Page 1
FLOW CALIBRATION OF THE BRYAN
CANAL RADIAL GATE AT THE UNITED
IRRIGATION DISTRICT
USCID Water Management Conference
Meeting Irrigation Demands in a Water-Challenged Environment
SCADA and Technology: Tools to Improve Production
Fort Collins, Colorado September 28 - October 1, 2010
Page 2
Gabriele Bonaiti
Eric Leigh
Askar Karimov Extension Associates
Guy Fipps
Professor and Director
Irrigation Technology Center
Texas AgriLife Extension Service
Department of Biological and Agricultural Engineering
Texas A&M University, College Station
Project based upon work supported by the Cooperative State Research,
Education, and Extension Service, U.S. Department of Agriculture, under
Agreement No. 2005‐45048‐03208
For program information, see http://riogrande.tamu.edu
Page 3
LOWER RIO GRANDE VALLEY, TEXAS
Page 4
UNITED I.D.
37,800 acres
47% urban expansion
57,000 acre-feet Class A water rights
50 miles open canals, 115 miles of gravity fed pipeline
Bryan canal
carries 30% of total water distributed: › Irrigation: 10,000 acre-
feet/yr
› Municipal: 23,000 acre-feet/yr
Mission
Mc Allen
Page 5
DEMONSTRATION PROJECT
Purpose
Improve the efficiency of Bryan Canal management by:
› Replacing the old, dilapidated sliding gate
› Establishing telemetry and remote control capabilities
› Giving the District the ability to control the canal
system based on flow
Page 6
DEMONSTRATION PROJECT
Objectives and issues related to gate
calibration for flow
Calibrate a locally manufactured radial gate for flow
Develop an easy to use set of procedures so the district
can calibrate their own gates
Determine best equation for local flow conditions
Determine how are choice of actuators affects the
process
Demonstrate the use of inexpensive data loggers (“non
SCADAPACK”) and other components
Page 7
Replacing head vertical slide gate in poor condition…
Trapezoidal concrete-
lined canal
225 cfs
capacity
7 ft deep,
18 ft top
width, 4 ft
bottom width, side
slopes of
1:1
Page 8
…with new radial gate
Advantages:
Small force for lift and operation,
Good hydraulic discharge for
calibrating a gate to serve as a flow measurement device
Rectangular structure
24 ft long and 10 ft
wide
Radius of 7 ft, vertical
height of 7 ft, pinion height of 3.5 ft.
Opening 0 - 5 ft
Gate leaf edges have
hard bar-shaped rubber seals
Page 9
Equipment and Instrumentation
Pressure transducer sensors to measure water level
Actuator to control manually and remotely the gate opening
Doppler flow meter to measure flow
Page 10
Calibration of a Doppler Flow Meter
Methods:
› Price Type AA as current meter
› Tests were carried out at high and low flows
› USGS recommended procedures using the two-point
method
Results:
› Doppler max. flow = 116 cfs (1.7 ft/s)
› Doppler average flow < current meter flow by 3%
Doppler data were corrected accordingly
Page 11
Calibration of Vertical Gate Opening vs Actuator
Position and Input Voltage
Methods
› 4-20 mA input to the actuator was set to a 0-100%
scale, which corresponds to 0 to 5’ gate opening
› Measurements were taken every 0.5’during:
opening of gate
closing of gate
› Measurements:
vertical opening of the gate
position of actuator
Hysteresis: “from the Greek for
shortcoming, to be late, to fall short…”
Page 12
Results: Hysteresis between Actuator position and Gate opening
y = 0.0176x2 + 0.9057x - 0.029
R² = 0.9996
y = 0.0328x2 + 0.8389x - 0.0352
R² = 0.9999
0
1
2
3
4
5
0 1 2 3 4 5
Ga
te o
pen
ing
(ft
)
Actuator position (ft)
Opening motion
Closing motion
Page 13
Two different equations were used to convert actuator signal into gate
opening
R² = 0.9996
R² = 0.9999
0
1
0 1
Ga
te o
pen
ing
(ft)
Actuator position (ft)
Opening motion
Closing motion
0.1 ft
difference =
4.4 cfs in
average
Page 14
Results: Hysteresis between Data logger input and Actuator position
Disregarded data logger input data for flow calibration purposes
y = 0.0128x2 + 0.9535x - 0.0133
R² = 0.9998
y = -0.0146x2 + 1.0888x + 0.056
R² = 0.9998
0
1
2
3
4
5
0 1 2 3 4 5
Actu
ato
r p
osi
tion
(ft
)
Data logger input (ft)
Opening motion
Closing motion
0.16 ft
difference =
7.1 cfs in
average
Page 15
Calculated vs Measured Flow Rate
Measured data
0
10
20
30
40
50
60
70
0
1
2
3
4
5
6
7
01/12/09 02/09/09 03/09/09
Flo
w (
cfs)
Fee
t (f
t)
Upstream level Downstream level
Gate opening Measured flow
Page 16
Range in conditions that were observed
› Flow conditions were always submerged
Parameter Max Average Min
Opening (ft) a 1.5 0.5 0.0
Upstream level (ft) h1 7.0 5.4 3.9
Downstream level (ft) h2 5.4 4.2 2.2
Head differential(ft) h1-h2 2.8 1.2 0.0
Page 17
Calculating Flow › submerged flow orifice equation
Q = discharge Cq = discharge coefficient a = gate opening L = gate width g = gravitational constant h1 = upstream level h2 = downstream level h1 - h2 = head differential
› This equation is commonly used to calculate flow through an underflow gate for slow and very submerged flow
Values were selected for Cq
until the calculated cumulative flow converged with measured cumulative flow
Page 18
Results: A single discharge coefficient Cq = 0.70
matched measured flow with an average error of
2.9 % (relative standard deviation calculated when flow > 2 cfs)
0
600
1,200
1,800
2,400
3,000
3,600
4,200
0
10
20
30
40
50
60
70
01/12/09 02/09/09 03/09/09
Cu
mu
late
d flo
w (ac
ft)
Flo
w (cf
s), R
el. s
t. d
ev. (
%)
Calculated flow Measured flow Relative standard deviation
Page 19
Results: Linear regression interpolating all events
y = 1.0197x - 0.513
R² = 0.9943
-10
10
30
50
70
-10 10 30 50 70
Cal
cula
ted flo
w (cf
s)
Measured flow (cfs)
Page 20
Results: Effects of correction with hysteresis
Page 21
Results: effects of correction with hysteresis
The linear regressions
fitting the two groups of
data were tested for
parallelism, and the
slopes resulted
statistically different
from each other
Page 22
CONCLUSIONS
It was possible to calibrate a locally manufactured
gate for flow
The doppler flow meter was useful as the reference
flow because of high accuracy and large amount
of data that was processed
Further work:
› Test the submerged orifice equation in different conditions
of flow
› Test other equations
› Apply the same procedure to other gates
› Test different actuator types and data logger set up to
limit hysteresis issues