25-05-2018

1

Training for Asian Young Professionals on

Performance Assessment of Irrigation Systems

( 9th to 13th April 2018 in Beijing, China)

Structural interventions in modernization of irrigation systems-

Canal lining, Long crested weir, Baffle distributor, Measuring devices etc.

By

M G Shivakumar Chief Engineer

VJNL, UBP Zone, ChitradurgaWater Resources Department , Karnataka, India

Contents

• What are structural interventions?

• Canal lining

• Long Crested Weir

• Baffle distributor

• Measuring device

• Other structures (offtakes, cross regulators,escapes etc)

LINING IRRIGATION

CANALS

CANAL LINING - TO LINE or NOT TO LINE

• Canal lining is a very expensive element in canal construction - Cost of lining typically represents about 40 percent of the total cost.

• Before making such a large investment, there must be a clear idea of the benefits to be obtained

• The reduction of seepage losses is often assumed to be constant for the expected life of the lining to have a chance of achieving a favorable economic return

Issues of Canal Lining : Cost/Benefit

25-05-2018

2

• Seepage losses typically represent 10 to 40 percent of diverted water

• Sedimentation reduces seepage

• Within 6 years of canal lining, losses return to around 60 % of the initial seepage (WB)

Issues of Canal Lining: Seepage losses

Benefits of Canal Lining

• reduces seepage losses

• improves canal hydraulics

• reduces water-logging and salinization

• reduces weed growth

• improves equity and reliability of water

distribution?

• reduces maintenance (?)

Benefits of Canal Lining: Reduction of

weed growth

this benefit is questionable

in some projects with old

lined canals and poor

construction joints

Quality of construction and concrete

Saturation of the soils after

commissioning of the canal

may result:

– cracks in the

slabs

– opportunity for

seepage losses

under most of the

slab area

25-05-2018

3

Estimates of Seepage Losses:

a source of doubt and concern

• Difficult to compare the results of tests and

estimates because of number of variables

• Checking seepage losses independently of

operation records are expensive

• Published records often fail to record key

variables

Seepage Variables

• characteristics of the soils:

– porosity, permeability,

– chemistry, granulometry,

– stratigraphy

• geometry of canals

• position of water table

• chemistry of water and soils

• sedimentation: self sealing effect

Seepage Variables (cont.)

• quality of construction

• age of the canals

• cycles of filling/draining

• maintenance

Seepage Values: Pakistan

• Normally assumed 8 cfs/million ft2 (0.23m/day)

• Estimates of average losses up to outlets: 20 to

25 % of inflows at head of main canals

• range of measurement values: 0.15 to 0.30 m/day

(0.06 to 0.57m/day)

• ISRIP and IWMI measured 15 to 25 % average

losses from the main and secondary system –

CRBC, Fordwah, Eastern Sadiqia (1990-1995)

25-05-2018

4

Measurements using ponding method

(Singh, 1987)

•unlined canal 0.3 m/day

•lined canal 0.03 m/day

These data come from controlled tests in

laboratories and experimental sections with

excellent maintenance

INDIA: PUNJAB

brick lined canals

•Sunam branch ( four-month old)

0.06 m/day

•Mudki distributary (15 km, 5-year old)

0.29 m/day

•Mukstar distributary (6-year old)

0.49 m/day

Watercourses (24): seepage losses from channel

older than 4 years are comparable to losses from

unlined channels

Canal Lining: A way to modernize the

irrigation system?

• it is generally advocated by Irrigation Agencies to

reduce seepage

• there are obstacles to canal lining : e.g it

interrupts irrigation services

Line or not to Line: The Debate

• Since seepage recharges the

groundwater, overall water

saving could be marginal

• Cost of pumping

• Deterioration of water

quality, Waterlogging and

salinization of adjacent

lands

• Farmers use groundwater

more efficiently

25-05-2018

5

THE TECHNICAL CASE FOR CANAL

LINING IS NOT STRAIGHTFORWARD

According to the WB-Experts

•Canal lining with 1% crack area has a seepage rate

of 70% of that for unlined conditions (depth to

water table: 8m)

•There is now strong evidence that hard surface

linings deteriorate within a few years until seepage

losses return to that for an unlined canal

THE TECHNICAL CASE FOR CANAL

LINING IS NOT STRAIGHTFORWARD

1% crack area = 70% seepage of unlined conditions (depth

to water table: 8m) HOW POSSIBLE?

Water tableWater table

Unlined Cracked Lining

Answer lies in thewetted perimeter

Head makes thin concrete cracks watering a much larger strip behind the concrete layer.

Canal Lining: Questions ?

• Do we really need lined canals to improve water delivery service to water users?

• Do we want to carry out this very expensive option and risk losing the valuable money that could be spent more effectively on other interventions?

Canal Lining: Options

• Avoid systematic canal lining as a norm !

• Serious documentation for lining to be done !

• Wherever canal lining is justified/needed use only

high standards techniques [geo-membrane,

reinforced concrete]

• Adjust operation and maintenance to the needs of

lined canals.

25-05-2018

6

Long Crested Weir - Design

25-05-2018

7

Variables? Length - Height

Provision for flushing !

h

Q= c. Lc. [h]3/2

(c = 1 to 1.9)

LCW= Overshot structure

25-05-2018

8

Exercise: Crest length of a Long crested weir ?

Q= c. Lc. [h]3/2

Objective of control

Q1 ➔ H1

Q2 ➔ H2

Variation H= H2-H1

2/3

21

3/22

3/21

Chh

cQ

cQ

L

−

−

=

Crest lenth variation for Q=2 to 5 m3/s

0

10

20

30

40

50

60

70

80

90

100

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

Water level variation

Cre

st

len

gth

Crest length of a Long Crested weir ?

H1-H2= 0.15 m

Then Length crest = 17 m

Objective: Ensure that flow is not submerged for Q max!

Height of a Long Crested weir

Partially submerged Drop is insufficient

Crest should be slightly higher than normal flow to allow a non submerged flow on the crest

Height of a Long Crested weir

25-05-2018

9

H1

H crest = H normal + ∆hcrestH flow = H crest+ H1Head losses = H1+ ∆hcrest

Head losses

H normal

Head losses

∆hcrest

∆hcrest ?0.10 0.15 m

Backwater effect = Head losses/slope

Good Design for flushing

Gate closed (periodic flushing)Part of Q2 through the gate (permanent flushing)

Gate

Significant reduction of Crest Length

Part of ∆Q (Q1-Q2) is taken care by periodic gate adjustments thus H1-H2 is minimum and Length crest also.

Advantage of Mixed CRAdapting existing CR as Mixed Regulator with the construction of a Weir to control the water level. Advantage: no need to have a permanent operator

Mixed CR: a Very good solution for improvement !

25-05-2018

10

25-05-2018

11

Canal Flow Measurement

25-05-2018

12

Basic Options1. Weirs

2. Flumes

3. Electronic devices

4. Rated sections

Weirs and flumes are based on Critical Flow Devices

For one depth, there is one corresponding flow rate

Critical Flow Devices

For one depth, there is one corresponding flow rate

0

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

6 0 0

7 0 0

5 6 7 8 91

0

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

H e ig h t c m

Dis

ch

arg

e i

n l

/s

1. Weirs

25-05-2018

13

90°H

1

4 H

L

H

L

TRIANGULAR (90° V-NOTCH) TRAPEZOIDAL (CIPOLLETTI) RECTANGULAR

Types of Weirs

Head

Staff Gauge

Weir Crest

Free Flow

Note: Air pocket behind crest

Weir Crest Note: Downstream water

level above crest

Submerged

2. Flumes

25-05-2018

14

Weir

Flume

Y1 Y2 Headloss = Y2 - Y1

Head

Headloss

Diff. Weir FlumesParshall Flumes

Replogle Flumes (RBC)

TADLA MoroccoPK Albania

TADLA Morocco

25-05-2018

15

6:1 downstream ramp

6:1 downstream truncated ramp

FLOW

FLOW

Replogle Flume - Ramp Recommendation

4-in Diameter Pipe

Install flush to concrete on both ends and flush to the bottom of the canal.

Replogle Flume - Flush Pipe

25-05-2018

16

25-05-2018

17

WinFlumeUSBR-Denver www.usbr.gov/pmts/hydraulics_lab/winflume

Flow Measurements

Accuracy !!!

HH

QQ = H=Q

A key issue:

Accuracy of gauge reading ∆H

25-05-2018

18

Accuracy of gauge reading:turbulence and waves

Higher the Head higher the accuracy

HH

QQ =

25-05-2018

19

Accuracy of measurement

inversely proportional to flow width

TOO WIDEHead is low particularly at low discharge

Head max = 0.60 mAccuracy at Q max=250 l/s for 2cm

Turbulence

HH

QQ =

Accuracy of measurement

inversely proportional to flow width

OK

0

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

6 0 0

7 0 0

5 6 7 8 91

0

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

H e ig h t c m

Dis

ch

arg

e i

n l

/s

153

194

421

481

25-05-2018

20

Accuracy of weir measurements

increases with available head

High Head ==> low width

Low head ==> high width

Accuracy of weir measurements

decreases with width of flow

3. Velocity sensorDoppler

25-05-2018

21

25-05-2018

22

4. Rated sections

Using rated sections

No measurement without periodic

calibration

THANK YOU