(b) at the outlet of a siphon to control upstream submergence

• Operation of a gate at the entrance of a siphon may ensure hydraulic seal, but
will not ensure full pipe flow in the downhill section(s) of pipe at discharges below
the design value
• It is not common to install a gate at the outlet of a siphon (this is never done in
USBR designs)
4. Collars

• Collars may be used, as with culverts, to prevent “piping” and damage due to
burrowing animals
• However, with siphons they are not always necessary because the inlet and
outlet structures should be designed and built to direct all water into the entrance
and exit all water to the downstream channel
BIE 5300/6300 Lectures 231 Gary P. Merkley
 5. Blowoff and Vent Structures

• A “blow-off” structure is a valved outlet on top of the pipe at a low point in the
siphon
• Smaller siphons often do not have a blow-off structure
• These structures are used to help drain the siphon in an emergency, for routine
maintenance, or for winter shut-down
• Blow-off structures may have man holes (or “person access holes”) on large
siphons to allow convenient entry for manual inspection
• Blow-off structures can be used to periodically remove sediment from the pipe
• Continuous-acting vents are installed in some siphons to remove air during
operation (this is usually an after thought when “blow-back” (surging) problems
are manifested)
• Others have simple vertical pipes to vent air from the pipe, but these can have
the opposite effect
• Air can become trapped, especially in long siphons, during filling; filling of the
siphon should be gradual, not sudden
• Some blow-off structures are of the “clamshell” type, with top and bottom leafs off
of a tee at the bottom of the siphon
• Clamshell blow-offs are not so common, but have definite advantages in terms of
avoiding cavitation (handling high velocity flows at large heads) compared to
butterfly valves, for example

6. Canal Wasteways

• A wasteway (side-spill weir) is sometimes built in the canal just upstream of a

siphon inlet to divert the canal flow in the case of clogging of the siphon or other
emergency situation
• Also, the inlet to a siphon should always have trash racks and/or screens to
prevent rocks and other debris from entering the pipe
• If the inlet is at a canal turnout, the design should have a forebay to calm the
turbulence before entering the siphon; otherwise, air entrainment may be
significant

7. Safety Features

• In operation, siphons can appear to be harmless, especially in a large canal, but

can be deadly
• Just upstream of the siphon entrance the following may be used:

• posted signs with warnings

• ladder rungs on the canal banks
• steps on the canal banks
• cable with floats across the water surface
• safety net with cables and chains
• gratings or trash racks
Gary P. Merkley 232 BIE 5300/6300 Lectures
III. Design of Siphons

• The design of siphons has many similarities to the design of culverts; however,
unlike the design of culverts:

1. siphons are usually designed for full pipe flow

2. siphons are usually designed to minimize head loss
3. siphons take the water down, then back up

• USBR siphon designs are generally for an assumed 50-year useful life

1. Pipe Velocity Limit

• According to the USBR, pipe velocities at design discharge should be between

3.5 and 10 fps
• Recent USBR designs have mostly called for 8 fps velocity
• Many small culverts are designed for 10 fps
• In general, lower pipe velocities are fine for small siphons, but in large capacity
and or long siphons it is justifiable to design for higher velocities
• Long siphons can cost much less with even a slightly smaller pipe size

2. Head Losses

• Culverts are usually designed for full pipe flow from inlet to outlet (the outlet is
almost always submerged, and it is highly unlikely that open channel flow would
prevail throughout the siphon – the change in elevation is usually too great)
• Total head loss is the sum of: inlet, outlet, pipe, and minor head losses
• Convergence losses at the entrance are usually negligible, but divergence losses
at the exit (outlet) can be significant
• Most of the loss in a siphon is from pipe friction
• Outlet losses are typically about twice the inlet losses
• Minor losses in pipe bends are usually insignificant
• Most siphons are designed to carry the full design discharge without causing an
“M1” profile (backwater) in the upstream channel – to achieve this, it is important
to carefully estimate head losses
• If the total siphon head loss at the design discharge exceeds available head
(difference in upstream and downstream canal elevations and water depths) the
siphon will operate at a lower discharge and cause the upstream water level to
increase
• A hydraulic jump in the descending part of the siphon (upstream side) will greatly
increase the head loss, and may cause problems of surging and “blow-back”
• Blow-back occurs when air is entrained into the water due to a hydraulic jump in
the pipeline, or due to movement of a hydraulic jump within the pipe; water and
air periodically surge backwards through the inlet

BIE 5300/6300 Lectures 233 Gary P. Merkley

Gary P. Merkley 234 BIE 5300/6300 Lectures
 • Blow-back is usually more problematic in siphons with relatively flat descending
(upstream) slopes -- a small change in the downstream head can cause a
hydraulic jump to move within the pipe at the upstream end

3. Hydraulic Seal

• The “hydraulic seal” is the minimum required upstream head, relative to the
upper edge of the siphon pipe at the siphon inlet, to prevent the entrainment of
air at that location
• The hydraulic seal recommended by the USBR is equal to 1.5∆hv, where ∆hv is
the difference in velocity heads in the upstream open channel and in the pipe
(when flowing full)
• For a more conservative value of the hydraulic seal, use 1.5hpipe, where hpipe is
the velocity head in the siphon pipe when flowing full

hydraulic seal

siphon
inlet

4. Design Steps

• Determine the route that the siphon will follow

• Determine the required pipe diameter according to the design discharge and
allowable velocity
• Determine the appropriate transition structure types at the inlet and outlet, or
design custom transitions for the particular installation
• Design the siphon layout according to the existing terrain, and the proposed (or
existing) canal elevations at the inlet and outlet
• Determine the pressure requirements of the pipe according to the head (at the
lowest point) during operation
• Determine the total head loss in the siphon at the design discharge
• If the head loss is too high, choose a larger pipe or different pipe material; or,
consider adjusting the canal elevations at the inlet and outlet
• Use Fig. 2-7 on page 30 of Design of Small Canal Structures (USBR) to
determine whether blow-back might be a problem, and make adjustments if
necessary
BIE 5300/6300 Lectures 235 Gary P. Merkley
IV. Siphon Pressure Rating

• What is the maximum pressure in the inverted siphon pipe? This must be
calculated at the design stage so that a suitable pipe is selected.
• The maximum pressure is equal to the maximum of:

1. Water surface elevation at the outlet minus the elevation of the

lowest point in the siphon (unless there is a closed valve at the
outlet); or
2. Water surface elevation at the inlet minus the elevation of the
lowest point in the siphon, minus the friction loss from the
entrance to the low point.

• In any case, the maximum pressure will be at, or very near, the location of
minimum elevation in an inverted siphon
• If a gate or valve is at the siphon exit, and it is completely closed, the maximum
pressure will be according to #2, without subtracting friction loss (i.e. full pipe,
zero flow condition); otherwise, the zero flow condition pertains to #1
• Note that the above assumes that an open channel is upstream of the siphon
entrance, and an open channel is at the siphon exit
• Note that in order to calculate friction loss, you need to assume a pipe diameter
(ID) and a pipe material
• Due to possible water surging in the pipe, the pressure may be somewhat higher
than that calculated above, so consider adding a 10% safety factor

References & Bibliography

USBR. 1978. Design of small canal structures. U.S. Government Printing Office, Washington, D.C.
435 pp.

Gary P. Merkley 236 BIE 5300/6300 Lectures

Lecture 21
Culvert Design & Analysis
Much of the following is based on the USBR publication:
“Design of Small Canal Structures” (1978)

I. Cross-Drainage Structures

• Cross-drainage is required when a canal will carry water across natural drainage
(runoff) channels, or across natural streams; otherwise, the canal may be
damaged
• In some cases, cross-drainage flows are collected in a small channel paralleling
the canal, with periodic cross-drainage structures over or under the canal; this is
especially prevalent where there are poorly defined natural drainage channels
• In culvert design for carrying runoff water, usually one of the big questions is
what the capacity should be
• When the canal capacity is less than the natural channel capacity, it may be
economical to build an inverted siphon so the canal crosses the natural channel
• With siphon crossings, it is not nearly as important to accurately estimate the
maximum flow in the natural channel because the structure is for the canal flow
• In other cases, it may be more economical to provide cross-drainage by building
a culvert to accommodate natural flows after the canal is constructed
• In these cases, the cross-drainage structure does one of the following:

1. Carry water under the canal

2. Carry water over the canal
3. Carry water into the canal

• Here are the common cross-drainage solutions:

1. Culverts

• These are often appropriate where natural flows cross a fill section of the
canal
• Culverts may tend to clog with weeds, debris, rock, gravel, and or
sediments, especially at or near
the inlet

2. Over-chutes

• These are appropriate where

the bottom of the natural
channel is higher than the full
supply level of the canal
• For example, over-chutes might
be used in a cut section of the
canal
Canal over-chute & bridge

BIE 5300/6300 Lectures 237 Gary P. Merkley

 • Open-channel over-chutes can carry debris and sediment that might clog
a culvert, but pipe over-chutes may be equally susceptible to clogging

3. Drain Inlets

• With these structures, the flow of the natural channel is directed into the
canal
• These may be appropriate where the natural flows are small compared
to the canal capacity, and or when the natural flows are infrequent
• These may be appropriate when the canal traverses a steep slope, and
cross-drainage might cause excessive downhill erosion, compromising
the canal
• These may be less expensive than over-chute or culvert structures, but
may require more frequent maintenance of the canal
• Drain inlets may be problematic insofar as rocks, sediment and other
debris can clog the inlet and or fill the canal near the inlet, obstructing
the canal flow

II. Alignment

• Align the culvert along natural open channels where possible so that the natural
runoff pattern is not disturbed any more than necessary
• If the natural drainage channel is not perpendicular to the canal, it is best to have
a skewed alignment of the culvert
• One or more bends in the culvert can be used to help follow the natural channel,
especially in longer culverts
• If there is no apparent natural runoff channel, consider using the shortest straight
path from inlet to outlet
• In some cases it may be unnecessary or undesirable to follow a natural channel

III. Barrel Profile

• Knowing the inlet and outlet locations will determine the length and slope of the
culvert
• The invert of the inlet and outlet should correspond approximately to the natural
ground surface elevations at the two respective locations -- otherwise,
sedimentation and or erosion will likely occur, requiring maintenance
• However, a compound slope may be needed if:

1. The culvert would not have enough vertical clearance

under a canal (about 2 ft for an earth canal, or 0.5 ft for a
concrete canal), road, etc.;
2. The slope of the culvert would cause supercritical open-
channel flow, which might require a downstream energy
dissipation structure (making the design more costly); or,
3. You want to force a hydraulic jump to dissipate energy.

Gary P. Merkley 238 BIE 5300/6300 Lectures

 • The USBR recommends, in general, a minimum slope of 0.005 and a maximum
slope of somewhat less than the critical slope (maintain subcritical flow)
• The minimum slope is imposed in an effort to prevent sediment deposition in the
culvert barrel
• The barrel of the culvert is usually circular (perhaps corrugated pipe) or
rectangular
• The maximum slope is imposed in an effort to avoid the additional cost of an
energy dissipation structure at the outlet (channels upstream and downstream of
culverts are typically unlined, although there may be some riprap)
• With a compound slope, the upstream slope is steeper than critical, and the
downstream slope is mild, thereby forcing significant energy dissipation through a
hydraulic jump in the vicinity of the break in grade, inside the barrel

IV. Inlets and Outlets

• USBR Culvert Inlets

Type 1: “broken-back transition”, appropriate for natural channels with well-

defined upstream cross-section (USBR Figs. 7-1 & 7-2)
Type 2: suitable for wide natural channels with poorly-defined upstream
cross section (USBR Fig. 7-4)
Type 3: “box inlet”, also for use in a poorly-defined natural channel, but
allows for a lower barrel invert at the inlet (USBR Fig. 7-5)
Type 4: similar to Type 3, but with a sloping invert, allowing for an even
lower barrel inlet (USBR Figs. 7-6 & 7-7)

• USBR Culvert Outlets

1. With energy
dissipation structure
2. Without energy
dissipation structure

• There are other USBR

standard inlet designs
(besides the above
four)
• USBR-type culvert Type 1 Transition (USBR)
inlets and outlets are
made almost exclusively of concrete
• Some corrugated metal culverts have a circular or elliptical cross section with
smooth metalic inlet and outlet transitions
• Use standard inlet & outlet designs if possible to save time and to avoid
operational and or maintenance problems

BIE 5300/6300 Lectures 239 Gary P. Merkley

 Type 3 Transition (USBR)

Type 4 Transition (USBR)

Gary P. Merkley 240 BIE 5300/6300 Lectures

V. Pipe Collars

• Pipe collars are used to prevent “piping” along

the outside of the barrel and or damage by
burrowing animals
• For culverts under canals, the typical USBR
design calls for three collars: one under the
center of the upstream canal bank, and two
under the downstream canal bank
• A “short path” between two adjacent collars
means that the collars are too close together
and or their diameters are too small
• The USBR recommends the following for minimum collar spacing:

Xmin = 1.2 Y (1)

VI. Basic Design Hydraulics

• Culverts are typically designed for full- Culvert with collars (USBR)
pipe flow in the barrel at the design
discharge value
• This means that pressurized pipe flow is
impending at the design discharge, but
at lower flow rates open-channel flow
exists in the barrel
• The upper limit on barrel velocity is
usually specified at about 10 fps, or
perhaps 12 fps with an energy
dissipation structure at the outlet
• For full pipe flow without inlet and outlet
structures, in which case the culvert is
simply a buried pipe, you can use a limit
of 5 fps

• Knowing the design discharge and the

velocity limit, the diameter (circular
barrels) for full pipe flow can be directly
calculated
• For rectangular barrel sections, you need to determine both width & height

• Discharge capacity can be checked using the Manning (or Chezy) equation for a
circular section running full (again, impending pressurization)
• For new pre-cast concrete pipe, the Manning “n” value is about 0.013, but for
design purposes you can use a higher value because the pipe won’t always be
new

BIE 5300/6300 Lectures 241 Gary P. Merkley

 • You can also check the discharge using the Darcy-Weisbach equation, with
specified values for upstream and downstream water surface elevations in the
inlet and outlet structures, respectively

• The head loss through a typical inlet structure with inlet control can be estimated
as a “minor loss” by:
V2
hf = K (2)
2g

where the coefficient K may vary from 0.05 for a smooth, tapered inlet transition,
flush with the culvert barrel, to 0.90 for a projecting, sharp-edged barrel inlet
• Note that the inlet and or outlet losses may or may not be “minor” losses when
dealing with culverts, especially when the barrel is short

• For outlet control, the head loss is estimated as in the above equation for inlet
control, except that there will also be expansion losses downstream
• For barrel control, the head loss is the sum of the inlet, barrel, and outlet losses

References & Bibliography

USBR. 1978. Design of small canal structures. U.S. Government Printing Office, Washington, D.C.
435 pp.

Gary P. Merkley 242 BIE 5300/6300 Lectures

Lecture 22
Example Culvert Design
Much of the following is based on the USBR technical publication
“Design of Small Canal Structures” (1978)

I. An Example Culvert Design

• Design a concrete culvert using the procedures given by the USBR (Design of
Small Canal Structures, USBR 1978)
• The culvert will go underneath a concrete-line canal as shown in the figure below,
perpendicular to the canal alignment (shortest path across)

• The outside slope of the canal banks is 1.5:1.0 on both sides (H:V)
• The inside side slope of the concrete lining is 1:1 (both sides)
• The concrete lining thickness is 0.05 m
• Elevation of the top of the canal banks is given as 134.00 m, with elevations of
the original ground surface at intersections with canal banks given in the figure
• The berm on both banks is 3.0 m wide
• The depth of the canal from the bottom to the top of the berms is 4.0 m, with the
upper 30 cm unlined

• The barrel will be pre-cast circular concrete pipe, and the inlet and outlet
structures can be specified as Type 1, 2, 3, or 4, as given by the USBR
• The available concrete pipe has an inside diameter of: 60, 70, 80, and 90 cm
• You must select from one of these diameters for the culvert barrel
• Length of the pipe is 1.5 m per section

• The hydrological assessment of the area came up with a maximum surface

runoff of 2.4 m3/s for a 20-year flood in the area uphill from the canal
• This is the runoff rate that the culvert must be designed to carry

• The upstream and downstream natural channels are wide and poorly defined in
cross-section, and no effort will be made to develop prismatic channels upstream
of the culvert inlet, nor downstream of the culvert outlet

BIE 5300/6300 Lectures 243 Gary P. Merkley

II. A Culvert Design Solution

1. Determine the Horizontal Distance

• The horizontal distance from the culvert inlet to the pipe outlet is (from left to
right):

1.5*(134.00-132.12)+3.0+4.0+5.5+4.0+3.0+1.5*(134.00-128.06) = 31.23 m

2. Determine the Required Pipe Size

• Use a maximum average barrel velocity of 3.0 m/s

• Then, for the design discharge of 2.4 m3/s:

4Q 4(2.4)
D= = = 1.01 m (1)
πV π(3.0)

• The largest available pipe size is 90 cm; therefore, two or more pipes are needed
in parallel for this culvert design

• For half the design discharge, 1.2 m3/s, the required diameter is:

4Q 4(1.2)
D= = = 0.71 m (2)
πV π(3.0)

• Then, we can use two 80-cm ID pipes at a full pipe flow velocity of 2.39 m/s
• It would also be possible to use three 60-cm ID pipes at a full pipe flow velocity of
2.83, which is closer to the maximum velocity of 3.0 m/s
• But, choose two 80-cm ID pipes because it will simplify installation, require less
excavation work, and may reduce the overall pipe cost

3. Determine the Energy Loss Gradient

• With the full pipe flow impending, the energy loss gradient can be estimated by
the Manning equation for open-channel flow
• Use a Manning n value of 0.015 for new concrete pipe, with a slight safety factor
for aging (typical useful life is estimated as 40 to 50 years)
• Use half the design discharge because two 80-cm ID pipes will be installed in
parallel

2 2 4/3
Sf =
Q2n2 Wp4 / 3
=
(1.2 ) ( 0.015 ) ( 2.51)
= 0.011 m / m (3)
A10 / 3 ( 0.503 )10 / 3

Gary P. Merkley 244 BIE 5300/6300 Lectures

 where the wetted perimeter, Wp, for full pipe flow is πD; and the area, A, is πD2/4,
for an inside diameter of 0.80 m and half the design discharge, 1.2 m3/s

4. Determine the Critical Slope

• For critical flow, the Froude number is equal to unity:

Q 2T
Fr2 = = 1.0 (4)
gA 3

• For circular pipes, the following definitions apply:

⎛ 2h ⎞
β = 2cos−1 ⎜ 1 − ⎟ (5)
⎝ D⎠

⎛β⎞ βD
T = D sin ⎜ ⎟ and, Wp = (6)
⎝2⎠ 2

D2
A= (β − sin β ) (7)
8
• Solve for depth, h, such that Fr2 = 1.0 for Q = 1.2
m3/s and D = 0.80 m
• Using the Newton method, hc = 0.663 m

• Calculate the energy loss gradient (critical slope)

corresponding to this depth
• For a depth of 0.663 m, the flow area is 0.445
m2, and the wetted perimeter is 1.83 m
• Applying the Manning equation:

2 2 4/3
(1.2 ) ( 0.015 ) (1.83 )
( Sf )crit = = 0.011 m / m (8)
( 0.445 )10 / 3
• This is essentially the same loss gradient as for impending full pipe flow, but
note that the critical flow depth is 83% of the pipe ID
• If the slope of the pipe is 0.011 m/m or greater, critical flow can occur

BIE 5300/6300 Lectures 245 Gary P. Merkley

5. Determine the Minimum Upstream Pipe Slope

• The upstream pipe will be situated so as to begin at Elev 132.12, and just
clear the canal base at the left side
• The elevation of the canal base is 134.00 - 4.0 = 130.00 m

• The horizontal distance from the culvert inlet to the left side of the canal base
is 1.5*(134.00-132.12)+3.0+4.0 = 9.82 m
• The pipe must drop at least 132.12-130.00+0.05+0.2 = 2.37 m over this
horizontal distance
• This corresponds to a pipe slope of 2.37/9.82 = 0.24 m/m (24%)

• The critical slope is 1.1% (< 24%), so the culvert will have inlet flow for the
design discharge (and for lower discharge values)
• At the design discharge, we will expect a hydraulic jump in the pipe upstream
of the bend, because the pipe slope will be lower in the remaining
(downstream) portion of the culvert

• It is necessary to check that the slope of the downstream pipe does not
exceed the critical slope

6. Determine the Downstream Pipe Slope

• The downstream part of the culvert barrel will travel a horizontal distance of
31.23 - 9.82 = 21.41 m
• The change in elevation over this distance will be 129.75 - 128.06 = 1.69 m
• Then, the slope of the downstream part of the pipe will be 1.69/21.41 = 0.079
m/m (7.9%)

• This slope is greater than the critical slope, and is not acceptable because it
would cause supercritical flow throughout, from inlet to outlet, causing erosion
downstream (unless erosion protection is used)
• Use the USBR recommended downstream slope of 0.005 (0.5%), which is
less than the critical slope of 1.1%

• To accomplish this, the upstream (steep) portion of the culvert pipe can be
extended further in the downstream direction (to the right)
• Equations can be written for the tops of the upstream and downstream pipes:

Upstream: .................................... y = −0.24x + 132.12

Downstream:.............................. y = −0.005x + 128.22

[Note: 128.06+(31.23)(0.005) = 128.22]

Gary P. Merkley 246 BIE 5300/6300 Lectures

 • Solving the two equations for distance, a value of x = 16.60 m is obtained
• This is the distance from the inlet at which the top of the upstream pipe
intersects the top of the downstream pipe
• The elevation of the intersection point is y = -0.24(16.60) + 132.12 = 128.14
m
• The minimum clearance of 0.2 m under the canal bed is still provided for, but
the excavation for the culvert will require more work
• Note that in many locations the natural ground slope is insufficient to justify a
critical slope on the upstream side, and a subcritical slope on the downstream
side of the culvert

7. Determine the Pipe Lengths

• The approximate length of the upstream (steep) pipe is:

Lus = (16.60 )2 + (132.12 − 128.14 )2 = 17.07 m (9)

• The approximate length of the downstream pipe is:

Lds = ( 31.23 − 16.60 )2 + (128.14 − 128.06 )2 = 14.63 m (10)

• At 1.5 m per pipe, this corresponds to (17.07 + 14.63)/1.5 ≈ 21 pipes

• For a double-barreled culvert, there must be about 42 pipe lengths

8. Specify Inlet and Outlet Types

• The inlet and outlet can be USBR Type 2, 3, or 4

• Type 1 would not be appropriate because the upstream and downstream
channels are not well defined
• An energy dissipation structure at the outlet is not needed (outlet velocity will be
< 15 fps)

BIE 5300/6300 Lectures 247 Gary P. Merkley

9. Specify Collar Placement and Size

• Use the standard USBR culvert design, calling for two collars under the downhill
canal bank, and one collar under the uphill bank
• The distance between the two collars under the downhill bank will be
approximately 3.0 m plus 2.0 ft, or 3.61 m
• Then, the Y value is X/1.2, or Y = 3.61/1.2 = 3.0 m
• This gives very large collars
• There are other methods for determining collar size, but in this case the Y value
can be taken as 1.0 m, which would be only about one meter below the uphill
canal berm
• More information about the site and soil would be required to verify the adequacy
of the collar design
• Many culverts don’t have collars anyway, and in some cases they are
problematic because they impede effective soil compaction – “piping” may be
worse with the collars

References & Bibliography

USBR. 1978. Design of small canal structures. U.S. Government Printing Office, Washington, D.C.
435 pp.

Gary P. Merkley 248 BIE 5300/6300 Lectures

Lecture 23
Culvert Hydraulic Behavior

I. Open-Channel Definitions

• The following open-channel definitions apply in most cases:

Canal Open channel of mild slope (subcritical flow) and relatively long;
could be lined or unlined
Flume A channel built above the natural ground surface, usually of mild
slope and rectangular or circular cross section, crossing a
depression or running along the contour of a hillside
Chute Like a flume, but having a steep slope (supercritical flow; Fr2 > 1.0)
and usually with some type of energy dissipation structure on the
downstream side (outlet)
Culvert One or more circular or rectangular pipes/conduits in parallel,
crossing under a road, canal, or other structure, either flowing full
(pressurized) or part full (open-channel flow); often used as a
cross-drainage structure
Prismatic This means a constant cross-sectional shape with distance,
constant and uniform bed slope, and straight channel alignment,
applied to any of the above

II. Culvert Hydraulics & Flow Regimes

• A culvert can serve as a combination open-channel and closed conduit

structure, depending upon the type of flow condition in the culvert
• Most of the research involving the hydraulics of culverts has been concerned
with the use of such structures under highways
• Some of the research has focused on inlet control (free orifice flow) and
submerged outlet control (submerged orifice flow)
• For culverts placed in an irrigation conveyance channel (i.e. not as a cross-
drainage structure), free surface (open-channel) flow usually occurs in the
culvert
• Culverts as part of an irrigation canal often pass under vehicular bridges
• Typically, downstream hydraulic conditions will likely control the depth of flow
and the discharge in the culvert
• Culvert hydraulics can be much more complex than might appear at first
glance

III. Flow Regimes

• The classification of the hydraulic behavior of culverts can take several forms
• Three primary groupings can be used to describe the hydraulics of culverts
• These groups are based on the three parts of the culvert that exert primary
control on culvert performance and capacity:

BIE 5300/6300 Lectures 249 Gary P. Merkley

 (1) the inlet;
(2) the barrel; and
(3) the outlet

• Usually, one of these three primary controls determines the performance and
capacity of the culvert
• An example of this is a projecting, square-edged inlet with the barrel on a
steep slope (Fr2 > 1.0) and flowing partly full: if the inlet is not submerged, the
upstream water level (headwater) is determined by the inlet characteristics
alone
• In other cases, two or even all three primary controls can simultaneously
affect the performance and discharge capacity

• For example, if the inlet and outlet are submerged and the barrel is full, then a
designer can determine the headwater elevation by adding the outlet losses,
the barrel friction losses, and the inlet losses to the tailwater (downstream)
elevation (assuming the same specific energy in both the upstream &
downstream open channels)

• The classification is further subdivided under each main group, as shown in

the table below (Blaisdell 1966)
• The classification indicates the number of factors the designer must consider
when determining the performance of a culvert and computing its capacity
under different regimes
• Only those items that exert a control on the hydraulic performance of a culvert
are listed in the table
• Many alternatives are possible for each control

• For example, each type of inlet will have a different effect on the culvert
performance
• Many of the items listed in the table are inter-related, which further
complicates an already difficult problem
• For instance, the depth of the flow just inside the culvert entrance depends on
the inlet geometry
• If this depth is less than the normal depth of flow, a water surface profile must
be computed beginning with the contracted depth of flow to determine the
flow depth at the culvert outlet

Gary P. Merkley 250 BIE 5300/6300 Lectures

 INLET CONTROL

A. Unsubmerged (Free Surface)

1. Weir (Modular Flow)
2. Surface profile (Non-Modular Flow)
B. Submerged (Inlet Crown Under Water)
1. Orifice (Free Orifice Flow)
2. Vortex (Non-Aerated Jet)
3. Full (Submerged Orifice Flow)

BARREL CONTROL

C. Length
1. Short
2. Long
D. Slope
1. Mild
i. Barrel slope less than critical slope
a. Part full, normal depth greater than critical depth
b. Full, not applicable
ii. Barrel slope less than friction slope
a. Part full, depth increases along barrel
b. Full, barrel under pressure
2. Steep
i. Barrel slope steeper than critical slope
a. Part full, normal depth less than critical depth
b. Full, not applicable
ii. Barrel slope steeper than friction slope
a. Part full, depth decreases along barrel (increases if
the inlet causes the depth inside the inlet to be less
than normal depth)
b. Full, barrel under suction
E. Discharge
1. Partially Full (Free-Surface Open-Channel Flow)
2. Slug and Mixture (Unsteady Flow)
3. Full (Closed Conduit Flow)

OUTLET CONTROL

F. Part Full (Free Surface Open Channel Flow)

1. Critical Depth (Free Flow)
2. Tailwater (Submerged Flow)
G. Full (Closed Conduit Flow)
1. Free (Free Orifice Flow)
2. Submerged (Submerged Orifice Flow)

BIE 5300/6300 Lectures 251 Gary P. Merkley

IV. Hydraulically Short & Long

• If the computed outlet depth exceeds the barrel height, the culvert is hydraulically
long, the barrel will fill, and the control will be the inlet, the barrel, and the outlet
• If the computed depth at the outlet is less than the barrel height, the barrel is only
part full and the culvert is considered hydraulically short, will not fill, and the
control will remain at the inlet
• Whether a culvert is hydraulically long or short depends on things such as the
barrel slope and the culvert material
• For example, changing from corrugated pipe to concrete pipe can change the
hydraulic length of a culvert from long to short
• A similar effect could result from a change in the inlet geometry
• Flow in culverts is also controlled by the hydraulic capacity of one section of the
installation
• The discharge is either controlled at the culvert entrance or at the outlet, and is
designated inlet control and outlet control, respectively
• In general, inlet control will exist as long as the ability of the culvert pipe to carry
the flow exceeds the ability of water to enter the culvert through the inlet
• Outlet control will exist when the ability of the pipe barrel to carry water away
from the entrance is less than the flow that actually enters the inlet
• The location of the control section will shift as the relative capacities of the
entrance and barrel sections change with increasing or decreasing discharge
• This means that it cannot be assumed that a given culvert will always operate
under the same hydraulic regime

V. Three Hydraulic Classifications

Inlet Control

• Inlet control means that the discharge capacity of a culvert is controlled at the
culvert entrance by the depth of headwater and the entrance geometry, including
the barrel shape and cross-sectional area
• With inlet control, the roughness and length of the culvert barrel, as well as outlet
conditions (including depth of tailwater), are not factors in determining culvert
capacity
• An increase in barrel slope reduces the headwater (inlet) depth, and any
correction for slope can be neglected for conventional or commonly used culverts
operating under inlet control

Barrel Control

• Under barrel control, the discharge in the culvert is controlled by the combined
hydraulic effects of the entrance (inlet), barrel length & slope, and roughness of
the pipe barrel
• The characteristics of the flow do not always identify the type of flow
• It is possible, particularly at low flows, for length, slope, and roughness to control
the discharge without causing the pipe to flow full
Gary P. Merkley 252 BIE 5300/6300 Lectures
 • But, this is not common at design discharges for most culverts
• The usual condition for this type of flow at the design discharge is one in which
the pipe cross section flows full for a major portion (but not all) of the length of
the culvert
• The discharge in this case is controlled by the combined effect of all hydraulic
factors

Outlet Control

• Culverts flowing with outlet control can have the barrel full of water or partly full
for either all or part of the barrel length
• If the entire cross section of the barrel is filled with water for the total length of the
barrel, the culvert is said to be flowing full

VI. Culvert Flow Regimes

• The following is a slightly different classification of culvert flow

• The flow through culverts can be divided into six categories (French 1985;
Bodhaine 1976), depending on the upstream and downstream free-surface water
elevations, and the elevations of the culvert inlet and outlet
• The following categories are defined based on the design (maximum) discharge
capacity of a culvert

Type I Flow Inlet control. Critical depth occurs at or near the inlet:

(a) The slope of the culvert barrel is greater than the critical slope
(b) The downstream water surface elevation is lower than the elevation of the
water surface where critical flow occurs at the inlet
(c) The upstream water depth is less than approximately 1.5 times the barrel
height (or diameter)

Type II Flow Outlet control. Critical depth occurs at or near the outlet:

(a) The slope of the culvert barrel is less than critical slope
(b) The downstream water surface elevation is lower than the elevation of the
water surface where critical flow occurs at the outlet
(c) The upstream water depth is less than approximately 1.5 times the barrel
height (or diameter)

Type III Flow Barrel control. Subcritical barrel flow, a gradually-varied flow
profile:

(a) The downstream water surface elevation is less than the height (or
diameter) of the barrel, but is more than the critical depth at the outlet
(b) The upstream water depth is less than approximately 1.5 times the barrel
height (or diameter)

BIE 5300/6300 Lectures 253 Gary P. Merkley

 Type IV Flow Barrel control. Both the upstream and downstream ends of the
culvert are submerged, and the barrel is completely full of water.
The culvert behaves essentially like an orifice, but with additional
head loss due to the barrel.

Type V Flow Inlet control. The barrel flows partially full and supercritical flow
occurs in the barrel downstream of the inlet:

(a) The slope of the culvert barrel is greater than the critical slope
(b) The upstream water depth is greater than approximately 1.5 times the
barrel height (or diameter)

Gary P. Merkley 254 BIE 5300/6300 Lectures

 Type VI Flow Barrel control. The culvert is completely full of water:

(a) The upstream water depth is greater than approximately 1.5 times the
barrel height (or diameter)
(b) The outlet is unsubmerged (downstream depth less than the barrel height
or diameter)

VII. Additional Culvert Flow Regimes

References & Bibliography

Lindeburg, M.R. 1999. Civil engineering reference manual. 7th Ed. Professional Publications, Inc.,
Belmont, CA.

BIE 5300/6300 Lectures 255 Gary P. Merkley

Gary P. Merkley 256 BIE 5300/6300 Lectures
Lecture 24
Flumes & Channel Transitions

I. General Characteristics of Flumes

• Flumes are often used:

1. Along contours of steep slopes where

minimal excavation is desired
2. On flat terrain where it is desired to
minimize pumping, except perhaps at
the source
3. On flat terrain where it is desired to
avoid pumping, except maybe at the
water source
4. Where cross-drainage is required over Flume/chute on a steep slope
a depression

• A pipeline or siphon can be considered to be

an alternative to a flume in many cases, or to
a canal – these days the choice is essentially
one of economics
• Aesthetics may make an inverted siphon
preferable to a flume for the crossing of a
depression, even though the flume could be
less costly
• Unlike most canals, flumes seldom have
gates or other flow control structures; that is,
they are generally used strictly for Pipeline… aesthetics?
conveyance

• The average flow velocity in a flume is

higher than that for most canals
• But the flow regime in flumes is usually
subcritical, as opposed to chutes, which
usually operate under supercritical flow
conditions
• Flume cross-section shapes are typically
rectangular, but may also be semi-circular
or parabolic
• Several irrigation systems in Morocco have
networks of elevated semi-circular flumes Flume with parabolic section
• Flumes with non-rectangular sections are
usually pre-cast concrete, or concrete mixed with other materials
• Flumes may have under-drains, side inlet structures, and over-pass structures to
handle cross flows, especially for cross flows going down a slope

BIE 5300/6300 Lectures 257 Gary P. Merkley

 • Some flumes are elevated for crossing natural channels, depressions, roads,
railroads, etc.

II. Flume Design

• The hydraulic design of flumes involves calculations of maximum steady-state

capacity:

1. Determine the design discharge

2. Select an appropriate route with consideration to total length, construction
costs, longitudinal bed slope, need for cross-drainage, and other factors
3. Select a cross-sectional shape (usually rectangular, sometimes circular)
4. Select a material for the flume channel (this will determine the roughness)
5. Use the Chezy or Manning equations to determine the size of the cross
section (unless you expect nonuniform flow conditions)
6. Make adjustments as necessary to accommodate the design discharge and
other technical, aesthetic, safety, and economic factors

• As noted in a previous lecture, the hydraulic efficiency of the cross section may
be a consideration, semi-circular sections being the most efficient
• Inlet and outlet water levels may be a consideration in the flume design
• For example, the flume may be connected to a reservoir with a specified range of
water surface elevation
• As in any open-channel, avoid flume designs that would produce near-critical
flow conditions at the design capacity; attempt to arrive at a design with Fr < 0.9
• Consider USBR and or other guidelines for the inclusion of freeboard at the
design discharge
• Note that an overflow could quickly cause severe erosion under the flume

III. Flume Cross-Sections

• A variety of cross section shapes have been used for flumes

• The most hydraulically efficient section is that which has a maximum value of the
hydraulic radius for a given discharge, roughness and side slope
• The most hydraulically efficient rectangular section is that with a bed width to
depth ratio, b/h, of 2.0
• According to the USBR, the most cost-effective rectangular sections have a
range of ratios from 1 ≤ b/h ≤ 3

Gary P. Merkley 258 BIE 5300/6300 Lectures

 Flume Cross Section
0.30

0.25

0.20
Height (m)

0.15

0.10

0.05

0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Distance (m)

Cross section of a pre-cast concrete flume used in several irrigation

systems in the Dominican Republic

BIE 5300/6300 Lectures 259 Gary P. Merkley

IV. Function of Channel Transitions

• Channel transitions occur at locations of cross-sectional change, usually over a

short distance
• Transitions are also used at entrances and exits of pipes such as culverts and
inverted siphons
• Below are some of the principal reasons for using transitions:

1. Provide a smooth change in channel cross section

2. Provide a smooth (possibly linear) change in water surface
elevation
3. Gradually accelerate flow at pipe inlets, and gradually
decelerate flow at pipe outlets
4. Avoid unnecessary head loss through the change in cross
section
5. Prevent occurrence of cross-waves, standing waves, and
surface turbulence in general
6. Protect the upstream and downstream channels by reducing soil
erosion
7. To cause head loss for erosion protection downstream; in this
case, it is an energy dissipation & transition structure

V. General Comments about Transitions

• Transitions at pipe, flume, and canal outlets (ends) often have energy dissipation
structures included or added to the design
• Transitions are typically made of concrete or earth, the latter often having some
sort of riprap protection
• Earthen transitions between open-channel and pipe flow (culverts and siphons)
are often acceptable when the velocity in the pipe is less than about 3.5 fps (1
m/s)
• Transitions can have both lateral and vertical (bed) contraction or expansion

• The optimum angle of lateral convergence (at contractions) is given as 12.5° by

Chow (1959), corresponding to a 4.5:1 ratio

• The optimum angle of lateral divergence (at expansions) is often taken as

approximately 9.5°, or a 6:1 ratio

Gary P. Merkley 260 BIE 5300/6300 Lectures

 • This divergence ratio is used in a number of expansion transitions in open
channels, pipes, nozzles and other devices

• It is very difficult to design transitions that work well over a wide range of flows
• Consequently, transitions are often designed for specified maximum flow rates

• The theoretical analysis of hydraulics in transitions is limited, especially when the

analysis is for one-dimensional flow
• Three-dimensional analysis is usually required to predict the occurrence of waves
and to estimate head loss under different conditions, but even this is often
inadequate
• Physical models are required, in general, for reliable evaluation of the
performance of a particular transition design under various flow conditions
• Some day, mathematical models will be up to the task

VI. Standard USBR Transitions

• Standard designs are selected for many small transitions because the time and
effort to design a special transition for a particular case may not be justified, and
because the engineer may not know how to go about designing a transition
based on the application of hydraulic equations
• The most common type of transition used in small canal structures by the USBR
is called “broken back” (Type I), which has vertical walls on the
converging/diverging sides

• For small structures (about 100 cfs or less), the USBR usually applies one of five
standard transition designs
• Standard USBR transitions are given in Chapter VII of the “Design of Small
Canal Structures” (1978) book and other publications and design reports
• More sophisticated transitions may be designed for larger flow rates

• Many of the USBR transitions are inlet and outlet structures for pipes, not
transitions between channel cross section changes
• Many open-channel-to-pipe transition designs call for a transition length (in the
direction of flow) of about three times the diameter of the pipe

BIE 5300/6300 Lectures 261 Gary P. Merkley

 • The length can also be defined by a specified angle of convergence or
divergence, knowing the upstream and downstream channel widths
• For earthen transitions the minimum length may be given as about 5 ft or 2 m

• Inlet transitions to inverted siphons are generally designed such that the top of
the pipe is below the upstream open-channel water surface for the design
discharge (this is to help prevent the continuous entry of air into the siphon,
which reduces capacity)
• The USBR calls the difference in elevation between water surface and top of pipe
opening the “hydraulic seal”
• Inlet transitions for culverts are generally designed such that inlet control occurs,
possibly causing supercritical flow and a downstream hydraulic jump inside the
pipe (barrel)

VII. Computational Design Procedures

• A change in cross-sectional size or shape will generally cause a change in water

surface elevation
• However, in some cases it is desirable to avoid surface elevation changes and
backwater profiles
• This can be done by calculating one or more dimensions of the cross sections
such that the water surface elevation continues on a uniform downhill gradient
through the transition for a specified “design” discharge
• This is possible for subcritical flow only, not for supercritical flow

• The head loss in inlet transitions is typically taken to be about 0.1∆hv, where ∆hv
is the change in velocity head from upstream to downstream across the transition
• For outlet transitions, the loss is usually about double this, or 0.2∆hv
• In some cases, these losses may be twice these values, or more, but ∆hv is
usually very small anyway, compared to the specific energy

• Another approach, sometimes used in changes of width for rectangular sections,

is the reverse parabola transition, consisting of two parabolas which define the
bed elevation through the transition, or the channel width through the transition

• The second parabola is the same as the first, but inverted vertically & horizontally

Gary P. Merkley 262 BIE 5300/6300 Lectures

 • The transition length is determined before the parabola is defined, and the total
length of the transition is divided into two halves with the parabola in the second
half being the inverted parabola of the first half
• This procedure is described in some hydraulics books
• You can also use a unique 3rd-degree polynomial instead of two parabolas by
fixing the end points and specifying zero slope at each end
• In many cases the transition can be designed to give a very smooth water
surface
• The reverse parabola approach can also be applied to trapezoidal to rectangular
cross-section transitions, but the calculations are more involved
• A typical design approach is to use a continuous and uniform reduction in side
slope along the length of the transition, with a calculated bed elevation such that
the water surface continues as smoothly as possible from upstream normal depth
to downstream normal depth

BIE 5300/6300 Lectures 263 Gary P. Merkley

Gary P. Merkley 264 BIE 5300/6300 Lectures
Lecture 25
Design Example for a Channel Transition

I. Introduction

• This example will be for a transition from a trapezoidal canal section to a

rectangular flume section
• The objective of the transition design is to avoid backwater (GVF) profiles in the
transition, and upstream & downstream of the transition
• We will specify a length for the transition, but the total net change in canal invert
elevation across the transition will be defined as part of the solution
• The main design challenge will be to determine the shape (profile) of the canal
invert across the transition

II. Given Information

• The design flow rate is 15.0 m3/s

• The upstream trapezoidal section has 1:1 side slopes (m = 1)
• The bed slope of the upstream trapezoidal section is 0.000516 m/m
• The bed slope of the downstream rectangular flume is 0.00292 m/m
• The upstream and downstream channels are concrete-lined, as will be the
transition
• In this example, the length of the transition is specified to be L = 8.0 m; in other
cases the invert elevation change, ∆z might be specified
• Both L and ∆z cannot be specified beforehand because it would unnecessarily
constrain the solution

• The base widths and uniform flow depths for the upstream and downstream
channels are shown in the figure above; these were determined during the
design procedures for the respective channels (canal & flume)
• These calculations can be confirmed by applying the Manning or Chezy
equations

• The reduction in bottom width of the channel will be accomplished with a reverse
parabola, from b = 2.5 m to b = 2.0 m
• The reduction in side slope from m = 1 to m = 0 will be done linearly across the
length L of the transition
BIE 5300/6300 Lectures 265 Gary P. Merkley
III. Confirm Subcritical Flow

• In the upstream channel, for uniform flow, the squared Froude number is:

Q 2T Q2 (b + 2mh) (15)2 (2.5 + 2(1)(1.87))

Fr2 = = = = 0.262 (1)
g [h(b + mh)] 9.81 [1.87(2.5 + (1)(1.87))]
3 3 3
gA

• In the downstream channel (flume), for uniform flow, the squared Froude number
is:

Q2 T Q2b (15)2 (2.0)

Fr2 = = = = 0.577 (2)
gA3 g [hb] 9.81 [(2.15)(2.0)]
3 3

• Therefore, Fr2 < 1.0 for both the upstream canal and downstream flume
• Then, the flow regime in the transition should also be subcritical
• It would probably also be all right if the flow were supercritical in the flume, as
long as the flow remained subcritical upstream; a hydraulic jump in the transition
would cause a problem with our given design criterion

IV. Energy Balance Across the Transition

• For uniform flow, the slope of the water surface equals the slope of the channel
bed
• Then, the slope of the upstream water surface is 0.000516, and for the
downstream water surface it is 0.00292
• Since the mean velocity is constant for uniform flow, the respective energy lines
will have the same slopes as the hydraulic grade lines (HGL), upstream and
downstream
• For our design criterion of no GVF profiles, we will make the slope of the energy
line through the transition equal to the average of the US and DS energy line
slopes:

0.000516 + 0.00292
SEL = = 0.001718 (3)
2

Gary P. Merkley 266 BIE 5300/6300 Lectures

 • This means that the total hydraulic energy loss across the transition will be:

∆E = (0.001718)(8.0) = 0.0137 m (4)

where the length of the transition was given as L = 8.0 m

• The energy balance across the transition is:

Q2 Q2
hu + + ∆z = hd + + ∆E (5)
2gA u2 2gA d2

where hu is the upstream depth (m); Q is the design flow rate (m3/s); Au is the
upstream cross-sectional flow area (m2); ∆z is the total net change in canal invert
across the transition (m); hd is the downstream depth (m); Ad is the downstream
cross-sectional flow area (m2); and ∆E is the hydraulic energy loss across the
transition (m)

• The ∆z value is unknown at this point, but the slope of the water surface across
the transition should be equal to:

hu + ∆z − hd
S ws = (6)
L
where Sws is the (constant) slope of the water surface across the transition
(m/m); and L is the length of the transition (m)

• Combining Eqs. 5 & 6:

Q2 ⎛ 1 1 ⎞
⎜⎜ 2 − 2 ⎟⎟ + ∆E
2g ⎝ A d A u ⎠
S ws = (7)
L

• For Q = 15 m3/s; Ad = (2.15)(2.0) = 4.30 m2; Au = (1.87)(2.5)+(1.0)(1.87)2 = 8.172

m2; ∆E = 0.0137 m; and L = 8.0 m:

(15)2 ⎛ 1 1 ⎞
⎜ − ⎟ + 0.0137
2(9.81) ⎝ (4.3)2 (8.172)2 ⎠
S ws = = 0.0578 (8)
8.0

• Note that Sws ≠ SEL

BIE 5300/6300 Lectures 267 Gary P. Merkley

V. Change in Side Slope

• The side slope will change linearly from 1 to 0 over the length of the transition
• The equation for m, with x = 0 at the upstream end of the transition, is:

m = 1 − 0.125 x (9)

where 0 ≤ x ≤ 8 m

VI. Change in Bed Width

• The bed width decreases from 2.5 to 2.0 m over the length of the transition
• This reduction is specified to be a reverse parabola, defined over L/2 = 4.0 m
• Specific criteria could be used to define the shape of the parabola, but a
reduction of 0.5 m in bed width over an 8.0-m distance can be accomplished in a
simpler way
• Define the bed width, b, for the first half of the transition as follows:

x2
b = 2.5 − (10)
64

where 0 ≤ x ≤ 4 m

• For x > 4 m, the equation is:

(x − 8)2
b = 2.0 + (11)
64

where 4 ≤ x ≤ 8 m

• You can also do this with a 3rd-degree polynomial:

b = Ax3 + Bx 2 + Cx + D (12)

where A, B, C, D are fitted so that the slope is zero at x = 0 and at x = 8

• By quick inspection of Eq. 12, it is seen that b = 2.5 at x = 0, so D = 2.5

• And, at x = 8, b = 2.0
• The other two conditions are that the slope equal zero at the end points:

3Ax 2 + 2Bx + C = 0 (13)

where x = 0 and x = 8

Gary P. Merkley 268 BIE 5300/6300 Lectures

 • So, C must be equal to zero, and then A and B can be determined after a small
amount of algebra: A = 0.001953125, B = -0.0234375
• The results are not identical, but very close (see the figure below)

2.50

2.45

Two parabolas
2.40
3rd-degree polynomial
2.35
Base width, b (m)

2.30

2.25

2.20

2.15

2.10

2.05

2.00
0 1 2 3 4 5 6 7 8

Distance, x (m)

VII. Change in Bed Elevation

• The change in bed elevation can be determined by setting up and solving a

differential equation, or by the known change in velocity head across the
transition
• Setting up and solving the differential equation can be done, but it is easier to
apply the velocity head, which is the difference between the energy line (EL) and
the hydraulic grade line (HGL)
• The slope of the EL is SEL = 0.001718 in the transition, and the slope of the water
surface is Sws = 0.0578
• The velocity head can be described as follows:

V2 Q2 (15)2
= + x(S ws − SEL ) = + 0.0561x (14)
2g 2gA 2 2(9.81)(8.172)2

or,

BIE 5300/6300 Lectures 269 Gary P. Merkley

 V2
= 0.172 + 0.0561x (15)
2g
• And, the cross-sectional area of flow, A, is equal to Q/V, which equals h(b+mh):

Q
A= = h(b + mh) (16)
2g ( 0.172 + 0.0561x )

where b and m are defined as functions of x in Eqs. 9, 10, 11; and 0 ≤ x ≤ 8 m

• Eq. 16 is quadratic in terms of h:

−b + b2 + 4mA
h= (17)
2m
• Use Eq. 16 to calculate A as a function of x, then insert A into Eq. 17 and solve
for h at each x value
• Using an arbitrary invert elevation of 2.0 m at the transition inlet, the relationship
between depth of water, h, and canal bed elevation, z, across the 8-m transition
is:

h = 3.87 − S ws x − z(x ) (18)

where 0 ≤ x ≤ 8 m; and z = 2.0 at x = 0

• Once h is known, use Eq. 18 to solve for z, then go to the next x value

• The graph below shows the results of calculations using the above equations
• The numerical results are shown in the table below
• Note that the sum “z+h” decreases linearly through the transition (the water
surface has a constant slope)
• Note that the velocity head increases linearly through the transition
• Note that the summation, z+h+V2/2g, in the last column of the table (to the right)
decreases linearly at the rate of 0.001718 m per meter of distance, x, as we have
specified (see Eq. 3): the energy line has a constant slope

Gary P. Merkley 270 BIE 5300/6300 Lectures

 2.5

2.0

1.5
m (m/m)
b (m)
1.0 h (m)
z (m)

0.5

0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Distance (m)

9.0 4.050

8.0 Velocity head (m)

z+h (m) 4.045
7.0
Area (m2)
Head (m) and Area (m )
2

6.0 z+h+V2/2g 4.040

5.0 z+h+V /2g

2

4.035
4.0

3.0 4.030

2.0
4.025
1.0

0.0 4.020
0.0 2.0 4.0 6.0 8.0
Distance (m)

BIE 5300/6300 Lectures 271 Gary P. Merkley

 • Note that the bed elevation, z, increases with x at first, then decreases to the final
value of 1.26 m
• Note that the cross-sectional area decreases non-linearly from 0 to 8 m, but the
inverse of the area squared increases linearly, which is why the velocity head
also increases at a linear rate
• This transition design will produce a smooth water surface for the design flow
rate of 15 m3/s, but not for any other flow rate
• Below are the transition design results using an arbitrary invert elevation of 2.00
m at the inlet to the transition
• Why would you want to have a smooth water surface for the design flow rate in
such a transition?

Gary P. Merkley 272 BIE 5300/6300 Lectures

 Transition Design Results
2 2
x m b A h V /2g z z+h+V /2g
(m) (m/m) (m) (m2) (m) (m) (m) (m)
0.0 1.000 2.500 8.165 1.87 0.17200 2.00 4.0420
0.2 0.975 2.499 7.911 1.84 0.18322 2.02 4.0417
0.4 0.950 2.498 7.680 1.82 0.19444 2.03 4.0413
0.6 0.925 2.494 7.467 1.80 0.20566 2.04 4.0410
0.8 0.900 2.490 7.272 1.78 0.21688 2.05 4.0406
1.0 0.875 2.484 7.091 1.76 0.22810 2.05 4.0403
1.2 0.850 2.478 6.922 1.75 0.23932 2.05 4.0400
1.4 0.825 2.469 6.766 1.73 0.25054 2.05 4.0396
1.6 0.800 2.460 6.619 1.72 0.26176 2.05 4.0393
1.8 0.775 2.449 6.482 1.72 0.27298 2.05 4.0389
2.0 0.750 2.438 6.352 1.71 0.28420 2.05 4.0386
2.2 0.725 2.424 6.230 1.70 0.29542 2.04 4.0383
2.4 0.700 2.410 6.115 1.70 0.30664 2.03 4.0379
2.6 0.675 2.394 6.007 1.70 0.31786 2.02 4.0376
2.8 0.650 2.378 5.903 1.70 0.32908 2.01 4.0372
3.0 0.625 2.359 5.805 1.70 0.34030 2.00 4.0369
3.2 0.600 2.340 5.712 1.70 0.35152 1.99 4.0366
3.4 0.575 2.319 5.623 1.70 0.36274 1.97 4.0362
3.6 0.550 2.298 5.538 1.71 0.37396 1.95 4.0359
3.8 0.525 2.274 5.456 1.72 0.38518 1.93 4.0355
4.0 0.500 2.250 5.379 1.73 0.39640 1.91 4.0352
4.2 0.475 2.226 5.304 1.74 0.40762 1.89 4.0349
4.4 0.450 2.203 5.233 1.75 0.41884 1.87 4.0345
4.6 0.425 2.181 5.164 1.76 0.43006 1.84 4.0342
4.8 0.400 2.160 5.098 1.78 0.44128 1.82 4.0338
5.0 0.375 2.141 5.034 1.79 0.45250 1.79 4.0335
5.2 0.350 2.123 4.973 1.81 0.46372 1.76 4.0332
5.4 0.325 2.106 4.914 1.82 0.47494 1.74 4.0328
5.6 0.300 2.090 4.857 1.84 0.48616 1.71 4.0325
5.8 0.275 2.076 4.802 1.86 0.49738 1.68 4.0321
6.0 0.250 2.063 4.748 1.88 0.50860 1.65 4.0318
6.2 0.225 2.051 4.697 1.90 0.51982 1.62 4.0315
6.4 0.200 2.040 4.647 1.92 0.53104 1.58 4.0311
6.6 0.175 2.031 4.599 1.94 0.54226 1.55 4.0308
6.8 0.150 2.023 4.552 1.96 0.55348 1.51 4.0304
7.0 0.125 2.016 4.506 1.99 0.56470 1.48 4.0301
7.2 0.100 2.010 4.462 2.02 0.57592 1.44 4.0298
7.4 0.075 2.006 4.419 2.05 0.58714 1.40 4.0294
7.6 0.050 2.003 4.378 2.08 0.59836 1.35 4.0291
7.8 0.025 2.001 4.337 2.11 0.60958 1.31 4.0287
8.0 0.000 2.000 4.298 2.15 0.62080 1.26 4.0284

BIE 5300/6300 Lectures 273 Gary P. Merkley

Gary P. Merkley 274 BIE 5300/6300 Lectures
Lecture 26
Energy Dissipation Structures

I. Introduction

• Excess energy should usually be dissipated in such a way as to avoid erosion in

unlined open channels
• In this context, “excess energy” means excess water velocity which causes
erosion and or scouring in an open channel
• Erosive damage can occur even at low flow velocities when the water is swirling,
although at a slower rate
• Energy dissipation structures and other protective infrastructure are used at
locations that are prone to erosion

II. Locations of Excess Energy

• What are the locations of excess energy in open channels?

1. Channel constrictions (such as gates, weirs, others)

2. Steep longitudinal bed slopes
3. Drops in elevation

• Energy dissipation is almost always needed downstream of supercritical flow

sections
• Energy dissipation may also be desired in lined channels
• Energy dissipation structures are typically located at:

1. Sudden drops in bed elevation

2. Downstream ends of channel branches, flumes and chutes (especially
where discharging into earthen sections)
3. Outlets of culverts and inverted siphons
4. Structures causing supercritical flow (e.g. underflow gates)
5. Structures causing downstream turbulence and eddies

III. Energy Dissipation Structure Types

• Most energy dissipation structures in open channels are based on:

1. the creation of a stable hydraulic jump

2. head-on impact on a solid, immovable obstruction

• Both of these energy-dissipation structure classes can cause significant

turbulence, reducing the hydraulic energy
• USBR publications state that the impact-type energy dissipation structures are
more “efficient” than hydraulic-jump energy dissipaters (see Chapter VI in the
Design of Small Canal Structures book by the USBR)

BIE 5300/6300 Lectures 275 Gary P. Merkley

 • Here, a more “efficient” design is defined as one which results in a smaller and or
cheaper structure for the same energy dissipation capacity
• Design dimensions for energy dissipation structures are important because an
inappropriate design can worsen an erosion and or scouring problem, as has
been manifested in the field and in laboratory experiments
• There have been cases in which the installation of an energy dissipation
structure caused more erosion than that which occurred without it

• The USBR has published design specifications for:

1. baffled apron drops

2. baffled pipe outlets
3. vertical sleeve valve stilling wells

• The vertical sleeve structure is designed for energy dissipations at pipe-to-open

channel interfaces (flow from a pipe into an open channel)
• All three of the above USBR energy dissipation structures are of the impact type
• In practice, many variations of baffled energy dissipation structures can be found

IV. Hydraulic Jumps for Energy Dissipation

• In open channels, a transition from subcritical to supercritical flow regimes results

in very little localized hydraulic energy loss
• But, the opposite transition, from supercritical to subcritical, involves a hydraulic
jump and energy loss
• The energy loss through a hydraulic jump can be significant, so jumps can be
applied to energy dissipation applications in open channels

Higher velocity Hydraulic Jump turbulence Slower velocity

Upstream Downstream

subcritical
critical depth
supercritical depth
depth

conjugate depths

Figure 1. Side view of a hydraulic jump (flow is from left to right)

• The energy loss across a hydraulic jump (upstream to downstream) is equal to

the difference in specific energy:
Gary P. Merkley 276 BIE 5300/6300 Lectures
 ∆E = Eu − Ed (1)

• Energy loss can be calculated based on measurements of depth and flow rate
• In designs, hydraulic jump energy loss is unknown, so you must apply the
momentum function to determine a conjugate depth, then apply Eq. 1
• For a given Froude number, flow rate, and upstream depth:

1. a rectangular cross section gives the least energy loss

2. a triangular cross sections gives the greatest energy loss

• Cross sections with sloping sides provide more pronounced secondary currents
(essentially orthogonal to the stream-wise direction), which also help dissipate
hydraulic energy
• Thus, hydraulic jumps in trapezoidal cross sections give energy dissipation
magnitudes somewhere between the extremes of rectangular and triangular
cross-sectional shapes

• Some important hydraulic jump parameters, such as jump length and location,
are determined experimentally, not theoretically
• Thus, design procedures for hydraulic jump energy dissipaters always include
empirical equations
• The length of the “roller,” Lr, is always less than the length of the jump, Lj

Figure 2. Another side view of a hydraulic jump (flow is from left to right)

• Small (weak) hydraulic jumps do not have a roller

• The length of a hydraulic jump in a rectangular cross section can be
approximated by the following function:

1.01
L j ≈ 9.75hu (Fus − 1.0 ) (2)

where Fus is the Froude number on the upstream side of the jump

• There are several classifications for hydraulic jumps

BIE 5300/6300 Lectures 277 Gary P. Merkley

 • Procedures exist to determine the type of jump that might occur in a given
situation
• One classification groups hydraulic jumps into types “A” to “F”

V. Drop Spillways

• Drop spillways (also known as “drop structures”) are abrupt decreases in channel
bed elevation with a downstream stilling basin, used to dissipate hydraulic energy
in open channels
• Drop spillways often combine both hydraulic jump and impact features, although
not all design situations are associated with a hydraulic jump
• Much research and experimentation has been done on drop spillways in efforts
to adequately define design procedures and parameters
• Part of the reason for this is that, when incorrectly dimensioned, drop spillways
can actually worsen an erosion problem in the downstream channel
• Most drop spillways have the following basic features:

1. Inlet section
2. Drop section
3. Rectangular stilling basin
4. Outlet section

• The flow through a drop spillway:

1. Spills over a crest at a vertical headwall

2. Falls on a horizontal (level) apron
3. Impinges on floor blocks inside the basin
4. Exits the stilling basin over an end sill

• Energy dissipation occurs via:

1. Floor blocks
2. End sill at DS of basin
3. Turbulence in the “tail water”
4. Hydraulic jump (in some cases)

• The following drop structure design elements are adapted principally from
Donnelly & Blaisdell (1965) and involve mostly empirically-determined
relationships

Stilling Basin Length

• How long does the stilling basin need to be for effective energy dissipation?
• According to experimental results, a series of simple iterative calculations are
needed to answer this question
• Base dimensions on a design discharge and critical depth in a rectangular basin:

Gary P. Merkley 278 BIE 5300/6300 Lectures

 2
hc =3
( Q / b ) (3)
g

where hc is the critical depth of water in a rectangular open-channel section (m or

ft); Q is the flow rate (m3/s or cfs); b is the channel base width (m or ft); and g is
the ratio of weight to mass (9.81 m/s2 or 32.2 ft/s2)

• In the present context, b represents the width of the stilling basin

• Note that Eq. 3 is based on the squared Froude number, set equal to unity
• Critical depth, hc, may or may not actually occur in the stilling basin (if it does not,
there will be no hydraulic jump), but in any case the value of hc is still used in the
following design calculations

Where the Nappe Hits the Floor

• Consider the following figure where flow goes from left to right (note that the
coordinate origin is located at the brink of the overfall):

spillway crest
y
x
ht+ydrop
xt
ydrop
headwall

ht
su
free

bm

basin floor

xf
xa
xs
Figure 3. Side view of a drop spillway showing the free and submerged nappes
(flow is from left to right)

BIE 5300/6300 Lectures 279 Gary P. Merkley

 • This is the equation for the “free nappe” is:

xf ⎛ ydrop ⎞
= −0.406 + 3.195 − 4.386 ⎜ ⎟ (4)
hc ⎝ hc ⎠

where hc is as defined in Eq. 3; and the other variables are defined in Fig. 3

• Note that ydrop < 0, and hc > 0, in all cases

• This means that the ratio ydrop/hc is always negative
• Thus, xf increases with increasing absolute magnitude of ydrop
• Note also that xf defines the upper nappe surface
• Each of the terms in Eq. 4 are dimensionless

• This is the equation for the “submerged nappe:”

2
⎛x ⎞ ydrop
0.691 + 0.228 ⎜ t ⎟ −
xs
= ⎝ hc ⎠ hc
(5)
hc ⎛x ⎞
0.185 + 0.456 ⎜ t ⎟
⎝ hc ⎠

again, where hc is the critical depth, as defined in Eq. 3; xt is defined by Eq. 6;

and the other variables are defined in Fig. 3

• The variable xt is the distance to where the upper nappe surface plunges into the
tail water
• The nappe plunge location, xt, is defined by an equation which is similar to Eq. 4
for the free nappe:

xt ⎛ ht + ydrop ⎞
= −0.406 + 3.195 − 4.386 ⎜ ⎟ (6)
hc ⎝ hc ⎠

where ht is the tail water depth in the stilling basin, as seen in Fig. 3, and is
referenced to the stilling basin floor

• The term in parentheses in Eq. 6 will be positive in those cases in which the tail
water is above the spillway crest
• To avoid a negative square root term in Eq. 6, limit (ht + ydrop)/hc to a maximum of
0.7 when applying Eq. 6
• This is not a significant restriction because the required stilling basin length is not
affected when:

Gary P. Merkley 280 BIE 5300/6300 Lectures

 ht + ydrop
> 0.67 (7)
hc

• All water depths (including hc and ht) are greater than zero
• All “x” values downstream of the spillway crest are greater than zero
• But all “y” values are negative below the spillway crest, positive above (this
follows the convention introduced by Donnelly and Blaisdell), as seen in Fig. 3

• The average of the results from Eqs. 4 and 5 are used for drop structure design:

xa =
( x f + xs ) (8)
2
where the value of xa is can be determined mathematically (preferred) or
graphically, as shown in the following plot (Fig. 4) of the above equations

• The stilling basin length, L, will always be greater than xa (L > xa)

ht + ydrop = 0.7 0.65 0.6 0.55 0.5 0.4 0.3 0.2 0.1 0.0
10

-0.2
9
-0.5

8 -1.0

-2.0
7
-4.0
-8.0
6
xa/h c

1
0 1 2 3 4 5 6 7 8 9 10
-ydrop/hc

Figure 4. Plot of drop spillway design equations for determining the value of xa

BIE 5300/6300 Lectures 281 Gary P. Merkley

 Floor Blocks

• Floor blocks are usually included in drop structure designs to help dissipate
hydraulic energy before the flow exits the stilling basin
• There is a required minimum distance from xa to the blocks so the flow becomes
parallel to the floor before impinging on the upstream face of the blocks
• If the blocks are too close to the location of xa, water splashes (“boils”) off the
blocks, and may go over the sides of the stilling basin

floor
blocks
ydrop

xa xb
Figure 5. Side view of a drop spillway showing the recommended location of floor
blocks (flow is from left to right)

• If xb < ½hc, the floor blocks are mostly ineffective in terms of energy dissipation
• Thus, for stilling basin design, let

xb = 0.8hc (9)

• The recommended height of the floor blocks is 0.8 hc

• The recommended length of the floor blocks is 0.5 to 0.75 hc
• The recommended width of the floor blocks is also 0.5 to 0.75 hc
• Usually, make the floor blocks square (length = width)
• The upstream faces of the floor blocks should occupy from 50 to 60% of the
basin width for effective energy dissipation
• Use equal spacing of floor blocks across the width of the stilling basin, but make
slight adjustments as necessary to accommodate the total width, b

Longitudinal Sills

• Longitudinal sills are sometimes placed on the floor of the stilling basin, parallel
to the basin walls, as seen in a plan-view (Fig. 6)

Gary P. Merkley 282 BIE 5300/6300 Lectures

 1
1

longitudinal sill

end sill
b
longitudinal sill

w
in
gw
L

al
l
Figure 6. Plan view of a drop spillway showing longitudinal sills and square floor
blocks (flow is from left to right)

• These sills are unnecessary if the floor blocks are properly:

1. Proportioned
2. Spaced

• Longitudinal sills are sometimes included in a design for structural reasons

• If they are included, they should pass through (not between) the floor blocks, as
shown in Fig. 6

End Sill Location

• There is a minimum distance from the floor blocks to the end sill, which is located
at the downstream end of the stilling basin
• This minimum distance is intended to maximize the energy dissipation from both
the floor blocks and the end sill
• For design purposes, let:

x c ≥ 1.75hc (10)

where xc is defined in Fig. 7

BIE 5300/6300 Lectures 283 Gary P. Merkley
 floor
blocks end
ydrop sill

xa xb xc
stilling basin length, L
Figure 7. Side view of a drop spillway showing the location of the end sill and the
total basin length (flow is from left to right)

• However, in most design cases, xc is set equal to 1.75 hc

• In other cases, it may be necessary to provide a longer stilling basin length to
accommodate the site-specific conditions

Stilling Basin Length

• In summary, the stilling basin length is:

L = x a + xb + x c (11)

or,
L = x a + 2.55hc (12)

End Sill Height

• The end sill height is:

yend = 0.4 hc (13)

where yend is the end sill height, as shown in Fig. 8

• Observe that yend > 0 in all cases

Gary P. Merkley 284 BIE 5300/6300 Lectures

 ydrop yend

L
Figure 8. Side view of a drop spillway showing the height of the end sill

• The top of the end sill should be at or slightly above the invert (bottom) elevation
of the downstream channel (or downstream channel transition), as shown in the
following figure

floor invert of
blocks DS channel
end sill
ht
ydrop hds

L
Figure 9. Side view of a drop spillway showing the downstream channel invert

Tail Water Depth

• A minimum tail water depth is required in the design of a drop spillway

• To prevent downstream scouring, the tail water depth should be “about the
same” as the depth in the stilling basin
• If this is true, the hydraulic jump is submerged inside the basin length
• For design, let

ht ≥ 2.15hc (14)

where ht is from the downstream water surface to the stilling basin floor, as seen
in Fig. 9
BIE 5300/6300 Lectures 285 Gary P. Merkley
 • In most drop spillway designs, let

ht = 2.15hc (15)

• Note that the recommended ratio of ht/hc (= 2.15) is independent of the drop
height, ydrop
• There may be a hydraulic jump up to the tail water depth, in some cases

• If the tail water depth, ht, is too low (i.e. ht < 2.15 hc)

1. Increase the stilling basin width, b, which will decrease hc; or,
2. Increase |ydrop|, deepening the stilling basin floor

• An increase in |ydrop| and or b may increase construction and maintenance costs

• An increase in |ydrop| also increases the end sill height

• Note that the depth from the spillway crest to the stilling basin floor can be
increased not only by deepening the basin floor, but also by providing a weir at
the overfall location
• This solution can be convenient for the drop structure design, but care must be
taken with the freeboard in the upstream channel because increasing the
spillway crest height will result in a corresponding upstream water depth increase

• How to determine the value of tail water depth, ht?

• If uniform flow conditions prevail in the downstream channel, use the Manning or
Chezy equation to calculate hds
• Otherwise, apply gradually-varied flow analysis for subcritical conditions to
determine ht
• Thus, hds is calculated independently of the drop structure dimensions
• Finally,

ht = hds + yend (16)

Side and Wing Walls

• The tops of the sidewalls should be at least 0.85dc above the tail water surface
• Wing walls are DS of the end sill, at 45º angle, and with a top slope of 1:1
• Wing wall length depends on the width of the DS channel section
• Wing walls are not necessary if the DS channel is a lined canal

Gary P. Merkley 286 BIE 5300/6300 Lectures

 upp 0.85hc
er
na
pp
1

e
headwall

su
1

rf a
ht

ce
0.4hc wingwall

xa xb xc
L
Figure 10. Side view of a drop spillway showing footings and wing walls

Drop Spillway Construction

• Construction is usually of steel-reinforced concrete

• The basin floor should be level, both longitudinally and transversely
• Upstream and or downstream channel transitions may be needed
• Concrete floor and wall thickness is usually 5-8 inches (12-20 cm)
• The depth of the concrete footings should be 2-3 ft for most designs in small- and
medium-size channels
• May need riprap or other form of erosion protection upstream and downstream of
the drop structure where earthen channels exist
• The approach channel bed elevation should be the same as the spillway crest
elevation at the headwall
• The required headwall height at the crest location depends on the expected
upstream depth at the design discharge, plus freeboard
• The side walls slope down from the top of the headwall to the top of the wing
walls at the end sill location (see Fig. 10)
• In some cases it is convenient and appropriate to make the stilling basin width, b,
equal to the width of the upstream or downstream channel (may eliminate the
need for transitions)

VI. Drop Spillway Design Procedure

• The best design procedure depends on the given site conditions and
requirements for a particular location
• However, in general, the following procedure can be applied

1. Define the total available bed elevation change at the proposed drop
structure location.
2. Define the design discharge, Q.

BIE 5300/6300 Lectures 287 Gary P. Merkley

 3. Calculate hds based on the downstream channel conditions (cross
section, bed slope, roughness) using a uniform-flow equation, or the
gradually-varied flow equation, as appropriate.
4. Choose a reasonable value for the stilling basin width, b.
5. Calculate critical depth in the stilling basin, hc (Eq. 3).
6. Calculate yend (Eq. 13).
7. Calculate ht (Eq. 16).
8. Is Eq. 14 satisfied? If not, use Eq. 14 to recalculate the stilling basin
width, b, then go back to Step 5.
9. Calculate ydrop based on the total available bed elevation change and
yend (see Fig. 9), where ydrop should be less than zero. If ydrop ≥ 0,
consider raising the spillway crest by including a weir.
10. Calculate xf (Eq. 4).
11. Calculate xt (Eq. 6).
12. Calculate xs (Eq. 5).
13. Calculate xa (Eq. 8).
14. Calculate xb (Eq. 9).
15. Calculate xc (Eq. 10).
16. Calculate the stilling basin length, L (Eq. 11). If the length is not
acceptable, adjust b and go back to Step 5.
17. Calculate the floor-block dimensions and spacing.
18. Calculate the head wall height based on the upstream depth at Qmax,
plus freeboard
19. Calculate the height of the wing walls at the end sill (0.85 hc).
20. Prepare side view and plan view drawings of the drop spillway structure.

VII. Example Drop Spillway Design

Given:

• The design flow rate is Qmax = 9.0 m3/s

• There is a drop of 2.25 m in channel bed invert at this location
• The upstream channel is earthen, as is the downstream channel
• The upstream channel has a base width of approximately 5 m
• The downstream channel has an approximately trapezoidal cross section: base
width is b = 5 m, side slopes have m = 1.64, and the bed slope is So = 0.000112
• For the downstream channel, use a Manning n value of 0.019
• The depth in the downstream channel is at the uniform flow depth at Qmax

Solution:

1. The total available bed elevation change is given as 2.25 m.

2. The design discharge is given as 9 m3/s
3. Uniform flow conditions are expected in the downstream channel. Using the
ACA program, the normal depth in the downstream channel is 1.80 m at the
design capacity of 9 m3/s, with a Manning roughness of n = 0.019.
4. Try a stilling basin width of b = 5 m, matching the upstream channel base width
Gary P. Merkley 288 BIE 5300/6300 Lectures
 5. Critical depth in the stilling basin (Eq. 3):

(9 / 5)
2

hc = 3
= 0.691 m
9.81

6. The end sill height will be (Eq. 13):

y end = 0.4 ( 0.691) = 0.276 m

7. The tail water depth will be (Eq. 16):

ht = 1.80 + 0.276 = 2.076 m

8. Check to see if Eq. 14 is satisfied:

ht = 2.076 m
2.15hc = 2.15 ( 0.691) = 1.486 m
Thus,
ht > 2.15hc

and Eq. 14 is satisfied.

9. The value of ydrop will be:

y drop = −2.25 − 0.276 = −2.526 m

Notice that ydrop is negative (as required).

10. Calculate xf (Eq. 4):

⎡ ⎛ −2.526 ⎞ ⎤
x f = 0.691⎢ −0.406 + 3.195 − 4.386 ⎜ ⎟ ⎥ = 2.75 m
⎣⎢ ⎝ 0.691 ⎠ ⎦⎥

11. Calculate xt (Eq. 6):

⎡ ⎛ 2.076 − 2.526 ⎞ ⎤
x t = 0.691⎢ −0.406 + 3.195 − 4.386 ⎜ ⎟ ⎥ = 1.42 m
⎢⎣ ⎝ 0.691 ⎠ ⎥⎦

12. Calculate xs (Eq. 5):

BIE 5300/6300 Lectures 289 Gary P. Merkley

 ⎡ ⎛ 1.42 ⎞ ⎛ 2.526 ⎞ ⎤
⎢ 0.691 + 0.228 ⎜ 0.691⎟ + ⎜ 0.691 ⎟ ⎥
x s = 0.691⎢ ⎝ ⎠ ⎝ ⎠ ⎥ = 3.27 m
⎢ ⎛ 1.42 ⎞ ⎥
0.185 + 0.456 ⎜ ⎟
⎢⎣ ⎝ 0.691 ⎠ ⎥
⎦

13. Calculate xa (Eq. 8):

xa =
( 2.75 + 3.27 ) = 3.01 m
2
14. Calculate xb (Eq. 9):

x b = 0.8 ( 0.691) = 0.553 m

15. Calculate xc (Eq. 10):

x c = 1.75 ( 0.691) = 1.209 m

16. Calculate the stilling basin length (Eq. 11):

L = x a + x b + x c = 4.77 m

Notice that L < b in this design.

17. Floor block dimensions and spacing:

Floor block height: 0.8 hc = 0.8 (0.691) = 0.55 m

Floor block width: 0.5 hc = 0.8 (0.691) = 0.35 m
Floor block length: 0.5 hc = 0.8 (0.691) = 0.35 m

At 50% basin width, the required number of floor blocks is:

0.5b 2.5
N= = = 7.1 blocks
0.35 0.35
Round up to N = 8 blocks, giving a percent area of 56%. Placing a block against
each side wall of the stilling basin, the uniform spacing between the blocks will
be:
b − 0.35N
spacing = = 0.314 m
N −1
18. The height of the headwall, from the basin floor to the top, should be:

Gary P. Merkley 290 BIE 5300/6300 Lectures

 − y drop + 1.1(hds ) = 2.526 + 1.1(1.80) ≈ 4.50 m

where the coefficient 1.1 is to allow for freeboard.

19. Wing wall height at end sill:

0.85hc = 0.85(0.691) = 0.587 m

20. Design notes:

• Complete the design by specifying wall & floor thickness

• Specify the depth of footings (see Fig. 10)
• Specify the length of the wing walls (they should be at least long
enough to meet the side slopes of the downstream channel)
• Make design drawings (side and plan views)
• A more iterative design approach could be used to minimize the
size (b x L) of the drop spillway, thereby reducing its cost

References & Bibliography

Aletraris, S.S. 1983. Energy dissipation parameters for small vertical drop structures. Unpublished
M.S. thesis, BIE department, Utah State Univ., Logan, UT.
Donnelly, C.A., and Blaisdell, F.W. 1965. Straight drop spillway stilling basin. ASCE J. Hydraulics
Div., HY 3:101-131 (May 1965).
Peterka, A.J. 1964. Hydraulic design of stilling basin and energy dissipaters. Engr. Monograph 25.
USBR, Denver, CO. (September).
Rand, W. 1955. Flow geometry at straight drop spillways. ASCE J. Hydraulics Div., 81:1-13.
Schwartz, H.I., and Nutt, L.P. 1963. Projected nappes subject to transverse pressure. ASCE J.
Hydraulics Div., 89(HY4):97-104.
White, M.P. 1943. Energy loss at the base of a free overfall. ASCE Trans., 108:1361-1364.
(discussion of paper 2204).

BIE 5300/6300 Lectures 291 Gary P. Merkley

Gary P. Merkley 292 BIE 5300/6300 Lectures
Lecture 27
Protective Structures

I. Introduction

• Protective structures include energy dissipation and erosion control structures,

structures to divert excess water to prevent over-topping of canals, and others
• In this context, protective structures are for the canals and other infrastructure,
and not for the protection of animals and people
• For example, there should be a wasteway weir or other structure in a canal,
upstream of an in-line pumping plant -- the pump(s) could shut off unexpectedly,
possibly causing over-topping of the canal upstream
• An inverted siphon could become clogged, or a landslide could block the flow in
the canal, also causing the canal to overflow on the upstream side
• Of course, canal over-topping can also occur due to design deficiencies,
construction problems, and operational error
• This means that spillway structures are generally designed to carry the full flow of
the canal
• Side spillways sometimes have radial or vertical slide gates to facilitate
dewatering of the channel in emergencies, for sediment removal, and other
reasons
• Some in-line gate structures in canals have fixed-crest side weirs to allow for
water to pass downstream in the event of an operational error
• The flow from spillway structures is generally directed into a natural channel
which can safely carry the maximum spill rate away from the canal

II. Wasteway Weirs

• A wasteway weir is a sharp- or blunt-crested weir located along one bank of the
canal

BIE 5300/6300 Lectures 293 Gary P. Merkley

 • The weir may be designed to accept “stop logs” to allow for changes in the
operating level of the canal
• The crest of the weir is equal to the maximum operating level of the canal, taking
into consideration that a higher water surface elevation will result in the canal for
the full design discharge
• The wider the weir is, the less difference in water surface elevation between
impending spill and full-flow spill through the structure
• Wasteway weirs are located at places where the canal would be most likely to
overtop in the event of an operational error or the clogging of a flow control
structure (e.g. cross regulator):

1. Upstream of an inverted siphon entrance

2. Upstream of a gate or weir structure
3. Upstream of a pumping station

III. Siphon Spillways

• As opposed to inverted siphons, siphon spillways operate under negative

pressures (below atmospheric)
• Once a siphon spillway is primed, water will continue to flow through the structure
as long as the downstream water elevation is lower than that in the canal, or until
suction is broken by other means
• Siphon spillways are generally more expensive to build than side spillway weirs,
but for the same discharge they require a much smaller width
• Siphon spillways can discharge more water than a weir for a small increase in
upstream water surface elevation
• As the water level in the canal increases, the siphon spillway acts like a weir
• As the water level continues to increase, the siphon will become “primed” and
operate under full pipe flow conditions
• The USBR design calls for a crest elevation of about 0.2 ft above the normal
water surface (or full supply level) in the canal
• Note that the inlet to the siphon can serve as a sediment trap, requiring periodic
manual cleaning
• Also note that it is impractical to change the FSL of the canal once the siphon
spillway is installed – with side spill weirs, you can always raise the level, if
needed
• See Section 4-14 of the USBR Small Canal Structures book

Gary P. Merkley 294 BIE 5300/6300 Lectures

BIE 5300/6300 Lectures 295 Gary P. Merkley
 Air Vent for Breaking Suction

• An air vent is located just downstream of the top of the siphon spillway to break
suction when the upstream water level in the canal drops below a certain level
• This location is about 15° (from vertical) downstream of the top of the siphon
pipe, as shown in Figure 4-17 of the USBR Small Canal Structures book
• The upstream end of the vent is open at the normal water surface level (or FSL)
of the canal
• A potential operating problem with this type of structure is that when the suction
is broken, the discharge will suddenly cease, and this can cause surges in the
canal
• A pan can be attached to the upstream end of the vent to help prevent the vent
from acting as a siphon itself, possibly causing the water level in the canal to
drop below full supply level
• The pan helps reduce the amount of water level fluctuation in the canal

w
flo

FSL

Gary P. Merkley 296 BIE 5300/6300 Lectures

 15o

BIE 5300/6300 Lectures

12.00 ft

road

FSL
1.5:1
1.25:1 ft
2
0.
6.36 ft

6.16 ft
30o
ft high
2.00 ft
00 water

297
4.

4.0
2.00 ft 0f
t
sill
8.30 ft

3.85 ft 60o 1:1

3.0 ft

2.00 ft
5.00 ft 10.00 ft
Barrel ll all
f wa f w
utof tof
c cu

Gary P. Merkley
 Deflector at the Downstream Side of the Siphon

• A small angled deflector can be installed in the downstream end of the siphon to
help direct water up to the top of the barrel under non-full-flow conditions
• This helps to mix air and water and cause the siphon to prime to full flow quicker
• The roof of the siphon structure at the outlet should be above the expected
downstream water surface elevation
• This helps to evacuate air from the siphon

IV. Design of a Siphon Spillway

This design example is adapted from an example given by the USBR (1978)

Given:

Suppose there is a canal with a design discharge of 120 cfs in which an in-line pump
station is used to lift the water up to a downstream reach. The canal is trapezoidal in
cross-section, with a base width of 8.0 ft, side slope of 1½:1 (H:V).
The canal is at an elevation of about 6000 ft above msl. The available head across
the siphon spillway, H, is 6.0 ft.

Solution:

(a) Preliminary Calculations According to USBR recommendations, use a ratio RCL/D

of 2.0. Assume an initial value of D = 2.0 ft. Then, the radius of the centerline of the
siphon is RCL = 4.0 ft, and the radius of the siphon invert is RC = 4.0 ft - ½(2.0) = 3.0
ft. The radius of the top of the siphon is RS = 4.0 ft + ½(2.0) = 5.0 ft.

(b) Full Pipe Discharge Estimate the full pipe discharge by assuming: (1) orifice flow;
and (2) a discharge coefficient of 0.65. This will give a flow rate per “unit width” of
the barrel:

q = CdD 2gH = (0.65)(2.0) 2(32.2)(6.0) = 25.6 cfs / ft (1)

Note that this is an estimate, using an assumed Cd value, and D instead of area.

(c) Maximum Possible Discharge Now, estimate the discharge per unit width
according to the “vortex” equation, which takes into account the atmospheric
pressure available to “push” the water up over the invert of the siphon crest. The so-
called vortex equation looks like this:

⎛R ⎞
qmax = RC 2g(0.7h) ln ⎜ S ⎟ (2)
⎝ RC ⎠
where h is the available atmospheric pressure head, based on the density of water;
and qmax is the flow rate per foot of barrel width (cfs/ft).
Gary P. Merkley 298 BIE 5300/6300 Lectures
 The change in average atmospheric pressure with elevation can be approximated by
the following linear relationship:

Patm ≈ (33.9 − 0.00105Elev) / 2.31 (3)

where Patm is in psi and Elev is the elevation above mean sea level, in ft. For 6000-ft
of elevation, the atmospheric pressure is about 11.9 psi, or h = 27.6 ft of head
(water).

Then, qmax is equal to:

⎛ 5.0 ⎞
qmax = (3.0) 2g(0.7)(27.6) ln ⎜ ⎟ = 54.1 cfs / ft (4)
⎝ 3.0 ⎠
This means that the previously-calculated unit discharge of 25.6 cfs/ft (from the
orifice equation) is acceptable. If q were greater than qmax, it would have been
necessary to decrease H or change RCL.

(d) Barrel Width The width of the barrel is determined as:

Q 120 cfs
b= = = 4.7 ft (5)
q 25.6 cfs / ft

This value could be rounded up to provide a margin of safety, but we will leave it at
4.7 ft (at least for now).

(e) Vent Diameter The diameter of the siphon breaker pipe, Dp, should be such that
the cross-sectional area is at least 1/24th of the cross-sectional area of the barrel
(according to USBR guidelines). This gives an area of (2.0)(4.7)/24 = 0.39 ft2. The
corresponding ID is 0.70 ft, or 8.5 inches. Thus, use whatever pipe size would be
closest to this diameter (perhaps 9-inch nominal size), noting that steel pipe is
usually used (for strength).

(f) Outlet Sill Height The height of the deflector sill at the outlet of the siphon is given
as 1.5D, or 3.0 ft in our case.

(g) Outlet Ceiling Height The ceiling of the outlet is defined as h2. Referring to Figure
4-17, this is given by:

h2 = 1.5D + Ecritical + 1.0 ft (6)

where Ecritical is the specific energy for critical flow conditions, in feet. This is how the
dimensions are defined in the USBR design procedures.

BIE 5300/6300 Lectures 299 Gary P. Merkley

 The width of the downstream pool will be the same as the barrel width, or 4.7 ft, and
the section will be rectangular. Critical depth for the design discharge is:

Q2 (120)2
yc = 3 2
=3 2
= 2.7 ft (7)
gb (32.2)(4.7)

The velocity is Q/A = 120/(2.7*4.7) = 9.5 fps, so the velocity head is:

V2 (9.5)2
= = 1.4 ft (8)
2g 2(32.2)

Then, Ecritical is 2.7 + 1.4 = 4.1 ft. And, h2 = 1.5(2.0) + 4.1 + 1.0 = 8.1 ft.

(h) Other Design Details The inlet structure from the canal can be designed with a
height along the side slope of 2D (minimum). The inlet structure should provide a
minimum submergence of 1.5hv + 0.5 ft, where hv is the velocity at the inlet, and the
inlet area is at least 2Db.

Gary P. Merkley 300 BIE 5300/6300 Lectures

Lecture 28
Safety Considerations
“Here lies one whose name was writ in water”
John Keats (1821)
I. Introduction

• Canals and related infrastructure can be very dangerous to people and animals
• People drown in canals, inverted siphons and other facilities every year
• One of the most important considerations is the number of people that might be
exposed to dangerous facilities (canals, siphons, etc.) at a given site
• It is difficult to determine generally applicable design standards for safety
features because of many factors that should be considered
• Note that design engineers can be held legally liable for mishaps & accidents

II. USBR Hazard Classifications

• The kind of safety protection applied to a given canal and canal structures
normally depends on the safety classification:

Class A Canals nearby or adjacent to schools and recreational areas, or where

children are often present
Class B Canals nearby or adjacent to urban areas, county roads or highways
that would have frequent public access or recreational use
Class C Canals nearby or adjacent to farms, county roads or highways that
would have a possibility for children to occasionally be present
Class D Canals far from roads and houses that would usually not be visited by
the public
Class E Canals that might be a hazard to domestic animals
Class F Canals that would be very hazardous to large game animals

BIE 5300/6300 Lectures 301 Gary P. Merkley

III. Safety Devices for Canals

1. Preventative

• Fencing
• Sign Posting
• Guard Railings and other Barriers

2. Escape Devices (usually only upstream of the hazardous location)

• Safety Nets
• Ladders
• Cables with Floats
• Pipe Inlet Racks

Gary P. Merkley 302 BIE 5300/6300 Lectures

BIE 5300/6300 Lectures 303 Gary P. Merkley
Gary P. Merkley 304 BIE 5300/6300 Lectures
 BIE 5300/6300
Fall Semester 2004
Exam #1 – 19 Oct 04

Show your work neatly on this or separate pages. Show units for
all calculated values. Make note of any important assumptions.

Name_______________________________

1. (20 pts) A Cutthroat flume needs to be installed in an irrigation canal at a location

with a maximum flow rate of 50 cfs. The canal is rectangular in cross section, with a
base width of 6.0 ft, and a longitudinal bed slope of 0.00065 ft/ft. Normal depth at
the maximum flow rate is 2.40 ft.

(a) Select a standard Cutthroat flume size (specify L & W in feet).

(b) Will the canal cross section need to be modified to
accommodate the flume? Be specific.
(c) Determine the minimum height (in feet) of the flume floor with
respect to the channel bed such that the flume operates under
free-flow conditions at Qmax.

2. (15 pts) Current metering data is given below for an open channel with a top width
of 3.03 m. Complete the calculations to estimate the total flow rate in m3/s.

distance velocity (m/s)

from edge depth depth mean in mean in mean width area flow rate
at point
(m) (m) fraction vertical section depth (m) (m) (m2) (m3/s)
0.05 0.00 n/a 10% 0.019
0.105 0.090 0.400 0.036 0.0038
0.45 0.18 0.6 0.190 0.190
0.201 0.375 0.800 0.300 0.0602
1.25 0.57 0.2 0.208 0.211
0.8 0.214 0.215 0.700 0.500 0.350 0.0751
1.75 0.83 0.2 0.211 0.218
0.8 0.225 0.221 0.900 0.250 0.225
2.00 0.97 0.2 0.218 0.224
0.8 0.229 0.220 1.010 0.500 0.505
2.50 1.05 0.2 0.210 0.216
0.8 0.222 0.213 1.050 0.500
3.00 1.05 0.2 0.208 0.210
0.8 0.212
3.08 1.05 n/a

Totals:

1 of 7
3. (15 pts) You need to design a suppressed, rectangular sharp-crested weir for a
rectangular channel with a base width of 2.10 m, and a Qmax of 4.5 m3/s. A normal
depth of 0.678 m at Qmax was measured in the field. Manning roughness is 0.014.
Design the weir, specifying crest height, P, provided the conditions are appropriate.

4. (25 pts) A BCW, operating under free-flow conditions, has a flow rate of 17.00 m3/s
when hu (referenced from the sill) is 1.813 m. At this same flow rate, hd (referenced
from the downstream bed elevation) is 2.661 m. The upstream and throat sections
are rectangular with a width of 3.75 m. The sill height is zu = zd = 1.15 m. This BCW
has a downstream ramp with a 6:1 slope.

(a) Determine the upstream specific energy, Eu.

(b) Determine hc (referenced from the sill) for this flow rate.
(c) Estimate the hydraulic head loss in the diverging section, (hf)ds.

5. (25 pts) A circular pipe has a circular, sharp-edged orifice plate (D2 = 5.50 inches),
centered in the pipe cross section, which has an inside diameter of D1 = 7.90 inches.
Taps at D1 (upstream) and ½D1 (downstream) are connected to a manometer with
mercury, from which a head differential of 186 mm is measured. Water temperature
is 13°C. Ignoring thermal expansion adjustments:

(a) Determine the ratio β.

(b) Determine the kinematic viscosity, ν.
(c) Determine the discharge coefficient, Cd.
(d) Determine the upstream Reynolds number, Re.
(e) Determine the flow rate, Q.

2 of 7
Solutions:

1. (20 pts) A Cutthroat flume needs to be installed in an irrigation canal at a location

with a maximum flow rate of 50 cfs. The canal is rectangular in cross section, with
a base width of 6.0 ft, and a longitudinal bed slope of 0.00065 ft/ft. Normal depth
at the maximum flow rate is 2.40 ft.

(a) Select a standard Cutthroat flume size (specify L & W in feet).

For 50 cfs, select the W = 3 ft, L = 9 ft Cutthroat flume from the table in the
lecture notes, with Qmax = 56.9 cfs.

(b) Will the canal cross section need to be modified to

accommodate the flume? Be specific.

B = W + L/4.5 = 3 + 9/4.5 = 5 ft. Thus, no modifications to the channel

section (which is 6-ft wide) except for the inclusion of headwalls (US & DS) to
direct all flow through the flume.

(c) Determine the minimum height (in feet) of the flume floor with
respect to the channel bed such that the flume operates under
free-flow conditions at Qmax.

From Cutthroat flume table in lecture notes: St = 0.820, nf = 1.55, and Cf =

3.442 (ft & cfs) for the selected flume size. At 50 cfs:

1/ nf 1/1.55
⎛ Q ⎞ ⎛ 50.0 ⎞
hu = ⎜ f ⎟ =⎜ ⎟ = 2.77 ft
⎝ Cf W ⎠ ⎝ (3.442)(3.0) ⎠

Then,
(hd )max = Sthu = (0.820)(2.77) = 2.27 ft
Finally, the minimum floor height with respect to the channel bed is:

(hfloor )min = 2.40 − 2.27 = 0.13 ft

US Cutthroat flume DS
2.77 ft 2.27 ft
2.40 ft

0.13 ft

3 of 7
2. (15 pts) Current metering data is given below for an open channel with a top width
of 3.03 m. Complete the calculations to estimate the total flow rate in m3/s.

Note that there is a vertical wall at the 3.08 distance from the edge
(otherwise, the depth would be zero at this location). The ratio x/D is:

x (3.08 − 3.00)
= = 0.0762
D 1.05
Then,

Vx 0.65 + 10.52(0.0762)
= = 0.802
VD 1 + 10.676(0.0762) − 0.51431(0.0762)2

and,
0.65(0.210)
Vw = = 0.170 m/s
0.802
The rest of the calculations are simple and are given in the table below.

distance velocity (m/s)

from edge depth depth mean in mean in mean width area flow rate
at point
(m) (m) fraction vertical section depth (m) (m) (m2) (m3/s)
0.05 0.00 n/a 10% 0.019
0.105 0.090 0.400 0.036 0.0038
0.45 0.18 0.6 0.190 0.190
0.201 0.375 0.800 0.300 0.0602
1.25 0.57 0.2 0.208 0.211
0.8 0.214 0.215 0.700 0.500 0.350 0.0751
1.75 0.83 0.2 0.211 0.218
0.8 0.225 0.221 0.900 0.250 0.225 0.0497
2.00 0.97 0.2 0.218 0.224
0.8 0.229 0.220 1.010 0.500 0.505 0.1110
2.50 1.05 0.2 0.210 0.216
0.8 0.222 0.213 1.050 0.500 0.525 0.1118
3.00 1.05 0.2 0.208 0.210
0.8 0.212 0.190 1.050 0.080 0.084 0.0160
3.08 1.05 n/a 0.170

Totals: 3.030 2.025 0.427

4 of 7
3. (15 pts) You need to design a suppressed, rectangular sharp-crested weir for a
rectangular channel with a base width of 2.10 m, and a Qmax of 4.5 m3/s. A normal
depth of 0.678 m at Qmax was measured in the field. Manning roughness is 0.014.
Design the weir, specifying crest height, P, provided the conditions are appropriate.

Check the Froude number in this channel at Qmax.

Q2 T (4.5)2 (2.10)
F =
2
= = 1.50
gA 3 (9.81) [(2.10)(0.678)]3
r

Thus, the regime is supercritical, so it is suggested that this is not a good

location for a measurement weir.

4. (25 pts) A BCW, operating under free-flow conditions, has a flow rate of 17.00
m3/s when hu (referenced from the sill) is 1.813 m. At this same flow rate, hd
(referenced from the downstream bed elevation) is 2.661 m. The upstream and
throat sections are rectangular with a width of 3.75 m. The sill height is zu = zd =
1.15 m. This BCW has a downstream ramp with a 6:1 slope.

(a) Determine the upstream specific energy, Eu.

Q2
Eu = hu + =
2gAu2
(17.0)2
(1.813 + 1.15) + = 3.082 m
2(9.81) [(1.813 + 1.15)(3.75)]
2

(b) Determine hc (referenced from the sill) for this flow rate.

For critical flow, the Froude number equals unity. Thus,

Q2 T (17.0)2 (3.75)
F = 2
= =1
gA 3 9.81( 3.75hc )3
r

whereby hc = 1.280 m.

(c) Estimate the hydraulic head loss in the diverging section, (hf)ds.

See the equations in the lecture notes:

5 of 7
 log10 ⎡114.6 tan−1 (1/ 6 )⎤ − 0.165
ξ= ⎣ ⎦ = 0.638
1.742
The velocity at the critical-flow section is Vc = 17.0/((3.75)(1.280)) = 3.54 m/s.
The velocity at the DS section is Vd = 17.0/((3.75)(2.661)) = 1.70 m/s. Then,

ξ ( Vc − Vd ) 0.638 ( 3.54 − 1.70 )

2 2
(hf )ds = = ≈ 0.11 m
2g 2(9.81)

5. (25 pts) A circular pipe has a circular, sharp-edged orifice plate (D2 = 5.50 inches),
centered in the pipe cross section, which has an inside diameter of D1 = 7.90
inches. Taps at D1 (upstream) and ½D1 (downstream) are connected to a
manometer with mercury, from which a head differential of 186 mm is measured.
Water temperature is 13°C. Ignoring thermal expansion adjustments:

(a) Determine the ratio β.

D2 5.50
β= = = 0.696
D1 7.90

(b) Determine the kinematic viscosity, ν.

1
ν= = 1.198(10)−6 m2 /s
83.9192(13) + 20,707.5(13) + 551,173
2

(c) Determine the discharge coefficient, Cd.

This requires one or two iterations. The diameter of the orifice is: D2 =
0.3048(5.50/12) = 0.140 m. The cross-sectional area of the orifice opening is:

πD2 π(0.140)2
A2 = = = 0.0154 m2
4 4
Start with Cd = 0.6.

2g(0.186)(13.6 − 1)
Q1 = 0.6(0.0154) = 0.6(0.1194) = 0.0716 m3 /s
1 − (0.696)4

Then,

6 of 7
 4Q 4(0.0716)
Re = = = 543,500
πDν π(0.140)(0.000001198)

and,

Cd = 0.5959 + 0.0312(0.696)2.1 − 0.184(0.696)8

0.039(0.696)4 91.71(0.696)2.5
+ − 0.0158(0.696) +
3

1 − (0.696)4 (543,500)0.75
37.06
= 0.607 + 0.75
Re
= 0.609
The adjusted flow rate is:

Q2 = 0.609(0.1194) = 0.0727 m3 /s

Another iteration gives: Re = 551,900, and Cd = 0.609, so it is converged.

(d) Determine the upstream Reynolds number, Re.

Re = 551,900

(e) Determine the flow rate, Q.

Q = 0.0727 m3/s

7 of 7
 BIE 6300
Fall Semester 2004
Exam #2 – 14 Dec 04
Show your work neatly on this or separate pages. Show units for all
calculated values. Make note of any important assumptions.

Name_________________________________

An inverted siphon with a single barrel of circular concrete pipe (2.25-ft inside diameter) has a total
length of 635 ft. The descending part of the barrel has a length of 185 ft. The inlet and outlet
structures are ungated Type 1 USBR transitions. See the side view (profile) figure below:
upstream downstream
983.300 ft elev

975.950 ft elev

954.445 ft elev

The inverted siphon connects upstream and downstream open canals which have the same
trapezoidal in cross section: base width is b = 4.0 ft; inverse side slope is 1.00; and, longitudinal
bed slope is 0.000212 ft/ft. The Manning roughness is estimated to be n = 0.015 in the open
canals. Assume uniform-flow conditions in the open channels both upstream and downstream of
the inverted siphon.

a) Estimate the flow rate through the inverted siphon pipe under steady-state
flow conditions. Use the Darcy-Weisbach equation with friction factor “f”
defined by the Swamee-Jain equation:

0.25
f= 2
⎡ ⎛ ε 5.74 ⎞ ⎤
⎢log10 ⎜ + 0.9 ⎟ ⎥
⎣ ⎝ 3.75D NR ⎠ ⎦
where the roughness height is ε = 0.001 ft; and the kinematic viscosity of the
water is ν = 1.2(10)-5 ft2/s.

b) Calculate the maximum pressure in the inverted siphon pipe for steady-state
flowing conditions.
c) Calculate the maximum pressure in the inverted siphon pipe for non-flowing
conditions.
d) Calculate the uniform-flow depth in the upstream open canal.
e) Is the hydraulic seal sufficient at the inverted siphon entrance? (the barrel
invert has the same elevation as the canal invert at the siphon entrance).
Solution:

(a) Steady-State Flow Rate

• Recognize that the open canals US and DS of the inverted siphon have the same cross-
section, longitudinal bed slope, and roughness. Thus, if uniform-flow conditions prevail US
and DS of the inverted siphon, the depth just US of the siphon must be the same as the
depth just DS. This means that the head differential on the siphon is equal to the
difference in entrance and exit elevations:

∆Elev = 983.300 − 975.95 = 7.350 ft

• Under steady-state flow conditions, the velocity heads are the same in the open channels
just US and just DS of the inverted siphon. This means that the hydraulic energy loss
through the siphon under these conditions is very nearly equal to ∆Elev. Then,

∆Elev = hf
where hf is determined from Darcy-Weisbach and Swamee-Jain (English units):

L V 2 8 f LQ2 8 f (635)Q2
hf = f = = = 0.277 f Q2
D 2g gπ D 2 5
(32.2)π (2.25)
2 5

• The friction factor, f, is defined as:

0.25 0.25
f= 2
= 2
⎡ ⎛ 0.001 5.74 ⎞ ⎤ ⎡ ⎛ 5.74 ⎞ ⎤
⎢log10 ⎜ + 0.9 ⎟ ⎥ ⎢log10 ⎜ 0.000119 + 0.9 ⎟ ⎥
⎣ ⎝ 3.75(2.25) NR ⎠ ⎦ ⎣ ⎝ NR ⎠ ⎦

• The Reynolds number is:

VD 4Q 4Q
NR = = = −5
= 4.72(10)4 Q
ν νπD 1.2(10) π(2.25)
• Then:
0.25
f=
( )
2
⎡log10 0.000119 + 0.000357Q−0.9 ⎤
⎣ ⎦
• Combine the above equation with the equation for hf:

0.0693Q2
hf = = 7.350 ft
( )
2
⎡log10 0.000119 + 0.000357Q−0.9 ⎤
⎣ ⎦
where the only unknown is Q (cfs). Rearrange as follows:
 (
Q = −10.3 log10 0.000119 + 0.000357Q−0.9 )
• Using the above equation, make an initial guess of 10 cfs (for example), then iterate to
determine the solution to the equation:

Q next Q
(cfs) (cfs)
10.00 38.99
38.99 39.95
39.95 39.96

• From the above table, Q = 40.0 cfs (to three significant digits). This gives f = 0.0166.

(d) Uniform Flow Depth in the Canal

• Use the Manning equation:

[h(4.0 + h)]
5/3
1.49
40.0 = 0.000212
0.015 ⎡ 4.0 + 2h 1 + 1⎤ 2 / 3
⎣ ⎦
or,
[h(4.0 + h)]
5/3

27.7 =
[4.0 + 2.83h]
2/3

• Solving the above equation by iteration: h = 2.89 ft (to three significant digits).

(b) Maximum Pressure: Steady-State Flow

• In this case, the maximum pressure will occur at the lowest point in the inverted siphon, at
the end of the descending portion of the barrel. This pressure is equal to the change in
elevation from the US free water surface to the end of the descending part of the barrel,
minus the friction loss. Using Darcy-Weisbach for hf:

8f(185)(40.0)2
hmax = ( 2.89 + 983.300 − 954.445 ) −
(32.2)π2 (2.25)5
or,
hmax = 31.7 − 129 f

• Note that f is the same as previously calculated (same D, same Q, etc.): f = 0.0166. Then,

hmax = 31.7 − 129(0.0166) = 29.6 ft

and,
 29.6 ft
Pmax = = 12.8 psi
2.31 ft/psi
(c) Maximum Pressure: Zero Flow

• In this case, the maximum pressure will again occur at the lowest point in the inverted
siphon. For zero flow, there is zero friction loss.
• Assuming zero depth in the downstream channel:

hmax = 975.950 − 954.445 = 21.5 ft

and,
21.5 ft
Pmax = = 9.31 psi
2.31 ft/psi
• Thus, the pressure is very low under both steady and non-flowing conditions, well within
the pressure rating for concrete pipe joints.

(e) Hydraulic Seal at Siphon Entrance

• The hydraulic seal is the difference in elevation between the US free water surface and the
crown (highest point) of the barrel inlet. In our case, this is approximately equal to:

seal = 2.89 − 2.25 = 0.64 ft

• According to USBR design criteria, the required hydraulic seal is 1.5 times the difference
in velocity heads between the pipe and the open canal:

required seal = 1.5∆hv

• Velocity head in the pipe:

⎛ V2 ⎞ 8(40.0 cfs)2
⎜ ⎟ = = 1.57 ft
( )
2
⎝ 2g ⎠pipe (32.2) π(2.25)2

• Velocity head in the canal:

⎛ V2 ⎞ (40.0 cfs)2
⎜ ⎟ = = 0.0627 ft
⎝ 2g ⎠canal 2(32.2) ( 2.89(4 + 2.89) )
2

• Then,

required seal = 1.5 (1.57 − 0.0627 ) = 2.26 ft

• Thus, the actual hydraulic seal is much less than the required hydraulic seal.
 BIE 5300/6300 Assignment #1
Simple Flow Measurement and Flumes
02 Sep 04 (due 07 Sep 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments about each of the problems.

I. A float test was conducted in a straight reach of open channel. The rectangular
cross section had a base width of 1.55 m. A small piece of a wooden twig was
timed with a stopwatch over a distance of 10 m, and the three trials produced
transit times of: 7.42, 7.58, and 7.49 s. The depth of water in the center of the
channel was 1.32 m. Apply the float method to estimate the discharge in m3/s.

II. You have field data for a concrete-lined, rectangular irrigation canal. The base
width is 3.04 m. The data show that the channel is prismatic (straight in
alignment, with constant cross-sectional shape and size) and the water depth
was constant at 2.15 m along a 2.23-km reach. The water depth had not
changed over a period of two hours before the measurements were taken. If the
longitudinal bed slope of the channel is 0.00012 m/m, what is the estimated
discharge range (min & max), in m3/s, for this canal reach? (use the Manning
and or Chezy equations).

III. Data were taken using the dye method in an earthen canal. A slug of dye was
injected into the center of the stream and the leading (front) edge traveled a
distance of 12 m in 9.05 s, while the trailing edge crossed the 12-m distance in
10.11 s. If the cross-sectional area of the channel was 5.13 m2, what is the
estimated discharge in m3/s?

IV. The table below has free-flow calibration data for a Cutthroat flume with L = 3.0
ft and W = 8 inches. Analyze the data to determine the values of Cf and nf and
compare these values to those published in the lecture notes.

Q (cfs) hu (ft)
2.688 1.003
2.688 1.003
2.686 1.003
2.687 1.003
2.687 1.003
1.865 0.811
1.865 0.811
3.375 1.095
2.263 0.907
1.461 0.696
1.122 0.611
1.123 0.611
1.123 0.611
1.123 0.611
0.844 0.518
V. The table below has submerged-flow calibration data for a Cutthroat flume with L
= 3.0 ft and W = 8 inches. Analyze the data to determine the values of Cs and
ns, using the nf value from the previous calibration for the same flume size.
Compare these values to those published in the lecture notes.

Q (cfs) hu (ft) hd (ft)

2.687 1.005 0.575
2.687 1.008 0.602
2.687 1.009 0.627
2.687 1.011 0.660
2.687 1.013 0.689
2.687 1.021 0.739
2.687 1.030 0.781
2.685 1.055 0.863
2.685 1.083 0.931
2.685 1.120 0.993
2.683 1.183 1.093
1.865 0.813 0.581
1.865 0.815 0.606
1.865 0.820 0.639
1.865 0.828 0.673
1.863 0.840 0.702
1.863 0.905 0.826
1.86 0.870 0.763
1.86 0.953 0.893
1.86 1.077 1.025
1.858 1.015 0.968
1.123 0.613 0.443
1.123 0.617 0.468
1.123 0.618 0.485
1.123 0.623 0.523
1.123 0.638 0.560
1.123 0.666 0.608
1.123 0.698 0.658
1.12 0.828 0.805

VI. A Cutthroat flume is to be installed in an existing concrete-lined rectangular

canal with a base width of 5.0 ft. The maximum flow rate to be measured at that
location is 40 cfs. The longitudinal bed slope of the channel is 0.00045 ft/ft, and
the depth of the concrete lining is 3.7 ft. Assume a Manning roughness value of
n = 0.015.

(a) Select an appropriate Cutthroat flume size, in English units, from the table in
the lecture notes.
(b) Determine the minimum height of the Cutthroat flume floor, relative to the
existing canal bed, such that free-flow conditions prevail up to the maximum
discharge of 40 cfs. Do not specify a floor elevation below the existing canal
bed.
(c) Will the upstream canal banks need to be raised if the Cutthroat flume is
installed for free-flow conditions?
Solutions:

I. The average transit time of the three trials is: (7.42 + 7.58 + 7.49)/3 = 7.50 s.
The average estimated surface velocity is: 10 m/7.50 s = 1.33 m/s.
Alternatively, the average surface velocity can be taken as (10/7.42 + 10/7.58 +
10/7.49)/3 = 1.33 m/s (same result). The cross-sectional area of the channel is:

A = (1.55)(1.32) = 2.05 m2

The average depth in the channel is 1.32 m (because the section is rectangular).
Interpolating linearly (which is the simplest option, and probably as valid as
anything else in this case) in the table from the lecture notes, the surface
velocity coefficient is:

1.32 − 1.22 C − 0.72

=
1.52 − 1.22 0.74 − 0.72

which gives C = 0.73. The average velocity in the cross section is estimated as:
(0.73)(1.33 m/s) = 0.97 m/s. Finally, the flow rate is estimated to be:

( )
AV = 2.05 m2 ( 0.97 m/s ) ≈ 2.0 m3 /s

II. For this channel cross section size and type of lining, the Manning “n” value
might be in the range 0.012 < n < 0.018, depending on the condition of the
concrete lining and the presence (or absence) of vegetation and sediment. It
probably won’t be less than 0.012, but could be greater than 0.018.

The cross-sectional area is: A = (3.04 m)(2.15 m) = 6.54 m2. The wetted
perimeter is: Wp = 3.04 + 2(2.15) = 7.34 m. Applying the Manning equation:

1 ( 6.54 )
5/3
1 A5 / 3
Q= S = 0.00012 ≈ 5.5 m3 /s
0.012 ( 7.34 )
2/3 o 2/3
n Wp

at the lower range. Applying the equation again for n = 0.018, we get Q ≈ 3.7
m3/s. Thus, by the Manning equation, you might agree that: 3.7 < Q < 5.5 m3/s.

Note that the Manning “n” value has no more than two significant digits in this
case, so Q can have no more than that. You could also apply the Chezy
equation. These two equations will be discussed further in a future lecture.

III. With this method, you don’t apply a coefficient to the measured velocity, which in
this case is:
 12 m
V= = 1.25 m/s
0.5 ( 9.05 + 10.11)

Finally, the estimated discharge is:

( )
Q = AV = 5.13 m2 (1.25 m/s) ≈ 6.4 m3 /s

IV. The free-flow data were analyzed as shown in the table below, giving Cf = 2.71
and nf = 1.78. Three significant digits are the most that can be justified for this
kind of calibration. Note the high coefficient of determination, R2, of 0.998.
These results were done in MS Excel®, using the LINEST spreadsheet function.

Predicted Percent
Q (cfs) hu (ft) ln(Q) ln(hu) Q (cfs) Difference Linear Regression
2.688 1.003 0.9888 0.0025 2.7234 -1.31 1.7805 0.9974
2.688 1.003 0.9888 0.0025 2.7234 -1.31 0.0212 0.0070
2.686 1.003 0.9881 0.0025 2.7234 -1.38 0.9982 0.0198
2.687 1.003 0.9884 0.0025 2.7234 -1.35
2.687 1.003 0.9884 0.0025 2.7234 -1.35 Cf= 4.07
1.865 0.811 0.6233 -0.2097 1.8665 -0.08 nf= 1.78
1.865 0.811 0.6233 -0.2097 1.8665 -0.08
3.375 1.095 1.2164 0.0908 3.1868 5.74
2.263 0.907 0.8167 -0.0980 2.2773 -0.63
1.461 0.696 0.3791 -0.3626 1.4215 2.74
1.122 0.611 0.1151 -0.4929 1.1272 -0.46
1.123 0.611 0.1160 -0.4929 1.1272 -0.37
1.123 0.611 0.1160 -0.4929 1.1272 -0.37
1.123 0.611 0.1160 -0.4929 1.1272 -0.37
0.844 0.518 -0.1696 -0.6587 0.8390 0.59
Note: L = 3.0 ft and W = 8 inches

Thus,
Qf = 4.07 W h1.78
u

for W = 8/12 ft; hu in feet; and Qf in cfs.

V. The submerged-flow data were analyzed as shown in the table below, also in a
spreadsheet application, giving Cs = 1.60 and ns = 1.31. Three significant digits
are the most that can be justified for this kind of calibration. The free-flow
exponent, nf, was used in this calibration. Note the coefficient of determination,
R2, of 0.999.
 Predicted Percent Linear Regression
Q (cfs) hu (ft) hd (ft) S=hd/hu hu-hd LHS RHS Q (cfs) Difference -1.311555 0.468977
2.687 1.005 0.575 0.572 0.430 2.356 -1.417 2.61 2.85 0.007212 0.019525
2.687 1.008 0.602 0.597 0.406 2.449 -1.497 2.64 1.75 0.999184 0.030379
2.687 1.009 0.627 0.621 0.383 2.545 -1.575 2.66 1.01
2.687 1.011 0.660 0.653 0.351 2.685 -1.687 2.68 0.42 Cs= 2.40
2.687 1.013 0.689 0.680 0.324 2.813 -1.787 2.69 0.04 ns= 1.31
2.687 1.021 0.739 0.724 0.282 3.041 -1.965 2.70 -0.46
2.687 1.030 0.781 0.758 0.249 3.240 -2.118 2.71 -0.71
2.685 1.055 0.863 0.818 0.192 3.664 -2.441 2.70 -0.66
2.685 1.083 0.931 0.859 0.153 4.034 -2.720 2.69 -0.19
2.685 1.120 0.993 0.887 0.127 4.335 -2.954 2.71 -0.86
2.683 1.183 1.093 0.925 0.089 4.903 -3.380 2.68 0.09
1.865 0.813 0.581 0.715 0.232 2.992 -1.926 1.87 -0.22
1.865 0.815 0.606 0.743 0.209 3.158 -2.049 1.86 0.10
1.865 0.820 0.639 0.779 0.181 3.394 -2.224 1.85 0.82
1.865 0.828 0.673 0.812 0.156 3.635 -2.402 1.84 1.51
1.863 0.840 0.702 0.835 0.138 3.827 -2.549 1.84 1.44
1.863 0.905 0.826 0.913 0.079 4.731 -3.225 1.80 3.21
1.86 0.870 0.763 0.877 0.107 4.246 -2.868 1.83 1.55
1.86 0.953 0.893 0.937 0.060 5.178 -3.567 1.80 3.06
1.86 1.077 1.025 0.952 0.052 5.421 -3.846 2.04 -9.31
1.858 1.015 0.968 0.954 0.047 5.584 -3.890 1.83 1.32
1.123 0.613 0.443 0.723 0.170 2.987 -1.959 1.18 -5.20
1.123 0.617 0.468 0.759 0.148 3.207 -2.124 1.18 -4.79
1.123 0.618 0.485 0.784 0.133 3.380 -2.249 1.17 -3.88
1.123 0.623 0.523 0.838 0.101 3.833 -2.569 1.13 -0.52
1.123 0.638 0.560 0.877 0.078 4.242 -2.867 1.11 1.24
1.123 0.666 0.608 0.914 0.058 4.743 -3.238 1.09 2.63
1.123 0.698 0.658 0.944 0.039 5.365 -3.685 1.05 6.27
1.12 0.828 0.805 0.972 0.023 6.201 -4.389 1.15 -2.47
Notes: "LHS" is ln(Qs) - nf ln(hu-hd). "RHS" is ln(-log10(S))

Thus,
2.40 W ( hu − hd )
1.78

Qs =
( − log10 S )
1.31

for W = 8/12 ft; hu & hd in feet; and Qf in cfs.

VI. Referring to the Cutthroat flume table for English units in the lecture notes, and
the Cutthroat flume top view figure, it is seen that the width of the flume is B = W
+ L/4.5.

(a) In the table, the smallest Cutthroat flume with a capacity of at least 40 cfs is
for W = 3.333 ft and L = 7.50 ft. This gives B = 3.333 + 7.50/4.5 = 5.0 ft,
which is the same as the base width of the rectangular canal. This works out
just right in this case, so choose the W = 3.333 ft and L = 7.50 ft size. Note
that W/L = 4/9.
(b) From the calibration table (English units), St = 0.873 for our selected flume
size. Free-flow conditions at Qmax give:

1/ nf 1/1.57
⎛Q ⎞ ⎛ 40 ⎞
hu = ⎜ max ⎟ =⎜ ⎟ = 2.18 ft
⎝ Cf W ⎠ ⎝ (3.519)(3.333) ⎠
 Then, Sthu = (0.873)(2.18) = 1.90 ft, which is the maximum downstream
depth with respect to the upstream floor elevation.

Using the ACA program to determine the normal depth corresponding to the
given conditions, at Q = 40 cfs, it is found that the downstream depth would
be 3.07 ft. Thus, the floor of the flume must be at least 3.07 – 1.90 = 1.17 ft
above the existing canal bed. See the side view figure below.

(c) The depth of the concrete lining is given as 3.7 ft. The upstream depth at 40
cfs with the Cutthroat flume in place will be 2.18 + 1.17 = 3.35 ft, which is
less than 3.7 ft by a margin of 0.35 ft, which may be enough freeboard in this
case. The upstream canal banks do not need to be raised.
 BIE 5300/6300 Assignment #2
Current Metering Calculations
09 Sep 04 (due 14 Sep 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

I. You are given the electromagnetic current metering measurements below.

Calculate the total flow rate in the channel using a spreadsheet. Plot the cross-
section profile in a graph.

Distance Depth Depth Velocity (m/s) Average Width Area Flow

mean at subsection
(m) (m) Fraction At point vertical mean depth (m) (m) (m2) (m3/s)
Left
0.000 0.00 0.25

0.305 0.58 0.2 0.51

0.8 0.46
0.615 0.58 0.2 0.54
0.8 0.46
0.925 0.58 0.2 0.49
0.8 0.43
1.230 0.00 0.25
Right Totals:

II. You are given the electromagnetic current metering measurements below.
Calculate the total flow rate in the channel using a spreadsheet. Plot the cross-
section profile in a graph.

Distance Depth Depth Velocity (m/s) Average Width Area Flow

mean at subsection
Fraction At point vertical mean Depth
(m) (m) (m) (m) (m2) (m3/s)
Left
0.00 0.500

0.16 0.500 0.2 1.40

0.8 1.51
0.32 0.500 0.2 1.85
0.8 1.45
0.48 0.500 0.2 1.96
0.8 1.50
0.64 0.500 0.2 1.94
0.8 1.56
0.80 0.500 0.2 1.62
0.8 1.46
0.96 0.500
Right
Totals:
Solutions:

I. The calculations were performed in a spreadsheet and the results are:

Distance Depth Depth Velocity (m/s) Average Width Area Flow

mean at subsection
(m) (m) Fraction At point vertical mean depth (m) (m) (m2) (m3/s)
Left
0.000 0.00 0.25 0.250
0.368 0.290 0.305 0.088 0.033
0.305 0.58 0.2 0.51 0.485
0.8 0.46 0.493 0.580 0.310 0.180 0.089
0.615 0.58 0.2 0.54 0.500
0.8 0.46 0.480 0.580 0.310 0.180 0.086
0.925 0.58 0.2 0.49 0.460
0.8 0.43 0.355 0.290 0.305 0.088 0.031
1.230 0.00 0.25 0.250
Right Totals: 1.230 0.537 0.239

0.60

0.50

0.40

0.30

0.20

0.10

0.00
0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400
II. Again, the calculations were performed in a spreadsheet, applying the algorithm
for vertical walls (see the lecture notes), and the results are:

Distance Depth Depth Velocity (m/s) Average Width Area Flow

mean at subsection
Fraction At point vertical mean Depth
(m) (m) (m) (m) (m2) (m3/s)
Left x/D= 0.32
0.00 0.500 Vx/VD= 0.92 1.03
1.24 0.500 0.16 0.080 0.099
0.16 0.500 0.2 1.40 1.46
0.8 1.51 1.55 0.500 0.16 0.080 0.124
0.32 0.500 0.2 1.85 1.65
0.8 1.45 1.69 0.500 0.16 0.080 0.135
0.48 0.500 0.2 1.96 1.73
0.8 1.50 1.74 0.500 0.16 0.080 0.139
0.64 0.500 0.2 1.94 1.75
0.8 1.56 1.65 0.500 0.16 0.080 0.132
0.80 0.500 0.2 1.62 1.54
0.8 1.46 1.31 0.500 0.16 0.080 0.105
0.96 0.500 x/D= 0.32 1.09
Right Vx/VD= 0.92
Totals: 0.960 0.480 0.735

Note: Never show open-channel current metering results with more than two or
three significant digits – four or more significant digits cannot be justified.
 BIE 5300/6300 Assignment #3
Weir Calculations
24 Sep 04 (due 28 Sep 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

I. Hydraulically, what is the difference between a suppressed weir and a

contracted weir?

II. A rectangular sharp-crested weir is installed in a straight section of a rectangular

open channel with B = 10 ft, L = 7 ft, P = 3 ft, and hu = 1.37 ft. The downstream
water surface is well below the weir crest elevation.

(a) Do the stated conditions meet all the guidelines for setting and
operating weirs, as given in the lecture notes? If no, which are
violated?
(b) You are required to estimate the discharge over the weir for the stated
conditions. Assuming negligible approach velocity, estimate Qf in cfs.
(c) Estimate the discharge, Qf, without assuming a negligible approach
velocity (hint: use Hu instead of hu in the calibration equation).
(d) Now suppose hd = 0.31 ft and everything else is the same as given
above. Estimate Qs in cfs.

III. A Cipoletti weir is installed in an open channel. The approach velocity is

negligible. If the crest length is 0.90 m and the upstream depth, hu, is 0.22 m,
referenced to the crest elevation, what is the estimated free-flow discharge?

IV. An overshot gate with L = 8.0 ft and Gw = 12.0 ft is installed in a canal. At the
downstream side of the gate is a reservoir with a constant water surface
elevation which is 0.29 ft above the gate hinge. The irrigation district needs you
to develop and plot calibration curves for gate openings of: θ = 15, 20, 25, 30,
35, 40, 45, 50, 55, and 60 degrees. The plot should have hu on the abscissa
and Q on the ordinate. Both the abscissa and ordinate must start at zero. Each
curve on the plot must be labeled with its corresponding gate opening angle.

V. A new trapezoidal concrete canal with a base width of 2.0 m and inverse side
slopes of 1.5 has a total lined depth of 2.5 m. The bed slope of the canal is
0.00015 m/m and the length is 2.35 km, all straight in alignment (no curves or
bends). At the end of the section there is a sudden drop in the bed elevation of
3.5 m, then the same channel cross section continues downstream, with the
same bed slope. Design a sharp-crested weir, just upstream of the elevation
drop, for a maximum flow rate of 7.0 m3/s. Make sure the canal lining won’t be
overtopped upstream of the weir.
 BIE 5300/6300 Assignment #5
Open-Channel Constriction Calibrations
06 Oct 04 (due 12 Oct 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

You have measured data in a spreadsheet file (attached) for an open-channel

constriction (non-orifice flow). One data set is for free-flow conditions, the other for
submerged-flow.

I. Develop a free-flow rating for this constriction using the following equation:

Qf = Cf hnuf

(a) Determine Cf and nf.

(b) Make a graph with plotted symbols for (Qf)measured vs. (Qf)calculated. The
ordinate range should be the same as the abscissa range, with a
diagonal line representing (Qf)measured/(Qf)calculated = 1.0.
(c) Comment on the data fit, using correlation or other such indices, as
appropriate.

II. Develop a submerged-flow rating for this constriction using the following
equation:
Cs (hu − hd )
nf

Qs =
( − log10 S )
ns

(a) Determine Cs and ns, using nf from the free-flow rating.

(b) Make a graph with plotted symbols for (Qs)measured vs. (Qs)calculated. The
ordinate range should be the same as the abscissa range, with a
diagonal line representing (Qs)measured/(Qs)calculated = 1.0.
(c) Comment on the data fit, using correlation or other such indices, as
appropriate.

III. Solve for transition submergence, St, for the above calibration.

(a) Determine St.

(b) Make a graph of St (from 0.1 to 0.99 on the abscissa) vs. the function
value (Qf - Qs = 0) and indicate where the solution(s) exist, if any.
(c) If you don’t get any solution for St, try adjusting Cs slightly so that you
get a solution. If you do this, show the adjusted Cs value.
IV. Re-do the submerged-flow rating using the following equation:

Cs (hu − hd )
ns1

Qs =
( − log10 S )
ns 2

(a) Determine Cs, ns1 and ns2 based only on the submerged-flow data.
(b) Make a graph with plotted symbols for (Qs)measured vs. (Qs)calculated. The
ordinate range should be the same as the abscissa range, with a
diagonal line representing (Qs)measured/(Qs)calculated = 1.0.
(c) Comment on the data fit, using correlation or other such indices, as
appropriate.
(d) Comment on the data fit using this equation, as opposed to using the Qs
equation from (II) above.
Solutions:

I. Develop a free-flow rating for this constriction using the following equation:

Qf = Cf hnuf

Make two new columns for ln(Qf) and ln(hu) in the spreadsheet. Do a linear
regression using the LINEST spreadsheet function. The regression gives:

Cf= 5.57
nf = 1.61

for Qf in cfs; and hu in ft.

The R2 value is 0.999, indicating a very good fit, and this is also seen in the
comparison graph:

7.0

6.0

5.0
(Qf)calculated

4.0

3.0

2.0

1.0

0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
(Qf)measured

II. Develop a submerged-flow rating for this constriction using the following
equation:
Cs (hu − hd )
nf

Qs =
( − log10 S )
ns

Make two new columns for ln(Qs/(hu-hd)nf) and ln(-log10S) in the spreadsheet.
Do a linear regression using the LINEST spreadsheet function. The regression
gives:
 Cs= 2.62
ns = 1.43

for Qs in cfs; and hu & hd in ft.

The R2 value is 0.999, indicating a very good fit, and this is also seen in the
comparison graph:

7.0
6.5
6.0
5.5
(Qs)calculated

5.0
4.5
4.0
3.5
3.0
2.5
2.0
2.0 3.0 4.0 5.0 6.0 7.0
(Qs)measured

III. Solve for transition submergence, St, for the above calibration.

Use the equation for f(St) = 0, as shown in the lecture notes. This equation is
derived by setting Qf = Qs. Make a table of St versus f(St), then plot the results.
The only solution is for St = 1.00, which is mathematically correct, but physically
impossible.

Adjust Cs slightly, from 2.62 to 2.639, whereby a solution is found at about St ≈

0.79, as shown in the next graph.
 6.0

5.0

4.0
f(St)

3.0

2.0

1.0

0.0
0.0 0.2 0.4 0.6 0.8 1.0
St

0.016

0.014

0.012

0.010
f(St)

0.008
Solution at
0.006 St = 0.79

0.004

0.002

0.000
0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
St
IV. Re-do the submerged-flow rating using the following equation:

Cs (hu − hd )
ns1

Qs =
( − log10 S )
ns 2

Make three new columns for ln(Qs), ln(hu-hd), and ln(-log10S) in the spreadsheet.
Do a multiple linear regression using the LINEST spreadsheet function. The
regression gives:

Cs= 2.78
ns1 = 1.54
ns2 = 1.36

for Qs in cfs; and hu & hd in ft.

The R2 value is 0.996, indicating a very good fit, and this is also seen in the
comparison graph, which is very similar to the previous plot:

7.0
6.5
6.0
5.5
(Qs)calculated

5.0
4.5
4.0
3.5
3.0
2.5
2.0
2.0 3.0 4.0 5.0 6.0 7.0
(Qs)measured

Even though the R2 value is slightly lower than for the previous form of the
submerged-flow equation, the sum of absolute deviations in measured and
calculated discharges is less in this case.
 BIE 5300/6300 Assignment #6
Pipe Flow Measurement
13 Oct 04 (due 18 Oct 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments. E-mail the assignment to me, or drop it
by my office, by 12:00 pm Monday, then I will post my solutions at 12:30 pm
the same day so that you can study for the test on Tuesday.

I. You use a simple Pitot tube to measure the total head at the center of a circular
pipe with an inside diameter of 336 mm. The tip of the tube points in the
upstream direction. You find a total head of 42.35 m of water when connecting
the Pitot tube to a manometer. Separately, you measure the pressure in the
pipe at the same location, obtaining P = 413 kPa.

(a) Estimate the velocity in the pipe at the center of the cross section.
(b) Estimate the flow rate in the pipe, in liters per second.

II. You have a venturi connected to a manometer with mercury, whereby the
manometer is connected to an upstream tap, and to a tap just at the throat of the
venturi. The head differential on the mercury is 456 mm. The diameters are: D1
= 100 mm (upstream), and D2 = 50 mm (throat). The calibration coefficient for a
“machined inlet” is C = 0.995. Calculate the flow rate through the venturi.

III. You have a sharp-crested circular orifice at a gasketed pipe flange fitting. The
upstream pipe ID is D1 = 12.0 inches and the orifice diameter is 9.05 inches.
The orifice opening is centered in the pipe cross section. The upstream tap is at
a distance D1 upstream of the orifice plate, and the downstream tap is at a
distance ½D1 downstream of the plate. When the taps are connected to a
manometer with “blue” fluid (sg = 1.75), the head differential is observed to be
0.519 m. The water temperature is measured and found to be 8°C. Calculate
the flow rate to three significant digits, taking into account the Reynold’s number.

IV. You have to estimate the discharge from a partially-full horizontal pipe which
discharges freely into a canal. The end of the pipe is 20 cm above the water
surface in the canal. The pipe inside diameter is 35 cm, and the depth of water
at the pipe end is measured, giving 13 cm. Estimate the discharge in m3/s.

1
Solutions:

I. You use a simple Pitot tube to measure the total head at the center of a circular
pipe with an inside diameter of 336 mm. The tip of the tube points in the
upstream direction. You find a total head of 42.35 m of water when connecting
the Pitot tube to a manometer. Separately, you measure the pressure in the
pipe at the same location, obtaining P = 413 kPa.

(a) Velocity in the pipe at the center of the cross section.

Pipe area:
π(0.336)2
A= = 0.08867 m2
4
Pressure head:
P 413 kPa
= = 42.1 m
γ 9.81

Velocity head:
V2
= 42.35 − 42.1 = 0.25 m
2g

Velocity (at center):

V = 2.21 m/s

(b) Flow rate in the pipe, in liters per second.

The maximum flow rate would be:

Qmax = VA = (2.21)(0.08867) = 0.196 m3 /s

or, 196 lps. The true flow rate is probably slightly lower than this
because the velocity at the center of the cross section is greater than
the average velocity, even for fully turbulent flow.

II. You have a venturi connected to a manometer with mercury, whereby the
manometer is connected to an upstream tap, and to a tap just at the throat of the
venturi. The head differential on the mercury is 456 mm. The diameters are: D1
= 100 mm (upstream), and D2 = 50 mm (throat). The calibration coefficient for a
“machined inlet” is C = 0.995. Calculate the flow rate through the venturi.

First, the beta ratio is:

2
 D2 50
β= = = 0.50
D1 100

The cross-sectional area is:

π(0.05)2
A2 = = 0.001963 m2
4
The flow rate is:

2g∆h(sg − 1)
Q = Cd A 2 =
1 − β4
2g(0.456)(13.6 − 1)
0.995(0.001963) = 0.0214 m3 /s
1 − (0.50)4

or, 0.756 cfs.

III. You have a sharp-crested circular orifice at a gasketed pipe flange fitting. The
upstream pipe ID is D1 = 12.0 inches and the orifice diameter is 9.05 inches.
The orifice opening is centered in the pipe cross section. The upstream tap is at
a distance D1 upstream of the orifice plate, and the downstream tap is at a
distance ½D1 downstream of the plate. When the taps are connected to a
manometer with “blue” fluid (sg = 1.75), the head differential is observed to be
0.519 m. The water temperature is measured and found to be 8°C. Calculate
the flow rate to three significant digits, taking into account the Reynold’s number.

First, the beta ratio is:

D2 9.05
β= = = 0.7542
D1 12.00

The cross-sectional area is:

π(9.05 /12)2
A2 = = 0.4467 ft 2
4
or, 0.04150 m2.

The kinematic viscosity is:

1
ν= = 1.385(10)−6 m2 /s
83.9192(8) + 20,707.5(8) + 551,173
2

3
Assume that the linear expansion due to pipe and element temperature is
negligible (which it probably is). Next, assume a starting Cd value of 0.6.

2g(0.519)(1.75 − 1)
Q1 = 0.6(0.0415) = 0.0837 m3 /s
1 − (0.7542)4

The Reynold’s number is:

4Q 4(0.0837)
Re = = = 334,700
πDν π(0.2299)(0.000001385)

The first calculated Cd value is:

Cd = 0.5959 + 0.0312(0.7542)2.1 − 0.184(0.7542)8

0.039(0.7542)4 91.71(0.7542)2.5
+ − 0.0158(0.7542) +
3

1 − (0.7542)4 (334,700)0.75
= 0.609
The updated flow rate is:

2g(0.519)(1.75 − 1)
Q2 = 0.609(0.0415) = 0.0850 m3 /s
1 − (0.7542) 4

The updated Reynold’s number is:

4Q 4(0.0850)
Re = = = 339,900
πDν π(0.2299)(0.000001385)

The first calculated Cd value is:

Cd = 0.5959 + 0.0312(0.7542)2.1 − 0.184(0.7542)8

0.039(0.7542)4 91.71(0.7542)2.5
+ − 0.0158(0.7542) +
3

1 − (0.7542)4 (339,900)0.75
= 0.609
Thus, the coefficient has converged to within three significant digits after only
one iteration. The flow rate is 0.0850 m3/s (3.00 cfs).

4
IV. You have to estimate the discharge from a partially-full horizontal pipe which
discharges freely into a canal. The end of the pipe is 20 cm above the water
surface in the canal. The pipe inside diameter is 35 cm, and the depth of water
at the pipe end is measured, giving 13 cm. Estimate the discharge in m3/s.

Use the “California pipe method.”

a/D = (35 - 13)/35 = 0.629 (which is greater than 0.45… OK), and

2.48
1.88 ⎛ 0.35 ⎞
Q = 8.69 (1 − 0.629 ) ⎜ 0.3048 ⎟ = 1.90 cfs
⎝ ⎠

or, 0.0538 m3/s.

5
 BIE 5300/6300 Assignment #7
Earthen Canal Design
28 Oct 04 (due 2 Nov 04)

Show your calculations in an organized and neat format, including all relevant
calculations. Indicate any assumptions or relevant comments.

Given:

• An earthen canal is to be designed

• Accommodating the natural terrain, the longitudinal bed slope will be So = 0.0001
• Bed material is a non-cohesive fine sandy soil material with an average particle
diameter of 2 mm
• The angle of repose for wet material is 26 degrees
• Assume a Manning n of 0.019
• Design discharge is Qmax = 12 m3/s
• The source water has a low content of fine sediment (silt)
• The b/h ratio should be limited to values between 2 and 6
• The inverse side slope should be limited as follows: m < 3.5

Required:

• Design the earthen channel section by applying the tractive force method
• Compare your results with those using the assumption of a very wide channel, in
which the critical tractive force is γhSo
• Compare your design with the velocity as obtained from the Kennedy formula
• Compare your design with the velocity as obtained from the Lacey method
• Compare your design with the velocity as obtained from the maximum velocity
method, both for values by Fortier and Scobey, and by the USBR

A Design Solution:

Critical Tractive Force

• The critical tractive force is taken from Fig. 5 (non-cohesive material) of the lecture
notes, using the curve labeled “low content of fine sediment.”
• Instead of reading the graph by eye, use the appropriate equation from the lecture
notes:

Tc = 0.0756(2)3 − 0.241(2)2 + 0.872(2) + 2.26

Tc ≈ 3.64 N / m2

1
Angle of Repose

• The angle of repose, θ, is given as 26°

• Then, the ratio of Tside to Tbed is:

T sin2 φ
K = side = 1− 2
= 1 − 5.20 sin2 φ
Tbed sin θ

• Design requirements for this example call for a side slope between 0.0 & 3.5
• Then,
⎛ 1 ⎞
φmin = tan−1 ⎜ ⎟ = 15.9D
⎝ 3.5 ⎠

• Let φmin = 16° (round to nearest whole degree)

• θ = 26° is the upper limit for φ, so: 16° ≤ φ ≤ 26°

• Make a table of tractive force ratio, K, values for the acceptable range of φ:

φ (deg) m K
16 3.487 0.778
17 3.271 0.745
18 3.078 0.710
19 2.904 0.670
20 2.747 0.626
21 2.605 0.576
22 2.475 0.520
23 2.356 0.454
24 2.246 0.374
25 2.145 0.267
26 2.050 0.027

Maximum Shear Stress Fractions

b/h Kbed
• These are Kbed and Kside 2.0 0.881
• The range of inverse side slopes is: 3.487 ≤ m ≤ 2.050 2.4 0.906
• Recall that the range of bed width to depth is: 2.0 ≤ b/h ≤ 6.0 2.8 0.927
• For the bed, apply the equations from the lecture notes, where 3.2 0.946
Kbed is a function of the ratio b/h for trapezoidal cross sections… 3.6 0.963
4.0 0.969
4.4 0.971
4.8 0.973
5.2 0.975
• For the sides, apply the equations from the lecture notes, where 5.6 0.977
Kside is a function of inverse side slope, m, and the ratio b/h 6.0 0.980
2
 b/h
φ (deg) m 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
16 3.487 0.794 0.793 0.791 0.789 0.788 0.787 0.786 0.785 0.784 0.784 0.783
17 3.271 0.805 0.806 0.807 0.807 0.808 0.809 0.809 0.810 0.810 0.811 0.811
18 3.078 0.808 0.812 0.815 0.818 0.820 0.822 0.824 0.825 0.827 0.828 0.830
19 2.904 0.807 0.812 0.816 0.820 0.823 0.826 0.828 0.831 0.833 0.835 0.837
20 2.747 0.800 0.806 0.811 0.815 0.819 0.822 0.824 0.827 0.829 0.831 0.833
21 2.605 0.791 0.797 0.802 0.806 0.810 0.812 0.815 0.817 0.819 0.821 0.822
22 2.475 0.782 0.787 0.792 0.796 0.799 0.801 0.803 0.805 0.807 0.808 0.809
23 2.356 0.772 0.777 0.782 0.785 0.788 0.790 0.792 0.793 0.795 0.796 0.797
24 2.246 0.763 0.768 0.772 0.775 0.778 0.780 0.781 0.783 0.784 0.785 0.786
25 2.145 0.755 0.760 0.764 0.767 0.770 0.771 0.773 0.774 0.775 0.776 0.777
26 2.050 0.748 0.754 0.757 0.760 0.762 0.764 0.766 0.767 0.768 0.768 0.769
Note: values in italics are Kside

Maximum Allowable Water Depths

• These are hmax based on: (1) bed; and, (2) side slopes
• For water, use γ = 9,810 N/m3
• The following table has hmax values based on Kbed:

b/h
m 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
3.487 3.28 3.19 3.11 3.05 3.00 2.98 2.97 2.97 2.96 2.95 2.95
3.271 3.14 3.05 2.98 2.92 2.87 2.85 2.85 2.84 2.84 2.83 2.82
3.078 2.99 2.91 2.84 2.78 2.73 2.72 2.71 2.71 2.70 2.69 2.69
2.904 2.82 2.75 2.68 2.63 2.58 2.57 2.56 2.55 2.55 2.54 2.54
2.747 2.64 2.56 2.50 2.45 2.41 2.40 2.39 2.39 2.38 2.38 2.37
2.605 2.43 2.36 2.31 2.26 2.22 2.21 2.20 2.20 2.19 2.19 2.18
2.475 2.19 2.13 2.08 2.04 2.00 1.99 1.99 1.98 1.98 1.97 1.97
2.356 1.91 1.86 1.82 1.78 1.75 1.74 1.74 1.73 1.73 1.72 1.72
2.246 1.58 1.53 1.50 1.47 1.44 1.43 1.43 1.43 1.42 1.42 1.42
2.145 1.12 1.09 1.07 1.05 1.03 1.02 1.02 1.02 1.02 1.01 1.01
2.050 0.11 0.11 0.11 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10
Note: values in italics are hmax based on Kbed

• The next table has hmax values based on Kside:

3
 b/h
m 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0
3.487 3.63 3.64 3.65 3.66 3.66 3.67 3.67 3.68 3.68 3.68 3.69
3.271 3.44 3.43 3.43 3.43 3.42 3.42 3.42 3.42 3.41 3.41 3.41
3.078 3.26 3.24 3.23 3.22 3.21 3.20 3.20 3.19 3.18 3.18 3.17
2.904 3.08 3.06 3.05 3.03 3.02 3.01 3.00 2.99 2.98 2.98 2.97
2.747 2.90 2.88 2.86 2.85 2.84 2.83 2.82 2.81 2.80 2.79 2.79
2.605 2.70 2.68 2.67 2.65 2.64 2.63 2.62 2.62 2.61 2.61 2.60
2.475 2.47 2.45 2.44 2.42 2.42 2.41 2.40 2.40 2.39 2.39 2.38
2.356 2.18 2.17 2.15 2.15 2.14 2.13 2.13 2.12 2.12 2.12 2.11
2.246 1.82 1.81 1.80 1.79 1.78 1.78 1.77 1.77 1.77 1.77 1.77
2.145 1.31 1.30 1.30 1.29 1.29 1.28 1.28 1.28 1.28 1.28 1.28
2.050 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13
Note: values in italics are hmax based on Kside

• The above two tables show that hmax from Kside is greater for every combination
of b/h and m (within the given ranges)
• Therefore, use only the table for hmax from Kside to determine the maximum
allowable depth of water
• Note that the full range of “m” and the full range of “b/h” is represented in the
above two tables

Channel Base Width Limits

• Calculate the uniform flow (normal) depth for values of “m” from 2.050 to 3.487
(as shown in the above tables), and various values of base width, “b.”

base width, b (m)

m 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
3.487 1.93 1.88 1.83 1.79 1.74 1.70 1.66 1.62 1.58 1.55 1.51
3.271 1.96 1.91 1.86 1.81 1.76 1.72 1.68 1.64 1.60 1.56 1.53
3.078 2.00 1.94 1.89 1.83 1.79 1.74 1.70 1.65 1.62 1.58 1.54
2.904 2.02 1.97 1.91 1.86 1.81 1.76 1.71 1.67 1.63 1.59 1.55
2.747 2.05 1.99 1.93 1.88 1.83 1.78 1.73 1.69 1.64 1.60 1.56
2.605 2.08 2.02 1.96 1.90 1.84 1.79 1.75 1.70 1.66 1.61 1.58
2.475 2.11 2.04 1.98 1.92 1.86 1.81 1.76 1.71 1.67 1.63 1.59
2.356 2.13 2.06 2.00 1.94 1.88 1.83 1.77 1.73 1.68 1.64 1.60
2.246 2.16 2.09 2.02 1.96 1.90 1.84 1.79 1.74 1.69 1.65 1.60
2.145 2.18 2.11 2.04 1.97 1.91 1.85 1.80 1.75 1.70 1.66 1.61
2.050 2.21 2.13 2.06 1.99 1.93 1.87 1.81 1.76 1.71 1.66 1.62
Note: values in italics are depths, h, based on the Manning equation

• Next, divide the base width by each respective depth, over the range of inverse
side slope values, giving a table of b/h ratios:

4
 base width, b (m)
m 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
3.487 2.07 2.39 2.73 3.08 3.44 3.82 4.22 4.63 5.05 5.49 5.94
3.271 2.04 2.35 2.69 3.04 3.40 3.78 4.17 4.58 5.00 5.44 5.89
3.078 2.00 2.32 2.65 3.00 3.36 3.73 4.13 4.53 4.95 5.39 5.84
2.904 1.98 2.29 2.62 2.96 3.32 3.70 4.08 4.49 4.91 5.34 5.79
2.747 1.95 2.26 2.59 2.93 3.29 3.66 4.05 4.45 4.87 5.30 5.75
2.605 1.92 2.23 2.56 2.90 3.25 3.62 4.01 4.41 4.83 5.26 5.71
2.475 1.90 2.21 2.53 2.87 3.22 3.59 3.98 4.38 4.80 5.23 5.68
2.356 1.88 2.18 2.50 2.84 3.19 3.56 3.95 4.35 4.76 5.20 5.64
2.246 1.85 2.16 2.48 2.81 3.16 3.53 3.92 4.32 4.73 5.16 5.61
2.145 1.83 2.14 2.45 2.79 3.14 3.51 3.89 4.29 4.70 5.13 5.58
2.050 1.81 2.11 2.43 2.76 3.11 3.48 3.86 4.26 4.68 5.11 5.55
Note: values in italics are ratios of b/h, using h from the Manning equation

• These calculations show that the range of base widths is limited (approximately)
as follows (with b in meters):

4<b<9
whereby b < 4 gives b/h less than the minimum of 2, and b > 9 gives b/h greater
than the maximum of 6, as specified for this problem

Ratios of h to hmax

• Divide depths, h, from Manning by hmax based on Kside

• Only values less than unity are acceptable for this ratio
base width, b (m)
m 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
3.487 0.53 0.52 0.50 0.49 0.48 0.46 0.45 0.44 0.43 0.42 0.41
3.271 0.57 0.56 0.54 0.53 0.52 0.50 0.49 0.48 0.47 0.46 0.45
3.078 0.61 0.60 0.58 0.57 0.56 0.54 0.53 0.52 0.51 0.50 0.49
2.904 0.66 0.64 0.63 0.61 0.60 0.58 0.57 0.56 0.54 0.53 0.52
2.747 0.71 0.69 0.67 0.66 0.64 0.63 0.61 0.60 0.59 0.57 0.56
2.605 0.77 0.75 0.73 0.71 0.70 0.68 0.66 0.65 0.63 0.62 0.60
2.475 0.85 0.83 0.81 0.79 0.77 0.75 0.73 0.71 0.70 0.68 0.66
2.356 0.97 0.95 0.92 0.90 0.88 0.85 0.83 0.81 0.79 0.77 0.75
2.246 1.18 1.15 1.12 1.09 1.06 1.03 1.00 0.98 0.95 0.93 0.91
2.145 1.66 1.61 1.57 1.52 1.48 1.44 1.40 1.36 1.33 1.30 1.26
2.050 16.51 16.04 15.58 15.13 14.71 14.30 13.90 13.53 13.17 12.84 12.52
Note: values in bold are ratios of h/hmax that exceed unity, and are therefore unacceptable

• The above table shows that acceptable inverse side slopes for this design
problem are (rounding to two significant digits):

5
 2.3 < m < 3.5

• The above table also shows that the base width can be any value between 4
and 9 m
• These ranges of base width and inverse side slope represent the domain of
feasible design solutions for this channel
• It would often be best to limit the width of the channel, possibly choosing a value
of b = 4, and accepting a somewhat greater depth of water for uniform flow

Very Wide Channel

• With this assumption, Tc = γhSo

• Then,

Tc 3.64
hmax = = = 3.71 m
γSo 9,810(0.0001)

• The largest hmax value based on Kside, for the acceptable range of m and b, is
3.69 m (see the above table)
• Therefore, the “very wide channel” solution is less restrictive than the previous
solution and will not have any bearing on the range of feasible m and b values

Kennedy Formula

• For a “fine sandy soil,” C1 = 0.84

• For water containing “fine silt,” C2 = 0.64
• Applying the Kennedy formula:
0.64
⎛ h ⎞
Vo = 0.84(0.3048) ⎜ ⎟
⎝ 0.3048 ⎠

where Vo is the “regime” velocity (m3/s); and h is depth (m) from the Manning
equation

• Multiply Vo by area, A, to obtain flow rate according to the Kennedy formula,

where:

A = h(b + mh)

for a trapezoidal channel section

• The following table gives Q values (m3/s) based on depths from the Manning
equation and Vo from the Kennedy formula for the previously-established range
of acceptable b and m values:

6
 base width, b (m)
m 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
3.487 17.3 17.1 16.9 16.6 16.4 16.2 16.1 15.9 15.7 15.6 15.5
3.271 17.3 17.0 16.8 16.6 16.4 16.2 16.0 15.8 15.7 15.5 15.4
3.078 17.2 17.0 16.7 16.5 16.3 16.1 15.9 15.8 15.6 15.4 15.3
2.904 17.2 16.9 16.7 16.5 16.3 16.1 15.9 15.7 15.5 15.4 15.2
2.747 17.2 16.9 16.7 16.4 16.2 16.0 15.8 15.6 15.5 15.3 15.2
2.605 17.1 16.9 16.6 16.4 16.2 15.9 15.8 15.6 15.4 15.3 15.1
2.475 17.1 16.8 16.6 16.3 16.1 15.9 15.7 15.5 15.4 15.2 15.1
2.356 17.1 16.8 16.6 16.3 16.1 15.9 15.7 15.5 15.3 15.2 15.0
2.246 17.1 16.8 16.5 16.3 16.0 15.8 15.6 15.4 15.3 15.1 15.0
2.145 17.1 16.8 16.5 16.2 16.0 15.8 15.6 15.4 15.2 15.1 14.9
2.050 17.1 16.8 16.5 16.2 16.0 15.8 15.6 15.4 15.2 15.1 14.9
Note: values in italics are flow rate (m3/s) based on Vo from Kennedy (Q = VoA)

• Note that all flow rates in the above table are greater than Qmax (12 m3/s)
• This means that channel scouring would not be expected, at the design
discharge, according to the Kennedy formula
• However, some sediment deposition might occur, assuming the Kennedy
formula is correct for these site-specific conditions
• Recall that the Kennedy formula is 100% empirical

Lacey Method

• The average particle diameter is given as dm = 2 mm

• Then, the “f” value from Lacey is:

f = 1.76 2 = 2.49

• Longitudinal bed slope for “regime” flow:

⎛ f2/ 3 ⎞ ⎛ 2.492 / 3 ⎞
S = 0.000547 ⎜ 1/ 6 ⎟ = 0.000547 ⎜ = 0.00037
⎜Q ⎟ ⎜ (12 * 35.31)1/ 6 ⎟⎟
⎝ ⎠ ⎝ ⎠

• This is greater than the specified design bed slope of 0.0001

• Thus, Lacey’s method would predict some scouring at the design discharge
• Note that other relationships from the Lacey method could also be checked
• However, recall that the Lacey method is 100% empirical

Maximum Velocity Method

• Refer to the tables from the lecture notes

• From Fortier & Scobey, use the column for “water with colloidal silt”: the value
for fine sand is Vmax = 2.5 fps (0.76 m/s)
7
 • The following table gives average velocity (m/s) at a channel cross section for
uniform flow conditions, where V = Q/A, and Q = 12 m3/s:

base width, b (m)

m 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
3.487 0.58 0.58 0.57 0.57 0.57 0.57 0.57 0.56 0.56 0.56 0.55
3.271 0.59 0.58 0.58 0.58 0.58 0.58 0.57 0.57 0.57 0.56 0.56
3.078 0.59 0.59 0.59 0.59 0.58 0.58 0.58 0.58 0.57 0.57 0.57
2.904 0.60 0.60 0.60 0.59 0.59 0.59 0.58 0.58 0.58 0.58 0.57
2.747 0.61 0.60 0.60 0.60 0.60 0.59 0.59 0.59 0.58 0.58 0.58
2.605 0.61 0.61 0.61 0.60 0.60 0.60 0.60 0.59 0.59 0.58 0.58
2.475 0.62 0.62 0.61 0.61 0.61 0.60 0.60 0.60 0.59 0.59 0.59
2.356 0.62 0.62 0.62 0.62 0.61 0.61 0.61 0.60 0.60 0.59 0.59
2.246 0.63 0.63 0.62 0.62 0.62 0.61 0.61 0.61 0.60 0.60 0.59
2.145 0.63 0.63 0.63 0.62 0.62 0.62 0.61 0.61 0.61 0.60 0.60
2.050 0.64 0.64 0.63 0.63 0.63 0.62 0.62 0.61 0.61 0.60 0.60
Note: values in italics are average velocity (m/s)

• In this case, all average velocity values within the feasible range (tractive force)
of m and b are below 0.76 m/s
• Thus, no scouring is to be expected, although perhaps some sediment
deposition might occur, according to Fortier & Scobey
• Again, the maximum velocity method is 100% empirical and will not provide
accurate results in all cases
• From the USBR, an average particle diameter of 2 mm falls under the “coarse
sand” category, giving a maximum velocity of Vmax = 1.8 fps (0.55 m/s)
• This is more restrictive than the Vmax from Fortier & Scobey, and also exceeds
almost all of the velocity values from the above table, but not by very much

8
 BIE 5300/6300 Assignment #8
Culvert Design
16 Nov 04 (due 23 Nov 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments.

Given:

ƒ A culvert is to be designed to carry cross-drainage water under a new irrigation

canal (see the drawing below).
ƒ The outlet of the culvert will protrude through the retaining wall on the downhill
side.
ƒ The canal will be concrete lined, with a lining thickness of 2.5 inches.

ƒ The natural channel (a gulch) which carries the cross-drainage flow has a well-
defined and fairly stable cross-section.
ƒ The alignment of the gulch is orthogonal to the canal alignment.
ƒ The design flow rate for the culvert will be 100 cfs.
ƒ Circular concrete pipe will be used for the culvert barrel, and it comes with inside
diameters of 2.0 and 3.0 ft.
ƒ Let the upper top of the inlet to the culvert barrel be located at or just below
elevation 1,257.70 ft (this will require some excavation work just upstream of the
culvert inlet, but you don’t have enough information here to know how much
excavation will be needed).
ƒ Let the upper top of the outlet from the culvert barrel be located at or just below
elevation 1,249.63 ft.
Required:

ƒ Design a culvert for cross-drainage under the canal.

ƒ Follow USBR design guidelines, as given in the “Small Canal Structures” book.
ƒ Use English units for this design.
ƒ Attempt to make the design so that an energy dissipation structure is not required
at the culvert outlet. Don’t have supercritical flow at the outlet.
ƒ Determine the number of culvert barrels (in parallel) for this design.
ƒ Determine the slope(s) of the culvert barrel from inlet to outlet.
ƒ Specify collar size and locations.
ƒ Specify the total length of concrete pipe required for the barrel in your design.
ƒ Specify a standard USBR design for the inlet & outlet transitions.
ƒ Make a drawing showing the location of the culvert barrel relative to the canal
cross section and retaining wall, and the location of collars.
ƒ Show your calculations and units, indicate any important assumptions, and
provide any important comments you might have about the design.

Solution:

1. Determine the Required Pipe Size

• According to USBR design guidelines, use a maximum average barrel velocity of

10.0 fps
• Then, for the design discharge of 100 cfs:

4Q 4(100)
D= = = 3.57 ft (1)
πV π(10)

• The available pipe sizes are 2 and 3 feet, inside diameter.

• Try two 2-ft pipes:

0.5(100 cfs)
V= = 15.9 fps (2)
⎛ π(2)2 ⎞
⎜ 4 ⎟
⎝ ⎠

which is too high.

• Try two 3-ft pipes:

0.5(100 cfs)
V= = 7.07 fps (3)
⎛ π(3)2 ⎞
⎜ 4 ⎟
⎝ ⎠
 which is acceptable. Therefore, use two 3-ft pipes for this culvert design, giving
two barrels.

2. Determine the Energy Loss Gradient

• With the full pipe flow impending, the energy loss gradient can be estimated by
the Manning equation for open-channel flow, in which h = D
• Use a Manning n value of 0.015 for new concrete pipe, with a slight safety factor
for aging:

Wp = πD = π(3.0) = 9.425 ft (4)

πD2 π(3.0)2
A= = = 7.069 ft (5)
4 4
2⎛ Wp4 / 3 ⎞ ⎛ (50)(0.015) ⎞2 ⎛ (9.425)4 / 3 ⎞
⎛ Qn ⎞ ⎜ ⎟=
Sf = ⎜ ⎟ ⎟ ⎜⎜ 10 / 3 ⎟⎟
= 0.00744
⎜ A10 / 3 ⎟ ⎜⎝
(6)
⎝ 1.49 ⎠ ⎝ ⎠
1.49 ⎠ ⎝ (7.069) ⎠

where half of the design flow rate is used per barrel.

3. Determine the Critical Slope

• For critical flow, the Froude number is equal to unity:

Q2Tc
Fr2 = = 1.0 (7)
gA3c

• For circular pipes, the following definitions apply:

⎛ 2h ⎞
βc = 2cos −1 ⎜ 1 − c ⎟ (8)
⎝ D ⎠

⎛β ⎞
Tc = Dsin ⎜ c ⎟ (9)
⎝ 2⎠

D2
Ac = (βc − sinβc ) (10)
8
• Solve for depth, h, such that Fr2 = 1.0 for Q = 50 cfs and D = 3.0 ft (for English
units, g ≈ 32.2 ft/s2)
 • Using the Newton method, βc = 4.268 rad, and hc = 2.301 ft
• Calculate the energy loss gradient (critical slope) corresponding to this depth
• For a depth of 2.301 ft, the flow cross-sectional area is 5.818 ft2, and the
wetted perimeter is 6.402 ft
• Applying the Manning equation:

2⎛ 4/3 ⎞
⎛ (50)(0.015) ⎞ ⎜ ( 6.402 )
( Sf )crit =⎜ ⎟ ⎟ = 0.00850 (11)
⎝ 1.49 ⎠ ( 5.818 )
⎜ 10 / 3 ⎟
⎝ ⎠
• If the slope of the pipe is 0.00850 or greater, critical flow can occur

4. Determine the Minimum Upstream Pipe Slope

• According to USBR guidelines, the top of the culvert barrel (pipe) should clear
the bottom of the canal by at least 0.5 ft
• Note that if more information were provided about the uphill topography, it
would also be possible to move the culvert inlet uphill, away from the canal
berm
• The elevation of the inside top of the barrel at the culvert inlet is 1,257.70 ft,
as specified above
• Assume a pipe thickness of 2 inches
• Recall the 2.5-inch canal lining thickness (specified)
• The elevation of the inside top of the barrel at the right canal base:

1,252.80 - (2.5+2.0)/12 - 0.50 = 1,251.93 ft (12)

• The horizontal distance of the steep descending part of the culvert barrel:

1.732(1,261.08 - 1,257.70) + 10.00 + 0.268(1,261.08 - 1,252.80) = 18.07 ft (13)

• The minimum slope of the steep descending part of the culvert barrel:

1,257.70 − 1,251.93 5.77

slope = = = 0.319 (14)
18.07 18.07
• It is also necessary to check that the slope of the downstream pipe does not
exceed the critical slope

5. Determine the Downstream Pipe Slope

• The flatter downstream part of the culvert barrel would traverse a horizontal
distance of:

Length = 12.00 + 0.268(1,261.08 - 1,252.80) + 4.90 + 0.88 +

 (1,261.08 - 1,249.63)(3.28 - 0.88)/18.00 = 21.53 ft

• The change in elevation over this distance will be 1,251.93 - (1,249.63 - 2/12)
= 1,251.93 - 1,249.46 = 2.47 ft (where the 2/12 value is the assumed 2 inches
of pipe wall thickness)
• Then, the slope of the flatter downstream part of the barrel would be
2.47/21.53 = 0.115
• Also, the total horizontal distance is, then: 18.07 + 21.53 = 39.60 ft

• This slope is greater than the critical slope, and is not acceptable because it
would cause supercritical flow throughout, from inlet to outlet, causing erosion
downstream (unless erosion protection is used)

• Use the USBR-recommended downstream slope of 0.005 (0.5%), which is

less than the calculated critical slope of 0.850%
• To accomplish this, the upstream (steep) portion of the culvert pipe can be
extended further in the downstream direction (to the left)
• Linear equations can be written for the inside tops of the upstream &
downstream barrel segments:

Upstream: .................................. y = 0.319x + 1,245.07

Downstream:.............................. y = 0.005x + 1,249.46

where the intercept in the upstream equation is calculated as 1,257.70 -

(0.319)(39.60) = 1,245.07 ft

• Solving the two linear equations for distance:

1,249.46 − 1,245.07 4.39

x= = = 13.98 ft (15)
0.319 − 0.005 0.314

• This is the distance from the outlet at which the inside top of the upstream
pipe intersects the inside top of the downstream pipe
• The elevation of the intersection point is y = 0.319(13.98) + 1,245.07 =
1,249.53 ft, and the steep part of the barrel has the same slope as before
• Below, a profile of the inside top of the culvert barrel is shown
 1259

1258
Both steep
1257
Downstream mild
1256
Elevation (ft)

1255

1254

1253

1252

1251

1250

1249
0 5 10 15 20 25 30 35 40
Horizontal distance (ft)

6. Determine the Pipe Length

• The approximate length of the upstream (steep) pipe is:

Lus = ( 39.60 − 13.98 )2 + (1,257.70 − 1,249.53 )2 = 26.89 ft (16)

• The approximate length of the downstream (mild) pipe is:

Lds = (13.98 )2 + (1,249.53 − 1,249.46 )2 = 13.98 ft (17)

• A total length of 2(26.89 + 13.98) ≈ 82 ft of 3-ft diameter pipe is needed

7. Specify Inlet and Outlet Types

• The inlet and outlet can be USBR Type 1 (well-defined earthen section)
• The capacity of the downstream (mild) part of the barrel at impending full pipe
flow (but assuming open-channel flow) is approximated as:

1.49 (7.069)5 / 3
Q= 0.005 = 41 cfs (18)
0.015 (9.425)2 / 3
 • Then, at the design capacity of 50 cfs (per barrel), the downstream portion of the
barrel would flow full and there would be a hydraulic jump inside the upstream
(steep) part of the barrel
• The outlet velocity would be approximately 7.07 fps (see Eq. 3 above)
• An energy dissipation structure at the outlet is not needed (because the outlet
velocity will be < 15 fps)

8. Specify Collar Placement and Size

• The standard USBR culvert design, calling for two collars under the downhill
canal bank, and one collar under the uphill bank, is not appropriate in this design
because of the downstream retaining wall
• Use a single collar under the upstream canal bank, with Y = ½D = 1.5 ft, and a
thickness of 6 inches
• Note that some excavation will be required at the downhill side of the retaining
wall because the top of the culvert barrel has been set at the elevation of
1,249.63 (the inside top is about 2 inches lower, in this design)

9. Side View Drawing

4.90 ft 10.00 ft elev 1,261.08 ft

0.88 ft
elev 1,257.70 ft

r face
ground su
original barr
e l

elev 1,249.46 ft elev 1,249.53 ft

collar
barrel

13.98 ft

39.60 ft
Sprinkle & Trickle Irrigation
Lecture Notes
BIE 5110/6110
Fall Semester 2004

Biological and Irrigation Engineering Department

Utah State University, Logan, Utah
Merkley & Allen Page 2 Sprinkle & Trickle Irrigation Lectures
 Preface
These lecture notes were prepared by Gary P. Merkley of the Biological and
Irrigation Engineering (BIE) Department at USU, and Richard G. Allen of the
University of Idaho, for use in the BIE 5110/6110 courses. The notes are intended
to supplement and build upon the material contained in the textbook Sprinkle and
Trickle Irrigation by Jack Keller and Ron D. Bliesner (Chapman-Hall Publishers
1990). Due to the close relationship between the lecture notes and the textbook,
some equations and other material presented herein is taken directly from Keller
and Bliesner (1990) – in these instances the material is the intellectual property of
the textbook authors and should be attributed to them. In all other cases, the
material contained in these lecture notes is the intellectual property right of G.P.
Merkley and R.G. Allen.

Copyright Notice

This material has been duplicated by special permission of the copyright

holders. It is not to be duplicated or used for purposes other than learning resource
support for Utah State University. Any violation of this agreement is punishable
under existing copyright laws and may include severe fines and or imprisonment.

© Copyright 1990 to 2004

These lecture notes are formatting for printing on both sides of the page, with
odd-numbered pages on the front. Each lecture begins on an odd-numbered
page, so some even-numbered pages are blank.

Sprinkle & Trickle Irrigation Lectures Page 3 Merkley & Allen

Merkley & Allen Page 4 Sprinkle & Trickle Irrigation Lectures
 Contents
No. Lecture Title Page

Sprinkle Irrigation

1 Course Introduction.................................................................................... 9
2 Types of Sprinkler Systems; Soil-Water-Plant Relationships;
Planning Factors ...................................................................................... 11
3 Sprinkler Characteristics; Application Rates ............................................ 25
4 Set Sprinkler Uniformity & Efficiency........................................................ 39
5 Layout of Laterals & Mainline for Set Sprinklers; Pipe Hydraulics............ 51
6 Economic Pipe Selection Method............................................................. 59
7 Set Sprinkler Lateral Design .................................................................... 71
8 Set Sprinkler Lateral Design & Analysis................................................... 85
9 Mainline Pipe Design ............................................................................. 101
10 Minor Losses, Pressure Requirements & Pumps................................... 111
11 Pumps & System Curves; Affinity Laws & Cavitation ............................. 123
12 Center Pivot Design & Operation ........................................................... 145
13 Center Pivot Nozzling & Hydraulic Analysis ........................................... 155
14 Center Pivot Uniformity Evaluation; Linear Move Systems .................... 169
15 Maximizing Linear Move Field Length; Design Example ........................ 179

Trickle Irrigation
16 Components & Layout; Pressure Control & Flow Regulation ................. 189
17 Filtration for Trickle Irrigation Systems ................................................... 197
18 Trickle Irrigation Planning Factors; Salinity in Trickle Irrigation .............. 207
19 Water Requirements; Coefficient of Variation & System Capacity ......... 215
20 Emitter Selection & Design; Design Approach & Example ..................... 225
21 Pipe Specifications & Lateral Design; Manifold Location ....................... 231
22 Numerical Solution for Manifold Location; Derivations ........................... 241
23 Manifold Hydraulic Design ..................................................................... 253
24 Hydraulic Design of Mainline & Supply Line........................................... 275

Note: Equations are numbered consecutively in these lecture

notes as (xxx). Equations with the (xxx.xx) format refer to those
found in the textbook by Keller & Bliesner.

Sprinkle & Trickle Irrigation Lectures Page 5 Merkley & Allen

Merkley & Allen Page 6 Sprinkle & Trickle Irrigation Lectures
Units, Constants and Conversions
28.35 g/oz
15.85 gpm/lps (= 60/3.785)
7.481 gallons/ft3
448.86 gpm/cfs (= 7.481*60)
3.7854 litre/gallon

6.89 kPa/psi
1 cb = 1 kPa
10 mb/kPa, or 100 kPa/bar
2.308 ft/psi, or 9.81 kPa/m (head of water)
14.7 psi = 101.3 kPa = 10.34 m (head of water) = 1,013 mbar = 1 atm
62.4 lbs/ft3, or 1000 kg/m3 (max density of pure water at 4°C)
0.1333 kPa/mmHg

1 ppm ≈ 1 mg/liter (usually)

1 mmho/cm = 1 dS/m = 550 to 800 mg/liter

0.7457 kW/HP
1 langley = 1 cal/cm2
0.0419 MJ/m2 per cal/cm2

0.3048 m/ft
1.609 km/mile
2.471 acre/ha
43,560 ft2/acre
1,233 m3/acre-ft

57.2958 degrees/radian
π ≈ 3.14159265358979323846
e ≈ 2.71828182845904523536

ºC = (ºF – 32)/1.8
ºF = 1.8(ºC) + 32

Ratio of weight to mass at sea level and 45° latitude: g = 9.80665 m/s2

PVC = Polyvinyl chloride

PE = Polyethylene
ABS = Acrylonitrile-Butadiene-Styrene

Sprinkle & Trickle Irrigation Lectures Page 7 Merkley & Allen

Merkley & Allen Page 8 Sprinkle & Trickle Irrigation Lectures
Equation Chapter 1 Section 1Lecture 1
Course Introduction

I. Course Overview

• Design of sprinkle and trickle systems – perhaps the most comprehensive

course on the subject anywhere
• Previously, this was two separate courses at USU
• Everyone must be registered at least for audit
• Prerequisites: BIE 5010/6010; computer programming; hydraulics
• There will be two laboratory/field exercises
• Review of lecture schedules for sprinkle and trickle

II. Textbook and Other Materials

• Textbook by Keller and Bliesner

• Two textbooks are on reserve in the Merrill Library
• Lecture notes by Merkley and Allen are required
• We will also use other reference materials during the semester

III. Homework and Design Project

• Work must be organized and neat

• Working in groups is all right, but turn in your own work
• Computer programming and spreadsheet exercises
• Submitting work late (10% per day, starting after class)
• Sprinkle or trickle system design project

IV. Tests, Quizzes, and Grading Policy

• Maybe some quizzes (these will not be announced)

• Two mid-term exams
• Final exam is comprehensive

V. Units

• It is often necessary to convert units in design calculations

• Make it a habit to perform dimensional analysis when using equations; only
in some of the empirical equations will the units not work out correctly

VI. Irrigation Systems

• On-farm level (field)

• Project level (storage, conveyance, tertiary)

Sprinkle & Trickle Irrigation Lectures Page 9 Merkley & Allen

VII. General Types of On-Farm Irrigation Systems

Type U.S. Area World Area

Surface 65% 95%
Sprinkler 30% 3%
Micro Irrigation 3% 1%
Sub-Irrigation 2% 1%
These are approximate percent areas

VIII. Sprinkler Systems

Important Advantages

1. effective use of small continuous streams of water

2. greater application uniformity on non-homogeneous soils (provided there is
no appreciable surface runoff)
3. ability to adequately irrigate steep or undulating topographies w/o erosion
4. good for light and frequent irrigation where surface irrigation may be used
later in the growing season
5. labor is only needed for a short time each day (unless there are many fields)
6. labor can be relatively unskilled (pipe moving)
7. automation is readily available for many sprinkler systems
8. can be effective for weather (micro-climate) modification

Important Disadvantages

1. initial cost can be high (compared to surface irrigation systems) at $500 to

$3500 per ha
2. operating cost (energy) can be high compared to non-pressurized systems,
unless sufficient head is available from a gravity-fed supply
3. water quality can be a problem with overhead sprinklers if water is saline,
and in terms of clogging and nozzle wear. Also, some types of water are
corrosive to sprinkler pipes and other hardware
4. some fruit crops cannot tolerate wet conditions during maturation (unless
fungicides, etc., are used)
5. fluctuating flow rates at the water source can be very problematic
6. irregular field shapes can be difficult to accommodate
7. very windy and very dry conditions can cause high losses
8. low intake rate soils (< 3 mm/hr) cannot be irrigated by sprinkler w/o runoff

IX. Slides of Sprinkler Systems

[these will be shown in class]

Merkley & Allen Page 10 Sprinkle & Trickle Irrigation Lectures

Lecture 2
Types of Sprinkler Systems

I. Sprinkler System Categories

• Two broad categories: set and continuous-move

• Set systems can be further divided into: fixed and periodic-move

II. Set Systems:

Hand-Move

• very common type of sprinkler system

• costs about $30 - $90 per acre, or $75 - $220 per ha
• requires relatively large amount of labor
• laterals are usually aluminum: strong enough, yet light to carry
• usually each lateral has one sprinkler (on riser), at middle or end of pipe
• to move, pull end plug and begin draining of line, then pull apart
• lateral pipe is typically 3 or 4 inches in diameter
• usually for periodic move, but can be set up for a fixed system
• sprinklers are typically spaced at 40 ft along the pipe
• laterals are typically moved at 50- or 60-ft intervals along mainline

End-Tow

• similar to hand-move, but pipes are more rigidly connected

• tractor drags the lateral from position to position, straddling a mainline
• has automatically draining values (open when pressure drops)
• pipe is protected from wear during dragging by mounting it on skid plates or small
wheels
• least expensive of the mechanically-moved systems
• needs a 250-ft (75-m) “turning strip” about the mainline

Side-Roll

• very common in western U.S.

• costs about $150 - $300 per acre, or $360 - $750 per ha
• wheels are usually 65 or 76 inches in diameter

Sprinkle & Trickle Irrigation Lectures Page 11 Merkley & Allen

 • lateral is the axle for the wheels; lateral pipe may have thicker walls adjacent to a
central “mover” to help prevent collapse of the pipe during moving
• uses “movers” or motorized units to roll the lateral; these may be mounted in middle
and or at ends, or may be portable unit that attaches to end of line
• self-draining when pressure drops
• must drain lines before moving, else pipe will break
• windy conditions can cause difficulties when moving the lateral, and can blow empty
lateral around the field if not anchored down
• can have trail tubes (drag lines) with one or two sprinklers each
• need to “dead-head” back to the starting point
• mainline is often portable
• has swivels at sprinkler and trail tube locations to keep sprinklers upright
• low growing crops only (lateral is about 3 ft above ground)
• can be automated, but this is not the typical case

Side-Move

• almost the same as side-roll, but lateral pipe is not axle: it is mounted on A frames
with two wheels each
• clearance is higher than for side-roll
• not as common as side-roll sprinklers

Gun

• 5/8-inch (16 mm) or larger nozzles

• rotate by rocker arm mechanism
• discharge is 100 to 600 gpm at 65 to 100 psi
• large water drops; commonly used on pastures, but also on other crops

Boom

• have big gun sprinklers mounted on rotating arms, on a trailer with wheels
• arms rotate due to jet action from nozzles
• arms supported by cables
• large water drops; commonly used on pastures, but also on other crops

Other Set Sprinklers

• Perforated Pipe
• Hose-Fed Sprinklers
• Orchard Sprinklers

Fixed (Solid-Set) Systems

• enough laterals to cover entire field at same time

• will not necessarily irrigate entire field at the same time, but if you do, a larger
system capacity is needed
• only fixed systems can be used for: frost protection, crop cooling, blossom delay
• easier to automate that periodic-move systems

Merkley & Allen Page 12 Sprinkle & Trickle Irrigation Lectures

 • laterals and mainline can be portable and above ground (aluminum), or permanent
and buried (PVC or steel, or PE)

III. Continuous-Move Systems

Traveler

• could be big gun or boom on platform with wheels

• usually with a big gun (perhaps 500 gpm & 90 psi) sprinkler
• long flexible hose with high head loss
• may reel up the hose or be pulled by a cable
• large water drops; commonly used on pastures, but also on other crops
• some travelers pump from open ditch, like linear moves
• sprinkler often set to part circle so as not to wet the travel path

Center Pivot

• cost is typically $35,000 ($270/ac or $670/ha), plus $15,000 for corner system
• easily automated
• typical maximum (fastest) rotation is about 20 hrs
• don’t rotate in 24-hr increment because wind & evaporation effects will concentrate
• returns to starting point after each irrigation
• typical lateral length is 1320 ft (400 m), or ¼ mile (quarter “section” area)
• laterals are about 10 ft above the ground
• typically 120 ft per tower (range: 90 to 250 ft) with one horsepower electric motors
(geared down)
• IPS 6” lateral pipe is common (about 6-5/8 inches O.D.); generally 6 to 8 inches, but
can be up to 10 inches for 2640-ft laterals
• typical flow rates are 45 - 65 lps (700 to 1000 gpm)
• typical pressures are 140 - 500 kPa (20 to 70 psi)
• older center pivots can have water driven towers (spills water at towers)
• end tower sets rotation speed; micro switches & cables keep other towers aligned
• corner systems are expensive; can operate using buried cable; corner systems don’t
irrigate the whole corner
• w/o corner system, π/4 = 79% of the square area is irrigated
• for 1320 ft (not considering end gun), area irrigated is 125.66 acres
• with corner system, hydraulics can be complicated due to end booster pump
• center pivots are ideal for allowing for effective precipitation
• ignore soil water holding capacity (WHC)
• requires almost no labor; but must be maintained, or it will break down
• can operate on very undulating topography
• known to run over farmers’ pickups (when they leave it out there)!
• many variations: overhead & underneath sprinklers; constant sprinkler spacing;
varied sprinkler spacing; hoses in circular furrows, etc.
• sprinkler nearest the pivot point may discharge only a fine spray; constant radial
velocity but variable tangential speeds (fastest at periphery)
• some center pivots can be moved from field to field

Linear Move

• costs about $40,000 for 400 m of lateral

Sprinkle & Trickle Irrigation Lectures Page 13 Merkley & Allen
 • field must be rectangular in shape
• typically gives high application uniformity
• usually guided by cable and trip switches (could be done by laser)
• usually fed by open ditch with moving pump, requiring very small (or zero slope) in
that direction
• can also be fed by dragging a flexible hose, or by automated arms that move
sequentially along risers in a mainline
• need to “dead-head” back to other side of field, unless half the requirement is
applied at each pass
• doesn’t have problem of variable tangential speeds as with center pivots

IV. LEPA Systems

• Low Energy Precision Application (LEPA) is a concept developed in the mid

to late 1970s in the state of Texas to conserve water and energy in
pressurized irrigation systems
• The principal objective of the technology was to make effective use of all
available water resources, including the use of rainfall and minimization of
evaporation losses, and by applying irrigation water near the soil surface
• Such applications of irrigation water led to sprinkler system designs
emphasizing lower nozzle positions and lower operating pressures, thereby
helping prevent drift and evaporative losses and decreasing pumping costs
• For example, many center pivot systems with above-lateral sprinklers have
been refitted to position sprinkler heads under the lateral, often with lower
pressure nozzle designs
• The commercialization of the LEPA technology has led to many modifications
and extensions to the original concept, and is a term often heard in
discussions about agricultural sprinkler systems
• The LEPA concept can be applied in general to all types of sprinkler
systems, and to many other types of irrigation systems

Merkley & Allen Page 14 Sprinkle & Trickle Irrigation Lectures

Soil-Water-Plant Relationships

I. Irrigation Depth

MAD
dx = Wa Z (1)
100

where dx is the maximum net depth of water to apply per irrigation; MAD is
management allowed deficit (usually 40% to 60%); Wa is the water holding
capacity, a function of soil texture and structure, equal to FC – WP (field
capacity minus wilting point); and Z is the root depth

• For most agricultural soils, field capacity (FC) is attained about 1 to 3 days
after a complete irrigation
• The dx value is the same as “allowable depletion.” Actual depth applied may
be less if irrigation frequency is higher than needed during peak use period.
• MAD can also serve as a safety factor because many values (soil data, crop
data, weather data, etc.) are not precisely known
• Assume that crop yield and crop ET begins to decrease below maximum
potential levels when actual soil water is below MAD (for more than one day)
• Water holding capacity for agricultural soils is usually between 10% and 20%
by volume
• Wa is sometimes called “TAW” (total available water), “WHC” (water holding
capacity), “AWHC” (available water holding capacity)
• Note that it may be more appropriate to base net irrigation depth calculations
on soil water tension rather than soil water content, also taking into account
the crop type – this is a common criteria for scheduling irrigations through the
use of tensiometers

II. Irrigation Interval

• The maximum irrigation frequency is:

dx
fx = (2)
Ud

where fx is the maximum interval (frequency) in days; and Ud is the average

daily crop water requirement during the peak-use period

• The range of fx values for agricultural crops is usually:

0.25 < fx < 80 days (3)

Sprinkle & Trickle Irrigation Lectures Page 15 Merkley & Allen

 • Then nominal irrigation frequency, f’, is the value of fx rounded down to the
nearest whole number of days (
• But, it can be all right to round up if the values are conservative and if fx is
near the next highest integer value
• f’ could be fractional if the sprinkler system is automated
• f’ can be further reduced to account for nonirrigation days (e.g. Sundays),
whereby f ≤ f’
• The net application depth per irrigation during the peak use period is dn =
f’Ud, which will be less than or equal to dx. Thus, dn <= dx, and when dn = dx,
f’ becomes fx (the maximum allowable interval during the peak use period).
• Calculating dn in this way, it is assumed that Ud persists for f’ days, which
may result in an overestimation if f’ represents a period spanning many days

III. Peak Use Period

• Irrigation system design is usually for the most demanding conditions:

• The value of ET during the peak use period depends on the crop type and on
the weather. Thus, the ET can be different from year to year for the same
crop type.
• Some crops may have peak ET at the beginning of the season due to land
preparation requirements, but these crops are normally irrigated by surface
systems.
• When a system is to irrigate different crops (in the same or different
seasons), the crop with the highest peak ET should be used to determine
system capacity.
• Consider design probabilities for ET during the peak use period, because
peak ET for the same crop and location will vary from year-to-year due to
weather variations.
Merkley & Allen Page 16 Sprinkle & Trickle Irrigation Lectures
 • Consider deficit irrigation, which may be feasible when water is very scarce
and or expensive (relative to the crop value). However, in many cases
farmers are not interested in practicing deficit irrigation.

IV. Leaching Requirement

• Leaching may be necessary if annual rains are not enough to flush the root
zone, or if deep percolation from irrigation is small (i.e. good application
uniformity and or efficiency).
• If ECw is low, it may not be necessary to consider leaching in the design
(system capacity).
• Design equation for leaching:

EC w
LR = (4)
5ECe − EC w

where LR is the leaching requirement; ECw is the EC of the irrigation water

(dS/m or mmho/cm); and ECe is the estimated saturation extract EC of the
soil root zone for a given yield reduction value

• Equation 4 is taken from FAO Irrigation and Drainage Paper 29

• When LR > 0.1, the leaching ratio increases the depth to apply by 1/(1-LR);
otherwise, LR does not need to be considered in calculating the gross depth
to apply per irrigation, nor in calculating system capacity:

dn
LR ≤ 0.1: d= (5)
Ea

0.9 dn
LR > 0.1 d= (6)
(1 − LR)Ea

• When LR < 0.0 (a negative value) the irrigation water is too salty, and the
crop would either die or suffer severely
• Standard salinity vs. crop yield relationships (e.g. FAO) are given for
electrical conductivity as saturation extract
• Obtain saturation extract by adding pure water in lab until the soil is
saturated, then measure the electrical conductivity
• Here are some useful conversions: 1 mmho/cm = 1 dS/m = 550 to 800 mg/l
(depending on chemical makeup, but typically taken as 640 to 690). And, it
can usually be assumed that 1 mg/l ≈ 1 ppm, where ppm is by weight (or
mass).

Sprinkle & Trickle Irrigation Lectures Page 17 Merkley & Allen

V. Leaching Requirement Example

Suppose ECw = 2.1 mmhos/cm (2.1 dS/m) and ECe for 10% reduction in crop
yield is 2.5 dS/m. Then,

EC w 2.1
LR = = = 0.20 (7)
5ECe − EC w 5(2.5) − 2.1

Thus, LR > 0.1. And, assuming no loss of water due to application nonuniformity,
the gross application depth is related to the net depth as follows:

dn
d = dn + (LR)d = (8)
1 − LR
and,
dn
d= = 1.25dn (9)
1 − 0.20

See Eq. 5.3 from the textbook regarding nonuniformity losses.

Sprinkle Irrigation Planning Factors

I. Farm Systems vs. Field Systems

• The authors of the textbook only devote a few paragraphs to this topic, but it
is one of great importance
• A complete understanding of the distinctions between farm and field systems
comes only through years of experience
• Variability in design, operation and management conditions is limitless

“A poorly designed system that is well managed can often perform

better than a well designed system that is poorly managed”

• Farm systems may have many field systems

• Planning considerations should include the possibility of future expansions
and extra capacity
• Permanent buried mainlines should generally be oversized to allow for future
needs -- it is much cheaper to put a larger pipe in at the beginning than to
install a secondary or larger line later
• Consider the possibility of future automation
• Consider the needs for land leveling before burying pipes
• How will the system be coordinated over many fields?

Merkley & Allen Page 18 Sprinkle & Trickle Irrigation Lectures

 • What if the cropping patterns change? (tolerance to salinity, tolerance to
foliar wetting, peak ET rate, root depth, need for crop cooling or frost
protection, temporal shifting of peak ET period)
• What if energy costs change?
• What if labor availability and or cost change?
• What if the water supply is changed (e.g. from river to groundwater, or from
old well to new well)?
• What if new areas will be brought into production?

II. Outline of Sprinkler Design Procedure

1. Make an inventory of resources

• visit the field site personally if at all possible, and talk with the farmer
• get data on soil, topography, water supply, crops, farm schedules, climate, energy,
etc.
• be suspicious of parameter values and check whether they are within reasonable
ranges

2. Calculate a preliminary value for the maximum net irrigation depth, dx

3. Obtain values for peak ET rate, Ud, and cumulative seasonal ET, U (Table 3.3)
4. Calculate maximum irrigation frequency, fx, and nominal frequency, f’

• this step is unnecessary for automated fixed systems and center pivots

5. Calculate the required system capacity, Qs

• first, calculate gross application depth, d

• for center pivots use d/f = Ud, and T ≈ 90% of 24 hrs/day = 21.6

6. Determine the “optimum” (or maximum) water application rate

• a function of soil type and ground slope (Table 5.4)

7. Consider different types of feasible sprinkle systems

8. For periodic-move and fixed (solid-set) systems:

(a) determine Se, qa, nozzle size, and P for optimum application rate
(Tables 6.4 to 6.7)
(b) determine number of sprinklers to operate simultaneously to meet Qs
(Nn = Qs/qa) (Chapter 7)
(c) decide upon the best layout of laterals and mainline (Chapter 7)
(d) Adjust f, d, and/or Qs to meet layout conditions
(e) Size the lateral pipes (Chapter 9)
(f) Calculate the maximum pressure required for individual laterals
9. Calculate the mainline pipe size(s), then select from available sizes
10. Adjust mainline pipe sizes according to the “economic pipe selection method”
(Chapter 10)
Sprinkle & Trickle Irrigation Lectures Page 19 Merkley & Allen
11. Determine extreme operating pressure and discharge conditions (Chapter 11)
12. Select the pump and power unit (Chapter 12)
13. Draw up system plans and make a list of items with suggestions for operation

III. Summary

• Note that MAD is not a precise value; actual precision is less than two
significant digits; this justifies some imprecision in other values (don’t try to
obtain very precise values for some parameters when others are only rough
estimates)
• Why use f to determine Qs but f’ to determine net application depth?
(because Qs must be based on gross requirements; not irrigating 24 hrs/day
and 7 days/week does not mean that the crop will not transpire water 7
days/week)
• When determining the seasonal water requirements we subtract Pe from U.
However, to be safe, the value of Pe must be reliable and consistent from
year to year, otherwise a smaller (or zero) value should be used.
• Note that lateral and sprinkler spacings are not infinitely adjustable: they
come in standard dimensions from which designers must choose. The same
goes for pipe diameters and lengths.
• Note that design for peak Ud may not be appropriate if sprinklers are used
only to germinate seeds (when later irrigation is by a surface method).

IV. Example Calculations for a Periodic-Move System

Given:

Crop is alfalfa. Top soil is 1.0 m of silt loam, and subsoil is 1.8 m of clay loam.
Field area is 35 ha. MAD is 50% and ECw is 2.0 dS/m. Application efficiency is
estimated at 80%, and the soil intake rate is 15 mm/hr. Lateral spacing is 15 m and
lateral length is 400 m. Assume it takes ½ hour to change sets. Seasonal effective
rainfall is 190 mm; climate is hot. Assume one day off per week (irrigate only 6
days/week).

From tables in the textbook:

Hot climate, table 3.3 gives............Ud = 7.6 mm/day, and U = 914 mm/season
Top soil, table 3.1 gives ...........................................................Wa = 167 mm/m
Sub soil, table 3.1 gives ...........................................................Wa = 183 mm/m
Root depth, table 3.2 gives .........................................Z = (1.2 + 1.8)/2 = 1.5 m
Salinity for 10% yield reduction, table 3.5 gives ........................ECe = 3.4 dS/m
1. Average water holding capacity in root zone:

top soil is 1.0 m deep; root zone is 1.5 m deep...

Merkley & Allen Page 20 Sprinkle & Trickle Irrigation Lectures

 1.0 (167 ) + (1.5 − 1.0 )(183 )
Wa = = 172.3 mm/m (10)
1.5
2. Max net application depth (Eq. 3.1):

MAD ⎛ 50 ⎞
dx = Wa Z = ⎜ ⎟ (172.3 )(1.5 ) = 129.2 mm (11)
100 ⎝ 100 ⎠

3. Maximum irrigation interval (Eq. 3.2):

dx 129.2 mm
fx = = = 17.0 days (12)
Ud 7.6 mm/day

4. Nominal irrigation interval (round down, or truncate):

f ' = trunc ( fx ) = 17 days (13)

5. Net application depth:

dn = f 'Ud = (17 days ) ( 7.6 mm/day ) = 129.2 mm (14)

6. Operating time for an irrigation:

17 days is just over two weeks, and depending on which day is off, there
could be 3 off days in this period. So, with one day off per week, we will
design the system capacity to finish in 17 - 3 = 14 days. Thus, f = 14 days.
But, remember that we still have to apply 17 days worth of water in these 14
days (we irrigate 6 days/week but crop transpires 7 days/week)

7. Leaching requirement (Eq. 3.3):

ECw 2.0
LR = = = 0.13 (15)
5ECe − ECw 5 ( 3.4 ) − 2.0

LR > 0.1; therefore, use Eq. 5.3 b...

8. Gross application depth (Eq. 5.3b):

0.9dn 0.9 (129.2 )

d= = = 167.1 mm (16)
(1 − LR )(Ea /100 ) (1 − 0.13 )(0.8 )
Sprinkle & Trickle Irrigation Lectures Page 21 Merkley & Allen
9. Nominal set operating time:

With 167.1 mm to apply and a soil intake rate of 15 mm/hr, this gives 11.14
hrs minimum set time (so as not to exceed soil intake rate). Then, we can
make the nominal set time equal to 11.5 hours for convenience. With 0.5 hrs
to move each set, there are a total of 12.0 hrs/set, and the farmer can
change at 0600 and 1800 (for example).

At this point we could take the lateral spacing, Sl, sprinkler spacing, Se, and
actual application rate to determine the flow rate required per sprinkler.

10. Sets per day:

From the above, we can see that there would be two sets/day.

11. Number of sets per irrigation:

(14 days/irrigation)(2 sets/day) = 28 sets

12. Area per lateral per irrigation:

Lateral spacing on mainline is Sl = 15 m. Lateral length is 400 m. Then, the

area per lateral is:

(15 m/set)(28 sets)(400 m/lateral) = 16.8 ha/lateral

13. Number of laterals needed:

35 ha
= 2.08 laterals (17)
16.8 ha/lateral
Normally we would round up to the nearest integer, but because this is so
close to 2.0 we will use two laterals in this design.

14. Number of irrigations per season:

U − Pe 914 mm - 190 mm
= = 5.6 irrigations (18)
dn 129.2 mm/irrig

Thus, it seems there would be approximately six irrigations in a season. But,

the initial Rz value is less than 1.5 m, so there may actually be more than six
irrigations.

Merkley & Allen Page 22 Sprinkle & Trickle Irrigation Lectures

15. System flow capacity (Eq. 5.4):

with 11.5 hours operating time per set and two sets per day, the system runs
23 hrs/day...

Qs = 2.78
Ad
= 2.78
( 35 ha )(167.1 mm ) = 50.5 lps (800 gpm)
fT (14 days )( 23 hrs/day ) (19)

This is assuming no effective precipitation during the peak ET period.

Sprinkle & Trickle Irrigation Lectures Page 23 Merkley & Allen

Merkley & Allen Page 24 Sprinkle & Trickle Irrigation Lectures
Lecture 3
Sprinkler Characteristics

I. Hardware Design Process

1. Sprinkler selection
2. Design of the system layout
3. Design of the laterals
4. Design of the mainline
5. Pump and power unit selection

II. Classification of Sprinklers and Applicability

(see Table 5.1 from the textbook)

• Agricultural sprinklers typically have flow rates from 4 to 45 lpm (1 to 12

gpm), at nozzle pressures of 135 to 700 kPa (20 to 100 psi)
• “Gun” sprinklers may have flow rates up to 2,000 lpm (500 gpm; 33 lps) or
more, at pressures up to 750 kPa (110 psi) or more
• Sprinklers with higher manufacturer design pressures tend to have larger
wetted diameters
• But, deviations from manufacturer’s recommended pressure may have the
opposite effect (increase in pressure, decrease in diameter), and uniformity
will probably be compromised

• Sprinklers are usually made of plastic, brass, and or steel

• Low pressure nozzles save pumping costs, but tend to have large drop sizes
and high application rates
• Medium pressure sprinklers (210 - 410 kPa, or 30 to 60 psi) tend to have the
best application uniformity
• Medium pressure sprinklers also tend to have the lowest minimum
application rates
• High pressure sprinklers have high pumping costs, but when used in
periodic-move systems can cover a large area at each set
• High pressure sprinklers have high application rates

• Rotating sprinklers have lower application rates because the water is only
wetting a “sector” (not a full circle) at any given instance...
• For the same pressure and discharge, rotating sprinklers have larger wetted
diameters
• Impact sprinklers always rotate; the “impact” action on the stream of water is
designed to provide acceptable uniformity, given that much of the water
would otherwise fall far from the sprinkler (the arm breaks up part of the
stream)
• Check out Web sites such as www.rainbird.com
Sprinkle & Trickle Irrigation Lectures Page 25 Merkley & Allen
III. Precipitation Profiles

• Typical examples of low, correct, and high sprinkler pressures (see Fig 5.5).

Pressure is too low

Pressure is OK

Pressure is too high

• The precipitation profile (and uniformity) is a function of many factors:

1. nozzle pressure
2. nozzle shape & size
3. sprinkler head design
4. presence of straightening vanes
5. sprinkler rotation speed
6. trajectory angle
7. riser height
8. wind

• Straightening vanes can be used to compensate for consistently windy

conditions
• Overlapping sprinkler profiles (see Fig. 5.7)

uniform! uniform!
lateral

lateral

lateral

Merkley & Allen Page 26 Sprinkle & Trickle Irrigation Lectures

 • Simulate different lateral spacings by “overlapping” catch-can data in the
direction of lateral movement (overlapping along the lateral is automatically
included in the catch-can data, unless it’s just one sprinkler)

IV. Field Evaluation of Sprinklers

• Catch-can tests are typically conducted to evaluate the uniformities of

installed sprinkler systems and manufacturers’ products
• Catch-can data is often overlapped for various sprinkler and lateral spacings
to evaluate uniformities for design and management purposes
• A computer program developed at USU does the overlapping: CATCH3D;
you can also use a spreadsheet program to simulate overlapping (e.g. Ctrl-C,
Edit | Paste Special, Operation: Add)
• Note that catch-can tests represent a specific wind and pressure situation
and must be repeated to obtain information for other pressures or wind
conditions
• Typical catch-can spacings are 2 or 3 m on a square grid, or 1 to 2 m
spacings along one or more “radial legs”, with the sprinkler in the center
• Set up cans with half spacing from sprinklers (in both axes) to facilitate
overlap calculations
• See Merriam & Keller (1978); also see ASAE S398.1 and ASAE S436
Sprinkle & Trickle Irrigation Lectures Page 27 Merkley & Allen
V. Choosing a Sprinkler

• the system designer doesn’t “design” a sprinkler, but “selects” a sprinkler

• there are hundreds of sprinkler designs and variations from several
manufacturers, and new sprinklers appear on the market often
• the system designer often must choose between different nozzle sizes and
nozzle designs for a given sprinkler head design
• the objective is to combine sprinkler selection with Se and Sl to provide
acceptable application uniformity, acceptable pumping costs, and acceptable
hardware costs
• manufacturers provide recommended spacings and pressures
• there are special sprinklers designed for use in frost control

VI. Windy Conditions

• When winds are consistently recurring at some specific hour, the system can
be shut down during this period (T in Eq. 5.4 is reduced)
• For center pivots, rotation should not be a multiple of 24 hours, even if there
is no appreciable wind (evaporation during day, much less at night)
• If winds consistently occur, special straightening vanes can be used
upstream of the sprinkler nozzles to reduce turbulence; wind is responsible
for breaking up the stream, so under calm conditions the uniformity could
decrease
• For periodic-move systems, laterals should be moved in same direction as
prevailing winds to achieve greater uniformity (because Se < Sl)
• Laterals should also move in the direction of wind to mitigate problems of salt
accumulating on plant leaves
• Wind can be a major factor on the application uniformity on soils with low
infiltration rates (i.e. low application rates and small drop sizes)
• In windy areas with periodic-move sprinkler systems, the use of offset
laterals (½Sl) may significantly increase application uniformity
• Alternating the time of day of lateral operation in each place in the field may
also improve uniformity under windy conditions
• Occasionally, wind can help increase uniformity, as the randomness of wind
turbulence and gusts helps to smooth out the precipitation profile

Wind effects on the diameter of throw:

0-3 mph wind: reduce manufacturer’s listed diameter of throw by 10% for an
effective value (i.e. the diameter where the application of
water is significant)

over 3 mph wind: reduce manufacturer’s listed diameter of throw by an

additional 2.5% for every 1 mph above 3 mph (5.6% for every
1 m/s over 1.34 m/s)
Merkley & Allen Page 28 Sprinkle & Trickle Irrigation Lectures
 In equation form:

For 0-3 mph (0-1.34 m/s):

diam = 0.9 diammanuf (20)

For > 3 mph (> 1.34 m/s):

⎣ (
diam = diammanuf ⎡0.9 − 0.025 windmph − 3 ⎤
⎦ ) (21)
or,
diam = diammanuf ⎡⎣0.9 − 0.056 ( windm / s − 1.34 ) ⎤⎦ (22)

Example: a manufacturer gives an 80-ft diameter of throw for a certain sprinkler

and operating pressure. For a 5 mph wind, what is the effective diameter?

80 ft - (0.10)(0.80) = 72 ft (23)

72 ft - (5 mph - 3 mph)(0.025)(72 ft) = 68 ft (24)

or,
diam = 80(0.9-0.025(5-3))=68 ft (25)

VII. General Spacing Recommendations

• Sprinkler spacing is usually

rectangular or triangular
• Triangular spacing is more common
under fixed-system sprinklers
• Sprinkler spacings based on average
(moderate) wind speeds:

1. Rectangular spacing is 40% (Se)

by 67% (Sl) of the effective
diameter
2. Square spacing is 50% of the
effective diameter
3. Equilateral triangle spacing is
62% of the effective diameter
[lateral spacing is 0.62 cos (60°/2)
= 0.54, or 54% of the effective
diameter, Deffec]

• See Fig. 5.8 about profiles and spacings

Sprinkle & Trickle Irrigation Lectures Page 29 Merkley & Allen

VIII. Pressure-Discharge Relationship

• Equation 5.1:
q = Kd P (26)

where q is the sprinkler flow rate; Kd is an empirical coefficient; and P is the

nozzle pressure

• The above equation is for a simple round orifice nozzle

• Eq. 5.1 can be derived from Bernoulli’s equation like this:

P V2 q2
= = (27)
γ 2g 2gA 2

2gA 2P
= Kd P = q (28)
ρg

where the elevations are the same (z1 = z2) and the conversion through the
nozzle is assumed to be all pressure to all velocity

• P can be replaced by H (head), but the value of Kd will be different

• Eq. 5.1 is accurate within a certain range of pressures
• See Table 5.2 for P, q, and Kd relationships

• Kd can be separated into an orifice coefficient, Ko, and nozzle bore area, A:

q = KoA P (29)

whereby,
Ko = 2 / ρ (30)

where the value of Ko is fairly consistent across nozzle sizes for a specific
model and manufacturer

Merkley & Allen Page 30 Sprinkle & Trickle Irrigation Lectures

 • From Table 5.2 in the textbook, the values of Ko are as follows:

Flow Rate Head or Pressure Nozzle Bore Ko

q H or P d
lps m mm 0.00443
lps kPa mm 0.00137
lpm m mm 0.258
lpm kPa mm 0.0824
gpm ft inch 24.2
gpm psi inch 36.8

• Similar values can be determined from manufacturer’s technical information

• Note also that nozzle diameter (bore) can be determined by rearranging the
above equation as follows:

4q
d= (31)
πK o P

• The value of d can then be rounded up to the nearest available diameter

(64ths of an inch, or mm)
• Then, either P or q are adjusted as necessary in the irrigation system design
• Below is a sample pressure versus discharge table for a RainBird© sprinkler

Sprinkle & Trickle Irrigation Lectures Page 31 Merkley & Allen

Application Rates

I. Flow Control Nozzles

• More expensive than regular nozzles

(compare $0.60 for a brass nozzle to
about $2.70 for a flow control nozzle)
• May require more frequent
maintenance
• The orifice has a flexible ring that
decreases the opening with higher
pressures, whereby the value of
A P in the equation remains
approximately constant
• It can be less expensive to design
laterals and mainline so that these types of nozzles are not required, but this
is not always the case
• FCNs are specified for nominal discharges (4, 4.5, 4.8, 5.0 gpm, etc.)
• The manufacturer’s coefficient of variation is about ±5% of q; don’t use FCNs
unless pressure variation is greater than about 10% (along lateral and for
different lateral positions)

1.10P ≈ 1.05 P (32)

II. Low-Pressure Sprinklers

1. Pressure alone is not sufficient to break up the stream in a standard nozzle

design for acceptable application uniformity
2. Need some mechanical method to reduce drop sizes from the sprinkler:

• pins that partially obstruct the stream of water

• sharp-edged orifices
• triangular, rectangular, oval nozzle shapes

3. Some sprinkler companies have invested much into the design of such
devices for low-pressure sprinkler nozzles
4. Low-pressure nozzles can be more expensive, possibly with reduced
uniformity and increased application rate, but the trade-off is in operating cost

Merkley & Allen Page 32 Sprinkle & Trickle Irrigation Lectures

III. Gross Application Depth

dn
d= , for LR ≤ 0.1 (33)
Epa
where Epa is the “designer” application efficiency (decimal; Eq. 6.9). And,

0.9 dn
d= , for LR > 0.1 (34)
(1 − LR)Epa

• The gross application depth is the total equivalent depth of water which must
be delivered to the field to replace (all or part of) the soil moisture deficit in
the root zone of the soil, plus any seepage, evaporation, spray drift, runoff
and deep percolation losses
• The above equations for d presume that the first 10% of the leaching
requirement will be satisfied by the Epa (deep percolation losses due to
application variability). This presumes that areas which are under-irrigated
during one irrigation will also be over-irrigated in the following irrigation, or
that sufficient leaching will occur during non-growing season (winter) months.
• When the LR value is small (ECw ≤ ½ECe), leaching may be accomplished
both before and after the peak ET period, and the first equation (for LR ≤ 0.1)
can be used for design and sizing of system components. This will reduce
the required pipe and pump sizes because the “extra” system capacity during
the non-peak ET periods is used to provide water for leaching.

IV. System Capacity

• Application volume can be expressed as either Qt or Ad, where Q is flow

rate, t is time, A is irrigated area and d is gross application depth
• Both terms are in units of volume
• Thus, the system capacity is defined as (Eq. 5.4):

Ad
Qs = K (35)
fT
where,

Qs = system capacity;
T = hours of system operation per day (obviously, T≤ 24; also, t = fT)
K = coefficient for conversion of units (see below)
d = gross application depth (equals Ud/Eff during f’ period)
f = time to complete one irrigation (days); equal to f’ minus the days
off
A = net irrigated area supplied by the discharge Qs

Sprinkle & Trickle Irrigation Lectures Page 33 Merkley & Allen

 Value of K:

• Metric: for d in mm, A in ha, and Qs in lps: K = 2.78

• English: for d in inches, A in acres, and Qs in gpm: K = 453

Notes about system capacity:

• Eq. 5.4 is normally used for periodic-move and linear-move sprinkler

systems
• The equation can also be used for center pivots if f is decimal days to
complete one revolution and d is the gross application depth per
revolution
• For center pivot and solid-set systems, irrigations can be light and
frequent (dapplied < d): soil water is maintained somewhat below field
capacity at all times (assuming no leaching requirement), and there is
very little deep percolation loss
• Also, there is a margin of safety in the event that the pump fails (or the
system is temporarily out of operation for whatever reason) just when
MAD is reached (time to irrigate), because the soil water deficit is never
allowed to reach MAD
• However, light and frequent irrigations are associated with higher
evaporative losses, and probably higher ET too (due to more optimal soil
moisture conditions). This corresponds to a higher basal crop coefficient
(Kcb + Ks), where Ks is increased, and possibly Kcb too.
• When a solid-set (fixed) system is used for frost control, all sprinklers
must operate simultaneously and the value of Qs is equal to the number
of sprinklers multiplied by qa. This tends to give a higher Qs than that
calculated from Eq. 5.4.

V. Set Sprinkler Application Rate

• The average application rate is calculated as (after Eq. 5.5):

3600qR e
I= (36)
SeSl

where I is the application rate (mm/hr); q is the flow rate (lps); Se is the
sprinkler spacing (m); Sl is the lateral spacing (m); and Re is the fraction of
water emitted by the nozzle that reaches the soil (takes into account the
evaporative/wind loss)

• Re is defined in Chapter 6 of the textbook

• The instantaneous application rate for a rotating sprinkler (after Eq. 5.6):

Merkley & Allen Page 34 Sprinkle & Trickle Irrigation Lectures

 3600 qRe
Ii = (37)
⎛ S ⎞
πR 2j ⎜ a ⎟
⎝ 360 ⎠

where Ii is the application rate (mm/hr); Rj is the radius of throw, or wetted

radius (m); and Sa is the segment wetted by the sprinkler when the sprinkler
is not allowed to rotate (degrees)

• Note that due to sprinkler overlap, the

instantaneous application rate may actually be
higher than that given by Ii above
• For a non-rotating sprinkler, the instantaneous
application rate is equal to the average
application rate
• For a rotating sprinkler, the instantaneous
p ray
application rate may be allowed to exceed the s
basic intake rate of the soil because excess
(ponded) water has a chance to infiltrate while the
sprinkler completes each rotation
• See sample calculation 5.3 in the textbook
• Higher pressures can give lower instantaneous application rates, but if the
wetted radius does not increase significantly with an increase in pressure,
the instantaneous rate may increase
• The minimum tangential rotation speed at the periphery of the wetted area
should normally be about 1.5 m/s. For example, for 1 rpm:

(1.5 m / s)(60 s / min)

= 14.3m (radius) (38)
(1rev / min)(2π rad / rev )

• Thus, a sprinkler with a wetted radius of 14.3 m should rotate at least 1

rpm
• “Big gun” sprinklers can rotate slower than 1 rpm and still meet this
criterion

VI. Intake & Optimum Application Rates

• Factors influencing the rate at which water should be applied:

1. Soil intake characteristics, field slope, and crop cover

2. Minimum application rate that will give acceptable uniformity
3. Practicalities regarding lateral movement in periodic-move systems

Sprinkle & Trickle Irrigation Lectures Page 35 Merkley & Allen

 • Impact of water drops on bare soil can cause “surface sealing” effects,
especially on heavy-textured (clayey) soils
• The result is a reduction in infiltration rate due to the formation of a semi-
impermeable soil layer

• Sprinklers typically produce drops from ½ to 5 mm

• Terminal velocity of falling drops is from 2 to 22 m/s
• Water drops from sprinklers typically reach their terminal velocity before
arriving at the soil surface (especially sprinklers with high risers)
• See Tables 5.3 and 5.4 in the textbook

V. Approximate Sprinkler Trajectory

• The trajectory of water from a sprinkler can be estimated according to

physics equations
• The following analysis does not consider aerodynamic resistance nor wind
effects, and is applicable to the largest drops issuing from a sprinkler
operating under a recommended pressure
• Of course, smaller water drops tend to fall nearer to the sprinkler
• In the figure below, Rj refers to the approximate wetted radius of the sprinkler

• If the velocity in the vertical direction at the nozzle, Vy, is taken as zero at
time t1, then,
( Vy )t
1
= V0 sin α − g t1 = 0 (39)

where V0 is the velocity of the stream leaving the nozzle (m/s); α is the angle
of the nozzle; t1 is the time for a drop to travel from the nozzle to the highest
point in the trajectory (s); and g is the ratio of weight to mass (9.81 m/s2)

• Note that the term Vosin α in Eq. 37 is the initial velocity component in the
vertical direction, and the term gt1 is the downward acceleration over time t1
• The above equation can be solved for t1
• The initial velocity, V0, can be calculated based on the sprinkler discharge
and the nozzle diameter
• Values of α can be found from manufacturers’ information
Merkley & Allen Page 36 Sprinkle & Trickle Irrigation Lectures
 • Now, what is the highest point in the trajectory?
• First, solve for t1 in the previous equation:

Vo sin α
t1 = (40)
g
then,

g t12 V02 sin2 α

h1 = V0 sin α t1 − =
2 2g (41)

• Assuming no acceleration in the horizontal direction,

x1 = V0 cos α t1 (42)
solving for h2,
gt 22
h2 = hr + h1 = Vy t 2 +
2 (43)

where hr is the riser height (m); t2 is the time for a drop of water to travel from
the highest point in the trajectory to impact on the ground; and Vy = 0
• Then, x2 is defined as:
2 (hr + h1 )
x 2 = V0 cos α t 2 = V0 cos α (44)
g

And, the approximate wetted radius of the sprinkler is:

R j = x1 + x 2
(45)

• In summary, if air resistance is ignored and the sprinkler riser is truly vertical,
the theoretical value of Rj is a function of:

1. Angle, α
2. Nozzle velocity (qa/A)
3. Riser height, hr

• And, qa is a function of P

Sprinkle & Trickle Irrigation Lectures Page 37 Merkley & Allen

Merkley & Allen Page 38 Sprinkle & Trickle Irrigation Lectures
Lecture 4
Set Sprinkler Uniformity & Efficiency

I. Sprinkler Irrigation Efficiency

1. Application uniformity
2. Losses (deep percolation, evaporation, runoff, wind drift, etc.)

• It is not enough to have uniform application if the average depth is not

enough to refill the root zone to field capacity
• Similarly, it is not enough to have a correct average application depth if the
uniformity is poor
• Consider the following examples:

• We can design a sprinkler system that is capable of providing good

application uniformity, but depth of application is a function of the set time (in
periodic-move systems) or “on time” (in fixed systems)
• Thus, uniformity is mainly a function of design and subsequent system
maintenance, but application depth is a function of management

Sprinkle & Trickle Irrigation Lectures Page 39 Merkley & Allen

II. Quantitative Measures of Uniformity

Distribution uniformity, DU (Eq. 6.1):

⎛ avg depth of low quarter ⎞

DU = 100 ⎜ ⎟ (46)
⎝ avg depth ⎠
• The average of the low quarter is obtained by measuring application from a
catch-can test, mathematically overlapping the data (if necessary), ranking
the values by magnitude, and taking the average of the values from the low
¼ of all values
• For example, if there are 60 values, the low quarter would consist of the 15
values with the lowest “catches”

Christiansen Coefficient of Uniformity, CU (Eq. 6.2):

⎛
⎜ ∑
n
j=1 (
abs z j − m ) ⎟⎞
CU = 100 1.0 − (47)
⎜ ⎟
∑
n
⎜ z ⎟
⎝ j=1 j ⎠
where z are the individual catch-can values (volumes or depths); n is the
number of observations; and m is the average of all catch volumes.

• Note that CU can be negative if the distribution is very poor

• There are other, equivalent ways to write the equation
• These two measures of uniformity (CU & DU) date back to the time of slide
rules (more than 50 years ago; no electronic calculators), and are designed
with computational ease in mind
• More complex statistical analyses can be performed, but these values have
remained useful in design and evaluation of sprinkler systems
• For CU > 70% the data usually conform to a normal distribution, symmetrical
about the mean value. Then,

⎛ avg depth of low half ⎞

CU ≈ 100 ⎜ ⎟ (48)
⎝ avg depth ⎠
another way to define CU is through the standard deviation of the values,

⎛ σ 2⎞
CU = 100 ⎜ 1.0 − ⎟⎟ (49)
⎜ m π
⎝ ⎠

Merkley & Allen Page 40 Sprinkle & Trickle Irrigation Lectures

 where σ is the standard deviation of all values, and a normal distribution is
assumed (as previously)

• Note that CU = 100% for σ = 0

• The above equation assumes a normal distribution of the depth values,
whereby:

∑ z − m = nσ 2/ π (50)

• By the way, the ratio σ/m is known in statistics as the coefficient of variation
• Following is the approximate relationship between CU and DU:

CU ≈ 100 − 0.63(100 − DU) (51)

or,
DU ≈ 100 − 1.59(100 − CU) (52)

• These equations are used in evaluations of sprinkler systems for both design
and operation
• Typically, 85 to 90% is the practical upper limit on DU for set systems
• DU > 65% and CU > 78% is considered to be the minimum acceptable
performance level for an economic system design; so, you would not
normally design a system for a CU < 78%, unless the objective is simply to
“get rid of water or effluent” (which is sometimes the case)
• For shallow-rooted, high value crops, you may want to use DU > 76% and
CU > 85%

III. Alternate Sets (Periodic-Move Systems)

• The effective uniformity (over multiple irrigations) increases if “alternate sets”

are used for periodic-move systems (½Sl)
• This is usually practiced by placing laterals halfway between the positions
from the previous irrigation, alternating each time
• The relationship is:

CUa ≈ 10 CU
(53)
DUa ≈ 10 DU

• The above are also valid for “double” alternate sets (Sl/3)
• Use of alternate sets is a good management practice for periodic-move
systems
• The use of alternate sets approaches an Sl of zero, which simulates a
continuous-move system

Sprinkle & Trickle Irrigation Lectures Page 41 Merkley & Allen

IV. Uniformity Problems

• Of the various causes of non-uniform sprinkler application, some tend to

cancel out with time (multiple irrigations) and others tend to concentrate (get
worse)
• In other words, the “composite” CU for two or more irrigations may be (but
not necessarily) greater than the CU for a single irrigation

1. Factors that tend to Cancel Out

• Variations in sprinkler rotation speed

• Variations in sprinkler discharge due to wear
• Variations in riser angle (especially with hand-move systems)
• Variations in lateral set time

2. Factors that may both Cancel Out and Concentrate

• Non-uniform aerial distribution of water between sprinklers

3. Factors that tend to Concentrate

• Variations in sprinkler discharge due to elevation and head loss

• Surface ponding and runoff
• Edge effects at field boundaries

V. System Uniformity

• The uniformity is usually less when the entire sprinkler system is considered,
because there tends to be greater pressure variation in the system than at
any given lateral position.

⎡1
( ⎤
system CU ≈ CU ⎢ 1 + Pn / Pa ⎥
⎣2 ⎦
) (54)

⎡1
( ⎤
system DU ≈ DU ⎢ 1 + 3 Pn / Pa ⎥
⎣4 ⎦
) (55)

where Pn is the minimum sprinkler pressure in the whole field; and Pa is the
average sprinkler pressure in the entire system, over the field area.

• These equations can be used in design and evaluation

• Note that when Pn = Pa (no pressure variation) the system CU equals the CU
• If pressure regulators are used at each sprinkler, the system CU is
approximately equal to 0.95CU (same for DU)
Merkley & Allen Page 42 Sprinkle & Trickle Irrigation Lectures
 • If flexible orifice nozzles are used, calculate system CU as 0.90CU (same for
DU)
• The Pa for a system can often be estimated as a weighted average of Pn &
Px:

2Pn + Px
Pa = (56)
3

where Px is the maximum nozzle pressure in the system

VI. Computer Software and Standards

• There is a computer program called “Catch-3D” that performs uniformity

calculations on sprinkler catch-can data and can show the results graphically
• Jack Keller and John Merriam (1978) published a handbook on the
evaluation of irrigation systems, and this includes simple procedures for
evaluating the performance of sprinkler systems
• The ASAE S436 (Sep 92) is a detailed standard for determining the
application uniformity under center pivots (not a set sprinkler system, but a
continuous move system)
• ASAE S398.1 provides a description of various types of information that can
be collected during an evaluation of a set sprinkler system

Sprinkle & Trickle Irrigation Lectures Page 43 Merkley & Allen

VII. General Sprinkle Application Efficiency

The following material leads up to the development of a general sprinkle

application efficiency term (Eq. 6.9) as follows:

Design Efficiency:

Epa = DEpaReOe
(57)

where DEpa is the distribution efficiency (%); Re is the fraction of applied

water that reaches the soil surface; and Oe is the fraction of water that does
not leak from the system pipes.

• The design efficiency, Epa, is used to determine gross application depth (for
design purposes), given the net application depth
• In most designs, it is not possible to do a catch-can test and data analysis –
you have to install the system in the field first; thus, use the “design
efficiency”
• The subscript “pa” represents the “percent area” of the field that is
adequately irrigated (to dn, or greater) – for example, E80 and DE80 are the
application and distribution efficiencies when 80% of the field is adequately
irrigated
• Question: can “pa” be less than 50%?

Merkley & Allen Page 44 Sprinkle & Trickle Irrigation Lectures

 20% of area underirrigated

U
80% of area overirrigated

C
w
Lo
Relative Applied Depth

h CU
Hig

Desired Net Application Depth

1.0

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Area Receiving at Least the Desired Application

VIII. Distribution Efficiency

• This is used to define the uniformity and adequacy of irrigation

• DE is based on statistical distributions and application uniformity
• For a given uniformity (CU) and a given percent of land adequately irrigated
(equal to or greater than required application depth), Table 6.2 gives values
of DE that determine how much water must be applied in excess of the
required depth so that the given percent of land really does receive at least
the required depth

Percent area adequately irrigated (pa)

CU
95 90 85 80 75 70 65 60 55 50
94 87.6 90.4 92.2 93.7 94.9 96.1 97.1 98.1 99.1 100.0
92 83.5 87.1 89.6 91.6 93.2 94.7 96.1 97.5 98.7 100.0
90 79.4 83.9 87.0 89.4 91.5 93.4 95.2 96.8 98.4 100.0
88 75.3 80.7 84.4 87.3 89.8 92.1 94.2 96.2 98.1 100.0
86 71.1 77.5 81.8 85.2 88.2 90.8 93.2 95.6 97.8 100.0
84 67.0 74.3 79.2 83.1 86.5 89.5 92.3 94.9 97.5 100.0
82 62.9 71.1 76.6 81.0 84.8 88.2 91.3 94.3 97.2 100.0
80 58.8 67.9 74.0 78.9 83.1 86.8 90.3 93.6 96.8 100.0
78 54.6 64.7 71.4 76.8 81.4 85.5 89.4 93.0 96.5 100.0
76 50.5 61.4 68.8 74.7 79.7 84.2 88.4 92.4 96.2 100.0
74 46.4 58.2 66.2 72.6 78.0 82.9 87.4 91.7 95.9 100.0
72 42.3 55.0 63.6 70.4 76.3 81.6 86.5 91.1 95.6 100.0
70 38.1 51.8 61.0 68.3 74.6 80.3 85.5 90.5 95.3 100.0

• See Fig. 6.7

Sprinkle & Trickle Irrigation Lectures Page 45 Merkley & Allen
IX. Wind Drift and Evaporation Losses

• These losses are typically from 5% to 10%, but can be higher when the air is
dry, there is a lot of wind, and the water droplets are small
• Effective portion of the applied water, Re. This is defined as the percentage
of applied water that actually arrives at the soil surface of the irrigated field.
• This is based on:

• climatic conditions
• wind speed
• spray coarseness

• Figure 6.8 gives the value of Re for these different factors

• The Coarseness Index, CI, is defined as (Eq. 6.7):

⎛ P1.3 ⎞
CI = 0.032 ⎜
⎜ B ⎟⎟
(58)
⎝ ⎠

where P is the nozzle pressure (kPa) and B is the nozzle diameter (mm)

CI > 17 17 ≥ CI ≥ 7 CI < 7
fine spray between fine and coarse coarse spray

• When the spray is between fine and coarse, Re is computed as a weighted

average of (Re)fine and (Re)coarse (Eq. 6.8):

(CI − 7) (17 − CI)

Re = (Re )fine + (Re )coarse (59)
10 10

• Allen and Fisher (1988) developed a regression equation to fit the curves in
Fig. 6.8:

Re = 0.976 + 0.005ETo − 0.00017ETo2 + 0.0012 W

−0.00043(CI)(ETo ) − 0.00018(CI)(W) (60)
−0.000016(CI)(ETo )(W)

where ETo is the reference ET in mm/day (grass-based); CI is the

coarseness index (7 ≤ CI ≤ 17); and W is the wind speed in km/hr
• For the above equation, if CI < 7 then set it equal to 7; if CI > 17 then set it
equal to 17

Merkley & Allen Page 46 Sprinkle & Trickle Irrigation Lectures

X. Leaks and Drainage Losses

1. Losses due to drainage of the system after shut-down

• upon shut-down, most sprinkler systems will partially drain

• water runs down to the low elevations and or leaves through
automatic drain valves that open when pressure drops
• fixed (solid-set) systems can have anti-drain valves at sprinklers that
close when pressure drops (instead of opening, like on wheel lines)

2. Losses due to leaky fittings, valves, and pipes

• pipes and valves become damaged with handling, especially with

hand-move and side-roll systems, but also with orchard sprinklers and
end-tow sprinklers
• gaskets and seals become inflexible and fail

• These losses are quantified in the Oe term

• For systems in good condition these losses may be only 1% or 2%, giving an
Oe value of 99% or 98%, respectively
• For system in poor condition these losses can be 10% or higher, giving an Oe
value of 90% or less

XI. General Sprinkle Application Efficiency

• As given above, Eq. 6.9 from the textbook, it is:

Epa = DEpaReOe
(61)

where DEpa is in percent; and Re and Oe are in fraction (0 to 1.0). Thus, Epa
is in percent.

XII. Using CU or DU instead of DEpa

1. Application Efficiency of the Low Quarter, Eq

• Given by Eq. 6.9 when DU replaces DEpa

• Useful for design purposes for medium to high-value crops
• Only about 10% of the area will be under-irrigated
• Recall that DU is the average of low quarter divided by average
2. Application Efficiency of the Low Half, Eh

• Given by Eq. 6.9 when CU replaces DEpa

• Useful for design purposes for low-value and forage crops
Sprinkle & Trickle Irrigation Lectures Page 47 Merkley & Allen
 • Only about 20% of the area will be under-irrigated
• Recall that CU is the average of low half divided by average

XIII. Procedure to Determine CU, Required Pressure, Se and Sl for a Set

System

1. Specify the minimum acceptable Epa and target pa

2. Estimate Re and Oe (these are often approximately 0.95 and 0.99,
respectively)
3. Compute DEpa from Epa, Re and Oe
4. Using DEpa and pa, determine the CU (Table 6.2) that is required to achieve
Epa
5. Compute the set operating time, tso, then adjust f’ and dn so that tso is an
appropriate number of hours
6. Compute qa based on I, Se and Sl (Eq. 5.5)
7. Search for nozzle size, application rate, Se and Sl to obtain the CU
8. Repeat steps 5, 6 and 7 as necessary until a workable solution is found

XIV. How to Measure Re

• The textbook suggests a procedure for estimating Re

• You can also measure Re from sprinkler catch-can data:

1. Compute the average catch depth over the wetted area (if a single
sprinkler), or in the area between four adjacent sprinklers (if in a
rectangular grid)
2. Multiply the sprinkler flow rate by the total irrigation time to get the volume
applied, then divide by the wetted area to obtain the gross average
application depth
3. Divide the two values to determine the effective portion of the applied
water

XV. Line- and Point-Source Sprinklers

• Line-source sprinklers are sometimes used by researches to determine the

effects of varying water application on crop growth and yield
• A line-source sprinkler system consists of sprinklers spaced evenly along a
straight lateral pipe in which the application rate varies linearly with distance
away from the lateral pipe, orthogonally
• Thus, a line-source sprinkler system applies the most water at the lateral
pipe, decreasing linearly to zero to either side of the lateral pipe
• A point-source sprinkler is a single sprinkler that gives linearly-varying
application rate with radial distance from the sprinkler

Merkley & Allen Page 48 Sprinkle & Trickle Irrigation Lectures

 • With a point-source sprinkler, the contours of equal application rate are
concentric circles, centered at the sprinkler location (assuming the riser is
vertical and there is no wind)

Sprinkle & Trickle Irrigation Lectures Page 49 Merkley & Allen

Merkley & Allen Page 50 Sprinkle & Trickle Irrigation Lectures
Lecture 5
Layout of Laterals for Set Sprinklers

I. Selecting Sprinkler Discharge, Spacing, and Pressure

• In Chapter 6 of the textbook there are several tables that provide guidelines
for nozzle sizes for different:

• Wind conditions
• Application rates
• Sprinkler spacings

• For selected values of wind, application rate, and spacing, the tables provide
recommended nozzle sizes for single and double-nozzle sprinklers,
recommended sprinkler pressure, and approximate uniformity (CU)

• Table values are for standard (non-flexible) nozzles

• Table values are for standard sprinkler and lateral spacings
• More specific information can be obtained from manufacturers’ data

• Recall that the maximum application rate is a function of soil texture, soil
structure, and topography (Table 5.4)
• For a given spacing and application rate, the sprinkler discharge, qa, can be
determined from Eq. 5.5

I ( SeSl ) d S SO
qa = = n e l e (62)
3600 3600EpaSto

where qa is in lps; I is in mm/hr; dn is in mm; Sto is the operating time for each
set, in hours; and Sl and Se are in m

• Why is the Oe term included in the above equation? (because Epa includes
Oe, as previously defined, and must be cancelled out when considering an
individual sprinkler)

II. Number of Operating Sprinklers

• After calculating the system capacity and the design flow rate for sprinklers,
the number of sprinklers that will operate at the same time is:

Qs
Nn = (63)
qa

Sprinkle & Trickle Irrigation Lectures Page 51 Merkley & Allen

 where Nn is the minimum number of sprinklers operating, and Qs and qa have
the same units

• It is recommendable to always operate the same number of sprinklers when

the system is running. This practice can help avoid the need for pressure
regulation, and can avoid uniformity problems. It can also help avoid wasting
energy at the pump.
• For odd-shaped fields, and sometimes for rectangular fields, it is not possible
to operate the same number of sprinklers for all sets. In this case, pressure
regulation may be necessary, or other steps can be taken (multiple pumps,
variable-speed motor, variable application rates).

III. Lateral Design Criteria

• Lateral pressure varies from inlet to extreme end due to:

1. friction loss
2. elevation change

• The fundamental basis upon which sprinkler laterals are designed is:

“pressure head variation in the lateral should not exceed

20% of the average design pressure for the sprinklers”

• This is a design assumption that has been used for many years, and is
based on a great deal of experience
• The 20% for pressure variation is not an “exact” value; rather, it is based on
judgment and some cost comparisons
• A designer could change this value, but it would affect system performance
(uniformity), initial system cost, operating cost, and possibly other factors
• Computer programs could be written to search for an “optimal” percent
pressure variation according to initial and operating costs, and according to
crop value -- such an “optimal” value would vary from system to system

IV. Sprinkler Lateral Orientation

• It is usually preferable to run laterals on contours (zero slope) so that

pressure variation in the lateral pipes is due to friction loss only
• It is advantageous to run laterals downhill, if possible, because the gain in
energy due to elevation change will allow longer laterals without violating the
20% rule. But, if the slope is too steep, pressure regulators or flow control
nozzles may be desirable.
• If the ground slope is equal to the friction loss gradient, the pressure in the
lateral will be constant. However, the friction loss gradient is nonlinear
because the flow rate is decreasing with distance along the lateral.
Merkley & Allen Page 52 Sprinkle & Trickle Irrigation Lectures
Sprinkle & Trickle Irrigation Lectures Page 53 Merkley & Allen
 • It is usually not recommendable to run laterals in an uphill direction. In this
case:

1. both friction loss and elevation are working to reduce pressure toward
the end of the lateral, and length is more restricted if the 20% rule is
still used
2. However, for small slopes, running laterals uphill may be required to
reduce the total length of the mainline pipe

• Note that V2/2g in the lateral pipe is normally converted into total head as the
water flows through the nozzle body. Therefore, the velocity head (and EL)
should normally be considered in lateral design. However, since a portion of
the velocity head is lost during deceleration of the water at the entrance into
risers and as turbulence inside the sprinkler head, and since V2/2g in a
lateral pipe is typically small (< 1 ft of head, or 0.2 psi, or 0.3 m head, or 3
kPa), it is normally neglected during design, and the HGL is used.
• Aside from limits on pressure variation, laterals should be oriented so that
they move in the direction of the prevailing winds -- this is because of salinity
problems and application uniformity

• Figure 7.1 gives examples of layouts on different topographies

V. Lateral Sizing Limitations

• lateral pipes can be designed with multiple diameters to accommodate

desirable pressure distributions, but...
• hand-move laterals should have only one or two different pipe sizes to
simplify handling during set changes
• in practice, hand-move systems and wheel lines usually have only one size
of lateral pipe
• some wheel lines, greater than 400 m in length, may have 5-inch pipe near
the inlet and then 4-inch pipe at the end

Layout of Mainline for Set Sprinklers

I. Mainline Layout and Sizing

• if possible, run the mainline up or down slope so the laterals can be on

contours (lateral pressure variation due to friction loss only)
• can also run the mainline along a ridge so the laterals run downhill on both
sides (lateral friction loss partially offset by elevation change)
• should consider possible future expansions when sizing the mainline

Merkley & Allen Page 54 Sprinkle & Trickle Irrigation Lectures

 “Split-Line” Lateral Operation:

• laterals operate on both sides of the mainline

• the mainline can be sized for only half capacity halfway down the mainline if
laterals are run in different directions
• sometimes interferes with cultural practices
• it is convenient to have the water supply in the center of one side of the field,
but this is seldom a design variable (the well is already there, or the canal is
already there)
• may not need pumping if the water supply is at a higher elevation than the
field elevation (e.g. 50 psi = 115 ft or 35 m of head) -- when pumping is not
required, this changes the mainline layout and pipe sizing strategy
• in some cases it will be justifiable to include one or more booster pumps in
the design -- even when the water source is a well (the well pump may not
provide enough pressure for any of the lateral settings)
• we will discuss mainline economics in the next few lectures, then we will look
at mainline design in more detail later

II. Design Variables to Accommodate Layout

• Number of sprinklers operating

• Average application rate
• Gross application depth
• Average sprinkler discharge
• Sprinkler spacing
• Operating hours per day
• Irrigation frequency
• Total operating time (fT)
• System capacity
• Percent probability of rain during peak-use period
• MAD

• It may be necessary to adjust the layout if a suitable combination of the

above variables cannot be found
• Can also use flow control nozzles or pressure regulators to accommodate a
given layout

III. Sample Calculation

• Consider a periodic-move system with Sl = 50 ft, Se = 40 ft, f = 8 days, T =

11.5 hrs @ 2 sets/day, d = 2.7”, and qa = 4.78 gpm
• The field size is 80 acres (½ of a “quarter section”), 2,640 ft on one side and
1,320 ft on the other, rectangular
• The laterals will have to be 1,320-ft long

Sprinkle & Trickle Irrigation Lectures Page 55 Merkley & Allen

 • System capacity:

453(80 ac)(2.7 inch)

Qs = = 532gpm
(8 days)(2 sets / day)(11.5 hrs / set) (64)

• Number of sprinklers operating:

Qs 532
Ns = = = 111 sprinklers
qa 4.78 (65)

• Number of laterals,

1320 ft / lateral
= 33 sprinklers / lateral
40 ft / sprinkler (66)

111 sprinklers
= 3.36 laterals
33 sprinklers / lateral (67)

...so, round up to 4 laterals

• Thus, two laterals on each side of the mainline (symmetry)

1320 ft per lateral pair

= 26.4 (68)
50 ft / position

• Round this up from 26.4 to 27 positions per lateral pair

• This gives 2 x 27 = 54 total lateral positions, and 54/4 = 13.5 sets/lateral
• Use 13 sets for two laterals and 14 sets for the other two laterals
• Then, there will be 14 sets per irrigation, even though the last set will only
have two laterals operating
Merkley & Allen Page 56 Sprinkle & Trickle Irrigation Lectures
 • Adjusted irrigation frequency:

14 sets
f= = 7 days (69)
2 sets / day

• Note that the value of f was for an 8-day interval

• Thus, we need to increase Qs to complete the irrigation in less time

• Adjusted system capacity:

Qs = (4 laterals)(33 sprinklers / lateral)(4.78 gpm / sprinkler)

(70)
= 631 gpm

• Another way to adjust the system capacity:

⎛ 8 days ⎞
Qs = ⎜ ⎟ (532 gpm) = 608 gpm (71)
⎝ 7 days ⎠

• You might say that we are “effectively” finishing in somewhat less than 7
days, because the last set has only two laterals in operation, giving a system
capacity of 608 instead of 631
• Consider this calculation: there are 2 x 13 + 2 x 14 = 54 sets, but the last 2
sets have only 2 laterals. So, (52/54) x 631 = 608 gpm, as calculated above.
• Which is correct?

• There are (52/54)*(4 laterals) = 3.85 laterals operating on average during

each irrigation of the field
• However, you cannot always base the system capacity on the average
number of laterals operating
• The system capacity should be based on the “worst case”, which is when all
four laterals operate simultaneously
• This means that the required capacity is 631 gpm, not 608 gpm
• Note that many farmers will accept some increase in system capital cost to
provide more operational flexibility and safety
• In summary, we have essentially lowered f to accommodate the system
configuration (layout), but:

• same gross depth

• same number of hours per set
• same sprinkler flow rate
• same sprinkler spacing
• increased system capacity

Sprinkle & Trickle Irrigation Lectures Page 57 Merkley & Allen

Pipeline Hydraulics

I. Review

• Read Chapter 8 of the textbook to review the hydraulics of pipelines

• For pipe friction loss we will be using the Hazen-Williams and Darcy-
Weisbach equations
• Be familiar with the Moody diagram, for use with the Darcy-Weisbach
equation
• You can use the Swamee-Jain equation instead of the Moody diagram:

0.25
f= 2
(72)
⎡ ⎛ ε 5.74 ⎞ ⎤
⎢log10 ⎜ + 0.9 ⎟ ⎥
⎣ ⎝ 3.75D NR ⎠ ⎦

which is valid for turbulent flow in the range: 4,000 ≤ NR ≤ 1.0(10)8. The ratio
ε/D is called “relative roughness.” The roughness height, ε, varies widely

• We will also use the Blasius equation (Eq. 8.6) to determine the value of “f,”
in some cases, for “smooth pipes”

Merkley & Allen Page 58 Sprinkle & Trickle Irrigation Lectures

Lecture 6
Economic Pipe Selection Method
I. Introduction

• The economic pipe selection method (Chapter 8 of the textbook) is used to

balance fixed (initial) costs for pipe with annual energy costs for pumping
• With larger pipe sizes the average flow velocity for a given discharge
decreases, causing a corresponding decrease in friction loss
• This reduces the head on the pump, and energy can be saved
• However, larger pipes cost more to purchase

total

minimum
total
Cost

Energy costs =
annualized fixed costs

fixed energy

Pipe Size (diameter)

• To balance these costs and find the minimum cost we will annualize the fixed
costs, compare with annual energy (pumping) costs
• We can also graph the results so that pipe diameters can be selected
according to their maximum flow rate
• We will take into account interest rates and inflation rates to make the
comparison
• This is basically an “engineering economics” problem, specially adapted to
the selection of pipe sizes

• This method involves the following principal steps:

1. Determine the equivalent annual cost for purchasing each available pipe
size
2. Determine the annual energy cost of pumping
Sprinkle & Trickle Irrigation Lectures Page 59 Merkley & Allen
 3. Balance the annual costs for adjacent pipe sizes
4. Construct a graph of system flow rate versus section flow rate on a log-
log scale for adjacent pipe sizes

• We will use the method to calculate “cut-off” points between adjacent pipe
sizes so that we know which size is more economical for a particular flow
rate
• We will use HP and kW units for power, where about ¾ of a kW equals a HP
• Recall that a Watt (W) is defined as a joule/second, or a N-m per second
• Multiply W by elapsed time to obtain Newton-meters (“work”, or “energy”)

II. Economic Pipe Selection Method Calculations

1. Select a period of time over which comparisons will be made between fixed
and annual costs. This will be called the useful life of the system, n, in years.

• The “useful life” is a subjective value, subject to opinion and financial

amortization conditions
• This value could alternatively be specified in months, or other time period,
but the following calculations would have to be consistent with the choice

2. For several different pipe sizes, calculate the uniform annual cost of pipe per
unit length of pipe.

• A unit length of 100 (m or ft) is convenient because J is in m/100 m or

ft/100 ft, and you want a fair comparison (the actual pipe lengths from the
supplier are irrelevant for these calculations)
• You must use consistent units ($/100 ft or $/100 m) throughout the
calculations, otherwise the ∆J values will be incorrect (see Step 11 below)
• So, you need to know the cost per unit length for different pipe sizes
• PVC pipe is sometimes priced by weight of the plastic material (weight
per unit length depends on diameter and wall thickness)
• You also need to know the annual interest rate upon which to base the
calculations; this value will take into account the time value of money,
whereby you can make a fair comparison of the cost of a loan versus the
cost of financing it “up front” yourself
• In any case, we want an equivalent uniform annual cost of the pipe over
the life of the pipeline
• Convert fixed costs to equivalent uniform annual costs, UAC, by using the
“capital recovery factor”, CRF

UAC = P ( CRF ) (73)

Merkley & Allen Page 60 Sprinkle & Trickle Irrigation Lectures

 i (1 + i )
n
CRF = (74)
(1 + i) − 1
n

where P is the cost per unit length of pipe; i is the annual interest rate
(fraction); and n is the number of years (useful life)

• Of course, i could also be the monthly interest rate with n in months, etc.

• Make a table of UAC values for different pipe sizes, per unit length of pipe
• The CRF value is the same for all pipe sizes, but P will change depending
on the pipe size
• Now you have the equivalent annual cost for each of the different pipe
sizes

3. Determine the number of operating (pumping) hours per year, Ot:

(irrigated area)(gross annual depth)

Ot = = hrs / year (75)
(system capacity)

• Note that the maximum possible value of Ot is 8,760 hrs/year (for 365
days)
• Note also that the “gross depth” is annual, so if there is more than one
growing season per calendar year, you need to include the sum of the
gross depths for each season (or fraction thereof)

4. Determine the pumping plant efficiency:

• The total plant efficiency is the product of pump efficiency, Epump, and
motor efficiency, Emotor

Ep = EpumpEmotor
(76)

• This is equal to the ratio of “water horsepower”, WHP, to “brake

horsepower”, BHP (Epump ≡ WHP/BHP)

Sprinkle & Trickle Irrigation Lectures Page 61 Merkley & Allen

 • Think of BHP as the power going into the pump through a spinning shaft,
and WHP is what you get out of the pump – since the pump is not 100%
efficient in energy conversion, WHP < BHP
• WHP and BHP are archaic and confusing terms, but are still in wide use
• Emotor will usually be 92% or higher (about 98% with newer motors and
larger capacity motors)
• Epump depends on the pump design and on the operating point (Q vs.
TDH)
• WHP is defined as:

QH
WHP = (77)
102
where Q is in lps; H is in m of head; and WHP is in kilowatts (kW)

• If you use m in the above equation, UAC must be in $/100 m

• If you use ft in the above equation, UAC must be in $/100 ft

• Note that for fluid flow, “power” can be expressed as ρgQH = γQH
• Observe that 1,000/g = 1,000/9.81 ≈ 102, for the above units (other
conversion values cancel each other and only the 102 remains)
• The denominator changes from 102 to 3,960 for Q in gpm, H in ft, and
WHP in HP

5. Determine the present annual energy cost:

OtCf
E=
Ep
(78)

where Cf is the cost of “fuel”

• For electricity, the value of Cf is usually in dollars per kWh, and the value
used in the above equation may need to be an “average” based on
potentially complex billing schedules from the power company
• For example, in addition to the energy you actually consume in an electric
motor, you may have to pay a monthly fee for the installed capacity to
delivery a certain number of kW, plus an annual fee, plus different time-
of-day rates, and others
• Fuels such as diesel can also be factored into these equations, but the
power output per liter of fuel must be estimated, and this depends partly
on the engine and on the maintenance of the engine
• The units of E are dollars per WHP per year, or dollars per kW per year;
so it is a marginal cost that depends on the number of kW actually
required
Merkley & Allen Page 62 Sprinkle & Trickle Irrigation Lectures
6. Determine the marginal equipment cost:

• Note that Cf can include the “marginal” cost for the pump and power unit
(usually an electric motor)
• In other words, if a larger pump & motor costs more than a smaller pump,
then Cf should reflect that, so the full cost of friction loss is considered
• If you have higher friction loss, you may have to pay more for energy to
pump, but you may also have to buy a larger pump and/or power unit
(motor or engine)
• It sort of analogous to the Utah Power & Light monthly power charge,
based solely on the capacity to deliver a certain amount of power

Cf ($/kWh) = energy cost + marginal cost for a larger pump & motor

where “marginal” is the incremental unit cost of making a change in the

size of a component

• This is not really an “energy” cost per se, but it is something that can be
taken into account when balancing the fixed costs of the pipe (it falls
under the operating costs category, increasing for decreasing pipe costs)
• That is, maybe you can pay a little more for a larger pipe size and avoid
the need to buy a bigger pump, power unit and other equipment

• To calculate the marginal annual cost of a pump & motor:

MAC =
(
CRF $big − $small )
(
Ot kWbig − kWsmall ) (79)

where MAC has the same units as Cf; and $big -$small is the difference in
pump+motor+equipment costs for two different capacities

• The difference in fixed purchase price is annualized over the life of the
system by multiplying by the CRF, as previously calculated
• The difference in pump size is expressed as ∆BHP, where ∆BHP is the
difference in brake horsepower, expressed as kW
• To determine the appropriate pump size, base the smaller pump size on a
low friction system (or low pressure system)
• For BHP in kW:

QsHpump
BHP =
102Ep
(80)

Sprinkle & Trickle Irrigation Lectures Page 63 Merkley & Allen

 • Round the BHP up to the next larger available pump+motor+equipment
size to determine the size of the larger pump
• Then, the larger pump size is computed as the next larger available pump
size as compared to the smaller pump
• Then, compute the MAC as shown above
• The total pump cost should include the total present cost for the pump,
motor, electrical switching equipment (if appropriate) and installation
• Cf is then computed by adding the cost per kWh for energy

• Note that this procedure to determine MAC is approximate because the

marginal costs for a larger pump+motor+equipment will depend on the
magnitude of the required power change
• Using $big -$small to determine MAC only takes into account two (possibly
adjacent) capacities; going beyond these will likely change the marginal
rate
• However, at least we have a simple procedure to attempt to account for
this potentially real cost

7. Determine the equivalent annualized cost factor:

• This factor takes inflation into account:

⎡ (1 + e )n − (1 + i )n ⎤ ⎡ i ⎤
EAE = ⎢ ⎥⎢ ⎥ (81)
⎢ e −i ⎥ ⎢ (1 + i )n − 1⎥
⎣ ⎦⎣ ⎦
where e is the annual inflation rate (fraction), i is the annual interest rate
(fraction), and n is in years

• Notice that for e = 0, EAE = unity (this makes sense)

• Notice also that the above equation has a mathematical singularity for e =
i (but i is usually greater than e)

8. Determine the equivalent annual energy cost:

E ' = (EAE)(E) (82)

• This is an adjustment on E for the expected inflation rate

• No one really knows how the inflation rate might change in the future
• How do you know when to change to a larger pipe size (based on a
certain sectional flow rate)?

Beginning with a smaller pipe size (e.g. selected based

on maximum velocity limits), you would change to a
larger pipe size along a section of pipeline if the
Merkley & Allen Page 64 Sprinkle & Trickle Irrigation Lectures
 difference in cost for the next larger pipe size is less than
the difference in energy (pumping) savings

• Recall that the velocity limit is usually taken to be about 5 fps, or 1.5 m/s

9. Determine the difference in WHP between adjacent pipe sizes by equating the
annual plus annualized fixed costs for two adjacent pipe sizes:

E' (HPs1 ) + UACs1 = E' (HPs2 ) + UACs2

(83)
or,

∆WHPs1−s2 =
(UACs2 − UACs1)
E' (84)

• The subscript s1 is for the smaller of the two pipe sizes

• The units of the numerator might be $/100 m per year; the units of the
denominator might be $/kW per year
• This is the WHP (energy) savings needed to offset the annualized fixed
cost difference for purchasing two adjacent pipe sizes; it is the economic
balance point

10. Determine the difference in friction loss gradient between adjacent pipe sizes:

⎛ ∆WHPs1− s2 ⎞
∆Js1− s2 = 102 ⎜ ⎟
⎝ Qs ⎠ (85)

• This is the head loss difference needed to balance fixed and annual costs
for the two adjacent pipe sizes
• The coefficient 102 is for Qs in lps, and ∆WHP in kW
• You can also put Qs in gpm, and ∆WHP in HP, then substitute 3,960 for
102, and you will get exactly the same value for ∆J
• As before, ∆J is a head loss gradient, in head per 100 units of length (m
or ft, or any other unit)
• Thus, ∆J is a dimensionless “percentage”: head, H, can be in m, and
when you define a unit length (e.g. 100 m), the H per unit meter becomes
dimensionless
• This is why you can calculate ∆J using any consistent units and you will
get the same result

11. Calculate the flow rate corresponding to this head loss difference:

• Using the Hazen-Williams equation:

Sprinkle & Trickle Irrigation Lectures Page 65 Merkley & Allen

 1.852
∆J = Js1 − Js2
q⎞
= 16.42(10) ⎜ ⎟
⎝C⎠
6⎛
(Ds1−4.87 − Ds2−4.87 ) (86)

where q is in lps, and D is the inside diameter of the pipe in cm

• Or, using the Darcy-Weisbach equation:

∆J =
800 f q2
gπ 2 (Ds1−5 − Ds2−5 ) (87)

• Solve for the flow rate, q (with q in lps; D in cm):

0.54
⎡ ⎤
∆J
q = C⎢ ⎥
( )
(88)
⎢ 16.42(10)6 D−4.87 − D−4.87 ⎥
⎣ s1 s2 ⎦

• This is the flow rate for which either size (Ds1 or Ds2) will be the most
economical (it is the balancing point between the two adjacent pipe sizes)
• For a larger flow rate you would choose size Ds2, and vice versa

12. Repeat steps 8 through 11 for all other adjacent pipe sizes.

13. You can optionally create a graph with a log-log scale with the system flow
rate, Qs, on the ordinate and the section flow rate, q, on the abscissa:

• Plot a point at Qs and q for each of the adjacent pipe sizes

• Draw a straight diagonal line from lower left to upper right corner
• Draw a straight line at a slope of -1.852 (or -2.0 for Darcy-Weisbach)
through each of the points
• The slope will be different if the log scale on the axes are not the same
distance (e.g. if you do the plot on a spreadsheet, the ordinate and
abscissa may be different lengths, even if the same number of log cycles).
• In constructing the graph, you can get additional points by changing the
system flow rate, but in doing so you should also increase the area, A, so
that Ot is approximately the same as before. It doesn’t make sense to
change the system flow rate arbitrarily.
• Your graph should look similar to the one shown below

Merkley & Allen Page 66 Sprinkle & Trickle Irrigation Lectures

14. Applying the graph.

• Find the needed flow rate in a given section of the pipe, q, make an
intersection with the maximum system capacity (Qs, on the ordinate), then
see which pipe size it is
• You can use the graph for different system capacities, assuming you are
considering different total irrigated areas, or different crop and or climate
values
• Otherwise, you can just skip step 13 and just do the calculations on a
spreadsheet for the particular Qs value that you are interested in
• The graph is perhaps didactic, but doesn’t need to be constructed to
apply this economic pipe selection method

III. Notes on the Use of this Method

1. If any of the economic factors (interest rate, inflation rate, useful life of the
system) change, the lines on the graph will shift up or down, but the slope
remains the same (equal to the inverse of the velocity exponent for the head
loss equation: 1.852 for Hazen-Williams and 2.0 for Darcy-Weisbach).

2. Computer programs have been developed to use this and other economic
pipe selection methods, without the need for constructing a graphical solution
on log-log paper. You could write such a program yourself.

3. The economic pipe selection method presented above is not necessarily

valid for:
• looping pipe networks
• very steep downhill slopes
• non-“worst case” pipeline branches
Sprinkle & Trickle Irrigation Lectures Page 67 Merkley & Allen
 4. For loops, the flow might go in one direction some of the time, and in the
opposite direction at other times. For steep downhill slopes it is not
necessary to balance annual operation costs with initial costs because there
is essentially no cost associated with the development of pressure – there is
no need for pumping. Non-“worst case” pipeline branches may not have the
same pumping requirements (see below).

5. Note that the equivalent annual pipe cost considers the annual interest rate,
but not inflation. This is because financing the purchase of the pipe would be
done at the time of purchase, and we are assuming a fixed interest rate. The
uncertainty in this type of financing is assumed by the lending agency.

6. This method is not normally used for designing pipe sizes in laterals. For
one thing, it might recommend too many different sizes (inconvenient for
operation of periodic-move systems). Another reason is that we usually use
different criteria to design laterals (the “20%” rule on pressure variation).

7. Other factors could be included in the analysis. For example, there may be
certain taxes or tax credits that enter into the decision making process. In
general, the analysis procedure in determining pipe sizes can get as
complicated as you want it to – but higher complexity is better justified for
larger, more expensive irrigation systems.

Merkley & Allen Page 68 Sprinkle & Trickle Irrigation Lectures

IV. Other Pipe Sizing Methods

• Other methods used to size pipes include the following:

1. Unit head loss method: the designer specifies a limit on the

allowable head loss per unit length of pipe
2. Maximum velocity method: the designer specifies a maximum
average velocity of flow in the pipe (about 5 to 7 ft/s, or 1.5 to 2.0 m/s)
3. Percent head loss method: the designer sets the maximum pressure
variation in a section of the pipe, similar to the 20%Pa rule for lateral
pipe sizing

• It is often a good idea to apply more than one pipe selection method and
compare the results
• For example, don’t accept a recommendation from the economic selection
method if it will give you a flow velocity of more than about 10 ft/s (3 m/s),
otherwise you may have water hammer problems during operation
• However, it is usually advisable to at least apply the economic selection
method unless the energy costs are very low
• In many cases, the same pipe sizes will be selected, even when applying
different methods

• For a given average velocity, V, in a circular pipe, and discharge, Q, the

required inside pipe diameter is:

4Q
D= (89)
πV

• The following tables show maximum flow rates for specified average velocity
limits and different pipe inside diameters

Sprinkle & Trickle Irrigation Lectures Page 69 Merkley & Allen

 Gallons per Minute Litres per Second
Velocity Limit Velocity Limit
D (inch) A (ft2) 5 fps 7 fps D (mm) A (m2) 1.5 m/s 2 m/s
0.5 0.00136 3.1 4.3 10 0.00008 0.1 0.2
0.75 0.00307 6.9 9.6 20 0.00031 0.5 0.6
1 0.00545 12.2 17.1 25 0.00049 0.7 1.0
1.25 0.00852 19.1 26.8 30 0.00071 1.1 1.4
1.5 0.01227 27.5 38.6 40 0.00126 1.9 2.5
2 0.02182 49.0 68.5 50 0.00196 2.9 3.9
3 0.04909 110 154 75 0.00442 6.6 8.8
4 0.08727 196 274 100 0.00785 11.8 15.7
5 0.13635 306 428 120 0.01131 17.0 22.6
6 0.19635 441 617 150 0.01767 26.5 35.3
8 0.34907 783 1,097 200 0.03142 47.1 62.8
10 0.54542 1,224 1,714 250 0.04909 73.6 98.2
12 0.78540 1,763 2,468 300 0.07069 106 141
15 1.22718 2,754 3,856 400 0.12566 188 251
18 1.76715 3,966 5,552 500 0.19635 295 393
20 2.18166 4,896 6,855 600 0.28274 424 565
25 3.40885 7,650 10,711 700 0.38485 577 770
30 4.90874 11,017 15,423 800 0.50265 754 1,005
40 8.72665 19,585 27,419 900 0.63617 954 1,272
50 13.63538 30,602 42,843 1000 0.78540 1,178 1,571
1100 0.95033 1,425 1,901
Cubic Feet per Second 1200 1.13097 1,696 2,262
Velocity Limit 1300 1.32732 1,991 2,655
D (ft) A (ft2) 5 fps 7 fps 1400 1.53938 2,309 3,079
1 0.785 3.93 5.50 1500 1.76715 2,651 3,534
2 3.142 15.71 21.99 1600 2.01062 3,016 4,021
3 7.069 35.34 49.48 1700 2.26980 3,405 4,540
4 12.566 62.83 87.96 1800 2.54469 3,817 5,089
5 19.635 98.17 137.44 1900 2.83529 4,253 5,671
6 28.274 141.37 197.92 2000 3.14159 4,712 6,283
7 38.485 192.42 269.39 2100 3.46361 5,195 6,927
8 50.265 251.33 351.86 2200 3.80133 5,702 7,603
9 63.617 318.09 445.32 2300 4.15476 6,232 8,310
10 78.540 392.70 549.78 2400 4.52389 6,786 9,048
11 95.033 475.17 665.23 2500 4.90874 7,363 9,817
12 113.097 565.49 791.68 2600 5.30929 7,964 10,619
13 132.732 663.66 929.13 2700 5.72555 8,588 11,451
14 153.938 769.69 1,077.57 2800 6.15752 9,236 12,315
15 176.715 883.57 1,237.00 2900 6.60520 9,908 13,210
16 201.062 1,005.31 1,407.43 3000 7.06858 10,603 14,137
17 226.980 1,134.90 1,588.86 3100 7.54768 11,322 15,095
18 254.469 1,272.35 1,781.28 3200 8.04248 12,064 16,085
19 283.529 1,417.64 1,984.70 3300 8.55299 12,829 17,106
20 314.159 1,570.80 2,199.11 3400 9.07920 13,619 18,158

Merkley & Allen Page 70 Sprinkle & Trickle Irrigation Lectures

Lecture 7
Set Sprinkler Lateral Design

I. Basic Design Criterion

• The basic design criterion is to size lateral pipes so that pressure variation
along the length of the lateral does not exceed 20% of the nominal design
pressure for the sprinklers
• This criterion is a compromise between cost of the lateral pipe and
application uniformity in the direction of the lateral
• Note that the locations of maximum and minimum pressure along a lateral
pipe can vary according to ground slope and friction loss gradient

II. Location of Average Pressure in the Lateral

• We are interested in the location of average pressure along a lateral pipe

because it is related to the design of the lateral
• Recall that friction head loss along a multiple-outlet pipe is nonlinear
• The figure below is for a lateral laid on level ground – pressure variation is
due to friction loss only...

• For equally-spaced outlets (sprinklers) and approximately thirty outlets (or

more), three-quarters of the pressure loss due to friction will occur between
the inlet and the location of average pressure
• The location of average pressure in the lateral is approximately 40% of the
lateral length, measured from the lateral inlet
• If there were only one outlet at the end of the lateral pipe, then one-half the
pressure loss due to friction would take place between the lateral inlet and
the location of average pressure, as shown below
Sprinkle & Trickle Irrigation Lectures Page 71 Merkley & Allen
 • A computer program can be written to solve for the head loss in the lateral
pipe between each sprinkler
• Consider the following equations:

Total friction head loss:

(hf )total = ∑i=1(hf )i

n
(90)

Friction head loss to location of ha:

∑i=1∑ j=1(hf ) j
n i

(hf )a =
n +1 (91)

where n is the number of sprinklers; (hf)total is the total friction head loss from
0 to L; (hf)i is the friction head loss in the lateral pipe between sprinklers i-1
and i; and (hf)a is the friction loss from the lateral inlet to the location of ha

• As indicated above, (hf)a occurs over approximately the first 40% of the
lateral
• Note that between sprinklers, the friction head loss gradient is linear in the
lateral pipe
• Note also that (hf)0 = 0, but it is used in calculating (ha)f, so the denominator
is (n+1), not n

Merkley & Allen Page 72 Sprinkle & Trickle Irrigation Lectures

 • In applying these equations with sample data, the following result can be
found:
( h f )a
≈ 0.73 (92)
(hf )total
• This supports the above claim that approximately ¾ of the friction head loss
occurs between the lateral inlet and the location of ha
• Also, from these calculations it can be seen that the location of ha is
approximately 38% of the lateral length, measured from the inlet, for laterals
with approximately 30 or more sprinklers
• But, this analysis assumes a constant qa, which is not quite correct unless
flow control nozzles and or pressure regulators are used at each sprinkler
• We could eliminate this assumption of constant qa, but it involves the solution
of a system of nonlinear equations

III. Location of Minimum Pressure in Laterals Running Downhill

• The location of minimum pressure in a lateral running downhill is where the

slope of the friction loss curve, J, equals the ground slope

Sprinkle & Trickle Irrigation Lectures Page 73 Merkley & Allen

 • The above assertion is analogous to a pre-calculus “max-min problem”,
where you take the derivative of a function and set it equal to zero (zero
slope)
• Here we are doing the same thing, but the slope is not necessarily zero

Hazen-Williams Equation:

1.852
12 ⎛ Q ⎞
J = 1.21(10) ⎜ ⎟ D−4.87 (93)
⎝C⎠
for J in meters of friction head loss per 100 m (or ft/100 ft); Q in lps; and D in
mm

• In this equation we will let:

⎛q ⎞
Q = Ql − ⎜ a ⎟ x (94)
⎝ Se ⎠
for multiple, equally-spaced sprinkler outlets spaced at Se (m) from each
other, with constant discharge of qa (lps). Ql is the flow rate at the lateral
inlet (entrance).
• To find the location of minimum pressure, let J = S, where S is the ground
slope (in %, because J is per 100 m), which is negative for downhill-sloping
laterals
• Combining the two above equations and solving for x,

x=
Se ⎡
qa ⎣
(
Ql − 3(10)−7 C ( −S)0.54 D2.63 ⎤
⎦ ) (95)

where x is the distance, in m, from the lateral inlet to the minimum pressure
Merkley & Allen Page 74 Sprinkle & Trickle Irrigation Lectures
 • S is in percent; Se and x are in m; D is in mm; and Ql and qa are in lps
• Note that the valid range of x is: 0 ≤ x ≤ L, and that you won’t necessarily get
J = S over this range of x values:

• If you get x < 0 then the minimum pressure is at the inlet

• If you get x > L then the minimum pressure is at the end

• This means that the above equation for x is valid for all ground slopes: S = 0,
S > 0 and S < 0

IV. Required Lateral Inlet Pressure Head

• Except for the most unusual circumstances (e.g. non-uniform downhill slope
that exactly matches the shape of the hf curve), the pressure will vary with
distance in a lateral pipe
• According to Keller & Bliesner’s design criterion, the required inlet pressure
head to a sprinkler lateral is that which makes the average pressure in the
lateral pipe equal to the required sprinkler pressure head, ha

• We can force the average pressure to be equal to the desired sprinkler

operating pressure by defining the lateral inlet pressure head as:

3 1
hl = ha + hf + ∆he (96)
4 2
• hl is the required pressure head at the lateral inlet
• Strictly speaking, we should take approximately 0.4∆he in the above
equation, but we are taking separate averages for the friction loss and
elevation gradients – and, this is a design equation
• Of course, instead of head, h, in the above equation, pressure, P, could be
used if desired

Sprinkle & Trickle Irrigation Lectures Page 75 Merkley & Allen

 The value of ∆he is negative for laterals running
downhill

• For steep downhill slopes, where the minimum pressure would be at the
lateral inlet, it is best to let

hf = −∆he (97)

• Thus, we would want to consume, or “burn up”, excess pressure through

friction loss by using smaller pipes
• To achieve this equality for steep downhill slopes, it may be desirable to have
more than one pipe diameter in the lateral
• A downhill slope can be considered “steep” when (approximately)…

−∆he > 0.3ha (98)

• We now have an equation to calculate lateral inlet pressure based on ha, hf,
and he

Merkley & Allen Page 76 Sprinkle & Trickle Irrigation Lectures

 • However, for large values of hf there will be correspondingly large values of hl
• Thus, for zero ground slope, to impose a limit on hf we will accept:

hf = 0.20ha (for S = 0 only) (99)

• This is the same as saying that we will not allow pipes that are too small, that
is, pipes that would produce a large hf value
• An additional head term must be added to the equation for hl to account for
the change in elevation from the lateral pipe to the sprinkler (riser height):

3 1
hl = ha + hf + ∆he + hr (100)
4 2
or, in terms of pressure…

3 1
Pl = Pa + Pf + ∆Pe + Pr (101)
4 2
V. Friction Losses in Pipes with Multiple Outlets

• Pipes with multiple outlets have decreasing flow rate with distance (in the
direction of flow), and this causes the friction loss to decrease by
approximately the square of the flow rate (for a constant pipe diameter)
• Sprinkler and trickle irrigation laterals fall into this hydraulic category
• Multiply the head loss for a constant discharge pipe by a factor “F” to reduce
the total head loss for a lateral pipe with multiple, equally spaced outlets:

JFL
hf = (102)
100
where F is from Eq. 8.9a

1 1 b −1
F= + + (103)
b + 1 2N 6N2

for equally spaced outlets, each with the same discharge, and going all the
way to the end of the pipe.

• All of the flow is assumed to leave through the outlets, with no “excess”
spilled out the downstream end of the pipe
• N is the total number of equally spaced outlets
• The value of b is the exponent on Q in the friction loss equation

Sprinkle & Trickle Irrigation Lectures Page 77 Merkley & Allen

 • Darcy-Weisbach: b = 2.0
• Hazen-Williams: b = 1.852

• The first sprinkler is assumed to be located a distance of Se from the lateral

inlet
• Eq. 8.9b (see below) gives F(α), which is the F factor for initial outlet
spacings less than or equal to Se

NF − (1 − α)
F(α) = (104)
N − (1 − α)

where 0 < α ≤ 1

• Note that when α = 1, F(α) = F

• Many sprinkler systems have the first sprinkler at a distance of ½Se from the
lateral inlet (α = 0.5), when laterals run in both orthogonal directions from the
mainline

VI. Lateral Pipe Sizing for a Single Pipe Size

• If the minimum pressure is at the end of the lateral, which is the case for no
ground slope, uphill, and slight downhill slopes, then the change in pressure
head over the length of the lateral is:

∆h = h f + ∆h e (105)

If we allow ∆h = 0.20 ha, then

0.20ha = hf + ∆he (106)

JaFL
0.20ha − ∆he = (107)
100
and,
⎛ 0.20ha − ∆he ⎞
Ja = 100 ⎜ ⎟ (108)
⎝ FL ⎠

where Ja is the allowable friction loss gradient.

• Lateral pipe diameter can be selected such that J ≤ Ja

• The above is part of a standard lateral design criteria and will give a system
CU of approximately 0.97CU if lateral inlet pressures are the same for each
lateral position, for set sprinkler systems
Merkley & Allen Page 78 Sprinkle & Trickle Irrigation Lectures
 • If the lateral is sloping downhill and the minimum pressure does not occur at
the end of the lateral, then we will attempt to consume the elevation gain in
friction loss as follows:

hf = −∆he (109)

⎛ −∆he ⎞
Ja = 100 ⎜ ⎟ (110)
⎝ FL ⎠

• Note that in this case ∆h ≠ hf + ∆he. Rather, ∆h = hmax - hmin, where:

1. hmax is either at the lateral inlet or at the end of the lateral, and
2. hmin is somewhere between the lateral inlet and the end

• Given a value of Ja, the inside diameter of the lateral pipe can be calculated
from the Hazen-Williams equation:

0.205
⎡ K ⎛ Q ⎞1.852 ⎤
D=⎢ ⎜ l⎟ ⎥ (111)
⎢⎣ Ja ⎝ C ⎠ ⎥⎦

where Ql is the flow rate at the lateral inlet (Nqa) and K is the units coefficient
in the Hazen-Williams equation

• The calculated value of D would normally be rounded up to the next available

internal pipe diameter

VII. Lateral Design Example

VI.1. Given information:

L= 396 m (lateral length)

qa = 0.315 lps (nominal sprinkler discharge)
Se = 12 m (sprinkler spacing)
hr = 1.0 m (riser height)
slope = -2.53% (going downhill)
Pa = 320 kPa (design nozzle pressure)
pipe material = aluminum

Sprinkle & Trickle Irrigation Lectures Page 79 Merkley & Allen

 VI.2. Calculations leading to allowable pressure head loss in the lateral:

Nn = 396/12 = 33 sprinklers
F = 0.36
Ql = (0.315)(33) = 10.4 lps
∆he = SL = (-0.0253)(396) = -10.0 m

(Pf)a = 0.20Pa - ∆he = 0.20(320 kPa) – 9.81(-10.0 m) = 162 kPa

(hf)a = 162/9.81 = 16.5 m

VI.3. Calculations leading to required lateral pipe inside diameter:

0.3Pa = 0.3(320 kPa) = 96.0 kPa

0.3ha = 96.0/9.81 = 9.79 m

Now, 0.3ha < -∆he (steep downhill). Therefore, may want to use hf = -∆he.
Then, Ja is:

⎛ −∆he ⎞ ⎛ −( −10.0m) ⎞
Ja = 100 ⎜ ⎟ = 100 ⎜ ⎟ = 7.01m /100m (112)
⎝ FL ⎠ ⎝ (0.36)(396) ⎠

However, if 0.3ha > -∆he, Ja would be calculated as:

⎛ 0.20ha − ∆he ) ⎞ ⎛ 16.5 ⎞

Ja = 100 ⎜ ⎟ = 100 ⎜ ⎟ = 11.6m /100m (113)
⎝ FL ⎠ ⎝ (0.36)(396) ⎠

For now, let’s use Ja = 7.01 m/100 m. Then, the minimum pipe inside
diameter is (C ≈ 130 for aluminum):

0.205
⎡ 1.21E12 ⎛ 10.4 ⎞1.852 ⎤
D=⎢ ⎜ ⎟ ⎥ = 77.7mm (114)
⎢⎣ 7.01 ⎝ 130 ⎠ ⎥⎦

which is equal to 3.06 inches.

In the USA, 3” aluminum sprinkler pipe has an ID of 2.9” (73.7 mm), so

for this design it would be necessary to round up to a 4” nominal pipe size
(ID = 3.9”, or 99.1 mm).

However, it would be a good idea to also try the 3” size and see how the
lateral hydraulics turn out (this is done below; note also that for Ja = 11.6,
D = 70.0 mm).

Merkley & Allen Page 80 Sprinkle & Trickle Irrigation Lectures

 VI.4. Check the design with the choices made thus far

The real friction loss will be:

1.852
⎛ 10.4 ⎞
J = 1.21E12 ⎜ ⎟ ( 99.1 mm)−4.87 = 2.14m /100m (115)
⎝ 130 ⎠

JFL (2.14)(0.36)(396)
hf = = = 3.06 m (116)
100 100
The required lateral inlet pressure head is:

hl = ha + 0.75hf + 0.5∆he + hr
(117)
hl = 320 / 9.81 + 0.75(3.06) + 0.5( −10.0) + 1.0 = 30.9 m

Thus, Pl is (30.9)(9.81) = 303 kPa, which is less than the specified Pa of

320 kPa, and this is because the lateral is running downhill

VI.5. Calculate the pressure and head at the end of the lateral pipe

hend = hl − hf − ∆he = 30.9 − 3.06 − (−10.0) = 37.8 m (118)

which is equal to 371 kPa. Thus, the pressure at the end of the lateral
pipe is greater than the pressure at the inlet.

To determine the pressure at the last sprinkler head, subtract the riser
height to get 37.8 m – 1.0 m = 36.8 m (361 kPa)

VI.6. Calculate the location of minimum pressure in the lateral pipe

x=
Se ⎡
qa ⎣
(
Ql − 3(10)−7 C( −S)0.54 D2.63 ⎤
⎦ )
(119)
x=
12 ⎡
0.315 ⎣
(
10.4 − 3(10)−7 130(2.53)0.54 (99.1)2.63 ⎤ = −39.6 m
⎦ )
The result is negative, indicating that that minimum pressure is really at
the entrance (inlet) to the lateral pipe. The minimum sprinkler head
pressure is equal to hl – hr = 30.9 – 1.0 = 29.9 m, or 293 kPa.

Sprinkle & Trickle Irrigation Lectures Page 81 Merkley & Allen

 VI.7. Calculate the percent pressure variation along the lateral pipe

The maximum pressure is at the last sprinkler (end of the lateral), and the
minimum pressure is at the first sprinkler (lateral inlet). The percent
pressure variation is:

Pmax − Pmin 361 − 293

∆P = = = 0.21% (120)
Pa 320

That is, 21% pressure variation at the sprinklers, along the lateral

This is larger than the design value of 0.20, or 20% variation. But it is
very close to that design value, which is somewhat arbitrary anyway.

VI.8. Redo the calculations using a 3” lateral pipe instead of the 4” size

In this case, the location of the minimum pressure in the lateral pipe is:

x=
12 ⎡
0.315 ⎣
(
10.4 − 3(10)−7 130(2.53)0.54 (73.7)2.63 ⎤ = 196 m
⎦ ) (121)

which is the distance from the upstream end of the lateral.

There are about 196/12 = 16 sprinklers from the lateral inlet to the
location of minimum pressure, and about 17 sprinklers from x to the end
of the lateral.

Friction loss from x to the end of the lateral is:

1.852
⎛ (17)(0.315) ⎞
Jx −end = 1.21E12 ⎜ ⎟ ( 73.7 )−4.87 = 2.65 m /100 m
⎝ 130 ⎠ (122)

(2.65)(0.38)(396 − 196)
(hf )x −end = = 2.01 m (123)
100
Friction loss from the inlet to the end is:

1.852
⎛ 10.4 ⎞
Jinlet −end = 1.21E12 ⎜ ⎟ ( 73.7 )−4.87 = 9.05 m /100 m (124)
⎝ 130 ⎠

Merkley & Allen Page 82 Sprinkle & Trickle Irrigation Lectures

 (9.05)(0.36)(396)
(hf )inlet −end = = 12.9 m (125)
100
Then, friction loss from inlet to x is:

(hf )inlet − x = 12.9 − 2.01 = 10.9 m (126)

The required lateral pipe inlet head is:

hl = ha + 0.75hf + 0.5∆he + hr
(127)
hl = 320 / 9.81 + 0.75(12.9) + 0.5(−10.0) + 1.0 = 38.3 m

giving a Pl of (38.3)(9.81) = 376 kPa, which is higher than Pl for the 4”

pipe

The minimum pressure head (at distance x = 196 m) is:

hx = hl − (hf )inlet − x − ( ∆he )inlet − x

(128)
hx = 38.3 − 10.9 − (−0.0253)(196) = 32.4 m

giving a Px of (32.4)(9.81) = 318 kPa, which is very near Pa.

The pressure head at the end of the lateral pipe is:

hend = hl − hf − ∆he = 38.3 − 12.9 + 10.0 = 35.4 m (129)

giving Pend of (35.4)(9.81) = 347 kPa, which is less than Pl. So, the
maximum lateral pipe pressure is at the inlet.

The percent variation in pressure at the sprinklers is based on Pmax = 376

-(1.0)(9.81) = 366 kPa, and Pmin = 318 - (1.0)(9.81) = 308 kPa:

Pmax − Pmin 366 − 308

% ∆P = = = 0.18 (130)
Pa 320

which turns out to be slightly less than the design value of 20%

Sprinkle & Trickle Irrigation Lectures Page 83 Merkley & Allen

 VI.9. What if the lateral ran uphill at 2.53% slope?

In this case, the maximum allowable head loss gradient is:

⎛ 0.20ha − ∆he ) ⎞
Ja = 100 ⎜ ⎟
⎝ FL ⎠
(131)
⎛ 0.2(320 / 9.81) − 10.0 ⎞
= 100 ⎜ ⎟ = −2.44m /100m
⎝ (0.36)(396) ⎠

which is negative because ∆he > 0.2ha, meaning that it is not possible to
have only a 20% variation in pressure along the lateral, that is, unless
flow control nozzles and or other design changes are made.

VI.10. Some observations about this design example

Either the 3” or 4” aluminum pipe size could be used for this lateral
design. The 4” pipe will cost more than the 3” pipe, but the required
lateral inlet pressure is less with the 4” pipe, giving lower pumping costs,
assuming pumping is necessary.

Note that it was assumed that each sprinkler discharged 0.315 lps, when
in reality the discharge depends on the pressure at each sprinkler. To
take into account the variations in sprinkler discharge would require an
iterative approach to the mathematical solution (use a computer).

Most sprinkler laterals are laid on slopes less than 2.5%, in fact, most are
on fields with less than 1% slope.

Merkley & Allen Page 84 Sprinkle & Trickle Irrigation Lectures

Lecture 8
Set Sprinkler Lateral Design

I. Dual Pipe Size Laterals

• Sometimes it is useful to design a lateral pipe with two different diameters to

accomplish either of the following:

1. a reduction in hf
2. an increase in hf

• In either case, the basic objective is to reduce pressure variations along the
lateral pipe by arranging the friction loss curve so that it more closely
parallels the ground slope
• It is not normally desirable to have more than one pipe size in portable
laterals (hand-move, wheel lines), because it usually makes set changes
more troublesome
• For fixed systems with buried laterals, it may be all right to have more than
two pipe diameters along the laterals
• For dual pipe size laterals, approximately 5/8 of the pressure loss due to
friction occurs between the lateral inlet and the location of average pressure

• Case 1: a lateral on level ground where one pipe size is too small, but the
next larger size is too big...

• d1 is the larger diameter, and d2 is the smaller diameter

• note that (hf)single is much larger than (hf)dual

Sprinkle & Trickle Irrigation Lectures Page 85 Merkley & Allen

 • Case 2: a lateral running downhill where one pipe size is too big, but the next
smaller size is too small...

• The composite friction loss curve for d1 and d2 more closely parallels the
ground slope than the curve with only d1, which means that the pressure
variation along the lateral is less with the dual pipe size design

II. Location of Average Pressure in Dual Size Laterals

• Do you believe that 5/8(hf)dual occurs between hl and ha?

• Consider the analysis shown graphically below (5/8 = 0.625)
• The plot is for a dual pipe size lateral with D1 = 15 cm, D2 = 12 cm, 100
equally-spaced outlets, 900 m total lateral length, Hazen-Williams C factor of
130, uniform sprinkler flow rate of 0.4 lps, and zero ground slope

Merkley & Allen Page 86 Sprinkle & Trickle Irrigation Lectures

• Notice where the ¾ value is on the left-hand ordinate
• Notice that the head loss from hl to ha is approximately 74% when the pipe size
is all D1 (1.0 on the abscissa) and when the pipe size is all D2 (0.0 on the
abscissa)
• The hl values (inlet pressure head) would be different for each point on the curve
if it were desired to maintain the same ha for different lateral designs
• Notice that the distance from the lateral inlet to the location of average pressure
head is roughly 40% of the total lateral length, but varies somewhat depending
on the ratio of lengths of D1 to D2 (in this example)

• These calculations can be set up on a spreadsheet to analyze any particular

combination of pipe sizes and other hydraulic conditions. Below is an example:

Section Flow Distance Diameter hf Sum (hf) d(he) head diff from hf/(hf)total
(lps) (m) (cm) (m) (m) (m) (m) ha (%)
1 40.00 9.00 15.00 0.31 0.31 0 49.69 12.08 0.016
2 39.60 18.00 15.00 0.31 0.62 0 49.38 11.77 0.032
3 39.20 27.00 15.00 0.30 0.92 0 49.08 11.47 0.048
4 38.80 36.00 15.00 0.29 1.21 0 48.79 11.18 0.064
5 38.40 45.00 15.00 0.29 1.50 0 48.50 10.89 0.079
6 38.00 54.00 15.00 0.28 1.79 0 48.21 10.61 0.094

Sprinkle & Trickle Irrigation Lectures Page 87 Merkley & Allen

III. Determining X1 and X2 in Dual Pipe Size Laterals

• The friction loss is:

⎛ J FL J F x ⎞ ⎛ J F x ⎞
hf = ⎜ 1 1 − 2 2 2 ⎟ + ⎜ 3 2 2 ⎟ (132)
⎝ 100 100 ⎠ ⎝ 100 ⎠
where,
hf = total lateral friction head loss for dual pipe sizes

J1 = friction loss gradient for D1 and Qinlet

J2 = friction loss gradient for D1 and Qinlet - (qa)(x1)/Se
J3 = friction loss gradient for D2 and Qinlet - (qa)(x1)/Se
F1 = multiple outlet reduction coefficient for L/Se outlets
F2 = multiple outlet reduction coefficient for x2/Se outlets
x1 = length of D1 pipe (larger size)
x2 = length of D2 pipe (smaller size)
x1 + x2 = L

• As in previous examples, we assume constant qa

• As for single pipe size laterals, we will fix hf by

∆h = hf + ∆he = 20%ha (133)

and,
hf = 20%ha − ∆he (134)

• Find d1 and d2 in tables (or by calculation) using Qinlet and...

( J)d1 ≤ Ja ≤ ( J)d2 (135)

for,
⎛ 20%ha − ∆he ⎞
Ja = 100 ⎜ ⎟
⎝ FL ⎠ (136)

• Now there are two adjacent pipe sizes: d1 and d2

• Solve for x1 and x2 by trial-and-error, or write a computer program, and make
hf = 0.20ha - ∆he (you already have an equation for hf above)

Merkley & Allen Page 88 Sprinkle & Trickle Irrigation Lectures

IV. Setting up a Computer Program to Determine X1 and X2

• If the Hazen-Williams equation is used, the two F values will be:

1 ⎛ 4 ⎞
F1 ≈ 0.351 + ⎜ 1 + ⎟ (137)
2N1 ⎝ 13N1 ⎠

1 ⎛ 4 ⎞
F2 ≈ 0.351 + ⎜1 + ⎟ (138)
2N2 ⎝ 13N2 ⎠

where
L
N1 = (139)
Se

L − x1
N2 = (140)
Se

• The three friction loss gradients are:

1.852
⎛Q ⎞
J1 = K ⎜ 1 ⎟ D1−4.87 (141)
⎝C⎠

1.852
⎛Q ⎞
J2 = K ⎜ 2 ⎟ D1−4.87 (142)
⎝ C ⎠

1.852
⎛Q ⎞
J3 = K ⎜ 2 ⎟ D2−4.87 (143)
⎝ C ⎠
where
⎛ L ⎞
Q1 = ⎜ ⎟ qa (144)
⎝ e⎠
S

⎛ L − x1 ⎞
Q2 = ⎜ ⎟ qa (145)
⎝ e ⎠
S

• The coefficient K in Eqs. 141-143 is 1,050 for gpm & inches; 16.42(10)6 for
lps and cm; or 1.217(10)12 for lps and mm
Sprinkle & Trickle Irrigation Lectures Page 89 Merkley & Allen
 • Combine the above equations and set it equal to zero:

f ( x1 ) = α1 ⎡α 2 − α3 (L − x1 ) F2 ⎤ − 0.2ha + ∆he = 0
2.852
(146)
⎢⎣ ⎥⎦

where
K
α1 = (147)
100 C1.852

1.852
⎛q L⎞
α2 = ⎜ a ⎟ D1−4.87F1 L (148)
⎝ Se ⎠
1.852
α3 = ( D1−4.87 − D2−4.87 ) ⎛ qa ⎞
⎜ ⎟
⎝ Se ⎠
(149)

• The three alpha values are constants

• Eq. 146 can be solved for the unknown, x1, by the Newton-Raphson method
• To accomplish this, we need the derivative of Eq. 146 with respect to x1

df(x1)
=
dx1
(150)
⎡ S (L − x1)0.852 ⎛ 8Se ⎞ ⎤
= α1α3 ⎢ 2.852F2 (L − x1)1.852 − e ⎜ 1 + ⎟⎥
⎢⎣ 2 ⎝ 13(L − x1) ⎠ ⎥⎦

• Note that the solution may fail if the sizes D1 & D2 are inappropriate
• To make things more interesting, give the computer program a list of inside
pipe diameters so that it can find the most appropriate available values of D1
& D2
• Note that the Darcy-Weisbach equation could be used instead of Hazen-
Williams
• In Eq. 146 you could adjust the value of the 0.2 coefficient on ha to determine
its sensitivity to the pipe diameters and lengths
• The following screenshot is of a small computer program for calculating
diameters and lengths of dual pipe size sprinkler laterals

Merkley & Allen Page 90 Sprinkle & Trickle Irrigation Lectures

V. Inlet Pressure for Dual Pipe Size Laterals

5 1
hl = ha + hf + ∆he + hr
8 2 (151)

• This is the same as the lateral inlet pressure head equation for single pipe
size, except that the coefficient on hf is 5/8 instead of 3/4
• Remember that for a downhill slope, the respective pressure changes due to
friction loss and due to elevation change are opposing

VI. Laterals with Flow Control Devices

• Pressure regulating valves can be located at the base of each sprinkler:

These have approximately 2 to 5 psi (14 to 34 kPa) head loss
• Also, flow control nozzles (FCNs) can be installed in the sprinkler heads
• FCNs typically have negligible head loss
• For a lateral on level ground, the minimum pressure is at the end:

Sprinkle & Trickle Irrigation Lectures Page 91 Merkley & Allen

 • The lateral inlet pressure head, hl, is determined such that the minimum
pressure in the lateral is enough to provide ha at each sprinkler...

hl = ha + hf + ∆he + hr + hcv (152)

where hcv is the pressure head loss through the flow control device

• For a lateral with flow control devices, the average pressure is not equal to
the nominal sprinkler pressure

havg ≠ ha
(153)

• If the pressure in the lateral is enough everywhere, then

2
⎛q ⎞
ha = ⎜ a ⎟
⎝ Kd ⎠ (154)

where ha is the pressure head at the sprinklers

• Below is a sketch of the hydraulics for a downhill lateral with flow control
devices

Merkley & Allen Page 92 Sprinkle & Trickle Irrigation Lectures

VII. Anti-Drain Valves

• Valves are available for preventing flow through sprinklers until a certain
minimum pressure is reached
• These valves are installed at the base of each sprinkler and are useful where
sprinkler irrigation is used to germinate seeds on medium or high value crops
• The valves help prevent seed bed damage due to low pressure streams of
water during startup and shutdown
• But, for periodic-move, the lines still must be drained before moving

Sprinkle & Trickle Irrigation Lectures Page 93 Merkley & Allen

Gravity-Fed Lateral Hydraulic Analysis

I. Description of the Problem

• A gravity-fed sprinkler lateral with evenly spaced outlets (sprinklers),

beginning at a distance Se from the inlet:

• The question is, for known inlet head, H0, pipe diameter, D, sprinkler spacing,
Se, ground slope, So, sprinkler discharge coefficient, Kd, riser height, hr, and
pipe material (C factor), what is the flow rate through each sprinkler?
• Knowing the answer will lead to predictions of application uniformity
• In this case, we won’t assume a constant qa at each sprinkler

II. Friction Loss in the Lateral

Hazen-Williams equation:

JL
hf = (155)
100

1.852
6⎛Q⎞
J = 16.42 (10 ) ⎜ ⎟ D−4.87
⎝C⎠ (156)

for Q in lps; D in cm; J in m/100 m; L in m; and hf in m.

Merkley & Allen Page 94 Sprinkle & Trickle Irrigation Lectures

 • Between two sprinklers,

1.852
JSe ⎛Q⎞
= 16.42 (10 ) Se ⎜ ⎟ D−4.87
4
hf =
100 ⎝C⎠ (157)
or,
hf = hw Q1.852 (158)

where Q is the flow rate in the lateral pipe between two sprinklers, and

hw = 16.42 (10 ) SeC−1.852D−4.87

4
(159)

III. Sprinkler Discharge

typically,
q = Kd h (160)

where q is the sprinkler flow rate in lps; h is the pressure head at the
sprinkler in m; and Kd is an empirical coefficient: Kd = KoA, where A is the
cross sectional area of the inside of the pipe

IV. Develop the System of Equations

• Suppose there are only four sprinklers, evenly spaced (see the above figure)
• Suppose that we know H0, Kd, C, D, hr, So, and Se

(Q − Q )
2 2
⎛q ⎞
q1 = K d H1 − hr → H1 = hr + ⎜ 1 ⎟ = hr + 1 2 2 (161)
⎝ Kd ⎠ Kd
(Q − Q )
2 2
⎛q ⎞
q2 = K d H2 − hr → H2 = hr + ⎜ 2 ⎟ = hr + 2 2 3 (162)
⎝ Kd ⎠ Kd
(Q − Q )
2 2
⎛q ⎞
q3 = K d H3 − hr → H3 = hr + ⎜ 3 ⎟ = hr + 3 2 4 (163)
⎝ Kd ⎠ Kd
2
⎛ q4 ⎞ Q42
q4 = K d H4 − hr → H4 = hr + ⎜ ⎟ = hr + 2 (164)
⎝ Kd ⎠ Kd

Sprinkle & Trickle Irrigation Lectures Page 95 Merkley & Allen

 • Pressure heads can also be defined independently in terms of friction loss
along the lateral pipe

H1 = H0 − hw Q11.852 − ∆he (165)

H2 = H1 − hw Q21.852 − ∆he (166)

H3 = H2 − hw Q31.852 − ∆he (167)

H4 = H3 − hw Q41.852 − ∆he (168)

where,
Se
∆he = (169)
So−2 +1

and So is the ground slope (m/m)

• The above presumes a uniform, constant ground slope

• Note that in the above equation, ∆he is always positive. So it is necessary to
multiply the result by –1 (change the sign) whenever So < 0.
• Note also that So < 0 means the lateral runs in the downhill direction
• Combining respective H equations:

( Q1 − Q2 ) 2 = H − hw Q11.852 − ∆he − hr (170)

0
K d2

( Q 2 − Q3 ) 2 = H 1.852
1 − h w Q2 − ∆he − hr (171)
K d2

( Q3 − Q 4 ) 2 = H − hw Q31.852 − ∆he − hr (172)

2
K d2

Q4 2
= H3 − hw Q41.852 − ∆he − hr (173)
K d2

Merkley & Allen Page 96 Sprinkle & Trickle Irrigation Lectures

 • Setting the equations equal to zero:

f1 = H0 −
( Q1 − Q2 )
2
−h 1.852
− ∆he − hr = 0
w Q1 (174)
K 2d

f2 =
( Q1 − Q2 )
2
−
( Q2 − Q3 )
2
−h 1.852
− ∆he − hr = 0
w Q2 (175)
K 2d K 2d

f3 =
( Q2 − Q3 )
2
−
( Q3 − Q4 )
2
−h 1.852
− ∆he − hr = 0
w Q3 (176)
K 2d K 2d

f4 =
( Q3 − Q4 )
2
−
Q 42
−h 1.852
− ∆he − hr = 0
w Q4 (177)
K 2d K 2d

The system of equations can be put into matrix form as follows:

⎡ ∂f1 ∂f1 ⎤ ⎡ ⎤ ⎡ ⎤
⎢ ∂Q ∂Q2 ⎥ ⎢ δQ1 ⎥ ⎢ f1 ⎥
⎢ 1 ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ∂f2 ∂f2 ∂f2 ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ∂Q ∂Q2 ∂Q3 ⎥ ⎢ δQ2 ⎥ ⎢ f2 ⎥
⎢ 1 ⎥ ⎢ ⎥ =⎢ ⎥ (178)
⎢ ∂f3 ∂f3 ∂f3 ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ δQ3 ⎥ ⎢ f3 ⎥
⎢ ∂Q2 ∂Q3 ∂Q4 ⎥
⎢ ⎥ ⎢ ⎥
⎢ ∂f4 ∂f4 ⎥ ⎢ δQ ⎥ ⎢ f4 ⎥
⎢ ⎥ ⎢ 4
⎥ ⎢ ⎥
⎣ ∂Q3 ∂Q4 ⎦ ⎣ ⎦ ⎣ ⎦

where the two values in the first row of the square matrix are:

∂f1 −2 ( Q1 − Q2 )
= − 1.852hw Q10.852 (179)
∂Q1 2
Kd

∂f1 2 ( Q1 − Q2 )
= (180)
∂Q2 K 2d

Sprinkle & Trickle Irrigation Lectures Page 97 Merkley & Allen

 The two values in the last row of the square matrix, for n sprinklers, are:

∂fn 2 ( Qn−1 − Qn )
= (181)
∂Qn−1 K 2d

∂fn −2Qn−1
= − 1.852hw Qn0.852 (182)
∂Qn 2
Kd

and the three values in each intermediate row of the matrix are:

∂fi 2 ( Qi−1 − Qi )
= (183)
∂Qi−1 K 2d

∂fi 2 ( Qi+1 − Qi−1 )

= − 1.852hw Qi0.852 (184)
∂Qi Kd2

∂fi 2 ( Qi − Qi+1 )
= (185)
∂Qi+1 K 2d

where i is the row number

• This is a system of nonlinear algebraic equations

• The square matrix is a Jacobian matrix; all blank values are zero
• Solve for Q1, Q2, Q3, and Q4 (or up to Qn, in general) using the Newton-
Raphson method, Gauss elimination, and backward substitution (or other
solution method for a linear set of equations)
• Knowing the flow rates, you can go back and directly calculate the pressure
heads one by one

• The problem could be further generalized by allowing for different pipe sizes
in the lateral, by including minor losses, by allowing variable elevation
changes between sprinkler positions, etc.
• However, it is still a problem of solving for x unknowns and x equations
• For pumped systems (not gravity, as above), we could include a
mathematical representation of the pump characteristic curve to determine
the lateral hydraulic performance; that is, don’t assume a constant H0, but
replace it by some function

Merkley & Allen Page 98 Sprinkle & Trickle Irrigation Lectures

V. Brute-Force Approach

• There is a computer program that will do the above calculations for a gravity-
fed lateral with multiple sprinklers
• But, if you want to write your own program in a simpler way, you can do the
calculations by “brute-force” as follows:

1. Guess the pressure at the end of the lateral

2. Calculate q for the last sprinkler
3. Calculate hf over the distance Se to the next sprinkler upstream
4. Calculate ∆he over the same Se
5. Get the pressure at that next sprinkler and calculate the
sprinkler flow rate
6. Keep moving upstream to the lateral inlet
7. If the head is more than the available head, reduce the end
pressure and start over, else increase the pressure and start
over

• Below is a screenshot of a computer program that will do the above

calculations for a gravity-fed lateral with multiple sprinklers

Sprinkle & Trickle Irrigation Lectures Page 99 Merkley & Allen

Merkley & Allen Page 100 Sprinkle & Trickle Irrigation Lectures
Lecture 9
Mainline Pipe Design

I. Split-Line Laterals

Laterals are usually distributed evenly along a mainline because:

• More equal pump load at different lateral positions

• Reduced mainline cost
• Don’t need to “dead-head” back when finished (cross over to other side)
• But, split-line laterals may interfere with cultural operations (wet areas at both
ends of field)

Consider twin laterals operating in the same direction:

Sprinkle & Trickle Irrigation Lectures Page 101 Merkley & Allen
 • In the above case, and for only a single lateral on the mainline, the design of
the mainline is relatively simple – it is easy to find the most extreme
operating position
• However, the friction loss along the mainline is about four times greater than
for the split-line configuration
• Note that for the above two configurations the first sprinkler on the laterals
would be at 0.5Se from the inlet, unless the mainline is laid upon a roadway
in the field

Twin split-line laterals with dual mainline...

• Same combined mainline length

• No valves on mainline -- elbow at each lateral inlet
• More labor required, but mainline costs less because no valves, and because
the mainline is sized for the flow rate of one lateral over the entire mainline
length (not half at twice the capacity)
• For different lateral positions, you remove pieces of the mainline from the
longer section and put on the shorter section

II. General Design Considerations

• Look at extreme operating conditions for the mainline by varying lateral

positions (this can be more complicated for irregular field shapes and non-
uniform field slopes)
• Can use economic pipe selection method, but don’t make a big sacrifice in
terms of pressure uniformity along the mainline to save pumping costs
• Buried mainlines do not obstruct traffic, nor do they remove any land from
production. But, they cannot be moved from field to field

Merkley & Allen Page 102 Sprinkle & Trickle Irrigation Lectures
 • Buried mainlines tend to last longer, because they are not handled and
banged up after installation

Uphill Split-Line Mainline Design

I. Definition of the Example Problem

• See example 10.1 from the textbook, an uphill mainline design for two split-
line laterals
• For design, consider the two extreme lateral positions:

1. Both laterals at position B (mid-point of mainline)

2. One lateral at position A and the other at position C

• Divide the mainline into two logical lengths, at the mid-point, according to the
two extreme lateral positions
• Determine the total allowable head loss due to friction in each of these logical
lengths, then find two adjacent pipe sizes for each length
• Determine the lengths of each pipe size so that the total head loss is just
equal to the allowable head loss
• This is somewhat analogous to the procedure for designing dual pipe size
laterals

• This is the system layout (shown with both laterals at position B):

Sprinkle & Trickle Irrigation Lectures Page 103 Merkley & Allen
 • Pump provides at least 172 ft of head at P
• Lateral inlet pressure head is given as 125 ft of head (Eq. 9.2)
• Supply line and mainline are to be aluminum, in 30-ft lengths
• The figure below shows the hydraulic schematic for this mainline, with
separate friction loss profiles for the two extreme lateral positions
• The mainline is tentatively divided into sizes D1 and D2 for the first half (L1),
and D3 and D4 for the second half (L2). So, there are potentially four different
pipe sizes in the mainline from A to C.

• We do not yet know what these pipe diameters will be

• We do not yet know the required lengths of the different diameter pipes

• hf1 is for the case when both laterals are at B

• hf2, hf3, and hf4 are for the case when one lateral is at A and the other at C

• hf1 is for 500 gpm over L1

• hf2 is for 250 gpm over L1 + L2
• hf3 is for 250 gpm over L2
• hf4 is for 250 gpm over L1

Merkley & Allen Page 104 Sprinkle & Trickle Irrigation Lectures
II. Select the Size of the Supply Line

• We need to select the size of the supply line to know what the head loss is
from P to A (pressure at P is given as 172 ft of head)
• Assume no elevation change between P and A

• From continuity, Q = A V, then for an allowable velocity of 5 ft/s:

4Q 4 (1.11 cfs )
D= = = 0.53 ft (186)
πV π ( 5 fps )

• This is 6.4 inches. In Table 8.4, the 6-inch pipe has an inside diameter of
5.884 inches. With this size, the velocity at 500 gpm would be 5.9 ft/s, which
we will accept (could use 8-inch pipe, but 6-inch is probably OK)

• From Table 8.4, the head loss gradient in the 6-inch supply line at 500 gpm is
2.27 ft/100 ft. Then,

(hf )P− A = ( 2.27 ft /100 ft ) ⎛⎜

440 ft ⎞
⎟ = 10.0 ft (187)
⎝ 100 ⎠

• This means that the pressure head at A is 172 ft - 10.0 ft = 162 ft

III. Determine D1 and D2 for both Laterals at B

• We will tolerate (hf)1 head loss over section L1 of the mainline when both
laterals are operating at B. This will give the required hl at B.
• We can see that (hf)1 is defined as:

hf1 = 162 ft − hl + 0.5 ∆he (188)

hf1 = 162 ft − 125 ft − 7 ft = 30 ft (189)

• The allowable loss gradient in section L1 for both laterals operating at B is

⎛ 30 ft ⎞
( Ja )L1 = 100 ⎜ ⎟ = 5 ft per 100 ft (190)
⎝ 600 ft ⎠

• From Table 8.4, this is between the 5- and 6-inch pipe sizes, which have
respective loss gradients of 5.54 ft/100 ft and 2.27 ft/100 ft for the 500 gpm
flow rate. Therefore, choose D1 = 6 inch and D2 = 5 inch.

Sprinkle & Trickle Irrigation Lectures Page 105 Merkley & Allen
 • Now we must find out how long D1 should be so that the friction loss is really
equal to 30 ft of head...

LD1 ( 2.27 ) + ( 600 − LD1 )( 5.54 ) = 100 ( 30 ft ) (191)

• Solving the above, LD1 = 99.1 ft

• Using 30-ft pipe lengths, we adjust the length to...

LD1 = 90 ft of 6" pipe ( 3 sections ) (192)

LD2 = 510 ft of 5" pipe (17 sections ) (193)

• With the adjusted lengths, we will get 30.3 ft of head loss over section L1 for
500 gpm (this is close enough to the allowable 30 ft)

IV. Determine D3 and D4 for Laterals at A and C

• We will tolerate (hf)3 + (hf)4 head loss over the whole length of the mainline
when one lateral is operating at A and the other at C
• We can calculate (hf)4 straight away because we already know the pipe sizes
and lengths in section L1...

hf 4 =
( 90 ft )( 0.63 ) + ( 510 ft )(1.53 ) = 8.37 ft (194)
100
where the friction loss gradients for 250 gpm are 0.63 ft/100 ft (6” size) and
1.53 ft/100 ft (5” size). These values were taken from Table 8.4.

• Now we need to know the allowable loss for (hf)3, such that the pressure in
the mainline at C will be equal to hl (we know that the pressure at A is 162 ft -
- it is more than enough)...

hf 2 = hf1 + 0.5 ∆he = 23.0 ft (195)

hf 3 = hf 2 − hf 4 = 23.0 ft − 8.37 ft = 14.6 ft (196)

• The allowable loss gradient in section L2 for laterals at A and C is

⎛ 14.6 ft ⎞
( Ja )L2 = 100 ⎜ ⎟ = 2.43 ft per 100 ft (197)
⎝ 600 ft ⎠

Merkley & Allen Page 106 Sprinkle & Trickle Irrigation Lectures
 • From Table 8.1, this is between the 4- and 5-inch pipe sizes, which have
respective loss gradients of 4.66 ft/100 ft and 1.53 ft/100 ft for the 250 gpm
flow rate. Therefore, choose D3 = 5 inch and D4 = 4 inch.
• Now we must find out how long D3 should be so that the friction loss is really
equal to 14.6 ft of head...

LD3 (1.53 ) + ( 600 − LD3 )( 4.66 ) = 100 (14.6 ft ) (198)

• Solving the above, LD3 = 427 ft

• Using 30-ft pipe lengths, we adjust the length to...

LD3 = 420 ft of 5 " pipe (14 sections) (199)

LD4 = 180 ft of 4 " pipe (6 sections) (200)

• With the adjusted lengths, we will get 14.8 ft of head loss over section L2 for
250 gpm (this is close enough to the allowable 14.6 ft)

V. Check this Mainline Design for an Intermediate Position

• Just to be sure, suppose that one lateral is operating halfway between A and
B, and the other halfway between B and C
• The allowable friction loss from point A to the furthest lateral is (hf)2 + ¼∆he,
or 23.0 ft + 3.5 ft = 26.5 ft. The actual friction loss would be:

hf = 0.01[(2.27)(90) + (5.54)(210) + (1.53)(600)] = 22.9 ft (201)

• OK, the head in the mainline at the furthest lateral is more than enough
• See the figure below for a graphical interpretation of the two laterals in
intermediate positions

Sprinkle & Trickle Irrigation Lectures Page 107 Merkley & Allen
VI. Comments About the Mainline Design

• Both D2 and D3 are the same size in this example

• If we were lucky, both D1 and D2 (or D3 and D4) could be the same size, but
that means the friction loss gradient would have to be just right

• The lateral inlet pressure will be just right when both laterals operate at B
• The lateral inlet pressure will be just right for a lateral operating at C
• The lateral inlet pressure will always be too high for a lateral operating
between A and B (the inlet pressure to the mainline, at A, is always 162 ft)

• We designed D1 and D2 for the condition when both laterals are at B. This is
a more demanding condition for L1 than when one lateral is at A and the
other at C (in this case, only half the system flow rate is in L1). So, we don’t
need to “check” D1 and D2 again for the case when the laterals are at A and
C.

• We didn’t consider the hydrant loss from the mainline into the sprinkler
lateral, but this could be added to the requirements (say, effective hl)

• This design could be also done using the economic pipe selection method (or
another pipe selection method. It would be a good idea to check to see if the
Merkley & Allen Page 108 Sprinkle & Trickle Irrigation Lectures
 172 ft at the pump (point A) could be reduced by using larger supply and
mainline pipes, thus reducing the annual energy costs. However, if the 172 ft
were due to gravity supply, the design would still be all right.
• However, the velocity in the 5-inch pipe at 500 gpm is too high, at 8.5 fps
(always check velocity limits when sizing pipes!)

VII. Both Laterals Operating at Point C

• How would the mainline design change if it were not split line operation, and
both laterals were operating at location C?
• In this case, intuition and past experience tells us location C is the critical
lateral position – if you don’t agree, then you should test other lateral
positions to convince yourself
• We will tolerate (hf)2 head loss over the entire 1,200-ft length of the mainline
when both laterals are operating at C. This will give the required hl at C.
• We can see that (hf)2 is defined as:

hf 2 = 162 ft − hl + ∆he (202)

hf 2 = 162 ft − 125 ft − 14 ft = 23 ft (203)

• The allowable loss gradient over the length of the mainline for both laterals
operating at C is

⎛ 23 ft ⎞
Ja = 100 ⎜ ⎟ = 1.92 ft per 100 ft (204)
⎝ 1,200 ft ⎠

• From Table 8.4, this is between the 6- and 8-inch pipe sizes, which have
respective loss gradients of 2.27 ft/100 ft and 0.56 ft/100 ft for the 500 gpm
flow rate
• Determine the respective pipe lengths so that the friction loss is really equal
to 23 ft of head...

L8 " ( 0.56 ) + (1,200 − L8 " )( 2.27 ) = 100 ( 23 ft ) (205)

• Solving the above, L8” = 248 ft

• Using 30-ft pipe lengths, we adjust the length to...

L8 " = 270 ft of 8" pipe ( 9 sections ) (206)

L6 " = 930 ft of 6" pipe ( 31 sections ) (207)

Sprinkle & Trickle Irrigation Lectures Page 109 Merkley & Allen
Merkley & Allen Page 110 Sprinkle & Trickle Irrigation Lectures
Lecture 10
Minor Losses & Pressure Requirements

I. Minor Losses

• Minor (or “fitting”, or “local”) hydraulic losses along pipes can often be
estimated as a function of the velocity head of the water within the particular
pipe section:
V2
hml = Kr (208)
2g

where hml is the minor loss (m or ft); V is the mean flow velocity, Q/A (m/s or
fps); g is the ratio of weight to mass (9.81 m/s2 or 32.2 ft/s2); and Kr is a
coefficient, dependent on the type of fitting (valve, bend, transition,
constriction, etc.)

• Minor losses include head losses through/past hydrants, couplers, valves,

pipe elbows, “tees” and other fittings (see Tables 11.1 and 11.2)
• For example, there is some loss when water flows through a hydrant, but
also some loss when water flows in a pipe past the location of a closed
hydrant
• Kr = 0.3 to 0.6 for flow in a pipeline going past a closed hydrant, whereby the
velocity in the pipeline is used to compute hml
• Kr = 0.4 to 0.8 for flow in a pipeline going past an open hydrant; again, the
velocity in the pipeline is used to compute hml
• Kr = 6.0 to 8.0 for flow from a pipeline through a completely open hydrant. In
this case, compute hml using the velocity of the flow through the lateral fitting
on the hydrant, not the flow in the source pipeline.
• For flow through a partially open hydrant, Kr increases beyond the 6.0 to 8.0
magnitude, and the flow rate decreases correspondingly (but not linearly)

• In using Tables 11.1 and 11.2 for hydrants, the nominal diameter (3, 4, 5,
and 6 inches) is the diameter of the hydrant and riser pipe, not the diameter
of the source pipeline
• Use the diameter of the hydrant for Kr and for computing Vr. However, for
line flow past a hydrant, use the velocity in the source pipeline, as indicated
above.
• Always use the largest velocity along the path which the water travels – this
may be either upstream or downstream of the fitting
• Do not consider velocities along paths through which the water does not flow

• In Table 11.2, for a sudden contraction, Kr should be defined as:

Sprinkle & Trickle Irrigation Lectures Page 111 Merkley & Allen
 ( )
2
Kr = 0.7 1 − Dr2 (209)

where Dr is the ratio of the small to large inside diameters (Dsmall/Dlarge)

• Allen (1991) proposed a regression equation for gradual contractions and
expansions using data from the Handbook of Hydraulics (Brater & King
1976):

( )
2
Kr = K f 1 − Dr2 (210)

where Kf is defined as:

K f = 0.7 − cos(f) ⎡⎣cos(f) ( 3.2cos(f) − 3.3 ) + 0.77 ⎤⎦ (211)

and f is the angle of the expansion or contraction in the pipe walls (degrees
or radians), where f ≥ 0

• For straight sides (no expansion or contraction), f = 0° (whereby Kf = 0.03)

• For an abrupt change in pipe diameter (no transition), f = 90° (whereby Kf =
0.7)
• The above regression equation for Kf gives approximate values for
approximate measured data, some of which has been disputed
• In any case, the true minor head loss depends on more than just the angle of
the transition

Expansion

Contraction
• For a sudden (abrupt) expansion, the head loss can also be approximated as
a function of the difference of the mean flow velocities upstream and
downstream:

Merkley & Allen Page 112 Sprinkle & Trickle Irrigation Lectures
 hml =
( Vus − Vds )
2
(212)
2g

• An extreme (albeit unrealistic) case is for Vds = 0 and hml = Vus2/2g (total
conversion of velocity head)
• Various other equations (besides those given above) for estimating head loss
in pipe expansions and contractions have been proposed and used by
researchers and engineers

Minor Loss Example

• A mainline with an open lateral hydrant valve has a diameter of 200 mm ID

upstream of the hydrant, and 150 mm downstream of the hydrant
• The diameter of the hydrant opening to the lateral is 75 mm
• Qupstream = 70 lps and Qlateral = 16 lps
• The pressure in the mainline upstream of the hydrant is 300 kPa

• The mean flow velocities are:

0.070 m3 / s
V200 = = 2.23 m / s (213)
⎛ π(0.200 m)2 ⎞
⎜⎜ ⎟⎟
⎝ 4 ⎠

0.070 − 0.016 m3 / s
V150 = = 3.06 m / s (214)
⎛ π(0.150 m)2 ⎞
⎜⎜ ⎟⎟
⎝ 4 ⎠

0.016 m3 / s
Vhydrant = = 3.62 m / s (215)
⎛ π(0.075 m)2 ⎞
⎜⎜ ⎟⎟
⎝ 4 ⎠
Sprinkle & Trickle Irrigation Lectures Page 113 Merkley & Allen
 • Note that V200 and V150 are both above the normal design limit of about 2 m/s
• The head loss past the open hydrant is based on the higher of the upstream
and downstream velocities, which in this example is 3.06 m/s
• From Table 11.1, the Kr for flow past the open hydrant (line flow; 6” mainline)
is 0.5; thus,

(3.06)2
(hml )past = 0.5 = 0.24m (216)
2(9.81)

• The head loss due to the contraction from 200 mm to 150 mm diameter (at
the hydrant) depends on the transition
• If it were an abrupt transition, then:

2
⎡ ⎛ 150 ⎞2 ⎤
Kr = 0.7 ⎢1 − ⎜ ⎟ ⎥ = 0.13 (217)
⎢⎣ ⎝ 200 ⎠ ⎥⎦

• And, if it were a 45° transition, Kf = 0.67, also giving a Kr of 0.13

• Then, the head loss is:
(3.06)2
(hml )contraction = 0.13 = 0.06m (218)
2(9.81)

• Thus, the total minor loss in the mainline in the vicinity of the open hydrant is
about 0.24 + 0.06 = 0.30 m (0.43 psi).

• The loss through the hydrant is determined by taking Kr = 8.0 (Table 11.1; 3”
hydrant):
(3.62)2
(hml )through = 8.0 = 5.3m (219)
2(9.81)

• This is a high loss through the hydrant (about 7.6 psi), so it may be advisable
to use a larger diameter hydrant.
• The pressure in the mainline downstream of the hydrant is (9.81 kPa/m):

⎛ V200
2 2 ⎞
− V150
P150 = P200 − γ (hml )past + γ⎜ ⎟
⎜ 2g ⎟
⎝ ⎠
(220)
⎛ (2.23) − (3.06)
2 2 ⎞
P150 = 300 − (9.81)(0.24) + 9.81⎜ = 295kPa
⎜ 2(9.81) ⎟⎟
⎝ ⎠

Merkley & Allen Page 114 Sprinkle & Trickle Irrigation Lectures
II. Total Dynamic Head

• The Total Dynamic Head (TDH) is the head that the pump “feels” or “sees”
while working, and is calculated to determine the pump requirements
• It includes the elevation that the water must be lifted from the source (not
necessarily from the pump elevation itself) to the outlet, the losses due to
“friction”, the pressure requirement at the outlet, and possibly the velocity
head in the pipeline
• For a sprinkler system, the value of TDH depends on the positions of the
laterals, so that it can change with each set. Pump selection is usually made
for the “critical” or extreme lateral positions, that is, for the “worst case
scenario”.
• Keller & Bliesner recommend the addition of a “miscellaneous” loss term,
equal to 20% of the sum of all “friction” losses. This accounts for:

1. Uncertainty in the Kr values (minor losses)

2. Uncertainty in the Hazen-Williams C values
3. Aging of pipes (increase in losses)
4. Wear of pump impellers and casings

• Losses in connectors or hoses from the mainline to laterals, if present, must

also be taken into account when determining the TDH
• See Example Calculation 11.2 in the textbook
• The next two lectures will provide more information about TDH and pumps

III. The System Curve

• The system curve determines the relationship between TDH and flow rate
• This curve is approximately parabolic, but can take more complex shapes
• Note that head losses in pipe systems are approximately proportional to the
square of the flow rate (Q2 or V2)
• For the Hazen-Williams equation, these losses are actually proportional to
Q1.852 or V1.852
• For standard, non-FCN, sprinkler nozzles, the head at the sprinkler is also
proportional to Q2
• Sprinkler systems can have a different system curve for each position of the
lateral(s)
• Defining the system curve, or the “critical” system curve, is important for
pump selection because it determines, in part, the operating point (TDH and
Q) for the system

Sprinkle & Trickle Irrigation Lectures Page 115 Merkley & Allen
IV. Valving a Pump

• A throttle valve may be required at a pump:

(a) Filling of the system’s pipes

• The head is low, and the flow rate is high

• Pump efficiency is low and power requirements may be higher
• Water hammer damage can result as the system fills
• Air vents and other appurtenances can be “blown off”
• For the above reasons, it is advisable to fill the system slowly

(b) To avoid cavitation, which damages the pump, pipes and

appurtenances
(c) To control the TDH as the sprinklers are moved to different sets

• Throttle valves can be automatic or manual

Pressure Requirements & Pumps

I. Types of Pumps

1. Positive Displacement

• Piston pumps
• Rotary (gear) pumps
• Extruding (flexible tube) pumps

2. Variable Displacement

• Centrifugal pumps
• Injector pumps
• Jet pumps

• The above lists of pump types are not exhaustive

• Positive displacement pumps have a discharge that is nearly independent of
the downstream (resistive) pressure. That is, they produce a flow rate that is
relatively independent of the total dynamic head, TDH

Merkley & Allen Page 116 Sprinkle & Trickle Irrigation Lectures
 Positive Displacement Pumps

Axial-Flow Impeller

Closed Centrifugal Pump Impeller

Jet Pump

Sprinkle & Trickle Irrigation Lectures Page 117 Merkley & Allen
 • But, with positive displacement pumps, the required pumping energy is a
linear function of the pressure
• Positive displacement pumps can be used with thick, viscous liquids. They
are not commonly used in irrigation and drainage, except for the injection of
chemicals into pipes and for sprayers
• Piston-type pumps can develop high heads at low flow rates
• Air injection, or jet pumps are typically used in some types of well drilling
operations. The air bubbles effectively reduce the liquid density and this
assists in bringing the drillings up out of the well. Needs a large capacity air
compressor.
• Homologous pumps are geometrically similar pumps, but of different sizes

II. Centrifugal Pumps

1. Volute Case This is the most common type of irrigation and drainage pump
(excluding deep well pumps). Produce relatively high flow rates at low
pressures.

2. Diffuser (Turbine) The most common type for deep wells. Designed to lift
water to high heads, typically using multiple identical “stages” in series,
stacked up on top of each other.

3. Mixed Flow Uses a combination of centrifugal and axial flow action. For high
capacity at low heads.

4. Axial Flow Water flows along the axis of impeller rotation, like a boat
propeller. Appropriate for high discharge under very low lift (head). An
example is the pumping plant on the west side of the Great Salt Lake.

5. Regenerative The characteristics of these pumps are those of a combination

of centrifugal and rotary, or gear, pumps. Shut-off head is well-defined, but
efficiency is relatively low. Not used in irrigation and drainage.

• In general, larger pumps have higher maximum efficiencies (they are more
expensive, and more effort is given toward making them more efficient)
• Impellers can be open, semi-open, or closed. Open impellers are usually
better at passing solids in the pumped liquid, but they are not as strong as
closed impellers
• Double suction inlet pumps take water in from both sides and can operate
without axial thrust

Merkley & Allen Page 118 Sprinkle & Trickle Irrigation Lectures
 Closed Impeller Semi-Open Impeller Open Impeller

Characteristic Curve

• The pump “characteristic curve” defines the relationship between total

dynamic head, TDH, and discharge, Q
• The characteristic curve is unique for a given pump design, impeller
diameter, and pump speed
• The characteristic curve has nothing to do with the “system” in which the
pump operates
• The “shut-off” head is the TDH value when Q is zero (but the pump is still
operating)
• The shut-off head can exceed the recommended operating pressure, or even
the bursting pressure, especially with some thin-wall plastic pipes
Total Dynamic Head, TDH

Efficiency
Cha
ra cte
risi
cC
urv
Shut-Off Head

Power

0
0 Flow Rate, Q

Sprinkle & Trickle Irrigation Lectures Page 119 Merkley & Allen
III. Centrifugal Pumps in Parallel

• Pumps in PARALLEL

Total Dynamic Head, TDH

means that the total
flow is divided into two
or more pumps
• Typical installations are
On Two
for a single inlet pipe, e in P
Pu aral
branched into two lel
m
pumps, with the outlets

p
from the pumps
converging to a single
discharge pipe
• If only one of the
pumps operates, some 0
type of valve may be 0 Flow Rate, Q
required so that flow does not flow backwards through the idle pump
• Flow rate is additive in this case

pump 1

pump 2
Two Pumps in Parallel

IV. Centrifugal Pumps in Series

• Pumps in SERIES means that the total flow passes through each of two or
more pumps in line
• Typical installations are for increasing pressure, such as with a booster pump
• Head is additive in this case

Merkley & Allen Page 120 Sprinkle & Trickle Irrigation Lectures
 Total Dynamic Head, TDH
Two
in Se
rie
s

One P
ump

0
0 Flow Rate, Q
• It is common for turbine (well) pumps to operate in series
• For centrifugal pumps, it is necessary to exercise caution when installing in
series because the efficiency can be adversely affected
• May need straightening vanes between pumps to reduce swirling
• Note that the downstream pump could cause negative pressure at the outlet
of the US pump, which can be a problem

pump 1 pump 2

Two Pumps in Series

Sprinkle & Trickle Irrigation Lectures Page 121 Merkley & Allen
Merkley & Allen Page 122 Sprinkle & Trickle Irrigation Lectures
Lecture 11
Pumps & System Curves

I. Pump Efficiency and Power

• Pump efficiency, Epump

water horsepower WHP

Epump = = (221)
brake horsepower BHP

where brake horsepower refers to the input power needed at the pump shaft
(not necessarily in “horsepower”; could be watts or some other unit)

• Pump efficiency is usually given by the pump manufacturer

• Typically use the above equation to calculate required BHP, knowing Epump
• Water horsepower is defined as:

QH
WHP = (222)
3956
where WHP is in horsepower; Q in gpm; and H in feet of head. The
denominator is derived from:

( 62.4 lbs/ft3 ) ( gal/min )( ft ) QH

γQH = ≈
( 33,000 ft-lbs/min-HP ) (7.481 gal/ft3 ) 3956
(223)

where γ = ρg, and ρ is water density. In metric units:

(1000 kg/m3 )(9.81 m/s2 ) (l/s )(m ) QH

WHP = ρgQH = =
( )
(224)
1000 l/m3
(1000 W/kW ) 102

where WHP is in kW; Q in lps; and H in meters of head

1 HP=0.746 kW (225)

• Total Dynamic Head, TDH, is defined as:

P V2
TDH = ∆Elev + hf + + (226)
γ 2g

Sprinkle & Trickle Irrigation Lectures Page 123 Merkley & Allen
 where the pressure, P, and velocity, V, are measured at the pump outlet, and
hf is the total friction loss from the entrance to the exit, including minor losses

• At zero flow, with the pump running,

P
TDH = ∆Elev + (227)
γ

but recognizing that in some cases P/γ is zero for a zero flow rate

• The elevation change, ∆Elev, is positive for an increase in elevation (i.e.

lifting the water)
• Consider a turbine pump in a well:

Merkley & Allen Page 124 Sprinkle & Trickle Irrigation Lectures
 Consider a centrifugal pump:

II. Example TDH & WHP Calculation

• Determine TDH and WHP for a centrifugal pump discharging into the air...

Head loss due to friction:

hf = hscreen + 3helbow + hpipe

(228)

for PVC, ε ≈ 1.5(10)-6 m, relative roughness is:

Sprinkle & Trickle Irrigation Lectures Page 125 Merkley & Allen
 ε 1.5(10)−6
= = 0.0000051 (229)
D 0.295

Average velocity,

Q 4(0.102)
V= = = 1.49 m/s (230)
A π(0.295)2

Reynolds number, for 10°C water:

VD (1.49 m/s )( 0.295 m )

NR = = = 336,600 (231)
ν 1.306(10)−6 m2 / s

• From the Moody diagram, f = 0.0141

• From the Blasius equation, f = 0.0133
• From the Swamee-Jain equation, f = 0.0141 (same as Moody)

Using the value from Swamee-Jain,

⎛ 1,530 ⎞ (1.49 )
2
L V2
hpipe =f = 0.0141⎜ ⎟ = 8.27 m (232)
D 2g ⎝ 0.295 ⎠ 2 ( 9.81)

Kinematic Viscosity
Water Temperature (°C)
(m2/s)
0 0.000001785
5 0.000001519
10 0.000001306
15 0.000001139
20 0.000001003
25 0.000000893
30 0.000000800
40 0.000000658
50 0.000000553
60 0.000000474

The values in the above table can be closely approximated by:

( )
−1
ν = 83.9192 T 2 + 20,707.5 T + 551,173 (233)

where T is in ºC; and ν is in m2/s

Merkley & Allen Page 126 Sprinkle & Trickle Irrigation Lectures
 From Table 11.2, for a 295-mm (12-inch) pipe and long radius 45-deg flanged
elbow, the Kr value is 0.15

V2 (1.49)2
helbow = Kr = (0.15) = (0.15)(0.11) = 0.017 m (234)
2g 2(9.81)
For the screen, assume a 0.2 m loss. Then, the total head loss is:

hf = 0.2 + 3(0.017) + 8.27 = 8.5 m (235)

With the velocity head of 0.11 m, the total dynamic head is:

TDH = 31 + 8.5 + 0.11 ≈ 40 m (236)

The water horsepower is:

QH (102 lps )( 40 m )
WHP = = = 40 kW (54 HP)
102 102 (237)

The required brake horsepower is:

WHP 40 kW
BHP = = ≈ 53 kW (71 HP)
Epump 0.76
(238)

• This BHP value would be used to select a motor for this application
• These calculations give us one point on the system curve (Q and TDH)
• In this simple case, there would be only one system curve:

System Curve
60

50
Total Dynamic Head (m)

40

30

20

10

0
0 20 40 60 80 100 120 140 160 180
Discharge (lps)

Sprinkle & Trickle Irrigation Lectures Page 127 Merkley & Allen
III. System Curves

• The “system curve” is a graphical representation of the relationship between

discharge and head loss in a system of pipes
• The system curve is completely independent of the pump characteristics
• The basic shape of the system curve is parabolic because the exponent on
the head loss equation (and on the velocity head term) is 2.0, or nearly 2.0
• The system curve will start at zero flow and zero head if there is no static lift,
otherwise the curve will be vertically offset from the zero head value

• Most sprinkle and trickle irrigation systems have more than one system curve
because either the sprinklers move between sets (periodic-move systems),
move continuously, or “stations” (blocks) of laterals are cycled on and off
• The intersection between the system and pump characteristic curves is the
operating point (Q and TDH)
• A few examples of system curves:

1. All Friction Loss and No Static Lift

Merkley & Allen Page 128 Sprinkle & Trickle Irrigation Lectures
 2. Mostly Static Lift, Little Friction Loss

3. Negative Static Lift

Sprinkle & Trickle Irrigation Lectures Page 129 Merkley & Allen
 4. Two Different Static Lifts in a Branching Pipe

5. Two Center Pivots in a Branching Pipe Layout

• The figure below shows two center pivots supplied by a single pump on a
river bank
• One of the pivots (#1) is at a higher elevation than the other, and is
further from the pump – it is the “critical” branch of the two-branch pipe
system
• Center pivot #2 will have excess pressure when the pressure is correct at
Center pivot #1, meaning it will need pressure regulation at the inlet to the
pivot lateral
• Use the critical branch (the path to Center pivot #1, in this case) when
calculating TDH for a given operating condition – Do Not Follow Both
Branches when calculating TDH
• if you cannot determine which is the critical branch by simple inspection,
you must test different branches by making calculations to determine
which is the critical one
• Note that the system curve will change with center pivot lateral position
when the topography is sloping and or uneven within the circle
• Of course, the system curve will also be different if only one of the center
pivots is operating

Merkley & Allen Page 130 Sprinkle & Trickle Irrigation Lectures
 Center pivot #1

Center pivot #2
275 kPa
833 m

275 kPa
750 m

308 m river
pump

6. A Fixed Sprinkler System with Multiple Operating Laterals

• The next figure shows a group of laterals in parallel, attached to a

common mainline in a fixed sprinkler system
• All of the sprinklers operate at the same time (perhaps for frost control or
crop cooling purposes, among other possibilities)
• This is another example of a branching pipe system
• Since the mainline runs uphill, it is easy to determine by inspection that
the furthest lateral will be the critical branch in this system layout – use
this branch to determine the TDH for a given system flow rate
• Hydraulic calculations would be iterative because you must also
determine the flow rate to each of the laterals since the flow rate is
changing with distance along the mainline
• But in any case, Do Not Follow Multiple Branches when determining the
TDH for a given system flow rate
• Remember that TDH is the resistance “felt” by the pump for a given flow
rate and system configuration

Sprinkle & Trickle Irrigation Lectures Page 131 Merkley & Allen
 lateral #5 (critical lateral)

lateral #4

lateral #3

mainline

uphill
lateral #2

lateral #1

pump

7. Two Flow Rates for Same Head on Pump Curve

• Consider the following graph

• “A” has a unique Q for each TDH value
• “B” has two flow rates for a given head, over a range of TDH values
• Pumps with a characteristic curve like “B” should usually be avoided

Merkley & Allen Page 132 Sprinkle & Trickle Irrigation Lectures
Total Dynamic Head, TDH

1 2
Unstable Stable
A
B
urv e
te mC
Sys

0
0 Flow Rate, Q

Affinity Laws and Cavitation

I. Affinity Laws

1. Pump operating speed:

2 3
Q1 N1 H1 ⎛ N1 ⎞ BHP1 ⎛ N1 ⎞
= =⎜ ⎟ =⎜ ⎟ (239)
Q2 N2 H2 ⎝ N2 ⎠ BHP2 ⎝ N2 ⎠

where Q is flow rate; N is pump speed (rpm); H is head; and BHP is “brake
horsepower”

• The first relationship involving Q is valid for most pumps

• The second and third relationships are valid for centrifugal, mixed-flow,
and axial-flow pumps

2. Impeller diameter:

Sprinkle & Trickle Irrigation Lectures Page 133 Merkley & Allen
 2 3
Q1 D1 H1 ⎛ D1 ⎞ BHP1 ⎛ D1 ⎞
= =⎜ ⎟ =⎜ ⎟ (240)
Q2 D 2 H2 ⎝ D2 ⎠ BHP2 ⎝ D2 ⎠

• These three relationships are valid only for centrifugal pumps

• These relationships are not as accurate as those involving pump
operating speed, N (rpm)

Comments:

• The affinity laws are only valid within a certain range of speeds, impeller
diameters, flow rates, and heads
• The affinity laws are more accurate near the region of maximum pump
efficiency (which is where the pump should operate if it is selected correctly)

• It is more common to apply these laws to reduce the operating speed or to

reduce the impeller diameter (diameter is never increased)
• We typically use these affinity laws to fix the operating point by shifting the
pump characteristic curve so that it intersects the system curve at the
desired Q and TDH

II. Fixing the Operating Point

Combine the first two affinity law relationships to obtain:

2
H1 ⎛ Q1 ⎞
=⎜ ⎟ (241)
H2 ⎝ Q2 ⎠

• If this relationship is plotted with the pump characteristic curve and the
system curve, it is called the “equal efficiency curve”
• This is because there is typically only a small change in efficiency with a
small change in pump speed
• Note that the “equal efficiency curve” will pass through the origin (when Q is
zero, H is zero)

• Follow these steps to adjust the: (1) speed; or, (2) impeller diameter, such
that the actual operating point shifts up or down along the system curve:

1. Determine the head, H2, and discharge, Q2, at which the

system should operate (the desired operating point)
2. Solve the above equation for H1, and make a table of H1 versus
Q1 values (for fixed H2 and Q2):

Merkley & Allen Page 134 Sprinkle & Trickle Irrigation Lectures
 2
⎛Q ⎞
H1 = H2 ⎜ 1 ⎟ (242)
⎝ Q2 ⎠
3. Plot the values from this table on the graph that already has the
pump characteristic curve
4. Locate the intersection between the pump characteristic curve
and the “equal efficiency curve”, and determine the Q3 and H3
values at this intersection
5. Use either of the following equations to determine the new
pump speed (or use equations involving D to determine the trim
on the impeller):

⎛Q ⎞ H2
Nnew = Nold ⎜ 2 ⎟ or, Nnew = Nold (243)
⎝ Q3 ⎠ H3

6. Now your actual operating point will be the desired operating

point (at least until the pump wears appreciably or other
physical changes occur)

• You cannot directly apply any of the affinity laws in this case because you will
either get the right discharge and wrong head, or the right head and wrong
discharge

Apply Affinity
Law from Here
e
rv
Cu
m
te
ys

3 S Operating Point
Head

Desired without Adjustment

Operating Point
2
correct head
Pu
incorrect discharge
rve

m
p
Cu

g!

C
ron
y
nc

ur
W
cie

ve
ffi

l E
ua
Eq
0
0 Flow Rate
Sprinkle & Trickle Irrigation Lectures Page 135 Merkley & Allen
III. Specific Speed

• The specific speed is a dimensionless index used to classify pumps

• It is also used in pump design calculations

Pump Type Specific Speed

Centrifugal (volute case) 500 - 5,000
Mixed Flow 4,000 - 10,000
Axial Flow 10,000 - 15,000

• To be truly dimensionless, it is written as:

2πN Q
Ns = (244)
(gH)0.75

where the 2π is to convert revolutions (dimensional) to radians

(dimensionless)

• Example: units could be N = rev/s; Q = m3/s; g = m/s2; and H = m

• However, in practice, units are often mixed, the 2π is not included, and even
g may be omitted
• This means that Ns must not only be given numerically, but the exact
definition must be specified

IV. Cavitation

• Air bubbles will form (the water boils) when the pressure in a pump or
pipeline drops below the vapor pressure
• If the pressure increases to above the vapor pressure downstream, the
bubbles will collapse
• This phenomenon is called “cavitation”
• Cavitation often occurs in pumps, hydroelectric turbines, pipe valves, and
ship propellers
• Cavitation is a problem because of the energy released when the bubbles
collapse; formation and subsequent collapse can take place in only a few
thousandths of a second, causing local pressures in excess of 150,000 psi,
and local speeds of over 1,000 kph
• The collapse of the bubbles has also been experimentally shown to emit
small flashes of light (“sonoluminescence”) upon implosion, followed by rapid
expansion on shock waves
• Potential problems:

1. noise and vibration

2. reduced efficiency in pumps
Merkley & Allen Page 136 Sprinkle & Trickle Irrigation Lectures
 3. reduced flow rate and head in pumps
4. physical damage to impellers, volute case, piping, valves

• From a hydraulics perspective cavitation is to be avoided

• But, in some cases cavitation is desirable. For example,

1. acceleration of chemical reactions

2. mixing of chemicals and or liquids
3. ultrasonic cleaning

• Water can reach the boiling point by:

1. reduction in pressure (often due to an increase in velocity)

2. increase in temperature

• At sea level, water begins to boil at 100°C (212°F)

• But it can boil at lower temperatures if the pressure is less than that at mean
sea level (14.7 psi, or 10.34 m)

Patmospheric

Pvapor

container with water

• Pump inlets often have an eccentric reducer (to go from a larger pipe
diameter to the diameter required at the pump inlet:

1. Large suction pipe to reduce friction loss and increase NPSHa, especially
where NPSHa is already too close to NPSHr (e.g. high-elevation pump
installations where the atmospheric pressure head is relatively low)
2. Eccentric reducer to avoid accumulation of air bubbles at the top of the
pipe

• See the following figure…

Sprinkle & Trickle Irrigation Lectures Page 137 Merkley & Allen
 Required NPSH

• Data from the manufacturer are available for most centrifugal pumps
• Usually included in this data are recommendations for required Net Positive
Suction Head, NPSHr
• NPSHr is the minimum pressure head at the entrance to the pump, such that
cavitation does not occur in the pump
• The value depends on the type of pump, its design, and size
• NPSHr also varies with the flow rate at which the pump operates
• NPSHr generally increases with increasing flow rate in a given pump
• This is because higher velocities occur within the pump, leading to lower
pressures
• Recall that according to the Bernoulli equation, pressure will tend to
decrease as the velocity increases, elevation being the same
• NPSHr is usually higher for larger pumps, meaning that cavitation can be
more of a problem in larger pump sizes

Available NPSH

• The available NPSH, or NPSHa, is equal to the atmospheric pressure minus

all losses in the suction piping (upstream side of the pump), vapor pressure,
velocity head in the suction pipe, and static lift
• When there is suction at the pump inlet (pump is operating, but not yet
primed), the only force available to raise the water is that of the atmospheric
pressure
• But, the suction is not perfect (pressure does not reduce to absolute zero in
the pump) and there are some losses in the suction piping

V2
NPSHa = hatm − hvapor − hf − hlift − (245)
2g

Merkley & Allen Page 138 Sprinkle & Trickle Irrigation Lectures
 Vapor Pressure Head

Atmospheric Pressure Head

Friction Loss

Static Lift

Velocity Head

Available NPSH

• If the pump could create a “perfect vacuum” and there were no losses, the
water could be “sucked up” to a height of 10.34 m (at mean sea level)
• Average atmospheric pressure is a function of elevation above msl

perfect
vacuum
10.34 m

sea level

water

• 10.34 m is equal to 14.7 psi, or 34 ft of head

• Vapor pressure of water varies with temperature

Sprinkle & Trickle Irrigation Lectures Page 139 Merkley & Allen
 11
10
9
Vapor Pressure Head (m)

8
7
6
5
4
3
2
1
0
0 10 20 30 40 50 60 70 80 90 100
Water Temperature (C)

• Herein, when we say “vapor pressure,” we mean “saturation vapor pressure”

• Saturation vapor pressure head (as in the above graph) can be calculated as
follows:
⎛ 17.27 T ⎞
hvapor = 0.0623exp ⎜ ⎟ (246)
⎝ T + 237.3 ⎠
for hvapor in m; and T in ºC

• Mean atmospheric pressure head is essentially a function of elevation above

mean sea level (msl)
• Two ways to estimate mean atmospheric pressure head as a function of
elevation:

Straight line:
hatm = 10.3 − 0.00105 z (247)

Exponential curve:

5.26
⎛ 293 − 0.0065 z ⎞
hatm = 10.33 ⎜ ⎟ (248)
⎝ 293 ⎠
where hatm is atmospheric pressure head (m of water); and z is elevation
above mean sea level (m)

Merkley & Allen Page 140 Sprinkle & Trickle Irrigation Lectures
 10.50

10.00
Mean atmospheric pressure (m)
9.50
Straight Line (m)
Exponential Curve (m)
9.00

8.50

8.00

7.50

7.00

6.50
0 500 1000 1500 2000 2500 3000
Elevation above msl (m)

V. Example Calculation of NPSHa

Sprinkle & Trickle Irrigation Lectures Page 141 Merkley & Allen
 1. Head Loss due to Friction

ε 0.2 mm
= = 0.000556 (249)
D 360 mm

viscosity at 20°C, ν = 1.003(10)-6 m2/s

flow velocity,
Q 0.100 m3 / s
V= = = 0.982 m/s (250)
π
A
( 0.36 ) 2
4
Reynold’s Number,

VD ( 0.982 )( 0.36 )
NR = = = 353,000 (251)
ν 1.003(10)−6

Darcy-Weisbach friction factor, f = 0.0184

velocity head,
V 2 (0.982)2
= = 0.049 m (252)
2g 2g

head loss in suction pipe,

L V2 ⎛ 8.1 ⎞
(hf )pipe =f = 0.0184 ⎜ ⎟ ( 0.049 ) = 0.0203 m (253)
D 2g ⎝ 0.36 ⎠

local losses, for the bell-shaped entrance, Kr = 0.04; for the 90-deg elbow, Kr =
0.14. Then,

(hf )local = ( 0.04+0.14 )(0.049 ) = 0.0088 m (254)

finally,

(hf )total = (hf )pipe + (hf )local = 0.0203 + 0.0088 = 0.0291 m (255)

Merkley & Allen Page 142 Sprinkle & Trickle Irrigation Lectures
 2. Vapor Pressure

for water at 20°C, hvapor = 0.25 m

3. Atmospheric Pressure

at 257 m above msl, hatm = 10.1 m

4. Static Suction Lift

• the center of the pump is 3.0 m above the water surface

• (the suction lift would be negative if the pump were below the water
surface)

5. Available NPSH
V2
NPSHa = hatm − hvapor − (hf )total − hlift −
2g (256)
NPSHa = 10.1 − 0.25 − 0.0291 − 3.0 − 0.049 = 6.77 m

VI. Relationship Between NPSHr and NPSHa

• If NPSHr < NPSHa, there should be no cavitation

• If NPSHr = NPSHa, cavitation is impending
• As the available NPSH drops below the required value, cavitation will
become stronger, the pump efficiency will drop, and the flow rate will
decrease
• At some point, the pump would “break suction” and the flow rate would go to
zero (even with the pump still operating)

Sprinkle & Trickle Irrigation Lectures Page 143 Merkley & Allen
Merkley & Allen Page 144 Sprinkle & Trickle Irrigation Lectures
Lecture 12
Center Pivot Design & Operation

I. Introduction and General Comments

• Center pivots are used on about half of the sprinkler-irrigated land in the USA
• Center pivots are also found in many other countries
• Typical lateral length is 1,320 ft (400 m), or ¼ mile
• The lateral is often about 10 ft above the ground
• Typically, 120 ft pipe span per tower (range: 90 to 250 ft), often with one-
horsepower electric motors (geared down)
• At 120 ft per tower, a 1,320-ft lateral has about 10 towers; with 1-HP motors,
that comes to about 10 HP just for moving the pivot around in a circle
• The cost for a ¼-mile center pivot is typically about $55,000 (about $435/ac
or $1,100/ha), plus about $20,000 (or more) for a corner system
• For a ½-mile lateral, the cost may be about $75,000 (w/o corner system)
• In the state of Nebraska there are said to be 43,000 installed center pivots,
about 15% of which have corner systems
• Center pivots are easily (and commonly) automated, and can have much
lower labor costs than periodic-move sprinkler systems
• Center pivot maintenance costs can be high because it is a large and fairly
complex machine, operating under “field” conditions
• The typical maximum complete rotation is 20 hrs or so, but some (120-acre
pivots) can go around in only about 6 hrs
• IPS 6” lateral pipe is common (about 6-5/8 inches OD); lateral pipe is
generally 6 to 8 inches, but can be up to 10 inches for 2,640-ft laterals
• Long pivot laterals will usually have two different pipe sizes
• Typical lateral inflow rates are 45 - 65 lps (700 to 1,000 gpm)
• At 55 lps with a 6-inch pipe, the entrance velocity is a bit high at 3 m/s
• Typical lateral operating pressures are 140 - 500 kPa (20 to 70 psi)
• The end tower sets the rotation speed; micro switches & cables keep other
towers aligned
• Corner systems are expensive; can operate using buried cable; corner
systems don’t necessarily irrigate the whole corner
• Without a corner system or end gun, π/4 = 79% of the square area is
irrigated
• For a 1,320-ft lateral (without an end gun), the irrigated area is 125.66 acres
• For design purposes, usually ignore soil WHC (WaZ); but, refill root zone at
each irrigation (even if daily)
• Center pivots can operate on very undulating topography
• Some center pivots can be moved from field to field
• Below are some sample center pivot arrangements

Sprinkle & Trickle Irrigation Lectures Page 145 Merkley & Allen
Merkley & Allen Page 146 Sprinkle & Trickle Irrigation Lectures
Sprinkle & Trickle Irrigation Lectures Page 147 Merkley & Allen
 • Some pivots have an end gun that turns on in the corners, in which all other
sprinklers shut off via individual solenoid-actuated valves. The pivot stops in
the corner while the end gun runs for a few minutes.
• Others just slow down in the corners, turning on an end gun, but leaving the
other sprinklers running (at lower discharges)
• Many farmers like extra capacity in the center pivot so they can shut off
during windy times of the day, and still complete the irrigations in time
• Corner systems have angle detectors so that sprinklers in the corner arm
turn on and off individually (or in groups) as the arm swings out and then
back in again
• Center pivots have safety switches to shut the whole thing off if any tower
gets too far out of alignment. Some also have safety switches to shut them
off if the temperatures gets below freezing (ice builds up and gets heavy,
possibly collapsing the structure). Some have safety switches connected to
timers: if a tower has not moved in a specified number of minutes, the
system shuts down. There may also be safety switches associated with the
chemical injection equipment at the lateral inlet location.

• Center pivots on rolling terrain almost always have pressure regulators at

each sprinkler
• Some engineers claim that center pivots can have up to about 90%
application efficiency

Merkley & Allen Page 148 Sprinkle & Trickle Irrigation Lectures
II. System Capacity

• The general center pivot design equation for system capacity is based on Eq.
5.4 from the textbook:

Ad R2d R2Udk f
Qs = K = = (257)
fT k1fT k1TEpa

where,

K is 2.78 for metric units and 453 for English units

k1 is (3,600 s/hr)/π = 1,146 for metric units; 30.6 for English units
kf is the peak period evaporation factor (Table 14.1 in the textbook)
A is area (ha or acre)
d is gross application depth (mm or inch)
f is frequency in days per irrigation
T is operating time (hrs/day)
R is the effective radius (m or ft)
Ud is the peak-use ET rate of the crop (mm/day or inch/day)
Qs is the system capacity (lps or gpm)

• The gross application depth, d, is equal to dn/Epa, where Epa is the design
application efficiency, based on uniformity and percent area (pa) adequately
irrigated
• The operating time, T, is generally 20-22 hrs/day during the peak-use period
• R is the effective radius, based on the wetted area from the center pivot
• The effective radius is about 400 m for many pivots
• R ≈ L + 0.4w, where L is the physical length of the lateral pipe, and w is the
wetted diameter of the end sprinkler
• This assumes that approximately 0.8 of the sprinkler radius beyond the
lateral pipe is effective for crop production
• Note that, for center pivots, Qs is proportional to Ud, and d and f are generally
not used, which is similar to drip irrigation design

III. Gross Application Depth

• If a center pivot is operated such that the water holding capacity of the soil is
essentially ignored, and water is applied frequently enough to satisfy peak-
use crop water requirements, then use dn/f = Ud, and

k f Ud k f Ud
d' = = (258)
Epa DEpaReOe

Sprinkle & Trickle Irrigation Lectures Page 149 Merkley & Allen
 where d' is the gross application depth (mm/day or inches/day); and kf is a
peak-use period evaporation factor, which accounts for increased soil and
foliage evaporation due to high frequency (daily) irrigation

• When LR > 0.1, the LR can be factored into the equation as:

0.9k f Ud
d' = (259)
(1 − LR)DEpaR eOe

which is the same as Eq. 14.1b from the textbook, except that DEpa, Re and
Oe are all as fractions (not percent)

• Values of kf can be selected for the peak period from Table 14.1 of the
textbook for varying values of frequency, f
• Values for non-peak periods can be computed as described in the textbook
on page 314:

(100 − PT ') / PT '

k'f = ( k f − 1) + 1.0 (260)
(100 − PT) / PT

where kf and PT are for the peak-use period (Table 14.1), and k'f and PT' are
the frequency coefficient and transpiration percentage (PT) for the non-peak
period
T
PT = (261)
ET
• PT and PT' can be thought of as the basal crop coefficient (Kcb), or perhaps
Kcb - 0.1 (relative to alfalfa, as per the note in Table 14.1)
• It represents the transpiration of the crop relative to an alfalfa reference

IV. Water Application along the Pivot Lateral

• A major design difficulty with a center pivot is maintaining the application rate
so that it is less than the intake rate of the soil
• This is especially critical near the end of the lateral where application rates
are the highest
• As one moves along the center pivot lateral, the area irrigated by each unit
length of the lateral (each 1 ft or 1 m of length) at distance r from the pivot
point can be calculated as:

a = π(r + 0.5)2 − π(r − 0.5)2 = 2πr (262)

which is equal to the circumference at the radial distance r

Merkley & Allen Page 150 Sprinkle & Trickle Irrigation Lectures
 • The portion of Qs (called q) which is applied to the unit strip at distance r is:

q a 2 πr 2r
= = = (263)
Qs A πR 2 R 2
or,
2rQs
q= (264)
R2
where q can be in units of lps per m, or gpm per ft

• This gives the amount of water which should be discharging from a specific
unit length of lateral at a radial distance r from the pivot point
• The q value at the end of the lateral (r = R) per ft or m is:

2Qs
qend = (265)
R
• Use q to select the nozzle size, where qnozzle = q Se

V. End-Gun Discharge

• This last equation is very similar to Eq. 14.20a, except for the omission of the
Sj term
• Equation 14.20b is for the end gun discharge, assuming that the end gun is
used primarily to compensate for the lack of pattern overlap at the end of the
lateral
• Equation 14.20b can be justified as follows:

• Assuming the “basic” circle discharge, Qb, includes the end gun discharge,
qg, we can write:
Qb qg
≈ (266)
πL2 ∆L ( 2πL' )

Sprinkle & Trickle Irrigation Lectures Page 151 Merkley & Allen
 or, perhaps more precisely,

Qb qg
≈ (267)
πL2 ∆L ( 2π (L '+ ∆L / 2 ) )

but ∆L/2 is generally very small compared to L’, and this is ostensibly
assumed in Eq. 14.20b, so after solving the above for qg you will arrive at Eq.
14.20b:

2L' ∆L
qg ≅ Qb ; for ∆L < 0.03L (268)
L2
VI. Application Rate

• For a center pivot, Se = 1 (based on a unit distance along the lateral) and Sl =
w (wetted width in the tangential direction), so the average application rate
(called AR) at a distance r along the lateral is:

k3 2r QsReOe 2πr k f dReOe 2πr U'd ReOe

AR = 2
= = (269)
R w 60 f T w 60 T w

where AR is the average application rate over width w (mm/min or inch/min);

k3 is 1.61 for English units and 60 for metric units; and f is the time to
complete one revolution (days)

• w is equal to the wetted diameter of the spray or sprinkler nozzles on the

lateral
• U’d is the gross daily irrigation water requirement (mm/day or inch/day) and
includes the effect of kf

k f d k f (Ud − Pe )
U'd = = (270)
f DEpa

• The three forms of the above equation assume a rectangular application

pattern across the width w (that is, the application rate is uniform across w)
• Note that AR is proportional to r and is at a maximum at the end of the lateral
• Note that if w could be equal to 2πr, the application rate would be equal to
the gross application depth divided by the hours of operation per day (just
like a fixed or solid-set sprinkler system) – but this is never the case with a
center pivot machine

Merkley & Allen Page 152 Sprinkle & Trickle Irrigation Lectures
 • At the end of the lateral (r = R), the average application rate can be
calculated as:

2πRU'd ReOe
ARr =R = (271)
60 T w

again, where a rectangular application pattern is assumed

VII. Application Rate with an Elliptical Pattern

• If the application pattern perpendicular to the lateral were elliptical in shape:

4 ⎛ 2k 3r QsReOe ⎞ 4 ⎛ 2 π r U'd R eOe ⎞ r U'd R eOe

AR x = ⎟ = π⎜ ⎟ = 7.5 T w
π ⎜⎝
(272)
2
R w ⎠ ⎝ 60 T w ⎠

where ARx is the maximum application rate (in the center of the pattern) (ARx
is in mm/min for U’d in mm/day)

• In the above equation, k3 is 1.61 for English units, or 60 for metric units
• It is usually a better approximation to assume an elliptical pattern under the
sprinklers than to assume a rectangular pattern, even though both are only
approximations

• For example, if U'd = 9 mm/day (which includes kf), T is 22 hrs/day, w is 30

m, R is 400 m, Re is 0.95 and Oe is 1.0, and the sprinkler application pattern
is elliptical, then the maximum application rate at the far end of the lateral is:

(400)(9)(0.95)(1.0)
AR x = = 0.69 mm / min (273)
(7.5)(22)(30)

• ARx is the peak AR (at the top of the ellipse, or directly beneath the lateral),
so an “average” (ARav) can be calculated, representing the average AR
beneath the wetted area perpendicular to the lateral pipe
• The calculated value of 0.69 mm/min is 41.4 mm/hr, which could be tolerated
only by a very sandy soil

Sprinkle & Trickle Irrigation Lectures Page 153 Merkley & Allen
 • For a rectangular pattern, ARav = ARx
• For an elliptical pattern, ARav = (π/4)ARx
• Therefore, in the example, ARav = (π/4)(0.69) = 0.54 mm/min
• If d were 10 mm, it would take tt = 10/0.54 = 18 minutes to apply the water at
the rate ARav. (may want to use d Re Oe instead of just d in such a
calculation)
• Re can be taken from Fig. 6.8 or from examples in Table 14.3
• Guidelines for determining CI are given in Table 14.4
• The center pivot speed (at the end of the lateral) is w/tt, where tt is the time of
wetting
• In the preceding example, w is 30 m and tt is 18 min
• Therefore, the speed should be about 30 m/18 min = 1.7 m/min at the end
• Note that with spray booms, w is larger, and AR is smaller for the same q
value

Merkley & Allen Page 154 Sprinkle & Trickle Irrigation Lectures
Lecture 13
Center Pivot Nozzling

I. Center Pivot Nozzling

• The wetted width of the application package can be reduced closer to the
pivot point because the towers are moving at a slower speed at inner points;
therefore, the application intensity (AR) is less (qr ∝ r)
• Generally, if spray booms are required near the end of the center pivot, spray
drops can be used toward the center, and the spray drops nearest the pivot
point will produce something like a fine mist
• At the far end of the lateral the application may be more like a torrential rain
• Generally, impact and spray sprinklers would not be mixed on a center pivot
because the pressure requirements are substantially different
• The minimum wetted width at any radius r along the pivot (for an elliptical
pattern) can be calculated as:

8r Ud
wr = (274)
60 T AR x DEpa

where ARx is the maximum permissible application rate (mm/min) according

to limits imposed by the soil, slope, and vegetative cover; Ud is in mm/day; T
is in hrs/day; and DEpa is expressed as a fraction

• Note that wr approaches zero for r → 0

• A suitable application device can then be selected for radius r such that the
wetted diameter of the device, wd, is greater or equal to wr (wd ≥ wr)
• The actual application rate (mm/min) at radius r at the ground using a device
with wetted diameter wd should be:

r U'dReOe
ARr ≤ (275)
7.5 T w d

• The term “ReOe” is included in the above equation to account for evaporation
and wind drift losses, and pipe leakage
• Note that 60/8 = 7.5 and that we are using f = 1 day
• Divide by Re in the above equation to obtain AR at the nozzle
• The wetting time at any radius r (assuming an elliptical pattern) along the
lateral is:
4D f
( t t )r = (276)
π ARr
where Df is the total cumulative application (d Re Oe)
Sprinkle & Trickle Irrigation Lectures Page 155 Merkley & Allen
II. Center Pivot Nozzling Example

Given:

ƒ Ud = 8 mm/day
ƒ Pe = 0 mm/day
ƒ T = 22 hrs/day
ƒ Qs = 73.6 lps
ƒ R = 400 m
ƒ Speed = 21.6 hrs/revolution
ƒ ARx = 2.3 mm/min (allowable)
ƒ kf = 1.02
ƒ DEpa = 0.74
ƒ Re = 0.94
ƒ Oe = 0.99

Calculations:

• Calculate Ud’:

k f (Ud − Pe ) 1.02 ( 8 − 0 )
U'd = = = 11.0 mm/day (277)
DEpa 0.74

• Tangential speed of the pivot at distance “r” along the lateral:

2πr
Sr = (278)
t

where Sr is the speed in m/min; t is the minutes per full-circle revolution; and
r is the radius from the pivot point in m

• The wetting time is, then:

wr
τ= (279)
Sr

where τ is wetting time in minutes; and wr is the minimum wetted width in m.

• The values of wd can be selected from available boom lengths, which in this
case is 6, 8, 10, and 12 m. For less than 4-m width, no boom is required.
Select wd values such that ARx is not exceeded.

Merkley & Allen Page 156 Sprinkle & Trickle Irrigation Lectures
 • Spreadsheet calculations:

r q wr wd ARr tau Qr
(m) (lps/m) (m) (m) (mm/min) (s) (lps)
0 0.0000 73.6
40 0.0368 1.14 3 0.829 928 72.9
80 0.0736 2.28 3 1.658 464 70.7
120 0.1104 3.42 6 1.244 619 67.0
160 0.1472 4.56 6 1.658 464 61.8
200 0.1840 5.70 6 2.073 371 55.2
240 0.2208 6.84 8 1.866 413 47.1
280 0.2576 7.98 8 2.177 354 37.5
320 0.2944 9.12 10 1.990 387 26.5
360 0.3312 10.26 10 2.239 344 14.0
400 0.3680 11.39 12 2.073 371 0.0

• Notes about the variables:

ƒ “q” is the sprinkler discharge per unit length of lateral at r

ƒ “wr” is the minimum wetted width at r
ƒ “wd” is the actual (selected) wetted width at r
ƒ “ARr” is the actual application rate at r
ƒ “wetting” is the wetting time at a given location
ƒ Qr is the flow rate in the lateral pipe at a given location

• Most manufacturers prefer to specify the actual nozzle sizing and spacing
along center pivots at the factory (rather than have the buyer specify these)
for reasons of liability (they have specialized computer programs which
attempt to maximize uniformity)
• Therefore, the designer will generally only specify the flow rate, pressure,
and type of nozzle (spray drop, booms, impacts, etc.), and the manufacturer
will specify individual nozzle sizes
• The following figure shows a center pivot with booms (the booms are greatly
exaggerated in width to show the concept)

Sprinkle & Trickle Irrigation Lectures Page 157 Merkley & Allen
III. Sprinkler/Nozzle Configurations

• Because the required discharge per unit of lateral is linearly proportional to

the distance along the lateral, discharge rates of sprinklers (or the sprinkler
spacing) need to be varied to match this linear variation in flow rate
• Various configurations of locating sprinklers along a center pivot lateral are
used

1. Uniform Sprinkler Spacing

• qa of nozzles increases with r

• Large heads are required near the end of the lateral
• A large w (due to large heads) results near the end which is good, but large
drop sizes may cause soil surface sealing and infiltration reduction
• The next figure shows a center pivot lateral with uniform sprinkler spacing
and increasing w values toward the downstream end of the lateral

Merkley & Allen Page 158 Sprinkle & Trickle Irrigation Lectures
 2. Uniform Sprinkler Size where the distance between sprinklers decreases
with r

• Sprinkler density is higher near the downstream end of the lateral

• May allow the use of small nozzles with lower pressures and less soil
sealing
• The value of w is smaller, so the soil must support a higher application rate
• See the following figure

3. Combination of 1 and 2

• This is the most common configuration because it

is easier to match the required q/m of the lateral
• Smaller drop size
• Intermediate wetted width, w
• May have uniform spacing of outlets on lateral spans, but unused outlets
are plugged (w/o any sprinkler), so pipes in each span are identical and
interchangeable

Sprinkle & Trickle Irrigation Lectures Page 159 Merkley & Allen
 Sizing Individual Nozzles:

1. Determine q for each spacing along the lateral (see Eq. 14.20a):

2Qs
qr = r Sr (280)
R2
where R is the maximum effective radius of the center pivot (approximately
equal to L + 0.4w); Sr is the sprinkler spacing at a distance r from the pivot
point; and r, Sr and R have the same units (m or ft)
2. Beginning at the design pressure at the end of the lateral, L (where q is
known), determine Pr:

Pr = (Pa )end + ( hfr )end + ( ∆He )end−r

(281)

where (hfr)end is the friction loss from point r to the far (downstream) end of
the lateral. Note that (∆He)end-r often averages out to zero as the pivot makes
its way around the circle, if the field slope is uniform (see the next figure).
Merkley & Allen Page 160 Sprinkle & Trickle Irrigation Lectures
 3. Select the best nozzle size to provide qr at a pressure of Pr
4. Return to Step 1 and repeat for the next r-Sr location
5. The required pressure at the pivot point is Pr = 0

no slope

uniform slope

centered on a hill

centered on a depression
IV. Trajectory Angles of Impact Sprinklers

• For center pivots, sprinklers with 6° to 18° trajectory angles (low angle) are
preferred because drift losses are minimized (see Table 14.3 in the textbook)
• Other things being the same, wind drift and evaporation losses can be higher
with center pivots than with other types of sprinkler systems because of the
relative height of the sprinklers above the ground
• But, you can use drop-down sprayers on a “goose-neck” pipe – some of
these may be only a few centimeters from the mature crop canopy

V. End Guns

• The discharge for an end gun can be computed as (see Eq. 14.21):
Sprinkle & Trickle Irrigation Lectures Page 161 Merkley & Allen
 ⎛ 2Sr Q ⎛ Sr ⎞⎞
⎜ 2 ⎜ L + ⎟⎟
R ⎝ ⎠⎠
Qg = ⎝
2
(282)
0.93
where L is the lateral length; R is the effective length of the pivot (R = L + Sr);
Qg is the end gun discharge; Q is the total center pivot flow rate (includes
Qg); Sr is R - L, which equals the effective wetted radius (or 75% of the gun
radius)

• The above equation is essentially Eq. 14.20a with r = L + Sr/2

• The 0.93 factor (taken as Oe) is to account for ineffective discharge beyond
75% of the gun's wetted width, w (see page 349 of the textbook)
• Also, Qg can be approximated as:

⎛ L2 ⎞
Qg ≈ Q ⎜ 1 − 2 ⎟ (283)
⎜ R ⎟
⎝ ⎠

where Q includes Qg and R = L + Sr

• Note that you can derive this last equation by substituting R - L for Sr in the
previous equation as follows:

2 (R − L ) Q [L + 0.5(R − L)]
=
0.93R2
2QL (R − L ) + Q (R − L )
2
(284)
2
0.93 R
Q ⎛ L2 ⎞
= ⎜ 1 − ⎟
0.93 ⎜⎝ R2 ⎟⎠

where the only difference is the 0.93 in the denominator

• A part circle rotation (typically about 150°) is generally used to achieve best
uniformity under the area covered by the gun sprinkler, which is beyond the
end of the lateral pipe
• If the rotation of the end gun covered 180° or more, it might make it too
muddy for the wheels of the end tower – so with 150° (or so) the path in front
of the end tower stays relatively dry

Merkley & Allen Page 162 Sprinkle & Trickle Irrigation Lectures
Booster Pump

• If an end gun is used primarily on corners, and if

Qg > 0.2Q, then a booster pump in parallel or
series may be necessary
• Without a booster, when the gun is turned on (at
the corner), the pressure along the lateral will
drop, and individual sprinkler flow rates will be
approximately q[Q/(Q+Qg)]
• An alternative to a booster pump is to
automatically decrease the center pivot speed at the corners: Scorner =
S[Q/(Q+Qg)], where S is the speed in all places except the corners. That is,
turn the gun on in the corners without any booster pump. This can help
provide for equivalent, more nearly uniform application in all parts of the field
(friction losses will be greater when the gun is operating; therefore, uniformity
will not be perfect)
• If the end gun requires higher pressure than the nozzles along the center
pivot lateral, then an electric-motor-driven booster pump may be mounted on
the last tower, upstream of the end gun. This pump will increase the
pressure to the end gun only. All other nozzles will be operated at lower
pressure.

Center Pivot Hydraulic Analysis

I. Center Pivot Hydraulics

• The total discharge in the lateral pipe (not the flow rate from sprinklers at r) at
any point r is approximately:

⎛ r2 ⎞
Qr = Qs ⎜ 1 − 2 ⎟ (285)
⎜ R ⎟
⎝ ⎠

where Qs is total system capacity (possibly including an end gun); r is

measured radially from the pivot point; and R includes Sr

Friction Loss

(1) For a center pivot without an end gun:

1.852
⎛Q⎞
hf = khFpL ⎜ ⎟ D−4.87 (286)
⎝C⎠

Sprinkle & Trickle Irrigation Lectures Page 163 Merkley & Allen
 where kh is 10.50 for hf and L in ft, Q in gpm, and D in inches; kh is
16.42(10)4 for hf and L in m, Q in lps, and D in cm

• Fp is the multiple outlet friction factor for a center pivot (see Fig. 14.12)
• Fp = 0.555 for a center pivot with a “large” number of outlets and no end gun
when using the Hazen-Williams equation.
• Other sources suggest using Fp = 0.543
• The value of C is about 130 for galvanized steel, or 145 for epoxy-coated
steel

(2) For a center pivot with an end gun:

Compute friction loss as though the center pivot were R m long rather than L,
and then subtract the non-existent friction past the point L, where R is the
effective (wetted) radius and L is the physical length of the lateral pipe.

A traditional way to consider the effects of an end gun on friction loss is:

1.852 1.852
⎛Q⎞ −4.87 ⎛ Qg ⎞
hf = khFpR ⎜ ⎟ D − khFg (R − L) ⎜ ⎟ D−4.87
⎝C⎠ ⎝ C ⎠ (287)

where Q is the total flow rate of the pivot plus the end gun; and Qg is the flow
rate of the end gun

• Fg should represent Fp for the distance R - L

• Base this on the bottom of page 355 of the textbook, where the number of
outlets is (R - L)/Se, where Se is the sprinkler spacing on the last
(downstream) span
• This Fg value will be conservative (it will underestimate the imaginary friction
loss, therefore, will overestimate the total friction in the pivot lateral), because
Fg will be greater than that stated
• Note that, near the end gun, the change in Q with distance is small; therefore,
the value of Fg will usually be near unity

• A different method for calculating friction loss is suggested in the textbook

(Eq. 14.26a)
• This may be a better method than that given above

Dual Pipe Sizes

• A center pivot may be assembled with dual pipe sizes (8- and 6-inch pipe, or
8- and 6 5/8-inch, for example)

Merkley & Allen Page 164 Sprinkle & Trickle Irrigation Lectures
 • One requirement to note is that the tower spacing must be closer for the 8-
inch pipe due to the added weight of water in the 8-inch pipe, and traction
problems from towers which are too heavy and sink into the soil
• Therefore, balance the savings in hf with added cost for the system
• Tower spacing is often 100 ft (30 m) for 8-inch pipe, and 150 ft (45 m) for 6-
inch pipe
• Weight per tower = Wt of tower + Wt of 1 span (steel) + Wt of water in the
span

Pressure Balance Equation

Hl = Ha + hf + ∆He + Hr + Hmin or (288)

where Hl is the pressure head required at ground level, at the pivot point; Ha is
the pressure head requirement of the last nozzle (or end gun); hf is the total
friction loss along the pivot lateral; ∆He is the elevation increase between the
pivot point and lateral end; Hr is the height of the lateral pipe less the vertical
length of any drop tubes; and Hminor is the sum of all minor losses along the
lateral

• Ha is, then, the pressure head requirement of the last nozzle

• If end guns require higher pressure than the nozzles along the lateral, then
an electric booster pump can be installed at the last tower on the pivot
• See Chapter 14 of the textbook for the hydraulic effect of turning end guns or
corner systems on and off, and for additional elevation effects

II. An Alternative Method for Determining Hf and Hl

• An alternative method to compute the exact value of Hl is to start at the pivot

end and progress toward the pivot, adding friction losses
• This is a "stepwise" computational procedure, and is generally more accurate
than using the JFpL equations
• Friction loss, hf, in pipe segments between nozzles on the pivot can be
computed using the Hazen-Williams or Darcy-Weisbach equations with Fp =1
• In this case, include ∆He if Hend is the minimum pressure head at the highest
elevation position of the lateral

n
( )
Hl = Hend + ∑ ⎡ hfi + ∆Hei + Hr + Hmin or ⎤
⎣ ⎦
i=0 (289)

where Hend is the desired nozzle pressure head at the pivot end; i is the outlet
number along the lateral (i = 0 at the end, and i = n at the pivot point); n is
the number of outlets (sprinklers) on the lateral; ∆Hei is the elevation
Sprinkle & Trickle Irrigation Lectures Page 165 Merkley & Allen
 difference between two adjacent points (i and i+1) along the lateral; and hfi is
the friction loss between outlet i and i+1, where i+1 is upstream of i. This last
term is defined as:

Ji ∆Li
hfi = (290)
100

where,
1.852
12 ⎛ Qi
⎞
Ji = 1.21(10) ⎜ ⎟ Di−4.87 (291)
⎝C⎠

with Qi in lps and Di in mm

and,
i
Qi = ∑ q j (292)
j= 0

in which j is the outlet number (j = 0 at the downstream end of the lateral);

and qj is a function of hj and Kd

• Hminor includes short hose connections between pipe segments (at towers)
• Therefore, actual computed hj values should be used with the selected
nominal nozzle size (or FCN size), where hj is the pressure head at outlet j
• The desired qj is:

⎛ 2Q ⎞ ⎛ 2Q ⎞
q j = rjSrj ⎜ 2 ⎟ = rj∆Li ⎜ 2 ⎟ (293)
⎝R ⎠ ⎝R ⎠

where ∆Li is the distance between outlet i and outlet i+1.

• ∆Li is constant for constant spacing, variable nozzle size

• ∆Li is variable for variable spacing, constant nozzle size

Point of “Average” Elevation

The point of “average” elevation along a lateral may be determined by weighting

elevations according to areas as (Allen 1991):

L 2π r (He ) r,α
Hew = ∫ ∫ 2πL2
dr dα (294)
r =0 α
Merkley & Allen Page 166 Sprinkle & Trickle Irrigation Lectures
 where r is a weighting term; and (He)r,α is the elevation at radius r and pivot
rotation angle α.

• Use α = 0 for the whole irrigated area

• Note that on uniform slopes, the weighted elevation, Hew, equals (He)pivot
• Note also that you probably won’t have data for elevations in polar
coordinates, so this equation may be rather “academic”
• See the textbook for additional equations which consider elevation effects

Sprinkle & Trickle Irrigation Lectures Page 167 Merkley & Allen
Merkley & Allen Page 168 Sprinkle & Trickle Irrigation Lectures
Lecture 14
Center Pivot Uniformity Evaluation

I. Introduction

• The calculation of an application uniformity term must take into account the
irrigated area represented by each catch container
• It is more important to have better application uniformity further from the pivot
point than nearer, because the catch containers at larger distances represent
larger irrigated areas
• If the catch containers are equally spaced in the radial direction, the area
represented by each is directly proportional to the radial distance

II. Equation for Center Pivot CU

• The equation for CU proposed by Heermann and Hein is (ASAE/ANSI S436):

⎛ ⎛ ⎞⎞
∑ i=1( diri ) ⎟⎟ ⎟⎟
n
⎜ n ⎜
⎜ ∑ r d −
i=1⎜ i i
∑i=1ri ⎟⎠ ⎟⎟
n
CU = 100 ⎜ ⎜ (295)
⎜ 1.0 − ⎝
⎜ ⎟
∑ i=1( diri )
n
⎜ ⎟
⎝ ⎠

where CU is the coefficient of uniformity; di is the depth from an individual

container; ri is the radial distance from the pivot point; and n is the number of
containers

• First calculate the summations:

∑i=1ri ∑i=1( dri i )

n n
and, (296)

• Then, perform the outer summation to determine the CU value

• That is, don’t recalculate the inner summation values for every iteration of the
outer summation – it isn’t necessary
• It is usually considered that a center pivot CU should be greater than 85%

• If the radial distances, ri, are equal, the sequence number of the can
(increasing with increasing radius) can be used instead of the actual distance
for the purpose of calculating application uniformity
• Consider the following two figures:

Sprinkle & Trickle Irrigation Lectures Page 169 Merkley & Allen
 slope?

catch containers

leg # #1
3 g
le
slope?

leg #
2

catch containers

level field?
no wind?...
CU = 100%

Merkley & Allen Page 170 Sprinkle & Trickle Irrigation Lectures
III. Standard Uniformity Values

• You can also calculate the “standard” CU or DU if you weight each catch
value by multiplying it by the corresponding radial distance
• To obtain the low ¼, rank the unweighted catches, then start summing radii
(beginning with the radius for the lowest catch value) until the cumulative
value is approximately equal to ¼ of the total cumulative radius
• This may or may not be equal to ¼ of the total catch values, because each
catch represents a different annular area of the field
• Finally, divide the sum of the catches times the radii for this approximately ¼
area by the cumulative radius
• This gives the average catch of the low ¼
• Don’t rank the weighted catches (depth x radius) because you will mostly get
the values from the low r values (unless the inner catches are relatively high
for some reason), and your answer will be wrong
• Don’t calculate the average of the low ¼ like this…
(because the lowest ¼ of the catches generally represents
something different than ¼ of the irrigated area):
• Actually, the equation at the right is all right, except for the
value “n/4”, which is probably the wrong number of ranked
values to use in representing the low ¼
• You can set up a table like this in a spreadsheet application:

Ranked Center Pivot Catches

Radius, r Cumulative r Depth, d d*r Cumulative d*r
smallest

largest
Totals: ---- ----

• Note that when you rank the depths, the radius values should stay with the
same depth values (so that the radius values will now be “unranked”; all
mixed up)
• To get the average weighted depth for the whole pivot area, divide the total
“Cumulative d*r” by the total “Cumulative r” (column 5 divided by column 2)
• Find the row corresponding closest to ¼ of the total “Cumulative r” value, and
take the same ratio as before to get the weighted average of the low ¼ area
• Look at the example data analysis below:

Sprinkle & Trickle Irrigation Lectures Page 171 Merkley & Allen
 Ranked catches
Radius, r Cum. r Depth, d d*r Cum. d*r
120 120 0.52 62.6 62.6
900 1,020 0.95 851.9 915
160 1,180 1.29 205.8 1,120
340 1,520 1.31 445.6 1,566
1000 2,520 1.46 1,456.3 3,022
1040 3,560 1.46 1,514.6 4,537
240 3,800 1.48 355.3 4,892
800 4,600 1.50 1,203.9 6,096
860 5,460 1.53 1,315.0 7,411
480 5,940 1.58 757.3 8,168
1280 7,220 1.58 2,019.4 10,188
980 8,200 1.60 1,569.9 11,758
540 8,740 1.63 878.2 12,636
360 9,100 1.65 594.2 13,230
460 9,560 1.67 770.4 14,000
880 10,440 1.70 1,495.1 15,496
320 10,760 1.72 551.5 16,047 1/4 area (11,055)
1140 11,900 1.75 1,992.2 18,039
1160 13,060 1.75 2,027.2 20,067
280 13,340 1.82 509.7 20,576
720 14,060 1.82 1,310.7 21,887
1300 15,360 1.82 2,366.5 24,253
200 15,560 1.84 368.9 24,622
420 15,980 1.84 774.8 25,397
440 16,420 1.89 833.0 26,230
1020 17,440 1.89 1,931.1 28,161
1200 18,640 1.92 2,301.0 30,462
600 19,240 1.94 1,165.0 31,627
640 19,880 1.94 1,242.7 32,870
1060 20,940 1.94 2,058.3 34,928
1100 22,040 1.94 2,135.9 37,064
220 22,260 1.97 432.5 37,497 1/2 area (22,110)
1080 23,340 1.97 2,123.3 39,620
380 23,720 1.99 756.3 40,376
740 24,460 1.99 1,472.8 41,849
920 25,380 1.99 1,831.1 43,680
1220 26,600 1.99 2,428.2 46,108
300 26,900 2.01 604.4 46,713
180 27,080 2.03 364.8 47,077
820 27,900 2.04 1,671.8 48,749
1260 29,160 2.04 2,568.9 51,318
660 29,820 2.06 1,361.7 52,680
1180 31,000 2.06 2,434.5 55,114
680 31,680 2.09 1,419.4 56,534
940 32,620 2.11 1,985.0 58,519
560 33,180 2.14 1,196.1 59,715
260 33,440 2.16 561.7 60,276
1120 34,560 2.18 2,446.6 62,723
700 35,260 2.23 1,563.1 64,286
760 36,020 2.23 1,697.1 65,983
100 36,120 2.25 224.5 66,208
960 37,080 2.26 2,167.0 68,375
520 37,600 2.28 1,186.4 69,561
620 38,220 2.28 1,414.6 70,976
1240 39,460 2.28 2,829.1 73,805
500 39,960 2.33 1,165.0 74,970
140 40,100 2.35 329.6 75,300
400 40,500 2.40 961.2 76,261
780 41,280 2.52 1,968.9 78,230
40 41,320 2.57 102.9 78,333
80 41,400 2.57 205.8 78,538
840 42,240 2.79 2,344.7 80,883
580 42,820 2.82 1,633.0 82,516
60 42,880 3.23 193.7 82,710
1320 44,200 3.79 4,998.1 87,708
20 44,220 3.83 76.7 87,784

Merkley & Allen Page 172 Sprinkle & Trickle Irrigation Lectures
 • Notice that the depth values (3rd column) are ranked from low to high
• Notice that the maximum value of cumulative r is 44,220 & maximum
cumulative d*r is 87,784. Then, the weighted average depth for the entire
center pivot is equal to 87,784/44,220 = 1.985 (whatever units)
• One quarter of 44,220 is equal to 11,055 which corresponds most closely to
the row in the table with depth = 1.72. For the same row, divide the two
cumulative columns (Col 5/Col 2) to get 16,047/10,760 = 1.491, which is
approximately the average of the low ¼.
• Finally, estimate the distribution uniformity for this data set as:

⎛ 1.491 ⎞
DU ≅ 100 ⎜ ⎟ ≅ 75% (297)
⎝ 1.985 ⎠

• Note that in this example, the average of the low ¼ was, in fact, based on
approximately the first n/4 ranked values

• Consider the weighed catch-can data plotted below:

14000

12000
Weighted Catch Values

10000

8000

6000

4000

2000

0
0 10 20 30 40 50 60 70

Distance from Pivot Point

• As in any application uniformity evaluation, there is no “right” answer. The

results are useful in a comparative sense with evaluations of other center
pivots and other on-farm irrigation systems.
• However, a plot of the catches can give indications of localized problems
along the center pivot radius

Sprinkle & Trickle Irrigation Lectures Page 173 Merkley & Allen
IV. The Field Work

• It may take a long time for the full catch in containers near the pivot point,
and because these represent relatively small areas compared to the total
irrigated area, it is usually acceptable to ignore the inside 10% or 20% of the
radius
• The pivot quickly passes the outer cans, but takes longer to completely pass
the inner cans, so you can collect the data from the outer cans sooner
• The pivot should not be moving so fast that the application depth is less than
about 15 mm
• Catch containers can be placed beyond the physical length of the lateral
pipe, but if they are so far out that the catches are very low, these can be
omitted from the uniformity calculations
• Catch containers should be spaced in the radial direction no further than
about 30% of the average wetted diameter of the sprinklers
• There is often an access road leading to the pivot point for inspection,
manual operation, maintenance, and other reasons
• If the crop is dense and fairly tall (e.g. wheat or maize) it will be difficult to
perform the evaluation unless the cans are placed on the access road
• Otherwise, you can wait until the crop is harvested, or do the test when the
crop is still small
• Some people recommend two radial rows of catch cans, or even two parallel
rows, to help smooth out the effects of the non continuous movement of
towers (they start and stop frequently to keep the pivot lateral in alignment)
• Some have used troughs instead of catch cans to help ameliorate this
problem.
• Note that if the field is sloping or undulating, the results from one radial row
of catch cans may be quite different from those of a row on another part of
the irrigated circle
• See Merriam and Keller (1978)

Merkley & Allen Page 174 Sprinkle & Trickle Irrigation Lectures
Linear Move Systems

I. Introduction

• Mechanically, a linear move system is essentially the same as a center pivot

lateral, but it moves sideways along a rectangular field, perpendicular to the
alignment of the lateral pipe
• The variation of flow rate in a linear move lateral is directly proportional to
distance along the lateral pipe, whereas with center pivots it is proportional to
a function of the square of the distance from the pivot point
• A center or end tower sets the forward speed of the machine, and the other
towers just move to keep in line with the guide tower (this is like the far end
tower on a center pivot)
• Usually, each tower is independently guided by cables and micro-switches as
for a center pivot – this keeps the lateral pipe in a straight line (aligned with
itself)
• Alignment with the field is usually not mechanically “enforced”, but it is
“monitored” through switches in contact with a straight cable along the center
of the field, or along one end of the field
• The center tower has two "fingers", one on each side of the cable, usually
slightly offset in the direction of travel (they aren’t side by side). The fingers
should be in constant contact with the cable – if one is lifted too far a switch
will be tripped, shutting the system down (because the whole lateral is
probably getting out of alignment with the field)
• If the cable is broken for any reason, this should also shut the system down
because the fingers will lose physical contact
• If the lateral gets out of alignment with the field and shuts off, it will be
necessary to back up one side and or move the other side forward until it is
in the correct position
• This can involve electrical “jumps” between contacts in the control box, but
some manufacturers and some installers put manual switches in just for this
purpose
• Some linear moves are fitted with spray nozzles on drop tubes or booms
• If they are spaced closely along the lateral, it may be necessary to put booms
out beyond the wheels at tower locations, either in back of the lateral or on
both sides of the lateral

Water Supply

• Water is usually supplied to the lateral via:

1. a concrete-lined trapezoidal-sectioned ditch, or

2. a flexible hose (often 150 m in length), or
3. automatic hydrant coupling devices with buried mainline

Sprinkle & Trickle Irrigation Lectures Page 175 Merkley & Allen
 • Hose-fed systems require periodic manual reconnection to hydrants on a
mainline – it is kind of like a period-move system, and you have to ask
yourself whether the linear move machine is worth the cost in this case!
• With the automatic hydrant coupling machines (see Fig. 15.3) there are two
arms with pipes and an elbow joint that bends as the linear move travels
down the field. The two arms alternate in connecting to hydrants so as not to
disrupt the irrigation nor the forward movement of the machine. These are
mechanically complex.
• The advantage of hose-fed and automatic coupling linear moves is that you
don’t need to have a small, uniform slope in the direction of travel, because
water is supplied from a pressurized mainline instead of an open channel
• On ditch-fed systems there can be a structure at the end of the field that a
switch on the linear move contacts, shutting down the pump and reversing
the direction of movement so that it automatically returns to the starting end
of the field.
• The advantages and disadvantages of the ditch-feed system are:

Pros
• Low pressure (energy) requirement
• Totally automated system
• More frequent irrigations than hose-fed, since no one needs to be
available to move the hose

Cons
• Trash and seeds and sediment pass through screen and may plug
nozzles
• The pump must be on (move with) the lateral, causing extra weight
• Should have uniform slope along the lateral route

Pros and Cons Compared to a Center Pivot

Pros
• Easy irrigation of a rectangular field (important if land is expensive, but
not important if land is cheap and water is scarce)
• Application rate is uniform over length of lateral, rather than twice the
average value at the end of the center pivot
• No end gun is required

Cons
• The lateral does not end up right back at the starting point immediately
after having traversed the irrigated area – you either have to
“deadhead” back or irrigate in both directions
• May be more expensive than a pivot due to extra controls, pump on
ditch feed, or more friction loss in the flexible feed hose (the hose is
fairly expensive)
Merkley & Allen Page 176 Sprinkle & Trickle Irrigation Lectures
II. System Costs

Relative costs for linear move systems are:

• $50,000 for a 1,280-ft hose-fed machine (perhaps 160 acres)

• $55,000 for a winch-tow hose fed 1,280-ft machine (perhaps 160 acres)
• $55,000 for a center pivot (1,280 ft) with a corner system (about 150 acres)
• $140,000 for a ½-mile (2,560-ft) linear with automated hydrant coupling
system (no ditch or hose required). The mainline is down the middle (a ¼-
mile lateral on each side). Perhaps 320 acres or more irrigated.

III. System Design

• A main strategy in linear move design is to minimize the cost per unit area.
This is done by maximizing the area covered per lateral (length of field)
• Generally, the lateral length is limited to 400 to 800 m. Therefore, the major
difficulties and objectives in linear design are to:

1. Maximize the irrigated area per lateral (this minimizes $/area). In other
words, how large a field can be irrigated by one machine?
2. Prevent runoff by matching ARx with Isoil + SS/ti (this tends to limit the field
length, because if AR is small, it won’t be possible to finish in f’ days),
where SS is the allowable surface storage in (mm or inches); and, ti is the
time of irrigation
3. Determine whether spray nozzles can be used without causing runoff
4. Minimize labor (for moving hoses and supervising)

as long as possible
• The allowable surface storage, SS, is the maximum amount of ponding
without incurring surface runoff
• SS is a function of the general topography and the microtopography, and of
the amount of foliar interception (water can “pond” on the crop leaves too)

Sprinkle & Trickle Irrigation Lectures Page 177 Merkley & Allen
 • SS is usually less than about 5 mm unless small basins are created along
furrows, for example
• ARx limits the field length because it corresponds to some minimum time to
finish an irrigation, for a given gross application depth, whereby a maximum
interval (f) is calculated in the preliminary design steps

Lateral Inlet Head

• This is the same as for periodic-move systems

• The pressure balance equation for linear move systems is similar to set
systems (both are linear, with uniform discharge from each outlet).

Pa 3 1
Hl = + hf + Hr + ∆He + (hf )minor + (hf )hose (298)
γ 4 2

• Or, if using flow control nozzles, with a minimum pressure required at the end
(assuming the minimum pressure occurs at the end):

Pend
Hl = + hf + Hr + ∆He + (hf )minor + (hf )hose (299)
γ

where Hr is the height of the lateral or spray boom above the ground; and,
(hf)minor are the hydrant coupler and tower connection losses.

• The parameter (hf)hose is the loss in the flexible hose connection on a hose-
fed system
• Note that (hf)hose may be a major loss, since the hose diameter is usually less
than 5” or 6".
1.852
⎛Q⎞
hf = khFL ⎜ ⎟ D−4.87 (300)
⎝C⎠
where kh = 10.50 for hf and L in ft, Q in gpm, and D in inches; Kh = 1.21(10)10
for hf and L in m, Q in lps, and D in mm. F is the multiple outlet friction factor
for a linear move (F ≈ 0.36).

• For hose-fed systems, the maximum hose length for dragging the hose is
220 ft. Therefore, there could be about 400 ft between hydrants.
• For hose-fed systems with a cable/winch system for assisting in dragging the
hose (towers only have a moderate amount of tractive power), the maximum
hose length is 330 feet (640 feet between hose hydrants).
• Flexible hoses normally come in 5-inch ($18/ft) and 6-inch ($25/ft) diameters
• The Hazen-Williams C value for the hose can usually be taken as 150

Merkley & Allen Page 178 Sprinkle & Trickle Irrigation Lectures
Lecture 15
Maximizing Linear Move Field Length

I. The Procedure

• The following procedure for maximizing field length is from Allen, 1983, Univ.
Idaho and Allen, 1990 (Irrig. Symp. Paper), and is used in the USUPIVOT
computer program
• The basic strategy is to examine different application depths and different w
values to maximize the area covered by the sprinkler system, and or to
minimize labor requirements

1. Calculate the maximum application depth per irrigation (dx ≤ MAD*Z*Wa).

Note that the maximum application depth may be less than MAD*Z*Wa
with an automatic system to maintain optimal soil water conditions and to
keep soil water content high in case of equipment failure (i.e. don’t need
to take full advantage of TAW)

f' = dx/Ud (round down to even part of day)

2. Calculate net and gross application depths:

dn = f (Ud)

d = dn/Epa

3. Calculate the (presumed) infiltrated depth per irrigation:

(Df)max = d.Re

where (Df)max is the maximum depth to be evaluated, and assuming no

runoff

4. For a series of 10 or so infiltration depths, df, beginning with df equal to

some fraction (say 1/10) of (Df)max:

df = (i/10)(Df)max where i = 1 to 10
and,
f' = df DEpa/(100 Ud)

f = f' - days off (days off may be zero because the system is automatic),
where f' = irrigation frequency for depth df. DEpa is used here (in percent)
because Ud is net, not gross

Sprinkle & Trickle Irrigation Lectures Page 179 Merkley & Allen
 5. Determine the maximum ARx for a particular df value using the following
two equations (assuming an elliptical pattern):

⎛ ( SF )( AR x ) ⎞ ⎛ n+1 n ⎞
1 n
⎜1− ⎜
⎟ k (n + 1)n+1 (D − SS − c )n+1 ⎟
⎝ k ⎠⎝⎜ ⎟
AR x = ⎠ (301)
2 2
⎛π⎞ ⎛ D ⎞
1.05 − 1.6 ⎜ ⎟ ⎜ − 0.5 ⎟
⎝ 2 ⎠ ⎝ df ⎠

where,

−n−1
⎛⎛ −0.5 ⎞
2 2⎞ ⎛ 2 ⎞
⎜⎜ ⎛ π ⎞ ⎛ D ⎞ ⎛ π ⎞ ⎛ D ⎞ ⎟
2 2 ⎟ ⎜ −1.6 AR2x ⎜ ⎟ ⎜ − 0.5 ⎟ ⎟ ⎟
⎜ ⎜ 1.05 AR x − 1.6 AR x ⎜⎝ 2 ⎟⎠ ⎜ d − 0.5 ⎟ ⎟ ⎜
⎜⎝ ⎝ f ⎠ ⎠ ⎝ ⎝ 2 ⎠ ⎝ df ⎠ ⎟⎠ ⎟
D=⎜ ⎟
⎛ ( SF )( AR x ) ⎞ ⎛ −1 ⎞ ⎛ 1 ⎞
⎜ ⎟
⎜
df n ⎜ 1 − ⎟ ⎜⎜ ( n + 1) n+1 ⎟⎟ ⎜ k n+1 ⎟ ⎟
⎜ ⎝ k ⎠⎝ ⎠ ⎝⎜ ⎟
⎠ ⎟
⎝ ⎠
+SS + c
(302)

and ARx is the peak application per pass (mm/min); D is the applied depth
at time t = ∫ (AR) dt (mm); SS is the allowable surface storage (after
ponding) before runoff occurs (usually less than about 5 mm); c is the
instantaneous soil infiltration depth, from SCS soil intake families (mm); k
is the coefficient in the Kostiakov-Lewis equation; and df is the total depth
of water applied to the ground surface (mm)

• The parameter “n” is defined as: n = a -1, where “a” is the Kostiakov
exponent (see NRCS soil curves at www.wcc.nrcs.usda.gov/nrcssirrig)
• Note that SS is a function of the field topography and micro-topography,
and is affected by foliar interception of applied water
• These last two equations have π in them because there is an inherent
assumption of an elliptical water application profile from the sprinklers or
sprayers
• Recall that ARav = (π/4)ARx for an elliptical pattern

• SF is a relative sealing factor (in terms of soil water infiltration), and may
have values in the range of 0 to about 0.36

Merkley & Allen Page 180 Sprinkle & Trickle Irrigation Lectures
 • The higher values of SF tend to be for freshly tilled soils, which are
generally most susceptible to surface sealing from the impact of water
drops
• Lower values of SF are for untilled soils and vegetative cover, such as
alfalfa or straw, which tend to reduce the impact of water drops on the soil
and help prevent runoff too
• If the linear move irrigates in both directions (no deadheading), then df is
one-half the value from these two equations

6. Compute the total wetting time, ti, in minutes:

df
ti = (303)
π
( AR x )
4
7. Compute the speed of the system for the required ti:

S = w/ti (m/min) (w is for a specific nozzle type)

If S ≥ Smax (this may occur for a high intake soil or for a very light
application with surface storage) then reduce the application rate and
increase time as follows:
w
ti = (304)
Smax

4 df
AR x = (305)
π ti

Thus,
S = Smax (306)

8. Calculate maximum field length, X:

8(a). For irrigation in one direction, only (dry return, or deadheading):

60 f T − 2 treset
X= (307)
⎛ 1 1 t ⎞
⎜⎜ + + hose ⎟
⎟
⎝ S wet Sdry 100 ⎠

where,

Sprinkle & Trickle Irrigation Lectures Page 181 Merkley & Allen
 X = maximum length of field (m);
f = system operating time per irrigation (days);
T = hours per day system is operated (21-23);
treset = time to reset lateral at each end of the field (min);
those = time to change the hose (min/100 m);
Swet = maximum speed during irrigation (m/min); and
Sdry = maximum dry (return) speed (m/min)

labor =
(
2 treset + 0.01X those + 2 t super ) (308)
60 f

where labor is in hrs/day; and tsuper is minutes of supervisory time per 100
m of movement

8(b). For irrigation in both directions (no deadheading):

60 f T − 2 treset
X= (309)
⎛ 1 t ⎞
2⎜ + hose ⎟
⎝ S wet 100 ⎠

and labor is calculated as above in 8(a)

9. Calculate the irrigated area:

XL
Areamax = (310)
10,000

where Areamax is in ha; and L is the total lateral length (m)

10. Labor per hectare per irrigation, Lha:

labor
Lha = (311)
Areamax

11. Repeat steps 5-10 for a different value of df

12. Repeat steps 4-11 for a new w (different application device or different
operating pressure)

13. Select the nozzle device and application depth which maximizes the field
length (or fits available field length) and which minimizes labor
requirements per ha
Merkley & Allen Page 182 Sprinkle & Trickle Irrigation Lectures
 14. System capacity:

π AR x w L
Qs = (312)
4k 3 Re

where k3 = 96.3 for L and w in ft, Qs in gpm, and ARx in in/hr; and k3 = 60
for L and w in m, Qs in lps, and ARx in mm/min

The system capacity can also be computed as:

df w L
Qs = (313)
ti k 3 R e

II. Assumptions & Limitations of the Above Procedure

• In the above procedure (and in the USUPIVOT computer program), when

designing for a system which irrigates in both directions, the second pass is
assumed to occur immediately after the first pass, so that the infiltration
curve is decreased due to the first pass before the ARx of the second pass is
computed
• This will occur near the ends of the field, where the design is most critical.
The proposed procedure assumes that:

• There is no “surge” effect of soil surface sealing due to a brief time period
between irrigation passes (when irrigating in both directions)
• The infiltration curve used represents soil moisture conditions
immediately before the initiation of the first pass
• The infiltration curve used holds for all frequencies (f) or depths (df)
evaluated, while in fact, as f↑, θ↓, so that the Kostiakov coefficients will
change. Therefore, the procedure (and field ring infiltration tests) should
be repeated using coefficients which represent the Kostiakov equation for
the soil moisture condition which is found to be most optimal in order to
obtain the most representative results.

Sprinkle & Trickle Irrigation Lectures Page 183 Merkley & Allen
Linear Move Design Example
I. Given Parameters

• Hose-fed linear move, irrigating in only one direction in a 64-ha field (400 m
wide and 1,600 m long)
• The pressure is 140 kPa (20 psi) for spray booms with a preliminary w if 10
m (33 ft)
• The soil infiltration characteristics are defined for the Kostiakov-Lewis
equation as:
Z = 5.43τ0.49 (314)

with Z in mm of cumulative infiltrated depth; and τ is intake opportunity time

in minutes. Other design parameters:

Ud = 7.7 mm/day
MAD = 50%
Z = 0.9 m
Wa = 125 mm/m
Oe = 1.00
Re = 0.94
Epa = 85%

• Maximum dry (returning) speed = 3.5 m/min

• Maximum wet (irrigating) speed = 3.0 m/min
• Reset time = 0.5 hours per end of field
• Hose Reset time = 10 min/100 m of travel distance
• Supervisory time = 5 min/100 m of travel distance

II. One Possible Design Solution

• This design will consider only spray booms with w = 10 m

• Note that the full procedure would normally be performed with a computer
program or spreadsheet, not by hand calculations

1. Calculate the maximum application depth per irrigation [dx = MAD(z)(Wa),

or less]

dx = (0.5)(0.9)(125) = 56 mm (315)

dx 56
f' = = = 7.3 ⇒ f ' = 7 days (316)
Ud 7.7

Merkley & Allen Page 184 Sprinkle & Trickle Irrigation Lectures
 2. Net and gross application depths:

dn = f Ud = (7)(7.7) = 54 mm (317)

dn 54
d= = = 64 mm (318)
Epa 0.85

3. Infiltrated depth at each irrigation:

D f = dR e = (64)(0.94) = 60 mm (319)

4. For a series of 10 infiltration values, calculate df, beginning with df = Df

/10:

⎛ i ⎞
df = D f ⎜ ⎟ (320)
⎝ 10 ⎠

where i = 1 to 10. For this example, let i = 4 and, df = (0.4)(60 mm) = 24

mm. Then,

df DEpa (24)(0.85 / 0.94)

f' = = = 2.8 days (321)
Ud 7.7

Assume no days off (no down time during the peak use period)

f = f '− days off = 2.8 − 0 = 2.8 days (322)

5. Determine the maximum ARx for the particular df depth:

From Eq. 282:

ARx = 0.97 mm/min

ARx reaching the soil surface = 0.97 (Re) = 0.91 mm/min

6. Compute the total wetting time, ti, in minutes:

4 df 4(24)
ti = = = 34 min (323)
π AR x π (0.91)

7. Compute the speed of the system for the required ti:

Sprinkle & Trickle Irrigation Lectures Page 185 Merkley & Allen
 w 10
S= = = 0.3 m / min (324)
ti 34

Thus, S < Smax (3.0 m/min), so this is OK.

8. Calculate maximum field length, X:

For irrigation in one direction, only (deadhead back):

60 f T − 2 treset
X= =
⎛ 1 1 t ⎞
⎜⎜ + + hose ⎟
⎟
⎝ S wet Sdry 100 ⎠ (325)
60(2.8)(22) − 2(30)
= = 970 m
⎛ 1 1 10 ⎞
⎜ + + ⎟
⎝ 0.3 3.5 100 ⎠

and, the labor requirements are:

(
2 treset + 0.01 those + 2 t sup er X ) =
60 f
(326)
2(30) + 0.01[10 + 2(5)][970]
= = 1.5 hrs / day
60(2.8)

where treset is the reset time at the end of the field (min); those is the hose
reconnection time (min/100 m); and tsuper is the “supervisory” time
(min/100 m)

9. Maximum irrigated area:

Areamax = XL/10000 = 970(400)/10000 = 38.8 ha

which is only about half of the actual field area!

10. Labor per ha per irrigation, L/ha:

L/ha = (labor/area)max = 1.5/38.8 = 0.039 hr/ha/day

11. Repeat steps 5 - 10 for a new df (not done in this example)

Merkley & Allen Page 186 Sprinkle & Trickle Irrigation Lectures
 12. Repeat steps 4-11 for a new w (different application device or operating
pressure). (not done in this example).

13. Select the nozzle, device and application depth that maximizes the field
length (or fits the available field length), and which minimizes labor
requirements per ha.

Note: 38.8 ha << 64 ha, which is the size of the field, (970 m << 1600 m
which is the length of the field). Therefore, it is important to continue
iterations (steps 11 and 12) to find an application depth and or new w
(different sprinkler or spray device) to reach 1600 m and 64 ha, if
possible.

Additional Observations:

• For a 6-m spray boom, applying a 12-mm depth per each 1.4 days
would almost irrigate the 64 ha. However, the labor requirement is
doubled, as the machine must be moved twice as often. This
additional cost must be considered and weighed against the larger
area irrigated with one linear move machine.
• If larger spray booms were used (w = 16 m rather than 10 m) (these
would be more expensive) then 18 mm could be applied each 2.1
days, and all 64 ha could be irrigated with one machine.
• If low pressure impact sprinklers were used (these would be less
expensive than spray booms, but energy costs would be higher), then
w = 22 m, and 30 mm could be applied each 3.5 days (more water
can be applied since the application rate is spread over a wider area
from the lateral), and all 64 ha could be irrigated. In addition, ETc
would be less since the soil would be wetted less often. Also, the soil
intake rate would be higher each irrigation because of a drier
antecedent moisture at the time of irrigation.
• Notice that required wetting time for rotation times (f) greater than 2
days are identical between all types of spray devices. This is
because, for the large depths applied, a minimum wetting time is
required. The system speed is adjusted to fit the w value of the water
application device.
• If no acceptable solution for this problem were found, then
alternatives to be evaluated would be to irrigate in both directions, or
to consider a ditch-fed linear move (this requires a leveled ditch, but
does not required time for moving hoses and hose friction losses).
• You could also consider a “robot” controlled machine that
automatically connects alternating arms to hydrants on a buried
mainline (but this is a very expensive alternative)
• You might begin to wonder whether an investment in a linear move
machine is justifiable when there is a significant labor requirement for
reconnecting the supply hose, resetting at the end of the field, and
Sprinkle & Trickle Irrigation Lectures Page 187 Merkley & Allen
 supervising operation. That is, why not put in a center pivot or a side
roll system instead?
• If one linear move cannot cover the entire field length in the available
period, “f” (days), you could consider two linear move machines for
the same field

14. System Capacity:

π AR x w L π (0.91)(10)(400)
Qs = = = 51 lps (809 gpm)
4k 3Re 4(60)(0.94) (327)

alternatively,

df wL (24)(10)(400)
Qs = = = 51 lps (809 gpm) (328)
ti k 3R e (33.6)(60)(0.94)

Note that the computed Qs is larger than one based strictly on Ud and T,
because the machine is shut off during reset and hose moving

For Qs based only on f, A, d and T, with no consideration for those,

Ad (38.3)(24)
Qs = 2.78 = 2.78 = 44 lps (700 gpm) (16)
fT (2.8)(22)(0.94)

But this flow rate is too low – it does not consider hose moving and
reset time. So, the 51 lps system capacity should be used for design

Merkley & Allen Page 188 Sprinkle & Trickle Irrigation Lectures
Lecture 16
Trickle Irrigation System Components & Layout

I. Introduction and Descriptions

• “Trickle” and “drip” are terms used to describe what can be generally called
“micro-irrigation systems”, in which water is applied in relatively precise
quantities and precise times and at precise locations
• Land-leveling costs notwithstanding, trickle irrigation systems are usually the
most expensive types of on-farm water application system to install
• They can also be expensive to operate and maintain
• Usually, trickle irrigation systems are installed in areas where water is scarce
and or expensive, crop value is very high, or topographical and other
conditions might preclude the successful use of other types of irrigation
systems
• Not all micro-irrigation systems are complex and expensive
• Labor-intensive forms of micro-irrigation continue to be practiced in many
areas of the world, especially for vegetable and other “cash” crops
• For example, people may carry water in buckets or shoulder harnesses to
carefully pour at each plant in a field
• Or, porous pots are buried at regular intervals along rows and filled with
water individually, which seeps out into the surrounding soil
• Sometimes water is merely splashed onto crop beds by hand

II. Advantages and Disadvantages of Trickle Systems

Advantages

1. Significant water, fertilizer, and operating cost (labor and

power) savings are possible
2. Ease of field operations due to reduced weed problems and non-wetted soil
surface (e.g. strawberries)
3. Ability to apply saline water because of frequent (daily) irrigation; thus, soil
water salinity is nearly the same as the irrigation water salinity
4. Ability to operate on steep slopes and rough terrain
5. The ratio of crop yield to evapotranspiration can be higher under trickle
irrigation because of reduced soil surface evaporation, continuously high soil
water (near FC), and lower root zone salinity due to frequent application
6. Relatively easy to automate the system
7. Can be less labor intensive than some other irrigation systems

Sprinkle & Trickle Irrigation Lectures Page 189 Merkley & Allen
 Disadvantages

1. Systems are expensive to purchase and install ($1,000 to $6,000 per ha)
2. Susceptibility to clogging of emitters, which usually have very small
openings – so, it is important to spend time and money on maintaining the
system, applying chemicals, and keeping filters clean
3. Possibly low distribution uniformity due to low operating pressures and
possibly due to steep slopes, especially along laterals, and due to clogging
4. Where laterals are on steep slopes, the water will drain out the downhill end
at every startup and shut-down.
5. Soils with very low intake rates will exhibit ponding and runoff
6. Salt tends to accumulate at the soil surface and around the wetted area --
when it rains, these accumulated salts may be driven into the root zone
7. These systems tend to require more capable and diligent management
because of the susceptibility to clogging, and because the systems are
usually designed to operate continuously during peak ET periods (can’t
afford to let the system shut down during these periods). These systems do
not usually take full advantage of the soil storage (buffer) capacity.

III. Trickle System Components

• The following is a list of many

of the common components
found in modern trickle
irrigation systems:

1. Pump & motor

2. Control head
• Valves
• Filters and
Screens
• Chemical Injection
Equipment
• Flow Rate or Volume Meters (U/S of acid injection)
3. Mainline
• Submains
• Manifolds
• Laterals
4. Water applicators
• Emitters
• Bubblers
• Sprayers
• Misters
• Others

Merkley & Allen Page 190 Sprinkle & Trickle Irrigation Lectures
 5. Other equipment
• Valves & air vents
• Vacuum Relief Valves
• Pressure Relief Valves
• Various Pipe Fittings and Appurtenances

• Not all trickle irrigation systems will have all of these components
• For example, some systems are gravity-fed and require no pumping
• Simple systems may not have submains and manifolds
• Some systems do not have pressure relief or other types of safety valves
• Systems with relatively dirty water will have multiple levels of filtration, others
may have only minimal screening

IV. Types of Water Applicators

• The basic purpose of water applicators in trickle irrigation systems is to

dissipate energy
• This is because lateral pressures must be high enough to provide adequate
uniformity, yet emitters must yield small flow rates
• Otherwise, one could simply punch holes in a plastic pipe (in some cases this
is exactly what people do)

1. Drip Emitters

• Emitters are the typical water application device in many systems

• Can be “on-line” (usually with barb) or “in-line” (inserted into tubing
by the tubing manufacturer)
• Various approaches are used to provide
energy dissipation: long-path (“spaghetti
tubing”), tortuous path (labyrinth), orifice,
vortex,
• Some emitters have flushing, or continuous
flushing features to help prevent clogging while
still having small flow rates (some designs are
very clever)
• Some emitters have pressure compensating
features to provide a more constant flow rate
over a range of operating pressures
• Some emitters are designed with multiple outlets

2. Line Source Tubing

• These are typically buried

• Single chamber with small orifices

Sprinkle & Trickle Irrigation Lectures Page 191 Merkley & Allen
 • Double chamber with orifices between chambers and orifices to
discharge water into the soil (acts something like a manifold to
control pressures and provide greater uniformity)

• Can be removed and reused next year (typically 4-5 years life)
• Can be “disked up” and left in the field as chunks of plastic
• Porous or “leaky” pipe, made from old tires or new materials

3. Micro-Sprayers

• These are like small sprinklers, but

may not overlap enough to wet the
entire ground surface
• Sometimes referred to as “spitters”
• A “gray area” between micro and
sprinkle irrigation – they have both
precise application and aerial +
soil distribution of water
• Larger wetted area per
applicator, compared to non-
spray emitters
• Less susceptible to clogging than most emitters

V. Types of Pipe Materials

• Laterals are usually polyethylene (PE)

• Laterals may be made from polybutylene (PB) or PVC
• Metal pipes are not used because of corrosion problems from injected
chemicals
• Non-buried lateral tubing should be black to discourage growth of algae and
other organic contaminants (don’t allow light to enter the tubing)
• Non-buried pipes preferably have ultraviolet light (UV) protection to prevent
rapid deterioration from exposure to sunlight

Merkley & Allen Page 192 Sprinkle & Trickle Irrigation Lectures
VI. Typical Trickle System Layouts

• Chemical injection should be upstream of filters to prevent system clogging

• Chemical tank and valves are often plastic to avoid corrosion and freeze-up
• A check valve will help prevent contamination of the water supply
• Manifolds are often the basic operational subunits
• Valves for flushing at ends of manifolds and laterals are often manually
operated
• Various mainline-manifold-header arrangements are possible

VII. Emitter Flow Rate versus Pressure

• The discharge equation for emitters is similar to that used for sprinkler
nozzles, but the exponent on the head or pressure term is variable
• An exponent of ½ corresponds to orifice flow, which is how some, but not all,
emitters are designed
• The general emitter equation is:

q = K dHx (329)

where q is the volumetric flow rate; Kd is the discharge coefficient; H is the

head (or pressure); and x is the exponent

• Note that the value of Kd is different depending on whether units of head or

pressure are used in the equation
• The exponent in the above equation is a function of the emitter design
Sprinkle & Trickle Irrigation Lectures Page 193 Merkley & Allen
 • For purely turbulent orifice flow the exponent is ½, assuming the pressure
head is fully converted to velocity head
• Pressure compensating emitters have x ≤ ½. A fully pressure compensating
emitter would have x = 0 (but these don’t really exist)
• Long-path, laminar-flow emitters typically have x ≈ 0.7 (if the flow were
completely laminar, the exponent would be 1.0)
• Vortex-type emitters typically have x ≈ 0.4

VIII. Emitter Design Objectives

• The two basic emitter design objectives, other than energy dissipation, are:

1. Have a low value of the exponent, x

2. Have flushing properties to prevent clogging

• The first objective provides for pressure compensating features, if the

exponent is less than 0.5 (i.e. compensates better than a simple orifice)
• The first and second objectives tend to be conflicting, because the more
pressure compensating the less ability to flush particles

IX. Field-Measured Uniformity

• To measure emission uniformity in the field you can use an equation

equivalent to the Distribution Uniformity (DU), as applied to sprinkler
systems:

⎛ q'n ⎞
EU' = 100 ⎜ ⎟ (330)
⎜ qa ⎟
⎝ ⎠
where EU’ is the field test emission uniformity; qn’ is the average discharge of
the low ¼ emitters from the sampling; and qa is the average discharge of all
emitters sampled

• EU’ should be at least 95% for properly designed and properly maintained
trickle irrigation systems
• Note that it is impossible to calculate EU’ based on field measurements if the
system is being designed (hasn’t been installed yet) – in this case there are
other equations to approximate EU (recall the design efficiency for sprinkler
systems)
• Most nonuniformity in micro irrigation systems is caused by: (1) emitter
plugging, wear, and manufacturing variations; and, (2) nonuniform pressure
distribution in pipes and hoses

Merkley & Allen Page 194 Sprinkle & Trickle Irrigation Lectures
X. Manufacturer’s Coefficient of Variation

• Emitters of the same type and manufacture have variations in discharge (at
the same operating pressure) due to small differences from manufacturing
tolerances. Some variation is allowed in the interest of cost savings.
• The manufacturer’s coefficient of variation is defined as:

s
ν= (331)
qa

where the standard deviation, s, is calculated from at least 50 samples (all at

the same pressure) of the same emitter type and model; and the
denominator is the average discharge of all emitters

XI. Emitter Layouts

• There are a number of configurations designed to increase the percent

wetted area, and still be economical
• There are tradeoffs between flow per lateral and total length of pipe and
tubing
• Below are some common emitter layouts

Sprinkle & Trickle Irrigation Lectures Page 195 Merkley & Allen
XII. Valves & Automation

• Valve automation can be accomplished with individual timers/devices, or with

a central controller

1. Volumetric Valves

• Manually turned on
• Automatically turn off

2. Sequential Operation

• Manually turned on
• Automatic sequencing from low to high elevation

3. Fully Automatic

• Time-based: soil moisture sensors determine whether to turn the

system on at a given time each day, then the system runs for fixed
duration in each subunit
• Volume-based: uses a flow meter to measure a specified volume
delivered to a subunit, then turns water off to that subunit
• Soil moisture-based: uses tensiometers, resistance blocks, or other
device to determine when to irrigate and for how long

• Note that time-based systems may give varying application depths over time
if the system flow rate changes due to clogging of filters
• This can be partially corrected by using pressure compensating emitters
• However, the use of a volume-based system with a flow meter may be best
because the flow rate measurement also gives an indication about filter
clogging

Merkley & Allen Page 196 Sprinkle & Trickle Irrigation Lectures
Lecture 17
Filtration for Trickle Irrigation Systems

I. Introduction

• Water for trickle irrigation systems can

come from open reservoirs, canals, rivers,
groundwater, municipal systems, and other
sources
• Solid contaminants can include both organic
and inorganic matter
• Examples of inorganic matter are sand, silt
and clay (soil particles), and trash floating in
the water
• Examples of organic matter are bacteria,
algae, moss, weeds & weed seeds, small fish, insect larvae, snails, and
others (see the figure below)

• Solid contaminants need to be removed from trickle irrigation systems

because they:

1. Cause clogging of emitters, which can lead to serious water deficits

and non-uniformity of the applied water
2. May cause wear on pump impellers, emitter outlets, and other
hardware

Sprinkle & Trickle Irrigation Lectures Page 197 Merkley & Allen
 3. Can provide nutrients which support the growth of bacteria in the
pipes
4. Can accumulate at the ends of pipelines and clog valves
5. Can contain weed seeds which aggravate weed control in the irrigated
area
6. Cost the farmer money

• All of the above problems translate to direct costs to the farmer

• These filters do not remove salts from the irrigation water (unless reverse
osmosis membranes are used, which are very uncommon in irrigation
systems and are not covered here)

• Groundwater usually requires less filtration than

surface water, but even groundwater should be
filtered
• The maximum allowable particle size in trickle
irrigation water is usually between 0.075 mm and
0.2 mm, so the water must be quite clean
• Filtration is almost always complemented by the
injection of various chemicals into the water to help
prevent clogging due to bacterial growth and
precipitation of solids from the water
• Solid particles smaller than emitter outlets can cause clogging when they
bridge at the opening (see the figure above)
• Some consultants recommend the removal of all particles larger than 1/10 of
the minimum outlet diameter for drip emitters, or about 1/7 of the minimum
outlet diameter for spitters, misters, and microsprayers
• Larger particles may be allowed with spitters, misters, and microsprayers
because of “shorter pathways” and sometimes larger openings

II. Types of Filtration

• The basic types of filtration used in trickle systems are:

1. Reservoirs (settling ponds)

2. Pre-screening devices
3. Sand separators
4. Sand (media) tanks
5. Gravity overflow screens
6. Tubular screens
7. Disc filters

Merkley & Allen Page 198 Sprinkle & Trickle Irrigation Lectures
III. Use of Reservoirs in pre-Filtration

• Some of the benefits of an open reservoir upstream of the pumps in a trickle

irrigation system are:

1. To buffer differences in supply and demand rates. The supply from a canal
or well seldom coincides exactly with the system requirements (flow rate and
duration), and the system requirements can change due to different numbers
of stations in operation, “down time”, and duration of sets.
2. To allow for settling of some of the suspended particles. In these cases the
reservoir serves as a “settling basin”. Precipitated sediment can be
periodically removed from the reservoir with equipment or manual labor.

Soil Texture Particle Size Vertical Settling

(microns) Velocity (mm/min)
Coarse sand > 500 38,000 (1½ mph)
Medium sand 250 - 500 22,000
Fine sand 100 - 250 5,000
Very fine sand 50 - 100 900
Silt 2 - 50 15
Clay <2 0.6 (very slow!!)

The barchart below has a logarithmic scale on the ordinate:

100,000

10,000
Settling velocity (mm/min)

1,000

100

10

0
Coarse sand Medium sand Fine sand Very fine Silt Clay
sand

Longer settling basins will allow more time for suspended particles to fall to
the bottom before arriving at the pump intake.

Sprinkle & Trickle Irrigation Lectures Page 199 Merkley & Allen
 3. To aerate water pumped from wells, thereby oxidizing and precipitating
manganese and iron out of the water (some groundwater has manganese
and iron, and these can cause plugging of emitters). Only 1.5 ppm of either
manganese or iron can cause severe clogging problems in trickle laterals
and emitters (see Table 18.1 in the textbook).
4. To allow for air to escape when the water comes from a “cascading” well, in
which air becomes entrained into the water. Air in pipelines can dampen the
effects of water hammer, but also causes surges and blockages of flow.
5. To allow oils to collect on the water surface. Oils can cause rapid clogging of
most types of filters, requiring special cleaning with solvents and possible
replacement of sand media. When pumping from a reservoir the inlet is
below the water, and oil does not enter.

IV. Pre-Screening Devices

• These screens are intended to

prevent fish, large debris and trash
from entering the pipe system,
upstream of the other filtration
devices
• Pre-screening is not necessary for
groundwater or municipal water
supplies
• Pre-screening devices often have
“self-cleaning” features, otherwise
they can clog up rapidly
• Horizontal grills
• Screen plates
• Rotating and self-cleaning screens
• Gravity screen filters

• If the inlet line upstream of a pre-screening device is pressurized, you will

lose all of the pressure and have to repressurize downstream of the screens
– this is a major disadvantage to pre-screening in such cases

V. Sand Separators

• Sand separators are used to remove sand (but not organic matter) from the
water
• Most work by spinning the water in an enclosed column (or cone) to remove
sand through a centrifuge-type action
• There are no moving parts
• Solid particles with a density of approximately 1.5 g/cm3 can be removed by
these devices (most sand has a density of about 2.65 gm/cm3)

Merkley & Allen Page 200 Sprinkle & Trickle Irrigation Lectures
 • Can remove from 70 to 95% of dense particles
• Periodic purging of accumulated sand (manual
or automatic) is necessary to maintain
performance
• Must have the correct flow rate through the
sand separator for proper operation, otherwise
less sand will be removed from the water
• Most sand separators have a pressure loss of
between 5 and 12 psi, from inlet to outlet. This
pressure loss does not change with time, only
with flow rate.
• Some sand separators are designed to fit down
into wells to protect the impellers and pump
bowls, but they are not as efficient as above-
ground sand separators
• Sand separators cannot remove all of the sand, and may pass large amounts
when the system is starting or stopping
• Therefore, screen filters should be installed downstream
• Sand separators are available but are not used as much as they were in the
past because people are using media tanks and other filters instead
• When taking water from a deep well, an alternative to using a sand separator
is to properly develop the well and use a good quality well screen

VI. Sand Media Filters

1. Introduction

• This type of filter is filled with sand, or some

other particles such as crushed (makes it
angular, traps debris better) granite,
crushed silica material (e.g. garnet)
• Some designers go by a uniformity
coefficient for the media, defined as the
ratio of 40% retained size (larger) to the
90% retained size (smaller) from a sieve analysis
• This uniformity should usually be between 1.0 (perfect!) and 1.5
• In most media filters the water passes through vertically, from top to bottom
• A drain screen (of which there are various types) at the bottom allows clean
water to exit, but prevents the media from leaving the filter tank.
• Many media filters have a layer of gravel around the bottom drain
• The inside top of the tank usually has some kind of “diffuser” to help spread
the dirty water evenly over the media surface
• Standard tank sizes are 24, 36 and 48 inches in diameter, with a media
depth of about 30 to 40 cm
• Tank walls are usually 3 - 5 mm thick, depending on the brand
Sprinkle & Trickle Irrigation Lectures Page 201 Merkley & Allen
 • The tanks are often made out of carbon steel, type 304 stainless steel, or
type 316 stainless steel (if water has high salt content)

• Every installation should have at least two tanks so that back-flushing can
occur during operation, but many designers recommend at least three tanks
in which only one is back-flushed at a time
• New media should be rinsed with clean water before placing it in the tanks
because it may have dust and other particles in it
• Some tanks have not performed well when the installers failed to rinse the
media first (resulting in fine particles passing into the irrigation system when
the tanks are first put into use)

2. Applicability of Media Filters

• These filters are very good for removing relatively large amounts of organic
and inorganic matter, but some pre-screening is usually necessary with
surface water supplies
• High volume filtration at 20 to 30 gpm/ft2 (1.3 to 2.0 cm/s)
• Some silt and clay particles can also be removed by sand media filters, but
not by most screen-type filters. However, much silt and clay can pass
through a media filter too.
• Large volumes of particle contaminants can be collected in the sand media
before the media must be cleaned, or “back-flushed”
• In some cases the water must be pre-cleaned before entering the sand tanks
to prevent rapid accumulation of particle contaminants
• Media filters can also remove some sand from the supply water, but this
sand cannot always be effectively back-flushed from the media -- for large
amounts of sand, there should be a sand separator upstream of the media
tanks
• Industrial media filters are often five feet deep (or more), but have smaller
flow rates and less frequent back-flushing than agricultural media filters,
which may be only 14 inches deep
• Many of the particles captured by agricultural media filters stay within the
upper few inches of the sand because they are back-flushed frequently

3. Back-Flushing the Tanks

• Back-flushing is required to clean the tanks

• Back-flushing can be performed manually or automatically, based on elapsed
time and or on a pressure differential limit across the tanks
• Typical pressure differential triggers are 5 to 10 psi (35 to 70 kPa) greater
than the pressure differential when the media is clean (clean media typically
has a pressure differential of 3 to 5 psi, or 20 to 35 kPa)
• Often, a timer is set to back-flush at least one time per day, even if the
pressure differential criteria (for flushing) is not met
Merkley & Allen Page 202 Sprinkle & Trickle Irrigation Lectures
 • Automatic back-flushing is recommendable, because labor is not always
reliable
• Some installations have view ports on the back-flush pipes so that an
operator can see if the water is clean (to know if the duration of the back-
flush is sufficient) and to see if any of the sand media is escaping during
back-flush
• The pressure differential during a backflush operation should be 7 - 10 psi
(50 - 70 kPa) – if it is greater than 10 psi, the flow rate is too high

VII. Secondary Filters

1. Tubular Screen Filters

• These are conventional screen filters, with two-dimensional surfaces and

little capacity to accumulate debris
• There are many different kinds and variations of these filters
• Primarily used as backup (safety) filters downstream of the primary filters
• If the screen becomes dirty and is not cleaned, the pressure differential can
become great enough to burst the screen. Or, the screen may stretch until
the openings expand enough to pass some of the debris (which is not
desirable)

• Flow through the filter is usually from inside to outside (debris is trapped on
the inside surface during operation) to prevent collapse of the screens
• Cleaning can be manual or automatic, and there are many varieties of
automatic cleaning methods
Sprinkle & Trickle Irrigation Lectures Page 203 Merkley & Allen
 • Some filter designs have a rotating suction mechanism to clean the dirty
(inner) side of the screen element
• Manually-cleaned filters can have slow or quick release cover latches -- the
slow release latches are preferred because the quick release version can
“explode” if opened while the system is at operating pressure (dangerous to
personnel)

2. Disc Filters

• Similar to a tubular screen filter, but using tightly packed plastic

disks for the filter media, with a deeper filter area
• Holds more contaminants than a regular screen filter without
clogging
• Often installed in banks (several filters in parallel)
• Often have automatic back-flushing features, requiring higher
pressure (about 45 psi minimum pressure) than normally available
for system operation, so there is a special booster pump for
cleaning
• Cleaning is often performed when the pressure differential across the filter
reaches 6 psi
• The pressure differential for a clean filter should be about 1 to 4 psi (unless
too much water is being pumped through the filter)
• These filters are not designed to remove sand from the water (sand gets
stuck in the grooves)
• These filters can have clogging problems with some kinds of stringy algae

VIII. Chemical Injection for System Maintenance

• It is usually necessary to use chemicals to maintain a trickle irrigation system

-- if not, the system will eventually become clogged
• Some chemicals are for the addition of plant nutrients, or fertilizers; this is
called “fertigation”
• Different chemicals may be used for pest control: herbicides, insecticides,
fungicides
• Other chemicals are used to kill bacteria and other organic contaminants,
and maintain a sufficiently low pH
• The uniformity at which chemicals are applied to the field can be assumed to
be equal to the emission uniformity of the emitters (assuming the chemicals
are water soluble, which they should be if injected into the irrigation system)
• Following are guidelines for clogging hazard of irrigation water in trickle
systems with emitter flow rates of 2 to 8 lph (after Bucks and Nakayama
1980):

Merkley & Allen Page 204 Sprinkle & Trickle Irrigation Lectures
 Kind of Problem Hazard Level
low moderate severe
suspended solids 50 ppm 50-100 ppm > 100 ppm
pH 7.0 7.0-8.0 > 8.0
salts 500 ppm 500-2,000 ppm > 2,000 ppm
bicarbonate -- 100 ppm --
manganese 0.1 ppm 0.1-1.5 ppm > 1.5 ppm
total iron 0.2 ppm 0.2-1.5 ppm > 1.5 ppm
hydrogen sulfide 0.2 ppm 0.2-2.0 ppm > 2.0 ppm
bacteria count 10,000/liter 10,000-50,000/liter > 50,000/liter

• Types of clogging in trickle systems that can be managed through the

injection of chemicals:

1. Slimy bacteria
These can grow inside pipes and inside emitters.
The chemicals used to kill this bacteria are
chlorine, ozone, and acids.

2. Iron and manganese oxides

Some kinds of bacteria can oxidize iron and manganese. Only small
amounts of iron or manganese are necessary to support bacterial
growth in the water. This can be treated with chlorine injection, acid
and chlorine injection, linear phosphate injection, and aeration in a
pond.

3. Iron and manganese sulfides

These are problematic with some groundwater. Dissolved iron and
manganese form a black, insoluble material. Manganese is toxic to
most plants in small concentrations, and this may become a problem
before clogging occurs. This can be treated by aeration,
chlorination, and acid injection.

4. Precipitation of calcium and magnesium carbonates

These are typically quantified by SAR (or adjusted SAR). You can
use a dropper to put some hydrochloric or muriatic acid on selected
drip emitters to test for the presence of carbonates and
bicarbonates. If present, the acid will react and cause a “fizzing”
sound and bubbles. One method to control precipitation of
carbonates is to continuously inject carbon dioxide gas into the
system, creating carbonic acid and lowering the pH. Sulfuric and
phosphoric acids are also used, as are SO2 generators. All of these
are designed to maintain the pH near (or slightly below) 7.0.

Sprinkle & Trickle Irrigation Lectures Page 205 Merkley & Allen
 5. Plant root entry into underground emitters
This is mainly a problem in permanent (several years) buried trickle
irrigation laterals. Can use acid injection at end of each season for
perennials to kill roots that are in the buried drip tubing. Or, use
herbicides to kill roots in the tubing without damaging the plants.
Some emitters and plastic drainage pipe have herbicide in the plastic
to discourage roots from entering.

• It is dangerous to experiment with chemical mixtures in trickle systems,

because some mixtures can cause clogging of the emitters
• Test the mixture in a glass container first, checking it after several hours or
more, and determining whether the chemical mixture is water soluble

• Injection of chemicals should be after the system starts, and stopping before
the system is turned off
• As a rule of thumb, one can assume an average pipe flow velocity of 1 fps, or
0.3 m/s, divide this into the longest pipe distance in the system (from pump to
farthest emitter), and determine the time
• This is the time to wait after starting the pump, and the time to allow for
flushing before turning the pump off

For example, if the farthest emitter is 600 ft from the

pump, the travel time can be estimated at 600 s, or 10
min. Thus, you should wait 10 min before beginning
chemical injection, and discontinue chemical injection 10
min before stopping the system, or before changing
stations.

• Chemicals should be injected on a mass basis per set, not time. Thus, one
would want to apply a certain number of lbs or kg of a chemical in an
irrigation set, and it does not matter that it is all applied quickly or over a long
time (provided that the starting and stopping delay discussed above is
adhered to)
• The minimum injection rate can be put into equation form:

qc =
Fr A
=
(kg / ha )(ha ) = lph (332)
ctr Ta ( kg / liter )( 0.8 )( hrs / set )

where Fr is the mass application rate per unit area; A is the area irrigated per
set; c is the concentration of the chemical; tr is some kind of uniformity ratio,
taken to be 0.8; and Ta is the hours per set, or hours of chemical injection, if
shorter than the set time

Merkley & Allen Page 206 Sprinkle & Trickle Irrigation Lectures
Lecture 18
Trickle Irrigation Planning Factors

I. Soil Wetted Area

• Trickle irrigation systems typically apply small amounts of water on a

frequent basis, maintaining soil water near field capacity
• But, usually not all of the soil surface is wetted, and much of the root zone is
not wetted (at least not by design) by the system
• Recall that the system is applying water to each individual plant using one or
more emission points per plant

Widely-Spaced Crops

• These include orchards and vineyards, for example

• According to Keller & Bliesner, for widely-spaced crops, the percent wetted
area, Pw, should normally be between 33% and 67%
• The value of Pw from the irrigation system can fall below 33% if there is
enough rainfall to supplement the water applied through the trickle system
• Lower values of Pw can decrease the irrigation system cost because less
emitters per unit area are required
• Lower values of Pw can allow more convenient access (manual labor &
machinery) for cultural practices during irrigation
• Lower values of Pw can also help control weed growth in arid and semi-arid
regions, and reduce soil surface evaporation
• Lower values of Pw carry the danger that the soil will dry to dangerously low
levels more quickly in the event the irrigation system goes “off-line” for any
reason (power failure, broken pipe, pump problems, labor shortage, etc.)
• With lower values of Pw, there is less storage of applied water in the root
zone, especially with light-textured soils (sandy soils)
• With tree crops, low values of Pw can lead to “root anchorage” problems, in
which root extension is insufficient to support the trees during winds

Closely-Spaced Crops

• These include most row crops

• Actual Pw values may be near or at 100% with row crops and subsurface drip
irrigation systems (in the USA rows are typically spaced from 30 inches to 60
inches)
• Larger values of Pw usually mean more extensive root development, and
enhanced ability for the plant to make use of any rain water that may come
• Figure 19.1 in the textbook shows a generalized relationship between Pw,
amount of rainfall, and crop production level – the figure implies that
maximum crop yield may be higher under a trickle irrigation system than with
other methods
Sprinkle & Trickle Irrigation Lectures Page 207 Merkley & Allen
 • Figure 19.1 indicates that 100% crop yield might be obtained, in general, with
Pw ≥ 33%

Wetted Soil Area, Aw

• The wetted soil area, Aw, is not measured at the soil surface, but from a
horizontal plane about 30 cm below the soil surface (actually, it depends on
root depth and soil type)
• The same is true for Pw
• The reason we are interested in Pw is to calculate the application depth “dx,”
as discussed in the following lecture
• This wetted area is distorted for sloping terrain, but the distortion is uniform
for uniform slopes (all other factors being the same)

• Wetted soil area can be estimated from empirical relationships and tables
(Table 19.1 in the textbook), but it is best to have site-specific field data in
which potential emitters are operated in the design area
• That is, test the emitter(s) and spacings in the field before completing the
irrigation system design
• Calculate percent wetted area, Pw, as follows:

⎛ N pS e w ⎞
Pw = 100 ⎜ ⎟ , for Se < 0.8w (333)
⎜ S pS rPd ⎟
⎝ ⎠
where Np is the number of emission points (emitters) per plant; Se is the
spacing of emitters along a lateral; w is the wetted width along the lateral; Sp
is the spacing of plants along a row; Sr is the spacing between rows; and Pd
is the fraction (not percent) of area shaded (see Lecture 19)

• Note that the numerator of Eq. 333 is wetted area, and the denominator is
actual plant area
• Note also that some emitters have multiple emission points

Merkley & Allen Page 208 Sprinkle & Trickle Irrigation Lectures
 • Se is the spacing between emitters on the lateral; however, if Se is greater
than 0.8w, then use 0.8w instead:

⎛ 0.8N p w 2 ⎞
Pw = 100 ⎜ ⎟ , for Se ≥ 0.8w (334)
⎜ S pS r Pd ⎟
⎝ ⎠

• Note that w is a function of the soil type

• Se’ is the “optimal” emitter spacing,
defined as 0.8w
• There are practical limitations to the value
of Se with respect to Sp, otherwise there
may not be enough emitters per plant
(perhaps less than one)

• Sample calculation:

• Suppose Sr = Sp = 3.0 m, Pd =
80%, and w = 1.1 m
• Determine Np for Pw ≥ 33%

Se ' = 0.8w = 0.8(1.1) = 0.88 m (335)

Np (0.88)(1.1)
0.33 = (336)
(3.0)(3.0)(0.80)
whereby Np = 2.45. Then,

3(0.88)(1.1)
Pw = = 0.40 (337)
(3.0)(3.0)(0.80)

• For double-lateral trickle systems, spaced Se’ apart, Pw is calculated as

follows (see Eq. 19.4):

⎛ N S' ( S' + w ) ⎞
Pw = 100 ⎜ ⎟ , for Se ≤ 0.8w
p e e

( )
(338)
⎜ 2 Pd S pS r ⎟
⎝ ⎠
or,
⎛ 1.44 w 2 N ⎞ 72 w 2 N
Pw = 100 ⎜ ⎟=
p p
, for Se ≤ 0.8w
( ) ( )
(339)
⎜ 2Pd S pS r ⎟ Pd S pS r
⎝ ⎠

Sprinkle & Trickle Irrigation Lectures Page 209 Merkley & Allen
 Se’=0.8w

0.5w
w
Se ’

0.5w

Se Se
Double laterals
• As in the previous equation, if Se > Se’, use Se’ instead of Se in the above
equation for double laterals
• In the above equation, the denominator has a “2” because Np for double
lateral systems is always at least 2
• For micro-spray emitters, the wetted area is greater than that measured at
the surface (because it is measured below the surface):

⎡ ⎛ Se ⎞ ⎤
⎢ N p ⎜ A s + (PS ) 2 ⎟ ⎥
Pw = 100 ⎢ ⎝ ⎠ ⎥ , for S ≤ 0.8 w (340)
e
⎢ S pS rPd ⎥
⎢ ⎥
⎣ ⎦
where As is the surface area wetted by the sprayer; and PS is the perimeter
(circumference) of the wetted surface area

• In the above equation for Pw, the term in the inner parenthesis is:

Se π w 2 π w Se π w ⎛ w ⎞
A s + (PS ) = + = ⎜ + Se ⎟ (341)
2 4 2 2 ⎝2 ⎠
where w is the diameter corresponding to As, assuming a circular area

• See Fig. 19.4 on sprayers

Merkley & Allen Page 210 Sprinkle & Trickle Irrigation Lectures
Salinity in Trickle Irrigation

I. Salinity in Trickle Systems

• Salinity control is specialized with trickle irrigation because (usually) less than
100% of the area is wetted, and because water movement in the soil has
significant horizontal components
• Irrigation water always contains salts, and fertilizers add salt to the crop root
zones -- salinity management in the crop root zone is a long-term
management consideration with trickle systems, as it is with any other
irrigation method
• Salts tend to accumulate, or “build up”, at the periphery of the wetted bulb
shape under the soil surface

1. Rain can push salts near the surface down into the crop root area (but
a heavy rain can push them all the way through the root zone)
2. If and when the irrigation system is not operated for a few days, there
can be pressure gradients in the soil that pulls salts from the periphery
up into the root zone

• The crop is depending on frequent irrigations (perhaps daily) to keep salt

build-ups from moving into the root mass
• It may be necessary to operate the trickle system immediately following a
light rain to keeps salts away from roots (even if the soil is at field capacity)

Sprinkle & Trickle Irrigation Lectures Page 211 Merkley & Allen
 • Annual leaching with surface irrigation or sprinklers (on a trickle-irrigated
field) may be necessary to clean salts out of the root zone, unless there is a
rainy period that provides enough precipitation to leach the soil
• If the irrigation water has high salinity, trickle systems can provide for higher
crop production because the frequent irrigations maintain the soil salinity
nearer to the ECw (this is often not the case with sprinklers and surface
irrigation systems - salinity concentrates due to ET processes between water
applications)

II. Yield Effects of Salinity

• According to Keller, the relative crop yield can be estimated as (Eq. 19.6):

Yactual (ECe )max − ECw

Yr = = (342)
Ypotential (ECe )max − (ECe )min

• This is the relative crop yield (or production) in terms of soil water salinity
only
• ECw is the electrical conductivity of the irrigation water
• (ECe)max is the zero yield point, and (ECe)min is the 100% yield threshold
value
• (ECe)max may be as high as 32, and (ECe)min can be as low as 0.9
• This is based on the linear relationship between relative yield and salinity as
adopted years ago by FAO and other organizations
• Of course, calculated Yr values must be between 0 and 1

• Salinity of the soil extract, ECe, is measured by taking a soil sample to the
laboratory, adding pure water until the soil is saturated, then measuring the
electrical conductivity -- most published crop tolerance and yield relationships
are based on the ECe as a standard reference
• Crops don’t instantly die when the salinity approaches (ECe)max; the osmotic
potential increases and roots cannot extract the water that is there
• There can also be specific toxicity problems with minerals at high salinity
levels
• According to Allen, the relative yield will be near 100% for ECw less than
about 2(ECe)min, provided that frequent irrigations are applied (maintaining
salinity concentrations in root zone)

III. Leaching Requirement

• According to Keller & Bliesner, the leaching requirement under a trickle

system in an arid or semi-arid region does not consider effective rainfall (arid
regions often have more serious salinity problems, but tropical regions are
also subject to salinity in low areas)
Merkley & Allen Page 212 Sprinkle & Trickle Irrigation Lectures
 • Look at Eq. 19.7:
ECw
LR t = (343)
ECdw

where LRt is the leaching requirement under trickle irrigation (fraction); and
ECdw is the electrical conductivity of the “drainage water”, which means the
water that moves downward past the root zone

• ECdw can be replaced by 2(ECe)max for daily or every-other-day irrigations

(keep water moving through the root zone), still obtaining Yr = 1.0

ECw
LR t = (344)
2 (ECe )max

IV. Allen’s Equation for LRt

• R.G. Allen suggests a more conservative equation for calculating the

leaching requirement under trickle irrigation:

1. For continuous trickle system operation (daily or once every two days),
the soil water in the root zone is maintained near field capacity, which can be
taken as approximately 50% saturation (θv) for many soils. Thus,

ECe = 0.5ECsoil (345)

(recall that ECe is measured after adding distilled water to the soil sample
until it is saturated)

2. Suppose the average ECsoil is taken as (0.667ECw + 0.333ECdw). Then,

for 100% relative yield at field capacity:

(ECe )min = 0.5 (0.667ECw + 0.333ECdw ) (346)

solving for ECdw,

ECdw = 6 (ECe )min − 2EC w (347)

3. Substitute this last equation into Eq. 19.7 from the textbook to obtain:

EC w
LR t = (348)
6 (ECe )min − 2ECw

Sprinkle & Trickle Irrigation Lectures Page 213 Merkley & Allen
 this is similar to the leaching requirement as calculated for sprinkler
irrigation in Eq. 3.3 (coefficients 5 and 1 instead of 6 and 2), except that
(ECe)min is for 100% yield rather than 10% reduction in yield

Merkley & Allen Page 214 Sprinkle & Trickle Irrigation Lectures
Lecture 19
Water Requirements in Trickle Irrigation

I. Trickle Irrigation Requirements

1. Daily Use Rate

• The daily transpiration rate under a trickle system is based on Ud and the
percent area shaded (covered) by the plant leaves. Eq. 19.9:

Td = 0.1 Ud Pd (349)

where Ud is as previously defined and Pd is the percent (0 to 100) shaded

area when the sun is overhead (or most nearly overhead, in temperate
zones)

• Note that when Pd = 0, Td = 0

• Note that when Pd = 100%, Td = Ud
• Note that Td is called “transpiration,” but it really includes evaporation too

1.0

0.9

0.8

0.7

Closely-spaced crops
0.1*sqrt(Pd)

0.6

0.5
pi/4 = 0.785

0.4

0.3

0.2

0.1

0.0
0 10 20 30 40 50 60 70 80 90 100
Pd

Sprinkle & Trickle Irrigation Lectures Page 215 Merkley & Allen
 • The reduction from Ud is justified by considering the typical reduction in wet
soil evaporation with trickle irrigation
• The maximum Pd for a mature orchard is usually about π/4 (0.785), which is
the ratio of the area of a square and the circle it encloses:

• Tree spacing is generally such that the trees do not compete for sunlight, and
the area of each tree is equal to the square of the spacing between them (for
a square spacing)

2. Seasonal Water Use

• This is calculated as for the peak daily use in Eq. 19.9:

Ts = 0.1 U Pd (350)

3. Seasonal Water Deficit

• To determine the seasonal water deficit, to be supplied from the irrigation

system, consider effective rainfall and initial soil moisture, in addition to
percent shaded area:

(
Dn = (U − Pe − Ms ) 0.1 Pd ) (351)

where U is used instead of Ts because Pe (effective precipitation) and Ms

(initial soil water content) are over the entire surface area

4. Net Depth per Irrigation

• This is the same as for sprinkle irrigation (or surface irrigation), but with an
adjustment for percent wetted area. Eq. 19.12 is:

Merkley & Allen Page 216 Sprinkle & Trickle Irrigation Lectures
 MAD Pw
dx = Wa Z (352)
100 100
• Essentially, the same net volume of water is applied as with other irrigation
methods, but on a smaller area of the surface (and subsurface)
• Then, the maximum irrigation interval is:

dx
fx = (353)
Td

and f’ (round down from fx to get whole number of days) is less than or equal
to fx, but often assumed to be 1 day for trickle system design purposes.
Then,

dn = Tdf ' (354)

II. Gross Irrigation Requirements

• The transmission ratio (peak use period) takes into account the two-
dimensional infiltration pattern, or bulb shape, under trickle irrigation
• Even if the net depth is exactly right, there will almost always be some deep
percolation (more than that which may be required for leaching purposes)
• The transmission ratio, Tr, is used as a factor to increase required gross
application depth from dn
• The transmission ratio is equivalent to the inverse of the distribution
efficiency, DEpa, as given in Chapter 6 of the textbook
• The transmission ratio is lower for heavy-textured (“fine”) soils because there
is more lateral water movement in the soil, and the bulb shape is flatter; thus,
potentially less deep percolation losses
• Table 19.3 gives approximate values of Tr for different soil textures and root
depths (1.0 < Tr < 1.1) – obtain more representative values from the field, if
possible
• Then, for LRt < 0.1, or Tr > 1/(1-LRt), Eq. 19.15a:

⎛d T ⎞
d = 100 ⎜ n r ⎟ (355)
⎝ EU ⎠

where EU is the emission uniformity (%), which can be taken as a field-

measured value for existing trickle systems, or as an assumed design value

• EU takes into account pressure variations due to friction loss and elevation
change, and the manufacturer’s variability in emitter production
• If f’ = 1 day, then dn can be replaced by Td in Eq. 19.15a
Sprinkle & Trickle Irrigation Lectures Page 217 Merkley & Allen
 • For LRt > 0.1, or Tr < 1/(1-LRt), Eq. 19.15c:

100dn
d= (356)
EU (1.0 − LR t )

• The difference in the above two equations is in whether LRt or Tr dominates

• If one dominates, it is assumed that the other is “taken care of” automatically

Gross Volume of Water per Plant per Day

• Equation 19.16:
d
G= SpSr (357)
f'
with d in mm; Sp and Sr in m; and G in liters/day

• Note that millimeters multiplied by square meters equals liters

• This equation does not use Pw because d is calculated for the entire surface
area, and each plant occupies an SpSr area
• Other versions of this equation are given in the textbook for gross seasonal
volume of water to apply

Required Application Time During Peak-Use Period

• Equation 20.11:
G
Ta = (358)
Npqa

where Ta is the required application (irrigation) time during the peak-use

period (hr/day), with G in litres/day, and qa in litres/hr

III. Coefficient of Variation

• This is a statistical index to quantify discharge variations in emitters, at the

same operating pressure, due to differences in the emitter construction
• The coefficient of variation is important in trickle system design and
evaluation because it can significantly affect the adequacy of the system to
irrigate the least watered areas of a field
• For statistical significance, there should be at least 50 measurements of
discharge from 50 individual emitters of the same design and manufacture

Merkley & Allen Page 218 Sprinkle & Trickle Irrigation Lectures
 2
1⎛ n ⎞
∑ (qi2 )
n
− ⎜ ∑ qi ⎟
i=1
n ⎜⎝ i=1 ⎟⎠ σ
ν= = (359)
⎛1 n ⎞ qavg
n − 1 ⎜ ∑ qi ⎟
⎜n ⎟
⎝ i=1 ⎠
or,
n
∑ ( qi − qavg )
2

1 i=1
ν= (360)
qavg n −1

where n is the number of samples; σ is the standard deviation; qi are the

individual discharge values; and qavg is the mean discharge value of all
samples

• Standard classifications as to the interpretation of ν have been developed

(Soloman 1979):

Classification Drip & Spray Emitters Line-Source Tubing

Excellent ν < 0.05 ν < 0.1
Average 0.05 < ν < 0.07 0.1 < ν < 0.2
Marginal 0.07 < ν < 0.11 ---
Poor 0.11 < ν < 0.15 0.2 < ν < 0.3
Unacceptable 0.15 < ν 0.3 < ν

• For a large sample (n > 50) the data will usually be normally distributed
(symmetrical “bell-shaped” curve) and,

68% of the discharge values are within............... (1±ν)qavg

95% of the discharge values are within............. (1±2ν)qavg
99.75% of the discharge values are within........ (1±3ν)qavg

IV. System Coefficient of Variation

• The system coefficient of variation takes into account the probability that the
use of more than one emitter per plant will cause an effective decrease in the
combined discharge variability per plant due to differences in the emitters
(not due to pressure variability due to pipe friction losses and elevation
changes)
• On the average, discharge variability due to manufacturer tolerances will
tend to balance out with more emitters per plant
Sprinkle & Trickle Irrigation Lectures Page 219 Merkley & Allen
 ν
νs = (361)
N'p

where Np’ is the minimum number of emitters from which each plant receives
water (see page 493 of the textbook)

• For a single line of laterals per row of plants,

L w = w + (N − 1)Se (362)

where Lw is the length of the wetted strip; and N is the number of emitters
(assumed to be evenly spaced). Then,

⎛L − w⎞
N = 1+ ⎜ w ⎟ (363)
⎝ Se ⎠
or,
⎛ Sp − w ⎞
Np ' ≈ 1 + ⎜ ⎟ (364)
⎝ Se ⎠

V. Design Emission Uniformity

• In new system designs it is not possible to go out to the field to measure the
EU’ (Eq. 17.2) – a different approach is required to estimate EU
• The design EU is defined as (Eq. 20.13):

qn
EU = 100 (1 − 1.27 νs ) (365)
qa

where qn is the minimum emitter discharge rate in the system, corresponding

to the emitter with the lowest pressure; and qa is the average emitter
discharge rate in the system, corresponding to the location of average
pressure in the system

• qn and qa are calculated (not measured) in new system designs by knowing

the topography, system layout, pipe sizes, and Qs
• Note that the value in parenthesis in Eq. 20.13 corresponds to the low one-
quarter emitter discharge

Merkley & Allen Page 220 Sprinkle & Trickle Irrigation Lectures
 • EU gives a lower (more conservative) value than EU’, and the equation is
biased toward lower discharges to help ensure that the least watered areas
will receive an adequate application
• Graphical interpretations of these relationships are given in Figs. 20.9 and
20.10

VI. Average of the low ¼

• Note that the inclusion percentages for 1, 2 and 3 standard deviations

correspond to any normally distributed data
• Note also that ν qavg = σ
• The textbook says that for a normal distribution, the average flow rate of the
low one-quarter of measured q samples is approximately (1 - 1.27 ν)qavg
• The 1.27 coefficient can be determined from the equation for the normal
distribution and tabular values of the area under the curve
• The equation is:
2
1 ⎛ q−qavg ⎞
− ⎜ ⎟
2⎝ σ ⎠
e
occurrences = (366)
σ 2π

• To use the tabular values of area under the curve (e.g. from a statistics
book), it is necessary to use qavg = 0 and σ = 1 (the alternative is to integrate
the above equation yourself, which can also be done)
• Actually, qavg never equals zero, but for the determination of the 1.27
coefficient it will not matter
• In the tables, for area = 75%, the abscissa value (q, in our case) is about
0.675
• The same tables usually go up to a maximum abscissa of 3.49 (recall that
99.75% of the values are within ±3σ, so 3.49 is usually far enough)
• Anyway, for 3.49, the area is about 99.98%, and that is from -∞ to +3.49 (for
qavg = 0 and σ = 1), for the high ¼
• For the low ¼, take the opposite, changing to q = -0.675 and q = -3.49
• In this case (qavg = 0 and σ = 1), the equation reduces to:

2
e −0.5q
occurrences = (367)
2π

Sprinkle & Trickle Irrigation Lectures Page 221 Merkley & Allen
 • For q = 0.675, occurrences = 0.31766718
• For q = 3.49, occurrences = 0.0.00090372
• Finally,

0.31766718 − 0.00090372
= 1.268 (368)
0.9998 − 0.7500

• Therefore, (1 - 1.268 ν)qavg is the average of the lowest 25% of measured

discharge values, for any given values of ν & qavg, and given normally-
distributed data

V. System Capacity

• The system capacity of a trickle system can be calculated by Eq. 20.15:

A N pq a
Qs = 2.78 (369)
N s S pS r

where Ns is the number of stations (sets); and A is the total net irrigated area.
Or,

A qa
Qs = 2.78 (370)
N s S eS l

Merkley & Allen Page 222 Sprinkle & Trickle Irrigation Lectures
 where the coefficient “2.78” is for Qs in lps; A in ha; qa in lph; and Sp, Sr, Se,
and Sl in m (10,000 m2/ha divided by 3,600 s/hr = 2.78)

VI. Operating Hours per Season

• The approximate number of hours the system must operate per irrigation
season (or per year, in many cases) is equal to the required gross seasonal
application volume, divided by the system flow rate:

Vs
Ot = K (371)
Qs

where K = 2,778 for Vs in ha-m; and Qs in lps; and Vs is gross seasonal

volume of irrigation water

Sprinkle & Trickle Irrigation Lectures Page 223 Merkley & Allen
Merkley & Allen Page 224 Sprinkle & Trickle Irrigation Lectures
Lecture 20
Emitter Selection & Design

I. Introduction

• There are hundreds of models, sizes and types of emitters, sprayers,

bubblers, and others, available from dozens of manufacturers
• Prices of emitters can change frequently
• Some emitters have longer life than others, but cost more
• Some emitters have better pressure compensating features, but cost more
• Some emitters have better flushing capabilities, but cost more
• It is very difficult to know which is the “correct” emitter for a particular design,
and usually there are a number of emitters that could work and would be
acceptable for a given system
• Thus, the selection of an emitter involves knowledge of the different types,
their prices, their availability, and their performance
• Experience on the designer’s part is valuable, and emitter selection will often
involve a process of elimination

II. Long-Path Emitters

• So-called “spaghetti” tubing is a typical example of a long-path emitter

• Long-path emitters also come in spiral configurations (Fig. 20.1 of the
textbook)
• These can be represented by an equation used for capillary flow under
laminar conditions:

g π D4 H
Ac = (372)
ν qK

where lc is the length of the flow path; D is the inside diameter; H is the
pressure head; ν is the kinematic viscosity (a function of water temperature);
q is the flow rate; K is for units conversion; and g is the ratio of force to mass

• The above equation is only approximately correct for long-path emitters

• The above equation is based on circular cross-sections, which is typical
• The above equation assumes laminar flow, which may not be the case
• Note that the flow rate is proportional to the fourth power of the diameter, so
the diameter is a very important dimension
• Note also that the flow rate is inversely proportional to the length (double the
length and get half the flow rate)

Sprinkle & Trickle Irrigation Lectures Page 225 Merkley & Allen
 • When is it valid to assume laminar flow? Consider that a Reynolds number
of 4,000 is probably as high as you can go without transitioning from laminar
to turbulent flow:

VD 4Q
= < 4,000 (373)
ν π νD

or, Q < 15D @ 10°C, with Q in lph and D in mm

cool warm hot

short
long

longest
• In black PE lateral hose, sunlight warms the water significantly as the velocity
slows down, and water viscosity decreases
• Long-path emitters would ideally be progressively longer along the lateral to
compensate and provide a more uniform discharge along the lateral

III. Tortuous- and Short-Path Emitters

• Tortuous-path emitters also have long paths, but not laminar flow. This is
because the path has many sharp bends, and is in the form of a maze
• Tortuous-path emitters tend to behave hydraulically like orifices, and so do
many short-path emitters
• Flow rate is nearly independent of the viscosity, at least over typical ranges
in viscosity
• Many short-path emitters have pressure compensating features

Merkley & Allen Page 226 Sprinkle & Trickle Irrigation Lectures
IV. Orifice Emitters

• Many drip emitters and sprayers behave as orifices

• The orifice(s) are designed to dissipate energy and reduce the flow rate to an
acceptable value
• Flow rate is approximately proportional to the square root of the pressure

V. Line Source Tubing

• Single-chamber tubing provides less uniformity than dual-chamber tubing

• In dual-chamber tubing, much of the head loss occurs through the orifices
between the two chambers. The outer chamber is somewhat analogous to a
manifold or header.
• The flow rate equation for dual-chamber tubing can be expressed as:

2gHno2
q = a'K
( )
(374)
1 + no2

where a’ is the area of the outer orifice; K is an empirical coefficient; H is the

pressure head; and no is the number of outer orifices per inner orifice (no >
1.0)

• See Fig. 20.2 in the textbook

VI. Vortex and Sprayer Emitters

• Vortex emitters have a whirlpool effect in which the water must exit through
the center of the whirlpool
• Energy is dissipated by the friction from spinning in a chamber, and from
exiting through an orifice in the center
• As mentioned in a previous lecture, the exponent on the pressure head is
approximately equal to 0.4 (in the discharge equation). Thus, these can
usually be considered to be (partially) pressure compensating

VII.Pressure Compensating Emitters

• Pressure compensating emitters usually have some flexible or moving parts

• These types of emitters tend to need replacement or repair more often than
most of the simpler emitter designs, therefore incurring higher maintenance
cost
• Figure 20.3 of the textbook shows one design approach for a pressure
compensating emitter

Sprinkle & Trickle Irrigation Lectures Page 227 Merkley & Allen
 • As defined previously, pressure compensating emitters always have a
pressure head exponent of less than 0.5 (otherwise they aren’t considered to
be pressure compensating)

VIII.Self-Flushing Emitters

• In this category there are continuous and periodic flushing emitters

• Periodic flushing emitters perform their self cleaning when the lateral is filled
(before it reaches full operating pressure), and when the lateral is emptied.
In other words, they typically flush once per day.
• Continuous-flushing emitters have flexible parts that can stretch to allow solid
particles to pass through
• Fig. 20.4 in the textbook shows an example of one such design
• These can be sensitive to temperature changes and are not normally
pressure compensating

IX. Calculating the Discharge Exponent

• You can calculate the exponent, x, based on a pair of measured flow rates
and pressure heads
• Recall a rule of logarithms: log (ax) = x (log a)
• The solution can be obtained graphically, but is more quickly accomplished
with calculators and electronic spreadsheets
• If you have more than two pairs of q and H, then you can take the logarithmic
transformation of the equation and perform linear regression; however, the
regression will be mathematically biased toward the smaller values

Design Approach & Example

I. Review of Example Designs

• We will review example designs in Chapter 21 of the textbook, and discuss

design alternatives and parameters affecting efficiency, etc

II. Summarized Trickle Irrigation Design Process

• These are 15 basic steps, following the material presented in Chapters 17-24
of the textbook, that can be followed for the design of many trickle systems
• These are basic steps and represent a summary of the generalized design
process, but remember that each design situation will have some unique
features

1. Collect data on the crop, climate, soil, topography, and irrigation water
quality, field shape & size, water availability.

Merkley & Allen Page 228 Sprinkle & Trickle Irrigation Lectures
 2. Select an emitter and determine an emission point layout such that 33% <
Pw < 67%. This will determine the number of emitters per plant, Np.
Emitter selection may involve field testing to determine the wetted width
(or diameter), w.

3. Calculate dx, fx, and Td. Note that fx will almost always be greater than
1.0.

4. Select a target value for EU (usually 70-95%; see Table 20.3) and
estimate the peak-use transmission ratio, Tr (usually 1.00-1.10; see Table
19.3).

5. Calculate the leaching requirement, LRt, based on crop type and irrigation
water quality.

6. Let f = 1 day (usually), then dn = Td. Calculate the gross application

depth, d.

7. Calculate the gross volume of water required per plant per day, G.

⎛ dS pS r ⎞
G = K⎜ ⎟ (375)
⎝ f ⎠

8. Calculate the daily hours of operation, Ta, (per station, or subunit) during
the peak-use period.

G
Ta = (376)
N pq a

9. Determine the number of operation stations based on Ta (with more

stations, the system capacity is lower).

If Ta = 24 hrs, then Ns = 1
If Ta = 12 hrs, then Ns = 1 or 2
If Ta = 8 hrs, then Ns = 2 or 3, and so on

10. Adjust Np and qa so that TaNs is equal to, or slightly less than, 90%(24
hrs/day) = 21.6 hrs/day. First, try adjusting qa because this is usually less
expensive than increasing Np. If the emitter is pressure compensating, or
if qa must be greatly altered, you may need to change Np (or you may
need to select a different emitter).

11. Having determined the value of qa, calculate the minimum allowable
emitter discharge, qn

Sprinkle & Trickle Irrigation Lectures Page 229 Merkley & Allen
 q a EU
qn = (377)
100 (1.0 − 1.27 νs )

Note that if EU is high and νs is high, it could be that qn > qa (but this
would not be a reasonable calculation result!)

12. Calculate the average (nominal) and minimum lateral pressure heads

1/ x
⎛ q ⎞
h=⎜ ⎟ (378)
⎝ Kd ⎠
1/ x
⎛q ⎞
hn = ha ⎜ n ⎟ (379)
⎝ qa ⎠

13. Calculate the allowable change in pressure head in an operating station

∆Hs = 2.5 (ha − hn ) (380)

14. Calculate Qs, Vs, and Ot.

15. Finally, size the laterals, headers, manifolds and mainline(s) according to
hydraulic design criteria.

Merkley & Allen Page 230 Sprinkle & Trickle Irrigation Lectures
Lecture 21
Pipe Specifications & Trickle Lateral Design

I. Plastic Pipe Specifications

• Trickle and sprinkle irrigation systems are

commonly built with plastic pipe, of which there
are various types and specifications
• It is important to understand how the technical
specifications affect design decisions (pipe
sizing)
• Standards for the design and operation of
pipelines are available from various professional
organizations such as ASAE (American Society
of Agricultural Engineers) and AWWA (American
Water Works Association)
• Some of the material below is taken from ASAE
standard S376.1 OCT92
• ASAE standard S435 pertains to the use of PE pipe for microirrigation
laterals

• Plastic pipe is now commonly used in irrigation and other pipelines

• Some of the most common types are PVC (polyvinyl chloride), ABS
(acrylonitrile-butadiene-styrene), and PE (polyethylene)
• PVC pipes are usually white, while ABS and PE are usually black
• ABS pipes are often used for buried drains and drainage pipes

• All of these pipe materials are called “thermoplastic” because the material
can be repeatedly softened by increasing the temperature, and hardened by
a decrease in temperature
• The pressure rating of plastic pipe (especially PVC) decreases rapidly with
increasing temperature of the pipe and or water
• For example, at about 43°C (109°F) the PVC pressure rating
drops to one-half of the nominal value at 23°C (73°F), and
almost the same amount for PE
• PE pipe temperature can easily reach 43°C on a sunny day

• Unlike most metal pipes, these plastic pipe materials are immune to almost
all types of corrosion, whether chemical or electrochemical
• The resistance to corrosion is a significant benefit when chemigation is
practiced in a pressurized irrigation system

• The dimension ratio (DR) of a plastic pipe is the ratio of average diameter (ID
or OD) to wall thickness
• PVC, ABS and some PE are OD-based, while other PE pipe is ID-based
Sprinkle & Trickle Irrigation Lectures Page 231 Merkley & Allen
 • Plastic pipe is currently manufactured up to a maximum diameter of 54
inches
• There are several standard dimension thickness
ratios (SDR) for several values, each with
its own pressure rating (at 23°C)
• Different types of PVC, ABS and PE
compounds exist, some of which are
stronger than others diameter
• Some plastic pipe is manufactured with
non-standard dimension ratios; in these
cases the ratio is called “DR” rather than
“SDR”
• Some pipe sizes are correspond to iron
pipe size (IPS), plastic irrigation pipe (PIP), and others
• These are different standards for indirectly specifying pipe dimension ratios
and pressure ratings

• The relationship between SDR, hydrostatic design stress (S in psi), and

pressure rating (PR in psi) for OD-based pipe is defined by ISO standard
161/1-1978
• The pressure rating (PR), which is the maximum recommended operating
pressure, is determined by the following equations:

2S 2S
PR = = (OD-based) (381)
SDR − 1 ⎛ OD ⎞
⎜ t − 1⎟
⎝ ⎠

2S 2S
PR = = (ID-based) (382)
SDR + 1 ⎛ ID ⎞
⎜ t + 1⎟
⎝ ⎠

where t is the pipe wall thickness

• Values of S can be obtained from published tables, as can values of PR for

given SDR and pipe material (plastic compound)
• Values of S vary from 6900 to 13,800 kPa for PVC, and from 3400 to 5500
kPa for PE

• Common terms used in the industry for PVC pipe include Class 160, Class
200, Schedule 40, Schedule 80 and Schedule 120 (in increasing strength
and decreasing SDR)
• With the “schedule” classification, the higher the schedule, the thicker the
walls, for a given nominal pipe diameter
Merkley & Allen Page 232 Sprinkle & Trickle Irrigation Lectures
 • The maximum allowable operating pressure is approximately equal to:

(schedule)SE
P= (383)
1000

where P is the operating pressure (psi); S is the allowable stress in the pipe
material (psi); E is the “joint efficiency”; and “schedule” is the schedule
number (e.g. 40, 80, 120, etc.)

• Joint efficiency (or “joint quality factor”) for PVC is approximately 1.00, due to
the fact that it is seamless

• Class 160 and 200 refer to 160 psi and 200 psi ratings, respectively
• The Schedule 40 and 80 specifications have carried over from classifications
used in iron pipes
• Schedule 80 is seldom used in irrigation because its pressure rating is much
higher than the maximum pressures found in most irrigation systems
• Schedule 40 is commonly used in irrigation

• Some specifications for the design and protection of pipelines depend on

whether the pressure is “low” or “high”
• Low pressure pipelines are generally considered to have operating pressures
less than about 80 psi
• The maximum working pressure in a plastic pipe should normally be about
70% of the pipe’s pressure rating, unless special care is taken in design and
operation such that surges and excessive pressure fluctuations will not occur

• Manufacturers and testing centers provide data on minimum bursting

pressures
• Depending on the SDR value, the minimum burst pressure for plastic pipes
should be between about 900 and 1800 kPa (130 and 260 psi), otherwise the
pipe does not meet standard specifications
• Below is a glossary of common pipe abbreviations and terms:

Sprinkle & Trickle Irrigation Lectures Page 233 Merkley & Allen
 Abbreviation Meaning
ABS Acrylonitrile-Butadiene-Styrene
DR Dimension Ratio
ID Inside Diameter
IPS Iron Pipe Size
ISO International Organization for Standardization
OD Outside Diameter
PE Polyethylene
PIP Plastic Irrigation Pipe
PR Pressure Rating
PVC Polyvinyl Chloride
SDR Standard Dimension Ratio

II. Trickle Irrigation Laterals

• Laterals are often above ground, but may be buried

• Drip “tape”, single- and dual-chamber laterals are usually buried a few
centimeters below the ground surface
• Above-ground laterals may be on the ground surface, or suspended above
the surface (e.g. in vineyards)
• Black polyethylene (PE) plastic pipe (or “hose”) is usually used for trickle
irrigation laterals
• Lateral pipes are typically about 0.5 or 1.0 inches in diameter
• Standard PE sizes are usually ID based and come in standard dimension
ratio (SDR) values of 15, 11.5, 9, 7 and 5.3
• Nominal PE pipe sizes for laterals are ½-inch, ¾-inch, 1-inch, and 1¼-inch
(all iron pipe size, or IPS)

• Laterals are usually single-diameter, but can be dual-sized

• Dual-sized lateral hydraulic analysis is essentially the same as previously
discussed for dual-sized sprinkler laterals

• To start a new system design, Keller & Bliesner recommend limiting the
lateral pressure variation to 0.5∆Hs, where ∆Hs is calculated from Eq. 20.14
• Then, 0.5∆Hs remains for the manifolds (if manifolds are subunits, or
“stations”)

• In lateral designs, the pipe diameter is usually chosen (not calculated), and if
the pressure variation or loss is “out of range”, then a different size can be
selected
• There are usually only a few lateral diameters to choose from

Merkley & Allen Page 234 Sprinkle & Trickle Irrigation Lectures
III. Trickle Lateral Hydraulics

• Friction loss gradients in plastic lateral pipe can be approximated by

combining the Darcy-Weisbach and Blasius equations (Eq. 8.7a):

7 Q1.75
J = 7.83(10) (384)
D4.75
for J in m/100 m; Q in lps; and D in mm

• The Blasius equation estimates the D-W “f” factor for smooth pipes
• If you want to calculate relative roughness, use ε = 1.5(10)-6 m

• It may be necessary to increase the J value because of emitter losses within

the lateral hose (barb, etc.) (see Fig. 20.8)
• Equation 22.1 is:

⎛S +f ⎞
J' = J ⎜ e e ⎟ (385)
⎝ Se ⎠
where fe is an equivalent length of lateral hose for each emitter, spaced
evenly at a distance of Se

• The fe pipe length is one way that minor hydraulic losses are calculated in
pipes
• From Eq. 8.7a, a dimensionless friction loss equation can be developed (see
Fig. 8.2), which is useful in semi-graphical hydraulic design work for trickle
irrigation laterals
• This is discussed in detail in the following lectures

• For a given lateral pipe size, lateral length, emitter spacing, and nominal
discharge per emitter, the lateral inlet pressure must be determined such that
the average lateral pressure is “correct”
• Then, the manifold can be designed to provide this lateral inlet pressure with
as little variation (with distance) as possible

• Figure 22.1 shows four different hydraulic cases for single lateral designs
• The design of pairs of laterals is essentially a compound single lateral
problem, with the added criterion that the minimum pressure be the same in
both laterals

• Not including riser height, the required lateral inlet pressure is (Eq. 22.6):

H l = H a + kh f + 0.5 ∆he (386)

Sprinkle & Trickle Irrigation Lectures Page 235 Merkley & Allen
 where k is 0.75 for single pipe size laterals, or 0.63 for dual pipe size laterals
(as in the design of sprinkler laterals); and ∆he is positive for laterals running
uphill

• The minimum pressure in a lateral is given by Eq. 22.7:

H'n = Hl − (h f + ∆he ) − ∆H c
(387)
H'n = H c − ∆H c

where Hc is the pressure head at the closed end of the lateral

IV. References (plastic pipe)

http://www.uni-bell.org/lit.cfm
http://www.dpcpipe.com/ag/pipirrig.html
ASAE Standards (1997). American Soc. of Agric. Engineers, St. Joseph, MI.
Handbook of PVC Pipe. (1979). Uni-Bell Plastic Pipe Association, Dallas, TX.

Trickle Manifold Location

I. Optimal Manifold Location

• If the ground slope along the direction of the laterals is less than 3% or so, it
is usually recommendable to run laterals off both sides (uphill and downhill)
of each manifold
• If the ground slope along the direction of the laterals is more than 3%, it may
be best to run the laterals only in the downhill direction
• The design objective for a pair of laterals is to have equal values of minimum
pressure, Hn’, in uphill and downhill laterals
• This means that the downhill lateral will always be longer for laterals of equal
pipe size on sloping ground
• The manifold should be located in-between rows of plants (trees), not over a
row
• For laterals on flat ground, the manifold goes in the center of the field (the
trivial solution)

II. Sample Graphical Solution for Manifold Location

• Use the dimensionless friction loss curves (Fig. 8.2) to locate the optimal
manifold position in a sloping field
• The laterals run along the 0.021 m/m slope
• The combined uphill + downhill lateral length is 315 m

Merkley & Allen Page 236 Sprinkle & Trickle Irrigation Lectures
 • The spacing of plants (trees) is
Sp = 4.5 m

• The spacing of emitters is Se =

1.5 m (thus, Np = 3 for single
line)
• The equivalent emitter loss is
fe = 0.12 m
• The nominal emitter discharge
is qa = 3.5 lph at 10 m head
(68.9 kPa)
• The lateral pipe ID is 14.7 mm

Solution for Manifold Location:

1. Number of emitters for the pair of laterals is:

315 m
= 210 emitters (388)
1.5 m/emitter
2. Total nominal discharge for the pair of laterals is:

Qpair =
( 210 emitters )( 3.5 lph/emitter ) = 12.25 lpm (389)
( 60 min/hr )
3. From Table 8.2 (page 141), J ≅ 13.3 m/100m. The adjusted J is:

⎛ S e + fe ⎞ ⎛ 1.5 + 0.12 ⎞
J' = J ⎜ ⎟ = 13.3 ⎜ ⎟ = 14.4 m/100 m (390)
⎜ Se ⎟ ⎝ 1.5 ⎠
⎝ ⎠
4. Multiple outlet factor, F = 0.36 for 210 outlets

5. Friction loss for the pair of laterals:

J'FL (14.4 )( 0.36 )( 315 )

(hf )pair = = = 16.3 m (391)
100 100
6. Elevation change for the pair of laterals:

( ∆he )pair = ( 315 m)( 0.021) = 6.62 m (392)

Sprinkle & Trickle Irrigation Lectures Page 237 Merkley & Allen
 7. Ratio of elevation change to friction loss for the pair:

⎛ ∆he ⎞ 6.62
⎜ ⎟ = = 0.41 (393)
⎝ f ⎠pair
h 16.3

8. From the nondimensional graphical solution (Fig. 8.2): x/L = 0.69. Then, x
= (0.69)(315 m) = 217 m. Look at the figure below:

How was this done?

• Looking at the above figure, a straight line was drawn from the origin
(0, 0) to (1.0, 0.41), where 0.41 is the ratio calculated above
• The nondimensional curve was overlapped and shifted vertically so
that the curve was tangent to the same straight line, then traced onto
the graph
• The nondimensional curve was then shifted vertically even more so
that the inverse half-curve (dashed) intersected the (1.0, 0.41) point,
also tracing it onto the graph
• The intersection of the two traced curve segments gave an abscissa
value of about 0.69, which is the distance ratio

Merkley & Allen Page 238 Sprinkle & Trickle Irrigation Lectures
 9. Finally, adjust x for tree spacing,

(217 m)/(4.5 m/tree) = 48.2 trees

• Therefore, round to 48 trees

• Then, x = (48 trees)(4.5 m/tree) = 216 m
• This way, the manifold lays buried halfway between two rows of trees,
not on top of a row

• This manifold position give the same minimum pressure in both the uphill
and downhill laterals
• Minimum pressure in the downhill lateral is located approximately
(0.35)(315 m) = 110 m from the closed end, or 216 - 110 = 106 m from
the manifold.

• This graphical solution could have been obtained numerically

• But the graphical solution is useful because it is didactic
• If you like computer programming, you can set up the equations to solve
for the lateral hydraulics based on non-uniform emitter discharge (due to
pressure variations in the laterals), non-uniform ground slope, etc.

• Note that this procedure could also be used for sprinklers, but it would
probably only be feasible for solid-set, fixed systems

Sprinkle & Trickle Irrigation Lectures Page 239 Merkley & Allen
Merkley & Allen Page 240 Sprinkle & Trickle Irrigation Lectures
Lecture 22
Numerical Solution for Manifold Location

I. Introduction

• In the previous lecture it was seen how the optimal manifold location can be
determined semi-graphically using a set of non-dimensional curves for the
uphill and downhill laterals
• This location can also be determined numerically
• In the following, equations are developed to solve for the unknown length of
the uphill lateral, xu, without resorting to a graphical solution

II. Definition of Minimum Lateral Head

• In the uphill lateral, the minimum head is at the closed end of the lateral
(furthest uphill location in the subunit)
• This minimum head is equal to:

hn ' = hl − hfu − xuS (394)

where hn’ is the minimum head (m); hl is the lateral inlet head (m); hfu is the
total friction loss in the uphill lateral (m); xu is the length of the uphill lateral
(m); and S is the slope of the ground surface (m/m)

• Note that S must be a positive value

manifold
location

(hfd)1 hfu
(hfd)2
xm xu

h n’ h n’
• In the downhill lateral, the minimum head may be anywhere from the inlet to
the outlet, depending on the lateral hydraulics and the ground slope
• The minimum head in the downhill lateral is equal to:

hn ' = hl − (hfd )1 + (hfd )2 + x mS (395)

Sprinkle & Trickle Irrigation Lectures Page 241 Merkley & Allen
 where (hfu)1 is the total friction loss in the downhill lateral (m); (hfu)2 is the
friction loss from the closed end of the downhill lateral to the location of
minimum head (m); and xm is the distance from the manifold (lateral inlet) to
the location of minimum head in the downhill lateral (m)

• Combining Eqs. 1 and 2:

hfu + S ( xu + xm ) − ( hfd )1 + ( hfd )2 = 0

(396)

III. Location of Minimum Head in Downhill Lateral

• The location of minimum head is where the slope of the ground surface, S,
equals the friction loss gradient, J’:

S = J' (397)

where both S and J’ are in m/m, and S is positive (you can take the absolute
value of S)

• Using the Hazen-Williams equation, the friction loss gradient in the downhill
lateral (at the location where S = J’) is:

⎡ Se + fe ⎤ ⎡ ⎛ qa (L − xu − xm ) ⎞ ⎤
1.852
⎢ 10 −4.87 ⎥
J' = ⎢ ⎥ 1.212(10) ⎜ ⎟ D (398)
⎣ Se ⎦ ⎢⎣ ⎝ 3,600Se C ⎠ ⎥
⎦

where: J’ is the friction loss gradient (m/m);

Se is the emitter spacing on the laterals (m);
fe is the equivalent lateral length for emitter head loss (m);
qa is the nominal emitter discharge (lph);
L is the sum of the lengths of the uphill and downhill laterals (m);
xu is the length of the uphill lateral (m);
xm is the distance from the manifold to the location of minimum head
in
the downhill lateral (m);
C is approximately equal to 150 for plastic pipe; and
D is the lateral inside diameter (mm);

• The value of 3,600 is to convert qa units from lph to lps

• Note that xd = L - xu, where xd is the length of the downhill lateral
• Note that qa(L - xu - xm)/(3,600 Se) is the flow rate in the lateral, in lps, at the
location of minimum head, xm meters downhill from the manifold
• Combining the above two equations, and solving for xm:
Merkley & Allen Page 242 Sprinkle & Trickle Irrigation Lectures
 ⎛ 0.0129SeCD2.63 ⎞ ⎛ SeS ⎞0.54
xm = L − xu − ⎜ ⎟⎟ ⎜ ⎟ (399)
⎜ q ⎝ S + f ⎠
⎝ a ⎠ e e

where the permissible values of xm are: 0 ≤ xm ≤ xd

• Combine Eqs. 396 & 399, and solve for xu by iteration

• Alternatively, based on Eq. 8.7a from the textbook, xm can be defined as:

0.571
⎡ 3,600Se ⎤ ⎡⎛ SD4.75 ⎞ ⎛ Se ⎞ ⎤
xm = L − xu − ⎢ ⎥ ⎢⎜⎜ 5 ⎟⎟ ⎜ S + f ⎟
⎥ (400)
⎣ qa ⎦ ⎢⎣⎝ 7.89(10) ⎠ ⎝ e e ⎠ ⎥⎦

IV. Definition of Head Loss Gradients

• In the uphill lateral, the head loss is:

hfu = Ju 'Fu xu (401)

• In the downhill lateral, the head losses are:

(hfd )1 = Jd1 'Fd1 (L − xu ) (402)

and,
(hfd )2 = Jd2 'Fd2 (L − xu − xm ) (403)

• The above three “F” values are as defined by Eq. 8.9 in the textbook
• The friction loss gradients (in m/m) are:

Ju ' = K J ( xu )
1.852
(404)

Jd1 ' = K J (L − xu )
1.852
(405)

Jd2 ' = K J (L − xu − xm )
1.852
(406)

where, for the Hazen-Williams equation,

Sprinkle & Trickle Irrigation Lectures Page 243 Merkley & Allen
 ⎛ Se + fe ⎞ ⎡
1.852 ⎤
− ⎛ q ⎞
⎟ ⎢1.212(10) D ⎥
10 4.87
KJ = ⎜ ⎜
a
⎟ (407)
⎝ S e ⎠ ⎢ ⎝ 3,600S C
e ⎠ ⎥
⎣ ⎦

V. Solving for Optimal Manifold Location

• Using the definitions above, solve for the length of the uphill lateral, xu
• Then, xd = L - xu
• Note that you might prefer to use the Darcy-Weisbach and Blasius equations
for the manifold calculations; they may be more accurate than Hazen-
Williams
• The “OptManifold” computer program uses the Darcy-Weisbach & Blasius
equations

Merkley & Allen Page 244 Sprinkle & Trickle Irrigation Lectures
Where do these Equations Come From?

I. Derivation of Nondimensional Friction Loss Curves

• The nondimensional friction loss curves are

actually one curve, with the lower half
laterally inverted and shown as a dashed line
(Fig. 8.2)
• The dashed line is simply for flow in the
opposite direction, which for our purposes is
in the uphill direction
• We know from the previous lectures and
from intuition that the uphill segment of
lateral pipe will not be more than ½ the total
length, because it is equal to ½ for the case
where the ground slope is zero

• Following is the derivation for Eq. 8.10b, from

which Fig. 8.2 was plotted

• Darcy-Weisbach equation for circular pipes:

L V2
hf = f (408)
D 2g

• Blasius equation, for estimating f for small diameter (D < 125 mm) “smooth
pipes” (e.g. PE & PVC), and based on more complete equations that are
used to plot the Moody diagram

−0.25
f ≅ 0.32NR (409)

where NR is the Reynolds number, which for circular pipes is:

VD 4Q
NR = = (410)
ν νπD
• The kinematic viscosity, ν, is equal to about 1.003(10)-6 m2/s for water at
20°C
• Then, for this kinematic viscosity,

−0.25 −0.25
⎛ 4Q ⎞ ⎛Q⎞
f ≅ 0.32 ⎜ ⎟ ≈ 0.0095 ⎜ ⎟ (411)
⎝ νπD ⎠ ⎝D⎠

Sprinkle & Trickle Irrigation Lectures Page 245 Merkley & Allen
 • Putting the above into the Darcy-Weisbach equation:

−0.25
⎛Q⎞ L V2
hf = 0.0095 ⎜ ⎟ (412)
⎝D⎠ D 2g
or:
Q1.75
hf ≅ 0.00079L (413)
D4.75
where hf is in m; L is in m; Q is in m3/s; and D is in m

• Eq. 8.7a is obtained by having Q in lps, and D in mm, whereby the above
coefficient changes to 7.9(10)7

• Finally, in the above, use L(x/L) instead of L, and Q(x/L) instead of Q, and
call it “hfx”:
⎡Q ( x / L ) ⎤⎦
1.75
hfx ≅ 0.00079L(x / L) ⎣ (414)
D4.75
• Then,
hfx
= (x / L) ( x / L ) = (x / L)
1.75 2.75
(415)
hf

which is Eq. 8.10b and the basis for the nondimensional friction loss curves,
valid for plastic pipes with D < 125 mm

II. Derivation of Equation for ∆Hc

• The difference between the minimum pressure head and the pressure head
at the closed end of a lateral, ∆Hc, is used to calculate the minimum head in
the lateral, Hn’
• This is because the pressure head at the end of the lateral is easily
calculated as:

H c = H l − h f − ∆he (416)

where ∆he is negative for downhill slopes

• But the minimum pressure head does not necessarily occur at the end of the
lateral when the lateral runs downhill
• Thus, in general,

Merkley & Allen Page 246 Sprinkle & Trickle Irrigation Lectures
 H'n = H c − ∆H c (417)

• The above is from Eq. 22.7 in the textbook

• These concepts can also be interpreted graphically as in Fig. 22.1
• Following is a derivation of an equation for ∆Hc (based on Keller and Rodrigo
1979)

1. The minimum pressure in the lateral occurs where the ground slope (for a
uniform slope) equals the slope of the friction loss curve. The dimensionless
friction loss curve is defined as (Eq. 8.10b or Eq. 22.3b):

2.75
⎛ hfx ⎞ ⎛x⎞
⎜ ⎟ =⎜ ⎟ (418)
⎝ hf ⎠pair ⎝ L ⎠

2. The slope of this friction loss curve is:

⎛h ⎞
d ⎜ fx ⎟
1.75
⎝ hf ⎠pair ⎛x⎞
= 2.75 ⎜ ⎟ (419)
⎛x⎞ ⎝L⎠
d⎜ ⎟
⎝L⎠

3. The uniform ground slope on the dimensionless graph is:

⎛ ∆he ⎞ SL 100S
⎜ ⎟ = = (420)
⎝ hf ⎠pair ⎛ J'FL ⎞ J 'F
⎜ 100 ⎟
⎝ ⎠
4. Then,
100S
= 2.75 ( y )
1.75
(421)
J'F

in which y is the value of x/L where the minimum pressure occurs (0 ≤ y ≤ 1);
S is the ground slope (m/m); J’ is the friction loss gradient for the flow rate in
the pair of laterals (m/100 m); and F is the reduction coefficient for multiple
outlet pipes (usually about 0.36)

Sprinkle & Trickle Irrigation Lectures Page 247 Merkley & Allen
 5. Solve for y:
1/ 1.75
⎛ 100S ⎞
y=⎜ ⎟ (422)
⎝ 2.75 J'F ⎠
or,
1/ 1.75
⎛ 100S ⎞
y≈⎜ ⎟ (423)
⎝ J' ⎠
where F ≈ 0.36

Merkley & Allen Page 248 Sprinkle & Trickle Irrigation Lectures
 6. Referring to the figure on the previous page, the following equality can be
written:
∆Hc 100 y S
y 2.75 + = (424)
(hf )pair J'F

solving for ∆Hc,

⎛ 100 y S ⎞
∆Hc = (hf )pair ⎜ − y 2.75 ⎟ (425)
⎝ J'F ⎠

where,
J'FL
(hf )pair = (426)
100
and y can be approximated as in step 5 above (for F = 0.36)

7. After manipulating the equation a bit, the following expression is obtained:

−0.57
∆Hc = 8.9LS1.57 ( J' ) (427)

for ∆Hc in m; L in m; S in m/m; and J’ in m/100 m. Note that J’ and L are for
the pair of laterals, not only uphill or only downhill

III. Derivation of Equation for α

• The parameter α is used in the calculation of inlet pressure for a pair of

laterals on sloping ground where (Eq. 22.17):

⎛x ⎞
Hl = Ha + α (hf )pair + ⎜ − 0.5 ⎟ ( ∆he )pair (428)
⎝L ⎠

with,
1/ x
⎛q ⎞
Ha = ⎜ a ⎟ (429)
⎝ Kd ⎠

J'FL
(hf )pair = (430)
100

Sprinkle & Trickle Irrigation Lectures Page 249 Merkley & Allen
 100S
( ∆he )pair = (hf )pair = SL (431)
J'F

• Note that (∆he)pair must be a negative number

• The ratio x/L is the distance to the manifold, where L is the length of the pair
of laterals
• The following derivation is based on equations presented by Keller and
Rodrigo (1979):

1. Given that for a single lateral approximately ¾ of the friction loss occurs
from the inlet to the point where the average pressure occurs (multiple
outlets, uniform outlet spacing, constant discharge from outlets, single
lateral pipe size) we have the following:

(hf )downhill ⎛⎜ ⎞⎟ + (hf )uphill ⎛⎜ 1 − ⎞⎟

3 x 3 x
α (hf )pair = (432)
4 ⎝L⎠ 4 ⎝ L⎠

The above equation is a weighted average because the uphill lateral is

shorter than the downhill lateral

2. Recall that,
2.75
⎛ hfx ⎞ ⎛x⎞
⎜ ⎟ =⎜ ⎟ (433)
⎝ hf ⎠pair ⎝ L ⎠

Then,
2.75
⎛x⎞
(hf )downhill =⎜ ⎟ (hf )pair
⎝L⎠
(434)
2.75
(hf )uphill = ⎛⎜ 1 −
x⎞
(hf )pair
⎝ L ⎟⎠

3. Combining equations:

3 ⎡⎛ x ⎞ ⎛ x ⎞2.75 ⎛ x ⎞⎛ x⎞
2.75 ⎤
α (hf )pair = (hf )pair ⎢⎜ ⎟ ⎜ ⎟ + ⎜1− ⎟ ⎜1− ⎟ ⎥ (435)
4 ⎢⎣⎝ ⎠ ⎝ ⎠
L L ⎝ L ⎠⎝ L ⎠ ⎥⎦

3 ⎡⎛ x ⎞
3.75 3.75 ⎤
⎛ x⎞
α = ⎢⎜ ⎟ + ⎜1− ⎟ ⎥ (436)
4 ⎢⎝ L ⎠ ⎝ L⎠ ⎥⎦
⎣
Merkley & Allen Page 250 Sprinkle & Trickle Irrigation Lectures
 • This last equation for α is Eq. 22.25 from the textbook
• See the figure below

Sprinkle & Trickle Irrigation Lectures Page 251 Merkley & Allen
Merkley & Allen Page 252 Sprinkle & Trickle Irrigation Lectures
Lecture 23
Manifold Hydraulic Design

I. Introduction

• Manifolds in trickle irrigation systems often have multiple pipe sizes to:

1. reduce pipe costs

2. reduce pressure variations

• In small irrigation systems the reduction in pipe cost may not be significant,
not to mention that it is also easier to install a system with fewer pipe sizes
• Manifold design is normally subsequent to lateral design, but it can be part of
an iterative process (i.e. design the laterals, design the manifold, adjust the
lateral design, etc.)
• The allowable head variation in the manifold, for manifolds as subunits, is
given by the allowable subunit head variation (Eq. 20.14) and the calculated
lateral head variation, ∆Hl
• This simple relationship is given in Eq. 23.1:

( ∆Hm )a = ∆Hs − ∆Hl (437)

• Eq. 23.1 simply says that the allowable subunit head variation is shared by
the laterals and manifold
• Recall that a starting design point can be to have ∆Hl = ½∆Hs, and ∆Hm =
½∆Hs, but this half and half proportion can be adjusted during the design
iterations
• The lateral pressure variation, ∆Hl, is equal to the maximum pressure minus
the minimum pressure, which is true for single-direction laterals and
uphill+downhill pairs, if Hn’ is the same both uphill and downhill

II. Allowable Head Variation

• Equation 20.14 (page 502 in the textbook) gives the allowable pressure head
variation in a “subunit”
• This equation is an approximation of the true allowable head variation,
because this equation is applied before the laterals and manifold are
designed
• After designing the laterals and manifold,
the actual head variation and expected
EU can be recalculated

• Consider a linear friction loss gradient (no

multiple outlets) on flat ground:
In this case, the average head is
Sprinkle & Trickle Irrigation Lectures Page 253 Merkley & Allen
 equal to Hn plus half the difference in the maximum and minimum heads:

(
H max − H n = 2 H a − H n ) (438)

• Consider a sloping friction loss gradient (multiple outlets) on flat ground:

In this case, the average head occurs

after about ¾ of the total head loss
(due to friction) occurs, beginning
from the lateral inlet. Then,

(
H max − H n = 4 H a − H n ) (439)

• For a sloping friction loss gradient (multiple outlets) on flat ground with dual
pipe sizes, about 63% of the friction head loss occurs from the lateral inlet to
the location of average pressure. Then 100/(100-63) = 2.7 and,

(
H max − H n = 2.7 H a − H n ) (440)

• In summary, an averaging is performed to skew the coefficient toward the

minimum value of 2, recognizing that the maximum is about 4, and that for
dual-size laterals (or manifolds), the coefficient might be approximately 2.7
• The value of 2.5 used in Eq. 20.14 is such a weighted average
• With three or four pipe sizes the friction loss gradient in the manifold will
approach the slope of the ground, which may be linear
• Thus, as an initial estimate for determining allowable subunit pressure
variation for a given design value of EU, Eq. 20.14 is written as follows:

(
∆Hs = 2.5 H a − H n ) (441)

• After the design process, the final value of ∆Hs may be different, but if it is
much different the deviation should be somehow justified

Merkley & Allen Page 254 Sprinkle & Trickle Irrigation Lectures
III. Pipe Sizing in Manifolds

• Ideally, a manifold design considers all of the following criteria:

1. economic balance between pipe cost (present) and pumping

costs (future)
2. allowable pressure variation in the manifold and subunit
3. pipe flow velocity limits (about 1.5 - 2.0 m/s)

• From sprinkler system design, we already know of various pipe sizing

methods
• These methods can also be applied to the design of manifolds
• However, the difference with trickle manifolds is that instead of one or two
pipe sizes, we may be using three or four sizes
• The manifold design procedures described in the textbook are:

1. Semi-graphical
2. Hydraulic grade line (HGL)
3. Economic pipe sizing (as in Chapter 8 of the textbook)

Semi-Graphical Design Procedure

• The graphical method uses “standard” head loss curves for different pipe
sizes and different flow rates with equally-spaced multiple outlets, each outlet
with the same discharge
• The curves all intersect at the origin (corresponding to the downstream
closed end of a pipe)
• Below is a sample of the kind of curves given in Fig. 23.2 of the textbook

• Instead of the standard curves, specific curves for each design case could be
custom developed and plotted as necessary in spreadsheets
• The steps to complete a graphical design are outlined in the textbook
• The graphical procedure is helpful in understanding the hydraulic design of
multiple pipe size manifolds, but may not be as expedient as fully numerical
procedures

Sprinkle & Trickle Irrigation Lectures Page 255 Merkley & Allen
 3.8
3.6
3.4
3.2
3.0
Friction Head Loss (ft)

2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.25 inch
1.0
1.50 inch
0.8
2.00 inch
0.6
2.50 inch
0.4 3.00 inch
0.2 4.00 inch
0.0
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320

Flow Rate (gpm)

• The following steps illustrate the graphical design procedure:

Step 1:
flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd
Merkley & Allen Page 256 Sprinkle & Trickle Irrigation Lectures
 Step 2:

flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd

Step 3:

flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd

Sprinkle & Trickle Irrigation Lectures Page 257 Merkley & Allen
 Step 4:

flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd

Step 5:

flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd

Merkley & Allen Page 258 Sprinkle & Trickle Irrigation Lectures
 Step 6:

flow direction

(∆Hm)a

So ∆Em
1
0 manifold flow rate Qm
xd
HGL Design Procedure

• The HGL procedure is very similar to the graphical procedure, except that it
is applied numerically, without the need for graphs
• Nevertheless, it is useful to graph the resulting hydraulic curves to check for
errors or infeasibilities
• The first (upstream) head loss curve starts from a fixed point: maximum
discharge in the manifold and upper limit on head variation
• Equations for friction loss curves of different pipe diameters are known (e.g.
Darcy-Weisbach, Hazen-Williams), and these can be equated to each other
to determine intersection points, that is, points at which the pipe size would
change in the manifold design
• But, before equating head loss equations, the curves must be vertically
shifted so they just intersect with the ground slope curve (or the tangent to
the first, upstream, curve, emanating from the origin)
• The vertical shifting can be done graphically or numerically

Economic Design Procedure

• The economic design procedure is essentially the same as that given in

Chapter 8 of the textbook

Sprinkle & Trickle Irrigation Lectures Page 259 Merkley & Allen
 • The manifold has multiple outlets (laterals or headers), and the “section flow
rate” changes between each outlet
• The “system flow rate” would be the flow rate entering the manifold

IV. Manifold Inlet Pressure Head

• After completing the manifold pipe sizing, the required manifold inlet pressure
head can be determined (Eq. 23.4):

Hm = Hl + k hf + 0.5∆Em (442)

where k = 0.75 for single-diameter manifolds; k = 0.63 for dual pipe size
laterals; or k ≈ 0.5 for three or more pipe sizes (tapered manifolds); and ∆El
is negative for downward-sloping manifolds

• As with lateral design, the friction loss curves must be shifted up to provide
for the required average pressure
• In the case of manifolds, we would like the average pressure to be equal to
the calculated lateral inlet head, Hl
• The parameter ∆El is the elevation difference along one portion of the
manifold (either uphill or downhill), with positive values for uphill slopes and
negative values for downhill slopes

V. Manifold Design

• Manifolds should usually extend both ways from the mainline to reduce the
system cost, provided that the ground slope in the direction of the manifolds
is less than about 3% (same as for laterals, as in the previous lectures)
• As shown in the sample layout (plan view) below, manifolds are typically
orthogonal to the mainline, and laterals are orthogonal to the manifolds

Merkley & Allen Page 260 Sprinkle & Trickle Irrigation Lectures
 • Manifolds usually are made up of 2 to 4 pipe diameters, tapered
(telescoping) down toward the downstream end
• For tapered manifolds, the smallest of the pipe diameters (at the downstream
end) should be greater than about ½ the largest diameter (at the upstream
end) to help avoid clogging during flushing of the manifold

D1 D2 D3>0.5D1

• The maximum average flow velocity in each pipe segment should be less
than about 2 m/s
• Water hammer is not much of a concern, primarily because the manifold has
multiple outlets (which rapidly attenuates a high- or low-pressure wave), but
the friction loss increases exponentially with flow velocity

VI. Trickle Mainline Location

• The objective is the same as for pairs of laterals: make (Hn)uphill equal to
(Hn)downhill
• If average friction loss slopes are equal for both uphill and downhill manifold
branches (assuming similar diameters will carry similar flow rates):

Downhill side:

( ∆Hm )a = hfd − ∆E ⎛⎜
x⎞
⎟ = hfd − Y∆E (443)
⎝ ⎠
L
Uphill side:
L−x⎞
( ∆Hm )a = hfu + ∆E ⎛⎜ ⎟ = hfu + (1 − Y)∆E (444)
⎝ L ⎠

where x is the length of downhill manifold (m or ft); L is the total length of the
manifold (m or ft); Y equals x/L; and ∆E is the absolute elevation difference of
the uphill and downhill portions of the manifold (m or ft)

• Note that in the above, ∆E is an absolute value (always positive)

• Then, the average uphill and downhill friction loss slopes are equal:

Juphill = Jdownhill
hfu h (445)
= fd
L−x x
where J-bar is the average friction loss gradient from the mainline to the end
of the manifold (J-bar is essentially the same as JF)
Sprinkle & Trickle Irrigation Lectures Page 261 Merkley & Allen
 Then,
hfd = Jx
(446)
hfu = J(L − x)
and,
( ∆Hm )a = Jx − Y∆E
(447)
( ∆Hm )a = J(L − x) + (1 − Y) ∆E
then,
( ∆Hm )a + Y∆E
=J
x (448)
( ∆Hm )a − (1 − Y) ∆E
=J
L−x
• Equating both J-bar values,

( ∆Hm )a + Y∆E ( ∆Hm )a − (1 − Y) ∆E

= (449)
x L−x
• Dividing by L and rearranging (to get Eq. 23.3),

( ∆Hm )a + Y∆E ( ∆Hm )a − (1 − Y) ∆E

= (450)
Y 1− Y
or,
∆E 2Y − 1
= (451)
( ∆Hm )a 2Y(1 − Y)
• Equation 23.3 is used to determine the lengths of the uphill and downhill
portions of the manifold
• You can solve for Y (and x), given ∆E and (∆Hm)a = ∆Hs - ∆Hl
• Remember that ∆Hs ≈ 2.5(Ha – Hn), where Ha is for the average emitter and
Hn is for the desired EU and νs
• Equation 23.3 can be solved by isolating one of the values for Y on the left
hand side, such that:
⎛ 2Y − 1 ⎞ ⎛ ( ∆Hm ) ⎞
Y = 1− ⎜ ⎟⎜ a
⎟⎟ (452)
⎜ 2Y ⎟ ⎜ ∆ E
⎝ ⎠⎝ ⎠
and assuming an initial value for Y (e.g. Y = 0.6), plugging it into the right
side of the equation, then iterating to arrive at a solution
Merkley & Allen Page 262 Sprinkle & Trickle Irrigation Lectures
 • Note that 0 ≤ Y ≤ 1, so the solution is already well-bracketed
• Note that in the trivial case where ∆E = 0, then Y = 0.5 (don’t apply the above
equation, just use your intuition!)
• A numerical method (e.g. Newton-Raphson) can also be used to solve the
equation for Y

VII. Selection of Manifold Pipe Sizes

The selection of manifold pipe sizes is a function of:

1. Economics, where pipe costs are balanced with energy costs

2. Balancing hf, ∆E, and (∆Hm)a to obtain the desired EU
3. Velocity constraints

VIII. Manifold Pipe Sizing by Economic Selection Method

• This method is similar to that used for mainlines of sprinkler systems

• Given the manifold spacing, Sm, and the manifold length, do the following:

(a) Construct an economic pipe size table where Qs = Qm

(b) Select appropriate pipe diameters and corresponding Q values at locations
where the diameters will change
(c) Determine the lengths of each diameter of pipe (where the Q in the manifold
section equals a breakeven Q from the Economic Pipe Size Table (EPST)

⎛ Qbeg − Qend ⎞
LD = L ⎜ ⎟ (453)
⎝ Q m ⎠

where Qbeg is the flow rate at the beginning of diameter “D” in the EPST (lps
or gpm); Qend is the flow rate at the end of diameter “D” in the EPST, which is
the breakeven flow rate of the next larger pipe size) (lps or gpm); L is the
total length of the manifold (m or ft); and Qm is the manifold inflow rate (lps or
gpm). (see Eq. 23.7)

(d) Determine the total friction loss along the manifold (see Eq. 23.8a):

FLK ⎛ Q1a Qa2 − Q1a Q3a − Qa2 Qa4 − Q3a ⎞

hf = ⎜ + + + ⎟ (454)
100Qm ⎜⎝ D1c Dc2 D3c Dc4 ⎟⎠

where,

a= b+1 (for the Blasius equation, a = 2.75)

c= 4.75 for the Blasius equation (as seen previously)
Sprinkle & Trickle Irrigation Lectures Page 263 Merkley & Allen
 Q1 = Q at the beginning of the smallest pipe diameter
Q2 = Q at the beginning of the next larger pipe diameter
Q3 = Q at the beginning of the third largest pipe diameter in the manifold
Q4 = Q at the beginning of the largest pipe diameter in the manifold
F= multiple outlet pipe loss factor

• For the Hazen-Williams equation, F equals 1/(1.852+1) = 0.35

• For the Darcy-Weisbach equation, F equals 1/(2+1) = 0.33

L= the total length of the manifold

D= inside diameter of the pipe
K= 7.89(10)7 for D in mm, Q in lps, and length in m
K= 0.133 for D in inches, Q in gpm, and length in ft
hf = friction head loss

• The above equation is for four pipe sizes; if there are less than four sizes, the
extra terms are eliminated from the equation
• An alternative would be to use Eq. 23.8b (for known pipe lengths), or
evaluate the friction loss using a computer program or a spreadsheet to
calculate the losses section by section along the manifold
• Eq. 23.8b is written for manifold design as follows:

a −1 ⎛ a
FK Qm x1 x a2 − x1a x 3a − x a2 x a4 − x 3a ⎞
hf = ⎜ + + + ⎟ (455)
100La −1 ⎜⎝ D1c Dc2 D3c Dc4 ⎟⎠

where, x1 = length of the smallest pipe size

x2 = length of the next smaller pipe size
x3 = length of the third largest pipe size
x4 = length of the largest pipe size

• Again, there may be up to four different pipe sizes in the manifold, but in
many cases there will be less than four sizes

(e) For s ≥ 0 (uphill branch of the manifold),

∆Hm = hf + S xu (456)

For s < 0 (downhill branch of the manifold),

⎛ 0.36 ⎞
∆Hm = hf + S ⎜ 1 − xd
n ⎟⎠
(457)
⎝

Merkley & Allen Page 264 Sprinkle & Trickle Irrigation Lectures
 where n is the number of different pipe sizes used in the branch; and S is the
ground slope in the direction of the manifold (m/m)

• The above equation estimates the location of minimum pressure in the

downhill part of the manifold

(f) if ∆Hm < 1.1 (∆Hm)a, then the pipe sizing is all right. Go to step (g) of this
B B B B B B

procedure. Otherwise, do one or more of the following eight adjustments:

(1) Increase the pipe diameters selected for the manifold

• Do this proportionately by reselecting diameters from the EPST using a

larger Qs (to increase the energy “penalty” and recompute a new
B B

EPST). This will artificially increase the break-even flow rates in the
table (chart).
• The new flow rates to use in re-doing the EPST can be estimated for s >
0 as follows:
1/ b
⎛ hf ⎞
Qnew old
= Qs ⎜ ⎟ (458)
s ⎜ ( ∆Hm ) − ∆Em ⎟
⎝ a ⎠
and for s < 0 as:

1/ b
⎛ ⎞
⎜ hf ⎟
old ⎜ ⎟
Qnew
s = Qs (459)
⎜ ∆H ⎛ 0.36 ⎞ ⎟
⎜ ( m )a − ∆El ⎜ 1 − n ⎟ ⎟
⎝ ⎝ ⎠⎠
•The above two equations are used to change the flow rates to compute
the EPST
• The value of Qm remains the same B B

• The elevation change along each manifold (uphill or downhill branches)

is ∆El = sL/100
(2) Decrease Sm B B

• This will make the laterals shorter, Qm will decrease, and ∆Hl may B B B B

decrease
• This alternative may or may not help in the design process

(3) Reduce the target EU

• This will increase ∆Hs B B

Sprinkle & Trickle Irrigation Lectures Page 265 Merkley & Allen
 (4) Decrease ∆Hl (use larger pipe sizes) B B

• This will increase the cost of the pipes

(5) Increase Ha B B

• This will increase ∆Hs B B

• This alternative will cost money and or energy

(6) Reduce the manufacturer’s coefficient of variation

• This will require more expensive emitters and raise the system cost

(7) Increase the number of emitters per tree (Np) B B

• This will reduce the value of νs B B

(8) If Ns > 1, increase Ta per station

B B B B

• Try operating two or more stations simultaneously

• Now go back to Step (b) and repeat the process.

(g) Compute the manifold inlet head,

Hm = Hl + khf + 0.5∆Em (460)

where, k = 0.75 for a single size of manifold pipe

k = 0.63 for two pipe sizes
k = 0.50 for three or more sizes

• For non-critical manifolds, or where ∆Hm < (∆Hm)a, decrease Qs (or just B B B B B B B B

design using another sizing method) in the Economic Pipe Selection

Table to dissipate excess head
• For non-rectangular subunits, adjust F using a shape factor:

Fs = 0.38S1.25
f + 0.62 (461)

where Sf = Qlc/Qla; Qlc is the lateral discharge at the end of the manifold
B B B B B B B B

and Qla is the average lateral discharge along the manifold. Then,
B B

⎛ JL ⎞
hf = Fs F ⎜ ⎟
⎝ 100 ⎠ (462)

IX. Manifold Pipe Sizing by the “HGL” Method

• This is the “Hydraulic Grade Line” method

• Same as the semi-graphical method, but performed numerically

Merkley & Allen Page 266 Sprinkle & Trickle Irrigation Lectures
 (a) Uphill Side of the Manifold

• Get the smallest allowable pipe diameter and use only the one diameter
for this part of the manifold

(b) Downhill Side of the Manifold

Largest Pipe Size, D1 B B

• First, determine the minimum pipe diameter for the first pipe in the
downhill side of the manifold, which of course will be the largest of the
pipe sizes that will be used
• This can be accomplished by finding the inside pipe diameter, D, that will
give a friction loss curve tangent to the ground slope
• To do this, it is necessary to: (1) have the slope of the friction loss curve
equal to So; and, (2) have the H values equal at this location (make them
B B

just touch at a point)

• These two requirements can be satisfied by applying two equations,
whereby the two unknowns will be Q and D1 B B

• Assume that Ql is constant along the manifold…

B B

• See the following figure, based on the length of the downstream part of
the manifold, xd B B

• Some manifolds will only have a downhill part – others will have both
uphill and downhill parts

H flow direction

(∆Hm)a
hf
e
rv

D1 = ???
cu
ss

lo
i on
fr ict
So ∆Em
1
0
0 manifold flow rate Qm
xd
Sprinkle & Trickle Irrigation Lectures Page 267 Merkley & Allen
 • For the above figure, where the right side is the mainline location and the
left side is the downstream closed end of the manifold, the friction loss
curve is defined as:

JFL
H = ( ∆Hm )a + ∆Em − hf + (463)
100

where, using the Hazen-Williams equation,

1.852
⎛Q⎞
J = K⎜ ⎟ D−4.87 for 0 ≤ Q ≤ Qm (464)
⎝C⎠

1 1 0.852
F= + + (465)
2.852 2N 6N2

⎛ x ⎞⎛ Q ⎞
N = ⎜ d ⎟⎜ ⎟ for N > 0 (466)
⎝ Sl ⎠ ⎝ Qm ⎠

where N is the number of outlets (laterals) from the location of “Q” in the
manifold to the closed end
⎛ Q ⎞
L = xd ⎜ ⎟ (467)
⎝ Qm ⎠

For Q in lps and D in cm, K = 16.42(10)6 P P

• The total head loss in the downhill side of the manifold is:

1.852
J F x ⎛Q ⎞
hf = hf hf d = 0.01K ⎜ m ⎟ D−4.87Fhf x d (468)
100 ⎝ C ⎠

where Fhf is defined as F above, except with N = xd/Sl.

B B B B B B

• The slope of the friction loss curve is:

dH 1 ⎛ dJ dF dL ⎞
= ⎜ FL + JL + JF ⎟ (469)
dQ 100 ⎝ dQ dQ dQ ⎠

where,

Merkley & Allen Page 268 Sprinkle & Trickle Irrigation Lectures
 dJ 1.852KQ0.852
= (470)
dQ C1.852D4.87

dF xd ⎛ 1 0.852 ⎞
=− ⎜ + ⎟ (471)
dQ SlQmN2 ⎜⎝ 2 3N ⎟⎠

dL x
= d (472)
dQ Qm

• Note that dH/dQ ≠ J

• The ground surface (assuming a constant slope, So) is defined by: B B

⎛ Q ⎞
H = SoL = So x d ⎜ ⎟ (473)
⎝ Qm ⎠
and,
dH So x d
= (474)
dQ Qm

• Combine the two equations defining H (this makes the friction loss curve
just touch the ground surface):

⎛ Q ⎞ JFL
So x d ⎜ ⎟ = ( ∆Hm )a + ∆Em − hf + (475)
⎝ Qm ⎠ 100

• Solve the above equation for the inside diameter, D:

−0.205
⎡ 1.852 ⎛ So x dQ ⎞⎤
⎢ 100C ⎜ − ( ∆H )
m a − ∆ Em ⎟⎥
D= ⎢ ⎝ Qm ⎠⎥
( )
(476)
⎢ K Q 1.852 1.852
FL − Qm Fhf x d ⎥
⎢ ⎥
⎣⎢ ⎦⎥

• Set the slope of the friction loss curve equal to Soxd/Qm,

B B B B B B

So x d 1 ⎛ dJ dF dL ⎞
= ⎜ FL + JL + JF ⎟ (477)
Qm 100 ⎝ dQ dQ dQ ⎠

Sprinkle & Trickle Irrigation Lectures Page 269 Merkley & Allen
 • Combine the above two equations so that the only unknown is Q (note: D
appears in the J & dJ/dQ terms of the above equation)
• Solve for Q by iteration; the pipe inside diameter, D, will be known as part
of the solution for Q
• The calculated value of D is the minimum inside pipe diameter, so find the
nearest available pipe size that is larger than or equal to D:

D1 ≥ D & minimize(D1 − D) (478)

Slope of the Tangent Line B

• Now calculate the equation of the line through the origin and tangent to
the friction loss curve for D1 B B

• Let St be the slope of the tangent line

B B

⎛ Q ⎞
H = StL = St x d ⎜ ⎟ (479)
⎝ Qm ⎠
then,
⎛ Q ⎞ JFL
St x d ⎜ ⎟ = ( ∆Hm )a + ∆El − hf + (480)
⎝ Qm ⎠ 100

• Set the slope of the friction loss curve equal to Stxd/Qm, B B B B B B

St x d 1 ⎛ dJ dF dL ⎞
= ⎜ FL + JL + JF ⎟ (481)
Qm 100 ⎝ dQ dQ dQ ⎠

• Combine the above two equations to eliminate St, and B B

solve for Q (which is different than the Q in Eq. 476)

• Calculate the slope, St, directly B B

Smaller (Downstream) Pipe Sizes B

• Then take the next smaller pipe size, D2, and make its B B

friction loss curve tangent to the same line (slope = St); B B

JFL
H = H0 + (482)
100

where H0 is a vertical offset to make the friction loss curve tangent to the
B B

St line, emanating from the origin

B B

Merkley & Allen Page 270 Sprinkle & Trickle Irrigation Lectures
 • Equating heads and solving for H0, B B

⎛ Q ⎞ JFL
H0 = St x d ⎜ ⎟− (483)
⎝ Qm ⎠ 100

• Again, set the slope of the friction loss curve equal to St, B B

St x d 1 ⎛ dJ dF dL ⎞
= ⎜ FL + JL + JF ⎟ (484)
Qm 100 ⎝ dQ dQ dQ ⎠

• Solve the above equation for Q, then solve directly for H0 B B

• Now you have the equation for the next friction loss curve
• Determine the intersection with the D1 friction loss curve to set the lengthB B

for size D1; this is done by equating the H values for the respective
B B

equations and solving for Q at the intersection:

1.852
Hbig − Hsmall +
FLK ⎛ Q ⎞
⎜ ⎟
100 ⎝ C ⎠
(Dbig
−4.87 −4.87
− Dsmall )=0 (485)

where, for the first pipe size (D1): B B

Hbig = ( ∆Hm )a + ∆El − hf (486)

and for the second pipe size (D2): B B

Hsmall = H0 (487)

and F & L are as defined in Eqs. 437 to 439.

• Then, the length of pipe D1 is equal to:

B B

⎛ Q ⎞
LD1 = x d ⎜ 1 − ⎟ (488)
⎝ Qm ⎠

• Continue this process until you have three or four pipe sizes, or until you
get to a pipe size that has D < ½D1 B B

Sprinkle & Trickle Irrigation Lectures Page 271 Merkley & Allen
 Comments about the HGL Method B

• The above equation development could also be done using the Darcy-
Weisbach equation
• Specify a minimum length for each pipe size in the manifold
so that the design is not something ridiculous (i.e. don’t just
blindly perform calculations, but look at what you have)
• For example, the minimum allowable pipe length might be
something like 5Sl
B B

• Note that the friction loss curves must be shifted vertically

upward to provide the correct average (or minimum, if pressure regulators
are used) pressure head in the manifold; this shifting process determines
the required manifold inlet pressure head, Hm
B B

• Below is a screen shot from a computer program that uses the HGL
method for manifold pipe sizing

Merkley & Allen Page 272 Sprinkle & Trickle Irrigation Lectures
Sprinkle & Trickle Irrigation Lectures Page 273 Merkley & Allen
Merkley & Allen Page 274 Sprinkle & Trickle Irrigation Lectures
Lecture 24
Hydraulic Design of Mainline & Supply Line

I. Introduction

• Chapter 24 of the textbook contains a good summary and discussion of the

design process for trickle irrigation systems
• Keller & Bliesner divide the trickle irrigation design process into three phases:

1. Phase 1: Planning factors, gross application depth,

preliminary system capacity, filtration requirements, etc.

2. Phase 2: Hydraulic design of laterals, manifolds, and

subunit layout.

3. Phase 3: Hydraulic design of main and supply lines,

control head, and pumping plant.

II. Factors Affecting the EU

• Solomon (1985) studied the various factors affecting emission

uniformity and ranked them according to importance using model
studies and field surveys. This is his ranking:

1. Plugging of emitters and other system components

2. Number of emitters per plant, Np
B B

3. Emitter coefficient of variation, ν

4. Emitter exponent, x
5. Emitter discharge variation with temperature
6. Pressure head variations in subunits
7. Coefficient of variation on pressure regulators
8. Friction loss from manifolds to laterals
9. Number of pipe sizes in manifolds

• In general, plugging (or the lack thereof) has the greatest

influence on EU, except in cases where the system design is
very poor
• Plugging cannot be prevented by design only -- it is up to the
management to continuously maintain the system as necessary
• The above ranking implies that a poorly designed system might
perform better (in terms of application uniformity) than a well
designed system, provided that the system operation and
maintenance is given due attention and effort

Sprinkle & Trickle Irrigation Lectures Page 275 Merkley & Allen
III. Mainline, Supply, and Control Head Design

• After designing laterals and manifolds, the mainline, supply line

(if necessary), control head, and pump must be
designed/selected. This involves calculating the TDH at the
pump.
• Maximum system capacity would have already been determined
in the preceding design steps, based on the number of subunits,
their sizes, emitter discharge and spacing, and number of
emitters per plant.
• Subunits should be designed to have nearly the same discharge
so that the pump is not wasting energy in lower capacity
subunits; but this is not always possible.
• Mainline design is the same as in sprinkle systems, and various
pipe selection methods are available (including the economic
pipe selection method).
• The control head includes the filters, sand separators, chemical
injection equipment, flow rate and or volumetric meters, timers,
pressure gauges, and/or other hardware. Most of this equipment
is located in the same place in a trickle irrigation system.
However, other filters, screens, and gauges may be installed at
downstream locations in some cases.
• Sand media and other filters may have a combined head loss of
10-20 psi when “dirty”, and 3-10 psi when clean. However, if too
much flow is forced through the filters (i.e. not enough filter
capacity) the head loss can be higher, even when clean.
• Flow meters and chemical injection equipment may have1-5 psi
loss, and valves can have up to several psi loss (even when fully
open).
• The pressure changes from the control head to the subunits are
due to a combination of friction loss and elevation change: if the
subunit inlet is at a higher elevation than the control head, both
friction and elevation change contribute to a higher required
pressure at the control head, otherwise they tend to cancel out
(partially or completely).
• Mainline friction losses include minor (local) losses at bends,
through valves, and through screens.
• Losses at reductions in pipe diameter (in series) are usually
small, unless the reduction is sharp and abrupt. Diameter
transitions for PVC are often smooth.
• A “critical” subunit can be defined by calculating the combined
friction loss and elevation change to each subunit, and taking the
highest value plus the required inlet pressure head to the
manifold. The textbook calls this (Hm + Hfe)c.
B B B B B B

Merkley & Allen Page 276 Sprinkle & Trickle Irrigation Lectures
 • The critical subunit will define the “worst case” for which the
pumping unit should be designed. It may then be possible to
reduce pipe sizes in other subunits if they will have excess head
available.

IV. Trickle System Mainline Design

• Trickle mainlines can be sized using the same approach as is

used for sprinkle irrigation systems.
• Manifolds can be considered as stationary sprinkler laterals.
• You can use the EPST approach, where Ot depends (in part) on B B

the number of stations, Ns B B

Example calculation:

• Size sections A-B, B-C and C-D for a trickle irrigation mainline
having three manifolds
• First, decide on the number of stations, Ns B B

This could be 1, 3 or 6:

1. All laterals at the same time, or

2. Only laterals on one manifold at a time, or
3. Only laterals on one half of a manifold (on one side
of the mainline) at a time

• Set up a table to see the effect of Ns on the flow rates and Ot per B B B B

station (also see the figure below):

Number of Stations
1 2 3
Qs (lps) B B 30 30 30
Ot/station (hrs)
B B 1000 330 165
Ot, CD B B 1000 330 330
QCD B B 10 30 30
Ot, BC B B 1000 660 660
QBC B B 20 30 30
Ot, AB B B 1000 1000 1000
QAB B B 30 30 30

• Ot, CD, for example, is the number of hours that water is flowing in
B B

section CD per season

• This is the Ot that would be used in creating an economic pipe
B B

sizing table to determine the pipe size of section CD

• Ot varies with the number of stations and with the location of the
B B

mainline segment in the system

Sprinkle & Trickle Irrigation Lectures Page 277 Merkley & Allen
 • Therefore, the EPST must be developed several times (this is
easy in a spreadsheet, as you have seen)

• Notice that the product of Q and Ot is a constant, as this is

B B

proportional to the irrigation water requirement, which is the

same for each station and location in the field
• Pipe diameters are selected using QAB, QBC, and QCD as the
B B B B B B

break-even flow rates in the EPST; Qs remains constant

B B

The TDH is the sum of:

1. static lift
2. well losses (if applicable)
3. supply line losses (elevation and friction)
4. control head pressure losses
5. losses to the critical subunit plus inlet pressure, (Hm + B B

Hfe)c
B B B B

6. screen and valve losses (if applicable) at subunit inlets

• As in many hydraulic designs for pressurized pipe systems, it is

often recommendable to add a “safety factor” to the losses,
because losses are not precisely known, and they will probably
increase with time
• A common safety factor is to add 10% to the friction losses for
the calculation of TDH
Merkley & Allen Page 278 Sprinkle & Trickle Irrigation Lectures
 • A separate safety factor can be added, in some cases, to the
TDH for compensation for emitter plugging and degradation
• See sample calculations of TDH in Chapter 24

VI. We Live in an Imperfect World

• In many trickle systems, the ground slope

changes significantly between subunits, or
within subunits
• This complicates the system design and may
require more pressure regulation and larger
pipe sizes than would otherwise be necessary

• In many locations and countries the available

pipes, fittings, emitters, filters, and other
hardware are very limited
• Therefore, innovation, resourcefulness, and
improvisation may be very important
• If hardware is very limited, it may be best to consider another
type of irrigation system

• The system capacity for a micro-irrigation system can have a

safety factor added on to account for the possibility of:

1. slow, partial clogging of emitters and laterals

2. changes in crop type
3. inaccurate estimations of peak crop ET
4. more system “down time” than originally anticipated
5. various other factors

• It is desirable to minimize the hardware cost of an irrigation

system, but the cost of having an insufficient system capacity
may be many times higher than the marginal cost of larger pipes,
filters and valves
• When in doubt, it may be a good idea to increase the calculated
system capacity by 10 or 20%

• See Chapter 25 of the textbook for a thoughtful discussion of

factors in selecting a sprinkler or trickle system, or for selecting
another type of irrigation system

Sprinkle & Trickle Irrigation Lectures Page 279 Merkley & Allen
Merkley & Allen Page 280 Sprinkle & Trickle Irrigation Lectures
 Sprinkle and Trickle Irrigation Text Errata
September 1, 2004

p. 17: Second line add comma between wetted and may to read: --- “but lower leaves, if
wetter, may” -----.

p. 32: Equation 3.1. In definitions of terms, change “WA” to “Wa”; and for depth Z
change metric dimension from “mm” to “m”.

p. 38: In full sentence above Table 3.5, pluralize value to read: ---“the ECe values
presented in Table 3.5” ----.

p. 39 End of fifth line from top should be changed from LR < 0.1 to LR > 0.1.

p. 41: Add an another “e” in Hargr(e)aves in second line and in References.

p. 46: Third line from bottom of third paragraph, change “as” to “an” to read: ---
“making too sharp an “S” turn”----.

p. 65: In item 6, change “form” to “from”.

p. 70: Fig. 5.6. Two errors: a) the wind speed should be “5 m/s” not “1 m/s”; and b) the
sprinkler location “0” on the right hand (vertical) scale is not offset to reflect wind, it
should be shifted to where the lower “6” is located.

p. 83: Table 5.4. Change last number in first column from “(0.40)” to “(0.15)”.

p. 91: Fig. 6.2. Replace “Pivot” with “Pitot” in figure title.

p. 99: Table 6.2. For CU = 94% and pa = 80% the value given should be “94”.

p. 131: Sample Calculation 7.1. In last line of GIVEN: change “5-ft” to “50-ft”.

p.134: The value for the constant “K “ below Equation 8.1 for metric units would be
slightly more accurate if changed from 1.212 to 1.217 x 1012.

p. 135: On third line above Equation (8.2), change “13-m” to “13-mm”.

p.138: The value for the constant “K “ below Equation 8.7a for metric units would be
slightly more accurate if changed from 7.89 to 7.88 x 107.

p. 141: In the Flow rate column for (gpm): the 6.0 should be followed by “6.1” rather
than “0.1”.

1
p. 146: Two typos in the text one line above and two lines below Equation (8.9a), change
“form” to “from”.

p. 147: Sample Calculation 8.1. At end of last line, change “7.07” to “70.7”.

p. 164: Table 8.12. The heading for the first column should simply be “Method”, strike
out “(or size for C-D)”.

p. 167: Equation 8.19. Change “Table 8.11” to “Table 8.9” in definition for EAE(e).

p. 170: Fig. 8.8. The long FIXED PLUS OPERATING COSTS arrow should have had its
arrowhead at and terminated at the point where it crossed the solid line.

p. 196: In last line, change “Ha” to “Pa”.

p. 197: Fig. 9.7. In the left hand caption unit line, change “5 = 5%” to “S = 5%”.

p. 197: Sample Calculation 9.8. In third line of CALCULATIONS: change “two-end

outlets” to “two end-outlets”.

p. 209: Sample Calculation 10.2. In the line above the fourth equation from the bottom
of the page, the sentence should read, “The final hf2 with all 6-in. pipe is:”

p. 213: Equation 10.5. Change “LS” to ”Ls”.

p. 218: Sample Calculation 10.5. In GIVEN: under Economic relationships: change “Cp”
to “Cp”.

p. 219: Sample Calculation 10.5. In CALCULATION: Step 2, change the “0.0001” in the
equation to “0.001”.

p. 220: Sample Calculation 10.5. Clarification, in calculating L4, Js = J4 = 3.17 m/100m

and Jb = J6= 0.52 m/100m were interpolated from Table 8.5.

p. 222: Change “g = accumulation…” to “g = acceleration”.

p. 223: Table 11.2. Change the bottom Equation for “sudden contractions” to:

[
K r = 0.7 1 − (D r )2 ] 2

p. 226: Sample Calculation 11.1. In CALCULATIONS: in first line change “K” to “Kr”
and in fifth line from bottom, change “interests” to “intersects”.

p. 230: The fifth text line should read: “The static head ----- and C is:”

p. 231: Equation 11.4a. Change the “Qg” in the middle term on the right side to “Qs”.

2
p. 242: Near bottom, change “H = 1197 ft” to “H = 197 ft”

p. 256: Table 12.2. Change caption to read “.hp / 100 ft…” rather than “.hp / 1000 ft…”

p. 292: Below Equation 13.4. In the definition of T, change “180o $ T # 360o” to “180o
# T # 360o”.

p. 313: Table 14.1. Change the values in the “Irrigation interval” columns for both 5 and
6 days for Pasture - Peak from “0.02” to “1.02”.

p. 322: In first line, change “(0.3 in. / hr)” to “(0.4 in. / hr)”.

p. 324: Table 14.2. This table is out of date. The Irrigation Association’s Center Pivot
manual has an updated equivalent.

p. 343: Sample Calculation 14.6. Change the subscript of R’ from “R’e” to “R’n” in two
places a) in the sixth line from top of page; and b) in the line above the last equation.

p. 361: Error in four places, should read “…for the 168-mm (6 5/8-in.)…” (i.e., change
all occurrences of “6-in.” to “6 5/8-in.”).

p. 362: Error in eight places, change occurrence of “6-in.” to “6 5/8-in.”.

p. 363: Error in one place, change occurrence of “6-in.” to “6 5/8-in.”.

p. 363: Second equation should read:

(h f − h fj )6 = (h f )
− h fj J 6 / J 8 = 1.16 x 5.97 / 2.31 = 3.0 m
8

In other words, change the “+” separating hf and hfj to a “-“

Note that in the equation for (hf)8,6 that one should also subtract out for the hf due
to the end gun; however, this turns out to be small (0.04 m).

p. 395: In the line below Equation (15.5), change “LET” to “LET” and “indix” “index”.

p. 407: In eleventh line from bottom, change “drop” to “crop”; and in sixth line from
bottom, change “precipitaton” to “precipitation”.

p. 436: Fig. 17.5. In Part C, add value so (Pw ≅ 50%).

p. 443: Table 18.2. In first column change “Very find sand” to “Very fine sand”.

3
p. 456: Sample Calculation 19.1. In first equation “w” under CALCULATIONS: change
the “.15” to “1.5”, but the calculation is OK.

p. 457: Item 5 should read: “A 100-ml graduated cylinder;”

p. 462: Equation 19.3, 19.4, and 19.5. A “Pd” should be added to the denominator to
allow for applications to sparse plantings. As is, it applies only to dense mature orchards.
Otherwise, for sparse plantings, the canopy area should be used rather than Sp x Sr, but
this could become confusing. Therefore, the denominators should read: “Sp x Sr x
Pd/100”.

p. 492: At the end of last bullet near top of page, change “(1 + ν)” to “(1 " ν)”.

p. 496: Fig. 20.8. The three curves should be labeled from top to bottom: “LARGE ON-
LINE”; STANDARD ON-LINE”; AND “SMALL ON-LINE”.

p. 497: Errors in two places about mid-page, change “L / s” to “L / h”.

p. 512: In fourth line from top, change “dn” to “dn”.

p. 512 The G calculated using Equation 19.16a be changed from 9.3 to 93.3 gal/day.

p. 513: In line above Equation 20.13a’, change “np” to “Np”.

p. 516: Near bottom of page in reference to “Seasonal Irrigation Efficiency” should refer
to Table 19.4 (rather than 19.3).

p. 521: In the second equation on the page for using Equation 19.11, change “[0.1 +
(75)0.5]” to “[0.1 (75)0.5]”, but calculation is OK.

p. 521: In line above last equation, change “Ot” to “Ot”.

p. 535: Comment: rearranged, Equation 22.11 becomes:

L (1 − F) a / b −1 / b
∆H c = S J
100

(see Equation 23.14b). This form is simpler and it might be more didactic to present both
forms. (S is absolute slope).

p. 537: Fig. 22.5. Insert arrows to show where “Av. hfp” refers to (left-hand side, middle
of figure). “Av. hfp” should refer to the vertical distance between the horizontal dashed
line and the lower end of the hfp “curve” (i.e., the vertical height of the hatched area that
is beneath the dashed line).

4
p. 538: Equation 22.19, 22.20. Add “-x/Lp ∆Ep” to each equation so that they read as:
2.75
⎛ ⎞
(H )d = h fp ⎜⎜ x ⎟
⎟
+ ∆H c + H 'n −
x
∆E p (22.19)
⎝ Lp ⎠ Lp

2.75
⎛ ⎞ ⎛ x ⎞⎟
(H )u = h fp ⎜⎜1 − x ⎟⎟ + ⎜1 −
⎜ ⎟
∆E p + H 'n (22.20)
⎝ Lp ⎠ ⎝ Lp ⎠

p. 549: Equation 22.17. Change “(y – 0.5)” to “(Y – 0.5)”.

p. 566: For clarification the bottom paragraph should read:

“Basic Equations. The elevation (relative to the datum at the minimum pressure head
but at the end of the pipe) of the hydraulic grade at any point x …..”

Furthermore, in the legend following Equation 23.14a, change explanation for Hx to:
“Hx = hydraulic head (minimum) at point x along a pipe-friction curve that is tangent to
the HGL, m (ft). (this is contrasted against Hx on page 577, which is a friction loss)“

p. 567: Equation 23.14b. Change the exponent on (x/L) to “a” instead of “1/a”.

p. 570: The numerical values in the second equation should be: 640/177 = 3.6 (not
640/17).

p. 581: Change referenced publication in next to last line to read: “Trickle Irrigation for
Crop Production “.

p. 598: Table 24.6. An “=” sign is missing in the equation in the middle of the table, it
should read: “(Hm + Hfe) = (50.2 + 7.4) =”.

5
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester 2004 – Exam #1
Include units in all results. Indicate any assumptions that you might make.

Name__________________________________

1. Preliminary Design Calculations (25 pts)

Almonds (deciduous trees, with no cover crop) have been planted in an 80-acre
orchard and will be irrigated with a set sprinkler system. The climate can be
characterized as “moderate.” The topsoil is 2.7 ft of silty clay loam, and below that are
3.1 ft of clay loam soil. The soil intake rate is 0.5 inch/hr.

Use an MAD value of 30%. ECw is 0.47 dS/m and the estimated water application
efficiency will be Ea = 85%. There is no effective rain during the peak-use period. For
these preliminary calculations, and based on the spacing of the trees, use a sprinkler
spacing of 40 ft x 50 ft.

a) Obtain the values for average Wa, average Z, Ud, U, and ECe from
the tables in Chapter 3 of the textbook.
b) Calculate the maximum net application depth per irrigation, dx.
c) Calculate the maximum irrigation interval, nominal irrigation
interval (whole number of days), and the net application depth per
irrigation, dn.
d) What is the gross depth to apply per irrigation?
e) What is the irrigation set time, Sto, in hours?

2. Economic Pipe Size Selection (25 pts)

Suppose you applied the economic pipe sizing method. What if you were using the
Hazen-Williams equation and based all your calculations on a system capacity of 100
lps, but now you realize a calculation mistake was made, and the system capacity
should really be 115 lps.

A section flow rate, q, (threshold between two adjacent pipe sizes) was 50 lps, but now
it needs to be adjusted for the new system capacity of 115 lps.

What is the new section flow rate for this system capacity?

1 of 8
3. Set Sprinkler Lateral Design (25 pts)

A fixed sprinkler system with buried IPS-PVC (thermoplastic pipe) laterals is to be

designed. The 304-m long laterals will be a -0.394% (downhill) slope. The nominal
sprinkler flow rate is 10 lpm at a pressure of 280 kPa, and the sprinkler spacing on the
laterals is 8.0 m. Riser height is 1.0 m. Let the allowable lateral ∆h be equal to 20% of
ha.

a) How many sprinklers will operate on the lateral if the first sprinkler
is spaced Se from the lateral inlet?
b) For a dual pipe size lateral, what is the allowable friction loss
gradient, Ja?
c) What two adjacent IPS-PVC thermoplastic pipe sizes would you
recommend? Specify the nominal diameters in inches according
to Table 8.3.
d) What are the respective lengths (x1 and x2) of the two pipe sizes?
Round the lengths to a multiple of Se for each size.
e) What is the required lateral inlet pressure head, hl?
f) What is the pressure at the downstream end of the lateral during
operation?
g) Is the mean velocity at the lateral inlet too high?

4. (25 pts) A portable aluminum sprinkler lateral has a nominal diameter of 4 inches and
goes downhill at a uniform slope of -0.5%. The inlet flow rate is 260 gpm, the lateral
length is 840 ft, and the sprinkler spacing along the lateral is 30 ft. The lateral inlet
pressure head is hl = 30.0 m.

a) What is the location of minimum pressure in the lateral pipe during

operation?
b) What is the minimum pressure in the lateral pipe during
operation?
c) Is the mean velocity at the lateral inlet too high?

2 of 8
Solutions:

1. Preliminary Design Calculations (25 pts)

a) Obtain the values for average Wa, average Z, Ud, U, and ECe from
the tables in Chapter 3 of the textbook.

• From Table 3.1, Wa range is the same for the topsoil & subsoil. Use Wa =
½(145 + 208) = 176.5 mm/m.
• From Table 3.2, for almonds, Z = ½(0.6 + 1.2) = 0.9 m.
• From Table 3.3, for deciduous orchard w/o cover crop in a “moderate” climate,
Ud = 4.8 mm/day, and U = 533 mm/season.
• From Table 3.5, ECe = 2.0 dS/m for almonds.

b) Calculate the maximum net application depth per irrigation, dx.

dx = MAD(Wa)(Z) = 0.30(176.5)(0.9) = 47.7 mm/irrig

c) Calculate the maximum irrigation interval, nominal irrigation interval

(whole number of days), and the net application depth per irrigation,
dn.

dx 47.7 mm/irrig
fx = = = 9.94 days/irrig
Ud 4.8 mm/day

Rounding down, f = 9 days. This is conservative because fx is nearly 10 days.

Then,

dn = ( 9 days/irrig )( 4.8 mm/day ) = 43.2 mm/irrig

d) What is the gross depth to apply per irrigation?

ECw 0.47
LR = = = 0.049
5ECe − ECw 5(2.0)-0.47

LR < 0.1, so,

dn 43.2 mm/irrig
d= = = 50.8 mm/irrig
Ea 0.85

e) What is the irrigation set time, Sto, in hours?

The soil intake rate is given, at 0.5 inch/hr, which equals 12.7 mm/hr. Then, the
minimum set time is:

3 of 8
 50.8 mm/set
( Sto )min = = 4.0 hrs/set
12.7 mm/hr

2. Economic Pipe Size Selection (25 pts)

Note the graph on page 67 of the lecture notes. The lines separating the adjacent
pipe sizes do not change because, in this problem, none of the economic
parameters have changed. The relationship between Q and q is fixed along the
slope of 2:1 on the log-log plot, or 1.852:1 in our case (with Hazen-Williams, as
specified).

We have:
Qs = K ( q−1.852 )

Then,
100 = K ( 50−1.852 )
115 = K ( qnew −1.852 )

or,
1/ −1.852
⎛ 115 ⎞
qnew = 50 ⎜ ⎟ = 46.6 lps
⎝ 100 ⎠

3. Set Sprinkler Lateral Design (25 pts)

a) How many sprinklers will operate on the lateral if the first sprinkler is
spaced Se from the lateral inlet?

304 m/lat
= 38 sprink/lat
8 m/sprink

b) For a dual pipe size lateral, what is the allowable friction loss
gradient, Ja?

Since there are more than 30 outlets along the lateral pipe, F = 0.36. The nominal
sprinkler pressure head is 280 kPa/9.81 = 28.5 m. Then,

⎛ 0.20ha − ∆he ⎞ ⎛ 0.20(28.5) − (304)( −0.00394) ⎞

Ja = 100 ⎜ ⎟ = 100 ⎜ ⎟ = 6.3 m/100m
⎝ FL ⎠ ⎝ (0.36)(304) ⎠

c) What two adjacent IPS-PVC thermoplastic pipe sizes would you

recommend? Specify the nominal diameters in inches according to
Table 8.3.

4 of 8
 The lateral inflow rate is:

QA = (38 sprink/lat)(10 lpm/sprink) = 380 lpm/lat (6.33 lps/lat)

With Ja = 6.3 m/100 m, Table 8.3 gives the following two adjacent PVC pipe sizes:
2” and 2½.” Note that the respective pipe IDs are 55.7 mm and 67.4 mm.

d) What are the respective lengths (x1 and x2) of the two pipe sizes?
Round the lengths to a multiple of Se for each size.

This problem requires a few iterations. Use C = 150 for PVC with Hazen-Williams.
Use units of lps and mm. From the lecture notes, we have:

1.217(10)12
α1 = 1.852
= 1.135(10)6
100(150)

1.852
⎡ (10 / 60)(304) ⎤
α2 = ⎢ ⎥⎦ (67.4)−4.87 (0.36)(304) = 4.152(10)−6
⎣ 8

1.852

α3 = ( 67.4 −4.87
− 55.7 −4.87
) ⎛⎜⎝ 108/ 60 ⎞⎟⎠ = −1.464(10)−12

Note that F1 = 0.36 due to the 38 sprinklers over the length L. F2 will depend on x1.
Make a table of values, arbitrarily choosing an initial x1 value of 150 m, and
searching for f(x1) = 0:

x1 (m) N F2 f(x1)
150 19 0.38 -1.10
100 26 0.37 0.19
125 22 0.37 -0.55
112 24 0.37 -0.19
105 25 0.37 0.03

Rounding, let x1 = 104 m, whereby x2 = L – x1 = 200 m.

e) What is the required lateral inlet pressure head, hl?

Calculate hf using F, as described in the lecture notes:

⎛x ⎞ ⎛ 200 ⎞
Q2 = ⎜ 2 ⎟ qa = ⎜ ⎟ (10 / 60) = 4.17 lps
⎝ Se ⎠ ⎝ 8 ⎠

5 of 8
 1.852
⎛ 6.33 ⎞
( 67.4 )
−4.87
J1 = 1.217(10)12 ⎜ ⎟ = 4.30 m/100m
⎝ 150 ⎠

1.852
⎛ 4.17 ⎞
( 67.4 )
−4.87
J2 = 1.217(10) ⎜
12
⎟ = 1.99 m/100m
⎝ 150 ⎠

1.852
⎛ 4.17 ⎞
( 55.7 )
−4.87
J3 = 1.217(10) ⎜
12
⎟ = 5.03 m/100m
⎝ 150 ⎠

where F1 = 0.36 and F2 = 0.37. Then,

471 − 147 + 372

hf = = 7.0 m
100

Finally,
5 1
hl = ha + hf + ∆he + hr
8 2

5 1
hl = 28.5 + (7.0) + ( −1.2) + 1.0 ≅ 33 m (320 kPa)
8 2

f) What is the pressure at the downstream end of the lateral during

operation?

hend = hl − hf − ∆he
= 33 − 7 + 1.2
≅ 27 m

or, 27 – hr = 27 – 1 = 26 m head at the last sprinkler.

g) Is the mean velocity at the lateral inlet too high?

Q 4(6.33 lps)
V= = = 1.77 m/s
A π(0.0674 m)2 (1,000 l/m3 )

Then, V < 2.0 m/s, so it is probably all right.

6 of 8
4. (25 pts) Portable aluminum sprinkler lateral.

• From Table 8.1, 4-inch pipe has ID = 3.9” (99.1 mm).

• Ql = 260 gpm, or 16.4 lps.
• Se = (30)(0.3048) = 9.14 m.
• N = 840/30 = 28 sprinklers.
• qa = 16.4/28 = 0.586 lps.
• For aluminum lateral pipe, C = 130.

a) What is the location of minimum pressure in the lateral pipe during

operation?

From the lecture notes, page 74:

⎛ 9.14 ⎞ ⎡
⎟ ⎣16.4 − 3(10) (130(0.5) (99.1) )⎤⎦ = 181 m
−7
x=⎜ 0.54 2.63

⎝ 0.586 ⎠

which is x = 595 ft.

b) What is the minimum pressure in the lateral pipe during

operation?

• L = 840 ft, or 256 m.

• F1 = 0.37 for 28 outlets (Table 8.7).
• Outlets beyond x = 181 m: N2 = (256 – 181)/9.14 ≈ 8.
• F2 = 0.42 for 8 outlets (Table 8.7).

1.852
⎛ 16.4 ⎞
( 99.1)
−4.87
J1 = 1.217(10) ⎜ 12
⎟ = 5.00 m/100 m
⎝ 130 ⎠

1.852
⎛ (8)(0.586) ⎞
( 99.1)
−4.87
J2 = 1.217(10) ⎜ 12
⎟ = 0.49 m/100 m
⎝ 130 ⎠

The friction head loss from inlet to x = 595 ft (181 m) is:

J1FL J F (L − x)
(hf )x =181m = 1
− 2 2
100 100

(5.00)(0.37)(256) (0.49)(0.42)(75)
(hf )x =181m = − = 4.6 m
100 100

or, 15 ft of head loss. Note that hl is given in m, while all other values are given in
English units. Finally,

7 of 8
 hx =181 m = hl − (hf )x =181 m − ∆he
= 30.0 − 4.6 − 181(−0.005)
≅ 26 m (86 ft, or 37 psi)

c) Is the mean velocity at the lateral inlet too high?

Q 4(16.4 lps)
V= = = 2.1 m/s
A π(0.0991 m)2 (1,000 l/m3 )

Whereby V > 2.0 m/s, so the entrance velocity in the lateral is higher than what we
might nominally allow. This suggests consideration of a larger pipe size.

8 of 8
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester 2004 – Exam #2
Include units in all results. Indicate any assumptions that you might make.
Don’t show more than three significant digits in the results.

Name__________________________________

1. (60 pts) Two gun sprinklers are supplied water from a pump at an open reservoir
(water surface at 203 ft above msl), as shown in the figure below:

gun #1
(elev = 219 ft)
reservoir
(elev = 203 ft)

t
5f
pump 450 f

36
(elev = 209 ft) t
“T” (elev = 217 ft)
ft
0
20

gun #2
(elev = 228 ft)
There is a common 4-inch (ID = 4.280”) PVC supply line from the pump to a “T”. The
supply line is 450 ft long.

A 3-inch (ID = 3.284”) PVC pipe goes 365 ft from the “T” to gun #1, and another 3-inch
pipe goes 200 ft from the “T” to gun #2. The sprinkler height above the buried lateral
pipe is hr = 8.00 ft for both guns.

The flow rate vs. pressure data for the gun sprinkler give the following relationship:

q = 11.7P0.49

for q in gpm; and P in psi.

The suction side of the pump has the same 4-inch PVC pipe as the supply line, 12 ft
in length, with two 45-degree long-radius, flanged elbows, a basket strainer, and a
foot valve.

1 of 7
 a) Develop one point on the system curve using the Hazen-Williams
equation (with C = 150) for friction losses. Use a flow rate of 80
gpm for gun sprinkler #1.
b) Calculate the flow rate for gun sprinkler #2.
c) Calculate the total system flow rate, Qs.
d) Calculate TDH for this flow rate. Show your calculations for minor
losses.

2. (20 pts) For the B2TPM Berkeley™ pump and 6-½” impeller, and the same system,
suppose now that the desired operating point is for Q = 150 gpm.

a) From the pump curves, determine NPSHr.

b) Calculate NPSHa (water temperature is 12°C).
c) Determine whether the pump is expected to cavitate at the
operating point.

3. (20 pts) For the same B2TPM Berkeley™ pump and 6-½” impeller, suppose that the
desired operating point is 150 gpm at a TDH of 150 ft. If the nominal pump speed is
3,600 RPM, what is the required speed for the desired operating point?

4. (5 bonus pts) Which of the following are a function of a center pivot’s radial speed?
(check all that apply)

… wetted width, w
… net application depth, dn
… average application rate, ARavg
… maximum application rate, ARx
… friction loss in the lateral pipe, hf

2 of 7
Solutions:

1. (60 pts) Two gun sprinklers are supplied water from a pump at an open reservoir
(water surface at 203 ft above msl), as shown in the figure below:

Move along the pipes from sprinkler #1 to the “T,” then to sprinkler #2 to determine the
flow rate there, then get the system flow rate (Qs = Q1 + Q2), and finally move to the
pump to determine Ppump.

I. Flow rate at sprinkler #2:

Pressure at gun sprinkler #1:

1/ 0.49
⎛ 80 ⎞
P1 = ⎜ ⎟ = 50.6 psi
⎝ 11.7 ⎠

Pressure head at gun sprinkler #1:

h1 = (50.6 psi)(2.31 ft/psi) = 117 ft

Pressure head at the “T”:

hT = h1 + hr + ∆he + hf
1.852
⎛ 80 ⎞
hT = 117 + 8 + (219 − 217) + 10.5(365) ⎜ ⎟ (3.284)−4.87
⎝ 150 ⎠
hT = 127 + 0.00109(80)1.852
hT = 131 ft

Pressure head at gun sprinkler #2:

h2 = hT + ∆he − hf − hr
1.852
⎛Q ⎞
h2 = 131 + (217 − 228) − 10.5(200) ⎜ 2 ⎟ (3.284)−4.87 − 8
⎝ 150 ⎠
h2 = 131 − 11 − 0.000599Q1.852
2 −8
h2 = 112 − 0.000599Q1.852
2

3 of 7
Flow rate at gun sprinkler #2:

0.49
⎛ 112 − 0.000599Q1.852 ⎞
0.49
⎛ h ⎞
Q2 = 11.7 ⎜ 2 ⎟ = 11.7 ⎜ 2
⎟
⎝ 2.31⎠ ⎝ 2.31 ⎠

giving Q2 = 77.7 gpm.

II. System flow rate:

Qs = Q1 + Q2 = 80 + 77.7 ≅ 158 gpm

III. Pressure at pump outlet:

Pressure head at pump outlet:

hpump = hT + ∆he + hf
hpump = 131 + (217 − 209) + 0.000371Q1.852
s

hpump = 139 + 4.38 = 143 ft

Pressure at pump outlet:

hpump 143 ft
Ppump = = = 61.9 psi
2.31 2.31

IV. Suction side of the pump:

From Table 11.2, for a 4-inch pipe:

Item Count Kr Total

Foot valve 1 0.80 0.80
Basket strainer 1 1.05 1.05
45-deg elbow 2 0.18 0.36
Total: 2.21

Velocity head:
8 ( Q / 448.86 )
2
V2 8Q2
= 2 4 = = 7.72(10)−6 Qs2
2g gπ D 32.2π (4.280 /12)
2 4

Minor losses (see the above table):

4 of 7
 ⎛ V2 ⎞
(h )
f,minor suction = 2.21⎜ ⎟
⎝ 2g ⎠

Pipe friction loss:

1.852
⎛Q ⎞
(hf )suction = 10.5(12) ⎜ s ⎟ (4.280)−4.87 = 9.89(10)−6 Q1.852
s
⎝ 150 ⎠

Static lift:
(hlift )suction = 209 − 203 = 6 ft
V. Total dynamic head (TDH):

Ppump V2
TDH = + (hf )suction + (hf,minor )suction + (hlift )suction +
γ 2g

where Ppump is the pressure at the pump outlet. Simplifying,

TDH = hpump + 9.89(10)−6 Q1.852

s + 2.48(10)−5 Qs2 + 6

then,

TDH = 143 + 9.89(10)−6 (158)1.852 + 2.48(10)−5 (158)2 + 6

TDH = 143 + 0.117 + 0.619 + 6
TDH = 150 ft

VI. System curve point:

TDH = 150 ft at Qs = 158 gpm

2. (20 pts) For the B2TPM Berkeley™ pump and 6-½” impeller, and the same system,
suppose now that the desired operating point is for Q = 150 gpm.

a) From the pump curves, determine NPSHr.

b) Calculate NPSHa (water temperature is 12°C).
c) Determine whether the pump is expected to cavitate at the
operating point.

From the pump curves, at 150 gpm, NPSHr ≈ 9 ft.

Mean atmospheric pressure head:

5 of 7
 10.3 − 0.00105(203 * 0.3048)
hatm = = 33.6 ft
0.3048
For water at 12°C,
⎛ 17.27(12) ⎞
hvapor = 0.0623 exp ⎜ ⎟ = 0.143 m (0.469 ft)
⎝ 12 + 237.3 ⎠
Velocity head:
V2
= 7.72(10)−6 Qs2 = 7.72(10)−6 (150)2 = 0.174 ft
2g

Minor losses:
⎛ V2 ⎞
(h )
f,minor suction = 2.21⎜ ⎟ = 2.21(0.174) = 0.384 ft
⎝ 2g ⎠

Pipe friction loss:

1.852
⎛ 150 ⎞
(hf )suction = 10.5(12) ⎜ ⎟ (4.280)−4.87 = 0.106 ft
⎝ 150 ⎠
Static lift:
(hlift )suction = 209 − 203 = 6 ft
Available NPSH:
V2
NPSHa = hatm − hvapor − (hf )total − hlift −
2g
NPSHa = 33.6 − 0.469 − 0.382 − 0.106 − 6 − 0.174
NPSHa = 26.5 ft

The pump is not expected to cavitate because NPSHa >> NPSHr.

3. (20 pts) For the same B2TPM Berkeley™ pump and 6-½” impeller, suppose that the
desired operating point is 150 gpm at a TDH of 150 ft. If the nominal pump speed is
3,600 RPM, what is the required speed for the desired operating point?

Follow the steps in the lecture notes. Make a table for the equal efficiency curve,
using Q2 = 150 gpm, and H2 = 150 ft:

6 of 7
 Q1 H1
(gpm) (ft)
100 66.67
120 96.00
140 130.67
160 170.67
180 216.00

Plot the equal efficiency curve and look for the intersection with the pump
characteristic curve, defining point (Q3,H3). From the graph, Q3 ≈ 158 gpm, and H3 ≈
167 ft. Then,

⎛Q ⎞ ⎛ 150 ⎞
Nnew = Nold ⎜ 2 ⎟ = 3,600 ⎜ ⎟ = 3,418 RPM
⎝ 3⎠
Q ⎝ 158 ⎠

4. (5 bonus pts) Which of the following are a function of a center pivot’s radial speed?
(check all that apply)

… wetted width, w
; net application depth, dn
… average application rate, ARavg
… maximum application rate, ARx
… friction loss in the lateral pipe, hf

7 of 7
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester 2004 – Final Exam
Include units in all results. Indicate any assumptions that you might make.
Don’t show more than three significant digits in any of the results.

Name__________________________________

1. (35 pts) A mature citrus orchard will be drip-irrigated (drip emitters) using a single
lateral per row of trees in a 132-ha field area. Other information:

• Tree spacing is 6.0 x 6.0 m.

• Peak daily ET is Ud = 5.1 mm/day.
• Seasonal water requirement: U = 660 mm.
• Effective rain, peak-use period: 1.5 mm/day (average, w/ 90% probability)
• Residual soil water in the spring: assume zero.
• Soil water holding capacity is 175 mm/m (medium texture).
• Irrigation water quality: ECw = 0.89 dS/m.
• Root zone depth is 1.5 m.
• Shaded area is 78%.
• Emitter equation:

q = 0.28P0.481
for q in lph; and P in kPa.

• Nominal emitter flow rate: qa = 3.85 lph.

• Manufacturer’s emitter coefficient of variation: 0.0487.
• Average wetted width at 3.85 lph: w = 2.01 m.
• Outlets per emitter: one.
• Use an MAD of 20%.

What you need to do:

1. Select an appropriate emitter spacing, Se.

2. Determine the number of emitters per tree, Np.
3. Calculate percent wetted area, Pw.
4. Calculate maximum net depth to apply per irrigation, dx.
5. Calculate the average peak daily “transpiration” rate, Td.
6. Calculate the maximum irrigation interval, fx. If fx ≥ 1 day, use f’ = 1day.
7. Calculate the net depth per irrigation, dn.
8. Select a reasonable target EU value.
9. Determine (ECe)max.
10. Determine the transmission ratio, Tr.

1 of 10
 11. Calculate the leaching requirement, LRt.
12. Calculate the gross depth to apply per irrigation, d.
13. Calculate the gross volume of water per tree per day, G.
14. Calculate ha, corresponding to qa = 3.85 lph, in m of water head.
15. Calculate the water application time, Ta.
16. Select the number of stations, Ns.
17. Determine the minimum number of emitters per tree, Np’.
18. Calculate the system coefficient of variation, νs.
19. Calculate the minimum allowable emitter flow rate, qn.
20. Calculate the allowable subunit pressure head variation, ∆Hs.
21. Calculate the system capacity, Qs.
22. Calculate the total gross seasonal depth to apply, Dg.
23. Calculate the gross seasonal volume of irrigation water, Vs.
24. Calculate the required number of operating hours per season, Ot.

2. (30 pts) A rectangular field of strawberries will be trickle irrigated. The laterals are 380-
m long in the direction of the 17.8-mm inside diameter PE laterals. Nominal emitter
flow rate is 2.75 lph at a pressure head of 11.5 m. The emitters are in-line, without any
barbs, spaced at 0.4 m along the lateral hose, which lies along a uniform ground slope
of 0.761%. The strawberries are spaced at 0.5 m in the field rows. The emitter
exponent is x = 0.544, and the system flow rate is 8.05 lps.

What you need to do:

1. Determine the optimal manifold location.

2. Determine the location of the minimum downhill lateral pressure.
3. Calculate the required lateral inlet pressure head, Hl.
4. Calculate the minimum uphill lateral pressure head, (Hn’)uphill.
5. Do your work neatly for full credit on this problem.

3. (35 pts) A trickle irrigation system with a manifold inflow rate of 8.4 lps has an allowable
subunit pressure head variation of ∆Hs = 4.72 m. The calculated pressure variation
along the lateral pipes is ∆Hl = 2.44 m, and the total length of the manifold will be 290
m. There is a uniform ground slope of 1.73% in the manifold direction. The following
PVC pipe sizes are available:

Size I.D.
(inches) (inches)
1.5 1.610
2.0 2.067
2.5 2.469
3.0 3.068
4.0 4.000
6.0 6.000

2 of 10
 What you need to do:

1. Design the manifold, using up to four different pipe diameters.

2. Use the attached friction loss curves for the six available pipe sizes.
3. Determine appropriate manifold pipe sizes and lengths.
4. Do not allow the maximum velocity in each pipe size to exceed 2.0 m/s.
5. Do your work neatly for full credit on this problem.

Solutions:

1. Trickle Design Calculations

1. Select an appropriate emitter spacing, Se.

S'e = 0.8w = 0.8(2.01) = 1.61 m

2. Determine the number of emitters per tree, Np.

Sp 6.0
Np = = = 3.73
Se 1.61

3. Calculate percent wetted area, Pw.

⎛N S w⎞ ⎛ (3.73)(1.61)(2.01) ⎞
Pw = 100 ⎜ p e ⎟ = 100 ⎜ ⎟ = 43.0%
⎜S SP ⎟ ⎝ (6.0)(6.0)(0.78) ⎠
⎝ p r d⎠

which is within acceptable limits.

4. Calculate maximum net depth to apply per irrigation, dx.

Table 3.2: Zavg (for citrus) = ½(0.9 + 1.5) = 1.2 m.

MAD Pw
dx = Wa Z = (0.2)(0.43)(175)(1.2) = 18.1 mm
100 100
5. Calculate the average peak daily “transpiration” rate, Td.

Td = 0.1Ud Pd = 0.1(5.1) 78 = 4.50 mm/day

6. Calculate the maximum irrigation interval, fx. If fx ≥ 1 day, use f’ = 1day.

3 of 10
 dx 18.1
fx = = = 4.0 days
Td 4.50

Then, let f’ = 1 day (for design).

7. Calculate the net depth per irrigation, dn.

dn = Td f ' = (4.5)(1) = 4.5 mm/day

8. Select a reasonable target EU value.

Table 20.3: “point-source” water applicators with Np > 3 gives recommended EU

range of 90 to 95%. Choose EU = 92%.

9. Determine (ECe)max.

From Table 19.2, for a citrus (e.g. orange) crop, (ECe)max = 8 dS/m.

10. Determine the transmission ratio, Tr.

From Table 19.3, for a “deep-rooted” (Z > 1.5 m) crop and a “medium-textured”
soil: Tr = 1.00.

11. Calculate the leaching requirement, LRt.

EC w 0.89
LR t = = = 0.056
2 (ECe )max 2(8)

12. Calculate the gross depth to apply per irrigation, d.

For LRt < 0.1, the following equation is applied:

⎛d T ⎞ ⎛ (4.50)(1.00) ⎞
d = 100 ⎜ n r ⎟ = 100 ⎜ ⎟ = 4.89 mm/day
⎝ EU ⎠ ⎝ 92% ⎠

13. Calculate the gross volume of water per tree per day, G.

d 4.89
G= SpSr = (6.0)(6.0) = 176 liter/day/tree
f' 1
14. Calculate ha, corresponding to qa = 3.85 lph, in m of water head.

Apply the given emitter equation, and use 9.81 kPa/m:

4 of 10
 1/ 0.481
⎛ 1 ⎞ ⎛ 3.85 ⎞
ha = ⎜ ⎟⎜ ⎟ = 23.7 m
⎝ 9.81 ⎠ ⎝ 0.28 ⎠

15. Calculate the water application time, Ta.

G 176
Ta = = = 12.3 hrs/day
Npqa (3.73)(3.85)

16. Select the number of stations, Ns.

Two stations would require 2(12.3) = 24.6 hrs/day. Thus, there can be only one
station (Ns = 1) in this design.

17. Determine the minimum number of emitters per tree, Np’.

Here is a new equation for Np’:

⎡ ⎛ S + w − 2Se ⎞ ⎤
N'p = 2 ⎢ trunc ⎜ 1 + p ⎟⎥ + 1
⎣ ⎝ 2S e ⎠⎦

In this problem,

⎡ ⎛ 6.0 + 2.01 − 2(1.61) ⎞ ⎤

N'p = 2 ⎢ trunc ⎜ 1 + ⎟⎥ + 1
⎣ ⎝ 2(1.61) ⎠⎦
= 2 [ trunc(2.49)] + 1
= 5 emitters

Note that Np’ must be an integer value.

18. Calculate the system coefficient of variation, νs.

ν 0.0487
νs = = = 0.022
N'p 5

19. Calculate the minimum allowable emitter flow rate, qn.

qaEU (3.85)(92)
qn = = = 3.64 lph
100 (1 − 1.27νs ) 100 (1 − 1.27(0.022) )

5 of 10
which corresponds to a head of:

1/ 0.481
⎛ 1 ⎞ ⎛ 3.64 ⎞
hn = ⎜ ⎟⎜ ⎟ = 21.1 m
⎝ 9.81⎠ ⎝ 0.28 ⎠

20. Calculate the allowable subunit pressure head variation, ∆Hs.

∆Hs = 2.5 (ha − hn ) = 2.5 ( 23.7 − 21.1) = 6.50 m

21. Calculate the system capacity, Qs.

ANpqa (132)(3.73)(3.85)
Qs = 2.78 = 2.78 = 146 lps
NsSpSr (1)(6.0)(6.0)

22. Calculate the total gross seasonal depth to apply, Dg.

Assume a TR value of 1.00 (Table 19.4). Then, Es = EU = 92%. Seasonal

effective rain is assumed to be zero because the value is not given. Residual
soil moisture is given to be zero. Thus,

( ) (
Dn = U 0.1 Pd = 660 0.1 78 = 583 mm )
Then, gross seasonal depth is:

100Dn 100(583)
Dg = = = 671 mm
Es (1 − LR t ) 92(1 − 0.056)

23. Calculate the gross seasonal volume of irrigation water, Vs.

Dg A (671)(132)
Vs = = = 88.6 ha-m
1000 1000

24. Calculate the required number of operating hours per season, Ot.

Vs ⎛ 88.6 ⎞
Ot = 2778 = 2778 ⎜ ⎟ ≈ 1,690 hrs/season
Qs ⎝ 146 ⎠

6 of 10
2. Optimal Manifold Location

Number of emitters for the pair of laterals:

380 m
= 950 emitters
0.4 m/emitter

Total nominal discharge for the pair of laterals:

(950 emitters)(2.75 lph/emitter)

Qpair = = 0.726 lps
3600 s/hr

Friction loss gradient (Eq. 8.7a):

(0.726)1.75
J = 7.83(10)7 = 51.4 m/100 m
(17.8)4.75

Note that this is a high value for J, and is beyond the values given in Table 8.2.
The multiple-outlet factor, F, is 0.36 for 950 outlets. Take fe = 0 since the emitters
are in-line.

Friction loss for the pair of laterals:

(51.4)(0.36)(380)
(hf )pair = = 70.3 m
100
which is equal to about 100 psi.

Elevation change for the pair of laterals:

( ∆he )pair = (0.00761)(380) = 2.89 m

Ratio of elevation change to friction loss for the pair of laterals:

⎛ ∆he ⎞ 2.89
⎜ ⎟ = = 0.041
⎝ hf ⎠pair 70.3

Plot the above value (0.041) on the ordinate of the dimensionless friction loss graph
(Fig. 8.2) at x/L = 1.0, then draw a straight line from the origin to this point. Slide
the dimensionless curve for the downhill portion of the laterals until it is tangent to

7 of 10
the line, then trace it. Slide the curves up further so that the uphill curve passes
through the plotted point. Determine the intersection of the two curves:

x/L ≈ 0.52

Then, the downhill portion of the pair of laterals will have a length of:

Ld = (380)(0.52) ≈ 198 m

The uphill portion of the pair of laterals will have a length of:

Lu = 380 − 198 = 182 m

Graphically, the location of minimum pressure in the downhill lateral would be:

Lmin = 0.1(380) = 38 m

as measured from the end of the downhill lateral, or 198 – 38 = 160 m from the
manifold location.

The required lateral inlet head is:

3
α= ⎡⎣(0.52)3.75 + (1 − 0.52)3.75 ⎤⎦ = 0.112
4
and,
⎛x ⎞
hl = ha + α (hf )pair + ⎜ − 0.5 ⎟ ( ∆he )pair
⎝L ⎠

hl = 11.5 + 0.112 ( 70.3 ) + ( 0.52 − 0.5 )( 2.89 ) = 19.5 m

which is 191 kPa, or 28 psi.

The minimum pressure head in the uphill lateral is:

Friction loss gradient (Eq. 8.7a):

1.75
⎛ (182)(2.75) ⎞
⎜ ⎟
7 ⎝ (0.4)(3600) ⎠
J = 7.83(10) = 14.2 m/100 m
(17.8)4.75

Friction loss:

8 of 10
 (14.2)(0.36)(182)
(hf )up = = 9.30 m
100
Minimum pressure head:

(hn ' )up = ha − (hf )up − ( ∆he )up

= 19.5 − 9.30 − (182)(0.00761)
= 8.81 m

3. Manifold Design

The elevation change over the manifold length is:

∆Em = (290)(0.0173) = 5.02 m

The allowable head variation in the manifold is:

( ∆hm )a = 4.72 − 2.44 = 2.28 m

The sum of elevation change and allowable head variation is: 5.02 + 2.28 = 7.30 m.

Draw a line from the origin to (8.4, 5.02). Draw a parallel line from (0.0, 2.28) to
(8.4, 7.30). These two lines define the band within which the friction loss curves
should be contained.

Look at the six friction loss curves on the attached graph. The smallest four sizes
are too small to fit within the band. The largest (6 inches) is too large because the
friction loss curve goes outside the band. Start with the 4-inch pipe size, but not at
the point (8.4, 7.30); start at (8.4, 5.02) such that the curve can fit within the band.
Trace the curve for the 4-inch pipe on the graph.

Entrance velocity would be:

4Q 4(0.0084)
V4 −inch = = = 1.04 m/s
πD 2 ⎡⎛ 4 ⎞ ⎤
2

π ⎢⎜ ⎟ (0.3048)⎥
⎣⎝ 12 ⎠ ⎦

which is within acceptable limits.

Note that the 4-inch pipe could be used for the whole length of the manifold and still
accommodate the allowable pressure head variation. But part of the manifold

9 of 10
should be 3-inch pipe, thereby reducing the head variation further, and also
decreasing the pipe cost.

Trace the 3-inch friction loss curve on the graph, making it tangent to the lower line
(ground slope line). This curve intersects the 4-inch curve at a flow rate of about
3.45 lps. The entrance velocity to the 3-inch pipe would be:

4Q 4(0.00345)
V3 −inch = = = 0.72 m/s
πD 2 ⎡⎛ 3.068 ⎞ ⎤
2

π ⎢⎜ ⎟ (0.3048)⎥
⎣⎝ 12 ⎠ ⎦

which is also within acceptable limits.

Then, use two pipe diameters (4- and 3-inch) with the following lengths of pipe in
the manifold:

⎛ 3.45 ⎞
L3 −inch = 290 ⎜ ⎟ = 119 m
⎝ 8.4 ⎠

L 4 −inch = 290 − 119 = 171 m

where, of course, the upstream end has the 4-inch pipe. These lengths could be
rounded as necessary to accommodate the lateral spacing.

10 of 10
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004
Assignment #1 (100 pts)
Due: 15 Sep 04

Given:

Data for a field in Cache Valley, Utah:

Crop= Sweet corn

Topsoil= Sandy loam
Topsoil depth (m)= 0.5
Subsoil= Silt Loam
Subsoil depth (m)= 1.4
Field area (ha)= 25
MAD (%)= 35
Irrigation water salinity, ECw (dS/m)= 1.02
Application efficiency= 88%
Soil intake rate (mm/hr)= 14
Time to change sets (hrs)= 0.5
Lateral length (m)= 180
Lateral spacing, Sl (m)= 12

Weather data:

1. Go to http://climate.usu.edu/
2. Click on "Utah Climate Center Data (Use Microsoft Explorer)"
3. Look at the instructions for selecting a region
4. Select a region which includes Utah
5. Click on the "Update Station List" button
6. At the upper left you see a list of stations
7. Click on the list and type "L"
8. Scroll down further to "logan usu exp stn"
9. Under "Element," select "Maximum Air Temperature"
10. Check the "Full period of record" box
11. Click on the "Export Data" button at the top
12. Click on the "Update Output" button at the lower left
13. Type "Ctrl-A" to select all, then "Ctrl-C" to copy
14. Paste it into Notepad or Word
15. Clean up the data and import to Excel
16. Go back and select “Minimum Air Temperature" and get that data
17. Go back and select “Total Precipitation" and get that data, too

Note that values with “-99999,M” are missing.

Required:

• Perform calculations to answer all of the questions as shown in the table format on the next page
• Show your steps in logical order, and write down your assumptions (if any) in determining the
respective values
• Do you work neatly
Notes:

1. Many years of weather data for Logan, Utah (Experiment Station site), are given on the web site.
2. Determine the mean monthly values (Jan – Dec) of maximum daily air temperature for the entire
period of record (about 34 years). To do this in Excel, you may want to use functions like
COUNTIF and SUMIF.
3. Plot the mean monthly values of maximum air temperature and determine which month is the
warmest; this month will be used below as the peak-use (peak ET) month.
4. Use the precipitation data to calculate the 75% rainfall probability value for the peak-use month to
determine the net crop ET requirement during that month. This means you need to calculate the total
rainfall (inches) for the peak-use month for the 34 years of record. You can consider that the 75%
rainfall value is all “effective” rainfall for that particular month.
5. You may notice some problems with the data sets from the Utah Climate Center web site. Document
these problems and describe how you have dealt with them.
6. Assume that there will be only six days of irrigation per week (one day off), even during the peak-use
period.
7. Obtain needed soil, root depth, EC and ET information from tables in the text (Chapter 3), or from
another source (if so, name that source). Use average values where max-min ranges are given in
the tables.
8. Use Eq. 3.1 to calculate the maximum net application depth per irrigation.
9. Use Eq. 3.2 to calculate the maximum irrigation interval, then to calculate the net application depth.
10. Use Eq. 3.3 to calculate the leaching requirement.
11. Use Eqs. 5.3a and 5.3b to determine the gross application depth.
12. Use Eq. 5.4 to calculate system flow capacity.
Table Format for BIE 5110/6110 Assignment #1 (Fall 2004)

Given Values:

Crop: .........................................................................
Topsoil depth (m):.....................................................
Subsoil depth (m): ....................................................
Location ....................................................................
Field area (ha): .........................................................
MAD (%): ..................................................................
Irrigation water salinity, ECw (dS/m):........................
Application efficiency (%): ........................................
Soil intake rate (mm/hr)............................................
Time to change sets (hrs): .......................................
Lateral length (m) .....................................................
Lateral spacing (m)...................................................

Obtained from Tables and Weather Data:

Average Wa of topsoil (mm/m):................................

Average Wa of subsoil (mm/m): ...............................
Average root depth, Z (m):.......................................
Seasonal effective rainfall at 75% prob. (mm): .......
Peak ET (mm/day): ..................................................
Seasonal ET (mm): ..................................................
Salinity of soil extract, ECe (dS/m): ..........................

Calculated Values:

Average Wa of root zone (mm/m): ...........................

Maximum net depth per irrigation (mm): .................
Maximum irrigation interval (days):..........................
Nominal irrigation interval (days): ............................
Net depth per irrigation (mm):..................................
Days off in each irrigation:........................................
Operating time per irrigation (days): ........................
Leaching requirement: .............................................
Gross application depth per irrigation (mm):...........
Minimum set operating time (hrs) ............................
Nominal set operating time (hrs)..............................
Number of sets per day:...........................................
Area per 200-m lateral per irrigation (ha): ...............
Number of 200-m laterals required:.........................
Approximate number of irrigations per year:...........
System flow capacity (lps):.......................................
Solution:

To determine the mean monthly values of maximum daily air temperature, the
SUMIF function was used in Excel to key on the column with the month names for the 34
years of record. The COUNTIF function was used in the same way, for each month, to
determine the number of records in each month. Finally, for each month, the sum of
temperature values was divided by the corresponding record count to arrive at an average
monthly value:

Month Count Sum Avg Temp

Jan 32 1,041 32.5
Feb 32 1,228 38.4
Mar 33 1,605 48.6
Apr 33 1,914 58.0
May 32 2,173 67.9
Jun 33 2,588 78.4
Jul 33 2,892 87.6
Aug 34 2,954 86.9
Sep 31 2,358 76.1
Oct 32 2,006 62.7
Nov 32 1,480 46.2
Dec 32 1,096 34.3

Logan, Utah (1969-2002)

90

80
Average Monthly Tmax (deg F)

70

60

50

40

30

20

10

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

It is seen that the month of July has the highest maximum monthly temperature for
the 34 years of record, and is closely followed by August. This will be the month for which
peak ET will occur in Logan, Utah.
 To determine the rainfall probability for the month of July, the total daily rainfall
values for that month were summed up for each year of record. Following are the
tabulated results:

July Rain July Rain

Year (inch) Year (inch)
1969 0.34 1986 1.79
1970 0.67 1987 1.65
1971 0.08 1988 0.00
1972 0.21 1989 0.11
1973 0.98 1990 0.31
1974 0.11 1991 0.16
1975 1.46 1992 0.99
1976 0.83 1993 3.21
1977 0.87 1994 0.02
1978 0.11 1995 0.72
1979 0.58 1996 1.44
1980 1.18 1997 2.30
1981 0.49 1998 0.12
1982 1.40 1999 0.57
1983 1.52 2000 0.07
1984 1.59 2001 missing
1985 1.50 2002 0.38

Rain (inch) Relative

From To Frequency Frequency Up Down
0.00 0.25 10 0.303 1.000 0.303
0.26 0.50 4 0.121 0.697 0.424
0.51 0.75 4 0.121 0.576 0.545
0.76 1.00 4 0.121 0.455 0.667
1.01 1.25 1 0.030 0.333 0.697
1.26 1.50 4 0.121 0.303 0.818
1.51 1.75 3 0.091 0.182 0.909
1.76 2.00 1 0.030 0.091 0.939
2.01 2.25 0 0.000 0.061 0.939
2.26 2.50 1 0.030 0.061 0.970
2.51 2.75 0 0.000 0.030 0.970
2.76 3.00 0 0.000 0.030 0.970
3.01 3.25 1 0.030 0.030 1.000
Totals: 33 1.000

As seen in the table above, based on the 34 years of record, there is a 69.7%
probability that the total July rainfall in Logan, Utah, will be 0.26 inches or more (column
“Up”). There are really not enough data points to determine the 75% level of confidence,
so use 0.26 inches of rain in July (the peak ET month) with an approximately 70% level of
exceedance. This is equivalent to 6.6 mm.

Given Values:
Crop: ......................................................................... Sweet corn
Topsoil depth (m):..................................................... 0.5 m
Subsoil depth (m): .................................................... 1.4 m
Location .................................................................... Logan, Utah
Field area (ha): ......................................................... 25 ha
MAD (%): .................................................................. 35%
Irrigation water salinity, ECw (dS/m):........................ 1.02 dS/m
Application efficiency (%): ........................................ 88%
Soil intake rate (mm/hr)............................................ 14 mm/hr
Time to change sets (hrs): ....................................... 0.5 hrs
Lateral length (m) ..................................................... 180 m
Lateral spacing (m)................................................... 12 m

Obtained from Tables and Weather Data:

Average Wa of topsoil (mm/m):................................ 125 mm/m (Table 3.1, average for sandy loam)
Average Wa of subsoil (mm/m): ............................... 167 mm/m (Table 3.1, average for silt loam)
Average root depth, Z (m):....................................... 0.5 m (Table 3.2, average for sweet corn)
Seasonal effective rainfall at 70% prob. (mm): ....... 6.6 mm (for month of July only, not season)
Peak ET (mm/day): .................................................. 6.4 mm/day (Table 3.3, corn in “moderate” climate)
Seasonal ET (mm): .................................................. 559 mm (Table 3.3, corn in “moderate” climate)
Salinity of soil extract, ECe (dS/m): .......................... 2.5 dS/m (Table 3.5, sweet corn)

Calculated Values:
Note: most of the following is specifically for the peak-use period and does not dictate what the irrigation
scheduling might be throughout the growing season. These are calculations leading to system design.

Average Wa of root zone (mm/m):

The topsoil depth is given as 0.5 m, and we have 0.5 m for the average effective root depth of sweet corn,
so the subsoil Wa is not considered herein.

Use Wa = 125 mm/m

Maximum net depth per irrigation (mm):

MAD ⎛ 35 ⎞
dx = Wa Z = ⎜ ⎟ (125 )( 0.5 ) = 21.9 mm
100 ⎝ 100 ⎠
Maximum irrigation interval (days):

For July, we have determined that there is a 70% probability of a monthly total of 0.26 inches (6.6 mm), or
more, of rain. This comes to an average of 6.6/31 = 0.21 mm/day, which is very little rain. This fact,
together with the realization that the rain might not fall during the peak-use period, may lead us to
conclude the safer choice is to assume zero effective rainfall during the peak-use period.

dx 21.9 mm
fx = = = 3.42 days
Ud 6.4 mm/day
Nominal irrigation interval (days):
 f ' = trunc ( fx ) = 3 days

Net depth per irrigation (mm):

dn = f 'Ud = ( 3 days )( 6.4 mm/day ) = 19.2 mm

Days off & operating time per irrigation:

The specification in this case is for one day off per week, but with f’ = 3 days, we can assume that the one
day off will not fall within the three-day interval during the peak-use period, which might involve only two or
three irrigations. Thus, let f = f’ = 3 days.

Leaching requirement:

ECw 1.02
LR = = = 0.075
5ECe − EC w 5 ( 2.5 ) − 1.02

LR < 0.1; therefore, use Eq. 5.3a...

Gross application depth per irrigation (mm):

dn 19.2
d= = = 21.8 mm
(Ea /100 ) 0.88
Minimum set operating time (hrs):

With 21.8 mm to apply and a soil intake rate of 14 mm/hr, this gives 1.56 hrs minimum set time (so as
not to exceed the soil intake rate).

Nominal set operating time (hrs)

Make the nominal set time equal to 2.0 hours for convenience. With 0.5 hrs to move each set, there
are a total of 2.5 hrs/set.

Number of sets per day:

With 24 hrs per day, there can be 24/2.5 = 9.6 sets/day. Round this down to a whole number: 8 sets per
day, giving a total daily operations time of 8(2.5) = 20 hrs.

Area per 200-m lateral per irrigation (ha):

(3 days/irrigation)(8 sets/day) = 24 sets/irrigation

Lateral spacing on mainline is Sl = 12 m. Lateral length is 180 m. The area per lateral is:

(12 m/set)(24 sets)(180 m/lateral) = 5.18 ha/lateral

Number of 180-m laterals required:

 25 ha
= 4.83 laterals
5.18 ha/lateral
Round up to a whole number to obtain 5 laterals for the system. Alternatively, round up to 6 laterals if
laterals will operate on both sides of the mainline, thereby balancing the number of laterals on each
side.

Approximate number of irrigations per year:

Assuming zero effective rainfall during the growing season:

U − Pe 559 mm - 0 mm
= = 29.1 irrigations
dn 19.2 mm/irrig
We could do a seasonal analysis of the probability of rainfall exceedance, but we can already surmise
that most of the rain will fall during times other than the peak-use period. So, to be conservative in our
design, we assume no seasonal effective rainfall. On the other hand, it is very unlikely that a farmer in
Logan, Utah would irrigate a corn field 29 times in a season, which could indicate a nonzero contribution
from rain and or the possibility that the effective root depth for sweet corn is somewhat more than 0.5 m.
Also, there may be a significant residual water content (from snowmelt and rain) in the soil at the
beginning of the growing season.

System flow capacity (lps):

At 8 sets/day and 2.0 hours set time, there are 16 hours of system operation per day:

Qs = 2.78
Ad
= 2.78
( 25 ha )( 21.8 mm ) = 31.6 lps
fT ( 3 days )(16 hrs/day )
This gives a capacity of 31.6/25 = 1.26 lps per ha, which is a reasonable value.
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #2 (100 pts)

Due: 29 Sep 04

Given:

A solution for the economic selection of pipe sizes is needed. The solution is to be
based on the following data:

ƒ interest rate, i = 2.4%

ƒ inflation rate, e = 0.1%
ƒ useful system life, n = 22 years
ƒ area irrigated, A = 58 ha
ƒ gross annual depth, d = 890 mm/year
ƒ maximum system capacity, Q = 210 lps
ƒ motor efficiency, Em = 94%
ƒ pump efficiency, Ep = 88%
ƒ avg cost of electricity for pumping = 0.06104 $/kWh
ƒ purchase cost of PVC pipe:

Size O.D. I.D. Wall Price

(inches) (inches) (inches) (inch) ($/20 ft)
6 6.625 6.031 0.297 35.10
8 8.625 7.943 0.341 53.22
10 10.750 9.976 0.387 75.81
12 12.750 11.890 0.430 98.08
14 14.000 13.072 0.464 116.95
16 16.000 14.940 0.530 154.33
18 18.000 16.809 0.595 198.39

ƒ In the above table: “O.D.” is outside diameter; “I.D.” is inside diameter;

and “Wall” is the pipe wall thickness

Required:

• Determine the cutoff flow rate values for each adjacent pair of pipe sizes (you are
not required to graph the pipe selection chart)
• The solution is to include the PVC pipe with nominal sizes 6, 8, 10, 12, 14, 16, and
18 inches, as shown above.
• You can use either the Hazen-Williams or Darcy-Weisbach equations.
• For PVC pipe, use a Hazen-Williams “roughness” factor of C = 150.
• Or, for Darcy-Weisbach, use the Blasius equation for smooth pipe.
• Assume that MAC is negligible.
Solution:

BIE 5110/6110
Economic Pipe Selection Method
Assignment #2, Fall 2004

Given data:
useful life: 22 years inflation: 0.001 per year motor efficiency: 0.94
interest rate: 0.024 per year area: 58 ha pump efficiency: 0.88
depth: 0.89 m/year electricity: 0.06104 $/kWh
capacity: 210 lps Hazen-Williams: 150 C factor

Pipe purchase prices:

Size O.D. I.D. Wall Price

(inches) (inches) (inches) (inch) ($/20 ft)
6 6.625 6.031 0.297 35.10
8 8.625 7.943 0.341 53.22
10 10.750 9.976 0.387 75.81
12 12.750 11.890 0.430 98.08
14 14.000 13.072 0.464 116.95
16 16.000 14.940 0.530 154.33
18 18.000 16.809 0.595 198.39

Capital Recovery Factor: CRF= 0.0590

Uniform annual pipe cost:

Size UAC
(inches) ($/20 ft/yr) ($/100 ft/yr)
6 2.07 10.36
8 3.14 15.71
10 4.48 22.38
12 5.79 28.95
14 6.90 34.52
16 9.11 45.56
18 11.71 58.56

Operating hours per year:

Ot= 683 hrs/year

Present annual energy cost:

E= 50.38 $/kW/yr (MAC said to be negligible)

Equivalent annual energy cost:

EAE= 1.0096
E'= 50.87 $/kW/yr
Difference in WHP between adjacent pipe sizes:

Sizes delta WHP

(inches) kW/100 ft
6 and 8 0.10515
8 and 10 0.13109
10 and 12 0.12923
12 and 14 0.10950
14 and 16 0.21691
16 and 18 0.25567

Difference in J between adjacent pipe sizes:

Sizes delta J
(inches) m/100 m (note: m/100 m is equivalent to ft/100 ft)
6 and 8 0.16756 (however, must use 102 for lps or 3,960 for gpm)
8 and 10 0.20889
10 and 12 0.20593
12 and 14 0.17449
14 and 16 0.34566
16 and 18 0.40743

Threshold flow rate between adjacent pipe sizes:

Sizes Section flow rate

(inches) lps gpm (note: Hazen-Williams equation)
6 and 8 11.2 177.4
8 and 10 27.4 434.3
10 and 12 53.8 852.8
12 and 14 99.1 1,570.0
14 and 16 160.0 2,535.6
16 and 18 260.9 4,134.8
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #3 (100 pts)

Set Sprinkler Lateral Design

Given:

• Sprinkler spacing, Se, of 12 m

• Lateral length of 252 m
• Riser height of 1.30 m
• Lateral will run downhill along a ground slope of 0.38%
• Nominal sprinkler discharge of 22 lpm at 2.08 atm pressure
• Nominal aluminum pipes sizes of 2, 3, 4 and 5 inches are available
• Assume a constant value of qa for each sprinkler along the lateral
• See Table 8.1 in the textbook (Keller & Bliesner) for pipe inside diameters

Required:

• Calculate the required aluminum lateral pipe size

• Round up to the nearest available pipe size
• Calculate the required lateral inlet pressure
• Calculate the location of minimum pressure in the pipe
• Calculate the pressure in the lateral pipe at the downstream end
• Calculate the percent pressure variation in the lateral pipe (%)
Solution:

If Hazen-Williams equation is used, the C value will be 130.

The number of sprinklers on the lateral is:

L 252 m
Nn = = = 21
Se 12 m

There will be 21 sprinklers per lateral. The multiple-outlet friction loss factor for 21
sprinklers is:

1 ⎡ 4 ⎤
F = 0.351 + 1+ = 0.38
2(21) ⎣ 13(21) ⎥⎦
⎢

The lateral inflow rate is:

(21)(22 lpm)
Ql = Nnqa = = 7.70 lps
60 s / min

The elevation change along the length of the lateral is:

⎛ −0.38 ⎞
∆he = ⎜ ⎟ (252 m) = −0.958 m
⎝ 100 ⎠

The nominal sprinkler pressure is ha = 2.08 atm, or,

ha = (2.08 atm)(10.34 m/atm) = 21.5 m

The allowable friction loss gradient is:

⎛ 0.20ha − ∆he ⎞
Ja = 100 ⎜ ⎟
⎝ FL ⎠
⎛ 0.2(21.5 m) − ( −0.958 m) ⎞
= 100 ⎜ ⎟
⎝ (0.38)(252 m) ⎠
= 5.49 m/100m

The minimum pipe ID is:

 0.205
⎡16.42(10)6 ⎛ 7.70 lps ⎞1.852 ⎤
D=⎢ ⎜ 130 ⎟ ⎥ = 7.27 cm
⎣⎢ 5.49 ⎝ ⎠ ⎦⎥

or 7.27 cm/2.54 = 2.86 inches.

From Table 8.1 (page 140) in the textbook, the nominal size of the aluminum lateral pipe is
the outside diameter, and the wall thickness is 0.05 inches. The 2” size has ID = 1.9”, the
3” size has ID = 2.9”, and so forth. From our calculated minimum diameter, we must
round up to the 3” nominal size. Then, the inside diameter is (2.9 inch)(2.54) = 7.37 cm.

The actual friction loss along the lateral will be:

1.852
⎛ 7.70 lps ⎞
J = 16.42(10) ⎜ 6
⎟ (7.37 cm)−4.87 = 5.22 m/100m
⎝ 130 ⎠

JFL (5.22)(0.38)(252 m)
hf = = = 5.00 m
100 100

The required lateral inlet pressure is:

hl = ha + 0.75hf + 0.5∆he + hr
= 21.5 m + 0.75(5.00 m) + 0.5( −0.958 m) + 1.3 m
= 26.1 m ( 2.52 atm )

The minimum pressure in the lateral pipe is located at a distance x from the inlet (qa is 22
lpm/60 = 0.367 lps; D is 73.7 mm):

x=
Se ⎡
qa ⎣ (
Ql − 3(10)−7 C( −S)0.54 D2.63 ⎤
⎦ )
=
12 ⎡
0.367 ⎣
(
7.70 − 3(10)−7 130(0.38)0.54 (73.7)2.63 ⎤
⎦ )
= 190 m

according to the equations in the lecture notes on pages 64 and 65. This “x” value
(distance) is confirmed (approximately) by assuming an inlet pressure head of 26.1 m, a C
value of 130, and so forth, calculating the head loss segment-by-segment in a
spreadsheet:
 Sprinkler Distance Q hf head
Position (m) (lps) (m) (m)
0 0 7.70 0.0000 26.10
1 12 7.33 0.6262 25.52
2 24 6.97 0.5721 24.99
3 36 6.60 0.5203 24.52
4 48 6.23 0.4707 24.09
5 60 5.87 0.4234 23.72
6 72 5.50 0.3785 23.38
7 84 5.13 0.3358 23.09
8 96 4.77 0.2955 22.84
9 108 4.40 0.2576 22.63
10 120 4.03 0.2221 22.45
11 132 3.67 0.1891 22.31
12 144 3.30 0.1585 22.20
13 156 2.93 0.1304 22.11
14 168 2.57 0.1048 22.05
15 180 2.20 0.0819 22.02
16 192 1.83 0.0615 22.00
17 204 1.47 0.0439 22.00
18 216 1.10 0.0290 22.02
19 228 0.73 0.0170 22.05
20 240 0.37 0.0080 22.09
21 252 0.00 0.0022 22.13

whereby the minimum head is found somewhere between 192 and 204 m from the lateral
inlet. The difference in the calculations are due to the F-factor for multiple outlets. It is also
seen that the maximum pressure head is at the lateral inlet (hmax = 26.1 m), which might
have been expected because the ground slope is very small.

Thus, the minimum pressure is at a distance of approximately 190 m from the lateral inlet,
which is nearest sprinkler number 190/12 ≈ 16.

The pressure in the pipe at the downstream end will be:

hend = hl − hf − ( −∆he )
= 26.1 − 5.0 + 0.958 = 22.1 m

The head loss due to friction from sprinkler 16 to the end of the lateral is estimated as (from
Table 8.7, F = 0.46 for 21 – 16 = 5 outlets):

1.852
⎛ (0.367)(5) lps ⎞
Jx −end = 16.42(10) ⎜
6
⎟ (7.37 cm)−4.87 = 0.366 m/100m
⎝ 130 ⎠

where the pipe length is 5(12 m) = 60 m,

 JFL (0.366)(0.46)(60 m)
(hf )x−end = = = 0.101 m
100 100

The friction loss from the inlet to distance x = 190 m is, then, 5.00 – 0.101 m = 4.90 m. The
elevation change from the inlet to 190 m is (-0.0038)(12)(16) = -0.730 m. Finally, the
minimum pressure head in the lateral pipe is:

hmin = 26.1 − 4.90 − ( −0.730) = 21.9 m

The pressure (or head) variation is:

hmax − hmin 26.1 − 21.9

∆P = = = 0.20
ha 21.5

which is 20%. This just meets our design criterion.

 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #4 (100 pts)

Mainline Design
Due: 13 Oct 04

Given:

• A large rectangular field, 1,200 m long and 1,000 m wide

• Periodic-move sprinkler laterals with buried mainline pipe
• Four sprinkler laterals operate, each covering ¼ of the field area
• Two laterals are on one side of the mainline, and two on the other side
• All four laterals move in the same direction when changing sets
• The mainline will run down the middle of the field, 1,200 m long
• The mainline will run uphill at a uniform slope of 0.387%
• System capacity is Qs = 135 lps
• PVC pipe sizes given in Table 8.5 are the available sizes
• Pressure available at the upstream end of the mainline is 389 kPa
• Required lateral inlet pressure is 275 kPa
• Laterals connect to mainline through hydrant valves
• Hydrant hydraulic loss to from mainline to lateral inlet is 25 kPa

Required:

• Design the mainline to consume all of the available head

• Use the basic procedures from Lecture 9 and Chapter 10
• Consider the critical lateral positions
• Determine mainline pipe diameters and respective lengths of each size
• Check operational velocity limits in the mainline pipe
• Do all calculations in metric units (lengths in m, flow in lps, and diameter in mm)
• Do your work neatly and logically – make it understandable to another engineer
• Include brief comments, as necessary, about design details and decisions
Solution:

(1) Extreme Lateral Position

• Recognize that the extreme lateral position is when two laterals are at the mid-point
of the mainline, and the other two at the uphill end of the mainline
• This is because the mainline runs uphill, so pressure must decrease monotonically
from upstream to downstream along the mainline pipe
• Also, the critical point is the uphill end of the mainline, because that must be the
location of minimum pressure in the mainline
• Thus, for the extreme lateral position, the two laterals at the mid-point will have
more than enough pressure if the last two laterals have just the required pressure
• Mainline design, then, should focus on providing Pmin = 275 + 25 kPa in the pipe
• These facts should be obvious

1,200 m

389 kPa Pmin

mainline

uphill @ 0.387%
(2) Allowable Loss due to Friction

• The allowable friction loss is the available pressure at the mainline inlet, minus the
elevation change, hydrant loss, and required lateral inlet pressure
• As in the example mainline design in the lecture notes, do not consider minor losses
due to flow past closed hydrants along the mainline
• And, as in the example problems in the lecture notes, it may help to look at this
problem using a schematic diagram
• First, convert pressures to heads and determine the elevation change along the
length of the mainline:

Mainline inlet head: 389 kPa/9.81 = 39.65 m

Elevation change:

∆he = 0.00387(1,200) = 4.64 m

Lateral inlet pressure plus hydrant loss:

275 + 25
hl + hhydrant = = 30.58 m
9.81

• Let L1 be the length of pipe diameter D1, and L2 the length for diameter D2

L1 L2
(hf)a
39.65 m
30.58 m

4.64 m
1,200 m
• Allowable loss due to friction along the entire length of the mainline pipe:

(hf )a = 39.65 − 30.58 − 4.64 = 4.43 m

(3) Selection of Mainline Pipe Diameters

• Calculate the required mainline pipe inside diameter assuming only one pipe size
• Recognize that at the extreme lateral position, the full system flow rate goes from
the beginning of the mainline to the mid-point, where only half the system flow rate
continues to the end of the mainline
• Make the allowable friction loss equal to the actual friction loss

J1 (L / 2 ) J2 (L / 2 )
( hf ) a = +
100 100
where L = 1,200 m; and (hf)a = 4.43 m

⎡ 1.217(10)12 D−4.87 ⎤ ⎡ 1.852 ⎛ Qs ⎞ ⎤

1.852

4.43 = 6 ⎢ ⎥⎢ s Q + ⎜ 2 ⎟ ⎥
⎣ C1.852 ⎦ ⎣⎢ ⎝ ⎠ ⎦⎥

where Qs = 135 lps; and C = 150 (plastic pipe).

• Solving the above equation, D = 326 mm (12.8 inches).

• This is slightly larger than the 12” pipe (ID = 308.1 mm) in Table 8.5.
• However, the velocity in the 12” pipe at 135 lps would be:

4Q 4(0.135)
V= = = 1.81 m/s
πD 2
π(0.3081)2

which is not too high, but as seen above, the friction loss would be too high

• It might also be noted that 135 lps is more than 2,000 gpm, a flow rate for which an
irrigation system would almost always use 12” or 15” nominal pipe size
• Based on the preceding, try 15” pipe (Table 8.6) for the first half of the mainline:

1.852
⎛Q ⎞ ⎛L⎞
(hf )15 " = 1.217(10) ⎜ s ⎟
10
D−4.87 ⎜ ⎟
⎝ C⎠ ⎝ 2⎠
1.852
⎛ 135 ⎞ ⎛ 1,200 ⎞
(hf )15 " = 1.217(10) ⎜10
⎟ (369.7)−4.87 ⎜ ⎟ = 1.88 m
⎝ 150 ⎠ ⎝ 2 ⎠
• This leaves (hf)a – (hf)15” = 4.43 – 1.88 = 2.55 m allowable head loss in the second
half of the mainline
• The allowable friction loss gradient in the second half of the mainline is:

⎛ 2.55 ⎞
Ja = 100 ⎜ ⎟ = 0.425 m
⎝ 600 ⎠

• At ½ Qs = 67.5 lps, this is very close to the J value for the 10” pipe in Table 8.5
• Using that 10” pipe,

1.852
⎛ 67.5 ⎞
(hf )10 " = 1.217(10) ⎜
10
⎟ (259.7)−4.87 ( 600 ) = 2.90 m
⎝ 150 ⎠

which is greater than the allowable loss of 2.55 m, but close

• Using the 15” pipe on the first half of the mainline, and 10” pipe on the second half,
the total friction loss would be: (hf)total = 1.88 + 2.90 = 4.78 m
• Then, the pressure at the last pair of laterals would be:

Pmin = 389 − 25 − 9.81(4.64 + 4.78) = 272 kPa

which is very close to the required 275 kPa at the lateral inlets

(4) Design Summary

• The design could involve three pipe sizes along the mainline
• But, in this case it works out well to use two sizes: 15” SDR 41 PIP for the first half
of the mainline, and 10” SDR 41 IPS for the second half of the mainline
• Thus, there will be 600 m of 15” mainline, and 600 m of 10” mainline
• The pressure will be just about right at the end of the mainline, only 3 kPa below
that which is required
• The pressure at other lateral positions will be more than enough

(5) Notes

• Tables 8.5 and 8.6 do not use the same friction loss equation
• Table 8.6 is based on Hazen-Williams with C = 155
• Minor losses past closed hydrants were not considered – if they were, it might be
necessary to use a combination of 12” and 10” pipe in the second half of the
mainline
• It might also be necessary to use a combination of 12” and 10” pipe in the second
half of the mainline if we include a safety factor for uncertainties
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #5 (100 pts)

Minor Losses & Pumps
Due: 13 Oct 04

Given:

• A rectangular field to be sprinkler irrigated

• There will be four laterals, two on each side of the mainline
• All of the periodic-move laterals will move in the same direction
• At the beginning of an irrigation, two laterals are at the upstream end of the mainline
(first hydrant) and the other two are at the mid-point of the mainline
• The nominal hydrant diameter, and the inside diameter, is 4 inches
• Each lateral irrigates ¼ of the field area, and each is 600 ft in length
• Required lateral inlet pressure head is Pl = 44 psi
• The mainline will be 5-inch aluminum (see Table 8.4), with hydrant values spaced at
Sl = 30 ft along the 1,320-ft length of the mainline
• The first hydrant is located 30 ft from the beginning of the mainline, and the last
hydrant is 1,320 ft from the beginning of the mainline
• The sprinkler system will operate 12 hrs/day, and the irrigation interval during the
peak-use period is 7 days
• The gross depth to apply per irrigation is 1.9 inches
• The mainline slopes downhill at a uniform slope of 0.18%
• See the table in Chapter 8 for the roughness height of aluminum pipe
• See Table 11.1 for minor loss coefficients, Kr

Required:

• What is the field area, in acres?

• What is the system capacity, in gpm?
• Will the velocity in the mainline acceptable, or too high?
• Use the Darcy-Weisbach equation for pipe friction loss
• Make a graph of required (minimum) pressure at the upstream end of the mainline
as a function of lateral position, such that the minimum lateral inlet pressure (for
each lateral position) is exactly 44 psi
• This means that one pair of laterals will have exactly 44 psi inlet pressure, while the
other pair will have a slightly higher inlet pressure, and this will be the case for each
lateral position
• Note that there will be 1,320/(2*30) = 22 different positions lateral for each pair of
laterals
• On the graph, you can call the lateral positions “1,” “2,” “3,” … “21,” and “22.”
• Make note of any assumptions and of references which you use to obtain data
Solution:

1. Field Area:

The field is given to be rectangular. Note that there are 43,560 ft2/acre. The irrigated
area is the length of the mainline (1,320 ft) multiplied by twice the length of one lateral
(2 x 600 ft):

(1,320)(1,200)
= 36.4 acres
43,560

2. System Capacity:

Use Eq. 5.4 and the given data:

Ad (36.4)(1.9)
Qs = 453 = 453 = 373 gpm
fT (7)(12)

3. Velocity Checks:

Table 8.4: 5-inch aluminum pipe has an inside diameter of 4.900 inches (0.408 ft).
Note that the maximum recommended velocity, in general, for sprinkler systems is 5 to
7 fps.

3(a). Full system capacity in the mainline:

Qs 4(373 gpm)
VQs = = = 6.36 fps
A π (60 s/min)(7.481 gal/ft 3 )(0.408 ft)2

3(b). Half system capacity in the mainline:

Qs 2(373 gpm)
VQs /2 = = = 3.18 fps
2 A π (60 s/min)(7.481 gal/ft 3 )(0.408 ft)2

3(c). Half system capacity through a hydrant valve:

Qs 2(373 gpm)
Vhydrant = = = 4.77 fps
2 A π (60 s/min)(7.481 gal/ft 3 )(0.333 ft)2

All of the above velocities are below 7 fps, so they are found to be acceptable.
4. Reynolds Numbers and Darcy-Weisbach f:

The Reynolds number for a circular pipe is defined as:

VD 4Q
Re = =
ν πD ν
Assume a water temperature of 10°C. From the table on page 126 of the lecture notes
(or from any other reference), the kinematic viscosity at this temperature is ν =
1.306(10)-6 m2/s.

4(a). Full system capacity in the mainline:

4(373 gpm)(0.3048 m/ft)2

( R e )Q = ≈ 184,000
s
π(448.86 gpm/cfs)(0.408 ft)(1.306E-6 m2 /s)

4(b). Half system capacity in the mainline:

2(373 gpm)(0.3048 m/ft)2

(Re )Q /2 = ≈ 92,000
s
π(448.86 gpm/cfs)(0.408 ft)(1.306E-6 m2 /s)

From the table on page 138 (Chapter 8) of the textbook, the roughness height of
aluminum pipe (with couplers as an equivalent length of pipe) is 0.005 ft. Then, from
the Swamee-Jain equation:

fQs = 0.0213
and,
fQs/2 = 0.0225

5. Velocity Heads:

There are three different velocity heads to be considered, based on the three velocities
given in 3(a) – 3(c) above. These are:

5(a). Full system capacity in the mainline:

VQ2s (6.36 fps)2

= = 0.628 ft
2g 2(32.2 ft/s2 )
 5(b). Half system capacity in the mainline:

VQ2s /2(3.18 fps)2

= = 0.157 ft
2g 2(32.2 ft/s2 )

5(c). Half system capacity through a hydrant valve:

2
Vhydrant (4.77 fps)2
= = 0.353 ft
2g 2(32.2 ft/s2 )

6. Minor Loss Coefficients:

From Table 11.2 for a 4-inch aluminum hydrant valve:

Flow Path Kr
Past closed hydrant 0.5
Past open hydrant 0.6
Through open hydrant 7.5

7. Calculating Hydraulic Losses:

At the start of an irrigation, one pair of laterals is at the first hydrant (#1), which is 30 ft
from the beginning of the mainline. The second pair of laterals is at a distance of 660 ft
from the first pair, at hydrant #23. At each subsequent set, the laterals move Sl = 30 ft
down the mainline.

The second set will find the first pair of laterals at hydrant #2, and the second pair at
hydrant #24. Finally, the last set of the irrigation will have the first pair at hydrant #22,
and the second pair at hydrant #44 (the last one on the mainline).

The minimum pressure head required in the mainline pipe at a open hydrant is the
required lateral inlet pressure head of (44 psi)(2.31 ft/psi) = 102 ft, plus the head loss
through the hydrant valve, which is:

V2 ⎛ 4.772 ⎞
(hf )hydrant = Kr = 7.5 ⎜ ⎟ = 2.65 ft
2g ⎝ 2(32.2) ⎠

Then, the minimum pressure head required in the mainline pipe at a open hydrant is:

hmain = 102 + 2.65 ≈ 105 ft

which is constant for any lateral position.

The following figure gives a schematic plan view of the field area:

1,320 ft

1,200 ft
mainline

downhill @ 0.18%

Note that for every one of the 22 lateral positions, the second pair of laterals is always
660 ft downstream of the first pair of laterals.

7(a). Considering the First Pair of Laterals

Make a table of lateral positions in which the number of upstream closed laterals
increases by one for each new lateral position (because the pair of laterals moves
further from the upstream end of the mainline). Thus, the pipe friction loss and the
minor losses due to flow past a closed hydrant increase with each lateral position. On
the other hand, the change in elevation partially offsets these friction losses. Note that
from the upstream end of the mainline to the first pair of laterals, the discharge is equal
to the entire system flow rate. Consider the following table:
 First Pair of Laterals
Lateral Distance to Elev Pipe hf Number of (hf)minor Req'd at Mainline
st
Position 1 Pair (ft) Change (ft) (ft) US hydrants (ft) Inlet (ft)
1 30 -0.054 0.98 0 0.00 105.9
2 60 -0.108 1.97 1 0.31 107.2
3 90 -0.162 2.95 2 0.63 108.4
4 120 -0.216 3.93 3 0.94 109.7
5 150 -0.270 4.92 4 1.26 110.9
6 180 -0.324 5.90 5 1.57 112.1
7 210 -0.378 6.88 6 1.88 113.4
8 240 -0.432 7.87 7 2.20 114.6
9 270 -0.486 8.85 8 2.51 115.9
10 300 -0.540 9.84 9 2.83 117.1
11 330 -0.594 10.82 10 3.14 118.4
12 360 -0.648 11.80 11 3.45 119.6
13 390 -0.702 12.79 12 3.77 120.9
14 420 -0.756 13.77 13 4.08 122.1
15 450 -0.810 14.75 14 4.40 123.3
16 480 -0.864 15.74 15 4.71 124.6
17 510 -0.918 16.72 16 5.02 125.8
18 540 -0.972 17.70 17 5.34 127.1
19 570 -1.026 18.69 18 5.65 128.3
20 600 -1.080 19.67 19 5.97 129.6
21 630 -1.134 20.65 20 6.28 130.8
22 660 -1.188 21.64 21 6.59 132.0

7(b). Considering the Second Pair of Laterals

Do the same thing as for the first pair of laterals, but considering that part of the
mainline has the full system flow rate, and part has only half of the system flow rate.
Also, the minor loss due to “line flow” past one open hydrant (location of the first pair of
laterals) must be added to the head losses.

The losses from the second pair of laterals to the upstream end of the mainline must be
added to the 105-ft head requirement (see above) in the mainline at the location of the
second pair of laterals. These losses include pipe friction and minor losses. Consider
the following table (next page):

7(c). Extreme Position

It is seen that for each of the 22 lateral positions, the second pair of laterals require a
higher pressure head at the upstream end of the mainline. This is because the downhill
slope of the mainline is very small, so the friction losses dominate the pressure variation
along the mainline. Thus, the following graph is for the required pressure head at the
upstream end of the mainline from the perspective of the second (downstream) pair of
laterals for each position, thereby giving more than enough pressure in the mainline at
the location of the first pair of laterals.
 Second Pair of Laterals
Distance to Distance to Elev Pipe Number of closed (hf)minor Req'd at Mainline
Position 1st Pair (ft) 2nd Pair (ft) Change (ft) hf (ft) US hydrants (ft) Inlet (ft)
1 30 690 -1.242 6.70 21 2.10 112.6
2 60 720 -1.296 7.68 22 2.42 113.8
3 90 750 -1.350 8.67 23 2.73 115.0
4 120 780 -1.404 9.65 24 3.05 116.3
5 150 810 -1.458 10.63 25 3.36 117.5
6 180 840 -1.512 11.62 26 3.67 118.8
7 210 870 -1.566 12.60 27 3.99 120.0
8 240 900 -1.620 13.58 28 4.30 121.3
9 270 930 -1.674 14.57 29 4.62 122.5
10 300 960 -1.728 15.55 30 4.93 123.8
11 330 990 -1.782 16.53 31 5.24 125.0
12 360 1,020 -1.836 17.52 32 5.56 126.2
13 390 1,050 -1.890 18.50 33 5.87 127.5
14 420 1,080 -1.944 19.48 34 6.19 128.7
15 450 1,110 -1.998 20.47 35 6.50 130.0
16 480 1,140 -2.052 21.45 36 6.81 131.2
17 510 1,170 -2.106 22.43 37 7.13 132.5
18 540 1,200 -2.160 23.42 38 7.44 133.7
19 570 1,230 -2.214 24.40 39 7.76 134.9
20 600 1,260 -2.268 25.39 40 8.07 136.2
21 630 1,290 -2.322 26.37 41 8.38 137.4
22 660 1,320 -2.376 27.35 42 8.70 138.7

140
Required US Mainline Head (ft)

135

130

125

120

115

110
0 2 4 6 8 10 12 14 16 18 20 22
Sprinkler Lateral Position
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #6 (100 pts)

Pump Characteristic Curves
Due: 20 Oct 04

Given:

• A graph of pump characteristic curves for Cornell Model 4HS

• Notice the abscissa scale and the note saying that the pump curves are “flat” from Q
= 0 to Q = 100 gpm
• The desired operating point on your system curve is Q = 550 gpm at TDH = 72 ft
• The pump will take water from a large concrete-lined canal with water surface which
can vary from 1,330 to 1,333 ft above mean sea level
• The suction side of the pump has a screen, 8.0 ft of 6-inch SDR41 IPS PVC pipe,
and one 45-degree elbow
• The pump itself is located at an elevation of 1,338 ft above mean sea level
• The crop to be grown in the irrigated field is alfalfa

Required:

1. What is the nominal pump speed (RPM)?

2. What is the pump efficiency at the desired operating point?
3. What is the WHP at the desired operating point?
4. What is the calculated BHP at the desired operating point? Does it match the BHP
value given by the manufacturer on the graph?
5. Select an impeller diameter (inches) from the options given on the graph
6. What is the required pump speed (RPM) such that the desired operating point is the
actual operating point?
7. What is the required impeller trim (reduction in D) if the pump speed cannot be
changed, such that the desired operating point is the actual operating point?
8. What is the NPSHr at the desired operating point?
9. What is the NPSHa at the desired operating point?
10. Do you expect the pump to cavitate at the desired operating point?
Solutions:

1. The nominal pump speed is listed on the Cornell sheet as 1200 RPM. The exact
speed for the pump curves is 1175 RPM.
2. From the graph, at 550 gpm and 72 ft head, the efficiency is approximately 75%, to
the nearest whole number; or more generally, the efficiency is between 74 and
75%.
3. At the desired operating point, the WHP is:

QH (550)(72)
WHP = = = 10.0 HP
3956 3956
or, 7.46 kW.

4. The calculated BHP is:

WHP 10.0
BHP = = = 13.3 HP
Epump 0.75

or, 9.94 kW. Interpolating on the graph, the BHP appears to be approximately 14
HP, or perhaps slightly less than 14 HP. Thus, the calculated value agrees fairly
well with the graphical value, given the need for interpolation “by eye.” Also, the 14
HP estimate is sufficient to determine a power unit (motor) to drive this pump.

5. Select the nearest pump curve which is above the desired operating point (unless a
pump curve is only slightly below the desired operating point). In this case, choose
the curve for the 14” nominal diameter.

6. For this, apply the procedure given in the lecture notes. Develop some points for
the “equal efficiency” curve, and make them near the desired operating point such
that the curve will intersect with the pump curve.

Q1 H1
(gpm) (ft)
560 74.6
570 77.3
580 80.1
590 82.9
600 85.7

After graphing the points from the above table, the intersection with the 14” pump
characteristic curve is approximately:

H3 = 77 ft
Q3 = 568 gpm
 Finally, reduce the pump speed as follows:

⎛ 550 ⎞
Nnew = 1175 ⎜ ⎟ = 1138 RPM
⎝ 568 ⎠

Note that the intersection (Q3, H3) is usually close to the actual operating point
without a change in speed, N, and the value of Q3 is not much different than Q2. In
this case, the adjustment in speed is slight and it might not be worth the expense to
gear down to Nnew; instead, it might be better to accept the actual operating point
without changing speeds.

To know the actual operating point without changing speeds, we would need
sufficient information to develop the system curve.

7. The intersection (Q3, H3) is already known from the previous calculations, so
calculate the required impeller diameter straight away as follows:

⎛ 550 ⎞
Dnew = 14 ⎜ ⎟ = 13.6 inches
⎝ 568 ⎠

where it is assumed that the nominal diameter is the actual standard diameter. The
table below the graph shows maximum impeller diameters, but to apply them it
would be necessary to consult the manufacturer to be sure.

Note that the above impeller trim (0.4”) is very slight.

8. Extrapolating in the manufacturer’s curves, at the desired operating point,

NPSHr ≈ 5 ft.

9. Determine NPSHa as given in the lecture notes:

Maximum static lift is given as:

(hlift)max = 1338 – 1330 = 8 ft

Average atmospheric pressure head:

(hatm)avg = 10.3 – 0.00105 (1,338)(0.3048) = 9.87 m (32.4 ft)

Assuming a water temperature of 10°C:

⎛ 17.27(10) ⎞
hvapor = 0.0623exp ⎜ ⎟ = 0.125 m (0.411 ft)
⎝ 10 + 237.3 ⎠
 From Table 8.5, the suction pipe has an ID of 6.301 inches. Then, pipe area is:

π(6.301/12)2
Apipe = = 0.217 ft 2
4
The flow rate is: Q = 550/448.86 = 1.23 cfs. And, the velocity head is:

V2 (1.23)2
= = 0.500 ft
2g 2(32.2)(0.217)2

Friction loss in the suction pipe, using Hazen-Williams:

1.852
⎛ 550 ⎞
hf = 10.5 ( 8.0 ) ⎜ ( 6.301)
−4.87
⎟ = 0.12 ft
⎝ 150 ⎠

Minor losses on suction side of pump: From Table 11.2 for a flanged “long-radius”
45-degree elbow, 6” nominal size, Kr = 0.17. Also from Table 11.2, for a “basket
strainer,” Kr = 0.85. Then,

(hf )minor = ( 0.17 + 0.85 )( 0.500 ) = 0.51 ft

Notice that the minor losses are greater than the pipe friction losses (the pipe is only
8-ft long, and it is PVC, so it’s smooth) . Note that the average velocity in the
suction pipe is 5.67 fps, which is OK. Note also that the static lift will have a much
greater influence on NPSHa than friction losses, and that the velocity head is on the
same order of magnitude as the friction losses. Finally,

NPSHa = 32.4 − 0.12 − 0.51 − 8.0 − 0.50 = 23.3 ft

10. This pump installation will not be expected to cavitate because NPSHa >> NPSHr.
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #7 (200 pts)

Operating Point for Fixed Sprinkler System
Due: 29 Oct 04

Given:

• A fixed sprinkler system (all sprinklers operate simultaneously) in an orchard.

• Buried IPS-PVC 1-½” lateral pipes (see Table 8.3 of the textbook).
• Rainbird® under-tree M20VH impact sprinklers with 5/64-inch SBN-1 nozzle (see
www.rainbird.com for technical specifications).
• Lateral spacing: Sl = 40.00 ft.
• Sprinkler spacing: Se = 40.00 ft.
• The field is trapezoidal in shape, as shown below:

0.18% downhill
800 ft

0.10% uphill
1,100 ft
mainline

pump
560 ft
• In the above figure, only some of the laterals and sprinklers are shown, but remember
that the field area is covered with sprinklers in a fixed, permanent system.
• There are 27 laterals along the mainline.
1
 • Note the field slopes along mainline and lateral directions.
• The suction (upstream) side of the pump has a 4-ft static lift from a pond and 10 ft of 8”
PVC pipe (ID = 8.205”) with one 90-degree elbow and a strainer screen at the inlet.
• The mainline is 8” PVC pipe (ID = 8.205”), and is 1,100-ft long.
• Sprinkler riser height is: hr = 3.0 ft.
• A Berkeley model 4GQH pump curve as shown on the following page. You will use
the characteristic curve for 1600 RPM.
• Ignore minor losses along the mainline and laterals.

Required:

1. Determine the irrigated area (acres).

2. Develop an equation for sprinkler flow rate (q) as a function of pressure (P), whereby q
= KdPx (you determine Kd and x based on manufacturer’s data). Note that for a
straight-bore nozzle, you would expect x to be very close to 0.5, as in Eq. 5.1a.
3. Determine the number of sprinklers along each lateral, where lateral #1 is the closest
to the pump, and lateral #27 is the furthest from the pump.
4. Develop at least five points on the system curve and present those points numerically
in a table. You should write a macro (Excel) or computer program to do this (don’t
attempt to do it by hand with only a calculator).
5. Plot the system curve points on the attached pump curve graph, and draw a smooth
curve through the points.
6. Determine the operating point (Q in gpm & TDH in ft) for this system.
7. Determine the average application rate for the whole field area.

Do you work in an organized, neat way. Make comments about assumptions

and other technical issues as appropriate. Turn in all your work, including the
code for your macro or computer program.

2
Solution:

1. The irrigated area is approximately:

0.5(800 + 560)(1100)
A= = 17.2 acres
43,560

Due to the trapezoidal field shape, and the fact that there are 27 laterals at Sl = 40 ft,
the effective irrigated area is slightly less than 17.2 acres.

2. A linear regression is performed on logarithms of the manufacturer’s data for P and q

(Rainbird® M20VH with 5/64-inch SBN-1 nozzle):

Pressure Flow
(psi) (gpm) ln(P) ln(q)
25 0.88 3.2189 -0.1278
30 0.97 3.4012 -0.0305
35 1.05 3.5553 0.0488
40 1.12 3.6889 0.1133
45 1.19 3.8067 0.1740
50 1.25 3.9120 0.2231

given an R2 of 0.9996, and the following equation:

q = 0.173P0.506

for q in gpm; and P in psi. Note that the exponent is close to 0.500, which is expected
for a straight-bore nozzle. But notice also that allowing for the flexibility in the
exponent, x, gives a better mathematical fit to the manufacturer’s data.

3. Let the first lateral be located at ½Sl from the lower edge of the field (where the pump
is located). Calculate the number of sprinklers per lateral by rounding the potential
lateral length by Se. The potential lateral length is calculated by linear interpolation
along the left side of the field area (see the figure given above). The equation is given
below, and the calculation results are shown in the following table.

⎛ 800 − 560 ⎞
L = 800 − (1,100 − y) ⎜ ⎟
⎝ 1,100 ⎠

where y is the distance along the mainline (ft); and L is the potential lateral length (ft).

3
 Distance Length No. of
Lateral (ft) (ft) Sprinklers
1 20 564.4 14
2 60 573.1 14
3 100 581.8 15
4 140 590.5 15
5 180 599.3 15
6 220 608.0 15
7 260 616.7 15
8 300 625.5 16
9 340 634.2 16
10 380 642.9 16
11 420 651.6 16
12 460 660.4 17
13 500 669.1 17
14 540 677.8 17
15 580 686.5 17
16 620 695.3 17
17 660 704.0 18
18 700 712.7 18
19 740 721.5 18
20 780 730.2 18
21 820 738.9 18
22 860 747.6 19
23 900 756.4 19
24 940 765.1 19
25 980 773.8 19
26 1,020 782.5 20
27 1,060 791.3 20

4. Develop a computer program to start with a given pressure at the furthest downstream
sprinkler on lateral #27, calculate the flow rate at that sprinkler, calculate the pressure
at the next upstream sprinkler, then the flow rate at that sprinkler, and so on, until
reaching the mainline. Calculate the pressure in the mainline at the location of lateral
#26, then iterate along lateral #26 to get the same pressure in the mainline at that
location. Repeat for all other laterals, moving in the upstream direction, until a
pressure is obtained for the upstream end of the mainline. The system flow rate is
known from these calculations (sum of all individual sprinklers). Assume pipe leakage
is zero.

Knowing the system flow rate, determine the losses in the suction side of the pipe, and
determine TDH by adding the velocity head at the beginning of the mainline, the
pressure head at the beginning of the mainline, the static lift on the suction side, and
the hydraulic losses on the suction side of the pump.
4
Preliminary calculations and assumptions:

Assume a water temperature of 10°C, giving a kinematic viscosity of: 1.306(10)-6 m2/s,
or 1.406(10)-5 ft2/s.

Use the Swamee-Jain equation with ε = 1.5(10)-6 m, or 4.92(10)-6 ft for PVC to obtain
the Darcy-Weisbach friction factor, f.

From Table 8.3, the ID of the lateral pipe is 1.754 inches (0.1462 ft). The ID of the
mainline pipe is 8.205 inches (0.6838 ft).

Once the pressure at the upstream end of the mainline is calculated, the additional
TDH values include static lift, velocity head, and losses in the suction side of the pump,
plus the riser height. Assume the pump outlet is at the same elevation as the
upstream end of the mainline (you don’t know if the mainline or laterals are buried, nor
how deep they might be, but this information could be used to develop a somewhat
more specific design).

V2
TDH = 2.308Pmain + hlift + hr + (hf )suction +
2g

where TDH is in ft; Pmain is the pressure at the upstream end of the mainline (psi); hlift is
given as 4.0 ft; hr is given as 3.0 ft; (hf)suction are the hydraulic losses in the suction pipe
(ft); and the last term is the velocity head (ft).

The iterative part is to determine Pmain; the rest of the TDH terms are easy to calculate
directly.

From Table 11.2 (minor loss coefficients):

8-inch “basket strainer”: Kr = 0.75

8-inch “regular 90-deg elbow”: Kr = 0.26

Velocity head in suction pipe:

V2 8Q2
= 2 4 = 0.1151Q2
2g gπ D

Friction loss in suction pipe:

L V2
hf = f = 1.684 f Q2
D 2g

5
 Putting it all together:

TDH = 2.308Pmain + 4.0 + 3.0 + 1.684 f Q2 + 0.1151Q2 (1 + 0.75 + 0.26 )

or,
TDH = 2.308Pmain + 7.0 + Q2 [1.684 f + 0.2314]

The results are given in the table below:

Qs Pmain TDH
P (psi) (gpm) (psi) Re f (ft)
20 367.2 21.8 108,347 0.01761 57.50
25 411.2 27.2 121,318 0.01721 70.09
30 451.0 32.7 133,061 0.01689 82.66
35 487.6 38.1 143,853 0.01663 95.21
40 521.6 43.5 153,902 0.01641 107.74
45 553.6 48.9 163,344 0.01622 120.26
50 583.9 54.3 172,277 0.01605 132.77
55 612.7 59.7 180,777 0.01590 145.28
60 640.2 65.1 188,901 0.01577 157.77

where P is the pressure at the furthest downstream sprinkler on lateral #27; Qs is the
total system flow rate; Pmain is the pressure at the upstream end of the mainline (just
downstream of the pump); Re is the Reynold’s number in the suction pipe; f is the value
from Swamee-Jain; and TDH is the total dynamic head.

5. The system curve (Qs versus TDH) is superimposed upon the pump manufacturer’s
curves, as shown in the figure below.

6. The operating point (intersection of the system curve and the 1600 RPM pump curve)
is seen to be approximately:

Qs = 568 gpm
TDH = 126 ft

7. The average application rate at this operating point is approximately:

(568)(12)(3600)
AR avg = = 0.075 inch/hr
(458)(40)(40)(448.86)

or 1.9 mm/hr.

6
 300
275
250
225
200
TDH (ft)

175
150
125
100
75
50
25
0 200 400 600 800 1000 1200 1400 1600
Flow rate (gpm)

The MS Excel VBA listing follows:

Const Pi = 3.141593
Const Se = 40# 'ft
Const Sl = 40# 'ft
Const LatCount = 27 'number of laterals
Const Dlat = 0.1462 'ft
Const Dmain = 0.6838 'ft
Const SoLat = -0.0018 'ft/ft
Const SoMain = 0.001 'ft/ft

Function hf(ByVal Q As Double, ByVal D As Double, ByVal L As Double) As Double

'------------------------------------------------------------------------------
' Returns friction loss in feet of water head (Darcy-Weisbach).
' If turbulent, uses the Swamee-Jain equation for the friction factor, f.
'------------------------------------------------------------------------------

Dim Re As Double, RelRough As Double, f As Double

viscosity = 0.00001406 'ft2/s

RelRough = 0.00000492 / D
Re = 4 * Q / (viscosity * Pi * D)

If Re > 4000 Then

f = RelRough / 3.75 + 5.74 / Re ^ 0.9 'Turbulent
f = WorksheetFunction.Log10(f)
f = 0.25 / f ^ 2
Else
f = 64 / Re 'Laminar
End If

7
 hf = 8# * f * L * (Q / Pi) ^ 2 / (32.2 * D ^ 5)

End Function

Function Lateral(ByVal P As Double, n As Integer, Qlat As Double) As Double

'------------------------------------------------------------------------------
' Calculates lateral inlet pressure head for "n" sprinklers.
' Pressure, P, is in psi.
'------------------------------------------------------------------------------

Dim Qs As Double, h As Double

Qlat = 0# 'cfs

For i = 1 To n
Qs = 0.173 * P ^ 0.506 'gpm
Qlat = Qlat + Qs / 448.86 'cfs
h = P * 2.308 + hf(Qlat, Dlat, Se) + SoLat * Se 'ft
P = h / 2.308 'psi
Next

Lateral = h

End Function

Function ParabolaFit(x, y, target As Double) As Double

'------------------------------------------------------------------------------
' Determines constants a, b, and c for a parabola through three points.
' Equation is: y = ax^2 + bx + c. If parabola impossible, uses bisection.
' Returns the x-value (P) which matches the specified target y-value (hmain).
' P is the pressure at the furthest DS sprinkler in the lateral.
'------------------------------------------------------------------------------

Dim Bisection As Boolean

Dim foo As Double, boo As Double
Dim a As Double, b As Double, c As Double

ParabolaFit = 0
foo = x(2) - x(1)
Bisection = Abs(foo) < 0.0000000001

If Not Bisection Then

C0 = x(1) ^ 2
c1 = (x(3) - x(1)) / foo
c2 = x(2) ^ 2 - C0

boo = x(3) ^ 2 - C0 - c1 * c2
Bisection = Abs(boo) < 0.0000000001

If Not Bisection Then

'------------------------
' Parabolic interpolation
'------------------------

a = ((y(1) - y(2)) * c1 - y(1) + y(3)) / boo

b = (y(2) - y(1) - a * c2) / foo
c = y(2) - x(2) * (a * x(2) + b)

c = c - target
ParabolaFit = (-b + Sqr(b * b - 4 * a * c)) / (2 * a)
Exit Function
End If
End If

8
 If Bisection Then

'--------------
' Use bisection
'--------------

If y(2) < target Then

ParabolaFit = (y(2) + y(3)) / 2
Else
ParabolaFit = (y(2) + y(1)) / 2
End If
End If

End Function

Function QsPmain(ByVal P As Double, Flow As Boolean) As Double

'------------------------------------------------------------------------------
' Iterates to determine the system flow rate for a given starting pressure.
' Returns either flow rate (Qs) or pressure (Pmain) at US end of the mainline.
'------------------------------------------------------------------------------

Dim n As Integer
Dim lat As Integer
Dim LatHead(1 To 3) As Double
Dim Pressure(1 To 3) As Double
Dim Qlat As Double, x As Double
Dim Qmain As Double, L As Double, NewP As Double, hmain As Double

hmain = 0
Qmain = 0

For lat = LatCount To 1 Step -1

'--------------------------------------------
' Determine number of sprinklers this lateral
'--------------------------------------------

x = 20 + (lat - 1) * Sl
L = 800# - (1100# - x) * 0.2181818
n = Round(L / Se, 0)

'--------------------------------------
' Calculate lateral inlet pressure head
'--------------------------------------

If lat = LatCount Then

'------------------------------------
' No need to iterate for last lateral
'------------------------------------

LatHead(2) = Lateral(P, n, Qlat)

Else

'------------------------------------------------
' Iterate to match lateral inlet & mainline heads
'------------------------------------------------

Pressure(1) = P / 4
Pressure(3) = 3 * P
Pressure(2) = (Pressure(1) + Pressure(3)) / 2

LatHead(1) = Lateral(Pressure(1), n, Qlat)

LatHead(3) = Lateral(Pressure(3), n, Qlat)
LatHead(2) = Lateral(Pressure(2), n, Qlat)

9
 If (LatHead(1) > hmain) Or (LatHead(3) < hmain) Then

'-------------------------------
' Failed to bracket the solution
'-------------------------------

QsPmain = -100
Exit Function
End If

For i = 1 To 50

'----------------------------------
' Search by parabolic interpolation
'----------------------------------

NewP = ParabolaFit(Pressure, LatHead, hmain)

If NewP < Pressure(2) Then

Pressure(3) = Pressure(2)
LatHead(3) = LatHead(2)
Else
Pressure(1) = Pressure(2)
LatHead(1) = LatHead(2)
End If

Pressure(2) = NewP
LatHead(2) = Lateral(NewP, n, Qlat)

If Abs(LatHead(2) - hmain) < 0.001 Then

'-------------------
' Solution converged
'-------------------

Exit For
End If
Next

End If

'-----------------------------------------
' Move upstream one hydrant along mainline
'-----------------------------------------

Qmain = Qmain + Qlat

hmain = LatHead(2) + hf(Qmain, Dmain, Sl) + SoMain * Sl
Next

'----------------------------
' Return the system flow rate
' or the mainline pressure
'----------------------------

If Flow Then
QsPmain = Qmain * 448.86 'gpm
Else
QsPmain = hmain / 2.308 'psi
End If

End Function

10
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #8 (100 pts)

Trickle System Design Calculations
Due: 01 Dec 04

Do you work in an organized, neat way. Write down any assumptions you make.

Given:

• A mature walnut orchard will be drip-irrigated.

• Use a single lateral per row of trees.
• Irrigated area is 44 ha.
• Tree spacing is 6.2 x 6.2 m.
• Peak daily ET is Ud = 4.9 mm/day.
• Seasonal water requirement: U = 541 mm.
• Effective rain, peak-use period: assume zero.
• Residual soil water in the spring: assume zero.
• Use an MAD of 25%.
• Soil water holding capacity is 178 mm/m.
• Water source: deep well with maximum discharge of 125 lps.
• Irrigation water quality: ECw = 0.61 dS/m.
• Root zone depth is 2.0 m.
• Shaded area is 75%.
• Emitter equation:

q = 0.32P0.53

for q in lph; and P in kPa.

• Nominal emitter flow rate: qa = 4 lph.

• Manufacturer coefficient of variation: 0.062.
• Average wetted width at 4 lph: w = 2.33 m.
• Outlets per emitter: one.

1
Required:

1. Use metric units in your calculations.

2. Select an appropriate emitter spacing, Se (m).
3. Determine the number of emitters per tree, Np.
4. Calculate percent wetted area, Pw. Use the equation from Lecture 18 (includes Pd in
the denominator). Make sure Pw is between 33% and 67%; if not, increase Np as
necessary.
5. Calculate maximum net depth to apply per irrigation, dx (mm).
6. Calculate the average peak daily “transpiration” rate, Td (mm/day).
7. Calculate the maximum irrigation interval, fx. If fx ≥ 1 day, then use f’ = 1day.
8. Calculate the net depth per irrigation, dn (mm).
9. Select a reasonable target EU value (Table 20.3).
10. Determine (ECe)max (Table 19.2).
11. Determine the transmission ratio, Tr (Table 19.3).
12. Calculate the leaching requirement, LRt.
13. Calculate the gross depth to apply per irrigation, d (mm).
14. Calculate the gross volume of water per tree per day, G (liter/tree/day).
15. Calculate ha, corresponding to qa = 4 lph, in m of water head (not kPa).
16. Calculate the water application time, Ta (hrs).
17. If Ta > 21.6 hrs/day, recalculate qa such that Ta = 21.6 hrs/day, then calculate ha
corresponding to the new qa value.
18. Determine the number of stations, Ns.
19. Determine the minimum number of emitters per tree, Np’.
20. Calculate the system coefficient of variation, νs.
21. Calculate the minimum allowable emitter flow rate, qn (lph).
22. Calculate the allowable subunit pressure head variation, ∆Hs (m).
23. Calculate the system capacity, Qs (lps). Is this less than or equal to the well capacity
of 125 lps?
24. Calculate the total gross seasonal depth to apply, Dg (mm).
25. Calculate the gross seasonal volume of irrigation water, Vs (m3).
26. Calculate the required number of operating hours per season, Ot (hrs/season). Make
sure it is not more than 8,760 hrs!

2
Required:

I. Emitter spacing

Use the “optimal” spacing: Se = 0.8w = 0.8(2.33) = 1.86 m

II. Emitters per tree

Sp 6.2
Np = = = 3.33
Se 1.86

III. Percent wetted area

⎛N S w⎞ ⎛ (3.33)(1.86)(2.33) ⎞
Pw = 100 ⎜ p e ⎟ = 100 ⎜ ⎟ = 50.1%
⎜S SP ⎟ ⎝ (6.2)(6.2)(0.75) ⎠
⎝ p r d⎠

IV. Maximum net application depth

MAD Pw
dx = Wa Z = (0.25)(0.501)(178)(2.0) = 44.6 mm
100 100
V. Average peak daily transpiration rate

Td = 0.1Ud Pd = 0.1(4.9) 75 = 4.24 mm/day

VI. Maximum irrigation interval

dx 44.6
fx = = = 10.5 days
Td 4.24

Then, use f’ = 1 day (for design purposes).

VII. Net depth per irrigation

dn = Td f ' = (4.24)(1) = 4.24 mm/day

VIII. Target EU

Table 20.3: “point-source” water applicators with Np > 3 gives recommended EU range of
90 to 95%. In this design iteration, choose EU = 92%.

3
IX. Maximum ECe

From Table 19.2, for a walnut crop, (ECe)max = 8 dS/m.

X. Transmission ratio

From Table 19.3, for a “deep-rooted” (Z > 1.5 m) crop and a “medium-textured” (see Wa
above) soil: Tr = 1.00.

XI. Leaching requirement

EC w 0.61
LR t = = = 0.038
2 (ECe )max 2(8)

XII. Gross application depth

For LRt < 0.1, the following equation is applied:

⎛d T ⎞ ⎛ (4.24)(1.00) ⎞
d = 100 ⎜ n r ⎟ = 100 ⎜ ⎟ = 4.61 mm/day
⎝ EU ⎠ ⎝ 92% ⎠

XIII. Gross volume of water per tree

d 4.61
G= SpSr = (6.2)(6.2) = 177 liter/day/tree
f' 1
XIV. Nominal emitter pressure head

Apply the given emitter equation, and use 9.81 kPa/m:

1/ 0.53
⎛ 1 ⎞⎛ 4 ⎞
ha = ⎜ ⎟⎜ ⎟ = 12.0 m
⎝ 9.81 ⎠ ⎝ 0.32 ⎠

XV. Water application time per irrigation

G 177
Ta = = = 13.3 hrs/day
Np qa (3.33)(4)

XVI. Number of stations

4
 Two stations would require 2(13.3) = 26.6 hrs/day. Thus, there can be only one station
(Ns = 1) in this design.

XVII. Minimum number of emitters per tree

See the figure below, showing a tree spacing of 6.2 m, and emitter spacing of 1.86 m, and
a wetted width of 2.33 m:

tree tree tree

It is seen that, on average, four emitters contribute some irrigation water to each tree.
Alternatively,

⎛ 5 m/tree ⎞
N'p = trunc ⎜ + 2⎟ = 4
⎝ 1.86 m/emitter ⎠

where “trunc” means to truncate (round down) to the nearest whole number.

XVIII. System coefficient of variation

ν 0.062
νs = = = 0.031
N'p 4

XIX. Minimum allowable emitter flow

qaEU (4)(92)
qn = = = 3.83 lph
100 (1 − 1.27νs ) 100 (1 − 1.27(0.031) )

which corresponds to a head of:

1/ 0.53
⎛ 1 ⎞ ⎛ 3.83 ⎞
hn = ⎜ ⎟⎜ ⎟ = 11.0 m
⎝ 9.81 ⎠ ⎝ 0.32 ⎠

5
XX. Allowable subunit head variation

∆Hs = 2.5 (ha − hn ) = 2.5 (12.0 − 11.0 ) = 2.5 m

XXI. System capacity

ANpqa (44)(3.33)(4)
Qs = 2.78 = 2.78 = 42.4 lps
NsSpSr (1)(6.2)(6.2)

which is less than the well capacity of 125 lps. Thus, the well has sufficient flow rate to
accommodate this design.

XXII. Gross season water application depth

Assume a TR value of 1.00 (Table 19.4). Then, Es = EU = 92%. Effective rain and
residual soil moisture are given to be zero. Thus,

( ) (
Dn = U 0.1 Pd = 541 0.1 75 = 469 mm )
Then, gross seasonal depth is:

100Dn 100(469)
Dg = = = 530 mm
Es (1 − LR t ) 92(1 − 0.038)

XXIII. Gross season application volume

Dg A (530)(44)
Vs = = = 23.3 ha-m
1000 1000

XXIV. Operating hours per season

Vs ⎛ 23.3 ⎞
Ot = 2778 = 2778 ⎜ ⎟ ≈ 1,530 hrs/season
Qs ⎝ 42.4 ⎠

6
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #9 (100 pts)

Trickle Manifold Location
Due: 3 Dec 04

Given:
A rectangular field of orchard trees, 550-m long in the direction of the PE laterals. The
preliminary design data are as follows:

lateral ID = 17.8 mm qa = 3.95 lph

Se = 2.5 m fe = 0.10 m
Sp = 5.0 m Ha = 13.0 m
x = 0.55 Qs =9.77 lps

Required:

Determine the following using either: (1) the semi-graphical, non-dimensional, design
procedure, or (2) a completely numerical design procedure:

1. Optimal manifold location

2. Required lateral inlet pressure head, Hl
3. Minimum lateral pressure head, Hn’

Show all of your work neatly, step by step. Adjust the manifold location uphill by as much as
0.75(Sp), or downhill by as much as 0.25(Sp) so that is is positioned midway between two plant
rows.
1. How many trees on the uphill side?
2. How many trees on the downhill side?
3. Is Hn’ the same on the uphill & downhill sides?
Solution:

• The solution can be obtained by different methods, as explained in class.

• The quickest and easiest solution is to use the New and Improved “OptManifold.exe”
computer program, as shown below:

• The New and Improved version of the program uses Hazen-Williams.

• Due to approximations and simplifying assumptions in the equations, the calculated value of
Hn’ is not exactly equal in the uphill and downhill parts of the lateral.
• The tree spacing is given as Sp = 5.0 m. Make the downhill lateral 375-m long, and the
uphill lateral will be 175 m in length (a slight adjustment on the calculation results).
• This gives 175/5 = 35 trees on the uphill side of the manifold, and 375/5 = 75 trees on the
downhill side.
 BIE 5110/6110
Sprinkle & Trickle Irrigation
Fall Semester, 2004

Assignment #10 (100 pts)

Trickle Manifold Pipe Sizing
Due: 13 Dec 04

Given:

• A trickle irrigation system with a manifold inflow rate of 5.00 lps.

• An allowable subunit pressure head variation of ∆Hs = 4.2 m.
• A lateral pressure variation of ∆Hl = 2.3 m.
• A uniform ground slope of 1.04% in the manifold direction.
• A total manifold length of 200 m.
• A lateral spacing of 2.0 m.
• The following PVC pipe sizes are available:

Size I.D.
(inches) (inches) (mm)
0.5 0.622 15.8
0.75 0.824 20.9
1 1.049 26.6
1.25 1.380 35.1
1.5 1.610 40.9
2 2.067 52.5
2.5 2.469 62.7
3 3.068 77.9
4 4.000 101.6
6 6.000 152.4
8 8.000 203.2
10 10.000 254.0

Required:

1. Design the manifold, using up to four different pipe diameters.

2. Use the Darcy-Weisbach and Blasius equations for friction loss.
3. Do not use any pipe diameter which is less than ½ the largest diameter.
4. Determine appropriate manifold pipe sizes and lengths.
5. Use either the semi-graphical method, or develop a computer program.
Solution:

• Assume a water temperature of 10°C, giving a kinematic viscosity of:

ν = 1.306(10)-6 m2/s.

• Then, the Reynolds number is:

4Q ⎛Q⎞
NR = = 9.75(10)5 ⎜ ⎟
νπD ⎝D⎠

• Plugging the above into the Blasius equation:

−0.25
⎛Q⎞
f = 0.0102 ⎜ ⎟
⎝D⎠

• Darcy-Weisbach:

⎛ Q1.75 ⎞
hf = 0.169 ⎜ 4.75 ⎟
⎝D ⎠

for L= 200 m; hf in m; Q in m3/s; and D in m.

• Elevation change over 200-m manifold length:

∆he = (200)(0.0104) = 2.08 m

• Allowable manifold pressure head variation:

( ∆hm )a = ∆Hs − ∆hl = 4.2 − 2.3 = 1.9 m

• In a spreadsheet, make a graph of hf versus Q (0 to 5 lps) with a separate curve for
each of the available pipe sizes (diameters).
• It is seen that the smallest five pipe sizes (1.5 inches and lower) have curves which
are obviously too steep for the range 0 < hf < (2.08 + 1.9), so these are omitted from
the graph.
• Plot straight lines which define the limits of the allowable manifold pressure head
variation. The lower line begins at the origin (0,0) and goes to (5,2.08). The upper
line goes from (0,1.9) to (5,2.08+1.9) = (5,3.98).
• Now it is seen that the largest four pipe sizes have head loss curves which are too
flat. These are eliminated from the graph. The graph has three pipe diameters:
 4.0

2 inch (52.5 mm)

3.5 2.5 inch (62.7 mm)
3 inch (77.9 mm)
Lower limit
3.0 Upper limit

2.5
Friction loss (m)

2.0

1.5

1.0

0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Manifold flow rate (lps)

• Make the head loss curve for the largest of the three diameters (3 inch) pass
through the point (5,3.98). Do this by observing that at 5 lps, the head loss from the
3-inch curve is 2.907 m. Add 3.98 – 2.907 = 1.073 m to all points on the curve for
the 3-inch pipe size, thereby shifting it vertically as needed, then draw a straight line
through the origin, tangent to this shifted curve (see the figure below).
• Next, vertically shift the curve for the 2.5-inch pipe so that it is tangent to the tangent
line, as shown in the second figure on the next page.
• Draw a vertical line at the intersection between these two curves, defining the length
of the 3-inch pipe size. The break point is at 1.71 lps. This gives the following
length for the 3-inch pipe size (assuming a linear change in flow rate along the
manifold):

⎛ 1.71 − 5.00 ⎞
L3 " = 200 ⎜ ⎟ ≈ 132 m
⎝ −5.00 ⎠

• Finally, make the 2-inch curve tangent to the tangent line (by vertical shifting), then
draw a vertical line at the intersection of this curve with the 2.5-inch curve. See the
final graph below.
 4.0
2 inch (52.5 mm)
2.5 inch (62.7 mm)
3.5
3 inch (77.9 mm)
Lower limit
3.0 Upper limit
Tangent line

2.5
Friction loss (m)

2.0

1.5

1.0

0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Manifold flow rate (lps)

4.0
2 inch (52.5 mm)
2.5 inch (62.7 mm)
3.5
3 inch (77.9 mm)
Lower limit
3.0 Upper limit
Tangent line

2.5
Friction loss (m)

2.0

1.5

1.0

0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Manifold flow rate (lps)
 4.0
2 inch (52.5 mm)
2.5 inch (62.7 mm)
3.5
3 inch (77.9 mm)
Lower limit
Upper limit
3.0
Tangent line
Break 1
2.5
Friction loss (m)

2.0

1.5

1.0

0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Manifold flow rate (lps)

4.0
2 inch (52.5 mm)
2.5 inch (62.7 mm)
3.5
3 inch (77.9 mm)
Lower limit
Upper limit
3.0
Tangent line
Break 1
2.5 Break 2
Friction loss (m)

2.0

1.5

1.0

0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Manifold flow rate (lps)
• The break point between the 2.5 and 2.0 curves is at 0.48 lps. This gives the
following length for the 2.5-inch pipe size:

⎛ 0.48 ⎞
L 2" = 200 ⎜ ⎟ ≈ 19 m
⎝ 5.00 ⎠

• Then, the 2.5-inch pipe size would have a length of 200 – 19 – 132 = 49 m.
• In summary:

Size Length Vmax

(inches) (m) (m/s)
3 132 1.05
2.5 49 0.55
2 19 0.22
Total: 200

• Note that the design would also be acceptable (within allowable limits) if only two
sizes were used: 3-inch and 2.5-inch. This would also simplify installation and
would not increase the cost significantly.
• Note also that the maximum average velocity in each of the pipe sizes is well within
acceptable limits.
• Finally, observe that the smallest pipe diameter (2 inches) is greater than ½ the
largest pipe diameter (3 inches), as desired for the manifold design.
• Note that if a different kinematic viscosity is applied, a different pipe sizing solution
might be obtained.
 BIE 5300/6300 Assignment #4
Broad-Crested Weir Calibrations
28 Sep 04 (due 05 Oct 04)

Show your calculations in an organized and neat format. Indicate any

assumptions or relevant comments. You can use ACA or WinFlume if you
like, or you can do calculations in a spreadsheet or other program.

I. You have to design a BCW for a concrete-lined trapezoidal canal with a bottom
width of 2.0 m, inverse sides slopes of 1.25:1 (H:V), lining depth of 2.7 m, and
maximum discharge of 12 m3/s. The Manning roughness is estimated to be
0.013, and the longitudinal bed slope is 0.000123 m/m. Make sure the BCW will
operate under free-flow conditions up to Qmax = 12 m3/s.

(a) Give the design dimensions for the BCW, and provide any relevant
comments about the design and your assumptions.
(b) What is the minimum flow rate which can be accurately measured with
your BCW design?
(c) Will the upstream canal walls need to be raised after installing the BCW
when operating at Qmax = 12 m3/s?
(d) Would you recommend including a DS ramp on the BCW? Why or why
not?
(e) Suppose you include two 1-inch diameter PVC drainage pipes at the
base of the BCW. What is the estimated discharge through the pipes at
the 12 m3/s capacity of the BCW? Is this a significant fraction of Qmax?

II. Use a spreadsheet or your own custom computer program to check the
calibration of your BCW design from the problem above, but based only on
energy-balance from upstream to the location of critical flow on the sill. In this
case, you will assume free-flow conditions at the BCW.

(a) Do the comparison for the full flow range of the BCW.
(b) Make a graph (two curves) of hu versus Q for the full calibration (from
the problem above) and for the simpler energy-balance calibration.
(c) Are the two calibrations significantly different in this case?

III. Suppose the Parshall flume at the location (Logan Canyon) of our field exercises
is getting badly deteriorated and needs to be replaced. A decision is made to
install a BCW instead of the Parshall flume, at approximately the same location.
Based on your lab data, and a Qmax of 40 cfs, what BCW design dimensions and
features would you propose?