Intl. J. River Basin Management Vol. 6, No. 1 (2008), pp. 23–30

© 2008 IAHR, INBO & IAHS

Sediment deposition in a rigid monsoon drainAMINUDDIN AB. GHANI, Professor and Deputy Director, River Engineering and Urban Drainage Research Centre (REDAC),Universiti Sains Malaysia, Engineering Campus, Seri Ampangan, 14300 Nibong Tebal, Penang, Malaysia.E-mail: redac02@eng.usm.my

NOR AZAZI ZAKARIA, Professor and Director, REDAC, Universiti Sains Malaysia, Engineering Campus, Seri Ampangan,14300 Nibong Tebal, Penang, Malaysia

MAHATHIR KASSIM, Postgraduate Student, REDAC, Universiti Sains Malaysia, Engineering Campus, Seri Ampangan,14300 Nibong Tebal, Penang, Malaysia

ABSTRACTField data collections to study the physical sediment characteristics and trends of sediment deposition were carried out at Raja River monsoon drainmade up of concrete channels for the period of 2000 and 2001. Assessments of the existing incipient motion equations developed from experimentalworks were made using the measured field data. The results show that equations by Novak & Nalluri (1975), El-Zaemey (1991) and Ab. Ghani et al.(1999) are able to predict satisfactorily the sediment deposition in rigid channel.

Keywords: Incipient motion; sediment deposits; rigid channel; storm drains.

1 Introduction

Rapid development in major towns in Malaysia results in theconstruction of new drainage system mainly open monsoon orstorm drains to cater the increase in surface runoff. Sedimentdepositions in these storm drains have been found to be a majorcause of flash flood due to the loss in the hydraulic capacity of thedrains. A constant minimum velocity of 0.9 m/s is recommendedby the Department of Irrigation and Drainage (DID), Malaysiato minimize sedimentation problems.

Recent studies (Ab. Ghani et al. 2000; Kassim et al. 2004;Kassim, 2005) in several major cities in Malaysia confirmthe presence of loose deposited beds of non-cohesive sedi-ments in rigid open storm drains with average sediment sizesbetween 0.35 mm and 2.40 mm. Ashley et al. (2004) foundout that these ranges of sediment sizes in Malaysia are simi-lar to European sediments which were collected from combinedsewers.

Kassim (2005) carried out field data collection along RajaRiver drainage system to identify the trend of sediment depo-sitions in open storm drain. This paper will highlight the resultof this new data collection programme and the assessments ofexisting incipient motion equations for rigid boundary channels(Delleur, 2001; Ashley et al. 2004).

Received on August 10, 2006. Accepted on July 21, 2007.

23

2 Existing incipient motion criteria for rigidboundary channel

The self-cleansing approach in terms of either minimum velocity(Vc) or shear stress (τc) is usually used to minimize sedimentproblems in urban drainage systems. The application of thisapproach is meant to avoid any deposition at any time or at leastover a long period of time no deposits will build up (i.e. high flowsflush the deposits). A number of experimental works have beencarried out for the past 30 years especially in Europe (Delleur,2001; Ashley et al. 2005) on sediment transport in urban drainagesystems. Field data measurements of sediment deposits in com-bined sewers were also made to study the physical characteristicsof sewer sediments and the trend of depositions (Bachoc, 1992;Blaszczyk & Ashley, 1996). The present study attempts to studythe nature of sediment deposition in open storm drains.

Experimental works at University of Newcastle upon Tyne,UK (Novak & Nalluri, 1975; Ojo, 1978; Novak & Nalluri, 1984;El-Zaemey, 1991; Nalluri et al. 1994; Nalluri & Ab. Ghani,1996) and recent works at Universiti Sains Malaysia (Salem,1998; Ab. Ghani et al. 1999; Ashley et al. 2005) suggest newmethods to improve the constant velocity and shear stress criteriaby taking into account several factors affecting sediment trans-port and incipient motion namely sediment characteristics (size,

24 Aminuddin Ab. Ghani et al.

sediment concentration) and drain characteristics (size, slope androughness).

Table 1 shows a list of available incipient motion criteria fromexperimental works (Kassim, 2005) used to predict sedimentdeposition along the Raja River channel. The required parametersfor the applications of these equations are the average sedimentsize (d50), the sediment specific gravity (Ss = 2.65), and the flowdepth (yo). The hydraulic radius, R is determined from the flowdepth and cross section of the drain, and the dimensionless sedi-ment size, Dgr(= d50(g(Ss − 1)/ν2)1/3; ν the kinematic viscosity

Table 1 Existing incipient motion criteria from experimental works (Kassim, 2005).

Researcher Equation Cross section Equation No.

Novak & Nalluri (1975) Vc = 0.17(Ss − 1)1/2d0.2450 Rectangular 1

Novak & Nalluri (1975) Vc/(gd50)1/2 = 0.61(Ss − 1)1/2[d50/R]-0.27 Rectangular; Circular 2

Novak & Nalluri (1975) τc = 0.128(Ss − 1)d0.450 Rectangular 3

Ojo (1978) Vc = 0.184(Ss − 1)1/2d0.2450 Rectangular 4

Ojo (1978) Vc = 0.65(gd50)1/2(Ss − 1)1.2[d50/R]-0.28 Rectangular 5

Ojo (1978) τc/ρgd50(Ss − 1) = 0.062(U∗/νd50)−0.54 Rectangular 6

Novak & Nalluri (1984) Vc/[gd50(Ss − 1)]1/2 = 0.5[d50/R]−0.4 Rectangular; Circular 7El-Zaemey (1991) Vc/[gd50(Ss − 1)]1/2 = 0.75[d50/R]−0.34 Deposited Circular Bed 8El-Zaemey (1991) Vc/[gd50(Ss − 1)]1/2 = 0.80[d50/R]0.325[y0/B]0.04 Deposited Circular Bed 9Ab.Ghani et al. (1999) Vc/[gd50(Ss − 1)]1/2 = 1.07[d50/R]−0.23 Rectangular 10Ab.Ghani et al. (1999) τc/[ρgd50(Ss − 1)]1/2 = 0.17D−0.57

gr Rectangular 11

STUDY AREA

Sg.Raja Pumping Station SR1

SR9

SR2

SR3

SR4

SR5

SR6

SR7SR8

SR10

Figure 1 Study area – Raja River urban drainage system for Alor Setar City.

of water) is determined from the average sediment size. Thewater density (ρ) is assumed as 1000 kg/m3 and the gravitationalconstant (g) is taken as 10 m2/s. The width of deposition (B) ismeasured across the section.

3 Study area

The Raja drainage catchment is part of the Alor Setar CityDrainage scheme, Phase I (Figure 1). The scheme comprises

Sediment deposition in a rigid monsoon drain 25

three catchment basins (Raja, Langgar & Putera) with a totalcatchment area of 300 ha and was proposed as part of a flood alle-viation programme for the region. The Raja basin is the largestin the area (233 ha) and is also the most prone to problems offlooding. Flows from the Raja basin are drained by gravity to apump station before being discharged via pumping mains to theRiver Kedah waterway.

The main lengths of the drainage system take the form of trape-zoidal/rectangular, open channel sections. The sections are linedin concrete, vary in width from 4 to 16 m and have been designedto a minimum velocity criterion of 0.9 m/s for the purpose ofsediment self-cleansing.

The study area has two typical monsoons; namely, the north-east monsoon and southwest monsoon. The northeast monsoonusually occurs from November to February. The southwest mon-soon usually reaches the west coast of Peninsular Malaysia fromthe Indian Ocean and prevails over Peninsular Malaysia fromMay to August. In the transition period between the above twomonsoons, from September to November, the western windprevails and causes the heaviest rainfall in the study area in ayear. Thus, the study area tends to have two rainy seasons ina year: one from April to May, and another from September toNovember. The annual rainfall depth in the study area is about2,000 to 3,000 mm, while the temperature is about 27 Celsius onaverage.

4 Field data collection and results

A sediment data collection programme (Kassim, 2005) was car-ried out in 2000 and 2001 whereby measurement of sedimentdeposition were made at ten stations along Alor Derga and AlorSiam forming the upstream branches Raja River and the mainreach of the Raja River (Figure 1). The first measurementswere carried out before the 2000 rainy season while the second

Figure 2 Sediment deposit thickness survey at station SR1 (10th June 2000).

measurements were conducted after 2001 rainy seasons. Thusa comparison can be made on the sediment characteristics anddeposition for conditions before and after rainy seasons.

For each station (Figure 2), the thickness of sediment depositsand flow depths along the channel was measured over a 20-metredistance at an interval of 2 m spacing (Figure 3). The accuracyof the sediment thickness survey is ±5 mm. Two persons at bothbanks of the drain were required to hold the survey staff verticalvia a rope throughout the measurement. A plate was attached tothe survey staff at the bottom to ensure that the staff remainedon the top of the sediment deposition. For each interval, threemeasurements of sediment thickness were made. Samples ofsediment (Figure 4) were also collected by grab during the mea-surement of sediment profiles. Figure 5 shows an example ofthe slope of sediment deposit (Sb) as compared to the as-builtslope (So). Similar trends of sediment deposition were found atall stations. All the measurements were made at low dischargeto make sure no sediment transport occurred during the fieldwork. This would allow the highest depth of deposition to bemeasured.

Tables 2 to 4 give the summary of the sediment deposit andhydraulic characteristics during the field works carried out in2000 and 2001. The ranges of data for the field work are 0.18 <

V(m/s) < 0.25, 0.45 < d50(mm) < 1.80, Ss = 2.50, 0.24 <

yo(m) < 1.23, and 0.36 < ys(m) < 1.25 where V is the flowvelocity, and ys the main channel sediment deposit thickness. Thesediment size distributions for both periods of measurements aregiven in Figure 6 indicating that they are all non-cohesive. Theflow velocity was computed from Manning equation based on themeasured flow depth and main channel sediment slope. Takinginto account the presence of the sediment deposits, a value of0.030 for the coefficient of Manning (n) was assumed for allstations in computing the flow velocity.

It can be seen from Tables 2 and 3 that coarser sediments arefound in the bed after rainy season (Table 3) compared to before

26 Aminuddin Ab. Ghani et al.

x x x x x x x x x x x

x x x x x x x x x x x

x x x x x 2m x x x x x x

Sediment deposit sampling

X survey points

Cross section

SR7

0 1

B

Q

Study Reach L= 20m

2 3 4 5 6 7 8 9 Cross section

10

Figure 3 Measurement points at a station (SR 7).

Figure 4 Sediment deposit sampling at station SR6 (20th June 2000).

Sediment deposition in a rigid monsoon drain 27

(a) Profile of Sediment Deposition

-1.6

-1.4

-1.2

-1

-0.8

-0.6

-0.4

-0.2

00 5 10 15 20 25

Distance (m)

Ele

vati

on

(m

)

sediment as-built (center)

Flow Direction

So = 0.0004

Sb = 0.0008

Channel Width (m)

-1.5

-1

-0.5

02 4 6 8 10 12 14

Ele

vati

on

(m

)

Sediment

(b) Thickness of Sediment Deposition

0

Figure 5 Sediment deposition at station SR1 (10th June 2000).

Table 2 Sediment deposit and hydraulic characteristics for 2000 data.

Station Yo (m) A (m2) P (m) R (m) V (m/s) d50 (mm) Sediment thickness (m)

Left bank Main channel Right bank

SR1 0.9685 11.405 13.793 0.8269 0.2130 0.9000 0.9339 1.1130 0.8587SR2 1.0060 11.641 13.751 0.8465 0.1800 0.4500 0.8724 1.2540 0.8784SR3 1.2256 14.568 14.358 1.0146 0.2070 0.8000 1.1789 1.3190 1.1789SR4 0.5432 6.0810 12.795 0.4752 0.2370 1.4000 0.3625 0.8350 0.4322SR5 0.8070 6.2600 9.4190 0.6646 0.1850 0.5000 0.7070 1.0020 0.7119SR6 0.9346 7.1960 9.5890 0.7504 0.1810 0.4600 0.7940 1.2150 0.7948SR7 0.6208 3.9990 6.2140 0.6433 0.2280 1.2000 0.6197 0.6480 0.5947SR8 0.4020 2.4720 5.2760 0.4684 0.2130 0.9000 0.4002 0.4300 2.4720SR9 0.4436 2.6880 5.3290 0.5044 0.2440 1.6000 0.4160 0.5020 0.4129SR10 0.2626 1.3150 3.7190 0.3534 0.2330 1.3000 0.2139 0.3690 0.2048

rainy season. This indicates that bed erosion has occurred duringhigh flow in the rainy season and carried along the finer sediments.Table 4 also shows that milder sediment slopes are present beforerainy season (0.0002–0.0033) as compared to after rainy season(0.0002–0.0036). All sediment bed slopes differ than the as-built

slopes indicating that attempts should be made to compute pos-sible trend of sediment deposition. Hence the assessment of theexisting incipient motion equations are greatly required sincealmost all the available equations were derived from experimentalworks.

28 Aminuddin Ab. Ghani et al.

Table 3 Sediment deposit and hydraulic characteristics for 2001 data.

Station Yo (m) A (m2) P (m) R (m) V (m/s) d50 (mm) Sediment thickness (m)

Left bank Main channel Right bank

SR1 1.0415 12.253 13.920 0.8801 0.2100 0.8500 0.9343 1.2040 0.9861SR2 0.7971 9.1740 13.332 0.6881 0.1970 0.6500 0.6661 1.0590 0.6663SR3 0.9709 11.339 13.735 0.8256 0.2090 0.8400 0.8675 1.1780 0.8672SR4 0.6544 7.4530 13.043 0.5714 0.1970 0.6500 0.5035 0.9200 0.5397SR5 0.8864 6.8570 9.5390 0.7188 0.2040 0.7500 0.7685 1.1200 0.7706SR6 0.7687 5.9360 9.3230 0.6366 0.2170 0.9800 0.6615 0.9830 0.6617SR7 0.8765 5.6430 6.7130 0.8404 0.2510 1.8000 0.8769 0.9160 0.8365SR8 0.5145 3.1720 5.4990 0.5768 0.2440 1.6000 0.5080 0.5450 0.4906SR9 0.3034 1.8180 5.0530 0.3598 0.2280 1.2000 0.2750 0.3570 0.2781SR10 0.2432 1.1890 3.6700 0.3238 0.2510 1.8000 0.1845 0.3600 0.1851

Table 4 Comparisons between measured sediment slopes and as-built slopes.

Station SR1 SR2 SR3 SR4 SR5 SR6 SR7 SR8 SR9 SR10

Sediment deposit slope, Sb (2000) 0.0008 0.0007 0.0002 0.0002 0.0003 0.0001 0.00025 0.0027 0.0003 0.0033Sediment deposit slope, Sb (2001) 0.0020 0.0003 0.0003 0.0002 0.0005 0.0005 0.0030 0.0030 0.0012 0.0036As-built slope So 0.0004 0.0005 0.0005 0.00035 0.00035 0.00035 0.0010 0.0010 0.0010 0.0005

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

100.00

0.01 1 100

Size (mm)

Perc

enta

ge P

assin

g (

%)

SR1 SR2 SR3 SR4 SR5

SR6 SR7 SR8 SR9 SR10

.

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

100.00

0.01 1 100

Size (mm)

Pe

rce

nta

ge

Pa

ssin

g

(%)

SR1 SR2 SR3 SR4 SR5

SR6 SR7 SR8 SR9 SR10

.

(a) 2000

(b) 2001

Figure 6 Sediment distribution curves for all stations, Raja River.

5 Assessments of incipient motion equations

Tables 5 and 6 give the results of comparison between the calcu-lated sediment deposit slopes and the observed ones for the tenstations during measurement period in 2000 and 2001. For thevelocity critical equations (Equations 1, 2, 4, 5 and 7–10), thecomputed slopes were obtained by the use of Darcy-Weisbachequation. Whereas for the case of critical stress equations (Equa-tions 3, 6 and 11), the computed slopes were obtained from thedefinition of shear stress (τc = ρgRSb). The criterion used toselect the best equation is the discrepancy ratio namely the ratioof calculated slopes using the equations listed in Table 1 over theobserved slopes. If the discrepancy ratio is between 0.5 and 2.0,the equation is deemed to be accepted for use (Yang 1996).

The summary of the assessment of the existing incipientmotion criteria is given in Table 7. The results show that the crit-ical velocity criterion given by Novak and Nalluri (Equation 2)seems to give the best prediction of the sediment deposit slopesfollowed by El-Zaemey (Equation 8) and Ab. Ghani et al (Equa-tions 10 and 11). The highest percentage of discrepancy value inthe accepted range of 0.5 and 2.0 is 55% as given by Equation 2(Figures 7 and 8). The results suggest the different conditionsat field compared to the controlled condition in the laboratory.Revision of the existing equations may be needed to improve thesimulation of deposition in actual storm drains.

6 Conclusions

A detailed study on sediment deposition trend was carried out in2000 and 2001 whereby measurements of sediment deposition

Sediment deposition in a rigid monsoon drain 29

Table 5 Discrepancy ratio for 2000 data.

Station Equation

1 2 3 4 5 6 7 8 9 10 11

SR1 0.066 0.313 0.031 0.003 0.821 0.040 1.239 1.229 0.932 0.558 1.852SR2 0.052 0.255 0.026 0.002 0.679 0.022 1.217 1.108 0.825 0.430 0.904SR3 0.189 1.008 0.097 0.008 2.662 0.116 4.341 4.143 3.170 1.751 5.335SR4 0.680 2.380 0.259 0.029 6.121 0.435 7.275 8.134 6.066 4.594 20.070SR5 0.177 0.757 0.082 0.008 2.001 0.074 3.301 3.133 2.391 1.311 3.414SR6 0.434 1.985 0.210 0.018 5.269 0.181 9.128 8.454 6.489 3.380 8.344SR7 0.337 1.395 0.144 0.014 3.620 0.220 4.802 5.082 4.053 2.596 10.16SR8 0.042 0.146 0.016 0.002 0.378 0.021 0.497 0.528 0.411 0.271 0.970SR9 0.447 1.609 0.172 0.019 4.133 0.312 4.825 5.444 4.330 3.125 14.4SR10 0.059 0.176 0.021 0.003 0.451 0.033 0.509 0.584 0.456 0.346 1.523

Table 6 Discrepancy ratio for 2001 data.

Station Equation

1 2 3 4 5 6 7 8 9 10 11

SR1 0.024 0.116 0.011 0.001 0.305 0.014 0.474 0.464 0.352 0.205 0.657SR2 0.192 0.831 0.088 0.008 2.187 0.093 3.415 3.331 2.476 1.465 3.749SR3 0.170 0.809 0.081 0.007 2.126 0.100 3.262 3.210 2.429 1.435 4.590SR4 0.368 1.445 0.159 0.016 3.787 0.168 5.656 5.642 4.150 2.585 7.751SR5 0.116 0.514 0.053 0.005 1.351 0.062 2.060 2.034 1.579 0.914 2.842SR6 0.155 0.641 0.067 0.007 1.669 0.091 2.318 2.398 1.863 1.174 4.192SR7 0.024 0.113 0.011 0.001 0.293 0.021 0.376 0.405 0.333 0.213 0.972SR8 0.037 0.145 0.015 0.002 0.373 0.027 0.449 0.499 0.400 0.278 1.260SR9 0.153 0.461 0.054 0.006 1.182 0.082 1.364 1.548 1.196 0.898 3.790SR10 0.071 0.201 0.024 0.003 0.511 0.046 0.521 0.629 0.497 0.409 2.114

Table 7 Summary of discrepancy analysis for existing incip-ient motion equations (2000 and 2001 data).

Equation No. No. of data within 0.5–2.0 Percentage (%)discrepancy ratio

1 1 52 11 553 0 04 0 05 7 356 0 07 7 358 9 459 6 30

10 8 4011 8 40

were made at ten stations along Raja River drainage systemin Alor Setar. An assessment of existing incipient motion cri-teria for rigid boundary channel developed from laboratoryworks shows that the equation developed by Novak and Nalluri

Figure 7 Assessment of equation 2 for 2000 data.

(Equation 2) seems to give the best prediction of the sedi-ment deposit slopes followed by El-Zaemey (Equation 8) andAb. Ghani et al. (Equations 10 and 11).

30 Aminuddin Ab. Ghani et al.

Figure 8 Assessment of equation 2 for 2001 data.

Notations

A Flow cross-sectional area (m2)

B Deposition width (m)d50 Average sediment size (mm)Dgr Dimensionless sediment size (= [(Ss − 1)g/ν2]1/3d50)

g Gravitational constantn Manning roughness coefficientP Flow wetted perimeter (m)Q Flow discharge (m3/s)R Flow hydraulic radius (m)So As-built Slope of ChannelSb Measured Slope of depositionSs Sediment specific gravity (= ρs/ρ)

ys Main channel sediment bed thicknessV Average flow velocityVc Critical velocityyo Average flow depthγs Specific weight of sedimentν Kinematic viscosity of waterρ Density of waterρs Density of sedimentτc Critical shear stress

References

1. Ab. Ghani, A., Zakaria, N.Z., Kassim, M. and AhmadNasir, B.(2000). “Sediment Size Characteristics of UrbanDrains In Malaysian Cities,” Journal of Urban Water, 2(4),335–341.

2. Ab. Ghani, A., Salem, A.M., Abdullah, R., Yahaya,A.S. and Zakaria, N.A. (1999) “Incipient Motion of Sedi-ment Particles Over Deposited Loose Beds in a RectangularChannel,” 8th International Conference on Urban StormDrainage, Sydney, 30 August– 3 September.

3. Ashley, R.M., Bertrand-Krajewski, J.L., Hvitved-Jacobsen, T. and Verbanck, M. (2004). “Solids inSewers – Characteristics, Effects, and Control of SewerSolids and Associated Pollutants,” IWA Publishing, Lon-don, 340pp. Scientific and Technical Report No. 14,ISBN 1 900222 91 4.

4. Ashley, R.M., Bertrand-Krajewski, J.L. and Hvitved-Jacobsen, T. (2005). “Sewer Solids – 20 Years of Investi-gation,” Journal of Water Science and Technology, 52(3),73–84.

5. Bachoc, A. (1992). “Location and General Character-istics of Sediment Deposits into Man-Entry CombinedSewers,” Journal of Water Science and Technology, 25(8),47–55.

6. Blaszczyk, P. and Ashley, R.M. (1996). “Applicationof New Criteria to Control Sediment Problems in Com-bined Sewers in Poland,” Journal of Water Science andTechnology, 33(9), 245–252.

7. Delleur, J.W. (2001). New Results and Research Needson Sediment Movement in Urban Drainage,” Journal ofWater Resources Planning and Management, ASCE, 127(3),186–193.

8. El-Zaemey, A.K.S. (1991). “Sediment Transport OverDeposited Beds in Sewers.” Ph.D. Thesis, Dept Civil Eng.,University of Newcastle upon Tyne, England.

9. Kassim, M. (2005). Sediment Deposition in a Rigid Mon-soon Drain: A Case Study of Raja River, Alor Setar,MSc Thesis, School of Civil Engineering, Universiti SainsMalaysia.

10. Kassim, M., Ab. Ghani, A., Abdullah, R. and Zakaria,N.A. (2004). Prediction of Sediment Deposition in RajaRiver Concrete Drainage System: A Case Study, Water andEnvironmental Management Series: Sewer Networks andProcesses Within Urban Water System, International WaterAssociation (IWA), pp. 59–65.

11. Nalluri, C. and Ab. Ghani, A. (1996). “Design OptionsFor Self-Cleansing Storm Sewers,” Journal ofWater Scienceand Technology, 33(9), 215–220.

12. Nalluri, C., Ab. Ghani, A. and El-Zaemey, A.K.S.(1994). “Sediment Transport Over Deposited Beds in Sew-ers,” Journal of Water Science and Technology, 29(1–2),125–133.

13. Novak, P. and Nalluri, C. (1975). “Sediment Trans-port in Smooth Fixed Bed Channels,” Hydraulics Division,101(HY9), 1139–1154.

14. Novak, P. and Nalluri, C. (1984). “Incipient Motion ofSediment Particles Over Fixed Beds,” Hydraulic Research,22(3), 181–197.

15. Ojo, S.I.A. (1978). “Study of Incipient Motion and Sedi-ment Transport Over Fixed Beds,” Ph.D. Thesis, Dept CivilEng., University of Newcastle upon Tyne, England.

16. Salem, A.M. (1998) “Incipient Motion Over LooseDeposited Beds in a Rigid Rectangular Channel,” M.Sc.thesis, School of Civil Engineering, University of Science.

17. Yang, C.T. (1996). Sediment Transport – Theory andPractice, McGraw–Hill, New York.