Title Experimental Study on Sedimentation over the Floodplain due to River...

Title Experimental Study on Sedimentation over the Floodplain dueto River Embankment Failure

Author(s) ISLAM, Md. Zahurul; OKUBO, Kenji; MURAMOTO,Yoshio; MORIKAWA, Hiroshi

Citation Bulletin of the Disaster Prevention Research Institute (1994),44(2): 69-92

Issue Date 1994-06

URL http://hdl.handle.net/2433/125001

Right

Type Departmental Bulletin Paper

Textversion publisher

Kyoto University

Bull. Disas. Prey. Res. Inst., Kyoto Univ., Vol. 44, Part 2, No. 380, June, 1994 69

Experimental Study on Sedimentation over the Floodplain

due to River Embankment Failure

By Md. Zahurul ISLAM, Kenji OKuBO, Yoshio MURAMOTO and Hiroshi MORIKAWA

(Manuscript recieved on June 2, 1994, revised on September 21, 1994)

Abstract

River embankment failure and floodplain sedimentation are common and frequent natural disasters in alluvial rivers. In Bangladesh, for instance, every year hundreds of embankments are damaged by flood water resulting in crop damage due to inundation and sedimentation. To elucidate the process of sedimentation over the vegetated floodplain due to embankment failure, a basic series of laboratory experiments was carried

out with different lengths of embankment breach and different heights of model plant in the floodplain. The sediment volume and thickness over the floodplain increased with the increase of the breach length in the embankment. The maximum thickness upto a definite limit attained in the early stage was almost equal to or less than the water depth in the floodplain. Sedimentation was also observed to be greater in the floodplain with the more retardative jute field than in that with the rice field. Volume and area of sedimentation increased significantly with the increase of sediment inflow into the floodplain, namely the flood duration, the discharge and sediment concentration in the river.

1. Introduction

Bangladesh is a disaster prone country where flood and river embankment failure resulting in sedimentation over the floodplain are common and frequent natural hazards. The major river system, the Ganges-Brahmaputra-Meghna and their tributaries and distributaries carry 13 millions of tons of suspended sediments per day

(Coleman, 1969). Choudhury (1990) noted that about 80% of the surface flows con-centrate during the monsoon (June to September) that carry about 2.4 billion tons of

sediments a year through the river system of Bangladesh. These sediments settle over the river beds, floodplains and low lying areas during the inundation and with the reces-sion of flood water. As a result river channels and their distributaries are silted up with sediments composed of fine sands and silts causing drainage congestion and overbank flow resulting in river embankment failure and sedimentation over the floodplain caus-ing a severe damage to agricultural crops (Ahmed et al., 1992, Bangladesh Water Development Board, 1987; Hogue et al., 1991; Klaassen et al., 1988 and Pramanik, 1993). An investigation carried out in Institute of Flood Control and Drainage Research, Bangladesh University of Engineering and Technology, showed that only ten breaches of eight embankments caused sediment deposition to amount to more than 16 million m3

7 0 Z. Islam, K. Okubo, Y. Muramoto and H. Morikawa

over the floodplain vegetated with young rice and jute plants causing damage to agricultural crops of about 66,700 hectares (Islam, 1991). The simply averaged thickness of the sedimentation within the flood affected areas was obtained as 2.4 cm. Meanwhile it was found from the field survey after the flood that the actual thickness of near breach deposition in the vicinity of embankment was at least ten times higher than the average. It was also noticed that the deposited material was rather silty and river-born. Little is known, however, about the sediment inflow condition through the em-bankment when the breach was actively taking place. Therefore, it seemed that some basic experiments would be needed employing a physical model to observe the whole

process of sedimentation over the floodplain due to river embankment failure with different breach lengths and vegetation types, by controlling flood duration, discharge and sediment concentration in the river channel.

This paper describes the results of the experiments carried out using a physical model in the hydraulic laboratory at Disaster Prevention Research Institute of Kyoto

University with the following objectives: i) To study the effects of different breach lengths in the river embankment and of

different heights of vegetation on the floodplain sedimentation, ii) To observe the whole process of floodplain sedimentation by changing the duration,

discharge and sediment concentration in river channel. This kind of study is rare in the available literature except for a few recent studies;

Takahashi et al.(1988) and Muramoto et al.(1989) demonstrated overbank flows and sedimentation processes due to floods of rivers. Nakagawa et al.(1989) and Shieh (1989) described the sedimentation due to sediment laden jets entering the basins at a right-an gle .

2. Experimental Set-up

The experiments were carried out by using a distorted model according to the scheme shown in Table 1. The plan and side views of the model are shown in Fig. 1. The working section connecting the up and downstream basins is 200 cm long with a constant longitudinal slope of 1/300. The model consists of the 15 cm wide river chan-nel and the right side floodplain with a width of 60 cm. The floodplain is vegetated with the assumed rice and jute plants artificially made with thin plastic sticks with the plant heights (Hg) of 5.5 mm and 25 mm, respectively. The water depth (dw) in the river and floodplain was kept always 3.6 cm.

Both rice and jute are common crops in Bangladesh, with both plants being about 100 times higher than the corresponding artificial materials selected for the present ex-

periment. Thus, the vertical length scale ratio was taken as 100 and the corresponding water depth in the field is 3.6 m, which is a typical flood depth in this country. On the other hand, the horizontal length scale was based on the channel width, assuming the length ratio of 1,000. Then the corresponding width of flooded river becomes 150 m. The Froude similarity is expected for inflow discharges at breached sections, where the velocity is proportional to the square root of water depth, which reduces the velocity ratio

Experimental Study on Sedimentation over the Floodplain due to River Embankment Failure 71

Table 1. Scheme of the experiments

LTfHv Run Experiment b, Type (cm) (min.) Vegetatioon (mm)

1 EBSR1 5 20 rice 5.5

2 EBSR2 10 20 rice 5.5

3 EBSR3 15 20 rice 5.5

4 EBSR4 20 20 rice 5.5

5 EBSJ1 5 20 jute 25

6 EBSJ2 10 20 jute 25

7 EBSJ3 15 20 jute 25 8 EBSJ4 20 20 jute 25

9 EBSJ2.TDP 10 100 jute 25

10 EBSJ3.TDP 15 100 jute 25

11 EBSR2.TD1 10 20 rice 5.5

12 EBSR2.TD2 10 40 rice 5.5

13 EBSR2.TD3 10 100 rice 5.5

14 EBSJ2.TD1 10 20 jute 25

15 EBSJ2.TD2 10 40 jute 25

16 EBSJ2.TD3 10 100 jute 25

17 EBSJ4.DD1 20 20 jute 25

18 EBSJ4.DD2 20 20 jute 25

19 EBSJ4.DD3 20 20 jute 25

EBSR: Embankment Breach resulting in Sedimentation over Rice field. EBSJ: Embankment Breach resulting in Sedimentation over Jute field.

TDP: Time Dependent, Preliminary Experiment. TD: Time Dependent Experiment.

DD: Discharge Dependent Experiment.

0_, ti3

Floodplain vegetated with 0_2 Downstream basin Upstream basin"r)

II %),41model rice or jute plants C4

c) IVBreach Embankment 4 N1/4,4" o— n

m River channel 15 cm -- -..---- =I

.1 1 .111--•

t

PLAN VIEW 200 160 120 80 40 0

1.-----80 cm 1 1111.111.1.80 cm x (cm)

Sediment feeder

i . -------

0 T 1 ... i

1

IA , 4,4 1{ 1 Pump 1 --LI

1 1

1 I

Pump 2

SIDE VIEW

Fig. 1. Experimental set-up for sedimentation over the floodplain due to embankment failure.

72 Z. Islam, K. Okubo, Y. Muramolo and H. Morikawa

of 10 and the advective time and discharge ratios become 100 and 106, respectively. The model embankment was constructed by plastic plates with 7 mm in thickness,

dividing the floodplain and the river channel with breach lengths (Lb) of 5, 10, 15, and

20 cm and with an exit of 20 cm at the downstream of the embankment to deliver water from the floodplain into the channel. Total discharge at the upstream section of river channel for the experiments of EBSR and EBSJ in runs 1-8 was 2,160 cm3/s, and the sec-tional mean velocity was 40 cm/s. Discharge into the river channel was kept constant, while the net discharge onto the floodplain through the embankment breach increased as the breach length increased.

The downstream ends of the breaches in the embankment were fixed at x=80 cm. Before the commencement of the experiments, water was supplied at the up and downstream basins of the model until the attainment of 3.6 cm in depth on the floodplain

as well as in the river channel. A steady state was formed in both the floodplain and river channel when the pumps were started for recycling of water in the model. Cons-tant sediment supply at the upper end of the river channel was ensured with the help of a sediment feeder. Sediment mixers were used to mix the sediments and water in both the up and downstream basins. Turbidity meter (Model PC 106, Tokyo Keisoku Co. Ltd.) was fixed at the upstream end of the river channel to obtain the record of sediment con-centration in the river water. Average sediment concentration was 9,400 mg/I for all of the EBSR and EBSJ experiments and 12,000 mg/1 for the time dependent experiments

(EBSJ2.TDP, EBSJ3.TDP, EBSR2.TD and EBSJ2.TD in runs 9-16), while that for the discharge dependent experiments, EBSJ4.DD in runs 17-19, was kept 11,200 mg/l. Fig. 2 shows the grain size distribution of the used sediment materials with the medium diameter of about 100 micra.

After the former part of the experiment with the elapsed time (TO of 20 minutes, the experiment was carried out for an additional 60 minutes following the same pro-

100'

80

60 -

]40-

20-

01I I II. I I 10 100 1000

d Cum )

Fig. 2. Grain size distribution curve of the sediment materials.


cedure in runs 1-8. To check the results of the earlier runs, the time dependent (TD) experiments, EBSJ2.TDP and EBSJ3.TDP with the embankment breach length of 10 cm and 15 cm, respectively, were carried out with the duration of 100 minutes in runs 9-10, and the experiments of EBSR2.TD1-3 and EBSJ2.TD1-3 with the embankment breach length of 10 cm were carried out for durations of 20, 40 and 100 minutes in runs 11-16. For the discharge dependent (DD) experiments of EBSJ4.DD in runs 17-19, varying discharges of 1,080 cm3/s, 1,620 cms/s and 2,160 cm3/s were supplied in the river channel.

Measurement of sediment thickness for all the experiments was made laterally and longitudinally after every two centimeters at the points of sediment deposition over the floodplain vegetated with a model rice or jute field. It was done using a point gauge and later with the help of the laser beam gauge as a displacement sensor (Model LB10 and the amplifier Model LB60, Keyence). The thickness was measured from the base level of the vegetated floodplain. Reduction in total deposited volume due to the plants was taken into account for later discussion, which were approximated by uniform fractions of

4% for the rice plant and 12% for the jute plant to the total volumes. For determination of length (Ls) and width (W,) of scour holes that formed immediately beyond the breach, several measurements were taken longitudinally and laterally, and the average length and width of the scour holes were obtained.

3. Sedimentation Patterns over the Floodplain

3.1 Sedimetation patterns affected by vegetation

Sedimentation views and their patterns over the floodplain in EBSR and EBSJ are shown in Photos 1 and 2, and Figs. 3 and 4, respectively. It is seen from the photos and figures that both the area and volume of sedimentation increased when the breach length was increased as 5, 10, 15, and up to 20 cm. Similar sediment deposition for EBSR and EBSJ was observed to be higher in EBSJ than in EBSR.

In EBSR, the sediment distribution patterns were observed to be longer in shape over the floodplain along the flow axis of the sediment laden jet through the breaches for-ming long scouring holes immediately beyond the breaches in the land-side. The deposi-tion was closer to the model boundary wall opposite to the embankment across the floodplain. In Photo 1, the sediment deposition looks wavy and covered with ripples

and holes formed over the deposited area. As the sedimentation increased, rice plants became less visible losing their roughness effect. In the midstream of the floodplain, sediment transport lasted, and turbid water came out at the exit of the floodplain.

In EBSJ, in contrast, the main part of the sediment deposition over the vegetation was delineated by a sharp edge beyond which the deposition height rapidly decreased, and was found to be almost smooth and flat without any ripple formed, over which a shallow flow of water was observed. The shapes were almost round forming very minor scouring holes just beyond the breach in the land-side. But the smooth, flat surface of the main deposition would sometimes be disturbed by fine scale channel formation at later stages. Particles were trapped on the outer slope of the main deposition, so that

74 Z. Islam, K Okubo, Y Muramoto and H Monkawa

. , . -. .•' : , .

I,.. _:-• :, .,, ,.... . _ i. , • ' i • ::_-:;.-.-.7:'''''; ' ',--'7:i .r..--,/, .. -'• •• . - ̀ '.. ' i:, - ,;...7.--, ' . "Lk .f., 11 •''','....7''''';''1-' .„,--4,,,-.,-...:,,.,7.,,,,,,, ',.,i''•••"--'le,'y ---'i1' .-' '

,0 . 4e ' 4' - . • - " : ' . . "-' '1. 4. 1„,,,.'-Fi.”;.q.„9, , - .., - .

, ,.`,..

-._ EBPR1

, • _ •,' _ fl I, ' ',',‘ ' ..';;,. '.: .''.::.t.,4

r F'.. ''44a '' '.."' -' 'II,

3iiiirk0,,, ..,,

0 ' ,

' ' - -:'T ̂ --- ' ' ' . ' i ''' i ' ,,,

;

,,

' • .

,

a) EBSR1 : 5 cmb) EBSR2 : 10 cm

' ;.•

/

4 . .

• .....

.,• 1 ,,,,,,, , ...-•,,,,;,,... ,,,,, .„

, ,,,._. „ ...,,,,f* ,. • • . '.- ;,..,,r i .. • • 4 .- ' ' . i" ̂4 • ^ ^

, rqr

.. . . : •

, • iN - .- ' \ - A' •-'14:.Z''''''''.-,•,•-•''''' ' EBSP.3\

• .. ,. .. - , .. „,-, •'''' Ok..;:::• _„-,..‘' ' - ••,-.., € •• -

'

...'''-ar EssR4

,-_- ,.. .. , ,. . .

.. . _

, .., . ,

. .

c) EBSR3 . 15 cmd) EBSR4 : 20 cm

Photo 1. Sediment distribution over the floodplain vegetated with model rice plants in the exp EBSR

Experimental Study on Sedimentation over the Floodplain due to River Embankment Failure75

, 1 ' , .• ' -4- .,

... ,- ,....„.„---

l. 1011°Pr"-

f !:- .", , ,, ,,,,

,,,,•7:44, , 1 -

:, ,,,,,,t,. ;"' ,,,,',.;.,;r-1 '''.-:::i!Eli,-....:.,.---..,c:-,:....", . ,,,-N,„,,,,,.........::',,,:67N,...,t ,'-:::....-::::::::::,:.:;::,-:L -',:;.:",

, . ,,, 4, •,, ' -.,-_,,,,-., ----,,,,,,::,...„:',,L: ,, .: -- .- - - _,-, , -:",,..;,:, ,,,,,,,;., ,,,,,-, ,,,,, ,- , , . :,.., , _ ,,.,,, ',,,-,,,,,,,,,,,,,. : - ..- - , , .,,,,,,,..: ,.,,,.,-,,..-,,-..•-•,,, :- ,

, _ - -, • 1,-,. - .= ';,,,,-:' . , L ,

, , L

,..,, .., .

a) EBSJ1 : 5 cm b) EBSJ2 • 10 cm

4 44 -.wriil 1- .-; - --'il

,. ,,,„,

, • Elr''?" , : I '. i I ' '

.,.-,L.4...,,

iN ....: ,,,,.n,,,,,,,

....,.,,

.. „.y1

. . .,, — .

. .

, • , ° ',-.. ., , ,

. _

i

c) EBSJ3 : 15 cm d) EBSJ4 20 cm

Photo 2. Sediment distribution over the floodplain vegetated with model jute plants in the exp EBSJ.

76 Z. Islam, K. Okubo, Y. Muramoto and H. Morikawa

0- . (mm)

Y 10. (it'A i

48

" on

- i (chi)31.01.1.

40.2 50 14:1:13INIA. 60 --.1—

0 20 40 60 80 100 120 140 160 180 200 x ( cm)

0- ,,,,,,,, 10-'`;''' ,.•"B Yon

(cm) 3D-b I , ill4 30-ii. ..

40- 2 50- EBSR2

60- 0 20 40 60 80 100 120 140 160 180 200 x(cm)

0- (mm)

Y 10-IQ ,,, ,ii1:

. 1 (cm) 146 20 1440J„..r 1110114 30-St

1_. %i 40-2 50- 113.1*1?.-INN;II"' ''PlI0 601I , I ,I1

020406040100120 140160180 200 x ( cm)

0 ) I0- limot 4.sr: (cm) 30-HIM :

40-'1 50- 10:1:1ZI11111

* 4,a*ir,116.IL..,1[11..11P.I. , 7.IIla

tl:.IllII

60- ,

0 20 40 60 80 100 120 140 160 180 200 X(cm) Fig. 3. Sediment distribution patterns in the floodplain vegetated with model rice plants.

downstream water of the floodplain was not turbid, suspending less sediment. The ratio of the sedimentation volume (Vsbc) beyond any corresponding contour

line to the total sediment volume (Vgd) are expressed as a percentage, Vsbc/Vsd, and that of the sedimentation area (Asbc) beyond any corresponding contour line to total sediment volume (Asd) are expressed as a percentage, Asbe/Asd. The relations between VsbeNscl,

Asbc/Asd and the dividing contour line (Zsi) are presented in Fig. 5. The figure confirms that Vsbc/Vsd and Asbe/Asd attained 100% earlier in most of EBSR when the Zd was only around 22 mm, while only about 50% was attained in EBSJ when the Zs1 was 22 mm. The ratios, Vsbc/Vsd and Agbc/Asd attained 100% at a later stage in most of EBSJ when the

Zsi was around 35 mm and was limited by the water depth. The ratios indicate some con-figurations such as degree of concentrated deposition in the confines of the field near a breach. Thus, vegetation would affect the sediment volume and thickness through its roughness effect. Less sedimentation both in volume and thickness was observed to be deposited until the height of vegetation, Hy, was reached, but then increased sharply


O. (mm) 10. •

(cm)LU•6 30. 40. I 2

50. EBSJ1 60.,

0 20 40 60 80 100 120 140 160 180 200 x(cm)

0 ( mm)

i 10 10 Y.1t (cm)LU

304i-r-1111 24 6 40

50 Es 60-•

b 20 40 60 80 100 120 140 160 180 200 x(cm)

in-0-!I'ref • Y,n (

coLU. 30- 11j-,1 4 40- 0 50- mg

60 0 20 40 60 80 100 120 140 160 180 200 x(ern)

-

.a ,n

( cm)CU., hl !I - 6 30. 40-

50. EBSJ4 601 , - -

0 20 40 60 80 100 120 140160 180 200 x(ern)

Fig. 4. Sediment distribution patterns in the floodplain vegetated with model jute plants.

after the height of 5.5 mm and the thickness of 25 mm was exceeded for rice and jute

plants in EBSR and EBSJ.

3.2 Effect of breach length on floodplain sedimentation

In the experiments, the downstream end of the embankment breach was fixed at the location of x = 80 cm, and the upstream end of the breach was shifted upwards at different locations of x=75, 70, 65, and 60 cm to form the embankment breaches of 5, 10, 15, and 20 cm, respectively.

Scouring holes just beyond the breach were seen especially in EBSR. Fig. 6 shows the relations between the breach lengths (Lb) in the embankment and the scour lengths (Ls) and widths (Ws) for EBSR. The L, and W, of the scouring holes just beyond the breaches in the land-side were found to increase with the increase of Lb in the embankments. On the upstream and downstream sides of a scouring, there were two sediment


100 " EBSR

ite 80 _, EBSR3 5-• - EBSR 40'

X40- g EBSJI. EBSJ2 20 - ESSJ3 ESSJ4

10 20 30 40 z.„ mm

100 EBSR- go _ EBSR2110 EBSR 3

- EBSR4-7 "60 .: •

; ; -r# EBSJ SEISJ2--• .4 29EBSJ3-. EBSJ4--,

10 29 30 40 za1, mm

Fig. 5. Relations between Vsb/V,,,, Asbc/Asd and Z01.

ridges with the upstream one being well developed. The sediment thickness in the upper reaches was observed to be higher than ones in the successive lower reaches.

Sediment thickness in the main deposition portion just beyond the embankment breach was observed to be limited up to dy, and did not take place further with the increase of Lb, while new development in sedimentation was observed in the upstream ridge as well as in the lower reaches of downstream with the increase of Lb. The breach length changed the net inflow discharges of water and sediment, the effect being similar to that of elapsed time, which will be discussed in the next chapter.

3.3 Profile, volume and thickness of sedimentation

The sediment deposition height was found to be higher at the upstream portion within x=60 to 120 cm, then gradually decreased towards downstream. The lateral maximum thickness of sediment deposition (Zmax) along x is presented in Table 2 and Fig. 7. In the figure, formation of the main deposition with flat top is clear in EBSJ

50-

48- •

(9 30 -

15 -

S 10 A 20 cm

E, Length of scour + Width of scour

Fig. 6. Relation between L,, W, and Lb in the experiments EBSR.


rather than in EBSR. The Z„,,, was observed to be the highest in EBSR4, and was 32.6 mm at x=70 cm, y=7 cm and were followed by EBSR3, EBSR2 and EBSR1. For

EBSJ, Zmax also did not exceed d„,, of 3.6 cm except in EBSJ4, where the highest Zmax was found to be 37.8 mm at x = 88 cm, y =1 cm. This was because of channel formation on the flat part of the main deposition near the breach.

The laterally averaged thickness of sediment deposition (Zmi) along x is shown in Fig. 8. The Zmi was observed to be the highest in EBSR4 and EBSJ4, 18.7 and 25.6 mm, respectively. Mean sediment thickness (Zsm) and the net volumes of total sediment depositions (Vsd) over the floodplain for EBSR and EBSJ are shown in Table 2 and Fig. 9. The highest Zs„, was observed to be 13.24 mm in EBSR4 and 14.53 mm in EBSJ4. The maximum Vsd were 9,200 cm3 and 7,942 cm3 in EBSR4 and EBSJ4, respectively. The results of analysis revealed the common trend of sediment deposi-tion, wherein Zsm and Vsd over the floodplain increased with the increase of Lb.

Table 2. Zni.,Z,„, and V,, over the floodplain

Zrna„ Run Experiment (mm) (mm) (cm3)

1 EBSR1 22.1 8.57 1,359 2 EBSR2 23.1 8.61 1,560 3 EBSR3 23.6 10.70 4,469 4 EBSR4 32.6 13.24 9,200 5 EBSJ1 29.0 11.32 1,980 6 EBSJ2 33.5 15.48 3,749 7 EBSJ3 34.4 14.01 6,074 8 EBSJ4 37.8 14.53 7,942

a] EBSR

30 - re*. f "641,%4 20 - ,00,4.% EtA

vikt. NI "d Le -

t„

•

0 42 62 02 102 722-14.2 162 -..1023..

x, CM

o asi21:5 cm + EB3Fc2:10 cm 6 E13883:35 cm a BER4:20 em

40 IDI EBSJ

30 - *f ?,), 4 i 4\

tattase4etta

I

16 a,

• „„„ • .

38 58 78 98 118 138 158 179 198 x, cm

0 Ma:5 cm + E0232:10 cm 6 E33.13:15 cm a EB314:20 om

Fig. 7. Zmax along x in the experiments EBSR and EBSJ.


Aaa] EBSR lA 16

ats,/-AABA *4" a oft 4:404,

' 0.'"itelegoNit 4,01141%,

1 T., g ^.,Zam

4 1

42 62 82 102 122 142 162 182

X, an o EBSR1:5 om + EB8R2:10 cm 0 E88R3:15 an A EBER4:20 an

30 - EBSJ

25 - +

fe47- tha 20 -

15

10 -AOT

_ J

0 38 se 489's lig

X, an E89J1:5 an E88.12:10 cm C EDST3:15 cal a EBS14:20 CM

Fig. 8. ZE,,/ along x in the experiments EBSR and EBSJ.

The relations between the volume of total sediment inflow (V81) into the floodplain and the volume of total sediment deposition (V.d) over the floodplain in the model are

presented in Fig. 10. Considering the fact that the constant river velocity was used instead of the net inflow velocity into the floodplain to estimate Vsi, the trap efficiency of sediment in the jute plants is rather high. The Vgd in both EBSR and EBSJ were ob-served to increase with the increase of V51. The comparison between EBSR and EBSJ shows that sediment distribution patterns greatly vary according to the different vegeta-

16

12 -

8 EBSR 1313S.3 4 -

0 I- 1'0 15 2.8

Lb. .11,

(x10') 10 -

8 -

6

>' 4 - O EBSR

+ EBSJ 2 - o

1 , 5 14 15 20

Lb. cm

Fig. 9. Relation between Zam, V,d and Lb over the floodplain.


10' :

Q.

A

m• 8-6

a/0 /• >.

/.

o

103.- . i 103 10' v", cm'

0 EBSR bi EBSJ Fig. 10. Relation between V,d and V.; in the floodplain.

tion involved. The efficiency of sediment deposition or trap rate as a whole as seen in Fig. 10 reveals that jute plants are more effective than rice plants in sediment deposi-tion over the floodplain.

After the completion of the former part of the experiments lasting 20 minutes, the ex-

periments were continued for 60 minutes more. Results in this latter part showed that the sediment deposition took place up to a definite limit of maximum thickness and was equal to or less than the cl,„ in the floodplain as inflow discharge through the floodplain reduced significantly due to the formation of a sediment deposition barrier in the flow. However, sediment deposition in volume and size was found to increase gradually over the floodplain in the latter part of the experiments than those in the former part, as Te increased.

4. Process of the Floodplain Sedimentation

4.1 Time evolution

To observe further the sediment deposition process over the floodplain due to an embankment breach of 10 cm, the further experiments, EBSR2.TD1, TD2 and TD3 and EBSJ2.TD1, TD2 and TD3, were carried out with durations of 20, 40 and 100 minutes. Perspective views and evolution of patterns of the sedimentation process in EBSR2.TD and EBSJ2.TD are shown in Photos 3 and 4 and Figs. 11 and 12. In EBSR2.TD, after 20 minutes, the deposited portion with the sharp edge similar to that in EBSJ experiment, was established immediately beyond the breach. This is clearly seen in Photo 3.

A comparison of the transverse deposition, Zsi.tdp and ZsLtd at x=80 and 94 cm in EBSJ2.TDP, EBSJ3.TDP and EBSJ2.TD is shown in Fig. 13. TDP in the figures means that a point gauge was used in the experiments instead of the laser beam gauge. The figure shows that the initial configuration of sedimentation was formed within

82 Z Islam, K. Okubo, Y Muramoto and H Morzkaz.va

a] EBSR2 TDB 20min' " Jr

r

I 100

b] EBSR2 TD2 40min Iltdrigwr.

' r

"-*

,Z7M:

f

( - TD3 : 10 in.,

_ .

Photo 3. Time dependent sediment distribution over the rice-vegetated floodplain due to embankment

breach

Experimental Study on Sedimentation over the Floodplazn due to River Embankment Failure83

'-1 al EBSJ2 . TD1; 2Orn,in,1-xi

.41.•••,',,,_f,k Is. ,-,'-#,,,J'., Li ,N:1,1,. . - ,4-,,,i .

-,

_

- '

......,

1 , --_-7_,- - ...i...1,

,,,,--'', ,,--Vef.re b1 EBSJ2 . TD2 ; 40min ,1,, -' "IP3, Ifill,,.iriki

'-,-,,filr'',,-...-.,,4,,,--,

i, • ,.. ..,,-,.-_,,,,,,,..,„..,-.. ,, -,, R:„1,...,4,-4-1---..,,,,.ii.41;ft ' 4,' '1'11.0T!Al',-2 4`+ti'i0 ,„' ,;1_,.,' i::.,

1it,,ip''I''' ri,,-,.,:./i,;, 4,-,'.:,q,1,,...

,_,i',..", '1t^LY'''L:'I^,''''1,

i,-,:n -r.*•*1r.-,,',41),","; '''''',,..:

t:,1- ? ,--114,;',e,-,1-,.,,Jl'',,6-r.,f,-'1?,,- ..i'',10N-',:''I,-HF.,:,',:-7' ,-.-ir ,4,,_I-•:A,, 0*

,'':,,„..?r.',,I.,,;,c ”'l•\._ d

‘44441111000 -,,q,,,pi- , 'N'tfi

i ' ''''V-...^.'k' : '',„ , , ,i',''l'/"')%1"'

.,J

i', _ '•, t'', ,,'',P''''..i -,; ..:'' '.A....'''''..,1cr, ',,,v.- .-_,,..,,,t,/

,f .,.... ,..' ,r,/fJ,''' 1 „

41:i 'i4mw i), 41-',',

d.'

1,1ST? i:

Photo 4. Time dependent sediment distribution over the jute-vegetated floodplain due to embankment breach.

84 Z Islam, K. Okubo, Y. Muramoto and H. Morikawa

0 (mm) 10 IV, .T5s1i110''9 ii1.0 Ye-, *

.I.: (cm)IIIL5 cu ),..,11.,.;,, 111 4 1 30 ,.'*1^! 1

.4---*ill a 40

a

50 14:30.111110 60 0 20 40 60 80 100 120 140 160 180 200 x( co

0 (m.)

Y 10 -,,,--31*11' i le v,,,,'NI(8 (cm) 20 .1:-:111111iill?,.1.:,;,6 30',',,

40iliNicIli-, 2 1l '''ilitirdiull 0 50 14:1:011110.111

60 "o 20 40 60 80 100 120 140 160 180 260 X(cm)

o io(—) L.,.,.„II].r.'ini.e8 Y on ,P 6 (cm)LU111/4,.,.l?„L,.ill Ii 1,111111:,-ti- "I_ WI 30 4

402 'LIi' _''!!:.111A a 50 10:11a1INIII ',..44-5 .'.'• ' ,11 60 "o 20

4U 60 80 100 120 14U 16U 18U 2UU x(cm)

Fig. 11. Time dependent sediment distribution patterns over the rice-vegetated floodplain due to embankment breach.

0 _ (mm) 10 ' '''''-.-',*'1•"e

on1,(cm)LU.1.1,„''''''?-11:1 IIIIII 30TIL.04- .,:.? .,-_iik,

40 el 50 14:311121101

60- 0 20 40 60 80 100 120 140 160 180 200 H(cm)

0(..) ,,,,riAi18 8 10 kr, "---,444,.,,tz,:d., Yon. .04, ,\',V)lillp'' (cm)Lu4,,1 ,^101 6 30 4

a

50 [0:31141111EN 60

0 20 40 60 80 100 120 140 16U 180 200 X(cm)

0 ,—,

Ye 10••: ,`.:''..7.„.:.:1.'',:;:•._10 („) 20'-:,1;',..:71:;, 115 A 6 30-4-

1-,,,:—..-.'44i11114 40-:-L _ET7-"--.i 0_r,...,2 50 VISION-III

60 0 20 40 60 80 100 120 140 160 180 200 x(cm)

Fig. 12. Time dependent sediment distribution patterns over the jute-vegetated floodplain due to embankment breach.


EBSJ2. ID • x=80 cm 0 x=94 cm

EBSJ2. TDP------',=St3tccmn,EBSJ3. TDP':,V4.1c.mm zsi Cm) ' Zs 1 Cmm)

40 40 10(3c....=.........515:10"lalt.!^,sg] 2: 108 min l--'7—..-\\ ]2@

80 sr 'rn7 ] 80 17 - ...A. 70/r-----------7\] 70 -7- -\\

60 .,/______ SO fe-7/ - '...\ ] 50 ''''N ] 58 7-77-------77\ ]

So ''''''.... 3e ---77----7S , °

. 20

ie ie ....(7='''Th ] ....._...

Iiiiiiiiiiiivi LiLiiiiiiiimil IiiiiiiiiiIiiitilittlii(Liiiii.1 60 40 Y(cm) 20 8 60 48 Y (cm) 28 0

(a) (b)

Fig. 13. Zs(.tdp and Zml.td along y at x=80 and 94 cm.

10-20 minutes and after that a gradual increase was observed in thickness. In EBSJ3.TDP, newly developed, remarkable sedimentation zone and increase in volume and area was observed with longer times. This process was similar to those of the length dependent basic experiments, EBSJ in runs 5-8.

The thickness of lateral maximum sediment depositions (Zmax.td) along x for EBSR2.TD and EBSJ2.TD is presented in Fig. 14. The highest Zmax.td as shown

40 ?..P .-404ewe's.. .) E3SR.TD

•-

.0-'.:044*+÷÷-H4N..

,,,..„4 .,:.,... Imo, .,_..-

.20 -: .i.: .„-.1t4e .; -..t ii :•

•

" le

• . 1 : 0

46 6606 lei 124

X, CM 0 EE822. TM (26 min) + ERS122.1D2(49 min) 0 E0R2.TD3(100 mini

50 b] EBSJ .TD

40 -4,00., lb..,

.-”..,1-,:...-1-4.4.-1.-.'44,474,

1 6 ...-..,; ' - , )_•.,Ii. -; . .. . -i.,

3 20 - : 0.%

. .

.: I 10- :,

-_,-------- sa'70 99 iie

x, cm 0 E33.12.TD1(20 min) + E138.12.TD2(40 min) . 1561 2.103(100 min)

Fig. 14. Zmax.td along x in the experiments EBSR.TD and EBSJ.TD.

86 Z, Islam, K. Okubo, Y. Muramoto and H. Morikawa

Table 3. Zmax.td, Vs1 td, Z,,,, ,d and Vsd.td over the floodplain

Run ExerimentTime Zmax.d vg.td Zsm.td V„,,d p (min) (mm) (cm2) (mm) (cm3)

11 EBSR2.TD1 20 30.7 72.8 18.42 2,610 12 EBSR2.TD2 40 35.6 110.9 16.80 3,180 13 EBSR2.TD3 100 39.4 137.5 19.17 6,738 14 EBSJ2.TD1 20 32.2 79.2 21.83 2,312 15 EBSJ2.TD2 40 35.7 109.9 25.23 3,628 16 EBSJ2.TD3 100 39.5 124.0 25.81 5,216

in Table 3 was observed to be 39.4 mm in EBSR2.TD3 and 39.5 mm in EBSJ2.TD3 in 100 minutes. These points were made in relation to the fore-mentioned fine channel formation on the major deposition, which made sediment transport outside the channels inactive and left some flat bars or islands. The increase of Zmax.td at 100 minutes in comparison to 40 minutes was 10.7% in EBSR2.TD and 10.6% in EBSJ2.TD.

The laterally integrated sediment volume per unit longitudinal length (Vg.,d) along x for EBSR2.TD and EBSJ2.TD are shown in Fig. 15. The highest vsud shown in Table 3 were 137.5 cm2 in EBSR2.TD3 and 124.1 cm2 in EBSJ2.TD3 at 100 minutes. The ratio of vsLtd at 100 min to those at 20 min in EBSR2.TD and EBSJ2.TD are 88.9% and 56.6%.

The mean thickness (Zsm,td) and net volume (Vsd.td) of total sediment deposition over the floodplain for EBSR2.TD and EBSJ2.TD are presented in Table 3 and Fig. 16. The increase of Zsm.td at 100 minutes in comparison to 20 minutes was negligible in EBSR2.TD and 18.2% in EBSJ2.TD. The net volume increased significantly at 100 minutes in comparison to that at 20 minutes by 158.2% in EBSR2.TD and 125.6% in

140 : ..,..6.6.60°., al EESR.TD ,.

120 : io 4, om * • A, tk . 100 :4.'it‘

44'

4

-,".4. 4\.,, l'e : 71'; /

f" !., .,4cit : .4, , _, 6 -16 ,64, .6,4. 206 ... .

e..._. , 1• a 46 66 96 ta6 124 x, cm

0 BER2.TD1(20 min) + E9BR2.TD2(40 min) m E3SR2.TD3(100 mini

120- ,*-4, ,6-...td EBSJ.TD ...4, - .'''.e.'4,.+4,4-4--...+.-e0. 100-' 4,,,k.,,,

N .4 16-6' , -- ,,../> ti

•'''' 1

- o'• -`1--r \ 20 - .4 I I - .

50 70 90 110 x, us

0 E08.12.TD1120 mini + EB932.TD2(40 min) 0 EBSJ2.TD3(100 mini

Fig. 15. Va).td along x in the experiments EBSR2.TD and EBSJ2.TD.


26

1 22 4-

14

10 20 4atie ee 100 T, min

o E0032.Th + EEM12.70

( 0103) 7

6 -

_

4 -

3 -

2 'L9 40 b0 ' r 100 7., min

13332.TD + E13.72.TD

Fig. 16. Relations between Z,„,td, V,d3d and Te.

EBSJ2.TD. Using the advective time ratio of 100, the time duration of 20 minutes in the ex-

periment corresponds to 33 hours in the field, which is within a possible range, and the experiment covers long enough time span.

4.2 Effect of river discharge

To observe the discharge dependent sediment distribution process over the floodplain due to an embankment breach of 20 cm, EBSJ4.DD1, DD2 and DD3 were carried out for 20 minutes each with a river discharge of 1,080 cm3/s, 1,620 cm3/s and 2,160 cm3/s, respectively. Perspective views and their patterns over the floodplain in EBSJ4.DD are shown in Photo 5 and Fig. 17. Sediment thickness was observed to be higher than 30 mm just beyond the breach. The difference between DD3 and DD2 in the upper portion was small, while that between DD2 and DD1 was remarkable.

The thickness of longitudinal sediment distributions (Zsimdd) at each different x was found to be higher at the upstream within around x=60-100 cm and gradually decreas-ed towards downstream within the sediment deposited area. The maximum sediment deposition was observed to be almost equal to or less than the ci,„ of 36 mm.

The lateral maximum thickness (Zmdd) for EBSJ4.DD is shown in Table 4 and Fig. 18, which shows the typical shape with the flat top for the case of DD3, and the highest Zniax.dd was 36.0 mm. The increase in Zraa,e.dd in DD3 was found to be 4.3% and 14.3% in comparison to those of DD2 and DD1, respectively. The sediment volume

per unit longitudinal length (vsi.dd) for EBSJ4.DD is shown in Table 4 and Fig. 19. The vsi.dd was observed to be the highest and increased significantly in DD3 by 28.1% and 75.0%, compared to those of DD2 and DD1, respectively. The figure shows increasing

88

, .L

Z Islam,K Okubo, Y Muramotoand H,Matzkatv‘a

ik.

1 FRq.1.4_ nni :1080 crr13, S::-.,',.. t.!„.•',',-.: ,Y

: .' 1.'4'.e'?' ,-'.:.,i':,.i.q.:,..,,4.-g.''"'''',,•la

' : ' ,; ' ' '•}2.;;;:;;;,"., 2. ;.....,1,':':''.437;,, 5 „."4, ' ..,.,:e'., ,,,'.41•"; \ 1 S

' '', , ', „ . if ,,I V,. -..- q•-,,, 1

„

,„., .,,,i,_ iiir

:, ,,i,,::k':...A..:;:-:.!:fi;:kS-i;---:-:1Vr- ,,,,:,,:,‘.,11),..,_:1::^;A:!;:.1i ' ,--:'!:' ,.:::.:'.;

,Ai..,: ,,:,,,..;. ' ,„,,:...,4„.;::.,,,e,..,,,

„

..',..':,.N.,', .y..,,,:,,.f.',;,,::':Cs' .,"V'' .'i!',.°'Y"' '.' ---4 :''S' f' ',Y;,:,:',R"',,'. feJ,'''''''

1620 c3c M/•- bl EBSJ4'DD2 :

lot ,_ „., 'N .

"gill

'Y rf

, r .., ,,,

,

n,m3ic 71—4,1141 c1 EBSJ4'DD3 =

AA

t-li

\

i

Photo 5. Discharge dependent sediment distribution over the jute-vegetated floodplain due to embankment breach.


0 . ( mm) 10 • 1.0. Y on ''''k h4IN-li711:1 17 11 101 r Fli 6

(cm) LL, '"1 ' I ' - 'lin* ' I k ii .1, ill, , -.,' :1111 i - .. IIIIIII 4 30 .

40

" 2 0

50 1:011111, ii I 60

0 20 40 60 80 100 120 140 160 180 200 X(ern)

O. - - . 10.lit ,.911.:-..7'I•-ii4i,,,••'14,...;.t.,,.1111 .1 y'

(cm) 20.'')-,,'-.11.''Ir.11:401liiiiilll' 6 30.70.71filitit'"::..:101Mit';1111111111 4 40. l'i I Q111,,:. ,' .14;p1[1111Q1 Q 2

50. EBSJ,004 0 60. .

0 20 40 60 60 100 12.0 140 160 160 200 x(cm)

0. -.(..) 10..,,tiNI.inBe B

e Y''''...-,'MIIIIIICiiiik)1-li-' (cm) 20.

.9'.*Ii.'3,- !ill i 110111W 1 kll', '.„ im 6 i 4

50.1:3511X1.1• 0 80.

0 20 40 60 60 100 In 140 160 180 200 x(cm) Fig. 17. Discharge dependent sediment distribution patterns over the jute-vegetated floodplain due to

embankment breach.

Table 4. Zmax.dd, Vsl.dd, Zsm.dd and Vsd.dd over the floodplain Discharge Z„,„...ddysLdd Zsiss.ddVsd.dd Run Experiment

(cmIs) (mm) (cm) (mm) (cm) 17 ESJ4.DD1 1,080 31.5 68.8 9.72 3,273 18 EBSJ4.DD2 1,620 34.5 94.0 13.69 6,045 19 EBSJ4.DD3 2,160 36.0 120.4 14.46 7,462

development towards downstream. The mean thickness (Z.m.cid) and total sediment volume (V.d.dd) of depositions over

the floodplain for EBSJ4.DD are presented in Fig. 20. A little increase in Zsm.dd was

48

,3° -

il5 4ki. j20 - i ,,,, -4*.

38 58 '78 98 118 138 188 178

x, cm

0 EBS14.1)81 + EBEL/4.1)32 * EJ3SJ4.1:03

Fig. 18. Z .dd along x in the experiments EBSJ4.DD.


120 - "so 1.* • 100 - 401044 fo 4•

Sae- 40' ecb

.

20

38 58 78 98 118 138 158 178

x,

^ EB8J4.001 + E05.14.1:02 o IEBSJ4.1:03

Fig. 19. v,j.dd along x in the experiments EBSJ4.DD.

20 - --

18

-

14

I12 -10

88- j6 -

4

2 —

idee 1420 2160 Discharge, cue/s

x10'

7 -

156 -

4 -

3 - 1880 1620 21'60 Discharge, ceis

Fig. 20. Relations between Z.,mdd, V.d cid and different discharges.

observed in DD3 in comparison to those of DD2 and DD1, while Vsd,dd increased significantly in DD3 by 128.0% in comparison to that of DD1. These characteristics will serve as reference to the sedimentation process during the rising stage of the flood with increasing discharge.

5. Conclusion

Based on the above discussion on the results of the model experiments the following conclusions are drawn:

(1) The sediment deposition over the floodplain due to the embankment breach increased in volume (V5d and vat) and thickness (Z31 and Zsm) with the increase of the breach length (Lb) in the failure embankments.

(2) Sediment deposition process over the floodplain can broadly be divided into two stages: i) formation of initial configuration and maximum thickness of sediment deposi-


tion upto a definite limit not exceeding the water depth in the floodplain at the main por-tion of deposition just after the breach attained in around 20 minutes; ii) gradual lateral and longitudinal development of sedimentation in new area over the floodplain extend-ing more towards the upstream. At this moment remarkable increase in sedimentation volume and area were observed with the duration of elapsed time. With the fixed downstream edge of embankment breach, formation of sharp ridges in sedimentation are observed upstream of the deposited area with the shifting of the upstream edge of the embankment breach to upstream.

(3) The sediment thickness (Ze, Zmax, Znil and Zsm) and volume (vsi and Vsd) are observed to be higher over the floodplain with the model jute plants in the experiments EBSJ in runs 5-8 than those with the rice plants in the experiments EBSR in runs 1-4. Vegetation resistance plays an effective and important role on the sedimentation over the floodplain. Sedimentation both in volume and thickness was observed to be less over the floodplain until it exceeded the vegetation height (HO and increased sharply after exceeding H. No difference was observed in the final stage of sedimentation over the floodplain irrespective of the different vegetation at the main portion of deposition immediately beyond the breach in the embankment.

(4) In the time dependent experiments, EBSR2.TD and EBSJ2.TD in runs 11-16, the sediment thickness (Z,Ltd, Zm .td, Zniud and Zsm.td) was attained in 20-40 minutes, increasing negligibly even in 100 minutes, while volume (vsLtd and V.d.td) and area of the sediment deposition increased significantly in 100 minutes in comparison to those of 20-40 minutes. Though the initial sedimentation depended on the Hy, the final volume and area of sedimentation depended on long term deposition time elapsed.

(5) In the discharge dependent experiments, EBSJ4.DD in runs 17-19, the sedi-ment thickness (Znu„,dd, Zradd and Z,dd) over the floodplain did not increase remarkably, but volume (Vd.dd and Vsd,dd) and size of the sediment deposition increased significantly with the increase of the discharge in the river channel.

(6) The time and discharge dependent sedimentation processes observed in EBSR2.TD, EBSJ2.TD and EBSJ4.DD, were similar to those of the length dependent basic experiments of EBSR and EBSJ in runs 1-8. The sediment deposition process over the floodplain due to the embankment breach depended upon the volume of sedi-ment inflow (Vgi) into the floodplain mainly composed of five hydraulic parameters: water depth in the floodplain, breach length in the embankment, sediment concentration in river water, discharge in the river channel, and duration of elapsed time.

Acknowledgement

The study was conducted under the financial support of the Japan Society for the Promotion of Science ( JSPS) in the form of a fellowship to the first author which is

gratefully acknowledged.


REFERENCES

1) Ahmed, S. M. U., M. M. Hogue and S. Hossain (1992): Floods in Bangladesh: A Hydrological Analysis, Final Report R01/92, Institute of Flood Control and Drainage Research (IFCDR), Bangladesh

University of Engineering and Technology(BUET), Dhaka, pp. 1-5. 2) Bangladesh Water Development Board: Flood in Bangladesh (1987): Investigation, Review and Recom-

mendations for Flood Control, Dhaka, pp. 31-39. 3) Choudhury, A. M. (1990): Use of Remote Sensing Techniques in Flood Study, Flood Study Forum

1990, 'Floods in Bangladesh: Bangladesh Views', Dhaka, pp. 7-13. 4) Coleman, J. M. (1969): Brahmaputra River: Channel Processes and Sedimentation, Sedimentary

Geology, vol. 3, Amsterdam, pp. 29-141. 5) Hogue, M. M., S. M. U. Ahmed and A. B. Siddique (1991): Effects of Different Planning Options on

Floods in Bangladesh: A Numerical Model Study, Partial Completion Report, R04/91, IFCDR, BUET, Dhaka, pp. 1-5.

6) Islam, M. Z. (1991): Failure of Flood Embankments: Case Studies of Some Selected Projects in Bangladesh, IFCDR, BUET, Dhaka, pp. 1-160.

7) Klaassen G. J., K. Vermeer and U. Nazim (1988): Sedimentological processes in the Jamuna River, Bangladesh, Proceedings of the International Conference on Fluvial Hydraulics, Budapest, pp. 1-2.

8) Muramoto, Y., Y. Fujita and K. Okubo (1989): Floodplain Erosion and Sedimentation due to Over- bank Flood Flow, Proceedings of the Japan-China (Taipei) Joint Seminar on Natural Hazard Mitigation,

Kyoto, pp. 429-438. 9) Nakagawa, H., T. Takahashi and M. Higashiyama (1989): Suspended Sediment Deposition in Flood

Plains due to Meandering River Bank Breach, Proceedings of the Fourth International Symposium on River Sedimentation, Beijing, pp. 587-595.

10) Pramanik, M. A. H. (1993): Floods and Cyclones Disasters in Bangladesh, Seisan-Kenkyu, Vol. 45, No. 3, Inst. of Industrial Science, Univ. of Tokyo, pp. 177-184.

11) Shieh C. L. (1989): A study on the Delta Formation Process in a Laterally Wide Basin, Ph. D. thesis, Department of Civil Engineering, Kyoto University, Kyoto, pp. 10-13.

12) Takahashi, T., H Nakagawa and M. Higashiyama (1988): Sedimentation in Flood Plains due to River Bank Breach, Proceedings of the 6th Congress, Asian and Pacific Regional Division, International Associa-

tion for Hydraulic Research, Kyoto, pp. 401-408.

Title Experimental Study on Sedimentation over the Floodplain due to River...

Documents

Transcript of Title Experimental Study on Sedimentation over the Floodplain due to River...