a study of hydraulic characteristics for ﬂow in equatorial...

Intl. J. River Basin Management Vol. 6, No. 3 (2008), pp. 213–223

© 2008 IAHR, INBO & IAHS

A study of hydraulic characteristics for flow in equatorial riversLAI SAI HIN, Lecturer, River Engineering and Urban Drainage Research Centre (REDAC), Universiti Sains Malaysia, EngineeringCampus, 14300 Nibong Tebal, Penang, Malaysia. E-mail: [email protected]

NABIL BESSAIH, Department of Civil Engineering, Faculty of Engineering, Universiti Malaysia Sarawak (UNIMAS), 94300 KotaSamarahan, Sarawak, Malaysia

LAW PUONG LING, Department of Civil Engineering, Faculty of Engineering, Universiti Malaysia Sarawak (UNIMAS), 94300Kota Samarahan, Sarawak, Malaysia

AMINUDDIN AB. GHANI, Professor, Deputy Director, REDAC, Universiti Sains Malaysia, Engineering Campus, 14300 NibongTebal, Penang, Malaysia. E-mail: [email protected]

NOR AZAZI ZAKARIA, Professor, Director REDAC, Universiti Sains Malaysia, Engineering Campus, 14300 Nibong Tebal,Penang, Malaysia. E-mail: [email protected]

MAH YAU SENG, Department of Civil Engineering, Faculty of Engineering, Universiti Malaysia Sarawak (UNIMAS), 94300 KotaSamarahan, Sarawak, Malaysia

ABSTRACTThis paper presents the results obtained from field measurements taken in several frequently flooded equitorial rivers, including velocity distributions,stage discharge relationships, roughness behaviours and discharge estimation. These have illustrated the large difference in velocity between the mainchannel and floodplain under flood conditions, and the effects of momentum transfer between deep and shallow flow, which include reduction in mainchannel velocity and discharge capacity, leading to a reduction in compound section capacity at depth above bankfull. Another significant characteristicthat has been found is that the floodplain regions behave as storage reservoirs (V = 0 m/s) in most cases due to high resistance of long and thick grassesalong the flood plains (n = 0.07−0.1). Flow resistance relationships have been presented in terms of Manning’s coefficient and Darcy-Weisbachfriction factor, showing the complex nature of flow resistance in the rivers and further explaining the danger inherent in the conventional practices ofextrapolating inbank data for the analysis of overbank flows. Results for discharge estimation have been shown for comparison with actual data, theerrors incurred by applying empirical methods to compound channel flows have been quantified and found to depend on the particular method used.

Keywords: Discharge estimation; flow resistance; equitorial river; overbank flow; velocity distribution.

1 Introduction

A large number of hydro-engineering problems are related toopen flow in compound channels. An understanding of flow incompound channels or rivers with floodplains is essential in prac-tical problems of flood mitigation and floodplain management. Itis therefore important for flow simulation to be correct not onlyon the water surface elevation, but also the sectional dischargeand velocity distribution, during the event of overbank flows.Unfortunately, most of the studies that have been carried out arebased on idealized experimental laboratory investigations. Fieldstudy is rare, partly because compound channel flow conditionsoccur typically under flood conditions when acquisition of datais difficult and sometimes dangerous. In the work presented, anattempt was made to focus on rivers under flood conditions.

Received on February 17, 2007. Accepted on October 31, 2007.

213

2 Open channel flow resistance

In open channel flow prediction, it is usually assumed that the flowis parallel and has a uniform velocity distribution (steady-uniformflow) and that the slope of the channel is small. Under such condi-tions, the convection acceleration is zero, and the streamlines arestraight and parallel. Because of the velocity does not change,the velocity head will be constant; therefore, the energy gradeline and water surface will have the same slope as the channelbottom.

Based on the above assumptions, a series of empirical meth-ods of discharge estimation in open channels and rivers havebeen developed. The simplest of these are uniform flow equationsattributed to Chezy and Manning, with parallel development inpipe flow leading to the Darcy-Weisbach equation. The uniform

214 Lai Sai Hin et al.

equations may be written as follow:

The Chezy equation gives

V = C(RSo)1/2 (1)

The Manning equation gives

V = (R2/3S1/2o )/n (2)

The Darcy-Weisbach equation for channel flow gives

V = [(8gRSo)/f)]1/2 (3)

where V is the average cross-sectional velocity, R is the hydraulicradius = A/P , A is the cross sectional area, P is the wettedperimeter, So is the bed slope, g is gravitational acceleration, C

is the Chezy roughness coefficient, n is the Manning roughnesscoefficient and f is the Darcy-Weisbach friction factor [1].

In analyzing the flow through open channels of regular sec-tional shape and hydraulic roughness, it is sufficient, in general,to use the overall hydraulic radius as the parameter, which char-acterizes the properties of the cross section. It is then possible tocalculate the discharge through the channel from one of a rangeof well-known uniform flow formulas in term of the channelroughness, slope and depth as given above.

However, if the cross-sectional shape is irregular, this couldlead to considerable errors [2–9]. One particularly importantexample of this occurs on the occasion of a compound sectionconsisting of a deep main channel with associated shallow flood-plains. In this case, a sudden change of depth would happen at thetransition between the main channel and the floodplain. More-over, the hydraulic roughness of the floodplain is often greaterthan that of the main channel. The combined effects of the greaterdepth of flow and smaller hydraulic roughness of the main chan-nel can lead to significantly higher velocity than those occurringon the floodplain. This velocity difference inevitably results ina lateral mass and momentum transfer mechanism, which cangreatly reduce the channel discharge capacity.

2.1 Coumpound channel flow resistance

Since many rivers assumed a compound shape at flood flows,it is clearly of considerable importance to have reliable meth-ods of channel analysis. This has prompted a significant researcheffort in the area of compound channels, aims at a fuller under-standing of the structure of flow as well as the development ofaccurate method for discharge estimation. Early work by Sellin[10], Zheleznyakov [11], Dracos and Hardegger [12] identifiedthe presence of a momentum transfer mechanism between themain channel and the floodplain flows. This takes the form of abank of vortices having vertical axes, which formed along themain channel/floodplain interface. The effect of the mechanismis to reduce main channel discharge capacity while increasing theflow on the floodplains. However, since the main channel takesthe majority of flow at depths just above bankfull, the net effectof the mechanism at such depth is to reduce the capacity of thecompound section when compared with that of a simple sectionat the same depth.

A number of studies have been aimed at quantifying the mech-anism in terms of an apparent shear force, which acts at themain channel/floodplain interface, the value of this apparentshear force has been shown to be many times greater than theaveraged boundary shear force. Studies of this type includedWormleaton et al. [2], Knight and Demetriou [3], Myers [13],Knight and Hamed [14], Christodolou and Myers [15]. A widerange of geometry and boundary roughness has been consideredand empirical expressions have been developed. However most ofthe above mentioned works have been laboratory based, usuallyconsidering smooth boundary straight channels with certain ide-alized conditions, and therefore none yet commands wide spreadacceptance.

More recently, much research effort on compound channelhas been directed towards a better understanding of the complexturbulent structure and secondary current, and moved towards thedevelopment of multiple dimensional (2 or 3-D) models [16–25],such as the k–ε, and algebraic stress models. While some successhave been achieved, these methods are too difficult to be used.The present stage of development does not yet encourage its usein normal engineering design, partly because of the complexity,but mainly because of uncertainty about the turbulence coefficientand any variation in it [26, 27].

Laboratory studies have succeeded in uncovering the funda-mental structure of flow in compound channels, but to be useful inproviding guidance for river engineers, such data must be verifiedby comparison with those obtained from full-scale compoundriver channels. In this case, field study would be the best wayto further understanding of flow in compound river channels,as well as evolving accurate methods of discharge prediction.The study presented in this paper is aimed at remedying to someextent the paucity of data from full-scale compound river chan-nels, thereby contributing to the understanding of flooding riverchannel hydraulics.

3 Field study and data collection

The present study [28] was carried out in three frequently floodedupstream reaches of Batang Samarahan River namely River Sen-ggai, River Senggi (B) and River Batu located in Kuching,the capitol city of Sarawak state, Malaysia. These rivers wereselected due to serious floods occurrence during Monsoon sea-son in the past few years. Figure 1 shows the locality map of therivers selected.

The rivers selected are shown in Figures 2–4. It presents thatthe rivers are almost straight and uniform in cross section, freefrom backwater and tidal effect. Table 1 shows the geometricalproperties and surface conditions of the rivers at the gaugingstations for comparison. The cross sections of these rivers areshown in Figures 5–7.

Flow gauging of the rivers was carried out from an adjustablebridge built across the rivers, using the velocity-area method. Aleveling staff has been used to measure the depth of flow, whereasan electromagnetic flow meter was used to measure point velocityat 20%, 40%, 60% and 80% of flow depth at up to 20 verticals

A study of hydraulic characteristics for flow in equatorial rivers 215

Figure 1 Samarahan River catchment (Source: DID Sarawak).

Figure 2 Morphological cross-section of River Senggai.

Figure 3 Morphological cross-section of River Senggi (B).

Figure 4 Morphological cross-section of River Batu.

across the sections. The flow depths and point velocities weremeasured to an accuracy of 0.0005 m (0.5 mm) and 0.0001 m/srespectively. For each measuring point, 3–5 reading were takenand averaged to give a mean point velocity to reduce the error dueto variation in water flow. Some 20 discharges were recorded foreach river, covering a wide range of inbank and overbank flows.

4 Velocity distribution

Velocity distributions at the gauging site of the rivers are shownin Figures 8–10. These figures clearly show that the maximumflow velocity occurs in the central of main channel region, whichdecreases towards the side banks and bottom directions, whereas,the flow velocity on the floodplains is found near to zero in allcases even at high overbank flow.


Table 1 Geometrical properties and surface conditions.

Geometrical properties River Senggai River Senggi (B) River Batu

Bankfull depth, h(m) 1.060 1.306 1.544Top width, B (m) 5.285 5.500 5.150Aspect ratio, B/h 4.986 4.211 3.335Bed slope — main channel , So 0.0010 0.0010 0.0016Bed slope — left floodplain, SL 0.0010 0.00085 0.0013Bed slope — right floodplain, SR 0.0010 0.00085 0.0013Surface condition — main channel Erodible soil Erodible soil large boulderSurface condition — side bank Erodible soil long vegetation Erodible soilSurface condition — floodplain long vegetation long vegetation long vegetation

00.5

11.52

2.53

20 40 60 80 100

Lateral Distance (m)

Figure 5 Lateral cross-section of River Senggai.

0

0.5

1

1.5

2

2.5

0510152025


Gro

un

d L

evel

(m

)

Figure 6 Lateral cross-section of River Senggi (B).

0

1

2

3

4

0 10 20 30 40 50 60 70 80


Gro

un

d L

evel

(m

)

Figure 7 Lateral cross-section of River Batu.

Figures 11–13 show the lateral distributions of averaged depthvelocity for the same rivers. These figures further illustrated theincrease of flow velocity with respect to flow depths. The large

0 2 4 6 8 10 12

Lateral Distance, (m)

0.5

1

1.5

Dep

th,

H (

m)

0m/s 0.2m/s 0.4m/s 0.6m/s 0.8m/s 1m/s

Figure 8 Velocity distribution for overbank flow of River Senggai, H = 1.658 m.

differences in velocity between the main channel and floodplainare due to the different in depth and surface roughness.

At the interface region between the main channel and flood-plain, the velocity is found to decrease rapidly, i.e. from veryhigh main channel velocity to near or sometimes smaller thanthe floodplain velocity. This is due to the significant momentumtransfer and apparent shear existed between the two zones charac-terized by a series of vortices (Vor1–Vor5) as shown in Figure 14.These interactions tend to retard the flow at the interface regionof main channel, while increasing the corresponding parameteron the floodplain.

On the floodplain region, flow velocity remained near to zeroin all observations even though under very high overbank flowconditions. This is due to the very rough surface and floodplainvegetations, which prevent it from flowing. As a result, the flood-plain regions were found to serve as storage reservoir at shallowoverbank flow instead of conveying access water.

When the main channel and the floodplain discharges obtainedfrom field measurements are divided by the respective area sub-jective to flow, the mean velocity for the main channel andfloodplain regions for each river are obtained and shown inFigures 15–17. These results further show that there is a large dif-ference in velocity between the flow in main channel and that onfloodplain. For River Senggai and River Batu, the velocities in themain channel increased rapidly with depth due to the decreasedof relative roughness in the main channels, e.g. the mean veloc-ity for the main channel of River Senggai has increased from0.277 m/s at bankfull depth (H = 1.06 m) to 0.749 m/s at depth,H = 1.658 m. However, for River Senggi (B), the mean veloc-ity is found to increase rather gradually compares to the othertwo rivers, from 0.258 m/s at bankfull depth, H = 1.306 m to


4 6 8 10 12 14 16 18


0

0.5

1

1.5

De

pth,

H(m

)

0m/s 0.1m/s 0.2m/s 0.3m/s 0.4m/s

Figure 9 Velocity distribution for overbank flow in River Senggi (B), H = 1.980 m.

58 59 60 61 62 63 64 65 66 67 68 69


0

1

2

Dept

h, H

(m)

0m/s 0.2m/s 0.4m/s 0.6m/s 0.8m/s 1m/s

Figure 10 Velocity distribution for overbank flow in River Batu, H = 2.42 m.

0

0.2

0.4

0.6

0.8

1

1.2

59 61 63 65 67 69 71 73Lateral distance (m)

Ave

rag

ed d

epth

vel

oci

ty, V

(m

/s)

flood plainmain channelflood plain

Figure 11 Averaged depth velocity for overbank flow of River Senggai.

0

0.2

0.4

0.6

0.8

1

1.2

2 4 6 8 10 12 14 16 18 20


Ave

rag

ed d

epth

vel

oci

ty,V

(m

/s)

H=0.698 H=1.088 H=1.35 H=1.475 H=2.13 H=2.278

flood plainmain channelflood plain

Figure 12 Averaged depth velocity for overbank flow in River Senggi (B).


0

0.2

0.4

0.6

0.8

1

1.2

1.4

54 56 58 60 62 64 66 68 70Lateral distance (m)

Av

era

ge

d d

ep

th v

elo

cit

y, V

(m

/s)

H=1.543 H=1.695 H=1.92 H=2.02 H=2.243 H=2.42

main channelflood plain flood plain

Figure 13 Averaged depth velocity for overbank flow in River Batu.

Figure 14 Series of vortices at the interface region of main channel andfloodplain for River Batu.

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2

Depth, H (m)

Ave

rag

e ve

loci

ty, V

(m

/s)

Main channel

Flood plain

Inbank Overbank

Figure 15 Average main channel and floodplain velocity for RiverSenggai.

0 0.5 1 1.5 2 2.5

0

0.2

0.4

0.6

0.8

1

Depth, H(m)

Ave

rag

ed v

elo

city

, V (

m/s

)

Main channel

Flood plain

Inbank Overbank

Figure 16 Average main channel and floodplain velocity for RiverSenggi (B).

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2 2.5

Depth, H (m)

Ave

rag

ed v

elo

city

, V

(m/s

)

Main channel

Flood plain

Inbank Overbank

Figure 17 Average main channel and floodplain velocity for RiverBatu.

0.263 m/s at higher flow of H = 2.185 m. This shows the signif-icant effects of the side banks vegetation in River Senggi (B) tothe overall flow rate.

On the other hand, the mean velocities for the floodplainregions are found to remain near to zero (< 0.1 m/s) in all casesdue to the ponding effects of the floodplain vegetation. Undersuch conditions, the floodplain regions are found to have littlecontribution to flood flow conveyance capacity.


5 Stage-discharge relationship

When the discharge obtained from measurements is plottedagainst depth of flow as shown in Figures 18–20, The graphs

0

1

2

3

4

5

6

0 0.5 1 1.5 2

Depth, H (m)

Dis

ch

arg

e,

Q (

m3/s

)

Main channelFlood plainTotal

Inbank Overbank

Figure 18 Stage and discharge relationship for River Senggai.

0

2

4

6

8

0 0.5 1 1.5 2 2.5

D

Dis

ch

arg

e,

Q (

m3 /s

)


Inbank Overbank

Figure 19 Stage and discharge relationship for River Batu.

0

0.5

1

1.5

2

2.53

3.5

4

4.5

5

0 0.5 1 1.5 2 2.5

Depth, H(m)

Dis

ch

arg

e, Q

(m3 /s

)


Inbank Overbank

Figure 20 Stage and discharge relationship for River Senggi (B).

Table 2 Contribution of main channel and floodplain in discharge capacity of flooding rivers.

River Senggai River Senggi (B) River Batu

(H-h)/H MC (%) FP (%) (H-h)/H MC (%) FP (%) (H-h)/H MC (%) FP (%)

0.0075 100.00 0.00 0.0333 100 0.00 0.0891 100 0.000.0603 100.00 0.00 0.0777 100 0.00 0.1192 100 0.000.1130 100.00 0.00 0.1153 100 0.00 0.1958 99.62 0.380.1770 100.00 0.00 0.1661 98.44 1.56 0.2142 99.29 0.710.2234 97.94 2.06 0.1894 98.11 1.89 0.2356 98.78 1.220.2838 97.73 2.27 0.2701 91.27 8.73 0.2665 97.81 2.190.3161 97.32 2.68 0.3059 89.79 10.21 0.3116 95.55 4.450.3321 95.71 4.29 0.3284 86.07 13.93 0.3382 94.10 5.900.3607 92.12 7.88 0.3873 84.07 15.93 0.3620 91.71 8.29

show that below bankfull level, the rating curves behave asexpected, in which the discharge increases accordingly withdepth of flow.

When the flow is overbank, all the plotted graphs have showna discontinuity, i.e. reduction of discharge when the flow isoverbank, due to the interaction between the main channel andfloodplain, following by a more rapidly increase of discharge atlarger depth due to larger areas subjected to flow. The interactioncan significantly reduces the main channel velocity when the flowis overbank. For River Senggai (Fig. 18) and River Batu (Fig. 19)with very obvious roughness differences between the main chan-nel and floodplain, the discontinuity starts at the bankfull level,in which the discharge at bankfull level is found larger than thosefor just overbank levels, even though it has a smaller flowing area.For example, the discharge for River Senggai at bankfull level(H = 1.06 m) is 0.903 m3/s, whereas, the discharge for overbankflows of H = 1.128 m and 1.155 m are 0.855 and 0.898 m3/srespectively.

For River Senggi (B) with similar roughness in the main chan-nel and that on the floodplain, the reduction of discharge in mainchannel is not clearly seen at the bankfull level but after a certainstage of overbank flow, i.e. (H-h)/H ≈ 0.2. The main reason forthis is that, when the flow is just overbank, the flow at both sidesof the interface region is very slow moving due to the side bankvegetation and that on floodplain. When this happened, the dif-ference in velocity is small at the interface region, and thus theinteraction effect is not clearly seen.

When the flow in the main channel and that on floodplainsare considered separately, Figures 18–20 together with Table 2further show that the main portion of discharge for overbank flowis carried by the main channel region, especially when the flow isjust overbank, e.g. the discharge on floodplains equal to zero forflow depth (H-h)/H ≤ 0.15, and >90% of the discharge is carriedby the main channel for depth ratio (H-h)/H ≤ 0.30.

The contribution of the floodplain regions in the total dischargeis also varied from river to river, for example, for a depth ratio(H-h)/H of 0.35, Table 2 shows that the contributions from thefloodplain regions of River Senggai, River Senggi (B), and RiverBatu are approximately 6.5%, 14.7%, and 7.0%, respectively.These results show that for the rivers investigated, the contri-bution of flow from the floodplain is minimal except for River


Senggi (B) with comparatively smaller main channel discharge,and thus the percentage of flood discharge becomes higher.

6 Flow resistance

The resistance to flow in the main channel region for each riverhas been calculated in terms of Manning roughness coefficient,n and Darcy-Weisbach friction factor, f . Selected graphs areshown in Figures 21 and 22. In each case, the graphs plotted canbe divided into two distinct zones:

The first zone is characterized by the inbank flow of the rivers, inwhich the Manning coefficient and Darcy-Weisbach friction fac-tor are decreasing linearly with flow depths due to the decreasedof relative roughness in the main channel region.

Generally, for the main channel regions, the Manning rough-ness coefficients are similar for River Senggai, River Senggi (B),and River Batu, with values range from 0.07 to 0.10, whereas thevalues of f calculated range from 0.54 to 0.89. This shows thatthe surface roughness for the main channel regions of the selectedrivers are much higher than that of laboratory compound channelsstudied before, which normally have main channel roughness,n ≤ 0.01.

The second zone is characterized by a sudden increased ofroughness value when the flow is overbank. As the surface

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

-0.2 -0.1 0 0.1 0.2 0.3 0.4

Depth ratio, (H-h)/H

Dar

cy-W

eisb

ach

, f

River Senggai

River Batu

Figure 21 Variation of Darcy-Weisbach friction factor with depth offlow.

0

0.02

0.04

0.06

0.08

0.1

0.12

-0.2 -0.1 0 0.1 0.2 0.3 0.4


Man

nin

g,

n

River Senggai

River Batu

Figure 22 Variation of Manning coefficient, n with depth of flow.

properties in the main channels remained the same, such an incre-ment is considered due to the interaction mentioned previously,which slow down the flow in main channel. For example, the n

and f values for River Senggai have increased from 0.078 and0.577 at the bankfull level to 0.092 and 0.771 at (H-h)/H = 0.082,before they continue to reduce at higher depths. This has pointout the danger of extrapolating the inbank flow resistance foroverbank flow estimation.

For the floodplain regions, the value of n and f for each riverhas also been calculated. However, as the velocities on the flood-plain are always very close to zero, so in this case, the value ofroughness determined is very big and practically meaningless forfloodplain analysis.

7 Discharge estimation

The results above show the complex nature of flow in floodedrivers, and to underline the danger of using inbank data as a guideto overbank flow behaviour, discharge estimation is carried outusing the various traditional methods, e.g. single channel method(SCM), vertical divided channel method (VDCM), and horizon-tal divided channel method (HDCM). The roughness coefficientused for the main channel region is that at bankfull depth, whichis likely to be the value chosen in the absence of data from over-bank flow, whereas for the floodplain region, a recommendedvalue by the Department of Irrigation and Drainage of Malaysia[29] of n = 0.25 has been used.

Discharge estimation is also carried using two experimentalmethods to review a possibility of application of these methodsderived on the basis of flume experiments, in more complicatedconditions of natural rivers with overbank flow. These methodsare: (1) Weighted divided channel method (WDCM) [8]; (2)AreaMethod [22].

The results obtained are plotted in Figures 23–25. Also plottedare the observed data for comparison. These results show thatfor inbank flow, the discharges estimated match closely to theobserved discharges, which imply that the inbank discharges areable to estimate accurately using traditional method, providedthat an accurate roughness coefficient is used.

When the flow is overbank, the discharges are over- or under-estimated depending on the method used. In most cases, theVDCM and Area Method are found to over-estimate the total

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

-0.2 -0.1 0 0.1 0.2 0.3 0.4

Depth Ratio, (H-h)/H

Dis

ch

arg

e, Q

(m

3/s

)

VDCM

HDCM

SCM

WDCM

Area Method

Q Observed

Figure 23 Comparison of observed and predicted discharge for RiverSenggai.


0

1

2

3

4

5

6

-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5


Dis

char

ge,

Q (

m3 /s

)

VDCMHDCMSCMWDCMArea MethodQ Observed

Figure 24 Comparison of observed and predicted discharge for RiverSenggi (B).

0

1

2

3

4

5

6

7

8

-0.2 -0.1 0 0.1 0.2 0.3 0.4


Dis

char

ge,

Q(m

3 /s)

VDCMHDCMSCMWDCMArea MethodQ Observed

Figure 25 Comparison of observed and predicted discharge for RiverBatu.

discharge with averaged errors of 13.04–26.48% and 10.99–20.575%; maximum errors of 24.14–70.58% and 21.85–59.19%,respectively depending on the geometrical and boundary condi-tions of the rivers. On the other hand, the HDCM method is foundto under-estimate the discharge with averaged errors of 8.11–8.95%, and maximum errors of 20.78–23.85%. Other methodssuch the SCM method is found serious under-estimate the dis-charge at low overbank flow, but in certain cases, becomes betterat larger depth of flow. The WDCM method is found to be able toproduce better estimations, with averaged errors of 3.68–11.43%,

Table 3 Discharge estimation for flooded rivers using conventional methods.

Name of river Method Max. Error (%) Ave. Error (%) RMSE (%)

River Senggai VDCM 26.766 14.865 16.893HDCM −20.781 8.953 10.959SCM −71.041 36.163 43.914

WDCM 7.821 6.943 8.677Area Method 25.545 10.989 12.875

River Senggi (B) VDCM 70.579 26.482 36.700HDCM −23.449 8.110 10.160SCM −64.824 25.569 33.526

WDCM 30.282 11.427 15.111Area Method 59.194 20.572 29.524

River Batu VDCM 24.139 13.043 16.137HDCM −23.850 8.734 11.488SCM −40.369 19.980 20.059

WDCM 7.832 3.681 4.500Area Method 21.848 11.376 13.924

and maximum errors of 7.82–30.28%. In general, as shown inTable 3, all the methods tested experienced a significant errors,and therefore, these methods should be used with extra cautionin overbank flow estimation.

8 Conclusions

Based on extensive overbank flow data collected from fieldstudy and results obtained, the following conclusions have beenmade.

1. Velocity distribution and stage discharge relationships confirmprevious laboratory findings of a reduction in main channelparameters due to the interaction between main channel andfloodplain, with a consequent reduction in compound sectioncapacity when floodplains are inundated.

2. The flow velocity on the floodplains remained near to zerodue to high resistance of surface vegetation, and therefore, ithas little contribution to overall discharge capacity of floodingrivers.

3. Flow resistance for the rivers has been illustrated using Man-ning coefficient, and Darcy-Weisbach friction factor, showinga much higher resistance in the main channel and floodplaincompared to those modeled in laboratory flumes. The resultsalso further confirmed the significant increased of flow resis-tance under flood conditions, and the consequent danger ofusing inbank data to overbank flows calculations.

4. The discharge in flooded river is either over-or-underestimatedusing the conventional and experimental methods, with a max-imum error >23%, depends on the methods used, and surfaceconditions of the rivers.

5. Further study on flooding river with different scale, sinuosity,and geometrical conditions has to be carried out in order toprovide more data to develop a reliable method for hydraulicanalysis under overbank flow conditions.


Notations

A =Wetted areaB = Top widthC = Chezy roughness coefficientf = Darcy-Weisbach friction factor

FP = Floodplaing = Gravitational accelerationh = Bankfull depthH = Depth of flow

HDCM = Horizontal Divided Channel MethodMC = Main channel

n = Manning’s coefficientP = Cross sectional wetted perimeterR = Hydraulic radius

SL = Longitudinal bed slope for left floodplainS0 = Longitudinal bed slopesSR = Longitudinal bed slope for right floodplain

SCM = Single Channel MethodV = Mean velocity

VDCM =Vertical Divided Channel MethodWDCM =Weighted divided channel method

References

1. Myers, W.R.C., Knight, D.W., Lyness, J.F. and Cassells,J.B. (1999). “Resistance Coefficients for Inbank and Over-bank Flows,” Proceedings of the Institution of Civil Engi-neers, Water, Maritime and Energy, 136, 105–115.

2. Wormleaton, P.R. and Hadjipanos, P. (1985). “Flow Dis-tribution in Compound Channels,” Journal of HydraulicEngineering, ASCE, 111(2), 357–361.

3. Knight, D.W. and Demetriou, J.D. (1983). “Floodplainand Main Channel Flow Interaction,” Journal of HydraulicEngineering, ASCE, 109(8), 1073–1092.

4. Wormleaton, P.R., Allen, J. and Hadjipanos, P. (1982).“Discharge Assessment in Compound Channel Flow,” Jour-nal of Hydraulic Division, ASCE, 108(HY9), 975–994.

5. Prinos, P. and Townsend, R. (1984). “Comparison ofMethods for Predicting Discharge in Compound OpenChannels,” Advances in Water Resources, 7(Dec), 180–187.

6. Myers, W.R.C. (1987). Velocity and Discharge in Com-pound Channels,” Journal of Hydraulic Engineering, ASCE,113, 753–766,

7. Myers, W.R.C. (1991) “Influence of Geometry on Dis-charge Capacity of Open Channels,” Journal of HydraulicEngineering, ASCE, 117, 676–680.

8. Lambert, M.F. and Myers, W.R. (1998). “Estimatingthe Discharge Capacity in Straight Compound Channels,”Proceedings of the Institution of Civil Engineering, Water,Maritime and Energy, 130, 84–94.

9. Myers, W.R.C., Lyness, J.F. and Cassells, J. (2001).“Influence of Boundary on Velocity and Discharge in Com-pound Channels,” Journal of Hydraulic Research, 39(3),311–319.

10. Sellin, R.H.J. (1964). “A Laboratory Investigation in theInteraction between Flow in the Channel of a River and thatof its Floodplain,” La Houille Blanche, 7, 793–801.

11. Zheleznyakov, G.V. (1971). “Interaction of Channeland Floodplain Streams,” Proceedings, Fourteenth Inter-national Conference of the International Association forHydraulic Research, Paris, France.

12. Dracos, T. and Hardegger, P. (1987). “Steady UniformFlow in Prismatic Channels with Flood Plains,” Journal ofHydraulic Research, 25(2), 169–185.

13. Myers, W.R.C. (1978). “Momentum Transfer in a Com-pound Channel,” Journal of Hydraulic Research, ASCE, 16,139–150.

14. Knight, D.D. and Hamed, M.E. (1984). “Boundary Shearin Symmetric Compound Channels,” Journal of HydraulicEngineering, ASCE, 110(10), 1412–1430.

15. Christodolou, G.C. and Myers, W.R.C. (1999). “Appar-ent Friction Factor on the Flood Plain-Main Channel Inter-face of Compound Channel Sections,” Proceedings of 28thCongress of the International Association for HydraulicResearch, Craz, Austria, SE3, A379, http://www.iahr.org/membersonly/grazproceedings99/doz/000/000/367.htm.

16. Nalluri, C. and Judy, N.D. (1985). “Interaction betweenMain Channel and Flood Plain Flow,” Proceedings of 21stIAHR Congress, Melbourne, Australia, pp. 378–382.

17. Pasche, E. and Rouve, G. (1985). “Overbank Flowwith Vegetatively Roughened Flood Plains,” Journal ofHydraulic Engineering, ASCE, 111(9), 1262–1278.

18. Radojkovic, M. and Djordjevic, S. (1985). “Computationof Discharge Distribution in Compound Channels,” Pro-ceedings of the 21st Congress of International Associationfor Hydraulic Research, Melbourne, Australia, 3, 367–371.

19. Tamai, N., Asaeda, T. and Ikeda, H. (1986). GenerationMechanism and Periodicity of Large Surface Eddies in Com-pound Channel Flow, 5th Congress of Asian and PacificDivision, IAHR, Seoul, Korea, 2, 61–74.

20. Krishnappan, B.G. and Lau, Y.L. (1986). “TurbulenceModeling of Flood Plain Flows,” Journal of HydraulicEngineering, ASCE, 112(4), 251–166.

21. Keller, R.J. and Rodi, W. (1988). “Prediction of FlowCharacteristics in Main Channel/Flood Plain Flows,” Jour-nal of Hydraulic Research, 24(4), 425–441.

22. Stephenson, D. and Kolovopoulos, P. (1990). “Effect ofMomentum Transfer in Compound Channels,” Journal ofHydraulic Engineering, V116(12), 1512–1522.

23. Tominaga, A. and Nezu, I. (1991). “Turbullent Structurein Compound Open-Channel Flows,” Journal of HydraulicEngineering, 117(1), 21–41.

24. Tominaga, A. and Nezu, I. (1991). “Turbulent Structurein Compound Open-Channel Flows,” Journal of HydraulicEngineering, ASCE, 117(1), 21–41.

25. Naot, D., Nezu, I. and Nakagawa, H. (1993). “Hydrody-namic Behavior of Compound Rectangular Open Channels,”Journal of Hydraulic Engineering, 119(3), 390–408.

26. Knight, D.W. (2001). Scooping Study on Reducing Uncer-tainty in River Flood Conveyance: Conveyance in 1-D River


Models, A Report for HR Wallingford and the EnvironmentAgency, London, UK

27. DAFRA/Environmental Agency, Flood and Costal DefenceR&D Programme (2003). Reducing Uncertainty in RiverFlood Conveyance: Interim Report 2: Review of Methodsfor Estimating Conveyance, Project W5A-057.

28. Lai, S.H. (2006) “Determination of Composite FrictionFactor and Manning Roughness Coefficient for Discharge

Estimation in Natural Compound Channels,” PhD Disserta-tion, University Malaysia Sarawak, Malaysia.

29. Department of Irrigation and Drainage, Malaysia (2000).Urban Stormwater Management Manual for Malaysia, Vol-ume 5 — Runoff Estimation, Kuala Lumpur, Malaysia,pp. 14–2.

a study of hydraulic characteristics for ﬂow in equatorial...

Documents