"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA"
(RAS/88/026) PROCEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCLUDING THOSE OF INTERNATIONAL RIVERS
1991.
한국건설기술연구원
※ 본 자료의 페이지 수는 원 자료와 다를 수 있습니다.
주관 수 행 기 관 : KOREA INSTITUTE OF CONSTRUCTION TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS)
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
VARIATION OF FROW AND R0UGHNESS CHRACITRISTICS IN FLUVIAL RIVIRS
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Professor, Hokkaido Institute of Technology, 참여연구원 : Japan Tutomu Kishi
Summary
The flow and roughness characteristics of fluvial rivers vary in accordance with the properties of the bed configurations. In field rivers, sand waves with different characteristic lengths are often formed
- 1 - simultaneously. This kind of bed configurations can hardly be observed in the flume experiments.
In the present research, the bed configurations are classified into three types by the type of bar and the characteristics of the resistance to flow are analysed in relation to the type of bar.
1. INTRODUCTION
Fluvial rivers acquire various bed configurations depending on the properties of sediment and the conditions of flow. Sand waves which are the constituent of the bed configurations affect the roughness of the bed, sediment transport and the variation of plane form of the river.
Fluvial rivers have three different characteristic length, that is the width, the depth and the grain size of the bed Materials. There are three kinds of sand waves which have the above characteristic lengths as the dominant wave length. Sand waves are classified into the meso-scale waves, the small-scale waves and the micro-scale waves in accordance with the respective characteristic lengths.
The representative examples of the meso-scale sand waves are the alternate bar, multiple row bars and braided bars. The representative examples of the small-scale sand waves are the sand dune, the transition bed, the flat bed and the anti-dune. The representative example of the micro-scale sand waves is the sand ripple.
The meso-scale sand waves give the river flow the meandering character. The small-scale and micro- scale sand waves act as the roughness elements to the flow.
Three kinds of sand waves are possible to coexist. In field rivers, we often observe the sand dunes superposed on a part of the bars. However, the flow and the roughness characteristics of fluvial rivers under the coexisting state of the meso-scale and small-scale waves have not been clarified. The reason is that the researches made so far have been based mainly on the results of flume experiments in which the coexisting state of the sand waves of the above different wave lengths is hardly to observe.
Therefore. the collection of field data and the classification of bed configurations have become important in analysing the variation of flow and roughness characteristics of fluvial rivers. In the present paper, the bed configurations of the fluvial rivers are classified in accordance with the type of bar and the characteristics of flow and roughness are analysed using the field data of Japanese rivers.
2. CLASSIFICATION OF BED CONFIGURATION BY TYPE OF BAR
The bed configurations of fluvial rivers of straight channel can be classified into three types by the type of bar which is formed on the bed.
- 2 -
The regime criteria of types of bar by Kuroki and Kishi (1982, 1984) is shown in Figure 1, in which the region where various types of bar are formed is given as a function of the shape parameter of the channel BS_{0}^{0.2}/D and the non-dimensional bed shear stress of the flow τ*. The notations used in the figure are as follows : B ＝ channel width, D ＝ mean water depth, S_{0} ＝ energy slope, τ* ＝ non-dimensional bed shear stress,τ* ＝ DS_{0}/(ds), d ＝ mean grain size of bed material, s ＝ (ρ_{s}-ρ)/ρ, ρ_{s} ＝ density of sand grain, ρ ＝ density of fluid. The results shown in Fig. 1 are applied favorably to the natural rivers when hydraulic quantities corresponding to the mean of the annual maximum discharges at respective gauging stations are used ［Yamaguchi and Kuroki(1981)］.
As to the bar height, the relation with the flow conditions was not clarified so far. Kishi and Kuroki(1986, 1987) have found, basing on the data of flume experiments, that the normalized height of alternate bar △/D can be expressed as a function of the shape parameter BS_{0}^{0.2}/D and the non-dimensional bed shear stress τ*. Empirical curves for △/D are given in Fig. 1 by fine solid curves, in which △ ＝ height of bar defined as the maximum value of the difference in the elevation of bed surface in a cross section. The tendency that △/D increases gradually from the boundary curve of incipience of the alternate bar toward the boundary curve of incipience of the multiple row bar is clearly recognized.
3. TYPE OF BAR AND THE CHARACTERISTICS OF RESISTANCE TO FLOW
3.1 Shear diagram and the law of hydraulic resistance
When the log-law for the rough turbulent open channel flow is approximated by the Manning- Strickler's expression, the resistance law is given by in which D'＝ effective water depth. Eq.(3) is a modified version of the Einstein-Barbarossa's expression, in which the equivalent sand roughness ks is evaluated as 2d. i.e. k_{s} ＝ 2d.
- 3 -
The term 7.66(D/2d)^{1/6} in the right hand side of Eq. (1) expresses the surface resistance due to sand grains and the term ( τ*'/τ*)^{⅔} expresses the increments of resistance due to sand waves. Thr relationship between τ*' and τ* is named as the shear diagram. In general, τ*' is a function of τ* and D/d.
3.2 Shear diagram and resistance law for small-scale sand waves
Kishi and Kuroki (1973) gave shear diagrams and resistance law for flow over small-scale sand waves. An example of the shear diagram is shown in Fig.2.
Shear diagram gives relations of τ*' v.s τ* for various sand waves. The τ*' -τ* relations for various bed forms are formulated as.
- 4 -
In Fig.2, the intersecting point of two lines for Dune Ⅰ and Dune Ⅱ is the incipient point of transition bed and the shearing stress at that point τ* is given by.
Observations in natural rivers, described in the next paragraph, gave the result that τ*_{t} does not exceed unity, that is τ*_{t} ＝ 1, for D/d ＞ 2.5 × 10³. This will suggest that the occurence of the suspending movement of the bed sediment promotes the transition of wave types. Meanwhile, the shearing stress for the incipient point of anti-dune τ*_{a} is given by.
When shearing stress of the flow τ* is between τ*_{t} and τ*_{a} the bed form is transition bed, in general, as are seen in Fig. 2 and τ*' is a multi-valued function for a given τ*. This suggests that the transition of bed from one form to the other depends on the size and size distribution of the bed material in addition to τ* and D/d.
3.3 Field observations of the resistance characteristics in the natural rivers with bar bed
Since the small-scale sand waves are often superposed on a part of the bar surface, the roughness of the natural rivers have to be affected by the interacting effect of sand waves with different lengths. Kishi and Kuroki(1988, 1990) have reported the results of analysis of the field data obtained in the 43 gauging stations of the 17 Japanese rivers. The data cover a wide range of the bedform chart shown in Fig. 1, i.e 3＜BSo^{0.2}/ D＜ 10², 0.05 ＜ τ* ＜ 4.0.
Examples which show the conditions of the gauging stations are given in Fig. 3. In the figure, ● mark expresses the hydraulic conditions corresponding to the mean of the annual maximum discharges and 0 mark expresses those corresponding to the maximum discharge during the observations.
The results of analysis of Kishi and Kuroki are summarized in Fig. 4. The characteristics of the resistance to flow in fluvial rivers with various bed configurations can be classified into three groups.
The resistance to flow at the stations contained in region Ⅰ in Fig. 4 is dominated by the small-scale sand waves. A typical example is shown in Fig. 5 which shows the shear diagram obtained at the station no.3 in Fig. 3. The shear diagrams of Fig. 5 agrees well with Eq. (5a), (5b) and (6).
The resistance to flow at the station contained in region Ⅲ in Fig. 4 is dominated by bar. A typical
- 5 - example is shown in Fig. 6 which shows the shear diagram obtained at the station no.9 in Fig. 3. The shear diagram in Fig. 6 shows for flow with alternate bar on which no small-scale sand waves are superposed. The reason of the small roughness effect of the alternate bar can be explained by the smaller values of (height/length) ratio comparing with those of small-scale sand waves.
The resistance to flow at the stations contained in region Ⅱ in Fig. 4 show transitional characters between regions Ⅰ and Ⅲ. Two typical examples are shown in Fig. 7 which shows the shear diagram obtained at the stations nos.7 and 8 in Fig.3.
The most remarkable feature observed in Fig. 4 is that the effect of the small-scale sand waves superposed on the bar surface becomes distinctive for τ* larger than around 0.4.
4. FLOW OVER ALTERNATE BARS
As stated above, the resistance to flow over bar bed is affected largely by the small-scale sand waves superposed on the bar surface. However, the meandering character of the flow is still maintained on altenate and multiply row bars. The small-scale sand waves do not change the flow pattern of large scale. Analytical solution for over altenate bars was given by Fukuoka(1989) .
Fig-1
- 6 -
Fig-2 Shear Diagram for D/d=1,000 (flume experiments for sand waves of small scale)
Fig-3
- 7 -
Fig-4
Fig-5
- 8 -
Fig-6
Fig-7
- 9 - "REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
FORECASTING IMPACTS OF HYCRAULIC STRUCTURES ON AN ALLUVIAL RIVIR
1991
주관 KOREA INSTITUTE OF CONSTRUCTION TECHNOLOGY(KICT), KOREAN ASSOCIATION 수 행 기 관 : OF HYDROLOGICAL SCIENCES(KAHS), IRTCES, IWHR, Tsinghua University Director,(IRTCES), (IWHR), Beijing, China 참여연구원 : Binpan Lin
INTRODUCTION
Two categories of methods may be applied by engineers of good experience to forecast the impacts of hydraulic structures on an alluvial river. These are the methods of mathematical and physical modeling. Chinese experience has indicated that physical and mathematical modeling, if properly applied by qualified engineers, could yield information of erosion and deposition well conforming to the observed data in prototypes. Thus a one-dimensional mathematical model has successfully reproduced the 19-year record of deposition in Danjiangkou reservior and physical modeling has well predicted the erosion and sedimentation in Gezhouba Project, both in China.
A one-dimensional mathematical model may be applied to predict the temporal development of longitudinal profiles of both the water surface and the bed of a river reach subsequent to the construction of a hydraulic structure. Since relatively less CPU time is required in one-dimensional computations, this model is particularly useful for the prediction of long-term changes in a long reach caused by the construction of a hydraulic structure. But as it gives only the average bed elevation in a
- 10 - cross section, there is no way to tell whether the cross section has a locally deeper part allowing the passage of a ship. Nor can one tell what changes a training work, e.g., a spur dike, might have brought about on the bathymetry of the reach. To provide answers to the latter questions, physical models or 2D mathematical models may be employed. Due to the existence of two time scales in a physical model of movable bed, there is a limitation to the length of a reach that can be modeled. According to the Chinese state of the art, a reach as long as 200 ㎞ has been studied with a physical model［1].
The length of a reach that may be modeled by a 2D mathematical model is, however, practically limited only by the CPU time that can be afforded. In China. the 2D mathematical model has so far been applied to the preliminary investigation of problems involving only relatively short reaches.
Both physical and mathematical modeling is based on observed data and has to be performed by personnel of good practical as well as theoretical background. Moreover, much schematization is needed in the application of either model, so that modeling is still more or less an art in the sense that it is not like the straight forward computation with a formula and that only people with proper experience can do the schematization properly.
Although all models, be it physical or mathematical, are designed or formulated by observing the fundamental equations of motion, they invariably contain some paramters that have to be calibrated against observed data. Thus either type of model is in a sense partially empirical .
ONE-DIMENSIONAL MATHEMATICAL MODEL
A one-dimensional mathematical model has been developed in China for alluvial rivers with essentially fixed banks. This model is mainly for the case where suspension is the principal mode of sediment transport, although the model as developed by Han [2,3] also contains subroutines for the computations involving bed load transport and the occurrence of density current. The principal part of this model is based on the following equations:
- 11 -
in which U is the mean velocity; h, the depth of the flow; Z, the bed elevation; So, the bed slope; S_{f}, the energy gradient; q, the discharge per unit width; p, the porosity; w, the fall velocity of sediment; a, a coefficient; C, the mean concentration of suspended sediment in a cross section; C*. the value of C when sediment transport is in equilibrium; x, the longitudinal coordinate and t, the time. Unlike the model for equilibrium transport of sediment, the formula for the capacity of sedimetn transport is replaced by eq. (4) which is the result of integrating the equation of sediment diffusion along the vertical. For detailed derivation of this equation, the reader is referred to ［2,3,4,5］ . This step of employing eq. (4) is more logical on two counts. (1) Transport of sediment in nature is seldom in equilibrium because of nonuniformity and unsteadiness of flow. (2) Foumulas for the capacity of transport are usually obtained under the condition of uniform and steady flow in laboratory flumes without taking into account the presence of 'wash load' in nature. Theoretically it is also unsound to evaluate the spatial and temporal derivatives from these formulas. Thus in the model developed by Han. eqs. (1) to (4) have been adopted. They will hold for the transport of suspended sediment under non-eqilibrium conditions, and may be applied to the computation of all cases of bed deformation under conditions of unsteady and non-uniform flow, including, among others, reservoir deposition, degradation downstream of a dam, re-entrainment of sediment from the upper surface of deposit during the drawdown of a reservoir and sedimentation in tidal flow. This model, as developed by Han ［2,3］, may be applied to compute erosion/deposition in all types of tranquil steady flows. In addition, it is also capable of computing the changes in the gradations of sediment both in suspension and on the bed under the conditions of either deposition or scour alone and simultaneous scour of finer particles and deposition of coarser particles from and to the bed. The capability to compute the varying gradations is of great importance and is not to be overlooked, if a realistic solution is to be obtained. For instance, change in gradation is one of the important factors affecting the headward extension of reservoir deposition and the scope and magnitude of downstream degradation. This model has been extensively verified. Particularly noteworthy is the verification against the 19 years' record of deposition of sediment in the Danjiangkou reservoir in Hubei, China (Fig-1) ［6］. It can be seen that the computed deposition in the reservoir at the end of every year conforms well to the observed data ［6］. This model is now being extensively applied to anticipate the quantity of deposition in the reservoir of the Three Gorges Project as well as the degradation and subsequent refilling in the channel downstream of the dam for a period of operation of over 100 years. Computations have shown that, by adopting a scheme of drawing down the reservoir during the flood season and filling up the reservoir during the low-flow season about 86％ of the flood control capacity and 90％ of the storage for power generation, navigation and water supply at the Three Gorges Project may be preserved permanently.
PHYSICAL MODELING
A notable advance in the art of physical modeling in China was achieved in the design of the 2800 MW Gezhouba Project on the main stem of the Yangtze River［7］. Physical models were built to study sedimentation in the backwater region and the general neighborhood up-and downstream of the dam, including sediment deposition at the entrances of the approach channels for the locks, sediment flushing structures for the approach channels, deposition in the fluctuating backwater region with special attention directed to the possible improvement of navigation, and problems caused by the deposition of sediment in the forebay of the powerhouse and its passage through the turbines and others. Model results have checked with the prototype data of the partially completed Gezhouba
- 12 - project collected from 1981 through 1983 ［7, 8］. It is worthwhile to note that recently a physical model was built for the specific purpose of verifying the technique of movable-bed model testing developed in China. The verification was based on the 14-year record of deposition in Danjiangkou Reservoir ［9］. The data of sediment and flow influxes for the years 1977 to 1985 were used for the calibration of the model and similar data for the years 1986 to 1990 were to be forecast by the model［9］. The model data of accumulated deposition come within about 20％ of the actual deposition observed, the discrepancy being smaller in narrow reaches and larger in wider reaches.
A physical model is usually designed according to the following criteria:
In these expressions, L denotes the scale ratio; n, the Manning coefficient; U_{0}, the incipient velocity of sediment; s, the sediment; r_{s} and r, the specific gravities of sediment and water respectively; G, the sediment discharge per unit width; G*, the sediment discharge per unit width under equilibrium conditions; and t', the time scale of bed deformation. Criteria (5) and (6) are derived from eqs. (1) and (2). Both criteria (3) and (4) may be derived from the simplified equation of sediment diffusion.
by assuming as ususal that K is the same as the coefficient of diffusion of momentum. Strictly speaking, both criteria (3) and (4) should be satisfied in order to achieve similarity of sediment
- 13 - diffusion. But it is impossible to satisfy both. In practice, if the sediment load is fine, then the criterion for settling, viz criterion (3), will be the more important. Thus it has been proposed that, when w/U＾{*} ＜ 1/40, the criterion for suspension may be neglected. Criterion (10) may be derived from eq. (4).
To save on laboratory space, a small distortion, say from 2 to 2.5, and a horizontal scale of 250 to 300 are often adopted. Moreover, it will be found that in order to satisfy the similarity requirements, one has to employ light- weight material as model sediment. An important experience of physical modeling in China is that the gradation of model sediment must follow that of sediment in the prototype as closely as practicable provided that flocculation does not occur. While the coarser part has important influence on the total deposition/erosion, the finer part, including what is normally considered as the wash load, may have considerable effects on the deposition on the flood plains and in the regions of slack water, including the deeper part of the reservoir. For studies involving repeated fluctuation of stages by large margins, it is desirable to avoid adopting a model sediment that would coagulate when exposed to the air by receding stages.
Where important structures are involved, such as the study of sedimentation in the neighborhood of a dam with ancillary structures such as ship locks, spillways, and sediment sluicing structures, a normal model without distortion is desirable, although slightly distorted model may also be used.
MATHEMATICAL MODEL IN TWO SPATIAL DIMENSIONS
In engineering applications, a depth-integrated model is often adopted. A "stream tube" model is sometimes also classified as a 2D model, but is actually a collection of one-dimensional models. It cannot very well furnish information on a more complicated flow pattern involving, for instance, circulation or slack water. When suspension is the principal mode of transportation, the basic equations of depth-integrated 2D model are as follows :
(k ＝ 1, 2‥‥‥n).
- 14 - where Q ＝ uh; P＝vh ; F＝2sin m R, m being the latitude and R, the speed of rotation of the earth; K, the coefficient of diffusion and f', the resistace coefficient. Other notations bear the same meanings as explained before.
In the case of 2D models, accurate simulation of the geometry of land boudaries is of great importance, for it determines to a great extent the proper reproduction of the pattern of erosion /deposition in the reach. Whereas the method of finite element can fit the boundary geometry very well, it usually requires more computation time than finite difference methods and is seldom employed in the already involved computations for movable bed in China. An accurate and convenient method for the satisfaction of conditions at solid boundaries of complicated geometry has been devised for use with finite difference computation in rectangular coordinates［10, 11］.
Several versions of depth-integrated 2D models for alluvial rivers have been developed in China, e.g., ［12-14］. All these versions still await further verification. The version by Zhou etal, like others, is a complete modal that can be applied to compute not only the distribution of deposition/erosion in a plan but also the accompanying changes in sediment gradations under various conditions of scour and deposition of the bed. Experience has indicated that unlike one-dimensional computations, preservation of continuity of discharge in mountainous streams is not to be taken for granted. Special care is needed to ensure conservation of discharge. Figures 2 to 5 show the computed patterns of flow and sedimentation on the bed, with and without regulating structures ---- in this case, a spur dike. Small numerals in figures 3 and 5 indicate the depths of deposition. Visual observation indicates that the computed configuration of the bed bears close similarity to the results of model tests. But direct comparison is yet to be made. What has been achieved so far, however, shows that 2D mathematical modeling is promising. It is under further development in China. Further verification will be sought.
RIVERS IN MOUNTAINOUS REGIONS
It is well known that rivers in mountainous regions are often characterized by steep gradients and swift and even torrential currents. It is also not uncommon that mountainous streams are alternately wide and narrow along its course. A wide reach upstream of a narrow one is often deposited during the flood seasons and scoured during the low flow. The opposite is often ture with the narrow reach downstream of a wide reach. Moreover, these rivers are often of rocky or gravel beds. Yet, depending on the geology of the basin, these rivers in flood seasons may carry heavy suspended load consisting of much fine material in addition to the bed load. Although under natural conditions much of the finer materials would behave as wash loads and pass through a reach with little exchange with the bed, some of it will be deposited locally in the slack water of separation zones formed by sudden local expansion of the banks. Moreover, after a reservoir is constructed, a part of the suspended load will behave as bed material load because of reduction in the velocity of flow in the backwater region of the reservoir. This meams that the bed of the reservoir will be gradually covered with finer matherial, accompanied by a change in bed roughness or resistance to flow, which must be correctly ascertained if the backwater curve of the reservoir is to be determined with a reasonalbe accuracy. It is usually relatively easy to determine the final roughness of the bed when the original rock outcroppings are entirely covered with sediment, for then the bed roughness would simply be the grain roughness of sediment covering the bed. The difficult part is how to estimate the roughness during the transition period when the rocky or gravelly bed is progressively covered by fine sediment until the bed becomes entirely alluvial. An analysis of this difficult problem has been performed by Han［15,］, who arrived at an interpolation formula for the transition period as follows:
- 15 -
where x ＝ (ak - a)/ak ; and nk is the roughness of the river when its bed is just covered with finer sediment and becomes alluvial. Let the cross sectional area of this be ak and the surface width Bk. Then if the surface width of the river is greater than 2Bk, nk will be taken as the grain roughness of the bed, otherwise side wall roughness should be taken into account. In this respect, a practice suggested by Han ［15］ is as follows:
(1) When the surface width is smaller than Bk, the roughness of both banks are to be taken into account in the computation of the composite roughness of the cross sections.
(2) When the surface width lies between 1 to 2 Bk, then the roughness of one of the banks should be taken into account.
Correct evaluation of roughness is important in that it will affect the stages of the backwater created by the reservoir which in turn will determine the extent of innundation.
For reservoirs constructed on mountainous streams carrying fine suspended loads, it is often advantageous to reduce reservoir sedimentation by drawing it down during flood seasons for sediment flushing and impounding it during the low-flow season to enhance power generation and navigation, if any. Under such circumstances, the headward extension of reservoir deposition is relatively small, because the equilibrium slope for the transport of fine sediment is small. In all cases. these changes may be anticipated with either mathematical or physical models.
CONCLUDING REMARKS
To determine the feasibility of a hydraulic project on an alluvial river, one must be able to anticipate its impacts on the river correctly even if only approximately. Past experience has indicated that an erroneous forecast could be extremely costly. Modeling, either physical or mathematical, has been so much developed in the last decades that it has now become a powerful tool for the correct forecasting of the impacts of a hydraulic project on an alluvial river.
It is to be stressed that the basis for correct modeling is the field data, including the conventional hydrologic data of water and sediment, particle gradations of suspended and bed load, cross sections of the river, boring of the overburden on the bed and others. Moreover, the principal modeler should be a hydraulic engineer of adequate experience, for in the course of modeling he will have to make much judgement in schematizing the boundary conditions, selecting the model materials in the case of physical models, evaluating the resistance coefficients of various reaches of the river and others. At times, he might even have to carry out comparative studies with data of similar rivers. Experience in China indicates that both mathematical and physical modeling if done by a right team headed by the right person can generate realistic information conforming substantially to the prototype. In other words, as pointed out in the introduction, modeling in a sense is still more or less an art and is partially empirical.
- 16 -
FIG-1 ANNUAL AND ACCUMULATED DEPOSITION IN DANJIANGKOU RESERVOIR FROM 1968. 1 to 1986. 12
FIG-2 COMPUTED FLOW PATTERN AT JIULONGPO ON 25 SEPTEMBER IN the 19th YEAR OF OPERATION, NORMAL POOL 175m Q in the Yangtze 7570 cu m/s, Q in the Jialing 12100 cu m/s, Stage at Tongluoxia 188.17m, Stage at Jiulongpo 170.50 m
FIG-3 COMPUTED PATTERN OF DEPOSITION AT JIULONGPO REACH ON 31 DECEMBER IN the 18th YEAR of OPERATION, Normal Pool 175m
- 17 -
FIG-4 COMPUTED FLOW PATTERN WITH SPUR DIKE AT JIULONGPO REACH ON 28 JULY IN the 12th YEAR OF OPERATION, NORMAL POOL 175M Q in the Yangtze 18870 cu m/s, Q in the Jialing 11390 cu m/s, Stage, at Tongluoxia 173.61 m, Stage, at Jiulongpo 175.98 m
FIG-5 COMPUTED PATTERN OF DEPOSITION WITH SPUR DIKE AT JIULONGPO REACH ON 4 OCTOBER IN THE 18th YEAR OF OPERATION, NORMAL POOL 175 m (dark crosses indicate projected filling)
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
- 18 - THE STATE AND THE PROSPECTS OF THE DIFECT SEDIMENT REMOVAL METHODS FROM RESERVOIRS
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Professor, Department of Civil Engineering, 참여연구원 : College of Engineering, Korea University, Seoul, Korea Yong Nam Yoon
Summary
The state of world-wide problems of reservoir sedimentation is first introduced. Presently available methods of direct sediment removal from reservoir are classified into four, i. e. flushing, dredging, dry- earth moving, and combination of flushing and dredging.
Among these four methods flushing and dredging methods, which are considered as large scale direct removal methods, are extensively reviewed in view of removal operations in general, and the type and size of equipments to be used. Since the most economic choice of methods and equipments depends on the reservoir sedimentation characteristics, the main characteristics of a specific reservoir are quantified according to subdivision of each characteristics.
The general operational procedure is introduced in the case of sediment slucing and sediment flushing along with the successful world experiences in sediment removal by flushing method.
The operational aspects of dredging in reservoirs, which determine the choice of equipments, are critically reviewed. The methods of excavation and transport of sediments, the types and main characteristics of dredgers presently available are also introduced along with the production capability of mechanical and hydraulic dredgers. Finally, a systematic approach is proposed for the best choice of dredging equipment for a particular reservoir with specific characteristics.
1. INTRODUTION
- 19 - Reservoirs are built in rivers to create water storage capacity for hydropower generation, water supply for irrigation or drinking water, industrial water, for flood control and the like. In rivers with considerable transportation of sediments such reservoirs function as sediment-traps and gradually silt up. Due to this process the active storage capacity of the reservoirs is reduced at a much higher speed than anticipated, and the functioning of the reservoirs impaired. In addition, several secondary impacts may extend a significant distance beyond the confines of the reservoir; that is, the aggradation of upstream river channel, changes in water quality and ecology within the reservoir, and the downstream channel changes. These have been described in detail in the Bulletin No.67 of the International Commission On Large Dams (ICOLD) which was published in 1989 as the product of 5 year joint study by the Technical Committee on Sedimentation of Reservoirs of ICOLO (ICOLD,1989).
An important question to ask is how serious is the sedimentation of reservoirs in various parts of the world. At the moment the total volume of all reservoirs around the world is roughly 5,000 ㎦. About 80% of this volume consists of reservoirs with a volume of more than 5 billion ㎥ each. This volume of 5,000 ㎦ is approximately equal to 13％ of the total annual river discharge in the world. For all these reservoirs together it has been estimated that annually 50 billion ㎥ of sediment is deposited, which means an annual reduction of the reservoir capacity by 1％ on the average. These figures are of course very general, but they indicate that the possibility of building new reservoirs, when the existing ones have to be abandoned after being silted up, is probably in many cases not a realistic solution.
For many cases the best solution would be to take further measures to reduce the sedimentation or to directly remove the deposited sediments from the reservoir. However, up till now, many reservoir owners have been of the opinion that direct removal of sediment is not a viable solution. and for that reason have not considered this possibility serious enough.
Against this background, the state and prospects of methodology for the direct sediment removal from reservoirs will be systematically presented in the present paper, which is a collection of the recent study done by the Reservoir Sedimentation Committee of ICOLD, of which the author has been a member for several years.
2. METHODS OF DIRECT SEDIMENT REMOVAL FROM RESERVOIRS
Four possible ways of physically removing the sediment from reservoirs are available at present, i. e. ; flushing, dredging, dry-earth moving, and combination of dredging and flushing.
Flushing means that reservoir-water is used to create a flow with high velocity through an outlet at lower locations in the dam in such a way that as much sediment be carried downstream as possible. To date flushing has been carried out successfully in reservoirs where sufficient water is available, but only on an experimental scale in dry areas with water shortage.
Dredging of deposited sediment in reservoirs has largely been limited to very small reservoirs where storage space is at a premium and to cleaning of intakes and outlets facilities. For large scale removal of sediments dredging has often been considered by reservoir owners to be uneconomical and causing too much loss of valuable water. However, due to the considerable recent developments in excavation techniques and transportation systems, dredging is becoming in many cases an
- 20 - economically attractive tool for sediment removal .
Dry-earth moving by excavation and transportation equipments can be effectively done for the removal of sediment deposits in the upstream section of the reservoir if the water level of the reservoir is kept lower than the maximum operation level for a considerable period of time.
Combination of dredging and flushing techniques may appear to be the most viable solution of sediment removal depending on the main characteristics of certain reservoirs.
If one of these four methods is to be applied for a certain reservoir in a most economical way the choice should depend on the characteristics of the reservoir in question. Therefore, the main characteristics of a specific reservoir which are important in determining the method of the type and size of equipments for sediment removal are first closely reviewed in the present paper, and then the present state of the method of flushing and dredging are to be seperately introduced in detail.
3. MAIN CHARACTERISTICS OF A SPECIFIC RESERVOIR
In view of the selection of the method of removing the sediment, the reservoir can be characterized by five main aspects of the reservoir under considerations, i. e. the availability of reservoir water, data on the sediment to be removed, water depth characteristics, disposal site and other general conditions.
3.1 The Availability of Reservoir Water
For dredging and flushing of sediment deposits considerable quantities of water are required in most cases and they are mostly lost. for reservoirs in dry regions the water availability becomes a very important criterion on the success of dredging and flushing.
3.2 Data on the Sediment to be Removed
With respect to the sediment to be removed three aspects concerning the deposited sediments should be considered; the sediment quantity, sediment properties, and the interval between successive removal activities.
The sediment quantity (or volume) to be removed depends completely on the removal policy to be followed, or on the relative size of the reservoir. The sediment properties may vary from very fine consolidated material in front of the dam to a variety of materials near the delta region, including trees, debris, gravels, rocks and so forth. With respect to the interval between successive removal activities two ways of sediment removal can be distinguished. In the case of maintenance dredging the capacity of the equipment has to be such that dredging can keep up with the rate of the annual sediment inflow in the reservoir, and hence, dredging equipment can be of limited size and can be
- 21 - operated on a continuous basis, whereas in the case of rehabilitation dredging the main objective is to restore the storage capacity by dredging in a short time period, and therefore, large equipment will be required only for a certain predetermined period of time.
3.3 Water Depth Characteristics
The magnitude of reservoir water depth is an important criterion for all methods of sediment removal. And, it is also important to know if the water level in the reservoir is constant throughout the year, or if there are certain periods with a considerably lower level. Both the magnitude and variation of water depth may have a significant influence on the choice of method and equipment for sediment removal.
3.4 Disposal Site Characteristics
The distance and elevation differences between the area of excavation and the disposal site have a considerable influence on the costs of removal. The location of disposal site may be near the upstream part of the reservoir, alongside of the reservoir, or downstream of the reservoir. The capacity of the disposal site available in the vicinity of the reservoir is also of great importance in dredging practices.
3.5 Other general Conditions
Among other general conditions in dredging operations the accessibility for equipments, workability, and navigability of barges are important for the most economical mix of equipments for the removal.
In order to be able to make a choice of equipment to be used it is important to know the key characteristics of the reservoir described above. For simplicity, the reservoir characteristics can be subdivided each into several choices as in Table 1.
(Stigter, et al, 1990)
4. REMOVAL OF SEDIMENTS BY FLUSHING
Flushing is a method of sediment removal from reservoirs in which reservoir-water is flushed to create a flow with high velocity to an outlet at lower portions of the dam. This method can be applied in two ways ; the sediment sluicing and sediment flushing(Thomas, 1982).
- 22 -
(White, et al, 1984)
(Chao, et at, 1985)
(Ackers, et al, 1987)
4.1 Sediment Sluicing through Reservoirs
Sediment sluicing means that sediment is sluiced downstream through the outlets of the dam before it deposits in the reservoir. Sediment sluicing is possible either by density currents or by turbulent transport.
Table 1. Relevant Sedimentation Characteristics of A Reservoir
- 23 -
Under certain favorable conditions the sediment-laden incoming flow does not mix with the reservoir water. At the reservoir entrance it rather dives as an compact jet underneath the clean reservoir water, following the previous river course at the bottom and finally proceeding towards the dam without major changes of its density. Through suitable outlets the sediment-water mixture can then be passed on downstream. The principle of sediment movement is shown in Fig. 1. The favorable conditions for sediment sluicing by density currents are met when the incoming water-sediment mixture is of high suspended sediment concentration, the river bed at the reservior entrance has a steep slope, the reservoir depth is large, and when the ground channel at the reservoir bottom is of straight alignment.
The conditions for the favorable development of density currents are rarely met in practice. In many cases, more effective sediment sluicing can be achieved by the turbulent transport of sediments
- 24 - provided that appropriate outlet facilities are available and used properly. The principle is shown in Fig. 2. The efficiency of sediment sluicing by turbulent transport can be increased by drawing down the water level of the reservoir as the sediment concentration of the lower layers will be higher than without drawdown.
4.2 Sediment Flushing from Reservoirs
Sediment flushing means the flushing of sediments which have been deposited already in the reservoir without loosening by a mechanical way, and it can be done either with or without water level drawdown. The efficiency of flushing is, of course, significantly dependent on the reservoir water level at the time of flushing.
In the case of flushing without water level drawdown the efficiency usually becomes very limited.
(Sheuerlein, 1986)
(Sheuerlein, 1987)
(Ackers, et al, 1987)
It is more or less restricted to the zone close to the outlet. As shown in Fig. 3 a funnel-shaped crator develops around the outlet within a short time period. However, as soon as the slopes of the orator have reached the magnitude of the angle of repose of the sediment, the flushing of sediments comes to an end. Therefore, the extent of sediment removal is usually very limited to the vicinity of the dam with a tolerable spill of reservoir water.
In the case of flushing with water level drawdown the efficiency improves to a great extent. The effectiveness depends upon the duration of flushing activity and the degree to which the reservoir is drawn down and, of course, the discharge capacity of the low level outlets. A channel is eroded through continuous back-cutting in the sediment depositions as in Fig. 4. The back-cutting usually develops along the longitudinal profile of the reservoir, and its lateral component is comparatively weak. According to Ackers et al (1987) effective flushing can only be observed where the drawdown level is below about half-height of the dam and where the discharging capacity at that level exceeds the mean annual flow by at least a factor of 2. In this case sediment flushing does not only act upon the deposits, but also affects the sediment which is still in suspension. For this reason it is often recommended to carry out flushing during the early part of the flood season in order to pass heavily sediment-laden water and to scour some of the deposits at the same time.
4.3 International Flushing Experiences
Some international experiences are available on the flushing method, major ones presented in Table 2(Farhoodi, 1988).
- 25 - Table 2. World Experiences in Sediment Removal by Flushing
It is noticable that such operations have been carried out under quite different conditions so that generally no similarity can be observed between any two case studies. Among the examples in Table 2 the Sefid Rud reservoir in Iran is known as the most typical and successful case in the removal of sediments from reservoirs by flushing method.
(MOE. Iran. 1984)
The live storage capacity of Sefid Rud reservoir is 1,800 million cubic meters(MCM), and about 800 MCM of sediments had been deposited in the reservoir during the operation period thereof, which started in 1962. This means about 40％ of the live storage had been lost. The average deposits in the reservoir, before the commencement of flushing operation in 1980, amounted to approximately 50 MCM per year. Table 3 shows the result of sediment removal by flushing during 1980 - 1984 through 5 bottom outlets, 2 mid-level outlets and 2 morning glory spillways, whose total discharges capacity amounts to 6,000 ㎥/sec. As can be seen in Table 3, the total weight of sediments removed reached to 391 million tons during 5 years, which is 21.7％ of the original reservoir capacity.
Flushing operation proved to be a very seccessful measure for the restoration of the live storage capacity of Sefid Rud reservoir. However, some limitations and side effects are involved in sediment flushing operations, which are mainly related to the reservoir, dam and appurtenance structures, downstream reaches, and the associated side effects.
Table 3. Sediment Removal by Flushing from Sefid Rud Reservoir, Iran
Concerning the reservoir, the risk of some slides in sediment masses or even in reservoir side walls are quite expectable due to the rapid draw-down of the water level. Another risk of the rapid draw- down of the water level would be the collapse of huge sediment masses in front of the bottom outlets which might clog the conduits.
Concerning the dam and appurtenance structures, the problems related to dam safety due to the
- 26 - rapid water level draw-down should first be considered. The damages caused to the bottom outlets and gate structures through corrosion, abrasion and cavitation due to high-velocity flow in the conduits become very serious so that intermittent repair works are compulsory.
The flushing operations may also cause some problems to the downstream reaches. The sediments flushed away from the reservoir may deposit in parts of the river course, causing the river bed aggradation and degradation. They may cause some damages to the irrigation network and structures through siltation. The sediments may also cause some problems by being deposited in the estuary when the reservoir is located near the river mouth.
The flushing operations may cause side effects, leaving unfavorable effects on the downstream river eco-system because the suspended sediment of high concentration may affect the water quality and the river biotope.
5. REMOVAL OF SEDIMENTS BY DREDGING
Dredging is the second major method of removing sediments from reservoirs. Basically, it is consisted of three main processes, i. e. : excavation, transportation, and disposal of the sediments. Excavation is the process of loosening the deposited sediment and putting it into or on the transportation equipment, which can take place mechanically or hydraulically. Transportation is the process of moving the excavated material to the area of deposit and can also take place mechanically or hydraulically. Disposal of the dredged material can be made either at the disposal site or to the river downstream of the dam.
5.1 Operational Aspects of Dredging in Reservoirs
The conditions for dredging operations in a reservoir can vary from case to case. Choice of dredging equipments has to be made taking into account all local circumstances as for the following operational aspects.
(Stigter and Korver, 1989)
Dredging location: There is quite a difference in conditions for dredging from the reservoir area in front of the dam, halfway the dam and the reservoir entrance, or near the reservoir entrance. When dredging from the area in front of the dam, consolidated sediment has to be dredged at a large water depth and the loss of water may cause problems, but the navigability is generally good. In the area halfway the dam and reservoir entrance, fine sediment is in general found at a variable water depth, and the loss of water is large or nill depending on the direction of disposal. The accessibility from land or the navigability can vary in the middle of a reservoir. In the area near the reservoir entrance the dredging is characterized by coarse sediment and debris with restricted depth, and the loss of water is eliminated because the surplus water flows back into the reservoir. The navigability is usually poor due to the small water depth, but accessibility from land is good in general.
Transport of Dredged Sediment: The water content of the dredged material is important in view of the
- 27 - most suitable way of transport, coarse or cohesive material dredged by mechanical dredgers can be dewatered and transported by conveyor belts. Hydraulic dredgers produce a soil/water mixture, which has to be transported by pipeline. Most dredgers can also be designed with a barge-loading system for transport of dredged sediment by a barge.
Sediment Disposal: Disposal or storage of the dredged material is one of the major problems encountered in reservoir dredging operations. In general, sediment disposal can be done by one of the four options, i. e. disposal over the dam, disposal through the dam, disposal upstream of the reservoir, or disposal alongside the reservoir.
Production Output Requirement : The dimensions and the rate of sediment deposition of the reservoir can be different for various projects. Dredging equipment has to be selected on the basis of the required capacity for sediment removal by dredging.
Dismountability of Dredger: Reservoirs, in general, and located in remote areas without the possibility of mobilization of the dredgers into the reservoir in floating condition. Therefore, two options are usually available. Dismountable dredgers are those which can be dismounted into parts of dimensions and weights suitable for transport over land, and they can be operated outside the reservoir later on. Whereas, local construction of the dredger by assembling the sections of a dredger on a building site next to the reservoir usually results in a dredger permanently stationed in the reservoir.
Cost of Dredging Operations: For the various types of dredgers the costs of dredging operations are different. In this respect, the total cost should be evaluated in terms of the investment cost for a dredger, operation cost, and the mobilization cost to the reservoir site.
5.2 Methods Excavation and Transport of Sediments out of Reservoirs
For the sediment removal from reservoirs the materials to be dredged should first be dislodged by excavation process which can be done by mechanical means, hydraulic means, or mechanical/hydraulic means. The mechanical means dislodge the consolidated sediments from the bed mechanically by some type of tool attached at the end of a dredger. Whereas, the hydraulic means apply jetwater to dislodge the compacted material to be dredged. The mechanical/hydraulic means combine both of the mechanical and hydraulic means.
Once the compacted materials are dislodged by one of the three means described above they can be transported to the reservoir surface either by mechanical device of the dredger or by suction pipe of the hydraulic dredger. Then, the dredged materials can further be disposed to the disposal site by barges, conveyor belts or aerial cableway.
5.3 Type and Main Characteristics of Dredgers
Dredgers can be classified as either mechanical dredger or hydraulic dredger based on the method of dislodging and transportation of deposited sediment.
There are three types of mechanical dredgers presently available in practice, i. e. bucket dredger, grab dredger and back-hoe dredger as shown in Fig.5, 6, and 7. The main characteristics of each
- 28 - dredger are summarized in Table 4.
(Stigter, et. al, 1990)
In this case of hydraulic dredger, there are presently five different types of dredgers available, i. e. : plain suction dredger, cutter suction dredger, wheel suction dredger, trailing suction hopper dredger, and dustpan dredger, which are shown in Fig.8-12.
The main characteristics of each of these dredgers are summarized in Table 5.
(Stigter, et al, 1990)
5.4 Transportation Devices
The transportation of dredged materials out of the reservoir can be made by hydraulic means in case of hydraulic dredger, and by mechanical means in case of mechanical dredger.
Dredged materials can be transported both in vertical and horizontal directions as a mixture with water by means of a centrifugal pump through a pipeline of the hydraulic dredger. Horizontal hydraulic transport over long distances can be effected by means of a floating or land-based pipeline connecting the dredger and the disposal area.
Table 4. Main Characteristics of Mechanical Dredgers
Table 5. Main Characteristics of Hydraulic dredgers
- 29 -
When dredging is done by mechanical dredger, barges, conveyor belts or aerial cableways can be used as the transportation devices of the dredged materials. Barges are the most common means of transporting dredged materials to the reclamation area and unloading can be done by a reclamation dredger or grabs. In case the dredged material does not contain much water or can be dewatered easily, the conveyor belt offers an alternative for transport over longer distances. In mountainous remote areas where elevation difference is important aerial cableway system could be developed.
5.5 Production by Dredging Equipments
The various types of dredging equipment for excavating the sediments and transporting the dredged material have been discussed so far. The production by each equipment varies widely according to its type and size and hence, the best-fit choice of the type and size of equipments should be made with due considerations to the main characteristics of the reservoir in question in order to guarantee the most economical way of dredging.
To assist the best choice of dredging equipment against the characteristics of the reservoir the various types of equipment which may be used for reservoir dredging can be subdivided each into three choices: small, average, and large. Table 6 and 7 show the production by the three sizes of mechanical and hydraulic dredgers, respectively, under some reservoir characteristics.
(Stigter, et al , 1990)
The hourly productions of dredgers presented in Table 6 and 7 are to be considered as very general. In reality, the sediment characteristics have a significant influence on the production.
5.6 A Systematic Approach for the Choice of Dredging Equipment for Each Specific Reservoir
The choice of the type and size of dredging equipment depends largely on the characteristics of the
- 30 - reservoir in question. However, the choice is usually quite complicated because of the fact that certain characteristics of the reservoir are very advantageous for a particular type of equipment, whereas other characteristics may be very disadvantageous for the same type of equipment.
Therefore, a systematic approach could be devised to develop a kind of scoring system as in Table 8, by which it becomes possible for each type and size of equipment to rate its performance for each characteristics of the reservoir. Eight relevant characteristics of the reservoir can be taken for the rating as shown in Table 8.
Table 6. Main Characteristics of Mechanical Dredgers
Table 7. Main Characteristics of Hydraulic Dredgers
- 31 - Table 8. Reservoir Sediment Removal Matrix
In the matrix presented in Table 8. the main types of equipment with different capacities and sizes are placed on the rows and the various relevant reservoir characteristics in the columns.
If the characteristics are known for a particular reservoir, it will be possible for each type and size of equipment to give ratings for each characteristics of the reservoir. By comparing the ratings of the various pieces of equipment a picture will be obtained, showing in main lines the best possible and economic spread of equipment to be applied for the reservoir under considerations.
This method will give a clear and realistic insight into the possibilities of removing sediment by the best type and size of dredging equipment. Thereafter, of course, further detailed plans have to be made by dredging experts.
- 32 -
6. CONCLUSION
In the present paper, the state of presently available methods of direct sediment removal from reservoir has been critically reviewed. Among the removal methods sediment flushing and dredging appear to be the most common practices for the restoration of active storage capacity of a reservoir. The most economic choice of methods and equipments depends on the reservoir sedimentation characteristics which, of course, are controlled by the major characteristics of a specific reservoir. The general operational procedure for flushing or dredging should be followed to cope with the sedimentation characteristics of the reservoir under considerations. A systematic approach proposed in the present paper can promise the best choice of dredging equipment for a particular reservoir with specific characteristics.
Although, up till now, the direct sediment removal methods have not been considered an practical measures of restoring reservoir capacity, the world experiences are proving the effectiveness and economic merit of the direct sediment removal methods by virtue of the fast improvements in the operational methodology for flushing and dredging, and the rapid advancements in manufacturing the large-scale dredging equipments.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
J. C. STEVENS AND THE SILT PROBLEM A REVIEW
1991
- 33 -
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Professor, Department of Civil Engineering, 참여연구원 : Colorado State University, Fort Collins. Colorado 80523, U.S.A. Carl F. Nordin Jr
INTRDOUCTION
Almost 60 years ago, J. C. Stevens identified some of the problems associated with the construction of dams and reservoirs on sediment-laden streams. and he issued a challenge to scientists and engineers to undertake the necessary studies and research to resolve these issues. Since Stevens' time, some 8,000 large dams and tens of thousands of small and medium size dams have been constructed throughout the world, many on alluvial streams in arid and semi-arid regions where the sediment problems are most severe. A majority of these projects have operated successfully, as designed, but quite a few have severe sediment problems. It is appropriate, therefore, to review the issues that Stevens raised so many years ago, and to try to determine how much progress has been made to resolve these issues.
BACKGROUND
In 1934, at the age of 58, John C. Stevens published his most important paper entitled "The Silt Problem" in the Proceedings of the American Society of Civil Engineers (ASCE). Two years later, the paper along with 35 pages of discussion by nine contributors and the author's closure were published in the Transactions of the Society (Stevens, 1936). Stevens was a consultant and a practical engineer, and he spent much of his career on practical problems associated with the consturction of dams and reservoirs and with other water resources developments. He had a clear understanding of the importance of good basic hydrologic data, perhaps stemming from his early training and experience with the U.S. Geological Survey, and throughout his life he had an active role in developing equipment and techniques for measuring hydraulic and hydrologic properties of rivers. He is probably best known as the inventor of one of the first continuous river stage recorders, for which he received a patent in 1911, the precursor to the models A-35 and A-71 Stevens stage recorders supplied by the company Leupold and Stevens, Inc. , which are used extensively throughout the world. When ASCE founded its Hydraulics Division in 1938, Stevens was appointed chairman of its Committee on Hydraulic Research, the Division's most active committee, and in 1945 he served as president of the Society.
In the synopsis to his paper, Stevens starts off with the statement "All the basic data that the writer could secure on the silting of reservoirs, where the actual capacity surveys have been made to determine the extent of silting, are contained in this paper". The paper itself is dry reading, containing a mass of statistics on the rates at which sediments were accumulating in reservoirs and summaries
- 34 - of the average sediment discharges of rivers of the world (but mostly rivers in the United States, for which Stevens had ready access to data) and only a few sections of the paper dealt with the processes involved and with the control of silt. The discussions of the paper provided substantial additional data for reservoirs in Europe and South Africa, prompting Stevens to remark in his closure that "In one paper will now be gathered practically all the basic data regarding silting of the important reservoirs of the world, as well as summarized data as to the quantity of sediment carried by most of its rivers".
Good basic data are always valuable, and Stevens' paper is cited even today (see for example, the recent lecture notes on reservoir sedimentation published by the International Research and Training Centre on Erosion and Sedimentation, IRTCES, 1985). In both the paper and the discussions, a good bit of attention was directed towards two projects in the semi-arid Southwest, the Elephant Butte Dam on the Rio Grande in New Mexico which was completed in 1916, and Boulder Dam on the Colorado River that was under construction when Stevens wrote his paper. The two rivers were notorious for their heavy sediments loads and the Rio Grande was exceptional for those times in that it possessed almost 40 years of continous records of both streamflow and sediment discharge. In the various sections of his report, Stevens deals briefly with the origin of silt, its transportation, the processes of sedimentation in reservoirs, the density of deposited sediments, and control of silt. In his section on the control of silt, he lists some examples that are of particular interest, the Aswan Dam on the River Nile, completed in 1902, Zuni Dam in New Mexico, completed in 1908, and Hamiz reservoir in Algeria completed before the turn of the century. Of the Aswan, he remarks:
No spillway was provided, but the dam has 180 sluice gates to pass a maximum flood of 500,000 cu ft per sec‥‥ These sluice gates are opened at the beginning of the flood period and the river flows through the reservoir practically as if no dam existed. During such flood periods, although the river is heavily charged with silt, no depositions occur in the reservoir. After the peak of the flood has passed, and the river begins to run clear, the sluice gates are gradually closed and the reservoir is filled for use during the subsequent Irrigation season. By this method of operation silting of the Aswan reservoir has been so far, and probably will be, entirely avoided.
For the Zuni Reservoir, he points out that in a period of about 25 years, the reservoir lost 80 percent of its capacity to deposited sediments: after watershed treatment to reduce the inflow of sediment and installation of a sluice to flush excess sediments from the reservoir, no further loss of capacity took place. For the Hamiz reservoir in Algeria, he noted that "periodic sluicing began in 1901, 23％ of the total water supply is used in sluicing out silt. No part of the silt deposited prior to 1901 has been removed, but further loss of capacity has been prevented.".
But the real heart of Stevens' contribution was contained in his definition of the silt problem and his challenge for engineers to deal with it. He proceeded as follows:
"Man cannot hope to halt the processes of mountain erosion and plains building. The land he cultivates could not exist except for these forces. He must expect that rains will gully his fields, or cover them with mountain debris, and that streams will continue to carry sediments that will fill the canals and reservoirs.
An empire exists below the reservoir that has been created by the Elephant Butte dam, on the Rio Grande in New Mexico. The land is phenomenally productive. The region embraces a substantial unit of civilization the very existence of which hangs upon the integrity of a storage reservoir to impound and deliver water so vitally necessary to life. The reservoir‥‥ is slowly being deprived of its ability to store water. Silt is being deposited at the rate of 20,000 acre-feet per yr. Its original capacity will be so depleted in two or three generations that the civilizations now dependent on it will have to seek other sources of water supply and storage."
Of the Boulder Dam Project, he remarks:
- 35 -
". . .Unless remedial measure are adopted, this reservoir will become virtually useless by the fifth generation. Fortunately, sites are available where other reservoirs may be constructed; and, after these are gone, other will doubtless be found‥‥but what of the ultimate future, when all available storage sites have been exhausted. Must these fertile areas revert, ultimately, to the sage-brush and the cactus? Will sedimentation, that made possible this vibrant civilization, ultimately sound its death knell?. and finally, the challenge:
"It is not the writers intention to paint a dark picture, but rather to stimulate a more intensive study and an intelligent research that will ultimately effect a practical solution of this problem. The menace exists; it is real; and unless something constructive can be evolved, civilization in these regions must eventually decline. It is unfair to sit complacent and pass this problem flippantly on to future generations. The engineer should be equal to the task of finding a solution, but it will take many years of experimentation and study, and he should be at the task, amply financed. ".
An interesting array of attitudes towards the silt problem were evidenced in the discussions to Stevens' paper. W. W. Waggoner, a mining engineer, proposed that "The answer to the question of preventing silting of a reservoir is to build another above it, to impound the silt". Phillip R. R. Bisschop, an irrigation engineer from South Africa, remarked that "There appears to be only one solution, namely to recognize frankly that some irrigation projects are doomed to a definite life span and will have to revert back to the original type and manner of flood irrigation‥‥" Frank E. Bonner. a consultant, observed that "the capital outlays will be amortized long before sedimentation of the reservoirs seriously impairs their usefulness. When the time comes, any desirable restoration of capacity will be a problem for the engineer of that far distant day‥‥" Morrough P. 0'Brien noted that" ... the time may come when perhaps bypass channels and other silt-controlling works, costing perhaps more than the dam itself, will be found to be economically justified. " Herman Stabler, with a poetic flair, declared that "In attempting to abolish the silt problem man can be but a Don Quioxote tilting at the windmill of Nature. " and he posed the question "should the energies of man be directed to prevention or cure?".
In his closure, Stevens does not attempt to pose solutions, but called again for a concerted effort to undertake the studies and research to solve the problem. He suggested that our effort should be directed towards prevention of the problem, and he rejected strongly the position advanced by Bonner that if the sediment problems do not affect the economy of the proposed project, they can safely be left for future generations to solve.
FROM PRACTICE TO THEORY
Developments to 1950
By 1934, a large amount of field and laboratory data on the erosion, transport, and deposition of sediments already existed, and at least in some cases, a few practical method had developed to deal with sediment problems. However, standard methods for collecting field data did not exist, there were
- 36 - no reliable methods to either measure or compute bedload, and the theoretical basis for estimating suspended sediment discharge was just coming to light.
(0'Brien, 1933)
Stevens' call for"... a more intensive study and an intelligent research that will ultimately effect a practical solution to this problem" either provoked or anticipated a flurry of activity that lead to fundamental advances in both practical and theoretical aspects of the silt problem. Within a few years, bed load samplers and bed load equations were developed in Europe; while in the United States, the important theoretical and experimental work on the suspended load equation was completed.
(Rouse, 1937;
(Vanoni, 1946)
The Federal Interagency Sedimentation Committee (FIASC) was formed to coordinate the activities of the many government agencies involved in sediment studies and to standardize equipment and techniques for sediment measurements, and by about 1946, the U.S. Gelolgical Survey expanded substantially its basic data collection program in anticipation of further developments on the Rio Grande and Missouri and Colorado River basins. In 1950, Einstein published his classic paper "The bed-load function for sediment transportation in open channel flow", which brought together most of what was known about river hydraulics, bed load, and suspended load and provided engineers with the first complete theory to predict the movement of bed material in alluvial channels. It is a tribute to Einstein's genius that today, some 40 years later, his theory is still a useful tool and widely used for predicting sediment transport.
By about 1950, a large amount of theoretical work on erosion, transport, and deposition of sediments had been accomplished. The theoretical basis for dealing with the silt problem was fairly well established at that time. However, the necessary empirical data, especialIy the long-term streamflow and sediment records required to determine both the mean values of flows and sediment discharges and their statistical distributions, including extreme values, still remained to be collected. Most of the continuous sediment records in the United States date from 1946-1950. Very long stage records exist for some rivers in China, and streamflow measurements and infrequent sediment sampling began on the Yellow River about 1919 and on the Yangtze River about 1922.
(Stroebe, 1925)
Around 1950, the hydrologic program expanded appreciably, and 40 or more years of continuous flow and sediment records exist for many river in China. However, in many areas of the world, parts of South America and Africa, for example, both streamflow records and sediment records are either short or non-existent, and many projects are being designed with sparse data. Some of thom, Santa Domingo in Venezuela and Amaluza in Equador, have severe sediment problems.
Reservoir Trap Efficiency
Trap efficiency, defined as the percent of incoming sediment retained in the reservoir, is a function mostly of the ratio of reservoir capacity to inflow, Figure 1. The empirical relations first were defined by Brune (1953) based on data from 41 reservoirs in the United States. Data from reservoirs in China,
- 37 - shown on the figure, serve to verify the relation. The capacity-inflow ratio is a convenient method to classify reservoirs according to size. Reservoirs with ratios of about one or greater are able to provide carry-over storage from one year to the next and may be classified as a large reservoir. Brune's trap efficiency curves also are useful for design purposes. The U.S. Bureau of Reclamation used these curves for many years as a basis for estimating reservoir sedimentation rates and although most estimates today are based on mathematical models, the empirical relations developed by Brune are still useful for rough approximations.
Sluicing Efficiency
Although Brune's curves are useful for planning and for estimating the life of a reservoir, they are based on long-term average values rather than seasonal or storm-induced runoff events when most of the sediment is likely to by transported, so very often they are not suitable for design purposes. Instead, sluicing efficiencies (one minus the trap efficiency) can be used that allow estimating the efficiencies by size class. Figure 2, for example, based on empirical data from Chinese reservoirs, gives the sluicing efficiencies as a function of the inflow, outflow, reservoir volume, particle size, or fall velocity, and concentration of the incoming flows. This relation was used in the design of Xiaolangdi Project on the Yellow River, where most of the severe sediment loads are carried by infrequent flood events with hyperconcentrated flows. Similar relations were developed from the mathematical model results for the Yangtze River Three Gorges Project, but normal sediment transport functions do not apply to the hyperconcentrated flows on the Yellow River, so the empirical relations shown in Figure 2 are considered more reliable.
Preserving the Long-term Storage of Reservoirs
Stevens pointed out the successful flushing of sediments at Aswan, and Harry G. Nickle, in his discussion, suggested that "in small reservoir (those in which the capacity is relatively small compared to the inflow during floods and in which stream conditions prevail during floods rather than reservoir conditions) many different effects may occur. The exact behavior of floods through these reservoirs, the deposition and picking up of materials in them ‥‥ are subjects that could be investigated more thoroughly‥‥".
In fact, though, not much attention was paid to this approach to sediment control until the severe sediment problems at Sanmenxia on the Yellow River resulted in the need for a major structural modifications to the dam and a complete revision of operating rules for the reservoir.
(Long and Zhang, 1987)
The Dam at Sanmenxia was completed in 1960 and began operation with a maximum pool level up to 332.6m. Within a few years, it was apparent that sediment deposits in the backwater area upstream of the reservoir and in the reservoir itself were more severe than anticipated, and that the primary function of the reservoir, to protect the Lower Yellow River from flooding, could not be realized. It was concluded that to preserve storage, it would be necessary to sluice sediments through the reservoir each year, and a reconstruction program was undertaken to provide large capacity low level outlets through the dam for that purpose. The operating rules of the reservoir also were changed. The current procedure is to lower the water level at the dam during the flood season, passing flow and sediment through the low level outlets. After the flood peak has passed and the sediment concentration is
- 38 - reduced, the reservoir is filled to provide regulation for power and water use. By operating the reservoir in this manner, a certain amount of storage can be preserved indefinitely.
Mostly as a result of experience with Sanmenxia and other reservoirs in the Yellow River Basin and the extensive work done on the feasibility and design studies for the Three Gorges Project and for Xiaolangdi , both the practical and theoretical aspects of preserving long-term storage in reservoirs is now well known, at least in China(see, for example, the reports by Xia and others, 1980, Zhang and Qian, IRTCES, 1985). In order to preserve the long term capacity of reservoirs, certain conditions have to be met. as follows:
1. The reservoir volume should be small relative to the mean annual flow. This means that a fairly large volume of the flow has to be spilled each year, and this excess flow can be used to flush sediment through the reservoir.
2. The river should have an excess capacity to transport its sediment through the reach the reservoir will occupy - that is, it should be able to transport its sediment load at a flatter lope then exists at present. Usually, this means that the reservoir is in a gorge section or in a reach of the river controlled by bedrock.
3. The reservoir width should be roughly comparable to the width of an equilibrium alluvial channel carrying the same sediment load.
The concepts of the reservoir operation are fairly simple, as sketched in Figure 3. During the flood season, the water level at the dam is kept at a low elevation, the flood control level(FCL). During the initial years of operation, the normal flood flows and some of the sediment load are passed through the reservoir. After the flood recedes, the decreasing flows with lower sediment concentrations are stored as the water level is raised to the normal pool level (NPL). The natural river has an excess capacity and slope Jo to carry its sediment load, as sketched in Figure 4. Eventually, a new equilibrium alluvial channel of width B and depth h will develop through the deposited sediments. In gorge sections, the channel width may be restricted by the valley walls, as sketched in Figure 4. Alluvial channels are self-adjusting, so the new equilibrium channel will develop a slope J smaller than the original channel slope that will allow the prevailing flows through the flood season to just transport the annual sediment load through the reservoir. After equitibrium is reached, there will be no net depositon or erosion through the reservoir. The volume occupied by the sediment deposits is lost, but any storage remaining between the deposited sediments and the NPL can be preserved indefinitely.
The water surface upstream of the dam and the upstream extent of the deposited sediments are determined by the water level at the dam during the flood season, so the most important design considerations in managing the sediment are selecting the flood control level and ensuring that the low level outlets have sufficient capacity to flush the annual sediment load with the excess flows that have to be spilled during the flood season.
The calculation of the equilibuium slope is straightforward, but it requires a good bit of information on the cross sections along the reservoir and estimated roughness values of the composite cross sections, which may be either in gorge sections or in wider sections containing a flood plain. as shown in Figure 4. If the time rate of deposition is required, this is best calculated with a mathematical model. With either form of calculation, the end result is the amount of deposition and its distribution along the reservoir and the amount of storage that can be preserved indefinitely.
- 39 - A FEW EXAMPLES
Sediment Problems
Almost 60 years have passed since Stevens issued his challenge, and in the light developments over that time, it is perhaps useful to evaluate again the severity the problem and the state of developments. Probably, if we were to pose the same question, how to solve the sediment problem, to dam designers today, we would receive about the same sorts of ideas expressed by the discussers of Stevens' paper. And in evaluating what has been done over the years, we would probably have to conclude that Mr. Bonner's point of view won the day. Some 8,000 large dams exist today, and in a great majority of cases the only real concern for sediment has been that the sediment problems not jeopardige the economic life of the project. Most of these projects are designed solely for hydropower, a few are multipurpose projects, but in almost all cases, the revenues from hydropower are the principal component that ultimately pay off the projects. Usually the economic lives of these projects are relatively short; that is, on the order of 20 to 30 years. Concerns for reservoir sedimentation problems, which often develop late in the project life, are relegated to a backseat position with the comfortable assurance that these problems can be passed on to future generations to resolve.
In the case of the Colorado River, the strategy was simply to build more reservoirs; the Glen Canyon Dam upstream of Lake Meade has enormous capacity and will safely retard sedimentation in Lake Meads for many generations. On the Rio Grande, additional reservoirs also have been constructed to help resolve the silt problem, but the results are not completely successful (Gorbach and Baird, 1991); sedimentation in the backwater reach of the reservoir is still a considerable problem that will net likely be resolved until the sediment yield from the Rio Puerco is reduced or controlled. The Rio Puerco is notorious for its high sediment load; it contributes about 5％ of the flow and 95％ of the sediment load to Elephant Butte reservoir. The project continues to function as designed, but maintaining the channel upstream of the reservoir become more difficult and expensive each year.
It might be useful to review a few examples of sediment problems at dams and reservoirs. Table 1 lists several reservoirs, their volumes, the mean annual flows and sediment loads, and some remarks on the status of the projects. Five of these are projects on which the writer has worked, the other examples are from the literature or from unpublished file data. The reservoirs are listed according to size, that is, the ratio of reservoir volume to mean annual inflow, and are plotted on Brune's trap efficiency curve in Figure 5.
In the previous discussion, it was noted that both the Three Gorges Project and the Xiaolangdi Project are designed and will be operated to preserve long-term storage in the reservoirs; that is, managing the sediment problems at these projects has a very high priority. This is because the highest priority of both projects is flood control rather than power production, and in order to insure adequate flood protection in the future, it is necessary to preserve long-term storage of the reservoir. At Xiaolangdi , additional sediment concerns involve sediment problems at the structure, including sediment damage to turbines, runners, gates, and other hydraulic equipment and erosion of the sluicing and power tunnels, and impacts on the channel downstream. By trapping mostly coarse sediments during the first few years of operation, there will be period of scouring and refilling along the downstream channel with some 20 years or more of zero net deposition, during which the costs of heightening the dikes along the lower river can be reduced substantially. After the reservoir come to equilibrium, there will still be some long-term benefits for sediment management because the flows during the flood season can be shaped to provide here efficient transport along the lower Reach.
Table 1. Status of reservoirs - some examples
- 40 -
Three Gorges is designed to reduce flooding, improve navigation, and generate power. If constructed, it will be the largest multi-purpose project in the world. Preserving long-term storage in the reservoir certainly is of primary concern, but numerous other potential sediment problems were identified in the feasibility study and preliminary design of the project, and manners to deal with these problems were identified. Navigation is an important component of the project, and conditions through the Gorges will be greatly improved with the reservoir in place. However. the upstream extension of backwater is near the port of Chongqing, so preserving navigation conditions in the vicinities of the ports and harbors around the city is of primary importance. After many years, depending on the flood control level at the dam, the sediment deposited through the reservoir will come to near equilibuium so that the annual sediment load can be passed through the reservoir with the flows that are spilled during the flood season. At this time, there may be sediment damage to turbines and sediment encroachment in the approaches to the lock structures. These problems were investigated though the use of extensive physical and mathematical modeling efforts. In many cases, the results from the one-dimensional math model were used as boundary conditions for the physical models. The sediment problems can be managed through a combination of proper design, careful operation of the reservoir, and some selective dredging and river training works.
The Rio Caroni drains the Guayana Shield, and the basin has a very low sediment yield; the ratio of reservoir volume to mean annual sediment load for Guri is on the order of 57,000 years (Table 1). No sediment problems were anticipated for this project. Recently, however, fairly extensive gold mining activities have developed in the basin, mostly using dredges or hydraulic mining techniques. The gold is recovered using mercury, and much of this accumulates in the sediments and moves into the reservoir. Already, mercury contamination may be a problem in fish populations of the reservoir, and studies are underway to determine the extent of the problem and methods to control the mercury usage. At this time, the hydraulic equipment must be flown into the basin by helicopter, so none of the mining operations are very large, but if large-scale hydraulic mining is ever allowed to develop, it is likely that large quantities of coarse sediments will be swept into the rivers and deposited in the backwater region upstream of the reservoir. This could have serious impacts on the economic feasibility of the proposed project at Tuyucai, just upstream of Embalsa do Guri, where a couple of
- 41 - meters increase in the tail water elevation on the turbines could results in substantial reductions in power benefits. EDELCA (C.V.G. Electification Del Caroni, C.A. ) has expanded their data-collection and monitoring program substantially since the mining began, and if necessary, restrictions may have to imposed to reduce both mercury contamination and sediment deposition problems.
The project at Carhuachi is under construction some 220 ㎞ downstream of Guri. The sediment load here is small, with average concentrations on the order of 10 to 20g/㎥, and is likely to remain small because most of the mining debris will be trapped upstream of Guri. There is some mining by dredges in the channel between Guri and Carhuachi, and there will be additional development in the small watersheds of tributaries entering the channel along the reach, but these are not likely to create any problems. The main sediment problem here arises during construction. The dam is designed with 12 low level outlets 10m wide and 11.5m high. Velocities through these outlets are about 28m/s. During the next stage of construction, a portion of the rock fill coffer dam will be removed and the river flows will be diverted through these outlet. Some of the debris from the coffer dam and other construction debris will be swept through the outlets, perhaps damaging the gates or concrete apron downstream. The question to be resolved is whether or not to install steel or other abrasion-resistant liners to reduce possible sediment damage. The design is not yet settled, but probably, there will be special protection provided for the gate sills and high-strength concrete with silicafume additive will be used on the spillway apron.
The volume of the reservoir upstream of Aswan High Dam is large relative to the mean annual flow and relative to the annual sediment loads, so no particular sediment problems are anticipated for this project. Nonetheless, sediment surveys of the reservoir are carried out each year because much of the sediment deposits in the live storage of the reservoir, between elevation 147 and 175, and in arid legions such as this. it is extremely important to have accurate stage-volume relations to plan the allocation of water. During dry years, such as were experienced during 1982-1988, shortages of even a few million cubic meters could have disastrous impacts on some farming communities. More recently, sediment surveys upstream of Aswan have been directed toward another important problem that was not anticipated at the time the dam was constructed. Some fairly small but fertile flood plains have developed in the backwater region of the reservoir and Nubians displaced when the Dam was constructed have moved back to these areas to settle. This resettlement apparently is encouraged by both the Sudanese and Egyptian governments. The flood plain are still developing, and their elevations will increase with time as additional deposition occurs in the backwater region of the reservoir. The High Aswan Dam Side Effects Research Institute (HADSERI) is presently trying to collect the data to develop a mathematical model of this region of the reservoir to predict future deposition and flood levels so that the Nubians can construct their homes, villages, and irrigation systems well above the future flood hazards.
At Tarbela, sediment is accumulating in the reservoir more rapidly than predicted, the delta front of the sediment deposit is advancing rapidly towards the dam, and in a few years sediment will encroach at the power inlets and jeopardize the power production. Possible solutions to the problem include dredging, constructing a coffer dam upstream to halt the advancing delta front, and constructing low level outlets sluice the sediment. So far as Ⅰ know, nothing has yet been settled on the issue, but any method used to deal with the problem is likely to be very costly.
Rosieres was designed with low level outlets for sluicing sediments, and the method of operation seems to have been successful in preserving about 2 billion cubic meters of storage, or about two- thirds of the reservoir's original capacity. However, no sediment sluices were provided under the turbines, debris piles up in front of the trash racks at the turbine inlets, and sediment deposits have forced closure of the hydropower on a number of occasions.
(Williams, 1991)
At present, the only solution to the problem is trash removal and dredging in front of the turbine inlets.
- 42 -
Lago Loiza is a water supply reservoir for the metropolitan area of San Juan. It is a small reservoir, with an original volume of only 7 percent of the meam annual flow. The stream has a heavy sediment load and during 38 years of operation, the reservoir has lost 70 percent of its capacity to sediment deposits (Collar and Guzman-Rios, 1991). In the original design, three low level outlets were provided for sluicing the sediments, but apparently these were covered with concrete during construction and are inoperable. Various alternatives are being examined to recover the storage capacity, and any course of action is likely to be expensive.
Amaluza on the Rio Paute was designed for power generation. The reservoir is small relative to the mean annual flow (3.2%) and also relative to the annual sediment load; the ratio of reservoir volume to volume of annual sediment load is only 32 years, Table 1. In some 9 years of operation, sediment deposits have overtopped the coffer dam left in place upstream of the main structure, and is encroaching on the low-level turbine inlet at the dam. Apparently, sufficient flow and sediment data existed at the time of design to anticipate this problem and design sluicing capabilities to manage the sediment and preserve long-term storage for this reservoir, but for some reason, this was not done. To maintain the power generation at the project, an expensive dredging program will have to be undertaken.
Rock Creek reservoir on the North Fork of the Feather River is the smallest of the reservoir listed in Table 1. Its volume is only 0.6 percent of the mean annual flow and about 18 times the volume of the annual sediment load. estimated from the reservoir deposits assuming a trap efficiency of 30％. The reservoir, which belongs to Pacific Gas and Electric Company, is used for peaking power with a maximum flow in the morning of about 80 ㎥/s. The reservoir is refilled each evening. After 40 years of operation, 60％ of the original capacity, about 3.5 million cubic meters, has been lost to sediment deposits. Plans are underway to evacuate about a million cubic meters of sediment and dispose of this off-site by trucking. The costs of these remedial measures are large, or the order of magnitude of the original investment in the project.
(Lee, 1991)
One of the criteria for preserving long-term storage is that the reservoir volume should be small relative to the mean annual flow. Only four of the reservoirs considered in these examples rate as large reservoirs, Figure 5, the other nine meet at least the size criterion. Probably, the sediment problems at some of these sites could have been managed more efficiently had the reservoirs been designed to preserve the long-term storage capacities.
Conflicts between Power Generation and Flood Control
In order to preserve the long-term storage on reservoirs, it is necessary to pull the reservoir down to a low level at the beginning of the flood season and to keep it at that level to the extent possible in order to pass the incoming flood and sediment load. The reservoir normally is surcharged infrequently and only in the case of extreme events where the downstream channel capacity is exceeded. On the other hand. in order to maximize the power production, it is desirable to maintain the water level over the turbines at the highest possible level. However, is this is done. as shown be some of the previous examples, the power benefits are ultimately lost and can be recovered only through expensive remedial measure. This inherent conflict between short-term power benefits and long-term flood control exists in most multi-purpose projects, so it is extremely important to sort out the priorities early in the feasibility stage of the project and design accordingly. However, it is important to keep in mind that although streamflow is a renewable resource and so is hydropower for low-head run of the river
- 43 - projects, storage is not a renewable resource. Once reservoirs are filled with sediment it is difficult and in many cases impossible to recover that storage.
SUMMARY AND CONCLUSIONS
Almost 60 years ago, J. C. Stevens pointed out some of the sediment problems associated with the construction of dams and reservoirs on sediment laden streams and he called on engineers and scientists to undertake the studies and research that would lead to practical solutions to the issues he raised. In this paper, I have attempted to review the progress that has been made over the past 60 years or so in dealing with reservoir sediment problems. It is clear that during this time a great amount of research has been carried out on the basic processes of sediment erosion, transport, and deposition, and a large amount of empirical evidence and practical experience has accumulated. Certainly, the theoretical and empirical basis for dealing with sediment problems now exists. However, judging from the examples and case studies compiled in Table 1, it is apparent that many problems still exist and that in a number of cases, these problems could have been avoided. A few general conclusions can be drawn from the history of development and the examples outlined in the previous sections:
1. Sediment problems associated with the construction of dams and reservoirs on sediment-laden streams generally cannot be solved; they have to be managed.
2. The theoretical and empirical basis to develop strategies for managing sediment problems now exists, for example, in the theory of preserving long-term storage of reservoirs and the use of mathematical models.
3. There is an inherent conflict between generating power for short-term benefits and preserving reservoir storage for long-term flood control and long-term power benefits.
4. Even today, reservoirs are being designed and constructed without proper considerations of the sediment problems. I believe that sediment engineers generally recognize that sediment problems cannot be solved, they have to be managed, and that we know how to develop appropriate management strategies. However. I believe that we have not communicated these ideas to the planners, dam designers, managers, and economists who ultimately make the major decisions on projects.
5. Finally, streamflow is a renewable resource, but hydropower is not if the power depends on storage of reservoirs that become filled with sediment. The conditions under which long-term storage of reservoir can be preserved indefinitely are rather restrictive, but clearly, where these conditions exist, the project should be designed to preserve that storage if it is at all feasible.
- 44 -
Figure 1. Reservoir trap efficiency. The data points are for reservoirs in China(after Ding Lianzhen, IRTCES, 1985).
Figure 2. Sluicing efficiencies (after Zhang Ren and Qian Ning. IRCES, 1985). Curve 1 is for median diameter, d. less than 0.04 ㎜ and concentrations, S, greater than 50 kg/m³ : curve 2 is for d less than 0.04 ㎜,S less than 50 kg/m³ : curve 3 is for d greater than 0.04 ㎜.
- 45 -
Figure 3. Ideal operating procedure to preserve long-term storage in reservoirs.
Figure 4. Definition sketch for preserving long-term storage in reservoirs.
- 46 -
Figure 5. Trap efficiencies for the examples listed in Table 1.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
SEDIMENTATION ASPECTS OF FLOOD - PLAIN MANAGEMENT IN BANGLADESH*
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Director, Surface Water Hydrology-2, Bangladesh 참여연구원 : Water Development Board(BWB) , Dhaka,
- 47 - Bangladesh Mohammad Alam Mial Deputy Director, Computer Centre, Surface Water Hydroloar-2, Bangladesh Water Water 참여연구원 : Development Board (BWDB). Dhaka, Bangladesh Md. Ali Husain
1. INTRODUCTION
Bangladesh is a Deltaic Flood-Plain Area situated at the foot hills of the highest mountain ranges of the world, The Himalayas and on the other hand the Bay of Bengal is situated at the South of the Country. Flood is a very sever and chronic problem of Bangladesh. The geographical location, flat and low topography, huge rainfall over the vast catchment areas of its river system, presence of tide in the Bay of Bengal in the South, silt laden huge discharge of water from upper young mountain ranges etc. have made the river system and sedimentation process very complex and unstable and consequently the flood problem also very complex and gigantic too. This country has been formed by the gradual deposition of easily erodible alluvial and recent deltaic sediments mostly carried by its three major international rivers, The Brahmputra, The Ganges and The Meghna. These rivers drain a vast catchment area of about 1.72 million sq. ㎞. which is approximately 12 times the size of the country(144,000 sq. ㎞). Only about 8 percent of this catchment area lie within the country and the rest are outside the country. The geographical location of Bangladesh with river network in shown in Exhibit-1. 1.
These processes of sedimentation have controlled the courses of the river system and formation of flood plain. The flood plain areas of this delta always remain under the active influence of these sedimentation and erosion processes. Every year, during flood season the waters overtop the river banks and flood some 20％ of the total land area of Bangladesh in a normal flood. The flood affected areas may be as high as 62％ of the total area as it happened in ever worst hit year 1988 which caused sever damage to agricultural products, homestead, loss of human lives and specially flood control and other development infrastructures.
Drought on the other hand in winter also affects the agricultural products and ultimately the economy of the country. River erosion is also a great problem for this country. Huge network of rivers, tidal and backwater effect under different conditions of rivers, the process of sedimentation and erosion and paradoxical problem of flooding in the monsoon and drought in winter etc. have made the hydrological system of the country very complex. This has necessitated to evaluate the performance and effect of development infrastructures and also to study the effect of different hydrological process including sedimentation, erosion, river morphology etc. on the flooding condition, drought condition and thus flood plain management. Comprehensive Flood Plain Management considering different hydro- morphological aspects are, therefore, necessary to mitigate these complex problems.
The paper is intended to study the sedimentation aspects of flood-plain management of Bangladesh.
2. GEOLOGY
- 48 - To understand the sedimentation process of the river system and also of flood plain, it is necessary to discuss the geology of the country and its land formation. The physiography of Bangladesh is closely associated with the formation of the Himalayan Mountains during the Tertiary period. Probably during the formation of the Himalayas, the ridges of the Chittagong Hill Tracts, Tippera Hills, Arakan and Naga HiIIs were also formd. Except for the Chittagong Region, Bangladesh has been formed by deposition of material transported by ancient and recent rivers. It is possible that these more resistant deposits influence the courses of the major rivers. The present Brahmaputra river used to flow through the old Brahmaputra course long before. The change of river course is illustrated in Exhibit- 2.1. Geomorphologic pattern of present day Bangladesh is shown in Exhibit-2.2.
3. HYDROLOGICAL SETTING
There are about 700 rivers in Bangladesh. The total river length in Bangladesh is about 22,000 ㎞. The topography of the land in Bangladesh is mostly flat with few hills in the north-east and south-east regions. The land relief constitute a gentle slope from north to south with some saucer shaped low pockets and depressions scattered all over the country. The land elevation varies from 1 metre in the south to about 90 metre above mean sea level in the north. About 50 percent of its land are below the elevation 12.5 m. Generalised relief contour map is presented in Exhibit-3.1. The river system carries huge discharge, both liquid and solid(sediment) during monsoon season, resulting from heavy rainfall of high intensity over the catchment and movement of easily erodible recent alluvial soil of the young mountain ranges and ultimately discharge into the Bay of Bengal. The combination of these huge flows and tide in the bay cause wide spread flooding in the country.
3.1 Rainfall
The rainfall in Bangladesh varies from place to place with maximum in the north-east and the minimum in western part of the country. The average annual rainfall varies from about 1000 ㎜ in the west to about 6000 ㎜ in the north-east region. The overall average annual rainfall of the country is about 2500 ㎜.
3.2 River Flow and Sediment Discharge
Particulars of river flows and sediment discharge of some major rivers are given below :
Table-3.1 River Flow
- 49 -
Table-3.2 Sediment Discharge
4. RIVER MORPHOLOGY AND SEDIMENTATI0N PROCESS
To study the flood plain management, it is necessary to understand the river morphology, sedimentation and erosion process of the flood plain.
4.1 Bank Stability
Bankline stability is very dependent on the behaviour of the river bed during flood and the subsequent fall of the river. In a meandering river, the changes and migration patterns are fairly predictable. In a braided channel, both banks may experience deposition or erosion simultaneously.
4.2 Deposition and erosion of the channel bed
- 50 -
River bank erosion at times created great problems. A rapid survey showed that rivers in Bangladesh were eroding at quite a large number of points over a long distance. Sedimentation also created serious problems in navigation, drainage and passage of flood water. River bank migration and changes in channel geometry seriously affect the operation and maintenance of Flood Control Drainage ＆ Irrigation (FCDI) projects and ultimately the flood plains. The change of river course of the Brahmaputra river from 1830 to recent years in presented in Exhibit-4.1.
5. FLOOD CONTROL PROGRAME OF BANGLADESH
5.1 Flood Vulnerable Area
About 66 percent of the total area or about 9.50 million hectares are cultivable land. The flood affected area may be as high as about 90000 Sq. ㎞. About 83700 Sq. ㎞. or about 60 percent of the country's total area is vulnerable to flood.
5.2 Flood Plain Area
Most people of Bangladesh live in rural flood plain areas. Due to various neceasities of live, the population, homestead, agricultural land have been gradually grown up in flood plain areas. Limited precipitation and flooding of land are necessary for various purposes like production and processing of agricultural crops, domestic work etc. When flooding exceeds the limit then it becomes a problem for every activity of life. Therefore, the necessity of controlling flood has been identified for the betterment of living condition of millions of people living in the rural flood plain areas.
5.3 Flood Control Programme
The Government have undertaken various flood protection programmes since long time. Other infrastructures have also been developed for common benefit and welfare of the people. The Government have so far implemented the following flood protection and flood plain management work to protect the flood plain areas from flooding.
- 51 -
5.4 Sedimentation Aspect of Flood Plain Management
The flood control and other infrastructural activities have helped to reduce the flood damages to some extent and on the other hand these have influenced the sedimentation and erosion process of rivers as well as flood plain areas.
The intensity, magnitude and frequency of flooding and also other natural calamities etc, appear to have increased in recent years. Many people have expressed various reasons for these calamities including environmental changes, global warming, deforestation, sea level rise, lack of coordination in infrastructure development, lessening of flood plain areas, sedimentation of rivers and flood plain areas etc.
Time trend analysis of low stage and low flow of the Brahmaputra river shows that the low water level has gone up considerably without any viable change in low flow regime (Exhibit-5.1). Similar trend of rise of high water level have also been noticed in other rivers, which is perhaps due to confinement of river bank and other reasons. Definite assessment of these changes have not yet been done. These should be done based on sound technical approach.
6. RECOMMENDATION FOR FLOOD PLAIN MANAGEMENT
6.1 Necessity of Flood Plain Management
- 52 -
In undertaking flood protection projects, both the advantages and the adverse effects of the projects should be considered. Flood plain management may also reduce flood damages specially in a poor deltaic agricultural country like Bangladesh. Proper handling of sedimentation and erosion process of the alluvial rivers will be definitely helpful in this respect.
6.2 Recommendations i) Adequate information and continuous research about sedimentation, erosion and morphological processes are necessary for better management of flood plain. ii) While constructing the flood protection embankment, sufficient setback distance, river training work with proper bank protection should be considered for the safety of embankment. Flood water at high stage should be allowed to enter the flood plain area to limited extent through regulated device for soil fertility with the deposition of these sediments. This should be done in balanced way so the agricultural, fisheries and other socio-economic activities are least affected. iii) Sound technical approach should be taken to assess the impacts of hydraulic structures on alluvial rivers. iv) Habitation of people in the river side of the polder should be discouraged Limited cultivation and plantation may be allowed in these areas by adjusting the cropping pattern so that least damage is caused to these corps. v) Inundation of flood protected area by internal precipitation should be prevented by drainage regulating devices and their effectiveness are to be reviewed and evaluated for proper improvement and maintenance. vi) Effect of sediment processes on different infrastructures are to be reviewed and studied. vii) Flood plain zoning may be considered to prevent loss of properties and human lives. The cropping pattern should be adopted according to flood plain zoning and depth of flooding in different areas. Industrial and other installations should be planned so that these are built to be flood proofing. viii) Land management should be practiced to reduce the sedimentation from upper reaches of river basin (by discouraging deforestation and encouraging afforestation etc). ix) Construction of embankment and polder in coastal areas have facilitate agricultural and fishing activities in these areas. Massive afforestation programme will help continue accretion of land by the sedimentation process and protect the agricultural and fishing farm and also the polders from devastation by cyclones and storm surge. The estuarine rivers should be maintained and protected against siltation of river bed so that drainage in least affected. x) International and regional co-operation in this respect should be explored for better management of whole basin area of rivers and transfer the technology in this field should be encouraged.
- 53 -
- 54 -
- 55 -
- 56 -
- 57 -
- 58 -
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
ON FAILURE OF RIVER TRAINING WORKS WITH REFERENCE TO AFNIKO HIGHWAY
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION
- 59 - OF HYDROLOGICAL SCIENCES(KAHS) Department of Hydrology and Meteorology, 참여연구원 : Kathmandu, Nepal K. p. Sharma Department of Hydrology and Meteorology, 참여연구원 : Kathmandu, Nepal S. R. Kansakar
Summary
Arniko Highway, joining Kathmandu and the Nepal-China border is one of the several highways constructed in the mountainous region of Nepal which is one of the most fragile mountains of the world. The road passes through steep mountains along the Sun Koshi bank in the Dolalghat-Kodari sector of the highway supported by different river training structures.
The failure of river training works along the sectors of Highway is a recurring problem generally occurring in the monsoons. Some historical floods such as the 1981 glacier lake outburst flood and the 1987 flood which was caused due to outburst of lakes formed by damming of river as a result of landslides were the two disastrous floods causing damage to the river training works, power plant and the road at several locations including bridges.
Basically the failures were associated with scouring around the structure, extra static and dynamic load on weak river banks and immense kinetic energy of the extreme floods. In addition to these criteria, the design of any structure along the river banks should consider the possibility of extraordinary floods which are associated with glacial or landslide lake outbursts.
INTRODUCTION
Nepal is a land locked mountainous country with more than 80％ of its area covered by hills and mountains. Physiographically Nepal can be divided into five regions, namely High Himalayan region, High Mountain region, Middle Mountain region. Siwalik region and Terai region. There is a wide variation in altitude from north to south. The altitude ranges from 8,848 m, the highest peak in the world to about 60 m in the country of average width of 193 ㎞. Due to the wide range of physiographic features, Nepal experience possibly all types of climatic features of the world.
The Himalayas include not only the highest peaks but also the deepest valleys of the world (Kaligandaki between Dhaulagiri and Annapurna and Arun between Khumbu and and Singalila). It is one of the youngest mountain ranges in the world, characterized by very steep landscapes with active erosion.
- 60 - The territory of Nepal is located in the upper and middle reaches of the major river systems of the Ganges basin. The upper reaches of the major rivers of Nepal comprise of the Tibetan plateau (watershed of Sapta Koshi, Gandaki and Karnali), Himalayan region, High mountain region and northern slope of Middle mountain region(Mahabharat). Most of the rivers originating in the southern slope of Middle mountain and Siwalik region are seasonal with flashy nature. The middle reaches of the major rivers pass through Terai region and wide river valleys in Middle mountain and Siwalik regions, whereas the middle reaches of the seasonal rivers are located mainly in Terai region. The middle reaches of the rivers are formed by the alluvial flood plains. A typical river profile is shown in Fig. 5.
Sediment load of all rivers in Nepal is very high especially during monsoon(rainy season). The size of sediment paticles varies from big boulders to fine silts and suspended clays. The Sun Koshi transports relatively large amount of sediment load since its watershed lies in highly erodible weak rock zones. The watersheds of two major river systems of Sun Koshi is shown in Fig. 1. The major tributary(Bhote Koshi ), originated in Thongla glacier in Tibet has most of its catchment area in the Tibet region. The Bhote Koshi flowing towards south meets a small tributary called the Sun Koshi at Barabise town in Nepal. The river is known as Sun Koshi in the downstream of the confluence until it meets the two major rivers Arun and Tamur of the Koshi basin. About sixteen kilometer downstream of Barabise. the Sun Koshi meets another river called Balephi which is drained out from Dorje Lakpa range of mountains. At Dolal Ghat, the Sun Koghi river meets the Indrawati river having catchment area of 1,225sq. ㎞. Further downstream the Sun Koshi river flows to the east nearly parallel to northern slope of middle mountain range (Mahabharat mountains). So the slope of the river at this part is relatively mild.
The Arniko highway, which joins Kathmandu valley with the Tibet autonomous region of China, follows along the Sun Koshi-Bhote Koshi river valley from Dolal Ghat to the Nepal-China border at Kodari. A run-of-river type hydropower plant with diversion canal scheme having capacity of 10. 05MW is located at Lamosangu, in between Balephi and Barabise section of the highway.
GEOLOGY AND TOPOGRAPHY
The Sun Koshi flows against the dip of the rock i. e. at right angle to the strike from kodari to Dolal Ghat. Dolomite is observed in the section from Kodari to Barabise, whereas the rest of the section consists of pre-cambrian gritty philite, schist and quartzite.
The river catchment considered in this study extends from Dolal Ghat to the northernmost end of Bhote Koshi in Tibet lying between the latitude of 27^{0} 39'N to 28^{0} 32' N and longitude of 85^{0} 40' E 86^{0} 18' E. The catchment area is elongated in north-south direction as shown in the Fig. 1. The elevation of the catchment area ranges from about 650 m at Dolal Ghat to 6,979 m at the Nepal- China border (Dorje). About 5％ of the basin is covered by permanent snow and glacier areas, whereas only about 10％ of the area is below 200 m elevation.
- 61 - STREAMFLOW CHARACTERISTICS
The flow in the region is maximum in July-August and minimum in February-March just before the snow melting season.
(Fig. 2)
The maximum discharge in July-August is the combined results of monsoon activity, maximum snow melt and high soil moisture storage.
The water levels recorded at lowest discharge condition for each year is plotted for Bhote Koshi at Barabise as shown in Fig.3, as it provides useful information on the stability of channel section. The figure shows that there is no significant change in section of Bhote Koshi until 1981 (before 1981 GLOF). The heavy silting due to 1981 GLOF was gradually removed in subsequent years. Significant alluvial deposits have been observed at the station since the 1987 floods.
No sediment sampling has been made in and around the study area. Comparing to the sediment loads of other river basins. which are similar to the river basin of Sun Koshi, the sediment load of the Bhote Koshi river at Barabise and the Balephi river at Balephi can be expected to be about 2,000tons/sq. m/yr and 3,000 tons/sq. ㎞/yr respectively. The bed load sampling haa been done nowhere in Nepal. The bed load is usually estimated in the range of 5％ to 25％ of the suspended sediment load.
DAMAGE TO STRUCTURES
The part of the road which is along the river valley is frequently affected by landslides, bank cuttings and floods. The severe floods including the GLOF had heavily damaged the Arniko highway in the past. Large scale damage however was recorded in 1981 and 1987.
The damage in 1981 was caused by a GLOF, which had bursting head of 35 m at the breach and the computed discharge at the source was 15,920 ㎥/s(3). The GLOF caused extensive damage to the Arniko higihway including washed away Friendship Bridge(bridge at the Nepal-China border) and Phulping Bridge.
The instantaneous peak discharge due to the GLOF at the Barabise gauging station was recorded 3,300 ㎥/s, which is about seven times the usual annual peaks observed at the station.
The disastrous Sun Koshi flood of 30 June 1987 is the historically recent flood of severe nature. The flash flood was caused due to outburst of lakes formed by damming of river as a result of land slides that occurred at various places on the upper parts of the Sun Koshi on 30 June 1987 following the torrential rains.
The flood badly damaged the Arniko highway, disrupted transportation system, swept away houses standing on both sides of the river and seriously damaged the Sun Kosih Hydro Power Plant. Two gates at the head work of the Sun Koshi HydroPower Plant were swept away. The diversion canal
- 62 - was damaged by the flood. The flood entered the Power Station and generators were filled with water. About 45 ㎝ high sand had accumulated on the upper floor. A 9m high flood flowed down the Sun Koshi river which was 1 m higher than the overflow dam built to check the discharge of 2,100 ㎥/s(2).
The photographs presented in Fig. 4. show some of the damaged portions of the Highway and river training works.
TYPES OF STRUCTURE
The river training structures in the study area are constructed for the purpose of the protection of highway. Bank erosion is one of the main causes of river instability. The process of bank erosion in this area is very active. Hence the necessity of the river training works for bank protection continues to be a recurring problem. The bank erosion is caused mainly.
- due to the disturbance of naturally stable slope by construction of road along the river bank,
- due to seismic disturbances.
- due to meandering of river.
Construction of road along the river bank increases static as well as dynamic load to the river bank causing its instability. Such effects can also be seen in different parts of other highways (Kathmandu - Pokhara, 200 ㎞ long) in the section from Naubise to Mugling (26-110 ㎞) is one of the notable examples.
Similarly, the seismic disturbances weaken the river banks. The Himalaya, one of the active seismic zones of the world is covered with criss-cross faults and thrust planes. Most of the river channels of Nepal follow either fault or fold.
The most common material used for river training works is gabions filled with stone. The following types of structures are constructed in the study area:
- dry stone and stone masonry revetment.
- dry stone, stone masonry and PCC spurs mainly normal to the bank with and without launching apron
- gabion mattress
- river bank protection with vegetative cover
- stone and vegetated stone riprap (lining)
- 63 -
CAUSES OF DAMAGE
River banks are the most fragile lands. Even a minor disturbance induces instabilities causing the bank failures. Construction of roads along with river training structures are the major factor disturbing the natrual stability of the river banks and river flow regime. After the initial river training measures taken during the construction of highway the river has undergone many changes. Since the flow condition of the river is changed by the structures, the stream creates a new equilibrium condition by scouring the bed, depositing sediment on bank, local scouring around structures and changing the form of river in plan. Most of the spurs and revetments are fornd to be damaged by local scouring around the structures because the foundations provided are not deep enough. Some of the structures were constructed without foundation. It is very important to have a good foundation for each hydraulic structure. It should be extended downward below maximum scour depth with some provision for safety.
The other major cause of failure is the improper design of the river training structures. Some of the sturctures were found to be damaged due to structural instability. The washing away of the Friendship and Phulping bridges and of different sections of road indicate that the road alignment was not sufficiently above the level of severe floods.
In some cases it is difficult to provide sufficient depty of foundation due to some constraints, which may be as follows:
- lack of flood data
- lack of sediment data
- lack of river behavior history of the site
- lack of high class technology
- limited budget
- urgency of the repair works
RECOMMENDATIONS
1. Construction of river training structures with lunching apron when the structures have to be constructed without foundation due to different reasons. In such a case, local scouring will take place below launching apron and the apron may settle down up to maximum scour depth which prevents the main structure from settling down.
- 64 - 2. Use of bioengineering measures in river training structures of dry stone masonry seems to be very successful from a technical, economical and environmental point of view. The main protection effect of plants in river training is caused by the gripping capacity of their roots. Two typical vegetated river training structures are presented schematically in Fig. 6.
There are different varieties of trees and bushes recommended for river training works for different types of climatic and soil condition of the site (1). Some of the varieties of trees and plants recommended for Nepalese context are Bainsh ( Salix tetrasperma), Lahare Pipal (Populus euramericana), Utis (Alnus nepalensis), Sissoo(Dalbergia sissoo), Seto Siris (Albizia procera), Dabdabe (Garuga pinnate) , Lankuri(Fraxinus floribunda), Tooni (Toona ciliata), Bayer (Ziziphus mauritania), Bihaja Bejarme ( Ipomea fistulosa), Aak (Calotropis gigantea), Assure (Adhatoda vasica), Hasua (Cestrum nocturnum), Dubo (Cynodon dactylon), Kikuyu (Pennisetum clandestinum), Molasses (Melinis minutiflora), Kansh (Saccharum spontaneum), Napier (Pennisetum) and others.
3. While selecting road alignment along river bank it is very much necessary to study on geology, soil stability and river behavior as well. the stability of river bank and subsequently the stability of bank protection structure should be checked by considering static as well as dynamic load of transportation and bank erosion capacities of the river. Lack of proper evaluation of all these factors is found to be a major cause behind the failures of bank protection works in Nepal.
Stability of structrues can also be increased by a suitable combination of revetment and riprap to match slope of the river bank.
4. The roads or any structures along river banks should be located sufficiently above the possible flood level. The possibility of GLOFs and landslide lake outburst floods should be properly evaluated for this purpose.
- 65 -
FIG 1
- 66 -
FIG 2
FIG 3
- 67 -
- 68 -
FIG 4
- 69 -
FIG 5
FIG 6
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
- 70 -
SEMIMENTATION ASPECTS OF FLOODPLAIN MANAGEMENT FOCUSED ON THE 1990 HAN RIVER FLOOD
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Researcher, Korea Institute of Construction 참여연구원 : Technology, Korea Kwonkyu Yu Research Fellow, Korea Institute of Construction 참여연구원 : Technology, Korea Hyoseop Woo Chief Hydrologist, Harza Hgineering Co., 150 S. 참여연구원 : Wacker Dr.,Chicago IL. 60606-4288, U.S.A Baum K. Lee Director, Korea Institute of Construction 참여연구원 : Technology, Korea. Byungha Seo
Summary
This paper presents sedimentation and erosion aspect of a recent disastrous flood that occurred during September 10-12 1990 in the Han River basin, a central part of the Korean peninsula. The flood caused massive sedimentation on the most of the flood plains and costly damages to the public facilities and infrastructures that had been constructed only recently. The flood also taught some valuable lessons regarding the sedimentation and erosion aspect to the concerned government authorities and engineers. They include; 1) A sound management of urban floodplain should encompass plans for alleviation of sediment deposition and for effective measure of sediment removal after flood; 2) Planning and design of floodplain modifications and developments must be based on careful studies on behaviors of movable-boundary rivers such as the Han River and of sediment inflow (mostly wash load) from the tributary watersheds to the Han River during extreme flood events; and 3) Sediment data as well as hydrologic data related to the flood of September 10-12 should be documented for improvement in the future floodplain management.
- 71 -
1. INTRODUCTION
Sediment deposits that occur on the floodplains in urban areas can cause serious problems on streets, highways, recreational facilities, and other developments. Maintenance and cleanup costs of sediment deposits on the urban floodplains are often unexpectedly high.
This paper deals with sedimentation and erosion of floodplain management for improvement of urban river environment. It focuses on lessons learned from the recent Han River flood that occured during September 10∼12, 1990, in the metropolitan area of Seoul, Korea. The following four subjects are discussed in this paper :
- Overview of the Han River flood of September 10∼12, 1990.
- Overview of the damage caused by the flood to the Lower Han River Development Project that had been completed recently.
- Estimation of sediment sources and sediment discharge during the flood in the river reach in Seoul and sediment volume deposited on the partially man-made floodplain which is an integral part of the Lower Han River Project.
A comparison of the observed sedimentation and erosion pattern due to the flood to those predicted by the physical model studies in 1983.
2. THE HAN RIVER AND FLOODPLAIN MODIFICATION
The Han River flows through the metropolitan area of Seoul, the capital city of Korea, and its political, economic, social and cultural centre. As shown in Fig. 1, the Han River basin has an area of 26,218 ㎢, about 27 percent of the national area. Over 12 million people, about 30 percent of the national population live in the basin. The river, about 470 ㎞ long, is the longest in South Korea. The river has two major tributaries - the Buk Han River and the Nam Han River, which merge at about 30 ㎞ east of Seoul and flows into the Yellow Sea.
The Han River is the main source of water supply, irrigation, and hydroelectric power to citizens of metropolitan Seoul and population in the basin. Besides this tangible importance, the Han River provides other benefits such as a scenic environment, pollution mitigation, recreational and athletic spaces. To maximize these benefits without any major damage to the natural environment, the Lower Han River Development Project was implemented in 1982, and completed in 1986. For this project, a total amount of about 400 billion won (570 million U.S. dollars) was invested.
As part of this project, nearly entire floodplains of the Han River in the metropolitan area of Seoul was developed in order to provide the public with sports and recreational facilities and grass lands. A total
- 72 - area of 693 million ㎡(69,300 ㏊) including sports and recreational areas of 310 million ㎡(31,000 ㏊) and grass lands of 383 million ㎡(38,300 ㏊) have been developed by the project.
3. THE HAN RIVER FLOOD
During September 10 through 12, 1990, the Han River basin received an areal average rainfall of more than 370 ㎜ with point rainfall-depths in some parts of the basin, exceeding 500 ㎜.
(Fig. 1)
This heavy precipitation caused an extreme flood all along the Han River reach and its numerous tributaries. As shown in Fig.3, the peak discharge at a gauging station on the Han River in Seoul reached nearly 30,000 ㎥/sec, which is the second highest in the 80 years of available record, only exceeded by the 1925 flood.
4. OVERVIEW OF DAMAGES CAUSED BY THE FLOOD
The September 1990 flood caused severe damages which included inundation of houses and paddy lands, riverbank failures, wrecks of several tour-boats of more than hundred tonnages, and especially sedimentation and erosion of the floodplains modified by the Lower Han River Development Project. Among these numerous damages caused by the flood, damages due to sedimentation in most areas of the floodplains and scourings in some parts, were very extensive. The floodplains, 50-500 meter wide and about 36 ㎞ long along both sides of the Han River reach in the Seoul metropolitan area, were modified and developed extensively during 1982-1986 as a central part of the project. It was designed to be inundated at a recurrence interval of 1.5 year. No serious consideration was given to sedimentation and/or erosion problems that could be caused by floods, when the planning of the Lower Han River Development Project was made.
Much of damages to the facilities and infrastructures in and near the floodplains were due to heavy sedimentation that required massive cleaning and/or excavation works. Fig. 1 shows an areal distribution of depth of sediment deposits on the floodplains developed along the Han River reach in the metropolitan area of Seoul. As shown in this figure, about 0.1 to 0.3 meters of sediment deposited on most part of the floodplains. Most sediment deposits are in the fine sand and silt range with D_{50} being equal to 0.1-0.3 ㎜. Fig.4 shows a concrete walkway next to the river buried under sediment of more than 1.5 meters that needs to be excavated. This area was especially deeply buried with sediments because it is located immediate next to a convex side of the meandering river reach, producing an eddy motion with a large amount of silting. As shown in Fig.5, however, some parts of the floodplains were scoured by the flood torrents, resulting in exposure of footings of playground facilities.
- 73 -
The overall effect of the sediment deposition on the floodplains was very detrimental . The sedimentation damages were mainly associated with cleaning sediments from recreational areas, parking lots, roads and highways, piers, and grass lands. The total cost of cleaning the entire sediment deposits was estimated to be about 5 billion won (7 million U.S. dollars). Due to this enormous amount of sedimentation, cleaning activities were limited to parks, piers, highways and access roads, and other facilities, which are essential to the public. The grass lands created on the floodplains covered by sediment deposits of more than 0.1 meters thick were left undone for re- seeding in the next spring season.
Another major detrimental effect of the sedimentation was temporary shutdowns of various facilities that were constructed on the floodplains. The shutdowns lasted from a few days to months, depending upon the amount of sediment to be removed.
5. SEDIMENT DISCHARGE AND DEPOSITS ON THE FLOODPLAINS DURING THE FLOOD
Total volume of bed sediment transported during the flood was estimated by using the flood hydrograph and bed material data of a gauging station in the Han River reach in Seoul , and by using some selected sediment transport formulas. These include Engelund ＆ Hansen's (1967) and van Rijn's formula(1954), which have been known generally reliable (White, et at., 1973; Woo and Yu, 1989), and Karim ＆ Kennedy's empirical formula(1983), which is based primarily on data collected from the Missouri River. The bed sediment characteristics of the Missouri River are similar to those of the Han River. The median diameter of the bed material collected at the gauging station immediately after the September flood ranges from 0.3 to 0.5 ㎜ with a gradation of about 1.5.
Fig.6 shows sediment-rating curves calculated from the three selected sediment transport formulas. The total sediment transported through the gauging station during the flood is estimated to be between 7.8 million metric tons by Karim and Kennedy's, and 2.7 million metric tons by van Rijn's formula. A bed material discharge of 4.0 million metric tons, which is estimated by Engelund and Hansen's formula, is judged to be reasonable. The average annual bed sediment transported through the gauging station is estimated to be about 2.5 million metric tons. This means that the bed sediment discharge during the flood period of only three days was nearly twice the average annual bed material discharge.
No field measurement for sediment transport was made during the flood in the Han River. Wash load, which can not be calculated by using the existing sediment transport formulas, therefore, was estimated to be about 30 to 50 percent of total sediment load. This estimate is based on an actual field measurement of sediment concentration in a tributary of the Han River during the flood. The total sediment load transported through the gauging station during the flood, therefore, is estimated to be about 8 million metric tons, which is equivalent to an average sediment discharge of 31 metric tons/sec, or an average sediment concentration of 1,500 ㎎/1.
Total volume of sediments deposited on the floodplains during the flood was estimated from field survey data shown in Fig.2. According to this figure, depth of sediment deposits on the floodplains varies widely, depending upon the configuration of the river reach. In general, however, it exceeds at least 0.1 meters throughout the whole area of the floodplains developed, except a few areas that
- 74 - indicated erosion during the flood. The total volume of sediment deposits was estimated about 1.5 million ㎥, which is equivalent to about 3.0 million metric tons by weight.
6. COMPARISON WITH RESULTS OF PHYSICAL MODEL TEST
A series of physical model test of movable-bed boundary on the Lower Han River had been conducted in 1983 before the initiation of the Lower Han River Development Project. The main objective of the test was to approximately identify scour and deposition pattern along the Han River reach that would be affected by the floodplain modification. During the test, sediment entering into the upstream test boundary was regulated so as not to cause an excessive deposition on the river bed tested. The scale of the model was 1/60 in the vertical and 1/200 in the horizontal. Bed materials used in the test was mainly diatom earth with median diameter of 0.4 ㎜ and specific gravity of 1.54.
The overall plane view of scour and deposition pattern along the tested river reach is shown in Fig.7. As shown in this figure, the scour and disposition Pattern appears to be related to the sinuosity of the river, with scours on the convex sides of the meandering reaches, and deposition on the concave sides. It appears that, however, the test result shown in Fig.7 hardly reproduces the prototype pattern caused by the actual flood shown in Fig.2. Except in few areas including those marked in "(－)" in Fig.2, most areas along the river reach received sedimentation during the tailing stage of the flood, while the model test had resulted scours to be prevalent all along the reach tested. This discrepancy appears to be due primarily to the following reasons:
- Most of the sediment deposited on the floodplain is in the fine sediment-or silt-size range, which implies that the depostion occurred during the subsiding period of flood. The behavior of the fine sediment, like wash load. was not considered in the model test, because the main interest of the test was to analyze changes in the river beds along the test reach, in which bed material movement was a main concern.
- The interruption of upstream sediment supply during the model test would have caused scours to be more prevalent than in reality.
It seems clear that inclusion of fine sediment load, or wash load, in the study of movable boundary rivers is essential in the planning and design of river training or floodplain modification. The effect of wash load and its sources during extreme flood events on floodplain management should be carefully studied.
7. LESSONS LEARNED FROM SEDIMENTATION ON FLOODPLAIN MANAGEMENT
Major lessons learned from the sedimentation on an extensively developed floodplains for urban uses include:
- 75 -
1) Neccesity for a long-range floodplain management must be recognized by policy makers, concerned government authorities, and practicing engineers. The sound management of urban floodplain must encompass plans for alleviation of sediment deposition during floods and for effective measure of sediment removal after floods.
2) Planning and design of floodplain modifications and developments must be based on careful studies on behaviors of movable-boundary rivers such as the Han River and of wash load and its sources during extreme flood events. Costs of such studies are insignificant as compared with the costs associated with floodplain restoration.
3) Hydrologic data related to the flood of September 10~12 should be documented for improvement in the future floodplain management. Such data should include detailed data of sediment discharge, sediment deposits, sediment sources, changes in rivercross section, as well as data of flood hydrograph, rainfall, and upstream reservoir operation. Video recording would have been very useful.
ACKNOWLEDGMENT
The data for this study were collected through a series of field survey during and after the flood. Some valuable data of sediment deposits on the floodplains, which were essential to this study, were provided by two metropolitan Seoul municipal officers, Mr. Sungun Nam, and Mr. Kwangbin Choi. Their help is greatly appreciated.
- 76 -
Fig.1 The Han River Basin: Isohyetal Map of Three-Day Rainfall during the September Flood.
Fig.2 Floodplains developed by the Lower Han River Development Project.
- 77 -
Fig.3 Rainfall Hyetograph and Runoff Hydrograph at Indogyo Gauging Station during the September Flood.
- 78 - Fig.4 Concrete Walkway buried under Sediments to be excavated (Chamwon District).
Fig.5 Playground scoured by the Flood Torrents (Yanghwa District).
Fig.6 Sediment-Rating Curves developed by Sediment-Transport Formulas and the Flood Hydrograph.
- 79 -
Fig.7 Plane View of Scoured and Deposited Area: Simulated by the 1983 Physical Model Test
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
RESPONSE OF THE CITANDUY RIVER DUE TO DEVELOPMENTS
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS)
- 80 - Chief of Sub Directorate of Planning and Design, Directorate of Rivers, Directorate General of 참여연구원 : Water Resources Development, Mini stry of Public Works, Indonesia Biswoko Sastrodihardic
Summary
The Citanduy River Basin is located on the southern coast of island of Java, and has a catchment area of about 3,000 square kilometers, as shown in Figure 1. The Citanduy Project Area consisting of some 4,460 square kilometers, is composed of the Citanduy River Basin and the Segara Anakan Lagoon and its tributaries (about 960 square kilometers).
In the past twenty years, the Citanduy River Basin has experienced extensive growth and development. This development has provided flood control and irrigation and has enhanced the quality of life of the people living in that area.
As a result of development (man-induced) and also natural influenced (e.g. eruption of volcano Salunggung) , the Citanduy River has responded to variations in its environment.
The sinuous alluvial reach of the Citanduy River has been relatively stable in spite of man-made cutoffs which have reduced the channel length from 98.7 ㎞ to 78.6 ㎞ over the past 80 years. When continuous levees were constructed on both banks, confining floods primarily to the main channel, bank erosion began in slightly more than half of the remaining 96 bends. The rates of erosion in many cases are not severe. The banks of the river are composed of hallyosite, a strong, light clay derived from volcanic rocks, which gives the river its stability. The eruption of volcano Salunggung in the headwaters in 1982 caused aggradation in a short upstream subreach of the alluvial river ; but farther downstream the effects of Salunggung sand were disappear due to the response of the river bed to levees. Based on the result of monitoring, there is a critical range of bend radii and deflection angle for which the pool depth is a maximum.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS
- 81 - INCUCING THOSE OF INTERNATIONAL RIVERS
PROBLEMS OF RIVERS IN KLANG RIVER BASIN
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Department of Irrigation and Drainage, Malaysia 참여연구원 : Ahmad Husaini Sulaiman
1. INTRODUCTION
Rivers have been very close to our human life since ancient time. They bring in the resources of irrigation, city-water, industrial water, hydro-power energy, navigation and drainage. Despite of all these, rivers are not always the mother of blessing. There is always this fear of floods in times of heavy rainfall. If no emphasis is given to cope with the problems of flood, any after-flood remedial works can be very costly.
Malaysian rivers are small by world standards. The fact that they are very steep and situated within the region of high torrential rainfall, high flood discharges are imminent.
This paper discusses the problems of Malaysian rivers due to high discharges with special attention to those in Klang Valley where urbanization is rapid.
2. THE KLANG RIVER BASIN
The Klang River basin is situated on the west coast of Peninsular Malaysia, covering an area of about 1,300 sq. ㎞.
- 82 -
(Fig. 1)
Its main river is the Klang river of length 120 ㎞ with tributaries, the Batu river, Combak river, Kerayong river, Damansara river, Kayu Ara river, Penchala river and others.
(Fig. 2)
The river basin is the most developed area in Malaysia. It consists of four major urban areas; the federal capital, Kuala Lumpur; Petaling Jaya; Shab Alam; and the Klang/Port Klang conurbation.
The climate of the Sg. Klang Basin is humid tropical with uniform temperature, high humidity and heavy rainfall. Average annual rainfall in the basin ranges from 1,900 ㎜ near the coast to over 2,600 ㎜ at the foothills in the east. The mean monthly temperature ranges from 26-28℃.
The upper port of the Sg. Klang is predominently underlain by igneous rocks, while the central area is underlain by limestone of the Kuala Lumpur Limestone formation. Some parts of the valley consist of schist, granite quartzite and phylite, while quartennary alluvium dominates the lower reaches, especially the river delta.
In terms of land use the river basin ranges from tropical forests to urban. Forests and swamps account for about 25％ of total basin area and are concentrated mainly in the upper parts of the basin area. Agriculture constitutes about 41％, consisting mainly rubber and oil palm plantations. Urbanized areas, consisting of residential, commercial, industrial, institutional and recreational areas forms about 29％ of the total basin area.
3. PROCESS OF DEVELOPMENT
The rapid increase of population in the urban areas has been the phenomena of the world since the middle of the century ago. Population migration from the rural to urban areas is certainly quite common in Malaysia. The process somehow promises good and better income to the people. This is very obvious, when agriculture in the rural areas faces the problem of labour shortage. What ever it takes, population concentration has to be on where there are better fortunes.
The Klang River Basin has all these promises that attract the influx of the rural population. As a result, the area becomes urbanized, where more housing and industrial development are needed. Whilst, development persists the whole ecology of the basin changes and this is apparent to all the rivers within the area. Indirectly, the hydrological regime of the rivers is changed.
The process of development, inevitably, increases the storm run-off in the rivers and very often than not, causing a number of floodings and bank erosions. The problems, in many ways, require serious attention and to be dealt with in order that the course of development is not impeded.
- 83 - For Klang River Basin the requirement to provide proper drainage for all development is a condition that must be heeded. An area that is to be developed into housing, for example, must comply to the drainage need of the area and the developer is to pay drainage contribution to the Authority to allow for the river to where its drainage system flows to, be upgraded to cater the increased storm run-off. The role of the developers in this respect has provide sufficient fund to improve these streams and rivers which are mainly the tributaries of the Klang river. The Klang river, being in itself the main supplier of all the flood discharges is also improved despite the space constraint due to part of the course that flows across the city centre of Kuala Lumpur.
(missing). concentration of run-off which are very much in excess of the capacities of the streams and rivers. The catchment area which once a forest and now intruded by uncontrolled development have resulted in the gradual deterioration of the rivers as efficient conduits of water and sediment. Thus, while previously, the occasional flooding was accepted as a way of life, it is now no longer true especially as floods become more destructive to public facilities, social and economic activities provided by the so-called socio-economic development.
In the history of Malaysia, the biggest flood occurred in 1926. Major floods were recorded subsequently in 1931, 1947, 1954, 1957, 1967 and 1971. Those were days when most areas were still undeveloped. Beyond ones inagination the kind of damages that can happen if the same magnitude flood would occur now.
(missing). three stretches, namely, an upstream stretch with steep slope of above 1/400 to 1/2300. a middle stretch between confluence with the Gombak River and a point at about 10 ㎞ downstream of the Puchong Drop. where a slope transition occurs(Fig.3).
The lower stretch is relatively gentle with slope between 1/2300 to 1/10,000.
Throughout the entire length of the Klang River there is a few active meanders. ocal erosion occurs at meanders along the middle stretch. There are also sediment deposits caused by soil erosion due to housing development. Some sections of the river- in the upper stretches have been canalized.
The Klang River has at some points form major constrictions to the flow of the flood run-off. This is especially common along the stretch in the city centre of Kuala Lumpur where many bridge crossings are present. At Sulaiman Bridge the flow is not to exceed 730 ㎥/s. Cost and space are generally the main hindrance to further improve the section to cater larger discharge.
Bank failures are not very serious for this river as most stretches are either concrete lined or built with stable side slopes consisting of berms and proper turfing.
5.2 Jinjang River
The river has a catchment area of 29.5 ㎢. Bank failures are common in some stretches because of its steep side slopes which are constantly eroded during heavy downpour. There are buildings built quite close to the river banks which are always threatened by bank failures.
- 84 -
Presently, the river section is able to cater a 10 year return period storm of about 10 ㎥/s.
5.3 Kerayong River
The river has a catchment area of 61.8 ㎢. Bank failures are also common. Reserves for widening the river section are available on most stretches.
5.4 Other Rivers
There are many other rivers that flow within the Klang River Basin which, generally, acquire similar characteristics. Their main problem is the inadequate flow capacity of the river channels, resultant from the increased run-off by developments in the upper catchments.
6. RIVER IMPROVEMENT
River improvement work is necessary, firstly to restore rivers to their natural efficiency and secondly to assist the river to discharge its flow and combat its tendency to meander, erode its bank and flood harmfully.
For Klang River and its tributaries, river improvement works consists of the follwoing requirements.
For other rivers such as Damansara, Jinjang, and Kerayong Rivers, the most frequent measure adopted is by deepening the rivers and canalising certain stretches. Concrete lined sections are also introduced in places where lands are costly and space for widening with earth sections is not available.
- 85 -
In some areas the construction of levees or embankments is often used in conjunction with the above measure and is aimed at confining the flood flow within two raised river banks.
7. SEDIMENT TRANSPORT
Annual suspended sediment load and yield in the Klang River Basin are high; ranging from 165-2,283 tons/㎢/yr. Although high by Malaysian standards, they are somewhat low compared to some rivers in this region. Sediment transport will not stop unless the river catchments reach an equilibrium of total urban because of the protection of soil by pavement and concrete.
Many studies had been carried out by various institutions and individuals to address the problem of sediment transport in this area but a discreet solution is yet to be adopted.
As it is upto now, frequent desilting works are carried out using excavator machines and in some wide river sections dredgers are used.
Below is a summary of suspended sediment loads and yield for some rivers in the Klang River Basin.
8. CONCLUSION
The problems of rivers in the Klang River Basin have been mainly focussed to their inadequate river channel capacity to cater the resultant discharge increase by rapid development in the area. Bank
- 86 - failures have become quite critical in some areas. Canalising including deepening of the rivers, constructing concrete lined sections and forming of levees are the counter-measure adopted to the river improvement works.
Sediment transport in the rivers is a phenomena that requires proper attention, to safe guard the stability of the river beds. Frequent desilting works are inevitable in this area so as to discourage siltation.
Fig. I LOCATION MAP OF THE KLANG RIVER BASIN
- 87 -
Fig. 2 LOCATION OF RIVERS
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
IMPACTS OF HYDRAULIC STRUCTURES ON ALLUVIAL RIVERS : RIVER PROCESSES OF TIDAL REACHES
1991
- 88 - 주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Executive Engineer, Bangladesh Water 참여연구원 : Development Board. Fellow No. 2347. Institution of Engineers, Bangladesh Dhali Abdul Qaium
Summary
This paper presents the experience of interaction of hydraulic structures on alluial rivers in deltaic coast of south-Western Bangladesh under tidal influence. Deltaic cost of Bangladesh possesses a unique location; meeting point of two distinct geological system; the Himalayan Drainage Basin and the funnel shaped Bay of Bengal. Huge volume of water (100,000 ㎥/sec. during flood) and sediment (1.5 to 2.4 billion tons/year ) carried by the Brahmaputra, Meghna and Padma/ Ganges from the catchment of Himalayan Drainage Basin plays a vital role in developing the coastal area of Bangladesh which is still active and under the process of development. Obviously, the coastal area has the lowest land elevation in the Himalayan Darinage Basin.
Coastal region of Bangladesh may broadly by divided into three marked geomorphological regions, the eastern, central and western region. The western region covers the coast line from the Tetulia river up to the border and lies in the south-western part of Bangladesh. According to agro-ecological zones study of Bangladesh based on the physiography, soil and flooding characteristic, the area falls under Ganges Tidal Flood Plain. The area, a very flat land with innumerable low lying beels, is cris- crossed by numerous tidal rivers and creeks. Under natural conditions, the land is subjected to periodic flooding at high tide. Till the early part of the fifties and before the Zamindari system was abolished, the local Zamidars and the people used to construct low height seasonal dykes along the river banks with wooden box sluices for restricting tidal water intrusion in agricultural land during pre ＆ post monsoon period to protect crops. Within the period of early sixties and early seventies, polders were constructed with drainage and flushing sluices in the Coastal Embankment Project (CEP) under the patronage of the Government to make a permanent solution. Immediate impact of CEP was tremendous on increased crop production.
While emplodering the low lying flat land, tidal rivers/creeks were blocked reducing the tidal prism flowing through the outfall channel. Moreover, diversion of the Ganges water at Farakka causes substantial reduction of upland from in these tidal channels as all these channels are interconnected with the Padma/Ganges. Channel geometry of alluvial rivers is affected by its discharge, sediment load, bank and bed resistance to flow and many other variables. In case of tidal river, minimum cross- sectional area of a tidal creek at entrance is linearly related to the tidal prism. Predication of response of alluvial tidal channel form due to changed situation is a very complex task. In case of CEP, reduction of tidal prism and up land flow in the outfall channel of empoldered area resulted in progressive siltation of the outfall channel causing drainage congestion within the polder. Restriction of tidal flow inside the polder has adversely affected the land formation process inside the polder and natural subsidence of the polder soil may be there causing detrimental effect on polder drainage. The problem is very acute is polder 25 of CEP where 19430 hectares of land out of 409,757 hectares of emplodered area under greater Khulna and Jessore district is facing water logging.
- 89 -
It is apprehended that the situation may lead to irreversible process forcing the inhabitants to migrate or to change profession.
1. INTRODUCTION
Flow phenomena in the deltaic reach of an alluvial river under tidal influence is highly complex. Man made interference, in the form of dam, barrage, lock, levee, intake, closure etc. on this natural process adds higher degree of complexity to it. Detailed study, based on investigation, analysis of long term data base and experience from similar work, should be carried out before undertaking any development work on a sensitive and intricate natural ecosystem comprising of alluvial rivers under tidal influence; otherwise, continuous interaction between manmade structures and natural processes may lead disastrious consequences; even, in some cases, the natural drainage system, where the structures are built, may stop functioning leading to complete change in the hydromorphological characteristics of the area.
2. OBJECTIVE OF THE PAPER
This paper presents the experience of interaction of hydraulic structures on alluvial rivers in deltaic coast of south-western Bangladesh under tidal influence.
3. THE STUDY AREA
3.1 Location
Bangladesh, a country in south Asia between latitude 20^{0} 3'N to 26^{0} 7'N and longitude 88^{0} E to 92^{0} 7' E having an area of 144,000 square kilometer, is bounded by the Indian states of West Bengal and Bihar on the west, Assam on the north, Assam and Tripura on the east, Burma on the extreme south-east and the Bay of Bengal on the south.
(Fig-1)
- 90 -
Depending on the hydromorphological condition, the country(Bangladesh) may be divided into four zones as northern zone, north-eastern zone, south-eastern zone and south-western zone.
(Fig-1)
The study area lies in the central part of south-western zone, in close proximity of Khulna town.
(Fig-1)
The area is located about 100 kilometer north of the Bay of Bengal and separated from the Bay by a large tidal swamp mangrove forest known as Sundarban.
3.2 Geololgy
Bangladesh, one of the biggest deltas, the largest tidal flat plain with magnificent mangrove forest ( the Sundarban), the deepest sedimentation basin and the largest deep seafan of the world (Anwar,1988) and part of the Bengal basin, possesses a unique physical geography. Himalays ; the youngest, the highest and the largest of mountain ranges on the north and the Bay of Bengal ; its funnel shape, shallow water depth in the north Bay, the swatch of no-ground (submarine canyon) and strong tidal and wind action on the south controls the hydrogeomorphological characteristics of the land mass of the country. The deltaic coast of the country, located at the meeting point of two distinct geological system, the Himalayan Drainage Basin and the Bay of Bengal and stretching along a length of about 710 kilometer (Nisat,1988) from Teknaf river in the south-east up to the mouth of Raimongal river in the south-west, is still very dynamic.
(Fig-2)
Coastal area of Bangladesh may. broadly, be divided into three distinct geomorphological region; the eastern, the central and the western region. The western region covers the coast line from the Tetulia river upto the international border and lies in the south-western zone of the country. The study area is part of the western coastal region.
According to the agro-ecological zones study of Bangladesh based on physiography, soil and flooding characterstics, the study area falls under Ganges Tidat Floodplain.
(Brammer, 1989)
Soil of the area is predominantly clay.
3.3 Topography
- 91 - Except the extreme south-eastern zone (Eastern Hills); extreme north-eastern zone (Northern Hills) and central part of north-eastern zone (Madhupur Tract), the entire country is almost flat, generally sloping from north to south. The land mass is cris-crossed by numerous rivers/Khals. Floodplain within the channels are saucer shaped; highest elevation prevails along the banks of the channels while the basins possess comparatively lower elevation.
Flat land, elevation ranging between ＋ 0.0m PWD to ＋ 3.0m PWD (ADC, 1986), with low tying depressions locally known as "beel " is prominent feature of the topography of the study area. Homesteads are situated on artificially raised ground and interconnected by earthern road.
3.4 Climate
Bangladesh enjoys tropical monsoon climate with three distinct seasons; the summer(March-April ), the rainy season (June-September) and the winter (November-February). Two transitional period; one from summer to rainy season and the other from rainy season to winter, are also noticable here. The summer is characterized by hot and little rainfall, the rainy season by hot and planty rainfall and the winter by cool and almost no rainfall. Same tropical monsoon climate, as elsewhere experienced in Bangladesh, prevails in study area. Important climatological data of the study area is presented below.
Table-1 Climatological data of the study area.
- 92 - Annual rainfall of the study area is 1900 ㎜, 74％ of the rainfall occurs within the period of June to September and 89％ within the period of May to October. Most striking feature of the climate in the study area is cyclonic storms and tidal surges that originate over the Bay of Bengal and hit the coastal belt causing colosial damage on lives and properties.
(Table-2)
3.5 Hydrology
The Brahmaputra, the Meghna, the Padma/Ganges; the major drainage artery of the Himalayan Drainage Basin, receive descharge from numerous tributaries throughout the course in addition to own flow. The rivers meet together in Bangladesh territory and discharges into the Bay of Bengal as the Meghna. Combined flood flow may exceed 1,900,000 ㎥/sec (Shah,1990) and annual sediment volume of the flow may range within 1.5 to 2.4 billion tons.
(Rahman,1988)
On the other hand, lean period discharge of the rivers is abnormally low; the Padma/Ganges at Harding bridge in Bangladesh recorded only 657 cubic meter per second which is only 0.93％ of the maximum peak discharge of the river at the same station.
(Chowdhury ＆ Khan 1981)
Table-2. Cyclonic storms and tidal surges on western coastal region of Bangladesh.
The enormous discharge and sediment load borne by the rivers is contributed from the watershed of about 1.7 million square kilometer (Fig-3). Bangladesh shares only 7％ of the total catchment area (Table-3).
Table-3 Catchment area of major rivers of Himalayan Drainage Basin.
- 93 -
River flow and sediment load is playing a vital role in developing the coastal area of Bangladesh which is still active and under the process of development.
Tidal influence along the entire coastal belt of Bangladesh is a prominent hydrological feature. Seasonal variation of semi-diurnal tides having an average period of 12 hours 25 minutes is noticed due to monsoon wind, strong surge in the ocean and up land flow. Sea level in the Bay increases about 0.6 metre from March to July due to on set of monsoon wind field.
(ADC, 1986)
The amplitude and phase of the tide differs remarkably at different places along the coast and within the estuarine rivers because of funnel shaped coastal geometry, uneven bottom topography, configuration of the estuarine channels and distribution of up land flow through the channels.
Amplitude of tide in Passur river system ranges between 3 metre to 4 metre while that of in Sibsa river system ranges between 4 metre to 5 metre.
(ADC, 1986)
The study area, located close to the Bay, characterized by flat tideland with numerous intricate system of rivers and creeks and possessing low land elevation, is subjected to periodic flooding. Water level and discharge of the rivers are dependent, mainly, on tides and up land fresh water flow. The Padma/Ganges is the only source of up land flow in the study area. Unilateral withdrawal of the Ganges water at Farakka reduces the dry season flow of the Padma/Ganges in Bangladesh territory. Consequently, the tidal rivers of the study area receive abnormally low up land flow.
- 94 - 4. DEVELOPMENT ACTIVITIES IN THE STUDY AREA
The coastal area suffer crop damage and loss of land fertility due to inundation by saline water and upland flood water. Construction of small dykes around individual land with wooden box sluices to stop intrusion of water and to drain internal rainfall has been practiced since 17th century by Zaminders. Zamindari system was abolished in 7％ of the total catchment area(Table-3).
(Table-3)
River flow and sediment load is playing a vital role in developing the coastal area of Bangladesh which is still active and under the process of development.
Tidal influence along the entire coastal belt of Bangladesh is a prominent hydrological feature. Seasonal variation of semi-diurnal tides having an average Period of 12 hours 25 minutes is noticed due to monsoon wind, strong surge in the ocean and up land flow. Sea level in the Bay increases about 0.6 metre from March to July due to on set of monsoon wind field.
(ADC, 1986)
The amplitude and phase of the tide differs remarkably at different places along the coast and within the estuarine rivers because of funnel shaped coastal geometry, uneven bottom topography, configuration of the estuarine channels and distribution of up land flow through the channels.
Amplitude of tide in Passur river system ranges between 3 metre to 4 metre while that of in Sibsa river system ranges between 4 metre to 5 metre.
(ADC, 1986)
The study area, located close to the Bay, characterized by flat tideland with numerous intricate system of rivers and creeks and possessing low land elevation, is subjected to periodic flooding. Water level and discharge of the rivers are dependent, mainly, on tides and up land fresh water flow. The Padma/Ganges is the only source of up land flow in the study area. Unilateral withdrawal of the Ganges water at Farakka reduces the dry season flow of the Padma/Ganges in Bangladesh territory. Consequently, the tidal rivers of the study area receive abnormally low up land flow.
The coastal area suffer crop damage and loss of land fertility due to inundation by saline water and upland flood water. Construction of small dykes around individual land with wooden box sluices to stop intrusion of water and to drain internal rainfall has been practiced since 17th century by Zaminders. Zamindari system was abolished in 1951 and since then, dykes deteriorated in absence of proper maintenance.
The Government entrusted Water and Power Development Authority (WAPDA) to reconstruct and develope the embankment. The authority started implementation of the Coastal Embankment Project (CEP) in 1961. The CEP program consists of 108 polders (LDL 1968) ; a land mass encircled by earthen embankment with drainage/flushing sluice at suitable location on the embankment for internal drainage/flushing.
(Fig-4)
The project area, about 1.3 million hectares of land stretching along a belt of about 530kilometer long
- 95 - and a few hundred meter to 80 kilometer in width is close to the Bay of Bengal in the estuarine and coastal area of Bangladesh.
(LDL, 1968)
The project includes about 7700 Kilometer of embankment and 900 sluices (MPO. 1984), 39 polders.
(Fig-5) were completed between 1965 to 1971 with 1744 kilometer of embankment, 391 drainage structures, 44 flushing structures and 16 drainage-cum-flushing sluices to benefit 409757 hectares of land under CEP programme in greater Khulna and Jessore district with an investment cost of about 4790.68 lac Taka.
(Rahman, 1990)
5. PRESENT SITUATION IN THE STUDY AREA
Impact of CEP on increased crop production was tremendous. The project contributed much to the "Green Revolution" of the sixties and seventies. The embankment also acts as a barrier against cyclonic storm and tidal surge reducing the damage.
However, some adverse environmental effects started appearing in the CEP area. Siltation of outfall channels and tidal rivers is a common problem; about 626 kilometer of channels are facing siltation in south-west zone of Bangladesh as per 1984 study (MPO. 1986). Loss of agricultural production in the study area was reported by the early 1980's due to drainage congestion (CKC,1991). Out of 409,757 hectares of study area; 31900 hectares under polder-25, 27/1, 27/2, 28/1 and 28/2 are affected badly by drainage congestion (Fig-6). About 10000 to 15000 hectares of land in 5 polders is facing water logging as in November, 1989 (CKC,1991). Among the polders, Folder-25 is facing the worst situation; large areas of the polder, specifically the beel Dakatia, remains permanently water logged throughout the year. Persistant water logging in beel Dakatia, especially during monsoon, compels the inhabitants to lead a subhuman life.
(Shahjahan ＆ Hossain, 1991)
The area of 5 polders is bounded by Hari/sree river on the west ＆ north-west, by railway line on the east, by Haria and Gallamari river on the south-east and by Bhadra, upper Salta, lower Solmari river on the south. The Hamkura river divides polder-27 and polder-28. Prior to construction of poIders, the area drained through internal rivers (Hamkura and upper Solmari ) to peripheral rivers (Sree/Harie, Bhadra, upper Salta, lower Solmari, Haria, Gallamare) and then outfall rivers (Kazibacha, Bhadra, lower Salta). The outfall channels are connected to two major tidal river system of the area; the Sibsa and the Passur.
(Fig-7)
- 96 -
; both of which has common outfall at Jefford point.
The Passur and the Sibsa are still active but the internal and peripheral rivers of the 5 polders are facing gradual siltation. The situation is so severe that some of the channels are silted up to high water level blocking internal polder drainage. At present, river beds of the Hamkura and upper Solmari are 0.60 to 1.2 meter higher than average ground level within the polder.
(Shahajan ＆ Hossain, 1991)
Attempts were made to keep the Hamkura river active by re-excavating a length of about 1.50 kilometer in 1982-83 but a depth of about 2.70 meter was deposited in the same place in next season.
(ADC, 1986)
Flow section of outfall channels of drainage sluices of 5 polders has been reduced to 53％ of the original section.
(ADC, 1986)
The depth of deep channel of upper Solmari river near junction is only 2.10 meter compared to 9.70 meter depth of lower Solmare river (ADC, 1986).
Siltation of outfall channels has created the polder drainage structures ineffective; DS-1 and DS-2 of polder-25 is completely blocked; DS-3 of the same polder is almost not functioning, DS-4, DS-5, D-11 to DS-13 of polder-25, DS-1, DS-5 and DS-6 of polder-27/1 and DS-1 to DS-5 of polder-28/1 are presently under the threat of siliation in near future.
(ADC, 1986)
Bhadra-Teligati junction was silted up except for a small high tide overflow channel in March, 1990
(CKC, 1991)
Fig-8 illustrates cross-section of Bhadra river near Domuria indicating chocked up situation.
Morirchap river in between polder-1 and polder-2 is also facing siltation as presented in Fig-9.
(Khan, 1985)
6. PROBABLE CAUSES OF SILTATION IN STUDY AREA
- 97 - Erosion and siltation is a natural process. Man-made interferance affects the natural process adversly causing ecological imbalance. Drainage congestion in study area may be attributed to the following causes; a) Embanking the channel.
Natural channels confined by embankment faces gradual siltation reducing its cross-sectional area in the form of raised river bed or siltation on the side slopes/berms. Khowai river in north-eastern zone of Bangladesh is an example of siltation due to confinement.
(Fig-9)
In CEP, the river is confined by embankment on both banks. Obviously, siltation is there. b) Reduction of upland flow.
Channel form of an alluvial river is controlled by numerous interrelated variables such as discharge, sediment load, longitudinal slope, bank and bed resistance to flow, vegetation etc. Prediction of response of alluvial tidal channel form due to changed situation is a very complex task. Response of channel form due to changing discharge and sediment load was studied by Lane (1955), Leopold and Moddack(1953), Khan (1971), Schumm(1971) and Santos ＆ Simons(1972) which indicates that depth of flow is dircetly proportional to water discharge and inversely proportional to sediment discharge while width of flow is directly proportional to both water discharge and sediment discharge (Khan, 1985). Unilateral with drawl of the Ganges water at Farakka deprives the tidal creeks of study area from upland flow. Response of upper reaches of the tidal channel due to reduced upland flow appears in the form of siltation (reduced flow depth and flow width). The process (siltation) propagates downstream which is a natural consequences. c) Closing the natural creeks/rivers.
0'Brien (1969) showed that minimum cross-sectional area of the entrance of tidal creek is linearly related to tidal prism. Eysink(1981) found the same relationship using the channel cross-sectional area below mean water level. Barua and Koch(1986) established the same relationship using the channel cross-sectional area below mean water level for Meghna estuary.
(Barua, 1990)
As a result of empoldering, tidal prism passing through the creeks is reduced significantly. Volume of tidal prism entering in polder-1 was 92.90 million cubic meter and 7.90 million cubic meter before and after the project respectively.
(MPO,1986)
Consequently, cross-sectional area of the creek is adjusted to cope with reduced tidal prism by siltation.
(Fig-9) d) Increased salinity in the creeks.
- 98 -
Upland water flow from the Padma/Ganges influences the salinity level of river water in study area.
(ADC,1986)
Diversion of the Ganges water at Farakka reduces the upland flow and increases the water salinity in outfall channels; the situation helps rapid sedimentation by coagulation and flocculation. e) Tide meeting point within the creeks.
The study area is interlaced with intricate system of tidal rivers and creeks. Empolderment of tidal creeks coupled with abnormal reduction of upland fresh water flow during dry season has changed the complex flow phenomena within the interconnected tidal rivers and developed some tide meeting points that faces siltation. Upper Salta river connecting Bhadra river and lower Salta river (Fig-7) facing gradual siltation due to tide meeting point in the river.
(ADC,1986) f) Subsidence of polder soil.
Natural subsidence of polder soil in the study area may help sinking the land but the process is not well documented. Subsidence will adversely affect the drainage system of the polder. g) Blocking the land formation process.
Tidal prism entering into the polder deposited sediment on the low depressions and helped in raising the land. Natural land formation process within the polder is halted by empowerment while the outfall channel bed is raised by siltation causing detrimental effect on polder drainage. h) Natural process of erosion and sedimentation.
Natrual channels are always in dynamic state. Channel forms of tidal creeks carrying cohesive and non-cohesive sediment load with the discharge are under constant change by erosion and sedimentation, a process very difficult to understand and explain.
The Koyra river between polder-13 and polder-l4/2 was silted up significantly though the channel is not embanked.
(LDL, 1968)
Erosion occurs on bed and banks of lower SoImari river, Chunkuri river and Mango river.
(ADC, 1986) i) Selection of empoldering level.
Selection of the height of land mass, which is proposed for empoldering, is a very important issue. A low height will face drainage congestion under gravity system
- 99 -
(Fig-10) while a high level may need long time to wait. Selection of empoldering level for the polders under CEP is not a well documented feature.
7. CONCLUSION
Natural process of siltation in the estuarine area can not be checked. However, development activities on the intricate natural eco-system; comprising of wondrous coastal belt with magnificent mangrove forest, a unique blessing of nature; should be prepared on the basis of comprehensive and integrated study and planning for the entire south-west zone of Bangladesh.
- 100 -
FIG - 1
- 101 -
FIG - 2
- 102 -
FIG - 3
FIG - 4
- 103 -
FIG - 5
- 104 -
FIG - 6
FIG - 7
- 105 -
FIG - 8
- 106 -
FIG - 9
FIG - 10
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF
- 107 - INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
THE EFFECTS OF UPSTREAM DAMS CONSTRUCTION AND TIDES TO THE DYNAMICS OF NATURAL PROCESSES IN THE TIDAL REACHES : EXAMPLES FROM THREE MAIN RIVERS IN JAVA. INDONESIA
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Research and Development Centre for 참여연구원 : Oceanotoar (CORD)-LIPI, Jakarta, Indonesia Otto S. R. Ongkosongo
Summary
Studies have been carried out in the tidal reaches of three main rivers in Java, Indonesia i. e. Citarum, Citanduy and Solo which have different types of upstream land and water management works.
As the most controlled, regulated or managed river in Java, Citarum has been dammed with electric generating plants in three sites i. e. Jatiluhur (1972), Cirata(1985) and Saguling (1987). The downstream waters have been regulated and diversified into vast irrigation canals for agriculture and drinking water supply for Jakarta City, in addition to the natural existing system with eight small to medium size distributaries forming a unique multiforms delta complex. Meanwhile exploitation of sand exists along river and the formed coastal sand bars. As results, during the last decades, most shorelines of the drainage area have been severely eroding and three distributaries have been abandoned to be filled up by fine grain sediments. On the other hand, during low tides, formation of river mouth sand bars always hamper the principal sea-land navigation connection the drainage
- 108 - routes with Jakarta. Dredging and straightening of river mouth in 1982 has not functioned well after three years of operation and the deeper channel has also been filled up by new sediments. Many parts of the river banks in the heavy river traffic route are significantly undercaved by boat-generated waves to such that river diversion prominently might occur any time.
The effects of tides to the dispersion of sediments, formations of turbidity maximum and salinity wedge in the three rivers are discussed. Unlike the two other rivers, the Citanduy which debouches in Segara Anakan coastal lagoon before entering the open deep waters of Indian Ocean, is almost untouched by salty marine water. Only the lagoon water mass is influenced causing diversion and retreat of sediments from the river to be partly deposited in its calmer environment. The 6,675 ㏊ lagoon water area in 1900 is estimated to be intensely filled by sediments remaining narrow to large tidal creeks in the end of this century.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
HYCROLOGY AND OCEANOGRAPH OF GOLOK RIVER
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Harbor Department, Bangkok (10100), Thailand 참여연구원 : Pramot Sojisuporn
- 109 -
Summary
Golok River is a drown-river-valley type estuary located between the border of Thailand and Malaysia. The bay is about 300 m wide at the mouth, is 2-3 m deep on the average, and experiences mixed tide. The river mouth is influenced by tidal forcing, river discharge, and monsoonal wind and wave. The river mouth experiences annual shallowness for a certain months after the northeast monsoon season due to low river discharge and northward sand drift along the coast. The building of sand in front of the river mouth poses difficulty in boat transportations in and out the river, and difficulty in defining international boundary. The purpose of this project is to investigate hydrology and oceanography of Golok River mouth for better understanding of the physical processes in the area. The study included depth sounding, current, salinity, temperature, and suspended solid measurements in the coastal area and in a river cross section during 24 hour periods, water level recording, wave measurement, and sand trap study. The data analysis to bring out salient characteristics of Golok River includes river discharge computation, horizontal and cross-sectional distributions of velocities and salinity, rate of water mixing, wave characteristics, and estimation of sand movement.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
SEDIMENT TRANSPORT PATTERN OF ALLUVIAL RIVERS IN KOREA
1991
- 110 -
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Deputy Director, River Planning Division, Water 참여연구원 : Resources Bureau, Ministry of Construction, Republic of Korea Sang-Tae Lee Professor and Head, Department of Civil 참여연구원 : Engineering, Yeungnam University, Republic of Korea Soontak Lee
Summary
Problems associated with sedimentation of alluvial rivers in Korea are summarized in this paper. First, hydrological and geographycal characteristics of Korea are briefly introduced. Then, typical problems of channel deposits associated with alluvial sedimentation in Korea are described. The characteristics of channel deposits in one of the typical alluvial rivers in Korea are also given from the measurement and analyses of sediment load at the river channels and estuary.
INTRODUCTION
Sediment transport in Korean rivers mainly occurrs due to soil erosion and the variation of alluvial river beds during floods. Most Korean river basins consist of mountainous areas with steep slopes upstream, and plains with mild slopes midstream and downstream. Therefore, continuous sediment transport in midstream and downstream areas brings about great changes in river beds every year.
In this respect, sedimentation problems aroused much interest in river management(Lee, et. al. 1982- 1988), and many studies have been carried out, centering around the rivers in Korea, about soil erosion on the slopes(Cho and Park, 1981), estimation of sediment transport (Ahn, 1980 and Kim, et. al. 1981 and 1982), the changes of river beds (Nahm, 1978), and the changes in sediment transport in rivers brought about by the barrage construction in estuary.
(ISWACO, 1982)
(Lee, 1976)
- 111 -
This paper introduces the characteristic of transport patterns of sediment load measured in the main- stream, tributaries and estuary of the Nakdong river which is a typical alluvial river in Korea.
HYDROLOGICAL AND GEOGRAPHYCAL CHARACTERISTICS
Korea belong to the moderately humid zone in Northeast Asia. It has a definite, seasonal climate which is greatly defined by dry, cold continental air masses in the winter and humid. warm air masses from the ocean in the summer. This climate influenced the development of the recent geomorphological features. Water is the main erosive force in Korea.
The yearly distribution of the precipitation is determined by westerly and north-westerly dry winds from the Asian continent in winter and north-easterly to south-easterly winds from the Pacific Ocean in summer. Thus the rainfall is concentrated in summer. In the rainy season (June to September). 66％ of the yearly rains (1,200 ㎜) occur. A precipitation of 16％ falls in the transition period during April and May, and 18％ of the precipitation falls in the remaining six months of the year(October to March).
The Nakdong river basin taken as an alluvial river basin is located, as is shown in Fig. 1, in the southeastern region of the Korean peninsula, and the river basin includes an area of 23,656 ㎢, a mean elevation of 284.9 m, and a length of 525.8 ㎞. The channel slope is very mild(0.00024). The soil structure in the upstream basin is of solid nature, while the midstream and dawnstream basins have comparatively soft deposits (In the utilization of land, the basin is made up of 20.3％ arable land, 69.9％ forest area, and 9.8％ urbanized area).
As for the channel characteristics of the basin, it has a rectangular form and tributaries flowing into the mainstream are short and have steep slopes while the mainstream has a very mild slope.
- 112 -
Fig. 1 Nakdong River Basin Especially the channel slope in the estuary is very mild and the river has a considerable width there. In point of hydrologic characteristics of basin, mean annual rainfall amounts to 1,100 ㎜. and its distribution is concentrated in a period of four months, that is, from June to September, during which four or five severe storms occur annually. The rest of the year is considered the dry season and has little rainfall. Thus, there are four or five flood periods every year and sea water backwashes into the river estuary during the dry season.
As a result, a barrage has been constructed in order to prevent the inflow of salty water and to secure a fresh water supply in the estuary area.
SEDIMENT TRANSPORT PATTERN
In Korea, there are numerous problems in relation with sedimentation of natural sediment among which channel deposits in alluvial rivers are only considered and those transport patterns are discussed in this paper. The main problems associated with channel deposits are river bed changes and meandering after large floods. These problems cause change in the bed-elevation and in the waters course downstream.
In the following sections, the characteristics of channel deposits in one of the typical rivers in Korea are given from the measurement and analyses of sediment load at the river channels and estuary.
1. Sediment Transport in the Channel
Three gauging stations along the mainstream(Sinsang, Hyunpung and Jindong) and two streams in tributaries (Wichun and Milyang-gang) were selected to measure sediment load and river discharge
- 113 - for analyses. Samples of suspended load and bed load were taken during the flood as well as non- flood seasons at the gauging stations where accurate stage records and rating curves were available, and analyses were made by measuring the concentration of the samples and estimating the amount of sediment load. Thus the relationships between sediment load and river discharge were obtained. The relationship between suspended load(Q*) and river discharge(Q) is shown in Fig. 2 (a).
Fig. 2 (a) Suspended Load-Discharge Relationship
FIg. 2 (b) Bed Load-Discharge Relationship These results are compared with theoretical values estimated by Einstein and Lane-Kalinske formulae in the same figure. In general, it was found that the measured values of the suspended load showed similar patterns to the theoretical values. But it is noteworthy that the values estimated by the Lane- Kalinske method are greater than other results.
Next, the relationship bed load(Q_{B}) and river discharge(Q) is show in Fig.2(b) in which the characteristics of bed load transport can be found. The characteristics of totat sediment load in the river basin were also analyzed from the relationships of river discharge to total sediment load and to the ratio of suspended load to total sediment load as shown in Fig.3(a) and Fig.3(b). These results show that most of total sediment load is composed of suspended load which is over 90％ of the total sediment load.
- 114 -
Fig. 3 (a) Total Sediment Load-Discharge Relationship
Fig. 3 (b) Q_{S}/Q_{T}-Q Relationship
2. Sediment Transpot in the Estuary
Estuary Characteristics
The estuary, located in the southeastern part of the Korean peninsula, drains the water from the Nakdong river to the sea. The tide generally penetrates approximately 60 ㎞ upstream of Handan up to Susan, whereas the tip of the salt wedge may penetrate some 30 ㎞ upstream of Hadan.
Approximately 3 ㎞ upstream of Handan the estuary is divided into two main channels by islands
- 115 - with a length of about 6 ㎞. Both channels are separately connected with the sea through the eastern part of a 3 ㎞ wide tidal flat area entering the sea through the tidal inlets.
(Fig. 4)
The bottom of the estuary consists of sand, silt and clay particles. The river bottom from Jindong (about 80 ㎞ upstream from the river mouth) down to the tidal flat area mainly consists of sand with hardly any silt and clay. The water movement in the estuary is determined by a number of phenomena, i. e. ; tide, river discharge, density currents, drift currents and waves. Upstream past the salinity penetration limit, the first two phenomena dominate, whereas in the tidal flat area all phenomena play an important role with respect to the motion of water and the sediment transports in the area. The Semi -diurnal tide penetrates into the tidal flat area and the Nakdong estuary through the tidal inlets, the main channels, and the tidal creeks intersecting the tidal flats. The flow velocities in the tidal inlets and the main channels to the estuary may go up to 0.75 m/sec during flood and 1.25 m/sec during ebb(spring tide) in the dry season. In the wet season the maximum ebb velocities will incerase whereas the flood velocitics will decrease to some extent due to the increased river discharges.
Fig. 4 Nakdong estuary
Sediment Transport
The water masses, which constantly are in motion in the Nakdong estuary and in the tidal flat area, continuously stir up and transport large amounts of sediments which are moved back and forth with the tide. On top of this the Nakdong river regularly supplies an amount of erosion material from its catchment area to the estuary and the tidal flat area where part of it will settle contributing to the gradual expansion of the delta and part of it will be flushed to the sea, particularly the fine sediments.
Quantitative information on sediment transports in the lower estuary has been collected during mean tide and spring tide.
(ISWACO, 1982)
Table 1 shows some data on the suspended sediment contents at the Hadan station in the Nakdong estuary in the dry season under mean tide conditions. Examination of sediment data showed that the
- 116 - sediment contents fluctuated over the tide and showed higher mean concentration levels under spring tide conditions than under mean tide conditions.
Table 1 Suspended Sediment Contents(mg/1) in the Nakdong Estuary(Hadan Station)
The measured sediment transports(T, ㎥/m day ) at Jindong station have been plotted against river discharges(Q, ㎥/sec) as is showen in Fig.5, where the average width of the cross sectional profile(A/h), approximate sediment contents(C, ㎏/l) and the sediment transport rate(T_{t}, ㎏/sec) also have been included. This figure illustrates that the average level of the sediment contents in the estuary and tidal flat area will depend on the river discharge. From the results of the sediment transport measurements and the data on sediment transports by the river it can be found that less than 5％ of the mean annual sediment supply by the river is transported by river discharges up to 400 ㎥/sec. The majority (85 to 90％) of this supply is transported by river discharges exceeding 800 ㎥/sec which, on the average, occurs only during 10％ of the time in a year, particularly in the wet season. From the examination of sediment transports it is derived that total sediment supply from the Nakdong river would amount to about 0.8 million ㎥ of sediment per year whereas the mean annual mud transport by the river is estimated at some 0.5 million tons.
Next, the sediment transports per unit of area(T, g/㎡. sec) measured during dry season in the lower estuary also have been plotted against the instantaneous current velocity(V, m/sec) as is showen in Fig. 6.
From this figure it is believed that the scatter in data ranging 0.57 to 2T is due to scatter in sediment contents of water sampled and possibly by hysteresis phenomena due to scour lag and sedimentation lag effect.
Fig. 5 Sediment Transport and Contents at Jindong
- 117 -
Fig. 6 Sediment Transport Rates in the Nakdong Estuary Related to Current Velocities
Impacts of the Barrage
As there is no significant change in river regime due to the barrage construction in the Nakdong estuary, the sediment supply and the bed shaping discharge rate of the river is not being affected. However, when river discharge flows into the river basin which was former tidal area, the cross section of the river increases and consequently the flow velocity reduces. Since the sediment transport capacity is proportional with the flow velocity to the power 3 as shown in fig. 6, this results in a rapid decrease of the sediment transports in the river basin. This basin, therefore. acts as an important trap for the sediments carried by the river, the coarser fractions being trapped first and the finer fractions more downstream.
After the construction of the barrage the tidal volumes in the tidal area were strongly reduced whereas the dimensions of the main channels were maintained. Hence, the current velocities in these channels are distinctly less than at the previous time. Consequently, the related sediment transport capacities were also reduced and the past (dynamic) equilibrium conditions are being disturbed.
The material for the sedimentation of the main channels in the tidal area are supplied from three sources; (a) the Nakdong river passing the barrage. (b) the outer deltas of the tidal inlets and the coastal area in front of the barrier islands, and (c) the tidal flats in the lower estuary.
CONCLUSIONS
Problems associated with sedimentation of alluvial rivers in Korea have been examined from the analyses of sediment load at the river channels and estuary of the Nakdong river.
It was found from these analyses that most of the total sediment load in the mainstream and tributaries consist of the suspended load which is over 90％ of total sediment load, and the total sediment supply from the Nakdong river would ammount to about 0.8 million ㎥ of sediment per year whereas the mean annual mud transport by the river is estimated at some 0.5 million tons.
In the lower estuary, the sediment transport capacity is propotional with the flow velocity to the power
- 118 - 3, and the related sediment transport capacities are being reduced with the disturbance of the past equilibrium conditions after the construction of the barrage.
In general, it can be concluded that most of alluvial rivers in Korea have sedimentation problems in downstream areas such as channel deposits and changes of water course.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
THE PROBLEMS OF FLOOD COTROL IN CHINA
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Senior Engineer and Professor, General Institute of Planning and Design in Water Resources and 참여연구원 : Hydropower, Ministry of Energy, Ministry of Water Resource, China Xiao Qiu He
Summary
- 119 -
The history of China over thousands of years and especially the experience of new China gained in practice in the past forty two years have proved that the national lands conservation and development program is of great significance to the development of China's national economy. Since the founding of the new China in 1949, a tremendous amount of river engineering work has been carried out on numerous rivers, particularly on the seven major rivers namely, the Yangtze, the Yellow, the Huai, the Hai, the Pearl, the Liao and the Songhua. All these works have contributed significantly to various aspects of the national economy. As river harnessing is an important constituent part of the national lands conservation and development program, it must meet some definite demands posed by the program. In addition, the work of developing and controlling big rivers is extremely complicated and difficult. Therefore, great care must be exercised in dealing with flood, water-logging, drought, alkalinization, siltation and pollution on the one hand, and with problems of irrigation, drainage, water supply, navigation, power generation, fishery and ecological conditions on the other. An outstanding difficulty encountered in regulating the Chinese rivers is the severe losses of water and soil from the watersheds of the big rivers, giving rise to heavy sediment load in the rivers. This complicates the river control work and causes a lot of troubles and even calamities.
This paper focuses mainly at discussing the specific problems of flood control in China. In which, the salient features and the present status of flood control are described. Some experiences and lessons learned in the past forty two years are also briefly summaried.
1. FLOOD CONTROL - A MAJOR ISSUE IN NATIONAL LANDS CONSERVATION AND DEVELOPMENT PROGRAM
Flood control being a major issue in China's national lands conservation and development program is determined by the specific features and social and historical conditions of the country.
A main feature of this country is the extremely uneven distribution of population as a result of the disproportionate combination of land and water resources. The annual precipitation decreases from more than 1,600 ㎜ in the southeastern coastal region to less than 200 ㎜ in the northwestern hinterland. The annual isohyet of 400 ㎜ forms a diagonal line crossing the mainland obliquely from northeast to southwest. To the northwest of this diagonal line, the regions are in shortage of water and sparse in population, but to the southeast of this line, the regions on the alluvial plains as well as those in the basins among mountains are densely populated and often subjected to the threat of flood.
On the middle and lower reaches of the seven major rivers, covering an area of about one tenth of the territory, are located more than half of the country's population and many important cities like Harbin, Shenyang, Beijing, Tianjin, Shijiazhuang, Jinan, Wuhan, Changsha, Nanchang, Hefei, Nanjing, Shangai, Zhengzhou, Hangzhou, Guangzhou, Nanning, ect. These are the political , economical and cultural centres of the country. However, the ground levels in these regions are generally lower than the flood stages of the rivers. Safety of these regions is ensured by dikes that have a total length of 200,000 ㎞. This poses a serious flood control problem specific in China.
- 120 - The disproportionate combination of land and water resources results in a shortage of arable land in China. The per capita cultivated area of about 1.5 mu(equivalent to 0.1 acre) in China is much less than that of 4.6 mu in the world. Inadequate land for cultivation will surely place a higher demand to the development of land and water resources. The Chinese agriculture has much more urgent demand on flood control and on irrigation and drainage than other country no only because of the uneven distribution of rainfall but also because of the need for higher specific yield. As a result of those, there had been a trend of excessive use of land and water resources at the expense of shrinking the forests, pasture, rivers and lakes. Such expansion, in return, aggravates the flood problem, thus forming a vicious circle.
Records indicated that the annual precipitation in many places varied greatly from year to year. In some years, it might be just right. In others, there might be severe floods or droughts. In the great alluvial plains, particularly in the flood-plains, the soil is so fertile that successive years of favourable weather would bring to the areas Prosperous economy, stable society and happy people. However, an extraordinary flood once occurred would kill numerous people, destroy everything and thus cause turmoil of the society. In addition, political factors also interacted with economic factors. Therefore, many rulers in the past dynasties considered flood control and river harnessing as important measures to stabilize their states.
What stated above shows wht flood control has long been a major issue in Chinese history and will continue to be a long-term major issue in the future.
2. PRESENT STATUS OF FLOOD CONTROL
After forty two years of hard work throughout the nation, 200,000 ㎞ of dikes of various kind have been built or renovated; 100,000 ㎞ of river channels have been improved; necessary flood-relief channels have been built to drain the Hai and the Huai River Basins. Altogether about 86,570 reservoirs of different sizes as well as 6.2 million storage ponds with total storage capacity of 450 billion ㎥ have been completed. In order to deal with excessive flood flow beyond the discharge capacity of some majors rivers, more than 100 flood diversion and detention basins with a total storage capacity of 120 billion ㎥ have also been constructed.
By constructing all these works, the degree of flood protection on middle and lower reaches of seven major rivers has been raised to a certain level. Benefits achieved are mainly as follows :
(1) With the help of numerous storage projects of various sizes, flows in rivers are partially under control , thus creating favourable conditions for flood prvention. Out of the total reservoir storage capacity of 450 billion ㎥, the large and medium size reservoirs on the seven major river basins account for 275 billion ㎥. The extent of control is different for different rivers. The gross storage capacity of all the large and medium size reservoirs on a river as a percentage of the annual runoff of the river is 9.6％ for the Yangtze River; 78％ for the Yellow River; 56％ for the Huai River; 85％ for the Hai River; 11％ for the Pearl River; 89％ for the Liao River; 24％ for the Songhau River.
(2) Through combined use of all above-mentioned engineering works, the vast plains in the middle and lower reaches of rivers are now safe from common floods. The damages of floods of rare
- 121 - occurrence can also be greatly reduced.
ㆍ The Yangtze River: The dikes along the middle and lower reaches of the main stream and lake areas can stand a 10-year to 20-year flood. If a big flood like the one in 1954 (corresponding to a 40- year flood) should happen again, the safety of the Jingjiang Dike (a most important dike along the middle reach) and Wuhan city can also be ensured.
ㆍ The Yellow River: The existing river control works can accommodate a flood as large as that occurring at Huayuankou in 1958. This flood with a maximum peak discharge of 22,000 ㎥/sec (corresponding to 60-year flood) was the biggest at Huayuankou since 1949.
ㆍ The Huai River: The middle reach of the main stream can safely discharge a flood as that of 1954 which was the largest since 1949 (corresponding to a 40-year flood). The main tributaries (The Yi, Shu and Si Rivers) can stand a flood varying from a 10-year to 20-year flood.
ㆍ The Hai River: The river system to the north can safely handle the flood of 1939 and the river system to the south the flood of 1963. Both floods correspond to a 50-year flood.
ㆍ The Pearl River: The main stream and the tributaries as well as the delta can accommodate 10-to 20-year floods .
ㆍ The Liao River: The dikes along the main stream and tributaries can stand a 10-year to 20-year flood, and cities of Shengyang, Fushun and Liaoyang are safe from a 100-year flood.
ㆍ The Songhua River: The farmland is safe from 10-year to 20-year floods. The cities of Harbin, Qiqihar and Jiamusi can stand a 100-year flood.
3. SOME EXPERIENCES AND LESSONS LEARNED IN PRACTICE IN CHINA
In reviewing the work done during the past 42 years, it is recognized that although much has been achieved in flood control in China but not without paying dearly. All practices no matter they are successful or not make us understand what should be done, and what should be particularly stressed, so that optimum results of disaster defence and derivation of benefits could be achieved at a least possible cost. Our main experiences and lessons gathered from those are briefly summed up as follows.
(1) There should be an overall planning for flood control construction with national lands conservation and development incorporated in the program.
ㆍ It is important to note that a fair guilding ideology for flood control in the overall program should be stressed both on refoming the nature and to daopt one's work to the nature. Everything done should
- 122 - be in line with the laws of nature and economics. Some silly things such as to destory forests and vegetations in mountainous areas, to reclaim indiscreetly the flood plains along the rivers, to construct wilfully polders along the lakes as well as to develop municipal and industrial infrastructures without taking into account the local condition of flood prevention would result in reaping half with twice the effort and not the other around as it should have been.
ㆍ It is observed also that only through such an overall planning can the flood control development be closely linked with the construction works in other field, and better coordinated with the country's social -economic developments.
ㆍ In this regard, due attention should be paid to select an adequate degree of protection for different protected regions in order to make an overall program practicable, affordable and justifiable. To set an improperly high degree of protection would always cost too much, however to set a too low degree of protection would always not commensurate with the socio-economic development in certain regions, According to China's experience, following criteria is recommended.
(2) Flood prevention projects should be integrated with a coherent package of measures and works, otherwise, the expected benefits will not accrue. Experiences show that the construction of flood prevention works will on their own bring about the benefits that justify the exercise. However, this usually is not the case in certain areas where agriculture is the dominant activity. For that case, program can only be successful provided an integrated approach is taken, which ensures a well combination of flood prevention works with associated regional socio-economic development.
In such an approach, both physical and non-physical dimensions are required.
ㆍ The physical dimension relates primarily to the infrastructures, comprising irrigation canals, drainage canals, sub-embankments, navigation and fishery facilities or even rural roads.
ㆍ The non-physical dimension relates to assistance to the farmers in adopting their practices to the changed conditions, thus helping them to fully exploit the land potential.
(3) Various structural flood prevention measures including reservoirs, flood diversion and detention basins, and embankments should be so selected as to strike a happy medium between the storage and discharge of flood on the upper, middle and lower reaches in accordance with the characteristics of the river basin.
ㆍ Upstream storage reservoirs are often used to impound part of the peak flow for later release, but such reservoirs will not normally be built for flood control alone. Their effect ofn flood control will also be constrained due to several other factors, such as the capacity of the reservoir, the location of the
- 123 - reservoir and the origination of the flood runoff.
Several other storage reservoirs upstream from the Three Gorges Project in the Yangtze River had been studied to compare with that of the Three Gorges Project. Because of the above factors, it is recognized that although the effective storage of those planned sites is comparable to that of the Three Gorges Project, they can't provide comparable flood control .
ㆍ Flood diversion and detention basins are often used also to divert the excess flood flow in China in order to prevent some important levee sections from bursting.
Since there are a lot of people living in the detention regions, such measure should be handled with great care. All planned projects or destined depressions can only be used after getting approval from higher authority. For the sake of safety of inhabitants living there, necessary platform or multi-storied building for refugee should be raised or constructed. The communication and warning system within the region should be established in advance. Meanwhile, other flood fighting and life-saving equipments should be arranged. In order to make use of these measures smoothly, now a strict control of population growth and land exploitation and utilization in those regions has also been put into practice in China.
ㆍ To construct embankments to keep high water level out has been applied elsewhere as well as in China for a long time. They have proved to he effective during most of the floods. Nevertheless, embankments have their limitations and drawbacks.
One of the most important impacts of embankment is the morphological changes of the river after their construction. They can lead to siltation in one place and increased scour in another; the overall effect of which could be increased local bank-cutting. Therefore, river training works should be carried out in line with the construction of embankment.
(4) The objective of river trainning in alluvial rivers is aimed at reducing the intensity and shifting scope of plane deformation so as to stabilize the plane morphology of the channel in phases through node control and necessary revetment. The study always includes; analysis of river pattern, analysis of fluvial process, selection of design elements and layout of river training works.
Major experiences related to training of the Yellow River and the Yangtze River can be summaried as follows:
ㆍ Node points control is a key measure for river training. The main function of which is to control the incoming flow from upstream, regulate and smooth out the flow passage downstream and control the shifting scope of bifurcated reaches. Nodes formed by fixed boundary due to strong anti-erosion ability of bank materials are popular in the Yangtze River, while nodes formed by man-made protection works or structures built on natural narrow sections are popular in the Yellow River.
ㆍ Guiding flow by bends usually protected by a group of spur dikes or groynes is a powerful measure for river pattern control . Their effectiveness has been proved in practice in the Yellow River. Since bends would be changed consequently in sequence if one bend changes, therefore training works should be started from the controlling point at the upstream and proceeded towards downstream.
ㆍ River which has no natural nodes is always trained in three stages. The first stage is aimed at protecting some important or vulnerable spots and controlling flood flow. The sceond stage is to adjust the flow passage, control the medium flow channel and regulate the river pattern. The last stage is to improve and consolidate the training works to establish a basically stable channel.
- 124 -
ㆍ To form a basic stable channel, engineering works should be accomplished in a certain length. In the lower reach of the Yellow River, even though 590 ㎞ of engineering works making up about 70％ of the length of entire reach has been built, it seems still not enough to make the medium flow channel stable. It needs to be emphasized that the planning of engineering works is not stagnant but an iterative process requiring revision as more insight is gained over time through tests and observations as the work proceeds.
ㆍ River training works designed according to dominant discharge can not only control the medium flow channel, but also stabilize the river pattern basically.
ㆍ The orientation of training works for different rivers is to mold the river into different pattern according to their nature. For the Yellow River, it was considered to control by bends and mold the river into single meandering one, while for the Yangtze River, it was considered to stabilize the mid- islands and to make the ratio of flow and sediment distribution stable. thus mold the river into a bifurcated one.
(5) For rivers with heavey silt load, flood control problem needs to be dealt integrately with both flood and silt.
ㆍ The Sanmenxia Reservoir in the Yellow River is one of the typical examples showing that a dam constructed on the river with very high silt content would not be problematic, if enough lower outlet works were provided, and if a flood control level of the reservoir during flood season were appropriately adopted. Since the completion of certain reconstruction works in late 70s, this reservior has basically remained in a silt-stable stage, keeping a storage capacity normally used for flood control and other purposes as expected.
ㆍ Although silt problem has long been taken as an origin of flood disaster in the Yellow River, yet there is no denying the fact that silt is also a kind of resources beneficial to flood control and improvement of saline land. During the last ten some years, silt has been widely used to widen and heighten the levees along the Yellow River through warping. After strengthening with silt, formerly weak levee sections have now become dependable. After raising land elevation along the countryside of levee, the salinization problem has also been solved.
(6) Non-structural measure including flood forecasting, flood warning and flood preparedness is of considerable significance both in unprotected flood prone areas as well as in the areas already protected for fighting against flood.
ㆍ There are more than 8,500 hydrological stations in China at present. A nationwide information- collecting network can basically satisfy the needs of flood control. With the spreading of computer technique, the receiving and processing system of real-time hydrological information with VAX-Ⅱ series computer or IBM-PC microcomputers as the processor have been developed and spread over the whole country in recent years. Meanwhile, the hydrological forecasting techniques have been developed greatly also. All these have played a very important role in flood fighting. In the light of a flood forecasting in 1975, the dam height of the Boshan Reservoir (earth dam with a total capacity of 0.62 billion ㎥) in Henan province was increased in time so as to prevent overtopping. Due to timely flood forecast in 1981 in upstream of the Yangtze River, 220,000 residents, many factories and government departments along the river had been transferred to safe areas prior to the coming of flood. thus made loss of damage minimal.
- 125 - ㆍ To set up an unblocked and reliable communication network is of great importance in flood warning system. It is worth mentioning that due to inefficiency of the warning system the economic losses in the Haihe River in 1963 and in the Huaihe River in 1975had become much more serious than those might be suffered from.
Since warning system should be so developed as to elicit at the village level, the meteorological transmitter at county level with a receiver in each village has been set up at present. This has been proved to be quite successful enjoyed by the masses.
ㆍ Flood preparedness activities comprising early warning, evacuation, flood fighting and organizing emergancy responses should be community-based and effectively coordinated among various agencies concerned.
Based on experiences in China in this respect, some specific measures besides others seem noteworthy.
- an evacuating and sheltering scheme be studied and identified in advance for each specific flood prone area.
- necessary laws and regultions be stipulated to prevent man-made destruction. In all cases, care should be taken not to induce harmful effects on river and cross-drainage flow.
- cuop and property insusance in flood prone areas be implemented gradually. Some pilot schemes already run have shown their effectiveness helpful for flood victims.
(7) Much attention should be paid to flood control management which usually consist of keeping constant maintainence and repair of all existing flood prevention work, fighting against emergency as well as taking urgent rescue and disaster relief. This has been proved to be absolutely necessary over ages in China.
Our experiences are mainly as follows :
ㆍ A powerful headquarter at various levels (the state, the province, the prefecture and the county) responsible for flood control management should be set up. Both national and local headquarters are generally composed of leading personal from all departments concerned, such as the water conservancy, meteorology, post and tele-communication, material and equipment supply, finance, agriculture, communication, and the People's Liberation Army(P.L.A).
ㆍ Flood fighting forces should be organized on the basis of the professional and non-professional, and of the army and civilian joint defence. In which army force should be the main one for taking emergencey actions.
ㆍ Measures for treating extraordinary flood should be planned in advance. When necessary, action to ensure security of important regions should be taken resolutely in view of an overall benefit.
ㆍ The capable youth living along the river should be imparted with technical knowledge and skill of flood control. Professional training on technique of fighting against emergency should be given to them so as to incerase their flood fighting ability.
- 126 - ㆍ The necessary laws and regulations to prevent embankments and other engineering works from man-made damage should be make public.
ㆍ The funds for annual maintenance and repair of the engineering works should be included in the national and local budgets.
(8) Any flood prevention projects should be so adopted as to make them sound and compatible from ecological and environmental points of view. It is generally recognized that structural measure for flood control can achieve a certain favourable environment for human habitation and agriculture, but some changes will inevitably lead to certain other changes. Improvement in one part of the region can degrade the environment elsewhere and at the same time one section of the community may benefit at the expenses of others.
Following are some of the major elements of environmental impact often occurred elsewhere as well as in China after the implementation of flood control scheme which should not be ignored.
ㆍ physical.
- hydrological impact.
- sediment movement, channel morphology.
- drainage conditions in the project area.
- ground water recharge portential.
- inland navigation and other water uses.
- soil physics and fertility.
ㆍ non-physical.
- land acquisition.
- agricultural pattern.
- flora and fauna changes.
- human life pattern.
(9) The objectives of a flood control scheme usually include; safeguarding lives and livelihoods, minimizing potential flood damages; creating flood free land and enhancing landuse; improving agro- ecological condtions for agricultural production; catering to requirements of navigations, communications, fishery and public health; creating a sense of confidence amongst people and stabilizing social equilibrium. It is obvious that the benefits accrued from the proposed scheme would be of immence value, however, different from the scheme of their purpose, the traditional benefit/cost analysis can hardly be used as a good tool to justify its feasibility. In China, the internal rate of return emerged out of the analysis is not considered as sole decision-making criterion for a flood control scheme. A rather low internal rate of return might be acceptable, if enough unquantified benefits could
- 127 - be identified.
The main reasons are mainly;
ㆍ To achieve the above objectives, only some of the benefits can be evaluated in precise quantification. Most of the socio-ecological benefits can't be quantified even they are so enormous.
ㆍ The dramatic damages caused by some catastrophic flood events would far outweigh the average annual damages as evaluated in the analysis.
ㆍ The indirect and intangible benefits in the long run which can hardly be foreseen would far outweigh the direct benefits in the short term.
ㆍ And above all, the social concern in a region under perpetual threat of flood is actually primarily one of risk aversion rather than profit maximization.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
DISCHARGE CHRACTERISTICS OF SUSPENDED SEDIMENT FOR FORESTED AND QUARRIED CATCHMENTS IN MOUNTAINOUS TERRAIN
1991
주관
수 행 기 관 : KOREA INSTITUTE OF CONSTRUCTION
- 128 - TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) 참여연구원 : Konkuk University, Korea Jong-Kwan Park
Summary
The problem of supply and transport of sediment from a mountainous catchment is very important in explaining dynamic geomorphology and the hydrological cycle. The discharge of suspended sediment is determined by a morphological system. Human interference to environment is also an important, not negligible factor in sediment production. Moreover, growing concern in recent years for the problems of nonpoint pollution and for the transport of contaminants through terrestrial and aquatic ecosystems has highlighted the role of sediment-associated transport in fluvial systems. This study was conducted in forested and quarried catchments in order to clarify the different discharge process and the mechanism of suspended sediment dynamics for each catchment.
As a forested catchment, the Yamaguchi River catchment which drains a 3.12 ㎢ area was chosen. On the other hand, the Futagami River basin, which is formed by three subbasins (1.07, 1.59 and 1.78 ㎢), as a quarried catchment was selected. These catchments are situated to the north and east of Mt. Tsukuba, Ibaraki, Japan. The discharge pattern of suspended sediment from the Futagami River basin is more unstable and irregular than that from forested catchment, the Yamaguchi River catchment. Under the similar rainstorm conditions, suspended sediment concentration from quarried catchment during a rainstorm event increases from 43 to 27,340 ㎎/1. However, in the case of the forested catchment it changes only from nearly zero to 274 ㎎/1. Generally, the supply source of suspended sediment is classified into two areas. the in-channel and non-channel source areas. As a result of field measurements, in the case of the forested catchment the in-channel (channel bed, channel bank and channel margin) is the main source area of suspended sediment. On the other hand, remarkable sediment source area on the quarried catchment is the non-channel that is unvegetated ground.
1. INTRODUCTION
In general, sediment load is classified into two groups by transport pattern, that is, bed load and suspended load. It is known that the major part of sediment load is transported by suspended sediment.
(Nippes, 1974)
- 129 - (Wood, 1977)
(Mizuyama, 1980)
(Park, 1990)
Suspended load is composed of fine particles and is predominantly a wash load; it is almost continually in suspension and is transported rapidly through the stream system.
Suspended sediment load is transported by running water in different path ways. Usually, it can be considered that the discharge process of suspended sediment is diverse in the artificially modified catchment. Suspended sediment discharge reflects the effect of human interference in the environment.
(Chow, 1967)
Walling and Gregory(1970) also indicated that the measurement of the effects of man upon drainage basin dynamics at present is important not only in explaining contemporary variations in water and sediment yield, but it is also an essential consideration in future predictions and in palaeohydrological reconsideration.
Nowadays, man's impact to the environment becomes a social issue for lack of the planning in utilization of the nature. The destruction of the nature such as the construction of golf links, in particular, becomes into big problems on the water quality, channel bed rise and frequent occurrence of the flood. The objective of this study is to investigate the discharge difference of suspended sediment between forested and modified catchment during rainstorm events and clarify the process and mechanism of the suspended sediment dynamics for each catchment.
2. TOPOGRAPHICAL CONDITIONS OF THE FIELD AREA
The River Futagami, as an example of modified catchment, drains a 4.44 ㎢ basin located in Tsukuba, Ibaraki Pref. , Japan (Fig. 1). The drainage basin is situated to the north of Mt. Tsukuba and to the west of Mt. Kaba. and can be subdivided into three catchments. For convenience, each catchment was named tentatively as the KITAZAWA catchment, the NAKAZAWA catchments and the MINAMIZAWA catchment. The longitudinal profile of channel in the three catchments was mapped as shown in Fig. 2. The longitudinal profiles of channel are approximately parallel to each other. The mean channel slope of each catchment has been estimated as between 0.174 and 0.232. The geology of this area is exclusively composed of granite. In the Futagami River basin, quarries are distributed widely. Man's interference has been a feature of the area for approximately 30 years. This study area, therefore, is a typical example of the modified catchment by human activities. The most modified catchment of the three, the KITAZAWA catchment, has been 16.7 per cent of the degree of human modification, with quarries occupying 12.9 per cent and roads covering 7.53 ㎞/㎢ of the catchment. The NAKAZAWA catchment, which is the largest catchment of the three, is 6.4 per cent modified by human activity, 4.3 per cent affected by quarrying, and the road density of 5.62 ㎞/㎢. The MINAMIZAWA catchment is the least modified, with 3.2 per cent of its catchment distributed by
- 130 - quarries and roads, covering 2.14 per cent and 2.7 ㎞/㎢ respectively. Vegetation consists of sparse deciduous trees approximately 15 m high and dense pinetrees with a dense ground cover of shrubs. There is no significant difference of vegetation among the three catchments. Soil type is a direct reflection of local lithology which ranges from silty-clay loams, volcanic ashes to residual deposits of granite.
On the other hand, the Yamaguchi River basin as an example of forested catchment was chosen.
(Fig. 3)
The drainage basin with an area of 3.12 ㎢ is located in the north of Mt. Tsukuba. The canopy density of the catchment is generally high. This study catchment is not nearly modified by human activity, but naturally preserved, though within the catchment the road had been constructed. It is not considered that sediment discharge from the road does not occur because the road was paved. Typical morphologic factors are shown in Table 1.
3. METHODOLOGY
The hydrologic response to the rainfall was monitored at the outlet in each catchment as shown in Fig. 1 and 3. Water discharge from the Futagami River basin was measured by using a weir and established constructures such as pipe. A stage/discharge relation for the weir has been derived from formulae supplemented by a current meter. It was not very effective, however, in shallow water. On the other hand, running water from the Yamaguchi River catchment was measured by a contracted rectangular weir. Three water stage recorders were established at the weir, which two of them are the digital type gauges and the other one is a daily chart type gauge. Therefore, discharge during rainstorm events could be measured very accurately.
Suspended load samples were collected by a hand sampling method using 1 liter polyester bottle at the each sampling site. An automatic pumping sampler could not be used because the water depth was too shallow. Sampling sites are shown in Fig 1. and 3. Dry weights of suspended load samples were measured later in the laboratory, and grain sizes were analyzed by a settling tube system, the efficiency of which was discussed in detail by Gibbs (1974) and Gibbs et al. (1971). Precipitation was measured by using a natural siphon rainfall recorder placed in each catchment.
4. MAGNITUTE AND PATTERN OF SUSPENDED SEDIMENT DISCHARGE
- 131 -
Concentrations and loads of suspended sediment
In the case of the Futagami River basin, many storms were recorded in the course of the study between June and October in 1987. The storm of June 20 which is the typical rainstorm in the measured events was selected. The total rainfall was 32.0 ㎜ and the peak rainfall intensity of the storm was 10.0 ㎜/hr. On the other hand, five rainstorm events were measured at the outlet of the Yamaguchi River basin for the analysis of the suspended sediment discharge. Table 2 shows the hydrologic characteristics and suspended sediment concentrations during the measured events. Total rainfall amount ranges from 30.5 ㎜ to 32.0 ㎜. Although it ranges almost same values, various sediment concentrations were observed because the rainfall intensity is different each other. Fig. 4 represents the relationships between the peak water discharge (Qp) and the suspended sediment concentration at Qp (Cp). Usually, it is known that the value of Qp and max value of suspended sediment concentration (Cmax) does not coincide. This phenomenon is, in general , called the hysteresis effect and it is found from the Table 2. Park (1991) also described that the loops of suspended sediment concentration in the Yamaguchi catchment showed the clockwise hysteresis. In the Futagami River basin, the hysteresis loops of suspended sediment concentrationare shown in Fig. 5. It shows clockwise hysteresis loops. From the Fig. 5, it is clear that the suspended sediment concentration during the rainstorm is in the decreasing order, the KITAZAWA, the MINAMIZAWA, the NAKAZAWA catchment. Fine-grained materials under 4 φ, in particular, were transported during rainstorms about 80 ％ in suspended sediment concentration. The transport rate of silt and clay is the highest value in the MINAMIZAWA catchment.
(Fig. 6)
On the other hand, the max values of suspended sediment concentration from the Yamaguchi River basin range from 103 to 274 ㎎/1, even if the values of the rainfall intensity have similar to that of the Futagami River basin.
(Table 2)
The suspended sediment concentrations in the forested catchment are low values about two orders of magnitute over that those in the artificially modified catchment.
Discharge process of suspended sediment
Universally, the supply source of suspended sediment loads during unrainy days is almost in-channel because suspended sediment transport from unvegetated ground does not occur. This regulation, of course, is not only applied in the forested catchment but in the quarried catchment. The in-channel process including in-channel sediment transport, and in-channel storage, is a very important factor in understanding suspended sediment dynamics. There is a great amount of fine-grained particles well sorted in the channel. Particularly, the parts of the channel margin which deposited much fine-grained particles influence the fluctuation of suspended sediment concentration during ordinary stages in the case of the modified catchment. Fig. 7 represents the suspended sediment concentration which was measured along the upstreams in the Futagami River basin. It is found that the values of suspended sediment on the falling stage, July 19, 1987 (total rainfall; 17.0 ㎜) is less than those of ordinary stages, July 7 and 8, 1987. In the case of the KITAZAWA catchment. sampling site K2, K3, K7 and of the NAKAZAWA, N2, N8, N9, of the MTNAMIZAWA, M1, M2 and M4, it can be seen that the values of
- 132 - suspended sediment concentration during unrainy days are large. It suggests that the suspended sediment concentration taken during ordinary stages is not always stable, and also indicates that the importance of the role of channel on the problem of suspended sediment transport has to be reappreciated anew.
On the other hand, the sediment source becomes to be changed during rainstorm events. Rainfall amount, rainfall intensity and raindrop impact are important as input factors of the suspended sediment delivery system. Especially, in the artificially modified catchment which unvegetated grounds range in basin, the pattern of suspended sediment discharge has some distinctive features. In the case of the Futagami River basin, sediment suppy from scattered waste heaps of earth in quarries has a direct effect on the fluctuations of suspended sediment concentration.
(Fig. 8)
It was confirmed, even if not recored, that surface flow from a quarry in the MINAMIZAWA catchment flowed into the forest along a sloping road.
(Fig. 9)
Moreover, unpaved road plays an important role of the increment of the suspended sediment concentration. Under the conditions of the concentrated heavy rains, running water produces rill or gully erosion and a great deal of fine materials is transported into the stream. Duck (1985) also described that the effect of road construction on sediment deposition has to be considered in the estimation of the process of sediment discharge. These road effect as one of the morphological factors in modified catchment can not be considered in natural condition.
The unvegetated ground is classified into two types, that is, active unvegetated ground and inactive unvegetated ground. Active ground is mainly formed by the heaps of earth with steep slope. On the other hand, inactive type is determined by not steep and hard top soil ground with few fine materials. These types must be classified in the estimation of the source of sediment supply.
5. CONCLUSIONS
This study on the suspended sediment dynamics was conducted in a forested catchment and quarried catchment. Suspended sediment discharge depends upon the morphological characteristics of catchment. Under the similar rainstorm conditions, suspended sediment concentration in the modified catchment during an event is higher than that in the forested catchment. And, the discharge pattern of suspended sediment from the artificially modified catchment is more unstable and irregular than that from the forested catchment. Moreover. in the case of the modified catchment the non- channel area is more important as the main source area of suspended sediment. The variation of suspended sediment concentrations reflects the difference of artificial rate to the catchment. And, the degree of catchment modification is also important to the sediment discharge.
- 133 - 6. ACKNOWLEDGEMENTS
The author wishes to acknowledge gratefully Prof. Shigemi Takayama and Associate Prof. Hiroshi Ikeda, the University of Tsukuba, for their advice and kind cooperation during the study.
TABLE 1. Typical morphometric indices for the study catchments.
TABLE 2. Hydrologic characteristics and suspended sediment concentrations during the measured events.
- 134 -
Fig. 1. The map of the Futagami River basin.
Fig. 2. Longitudinal profile of the main channel in the Futagami River basin
- 135 -
Fig. 3. The map of the Yamaguchi catchment.
Fig. 4. The relationship between the peak water discharge (Qp) and the suspended sediment concentration at Qp (Cp) in each catchment.
- 136 -
Fig. 5. The hysteresis loops of suspended sediment concentration in each catchment during rainstorms.
Fig. 6. Sand (<4 φ) percentage and silt-clay (>4 φ) percentage in suspended sediment concentration.
- 137 -
Fig. 7. The waste heaps of earth in the quarry.
Fig. 8. Surface flow flowed from a quarry Into forest along the sloping road
- 138 -
Fig. 9. Longitudinal change of suspended sediaent concentrations in the Futagami River basin. Number represents the sampling site of Fig. 1.
- 139 -
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
MATHEMATICAL SIMULATION OF NAVICATION CHANNEL CHANGES IN THE FLUCTUATING BACKWATER REGION OF THE THREEGORGES RESERVOIR
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Wuhan University of Hydraulic and Electric 참여연구원 : Engineering, Wuhan Hubei 430072,China Yitian Li Wuhan University of Hydraulic and Electric 참여연구원 : Engineering, Wuhan Hubei 430072,China Jianheng Xie Wuhan University of Hydraulic and Electric 참여연구원 : Engineering, Wuhan Hubei 430072,China Weiming Wu
Summary
- 140 -
In this paper, an one- and two- dimensional connected mathematical model is developed. The two- dimensional characteristics of flow and sediment transport are considered as much as possible in the model. The model is calibrated with field data and applied to simulate the river process of the fluctuating backwater region ofthe Three Gorges Reservoir. Some conclusions have been made based on the simulation results.
1. INTRODUCTION
The Three Gorges project is a huge multipurpose project. Navigation channel condition is one of the most concerned problem in the designning of the project. The fluctuating backwater region of the Three Gorges Reservoir is quite long, and situated in the Chuanjiang reach, a so-called golden navigation channel. The influence of sediment deposition on the navigation condition is very important, especially for some important reaches, for instance, Chongqing, Luoji-Wangjiatan and Qingyanzi etc.. Two-dimensional sediment mathematical model is one of the economic and effective methods for simulating a detail river bed deformation and navigation channel changes, but it is still relatively expensive. On the one hand, if the whole reach is calculated by the two-dimensional model, the computer storage requirement is quite large and computing time is quite long and expensive. But on the other hand, it is not necessary to calculate river bed deformation where navigation conditions are good in detail. Therefore an one- and two-dimensional connected model is used. The simulating practice shows this is a reasonable and economic measure.
2. GOVERNING EQUATIONS
Because an one- and two- dimensional connected model is used, the governing equations should include both one- and two- dimensional equations.
2.1 One-dimensional equations
Flow continuity equation :
- 141 -
Flow momentum equation :
For suspended load, sediment diffusion equation :
Bed deformation equation :
Sediment carrying capacity equation :
For bed load, bed load trasport equation :
Bed deformation equation :
in which A is cross-sectional area ; Q is water discharge; U and H are cross-sectional average velocity and water depth, respectively ; Z is water surface elevation ; n is Manning's roughness coefficient ; R is hydraulic radius ; B is water surface width ; S and S* are cross-sectional average suspended load concentration and sediment carring capacity respectively ; ρs' and ρb' are bulk density of suspended load and bed load deposit ; Q_{b} is bed load discharge ; d is sediment particle diameter ; α is a coefficient; ω is sediment settling velocity ; η_{s} and η_{b} are
- 142 -
2.2 Two-dimensional model
Flow continuity equation :
Flow momentum equation :
For suspended load, sediment diffusion equation :
Bed deformation equation :
Sediment carrying capacity equation :
For bed load, bed load transport equation :
- 143 -
Bed deformation equation :
in which x and y are two perpendicular horizontal coordinates ; u and v are depth-averaged velocities along the two coordinates axis ; ε is flow turbulence viscosity or sediment diffusion coefficient. Other symbols denote the same physical quantities as previously described, but being averaged in vertical instead of in cross section.
3. SUPPLEMENTARY RELATIONS
3.1 Coefficient of roughness
For one-dimensional model , the roughness coefficent varies from the beginning of reservoir operation to the bed that is fully covered by deposits. There, Han's formula (Han Qiwei, 1986) is used.
where n is a bed coefficient of roughness at any moment ; n_{0} is a bed coefficient of roughness at initial stage ; n_{k} is movabal bed coefficient at final stage which means that all the obstacales on the original bed have been fully covered by sediment deposition ; A is the deposition area at any moment; A_{k} is the deposition area at final stage.
For two-dimensional model, it is obvious that the coefficient of roughness across the channel is quite different from the average roughness coefficient of the small reach. In order to increase the accuracy of computations, Li and Xie (Li Yitian and Xie Jianheng, 1986) proposed a formula for calculating the Manning's roughness coefficent across the river.
- 144 -
where n is the roughness coefficient ; J is energy slop, f(η) is an emprical relation, η ＝ y/B is the relative position of the vertical, y is the distance of the vertical from one bank, B is width of the channel. The subscripts denote variables of he cross section, and those without subscript denote variables of each vertical.
3.2 Sediment carrying capacity
One-dimensional sediment carrying capacity is calculated by Zhang Ruijins's formula (Zhan Ruijing, 1961) which is widely used in China.
in which K and m are coefficients from the analysis of available data.
It has been verified that the sediment carrying capacity fomula for the whole cross section cannot be applied directly to a vertical.
(Li Yitian, 1988)
Based on data from field measurements, and expression for two-dimensional sediment carrying capacity calculation has been obtained.
in which, K is a cross section averaged coeffiecnt as stated above. Other symbols denote the same meanings as previously described, but averaged in vertical.
For non-uniform sediment carrying capacity, Li (Li Yitian, 1987) proposed a size distribution formula of sediment transported at its carrying capacity.
- 145 - in which σ_{v} is turbulent intensity of the vertical ; ΔP_{i} and ΔP_{bi} are the proportion of size fraction i in suspended load and bed mixture ; ω_{i} is settling velocity of partical size i ; u* is shear velocity. In this formula, the affection of bed mixture and flow intensity to the sediment carrying capacity is fully considered.
3.3 Bed load transport formula
Many formulas have been verified with the use of the Three Gorges river reach data, but all the calculated values are apart from the measurement. The values calculated by Dou's formula (Dou Guoren, etc., 1989) are agreed with the measurement relatively better, so Dou's formula is used herein.
where p^{b} is the propotion of bed load in the bed mixture : u_{c} is critical velocity when sediment begins to move; ω is settling velocity of bed load particle : k₁ is a coefficient.
4. BOUNDARY CONDITIONS AND NUMERICAL SOLUTION
Upstream boundary condition imposes water discharge and cross sectional averaged sediment concentration for one-dimensional model , and velocity distribution as well as sediment concentration distribution across the river for two-dimensional model ; Downstream boundary condition imposes cross sectional averaged water surface elevation for one-dimensional model, and water surface elevation distribution across the river for two-dimensional model. At the connecting section of one- and two-dimensional models, all the one-dimensional models calculated physical quantities are required to equate the cross sectional averaged values of the two-dimensional model calculated quantities. This is a very strict requirement, and it is hardly to be satisfied. If the connecting section is chosen at a relatively straight reach, the water surface elevation, water discharge and velocity distribution, sediment concentration, and etc. can be estimated with certain accuracy or approximately.
(Wu Weiming and Li Yitian, 1991)
For saving the time in computation and amount of data stored in the memory of computer, efforts have been made to introduce simplifications in numerical solutions of the governing equations. One of the major simplifications is that the coupled solution is approximated by uncoupled solution, which means that instead of solving simultaneously all the equations governing the phenomenon, the
- 146 - continuity equation and momentum equations of the water flow are solved first to obtain the hydraulic factors and then the bed load transport equation, the sediment carrying capacity equation, the sediment diffusion equation and the river channel deformation equations are solved to obtain the distribution of both suspended and bed load and consequently of the channel deformation. Such a simplification is permissible because that the bed level change within a time step is much smaller than the change of water depth. Another simplification is to take the unsteady flow as a steady one, in other words, turning the incoming water and sediment hydrographs into histograms with total quantities of water and sediment remaining unchanged. Some studies (Li Yitian and Yin Xiaoling. 1990) demonstrate that such a simplification exerts large influence on the computational results of flow discharge and water level as well as sediment transport and channel deformation during the rising period and peakflow of a flood. However, the influence on the quantity of accumulated channel deformation after the flood recession is just moderate, especially for computation of a long time series. It is out of question for our computation period is thirty years long.
The numerical method used in one-dimensional model is a standard routine. In two-dimensional, a modified finite analytic method is developed(Li Yitian, 1989).
This method can not only reduceing the computing time and with certain accuracy compared with finite difference and finite element method, but also fairly fit the irregular river bank.
5. SIMULATION OF THE NAVIGATION CHANNEL CHANGES
The upstream and downstream boundary condition proposed by the Yangtze Valley Planning Office (YVPO) is used herein.
(YVPO, 1988)
The computation reach is about 200 ㎞ in which the two-dimensional calculation reach is 70.5 ㎞, and the time period is 30 years. The reservoir operation scheme for the first 10 years is 156-135- 140(normal pool level, flood-control level, dry-season control level ), and for the other years is 175- 145-155. Based on the computer simulation, following conclusions could be made.
For the first 10 years, the backwater affection only reach Tongluoxia, the navagation channel in the Chongqing reach is still at its natural state. At Luoqi of Changshou reach, different paths followed by the main current in flood and dry seasons lead to the deposit of longitudinal underwater ridges. However the navigation depth can be kept, only in some dry years when discharge below 3500 cu m/s and the deposition on the shoal cannot be scoured in time, there would be some navigation problems. In the Qinganzi reach, some deposition would take Place and some navigation trouble may be occur, but not serious.
From list year, sediment deposition would take place in the whole backwater effect, the navigation conditions local deposition of sediment would cause navigation trouble. For instance, in Jiulongpo port of Chongqing reach, cumulative depositions of sediment in the left side bar would cause the navigation channel shift a distance. In Chaotianmen shipping terminal at Chongqing, the sediment deposition would occur in the natural navigation channel during dry season and the navigation channel would shift 125 m. All these river channel changes would introduce some trouble to navigation and shipping terminal operation. So some river trainning work and dredging operation should be taken.
- 147 -
Un Qingyanzi reach which contains a widened stretch on a curve with a central island, Jingchuanqi in the midstream. Under natural conditions, the flow during dry season, would follow the deeper branch along the concave bank to the right of the island. During the flood season, however, flows of large discharges would drown out the guiding boundaries and the large flow being of greater momenta or inertia would take a course of smaller curvature about 670 m to the left of its original course. A slack water region containing a slow vortex would then form over the right branch, giving rise to heavy deposition of suspended sediment in this region. Toward the end of the flood season, the flow in the river would gradually diminish and more flow would then be guided back to the right branch causing rapid scour in that region, so that eventually whatever deposition left over from the previous flood season would be scoured away and the right branch would be restored as main channel. After about 10years reservoir operation, deposition in the reservoir would cause the backwater stage to rise beyond 4 m, deposition in the right branch during flood season would be accelerated, while less sediment would be scoured away at the end of the flood season, accumulated depositions would take place in the right branch which would eventually blocked up. After about 17 years of reservoir operation, there would then be a switch of the main channel for navigation from right branch to left. As the new channel would pass over a bed full of large rock croppings the flow would be very rough and unsafe for navigation. Therefore either the bed should be cleared of undesirable protrusions or training work should be built to guide the flow into its original main channel. But the scale of construction is relatively small.
Sediment depositions in the other reach also cause some navigation trouble, either the requirement of navigation depth cannot be satisfied or the navigation channel shifts or switches, may be take place. But all these problems are not so serious and can be solved either by reservoir operation or engineering measure.
6. CONCLUSION
Extensive investigation on the mathematical simulation of sediment deposition in the backwater region of the Three Gorges reservoir has been made and an one- and two-dimensional connected sediment mathematical model is developed. Based on the computer simulation, some important observation and proposal for the navigation problems in the backwater region of the Three Gorges reservoir is made. Compared with the physical model test, the mathematical model calculated results are in accordence with the physical model test results both in qualitatively and quantitatively.
In this paper, an one- and two- dimensional connected mathematical model is developed. The two- dimensional characteristics of flow and sediment transport are considered as much as possible in the model. The model is calibrated with field data and applied to simulate the river process of the fluctuating backwater region of the Three Gorges Reservoir. Some conclusions have been made based on the simulationresults.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA"
- 148 - (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
A COMPUTER-AIDED GUIDELINE FOR THE SELECTION OF SEDIMENT TRENSPORT FORMILAS
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Research Fellow, Water Resources Division, 참여연구원 : Korea Institute of Construction Technology, Korea Hyoseop Woo 참여연구원 : Researcher. idea Kwonkyu Yu President, Korea Institute of Construction 참여연구원 : Technology, Korea Yun Sik Lee
Summary
A computer-aided guideline has been developed in order to provide river and canal engineers with a practical tool for the proper selection of sediment transport formula. This guideline is composed of a user-friendly, step-by-step, PC-oriented computer program. This guideline is based on the results of a careful review of the applicability and limitation set by the developers of each sediment transport formula, some existing performance tests of sediment transport formulas with field data; and a sensitivity analysis with the hydraulic and bed material variables, a comparative analysis with data of an imaginary channel with uniform sediment, and an extensive comparison of the sediment transport formulas with a field data of more than 500 data points selected from the literature.
- 149 -
This computer-aided guideline is named "GUIDE", and is stored in a HD diskettee and supplied with other auxiliary files necessary to the operation of it.
1. INTRODUCTION
Engineers engaged in river management and regulation and design and operation of canal system have great need for methods of estimating sediment transport discharge. Unfortunately, presently available methods or relations for estimating sediment discharge are far from completely satisfactory. At best these relations serve as guides to planning, not only because of the accuracy problem but also because of the difficulty in selecting proper sediment transport formulas.
Many sediment transport formulas have appeared in the literature since DuBoys(1879) presented his tractive force relation. The problem of the engineer is to select one or more of these for use in solving his particular problem. This selection.
Many sediment transport formulas have appeared in the literature since DuBoys(1879) presented his tractive force relation. The problem of the engineer is to select one or more of these for use in solving his particular problem. This selection is not straightfoward since the results of different formulas often differ drastically and it is not possible to determine positively which one gives the most realistic result.
A few guidelines for the selection of sediment transport formulas, if not comprehensive, are found in the literature, including the early work of Shen (1971) and a more recent work of Stevens and Yang (1989). These guidelines, however, lack comprehensiveness and generality because they only recommend one or two specific formulas according to some limited cases of the channel characteristics.
The purpose of this study, therefore, is to develope a comprehensive guideline for the selection of sediment transport formula and help river engineers select one proper to their particular problems. For this purpose, some existing sediment transport formulas whose computed results appear to be relatively reliable were selected preliminarily, and the performances of the selected formulas, including any limitation of their applicabilities, were analyzed by using the sensitivity analysis and dirct comparison with measured data.
Sediment transport formulas considered in this study are limited to ones developed in the Western countries including the United States and Western Europe, excluding Asian or Eastern Europian origins. It is because works of assessment of the sediment transport formulas developed in Asia or Eastern Europe, which is essential for the preliminary selection of sediment transport formulas that are to be tested in this study, are few, although the number of the formulas developed in non-Western regions is not small. Also, only total sediment transport formulas for alluvial streams with non- cohesive beds are considered in this study, excluding bed load or suspended load formulas, and the formulas for cohesive sediment transport.
- 150 - 2. METHOD AND PROCEDURE FOR THE STUDY
To develop the guideline, the following method and procedure were adopted in this study.
(1) Collection of existing sediment transport formulas.
Sediment transport formulas which were presented in the literature were collected and analyzed in respect of their formulations, applicabilities, and limitations.
(2) Collection of sediment discharge data.
Measured sediment discharge data in the field were collected through the literature review. Especially, more than 1,000 field and flume data points in Brownlie's compendium(1981b) were arranged into a database.
(3) Collection and analysis of research works regarding performance of sediment transport formulas.
More than 10 research works, including White, et al's work (1973), which assessed sediment transport formulas with measured sediment discharge data, were collected and analyzed.
(4) Preliminary selection of sediment transport formulas.
Through the above step (3), the total nine formulas were selected in order to be considered in the guideline. This preliminary selection is based on the followings.
① The formulas which seem relatively reliable;
Ackers ＆ White's(1973), Engelund ＆ Hansen's(1967),
Yang's(1973), and Shen ＆ Hung's(1971) formulas.
② The formulas which are known widely;
Einstein's(1950), Toffaleti's(1968), Colby's(1964) formulas.
③ The formulas which were developed recently;
Ranga Raju's(1981), and Rijn's(1984) formulas.
(5) Analyses of characteristics of the selected formulas.
① Through a careful examination of the condition in which parameters in each formula were calibrated, any limitation of applicability of each formula was analyzed.
② Through a parameter-sensitivity analysis, performance of each formula was examined in the physical viewpoint.
- 151 - ③ Each formula was applied to 519 field data points that were carefully selected from the database.
(6) Development of a computer-aided guideline.
Each formula considered in this study was programmed in FORTRAN. The program was thoroughly examined by using the existing sample calculations. A guideline program written in the dBASE Ⅲ＾{＋} language were finally developed.
3. RESULTS AND DISCUSSION
(1) Sensitivity analysis.
A parameter-sensitivity analysis was conducted to check the behavior of each selected transport formula with a change in the flow and bed sediment characteristics. This analysis is significant especially in the sense that if calculated values from a certain transport formula are too sensitive to a small change in a certain flow or sediment characteristics such as flow velocity or sediment size, that formula may not be acceptable. It is because with that formula, a small error in the measurement of flow or sediment characteristics would cause a large and susceptible difference in the prediction. For this sensitivity analysis, flow velocity, channel depth, slope, water temperature, and sediment size were considered as key parameters with uncertainties of measurement. Water discharge was not considered separately because it can be substituted with the flow velocity. A data collected by Nordin and Beverage (1965) from the Rio Grande River near San Ildefonso was used for the reference data. They are as follows; water discharge Q ＝ 41.4 ㎥/s, width B ＝ 34.2 m, depth d ＝1.0 m, mean velocity V ＝ 1.2 m/s, channel slope S ＝ 0.00098, water temperature T ＝ 15˚ C, sediment sizes for which 35, 50, 65, and 90％ of total sediment weight are finer, D ＝ 0.3 ㎜, 0.36 ㎜, 0.44 ㎜, and 1.31 ㎜, respectively. With each parameter changed from -50 ％ to ＋50 ％ from the reference value, changes in the sediment discharge calculated from each formula were plotted with changes in the parameter.
In general, all the transport formulas tested are more sensitive to the change in the flow velocity and sediment size, while they are less sensitive to the channel slope and depth. They are least sensitive to the change in water temperature.
As shown in Fig.1, Einstein's, Toffaleti's, and Colby's formulas are very sensitive to the change in the flow velocity. This result is predictable, because the former two transport formulas are based on the diffusion-advection theory where the flow velocity plays a dominant role, and the latter formula has been calibrated empirically with the flow velocity as a principal parameter. On the other hand, Engelund and Hansen's and Yang's formulas, among those tested, seem least sensitive to the flow velocity.
Next, a series of numerical experiments were conducted on the sediment transport formulas selected in this study. All the transport formulas selected were applied to an imaginary, wide channel with a uniform sediment. An empirical friction formula suggested by Brownlie (1981a) was used to simulate a change in bed roughness due to a change in flow discharge. Only lower flow regime and transition were considered. The unit water discharge was increased from 0.05 to 30 ㎥ /s/m with two different
- 152 - sizes of bed material of 0.25 and 1.0 ㎜. The bed slope and water temperature were assumed to be 0.0005 and 15 ℃, respectively. All the calculations in this test were conducted within the range of applicability of each transport formula, which had been suggested by the developer of the formula.
Fig.2 shows a part of the results of the test for the sediment size of 1.0 ㎜. As shown in this figure, sediment discharges calculated from each transport formula differ greatly in the margin of the order of the factor up to two even for this ideal condition, which clearly indicates that existing transport formulas lack consistency.
In general, exponential type of sediment transport formulas, such as Engelund and Hansen's and Shen and Hung's formulas, behave linearly in the log-log plot, while Ackers and White's and Colby's formulas, and the diffusion-advection type of transport formulas, such as Einstein's, Toffaleti's, and van Rijn's behave curvedly in the plot. The rough and sometimes abrupt variation of sediment discharge calculated using Toffaleti's formula appears to be due to the abrupt artificial variation of some empirical parameters used for the formula.
(2) Analysis with measured data.
An analysis of reliability of each sediment transport formula was conducted using a set of 519 field data points selected from the database. The discrepancy ratios, which are defined as the ratios between calculation and measurement, were plotted, respectively, with respect to water discharge, unit water discharge, flow depth, stream power, and dimensionless sediment diameter. A summary of the distribution of the discrepancy ratios is shown in Fig.3.
As a result, the performance of Einstein's formula seem to be poor. Colby's formula would overestimate sediment discharge in case of q ＞ 5.0 cms/m. Although Engelund ＆ Hansen's formula shows good results for the entire range of data, the result appears to be not encouraging for large-sized channeIs. Toffaleti's formula has a consistent trend of underestimation. Variations of the discrepancy ratio for Shen ＆ Hung's formula is large for large-sized channels.
Ackers ＆ White's formula shows good results, with its variation being somewhat larger than that of Engelund and Hansen's formula. Though Yang's formula would underestimate for large-sized channels, it shows a relatively good result for the other range of channel size. This result is in agreement with that of Rijn's study(1984). Variation of the discrepancy ratio of Ranga Raju's formula is very large. On the other hand, Rijn's formula would overestimate for small-sized channel such as q ＜0.2 cms/m, while it shows a very good result for medium- or large-sized channels.
In general, Engelund and Hansen's, Ackers and Whites', and van Rijn's formulas appear to be better than the others. On the other hand, Einstein's, Toffaleti's. Shen and Hung's, Yang's, and RR's formulas would underestimate sediment discharge. which is consistent to the result obtained by Brownlie (1981a).
4. GUIDELINE FOR THE SELECTION OF SEDIMENT TRANSPORT FORMULAS
- 153 - In this study, it is recommended for river and canal engineers facing estimating sediment discharge at a certain channel reach to select sediment transport formulas by the following steps. This guideline includes a basic information of sediment transport in its step-by-step procedure to help novice engineers who are not familiar with this kind of problems.
(1) step 1: Is the channel considered as an alluvial one?
Most existing sediment transport formulas can be used only for alluvial channels. Few formulas, therefore, can be applicable to non-alluvial channels, such as channels with cohesive- , rocky-, or concrete-bed.
(2) step 2: Is the reach straight and the flow condition steady and uniform?
Strictly speaking, no sediment transport formula can be applicable to a channel reach which meanders or whose flow condition is unsteady or non-uniform.
(3) step 3: Is any sediment rating curve available?
At present, the beat way to determine sediment discharge in an alluvial channel is to rely on direct measurement or use existing measured data for the channel. In case that any reliable sediment rating curve exists for the channel, therefore, it is recommended to use it first.
(4) step 4: Is washload dominant in the reach ?
Washload depends on stream sediment supply rather than hydraulic characteristics of the stream. No sediment transport formula can be applicable to streams with washload dominant.
Also, existing formulas, except Colby's one, can not be applicable to hyperconcentrated flow the sediment concentration of which exceeds about 100,000 ppm, where the presence of fine sediment (D_{50} ＜ 0.0625 ㎜) meterially affects fluid properties and bed material transport. Exceptionally, Colby's formula was formulated to be able to correct sediment discharge in the presence of fine sediment where the sediment concentration does not exceed 200,000 ppm.
(5) step 5: Are all the data necessary for calculating sediment discharge available?.
Existing sediment transport formulas necessarily require the flow depth (d), flow velocity (V) (or water discharge, Q), and energy slope (S) as the hydraulic characteristic data and median diameter (D_{50}) of bed material as the sediment characteristic data. No sediment transport formula can be used, when any one of the above variables is not available. Exceptionally, Colby's formula doesn't require data of the energy slope.
(6) step 6: Is the sediment size in the reach in the range in which the formulas can be applicable?
No sediment transport formula can be applicable to the channels with clay- or silt-beds (D_{50} ＜ 0.0625 ㎜) or cobble-bed (D_{50} ＞ 64 ㎜).
(7) step 7: Are the formulas reliable?
It is less recommended to use Einstein's, Toffaleti's, Shen and Hung's, and Ranga Raju's formulas, since they appear less reliable than the other formulas considered in this guideline.
- 154 -
(8) step 8: Is the sediment discharge calculated by the size fraction?
Among the sediment transport formulas considered in this study, only Einstein's, Toffaleti's, and Yang's ones can calculate sediment discharge by the size fraction.
(9) step 9: Does the hydraulic characteristics of the reach exceed the range in which the formulas can be applicable?
① The formulas that are applicable only to the flow where the Froude number (F_{r}) is less than 1.0, i.e., the subcritical flow, should not be applied for a supercritical or critical flow. For example, it was recommended to apply Ackeras and White's formula only to the flow with F_{r} ≤ 0.8. and van Rijn's one with F_{r} ＜ 0.9.
② Colby's and Yang's formulas seem to cause a large error when they are applied to large-sized channel. Especially for q ＞ 5 cms/m or d ＞ 5 m. Colby's formula would underestimate sediment discharge. On the other hand, van Rijn's one would overestimate for small-sized channels such as, q ＜ 0.2 cms/m or d ＜ 0.3 m.
③ Generally speaking, van Rijn's formula shows a relatively better result for large-sized channel (q ＞ 5 cms/m or d ＞ 5 m), and Engelund and Hasen's and Ackers and Whites' ones for small-sized channel. For medium-sized channel(0.5 ＜ q ＜ S cms/m or 0.5 ＜ d ＜ 5 m), these three formulas show a similar degree of reliability.
(10) step 10: Does the sediment characteristics exceed the range in which the formulas can be applicable?
① CoIby's formula can be applicable only for 0.1 ≤ D_{50} ≤l 0.8 ㎜.
② Rijn's one would overestimate sediment discharge for D_{50} ＞ 1.0 ㎜.
③ Ackers and Whites' formula overestimates drastically when the sediment size is smaller than that of the fine sand (D_{50} ＜ 0.25 ㎜).
(11) step 11: When no more information for the selection is available, it is recommended to use Ackers and Whites', Engelund and Hansen's, or van Rijn's formulas. The former two formulas were developed earlier and more widely used than the last one.
(12) step 12: Simplicity.
When the simplicity of calculation in situ is considered important, Colby's and Engelund and Hansen's formulas have a distinct merit. With these formulas, sediment discharge can be calculated using a simple chart and a calculator only.
The above procedure has been programmed by using the dBase Ⅲ＾{＋} language and the software developed thereby is named "GUIDE". This software, stored in a HD diskettee. is a user-friendly, step- by-step, PC-oriented computer program. The "GUIDE" is supplied without any charge to any
- 155 - individual or group entity who is interested in the selection of sediment transport formulas.
5. ACKNOWLEDGMENT
This paper is based on the research report entitled "Developement of a Guideline for the Selection of Sediment Transport Formulas", Report 89-WR-113, which was conducted in the Korea Institute of Construction Technology during the year of 1989.
The guideline suggested in this study is based on the 'average-sense' result. Any recommendation suggested by this guideline and the computer program "GUIDE", therefore, would not always necessarily reflect the best choice for a particular flow condition, and the Korea Institute of Construction Technology shall be under no "liability whatsoever to any individual or group entity by reason of any use made thereof.".
- 156 -
Fig.1 Sensitivity of sediment transport formulas with respect to change in flow velocity
- 157 -
Fig.2 Sediment-rating curves obtained by using sediment transport formulas (D = 1.0 ㎜)
- 158 -
Fig.3 Distrbution of discrepancy ratios
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF
- 159 - INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
PRELIMINARY STUDY ON SEDIMENT PROBLEM AT HOABINH RESERVOIR
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Hydrologist, Institute of Meteorology and Hydrology, Hydrometeorological Service, No. 4 참여연구원 : Bang Thai Than St., Hanoi, Vietnam Du Cao Dang Hydrological Engineer, Hydrometeorol ogical 참여연구원 : Service, No. 4 Dang Thai Than St., Hanoi . Vietnam Van Hui Dinh
1. INTRODUCTION
The Da River is one of the tributaries of Red River in Vietnam. which takes its source from China. Its drainage area is 51,800 sq. Kms (above HoaBinh dam), of which about 50％ belong to Vietnam.
Analyses of sediment regime have been made from observation data in station network. There are 9 sediment sampling stations in Da River basin (in Vietnam).
(Figure 1)
The HoaBinh Reservoir has been built for multipurpose. Former design features can be summarized as follow:
- 160 -
2. SEDIMENT REGIME
2.1 The long-term variability of sediment discharges is rather big.
The variability is defined by the statistical parameter, such as:
The coefficients of variation of sediment discharges are presented in table 1.
Table 1
- 161 - Figure 2 provides an example of variability of annual sediment yield at LaiChau station. We know that any man's activities on basin may lead to the change of sediment regime, but even under natural condition it is always changed.
Such high coefficients of variation demand long records for estimation of average annual sediment discharge.
The transported sediment yield is concentrated into flood season (from May to October). There are more than 90 ％ of the annual sediment load transported in this season. Especially in some high flood days, such as at HoaBinh there was 82 × 10^{6}tons of suspended sediment transported within 10 days (1969).
(Figure 3. a.)
The pattern of distribution of suspended sediment discharge is presented in figure 3b.
From these characteristics, if the pit mode of operation for discharging sediment through flow is chosen, it will be an effective measure to reduce the amount of deposited sediment in the reservoir. However, it's sometimes out of the prupose of flood control for downstream.
2.2 Distribution of sediment yield on various part of basin is different.
A convenient way supplying information regarding sediment yield for design is to analyze all available sediment data and regionalize it on a catchment scale. It's also to indicate maximum yield figures in order to make soil protection.
Because of limited gaging stations measuring sediment (9 out of 26 stations) and short observational periods (10 to 30 years) the data must be extrapolated and extended.
The relationship between sediment discharge Qs and water discharge Qw established for the above mentioned purpose is as follws :
Where a, b : coefficients.
(Fig 4)
Furthermore the extrapolation data were carried out by using soil erosion map established based on USLE.
Based on the results of analyses, the regionalized map of suspended sediment yield have been established.
(Fig 5)
- 162 - It can be seen that the main sediment sources are from the upstream part of basin.
The size of sediment particles is rather fine and it becomes tkner in flood periods.
(Fig 6)
(Fig 7)
3. THE ESTIMATION OF SEDIMENTATION IN THE RESERVOIR
The estimation of deposition in the reservoir includes the computation of the deposited volume and its distribution in the reservoir.
3.1 In computing volume deposited in the reservoir the equation of sediment balance can be used in the form[1]
(Long Yuqian, 1985)
Vi ＝ V_{o} ＋ V_{d}.
Where Vi : total of incoming sediment load (sediment yield from the drainage basin and from bank erosion),
Vo : sediment load passing through all the outlet structures (sluice gates, spilIways, turbine runners).
3.2 Distribution of deposited sediment :
As the HoaBinh is long, narrow and deep, its useful storage capacity may be deposited before its dead storage. This makes the negative influence for its operation.
The calculation of sediment distribution is carried out by empirical method(method of Borland and Miller [2]) and mathematical model method.
- 163 -
4. THE SEDIMENT CONTROL MEASURES
As analyzed above, the sediment in Da River catchment and HoaBinh reservoir has following characteristics :
- Sediment source are mainly supplied from upstream.
- Transported sediment is concentrated for the short flood time.
- The grain size of particles is rather fine.
- The ratio of the total storage capacity and annual water volume coming to the reservoir is rather small (＝ 1/6).
From those. the measures suitable for control of sedimentation in HasBinh reservoir may be :
(1) Soil conservation methods in watershed.
(2) Selection of operation scheme to release as much sediment as possible from the reservoir by making use of silt carrying capacity of the flood.
A project of soil conservation in catchment is carried out combined biological, agricultural and mechanical measures. The biological measure is concentrated on reforestation as forest bands to prevent soil erosion. Their effect perhaps in part depends on regional cooperation.
Table 2 The sediment balance in HoaBinh Reservoir (1983-1989)
(*) The estimation of V_{0} was carried out by using empirical method of Brune, Churchil and the method estimation of sediment discharge of density current.
- 164 -
Fan Jiahua, 1985)
Fig 1. A sketch map of Da River Basin end Hoa Binh Reservoir
Fig 2a - Variability of annual Sediment yield
- 165 - Fig 2. b - The double Mass curve of annual sediment yield Vs and water Sources. Vw
Flg. 3. a - The concentraflon of Sediment volume
Fig. 3. b - Distribution of Suspended Sediment
- 166 -
Fig. 4 - Relationship between S. Sediment yield and runoff
Fig 5 - Reglonailzed map of suspended Sediment yield
- 167 -
Fig. 6 - Grain size distribution at Hoa Binh
Fig. 7 - The variation of size of Sediment particles
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS
- 168 - INCUCING THOSE OF INTERNATIONAL RIVERS
RESERVOIR SEDIMENTATION AND WATER QUALITY IN KOREA
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Researchers, Water Resources Research 참여연구원 : Institute, Korea Water Resources Corporation, Taejon, Korea Kwang-Man Lee Professor, Civil Engineering Dept. , Chung-ang 참여연구원 : University, Seout, Korea Soo-Sam Kim Researchers, Water Resources Research 참여연구원 : Institute, Korea Water Resources Corporation, Taejon, Korea Hong-Sub Shin Researchers, Water Resources Research 참여연구원 : Institute, Korea Water Resources Corporation, Taejon, Korea Seok-Ku Ko
1. INTRODUCTION
Reservoir sedimentation has been regarded as one of the most important and complicated problems with regard to the reservoirs constructed on the alluvial channels. As the alluvial channel can be defined as a channel in which the scouring and deposition are continuously occurring, the construction of a dam can be resulted in causing sedimentation problems in the reservoir area created by the facility. Because most of the rivers in Korea are classified as alluvial channels. even if it is not serious, large amount of sediments are flowing into the reservoirs. Larger parts of the sediment inflow are occurring during the rainy season, but suspended solids are also flowing into reservoirs and deposited even during the low-flow period due to men's activities.
Fine sediment transport in a reservoir depends on the hydrologic, seasonal changes of thermocline, chemical and biological conditions. It is also influenced by water quality distribution including resuspension, deposition and flocculation of the sediments. For the large-scale reservoirs in Korea, sediments do not look like to cause any significant problem as regarding the lifetime of the structure and the elevated inundated area due to the sediments accumulation. However. fine sediments contained with large portion of nutrients flow into reservoirs, and these are the one of the nuisance
- 169 - sources in water quality management. Because these fine sediments contain a large amount of nutrients including phosphorus and nitrogen components, these cause overbreeding of photosynthetic algae resulting in near-eutrophication of the reservoir. This state can be found in the large-scale reservoirs constructed more than ten or fifteen years ago in Korea.
Judging from that the recent phenomena of red tides in some rivers or lakes are deeply dependent on rising and falling activities of the theromocline, the sediments are considered to be the principal sources of nutrients which are directly related to the phenomena. With respect that the seasonal changes of concentration of dissolved heavy metal's depend on the suspended particles, it is important management practice for the aquaenvironment to grasp the status of phosphorus and heavy metals existing in the sediments and to estimate their possible movability. Although many models have been developed to estimate the amount of pollutants which is discharged to the upper water from the bottom sediments, it is still difficult to estimate the characteristics of the pollutants in real rivers or reservoirs as the models were developed more based on theoretical point of view. It can be a very practical method to estimate the possibility of water pollution from the sediments through grasping the status of the pollutant existing in the sediments and understanding the factors of their movements.
(Jun, 1990)
This study aims at finding the relationship between sediments deposition and eutrophication of large- scale reservoirs in Korea, including some prospective solutions to the recent problems in the reservoir system management.
2. RESERVOIR SEDIMENTATION AND HYDROLOGICAL DATA
Korean peninsula belongs to the Asian monsoon climate and has the annual precipitation of 1,262 to 1,274 millimeters. Biennial variations of the rainfall are significant with the range of 854 to 1,683 millimeters during the year of 1905 to 1988, and the variations show the increasing trend in general. Rainfall distribution shows a remarkable distinction between rainy season and dry season with 70 percent of the yearly precipitation occurring June through September, and 20 percent October through the next March.
(KOWACO, 1990a)
Because of the seasonal extreme variation of rainfall, frequent floods occur in summer and drought during winter and spring. Especially, this country suffers from frequent flood disaster due to topographical rain and localized torrential down pours caused by typhoon. During this time a lot of suspended solids and sediments flow into reservoirs.
In Korea, several multi-purpose dams have been constructed for the major purposes of flood control and water supply including hydro power and water quality enhancement. Table 1 shows major large dams in Korea.
Table 1 Major Large Dams in Korea
- 170 -
Sediments deposition in a reservoir results in decrease in its capacity, and causes the rising backwater level in flood period due to a delta formed in the upper tail area. It also causes eutrophication in the reservoir due to nutrients containded in clayey and silty sediments. Therefore it greatly affects on water quality as well as on the change of physical phenomena of the reservoir.
It may be more realistic to use an empirical method based on the measured data as no mathematical model based on pure theoretical background can evaluate the factors quantitatively which affects on sediments deposition. An empirical approach is still regarded to be more acceptable than a theoretical method.
(KOWACO, 1983)
However, many scholars have been contributing themselves to improve the evaluation methodologies of sediments inflow and distribution in a reservoir. In 1983, Korea Water Resources Corporation (KOWACO), a semi-government organization, surveyed sediments deposition in several large-scale reservoirs, and estimated the deposited sedimentation amount as shown in Table 2.
(KOWACO, 1983a)
(KOWACO, 1983b)
(KOWACO, 1983c)
(KOWACO, 1983d)
The data can be some help to estimate sediments flowed in river as the dams are located throughout the major rivers.
For the Soyang reservoir located in the Han River, 500 cubic meters of sediment annually per one square kilo meters from upstream basin had been estimated to be deposited at the design stage. However, the surveyed result based on the 10 years of deposition after the completion of the project in 1973 shows 27 million cubic meters had been deposited which is equivalent to 1000 cubic meters of sediment annually per one square kilo meters had been deposited to the reservoir.
- 171 - (KOWACO, 1983b)
Even if it is not significant to the effective storage or life time of the project, the amount is two times greater than that estimated at the designed stage. Table 2 shows that the situations of the sediment deposition for the other surveyed reservoirs are similar to the Soyang reservoir.
Table 2 Sediments Deposition in the Major Reservoirs (Unit : m³/km²/Yr)
3. SEDIMENTS DEPOSITION AND WATER QUALITY IN A RESERVOIR
The amount of pollutants which are discharged into water body from the deposited sediments is dependent on the amount and the state of the pollutant contained in the sediment materials. And also it depends on the activities of benthos and hydrodynamic conditions near the surface of the sediment. Especially, the sediment which contains the easily soluble pollutants has large possibility of water pollution. It takes lots of time to be restored from the polluted state into good quality even if the source of pollution doesn't flow into a reservoir. In case the pollutants are flowed in a reservoir with sediment materials, then some parts of the pollutants are changed from the dissolved state into solid state, or some parts are adsorbed to the surface of fine particles and deposited to the bottom of the reservoir. And then the pollutants move up to the water body over a long time due to the turn-over phenomenon or due to the organic degradation or their chemical and physical changes.
(Jun, 1988)
Therefore, it is possible to estimate the future water quality by finding out these kind of activities of the reservoir sediments.
According to a study on the Soyang reservoir performed in 1989 (Jun and Park, 1989), it was estimated that the reservoir in general would be in between oligotrophic and mesotrophic state.
- 172 - However, the phenomena of remarkable water blooming have been observed in its tributary streams during Spring and Fall every year. Judging from that the water blooming have been arisen just after the reservoir water body turn-over period, these phenomena are deeply related to the movements of the thermal stratification of the reservoir. and it can be concluded that the deposited sediments are the major sources of nutrients-circulation in the reservoir.
KOWACO performed another study in 1990 on the water quality conservation using a ecosystem model for the Daechong reservoir.
(KOWACO, 1990b)
It was found that the sediments of which grain size smaller than silt were 72.5 percents including silt, clay and other organic matters. This fact informs that the deposited sediments causes the phenomena of deposition and releasing of the phosphorus contained in the deposited materials, and it shows a close relationship between reservoir sedimentation and water pollution.
Several models have been suggested to estimate the possible water pollution from the sediments, but there are some difficulties for practical application as they were mostly based on the theoretical application and far from real situations in some points of view. In this study more practical points were emphasized compared with the existing models. First it is necessary to understand the type or state of the pollutant contained in the sediment materials, and to obtain the ratio of the soluble and movable portion to the total amount of pollutant. Then it is possible to calculate the amount of organic matters which directly affects on the movability, and the carbon and nitrogen ration (C/N).
(Jun, 1988)
(Bostrom et al., 1985)
Figure 1 shows the process and procedure of the model considered in this study, and it illustrates the passes of phosphorus movement contained in the sediment. This figure was based on the Tchobanoglous and Schroeder's text (1987).
Figure 1 Movement Diagram of Phosphorus Contained in the Sediments
- 173 - 3.1 Characteristics of Sediments Deposited in a Reservoir
Deposition formation of the fine particles of the sediments depends on the kinds of suspended materials flowing into a reservoir, the velocity of the flowing current and the depth of water. This characteristics is one of the important factors to determine the concentration of organic matters, minerals, heavy metals and nutrients exist in the clayey materials. The smaller the grain size of the deposited materials, the more sediment becomes absorptive and contains more organic matters and micro size metal grains. This is because the smaller the grain size, the larger the surface area per unit weight becomes.
According to the previous research performed in 1989 on the Soyang reservoir (Jun ＆ Park, 1989), the sediment surface appeared yellowish in a large part of the crest of the deposited materials originated from the main stream. It was proved that thin layers of organic materials were covered on the deposited materials. Yellowish matters about one or two millimeters in thickness were frequently observed on the surface of the sediment in the parts originated from the branch streams, looking sometimes gray or brown to the about three centimeters in depth. However, the behind of this surface was remarkably distinguished from the upper parts appearing dark-gray or black colors. It was found that the surface of the deposited sediments in the Soyang reservoir was mostly considered to be mud in the parts originated from the main stream, and muddy sand in the parts originated from the branch streams. The components of clay, silt and sand had the percentages of 25.6, 46.3 and 28.2 respectively, which can be classified as Sansicl.
Table 3 shows the composition of the sediment of the Daechong reservoir which was investigated through KOWACO.
(KOWACO, 1990)
The results showed that the sediments were mostly composed of very small grains. The portion of clayey materials with very small grain size which mostly affects on the contents of organic matters, nutrients and metal grains was bigger in the midstream.
Table 3 Gradation Distribution of Daechong and Soyang Sediments
Samples of the sediments from ten points were selected within the Daechong reservoir to evaluate the contents of phosphorous according to the grain size distribution of the sediments.
(KOWACO, 1990)
The results of this test are shown on the Figure 2. The top and middle parts of this figure show the very positive relationships between phosphorous concentration and clay or silt contents of the sediment materials. However, the bottom part of this figure shows the negative relationship between phosphorous concentration and sand contents.
- 174 -
Figure 2 Relationships between Phosphorous Concentration and Sediment Contents, Daechong Reservoir in Korea
3.2 Sedimentation and Eutrophication in a Reservoir
The concentration and state of phosphorus in the sediment most greatly affect on phosphorous concentration in the aquatic ecosystem. Major source of phosphorus concentration in the aquatic ecosystem within a unpolluted reservoir is known to be sediments. The phosphorous as a factor limiting the primary production in the freshwater ecosystem is usually contained in the deposited sediments in a reservoir much higher than in water body. This phosphorous exist in a easily movable state to the water body subject to the environmental conditions. Phosphorus contained in the sediment usually moves from the sediment to the pore and then is released to water due to physical, biological and chemical interactions (Moniwa, 1978) as shown in Figure 1.
- 175 -
Because the phosphorous is the primary factor which controls the level of eutrophication of a reservoir, many models have been developed to estimate water quality of a reservoir using the phosphorus loading. According to a study on the effect of inland fish-farm on the reservoir water quality (KOWACO, 1991), the excessive critical loadings for the Soyang and the Daechong reservoir were calculated using the VollenweiderOECD model (1976) as 1.75 and 1.85 grams of phosphorous per square meters of reservoir surface area per year respectively. These amounts are equivalent to 78 tons of phosphorous per year to the Soyang reservoir, and 87 tons of phosphorous per a year to the Daechong reservoir. The possible inflows of total phosphorous may be increased to the much higher points than the critical values without controls on the inland fishfarms and erosion activities to prevent sediment inflows including waste water treatment to the extent of second level in order to remove phosphorous.
Estimation of possibility of water pollution by considering the reservoir sediments can be performed by measure of the types of phosphorus, the amount of phosphorus contained in the sediments. Table 4 shows phosphorus distribution classified by its type in the Soyang and Daechong reservoirs.
There are many types of phosphorus (P) contained in the sediments, but they are generally classified as absorbed-P which is absorbed to the surfaces of the very small sediment particles, nonapatite inorganic phosphorus (NAI-P) which is combined with ferrous or aluminum matters, residual-P which is directly related to the organic materials, and apatite-P which is usually contained in the apatite ores.
(Jun, 1988)
Table 4 Phosphorus Distribution Classified by its Type
In the respect of phosphorus type, there has been little studies on the movement of residual-P, but it has been a great concern because the phosphorus in the sediment is generally originated from the organic matters and much phosphorus can be released in accordance with degradation of organic materials. Especially, NAI-P has been the greatest concern in examining the activity of the circulating particles through the sediment and water body as it can be produced and released to a large amount even during a short period of breeding time. As shown in Table 4, the sediments of the Soyang and Daechong reservoirs contained 290.0 and 376.4 micro grams of NAI-P per one gram of dried sediment respectively. In other words, it shows that the chemical action is very important to eliminate the dissloved phosphorus in the reservoirs. In this case, there would be a great possibility of releasing much phosphorus to the water body as the acidity (pH) of the pore water or water body is increased because the NAI-P is adsorbed on the small particles with ferrous or aluminum matters.
The total amounts of phosphorus for the sediments of the Soyang and Daechong reservoir were 776.0 and 1206.1 micro grams per one gram of dried sediments respectively. These figures are lower than 1800 to 2530 for the Yunoko reservoir in Japan (Hosomi, 1981), and 2550 for the lakes in the Wisconsin area, USA.
- 176 -
(Filles 5, Swanson, 1975)
The total amounts of phosphorous for the sediments of other reservoirs in Korea are similar to the values of the Soyang reservoir.
Sediment of the Soyang reservoir indicates a little higher level of NAI-P contents than the others. especially the contents are much high on the top portion of the deposited sediments near the fish farms. The NAI-P is very sensitive in the anaerobic conditions and showed a great possibility of releasing from the sediments in accordance with exhaustion of dissolved oxygen in the main stream area. and with the rising of the acidity in the branch stream near the fish farm in the Soyang reservoir.
3.3 Nitrogen, Organic Carbon and Organic Matters in Sediment
Organic matters contained in the sediments are apt to supply the nutrients to the water body after decomposition, and cause the phosphorous to be discharged from the sediments to the water body by increasing the acidity (pH) of the water. The organic degradation which deeply affects on the eutrophication of a reservoir can be estimated by finding the carbon and nitrogen ratio (C/N ratio) to compare it with an average C/N ratio of photoplankton (5.6 for the Daechong).
If the ignition loss is high and the C/N ratio is small, then it means that the organic concentration of the deposited sediments is high and is less decomposed and it may cause bad results to the water quality. The bad effects are, for instance, some organic matters are released from the pores of the sediments to the water body, then increases the organic concentration and exhausts the dissolved oxygen resulting in anaerobic state. In this state, the pollutants such as heavy metals which were adsorbed in the surface of clay are released or the obnoxious odor comes out, and the eutrophication is speeded up by the released nutrients.
The C/N ratios of the deposited organic matters of the Soyang and Daechong reservoirs were 6.3 and 14.2 respectively. These figures show that organic degradation has been far more progressed in the Daechong than in the Soyang (Jun, 1989; KOWACO,1990). Therefore, sediments in the Soyang reservoir have the larger possibility to release the phosphorus originating from the organic degradation. Table 5 shows the results of the ignition loss test, the C/N ratio of the sediments of the Soyang and the Daechong reservoirs. The figures in this table were based on the average values from the ten sample tests. The C/N ratio of the organic matters in the sediments of the Daechong is much higher than that of the Soyang. This means that the decomposition of organic matters of the Daechong has almost completed and the phosphorous discharge is not significant as much to be worried about.
Table 5 Approximate Sediment Composition of Ignition Loss
- 177 -
4. CONCLUSION AND SUGGESTIONS
The problem due to the sediment deposition in a reservoir has been generally considered in the view of its quantity. However, the quality of the inflowing sediments hag to be included in addition of quantity problem in the planning or management stage of water resources facilities according to their given pruposes. For the large scale reservoirs in Korea, the quality problem has been becoming more important than quantity because the amount of sediment inflows is relatively small compared with the that of other significant countries or the designed life time of the water facilities, while water pollution problems like reservoir eutrophication have gathering strength due to very fast industrialization.
The followings are the results and conclusions of this study on the water quality assessment through reservoir sedimentation and its analysis. However, further investigations have to be continued in order to find out clear relations between water quality and reservoir sedimentation.
(1) Reservoir sedimentation is not significant in large scale reservoirs in Korea at least in a view point of quantity even if the recorded sediments are greater than the estimated values at the planning stage by varying one or two times.
(2) For the phosphorous which is deeply related to the eutrophication of a reservoir, it was found that a large amount might be released to the water body from the deposited sediments according to its existing states when aqua-environments are changed.
(3) It was found that the concentration of phosphorous would be affected by the level of degradation of the organic matters existing in the sediments. It is necessary to take into accout this point in reservoir water quality management.
(4) The concentration of phosphorous of the deposited sediment materials are dependent to the gradation of the sediments. The larger concentration of phosphorous are found in the case of the larger amount of smaller components like silt or clay.
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA"
- 178 - (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
RESERVOIR DEPOSITION AND ITS CONTROL, INCLUCING MITIGATION OF ANY ADVERSE EFFECTS ON THE RIPARIAN AREA
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Director General. Pakistan Water & Power Development Authority, Alluvial Channels 참여연구원 : Observation Project (ACOP), Lahore, Pakistan Ahmed Masud Choudri Secretary, Pakistan Council for Research in Water Resources (PERWR), Ministry of Science 참여연구원 : & Technology, Government of Pakistan, Islamabad. Pakistan Naseer A. Gillani
Summary
Reservoir sedimentation relates to age-old problem of loss of storage capacity in reservoirs that has assumed an enormous dimension with the accelerated construction of dams in the world since early fifties. With the increasing age of existing reservoirs and their added importance in meeting world's growing demand for agriculture and energy, a new emphasis is emerging in the planning and design of dams to provide some form of mitigation, if not elimination, of storage loss to sedimentation.
- 179 - The processes of sediment deposition in storage reservoirs are understood to the extent that engineering estimates for the gross rate of sediment accumulation can be made with some degree of certainty. The processes of re-entrainment of these deposits during flushing are, however, inadequately understood. Although sporadic efforts have been made at some dams to flush the accumulated deposits, the results have been uncertain mainly because the underlying phenomena have not been investigated. Presently, it is not possible to develop acceptable estimates of sediment sluicing at large reservoirs during operational or deliberate flushing cycles.
This paper discusses the processes that lead to reservoir sedimentation, a review of the approaches that have been developed in the past to predict the trap efficiency and the distribution of volume of deposits in reservoirs, various methods of mitigation of reservoir siltation, and also outlines the research needs particularly prototype research on the deposition, re-entrainment and sluicing of sediment from large storage reservoirs.
Also, a review of the sediment data for the Tarbela Reservoir (Pakistan) obtained to study sedimentation problems of the Tarbela Dam has been made.
1. INTRODUCTION
1.1 In the field of water resource engineering reservoir siltation stands out as one of the most serious economic problems. Storage reservoirs enhance usable water resources by re-regulating flood flows. At the present level of development, the reservoirs constitute a vital component of the world water resource. The volume of reservoirs in the world is around 4,100 cu.㎞ (about 3400 maf), and they regulate about 1,900 cu.㎞ (about 1600 maf) of flow volume (3). It has been said that all reservoirs are doomed to die. This is due to the loss of their storage capacity to sedimentation. Assuming a 100 year average life of reservoirs (Lake Mead; 350 ＋ and Tarbela ; 40 ＋ years), we are losing about 41 cu.㎞ of storage capacity per year. The average cost of storage has been estimated by Ambroggi (1) as US＄ 120 million per cu. ㎞. This seems on the low side because the better storage sites in the world have already been developed. Even at this level, the replacement cost of world average loss of reservoir capacity is about US＄ 13 million per day. This is a significant financial stake and consequently we see an increasing interest in sediment sluicing and scouring in reservoirs.
1.2 Unfortunately, it is not widely recognized that effective sediment sluicing requires a significant lowering of the reservoir level for prolonged periods of time. Thus, there is a trade off between the loss of power and storable quantity of water caused by the sluicing operation and the benefits accruing from the reclaimed storage capacity. The future will place greater stress on means of prolonging reservoir life in both the existing and planned projects. This also means a need for better understanding of the processes of reservoir siltation and re-entrainment of trapped sediment during sluicing or scouring operations.
1.3 The time-rate of siltation of reservoirs varies with their design and geographical location. Hoover dam, since its closure in 1935, has been losing gross storage capacity at the rate of 0.3 percent per year, whereas, measurements in Tarbela Dam (Pakistan) show that it is losing capacity at a rate of 1.5 percent a year. There have been some notable rates of reservoir siltation at other sites. For example, the 250 high Warsak Dam on Kabul river in Pakistan lost 18 percent of its storage voume in the very first year's operation. The world-wide average for the loss of reservoir storage due to sedimentation is estimated to be 1 percent with a consequent annual economic loss of around
- 180 - US＄ 5 billion in replacement costs alone.
1.4 The processes of sediment deposition in storage reservoirs are understood to the extent that engineering estimates for the gross rate of sediment accumulation can be made with some degree of certainty, see Brune (1963), Churchill (1947), Heinemann (1981) among others. The processes of re- entrainment of these deposits during flushing are, however, inadequately understood. Although sporadic efforts have been made at some dams to flush the accumulated deposits, the results have been uncertain mainly because the underlying phenomena have not been investigated. Presently, it is not possible to develop acceptable estimates of sediment sluicing at large reservoirs during operational and deliberate flushing cycles.
2. REVIEW OF RESERVOIR SEDIMENTATION
2.1 Storage reservoirs break the energy gradient of river flow and interrupt the downstream transportation of sediment load. The behaviour of sediment load as it enters a reservoir depends on its particle size distribution. The "coarser" fraction of load, which may include some silt fraction and flocculated clays settles first, The "finer" fraction, comprising fine silt and clay, deposits farther down in the reservoir. Under appropriate hydraulic and densimetric conditions. the finer fraction may even develop a density current and travel right up to the dam. As such the reservoirs exert a strong hydraulic sorting on the incoming sediment load. At any point in time, however, the reservoir deposits are characterized by two distinct features ; the delta and the bottom set beds.
(FIG.1)
The controlling processes for these two features are somewhat different. The development and growth of a delta is akin to the phenomenon of aggradation/degradation in a non-prismatic channel whereas, the bottom set deposits are controlled by a combination of "desilting basin" process and density currents.
2.2 The engineering interest in reservoir sedimentation concerns three physical aspects; overall volume of trapped sediment; distribution of deposit volume and distribution of sediment particle size within the reservoir. The loss of usable storage capacity due to sediment deposits reduces the efficacy of a reservoir to regulate the flow and to provide flood control. The distribution of volume of deposits determines the relative impact of trapped sediments on the usable storage and the distribution of particle size effects the density of deposits as well as the potential damage caused by the ingress of sediment into the power outlets. The reservoir deposits are never horizontal. The delta establishes a top slope that is flatter than the nascent bed slope of the river but it is nevertheless finitie. So that, the usable capacity is reduced even before the dead storage provided for the sediment deposits is exhausted. As an example, the 1980 survey of Tarbela reservoir showed that after 6 years of operation with more than 78 percent of dead storage still available, 44 percent of the deposit lay in the usable storage zone.
- 181 - 3. REVIEW OF APPROACHES FOR PREDICTION OF TRAP EFFICIENCY AND DISTRIBUTION
A number of approaches have been developed in the past to predict the trap efficiency and the distribution of volume of deposits in reservoirs. These are reviewed below.
3.1 Empirical Methods
The empirical approach to reservoir sedimentation is typified by methods of predicting trap efficiency such as Churchill (1947), Brune (1953) and Heinemann (1981). These methods are based on a qualitative understanding of the sedimentation process. The independent parameter is capacity: inflow ratio (Churchill uses a somewhat different and more relevant parameter - Sedimentation Index) which is used to graphically read the proportion of incoming sediment that will be trapped in the reservoir. Brune's curve, the more popular of these methods is based on the data obtained from 44 reservoirs covering drainage areas of 1.52 - 184,600 sq. miles. The capacity inflow ratio in the data varied from 0.0016 to 4.65 and the trap efficiency from 0 to 100 percent. There is some evidence (Heinemann, 1981) that Brune's curve overestimates the trap efficiency in small reservoirs. For large reservoirs on the other hand, it may under-estimate the trap efficiency. It is easy to see how the trap efficiency for a given capacity ; inflow will vary depending on the shape of reservoir. Brune (1953) recognized the role played by reservoir shape in determining its trap efficiency. But, in view of the scope of usually available data, his method does not provide for this variable.
It is apparent that the empirical methods for predicting trap efficiency of reservoirs are simplistic and will provide rough estimates within a range of ±20 percent. These methods are useful in approximate analyses and in the planning studies of small reservoirs and ponds that do not warrant detailed studies. They are, however, not adequate for the planning or design of large reservoirs.
Two other empirical methods are available for use in conjunction with the Brune's curve; a time- dependent dry density for the deposits (Lane and Koelzer, 1943) to convert the weight of incoming sediment to volume and a relative area method to estimate the distribution of volume of deposits(Borland and Miller, 1958 and Miller, 1962). Usual reservoir sediment surveys are concerned with the volume of deposits and they do not explicitly measure the specific weight of deposits. The average density of deposits in such a case is derived from a conversion of incoming sediment load to measured volumes of deposits.
It is concluded that the empirical methods are simple to apply but they may result in errors of up to ±20 percent in the trap efficiency. 16 percent in the initial density of deposits and a significant error in the distribution of deposits.
3.2 One-Dimensional Transient models.
The phenomenon of bed level transients has been studies by a number of workers(Mahmood, 1975; Mahmood and Ponce, 1976; Chen et al, 1975, Simons et al. 1975; Chen et al. 1978 among others). The models developed by these workers are based on the numerical solution of Saint Venant unsteady flow equations for the liquid phase and the mass conservation equation for the sediment load. The sediment transport function used in these methods is one of the commonly available
- 182 - functions (Mahmood, 1982) while, numerical solutions vary from coupled-implicit to un-coupled- explicit solutions. The most popular of the transient models is U.S. Army Corps of Engineer's HEC-6 model (1977). This model is of uncoupled explicit type and provides for variable stream geometry and re-entrainment of previously deposited sediments. In HEC-6, it is possible to study the transport and deposit of individual size fractions of the incoming load. This model has been modified (Pakistan Water and Power Development Authority, 1984) for reservoir sedimentation studies of proposed Kalabagh Dam.
The main difficulty in the available transient models arises from the fact that none of the available bed material load functions has been tested on deep reservoirs flows or for the degree of non-uniformity of flow experienced in large reservoirs. Another drawback in the one-dimensional transient Models is that they cannot realistically treat the finer sediment load fraction with their bed material load equations. In most sandbed rivers, this is a serious handicap because 50 percent or more of the total load in these streams lies in clay-silt size range. In general, the bed transient models will adequqately Simulate the sedimentation processes over the delta but downstream of that their reliability is questionable. These difficulties have given rise to another type of models that treat the reservoirs as desilting basins.
3.3 Desilting Basin Models
Hurst and Chao (1975) abandoned the one-dimensional transient model in their planning studies for Tarbela Dam. Instead they adopted Camp's (1944) trap efficiency curves for desilting basins. Such a model will most likely succeed in the early life of reservoirs that do not experience significant drawdown. When the delta has formed in the reservoir and at least part of the reservoir flow is of riverine type, the method will fail because desilting basin models are based on the assumption that the lower boundary of the basin is an absorbing boundary with no re-entrainment. The operational experience at Tarbela shows that Hurst and Chao's analysis under-estimated the stream-wise progression of delta. The actual delta crest after 9 years operation was located about 12 miles upstream of the dam instead of the predicted 30 miles. This is directly attributable to the aforementioned reason.
3.4 Sediment-Diffusion Models
A sediment diffusion model has been used by Merrill (1980) to simulate the sedimentation in three reservoirs in Nebraska and Illinois in which 90 percent of sediment load consists of clay-silt sizes. This model is based on two dimensional diffusion equation solved by an explicit numerical scheme. The conceptual approach of Merrill's study is appropriate and it shows that the diffusion model can be applied to reservoir sedimentation where the primary sediment load is in clay-silt range. However, the indirect empirical determination of key parameters of the model from observed data makes its application to new reservoirs difficult.
3.5 Density Currents.
Storage reservoirs frequently develop density stratifications due to temperature, salinity and turbidity
- 183 - differences between different layers. The river flow into a stratified reservoir may develop into an overflow, interflow or underflow depending on its density relative to that of different layer.
(FIG.2)
Of the various classes of density flows in a reservoir, turbidity currents are the most important from the point of view of reservoir sedimentation. For example, if the river flow develops into a density current that travels along the bed of the reservoir, it can deliver large concentrations of sediment to the downstream end which may then be ejected. Only a few measurements of turbidity currents have been reported in literature (Howard 1953 and Geza and Bogich, 1953) although, it is commonly assumed that the accumulations of fine sediment near the dams are caused by this phenomenon.
The possibility of tapping trubidity currents to pass sediment through reservoirs, led to an impassioned plea by Bell (1942) to use selective withdrawals from storage reservoirs to manage reservoir siltation as well as the fine material loading of the releases. A long term proven success of this method, however. has not been demonstrated so far.
4. RESEARCH NEEDS
As a result of the preceding review, a number of research and development problems suggest themselves.
These are listed below:
4.1 Sediment Yield
Sediment load carried by the flow is the primary variable that determines the rate of sedimentation in a reservoir. This is also the first area where research is needed to improve our understanding of processes involved in the generation and delivery of sediment from large basins. The role of sediment sources and sinks has not been studied in large basins and the effect of watershed management practices has not been critically evaluated by controlled experiments. Prototype research on the fate of eroded material in its journey to the outlet and the efficacy of both structural and non-structural measures is needed. This research will enhance the possibility of controlling sediment yield from the drainage basins.
4.2 Sediment Diffusion in Deep Flows
For want of any better information, the sediment transport and deposition functions used in the mathematical modelling of reservoirs are those developed from laboratory flumes, canals and rivers.
- 184 - Most likely, the decay of trubulance intensity significantly changes these processes in deep reservoirs. This would be especially true of the silt and clay particles, that dominate the sediment load in rivers. Measurements of flow field and sediment concentration profiles in large reservoirs are needed to develop appropriate hydrautic and sedimentation functions.
4.3 Sediment Re-entrainment
Sediment flushing is a useful method to rid of the existing deposits. It becomes more attractive when the silting up of a reservoir has reached an advanced stage. The efficacy of flushing depends on the rate with which the deposits can be reentrained by the flow. Existing knowledge, mostly gained from laboratory studies and theoretical investigations, suggests that rate of re-entrainment in reservoirs will be strongly effected by the clay content of deposits; mineralogy of clays and the chemical regime or water. for sand particles, the rate of re-entrainment depends on the velocity distribution within the reservoir and especially, that near the bed. The flow in reservoirs is strongly non-uniform, much more so than can be expected in streams. Processes of and relating to reentrainment of deposits have not been investigated in reservoirs. Prototype research in this area will be highly rewarding.
4.4 Density Currents
In the future, reservoirs will be monitored and operated to manage their thermal, salinity and sediment content in addition to the water flows. Theoretical aspects of density currents have been primarily developed from laboratory studies. Their validation on prototype structures has not been attempted so far. Field data on sediment related density currents are scarce. Research on the formation, behaviour and fate of density currents in reservoirs is needed. The results will be directly useful in alleviating the rate of sedimentation of existing reservoirs and will help in planning and design of future structures.
4.5 Empirical Methods.
Currently available empirical methods for the prediction of trap efficiency and distribution of deposits are 20-30 years old. In the meantime, and extensive data base has developed on the gross behaviour of reservoirs. Theoretical understanding of reservoir siltation has also improved in this period. Empirical methods will continue to be used to provide preliminary analysis for the large and the final analysis for small projects. The time is now ripe to develop a second generation of empirical methods with expanded scope and improved accuracy.
4.6 Mathematical Models
Presently available mathematical models for reservoir siltation are patterned after channel flow models. In general, the hydraulic and sedimentation processes in reservoirs are strongly three-
- 185 - dimensional and stratification can have a major effect on these processes. Due to their speed, declining costs of computer use and their potential to predict micro details, mathematical models will find much greater use in the future planning, design and operation of reservoirs. A need exists to develop more comprehensive mathematical models than the present one-dimensional variety.
5. REVIEW OF SEDIMENT DATA FOR TARBELA RESERVOIR (PAKISTAN)
5.1 General :
The 450 ft. high, Tarbela Dam in Pakistan with a total earth-rock fill volume of 186 million cu yds., is the largest dam in the world. The gross storage capacity of Tarbela lake, 11.1 maf (at construction) regulates about 60 maf of annual river of Indus at this site. During its annual cycle of operation, Tarbela reservoir level falls by more than 200 ft and produces a large scale redistribution of sediment deposits. The shape and migration rate of delta in Tarbela reservoir is unusual when compared with other dams and is entirely attributable to the reworking of deposits during prolonged periods of drawdown.
5.2 Tarbela Sedimentation :
High temperatures in summer cause snow and glaciermelt in the upper Indus catchment (about 65,500 sq. miles) and generate high discharges, which bring enormous amount of sediments to Tarbela Reservoir.
The sediment inflow at Tarbela reservoir is carried by River Indus and its tributories namely Shyok, Gilgit and Hunza. Figure-3 shows the contribution of sediments by these rivers in the Indus Drainage Basin.
Most of these sediments are trapped in the reservoir and deposited in the upper reaches. During subsequent months of low discharge, the reservoir depletes and the sediments deposited in the upper reaches are reworked and carried downstream within the reservoir. A part of these sediments goes out from the reservoir through the outlets while bulk of the sediments is being accumulated in the form of major delta. Total sediment load computed from mean inflow at Tarbela during the past 17 years(1974-1990) using Peshawar University Rating Curves.
각주 001) is 3,781 MST. The sediment load estimated from hydrographic survey for the same period is 3,231 MST. These figures are fairly different.
- 186 -
Nearly 42％ of this sediment has partially filled the dead storage of reservoir below EL. 1,300 in the form of a delta which is being built by the incoming sediments every year and advancing towards the Main Dam at an average rate of about half a mile per annum. During the year October 1989 to September 1990, the minimum reservoir level remained high at EL. 1,381.38 and the Hydrographic Survey showed no delta advancement.
5.3 Delta Advancement :
Delta advancement, period of minimum reservoir level below EL. 1,320, the annual sediment load computed from hydrographic survey and the river inflows from 1979 to 1980 is tabulated below.
It is evident from the above Table that in 1985 the longer duration, for which the reservoir remained below EL.1,320, had caused maximum advancement of delta towards the dam. and the shorter duration in 1990 has resulted in no advancement of delta.
- 187 -
FIG.1 PROFILE OF A TYPICAL RESERVOIR DELTA
FIG.2 UPSTREAM END OF RESERVOIR SHOWING FORMATION OF DENSITY CURRENT
FIG. 3
- 188 - "REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
RESERVOIR SILTING PATTERNS FROM A TWO-DIMENSIONAL HYDRODYNAMIC MODEL
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Asspciate Professor, Department of Agricultural 참여연구원 : Engineering, Seoul National University, Seoul, Korea Seung W. Park Graduate Student, Department of Agricultural 참여연구원 : Engineering, Seoul National University, Seoul , Korea Chang E. Park Senior Engineer, Development Planning Division, Rural Development Bureau, Mini stryof 참여연구원 : Agriculture, roreEtry and Fisheries in Korea Bong H. Lee
Summary
Sedimentation distributions within a reservoir are important to the change in the stage-capacity
- 189 - relation. Typical approaches to sedimentation pattern analyses are based on empirical relationships with judicious assumptions and simplifications. Apart from those approaches, an attempt was made in this study to apply a two-dimensional hydrodynamic model coupled with a hydrologic model to the simulation of reservoir deposition patterns some years after the construction. The hydrodynamic model depicts vertically averaged velocity distributions within a reservoir, which are used to define sediment transport and deposition rates. Seasonal changes in the inflow to and storage levels at a reservoir are defined using a hydrologic model. The combined effects of hydrodynamic and hydrologic conditions on silting patterns were simulated and the results compared with surveyed data from a small, hill-sided reservoir. Some discrepancies between the simulated and surveyed ones were observed in the silting depths. It was found, however, that the models are capable of providing with some insight to the silting patterns.
INTRODUCTION
Predicting deposition patterns in a reservoir is important to adequately define the stage-capacity relation, which is needed for proper management of the water resources. It is also applicable to identify the location of deposit.
Deposition patterns are closely related to the hydrology, hydrodynamics, sediment characteristics, and the bathymetry and their interactions. Sediment yield at a reservoir is primarily governed by the fluctuations and magnitudes of inflow rates. The flow characteristics are important to sediment transportation and deposition, which are dependent upon sediment characteristics such as sizes, weights, and shapes. Bathymetry controls the hydrodynamics of water body and eventually the silting processes.
(Thomas E. Croley 2, 1978)
Sediment yields for a given reservoir are often estimated using one of the following approaches; 1) gross erosion and sediment delivery ratio methods, 2) discharge-sediment rate curve methods, 3) empirical relationships based on reservoir sediment survey data. The yields are then multiplied by sediment trap efficiency to obtain the reservoir sediment yield to be deposited. However, the results do not provide with the distribution patterns within a reservoir.
(USBR, 1987)
U.S. Bureau of Reclamation proposed to use a Lara approach to predict the distribution of reservoir sediment deposits. Using a judicious assumption that there is a reasonable relationship between the percentage of total reservoir deph and sediment volume. an area-reduction procedure was proposed, which may be used to define the changes in the capacity curve of a reservoir.
(USBR, 1987)
Rational approaches have been introduced that consider simplified hydrodynamics of water body from sediment-laden inflow. These approaches are, however, limited to short time spans, while actual reservoir is subjected to historical realization of natural hydrologic phenomena. Thus, it appears that the approaches limit themselves purely to an event model rather than long-term changes in distribution patterns.
- 190 -
This paper attempts to couple hydrodynamic approach with hydrologic ones to predict silting patterns in a reservoir during and after the life time. The hydrological behaviours of a reservoir throughout its life time are first to be simulated using a reservoir operation study, and the results grouped into several combinations of daily inflow and storage levels. For each combination of inflow and storage level , hydrodynamics of water bodies are simulated with a sediment transport model, which result in accumulated silting rates for a few days. The silting rates for each event are again accumulated, based on the hydrologic characteristics, to obtain eventual silting depths within the reservoir. This paper also describes the results of the field application of the proposed approach to a small reservoir that has the sediment survey data.
TEST SITE
The Banweol reservoir was selected in this study as the test site. It is located about 30 ㎞ southwekt of Seoul, the Capital city. The watershed area is 1,220 hectares, which is divided into mountaineous area of 79 percent, paddy field of 20 percent, and rural village areas of 6 percent. The surface area at the full capacity is 46 hactres. The maximum storage capacity is 139.6 hectare-meter, and it supplies irrigation water to 405 hectares of paddy rice.
Hydrologic characteristics of the Banweol watershed have been monitored for past seven years, which include incremental rainfall data, streamflow data, reservoir water level fluctuations, and irrigation intakes. The details of hydrologic monitoring systems are depicted in Figure 1.
- 191 -
Figure 1. Banweol reservoir and hydrologic monitoring systems at the watershed The sediment survey for the Banweol reservoir was conducted two times during 28years after the construction. The annual average sediment yield was found to be about 20,000 ton/year.
DAILY RESERVOIR OPERATION SIMULATION
A mathematical model that can simulate daily water balance for a reservoir was developed and tested successfully with the data from the test site. The model constitutes three major components of daily inflow simulation, irrigation demand and intake simulation, and storage accounting for the reservoir. The brief descriptions of the components are as follows.
- 192 -
Daily Inflow Simulation
A hydrologic model was used in the study that can simulate daily inflow rates from a given watershed. A tank model was selected and modified, and the parameters calibrated. The schematic of the tank model used in this study may be schematically depicted in Figure 2, where the streamflow is conceptually divided into three components of surface flow, interflow, and base flow. Each component is controlled by the assumed storage level and parameters in a following form. where Q ＝ streamflow component, Ai ＝ parameter, ST ＝ storage parameter, H ＝ parameter for the outlet at the tank.
Figure 2. Tank model developed for daiIy-runoff relationship The storage level of a tank component at a given day is determined from the storage at the previous day minus the outflow, and evapotranspiration and infiltration lessee. The evapotranspiration is determined from daiIy potential value adjusted for surface cover conditions. The infiltration is defined as a function of the storage level similar to Equation 1.
Irrigation Intakes
Daily irrigation intakes form the reservoir was assumed as a function of irrigation demands, minimum releases for delivery losses, and farming requirements. Irrigation efficiency was also considered. Thus, the intake rates for a given day is defined in Equation 2.
- 193 -
where OR ＝ intake rates ( ㎥/day ), C ＝ coefficient for unit conversion, REQ ＝ irrigation demands ( ㎜/day ), MR ＝ minimum releases for delivery losses ( ㎜/hr ), TS ＝ farming requirements ( ㎜ ), L ＝ delivery loss rates ( ％ ), A ＝ irrigation areas( ㏊ ).
Simulation Results
The daily reservoir operation model was tested with the data from the teat site. The simulated daily inflow rates, intakes, and water level fluctuations were separately compared with the observed ones. A comparison between the simulated and observed daily fluctuations of the reservoir water levels is shown in Figure 3, which shows a good agreement between the two. Separate works have been done to test the components of the water balance model with a similar success.
Figure 3. Observed and simulated reservoir stages, Banwoel reservoir, 1986 Twenty-one year meteorological data from 1966 to 1986 were used to simulate reservoir operational characteristics. The resulting daily inflow rates and water level fluctuations were grouped as in Tables 1 and 2, using a state variable concept. Five storage states were used to classify daily water level fluctuations (Table 1) and eight states for inflow rates.
(Table 2)
Using Table 1 and 2, the combination of storage stages and inflow state for a given day is counted and the results are given in Table 3. Table 3 shows that the water levels were maintained at the level greater than 80 percent for more than 70 percent of the simulated periods, while inflow rates less than 5000 ton per day were predominant with 54 percent of the periods. Table 3 also indicates that the combinations of inflow rates and water levels vary considerably during the time span. Such the variability of the combination should be considered in the prediction of silting patterns, since the
- 194 - hydrodynamic characteristics at each combination may be significantly different.
Table 1. Stage classifications of water level and capacities for Banwoel reservoir
Table 2. State definition for inflow at Banwoel reservoir
Table 3. Number of days in each state and stage (1966 - 1986)
- 195 -
HYDRODYNAMIC SEDIMENT TRANSPORT MODEL
In an effort to simulate water and sediment circulations and silting rates within a reservoir, a depth- averaged two-dimensional hydrodynamic model was coupled with a sediment transport model. The details of the transport model will be described.
Basic Equations
Basic equations for a depth-averaged two dimensional flow are the continuity equation and momentum equations. And sediment transport may be described using a sediment continuity equation.
The continuity equation for a two-dimensional water circulation may be described by Equation 3.
where h ＝ water depth with respect to a reference level. u and v ＝ depth-averaged velocities in x- and y- directions, respectively, a ＝ verticql displacement of the reference level form a datum, t ＝ time.
- 196 -
The momentum equations for x- and y- directional flows are described in Equation 4 and 5.
where g ＝ gravitational acceleration, Ω ＝ Coriolis constant, V ＝ absolute flow velocity and V ＝ √(u² ＋ v²) , n ＝ Manning coefficient.
The sediment continuity equation may be expressed in Equation 6.
where C ＝ sediment concentration ( volume/volume ), and S ＝ sediment suppply rate from a source or sink.
The sediment transport may be limited either by the transport capacity of the flow or the sediment supply rates. Sediment rates exceeding the transport capacity are subject to deposition. The flow may pick up sediment from the bed when the transport capacity exceeds the available sediment transport rates. This concept of sediment continuity has been used in upland erosion simulation and channel bed stability studies.
The sediment transport capacity by the flow may be expressed using a sediment transport equation. In this study, Toffaleti equations are used as an index for the capacity. The equations define the total loads to be the sum of four sediment transport rates that describe the suspended sediment loads at upper, mid, and lower regions, and the bed loads. The details of Toffaleti are available in many of sediment transport textbooks.
(Vanoni, V. A, 1975)
The deposition rates are determined from the differences between the sediment transport capacity and the available sediment load.
- 197 - where S ＝ deposition rate, qs ＝ sediment load, Tc ＝ sediment transport capacity, αs＝ reaction constant ( 0 ＜ αs ,＜ 1.0 ). The reaction constant accounts for the physical responses of deposition such as the effects of fall velocities of sediment particle.
Numerical Approximations
An alternate direction implicit method was used to obtain the numerical solutions of the flow and sediment transport equations. Initial and boundary conditions were determined from the inflow and water level interaction matrix of Table 3. For each combination, initial conditions are the representing water level while flow velocities in the reservoir were set to zero. The boundary conditions were the inflow rates from three incoming streams while the water level is below the elevation of the spillway crest. When the water level is equal to or greater than the crest elevation, free falls were assumed at the spillway. For simplicity, the sediment concentrations of incoming flow were assumed to be 100 ppm, which may be seen as unrealistically low for high flow. The lack of measured sediment concentration data was one reason not to adjust the values as related to flow rates. It was believed. however, that adjustments would not insure better simulation results considering the limitations in the duplications of historical events.
The computational schemes may be seen in Figure 4. The computations begin with the determinations of flow velocities and water depths at each computational grid of 20 × 20 m in size. The results were then used to define the sediment transport and silting rates. Accumulated deposition depth at each grid was kept for each run for a period of 48 hours. The details of the simulation are available in Lee(1990).
(Lee, B. H, 1990)
- 198 -
Figure 4. Flow chart of two-dimensional sediment transport model The total depths of sediment deposition at all grids were defined from the weighted average values of accumulated silting depths. Converting the weighted values into the historical time spans resulted in the distributions of silting depths for the reservoir.
RESULTS AND DISCUSSIONS
Velocity Fields
An example of the flow velocities at the reservoir that were computed from simulated horizontal components is shown in Figure 5. Figure 5 is the case for the high water inflow to the full water level (stage 1 × state 1 in Table 3). It is shown that the incoming streamflows from three mouths at different locations which are identified as A, B, and C in Figure 5, induce overall circulation of water body. It is
- 199 - interesting that horizontal displacements of water body take place near the spillway, which contributes to the overall flow development in the reservoir.
Figure 5. Distribution of water velocity at stage 1 x state 1 (after 2 hours) The simulation results were subject to a verification. In this study, the simulated water levels at the gaging station was compared to the observed values. No direct comparisons between simulated and observed velocities were possible due to the lack of field data. However, an excellent agreement of the water level data may be sufficient for the verification purposes.
Sediment Concentration and Silting Patterns
The distributions of simulated sediment concentrations at the simulation run for the velocity fields described above may be seen in Figure 6. The simulated sediment concentrations were relatively high near the mouths of incoming streams and gradually decreased toward the center of the reservoir.
Figure 7 shows that the silting depths are in a pattern similar to the concentration distributions. Heavier silting is seen near the mouths, while it becomes smaller near the center. However, some degrees of silting were also observed near the spillway. It was apparently due to relatively strong velocity fields in the areas. This may indicate that the silting can be widespread within a reservoir of relative small size.
Comparisons to Sediment Survey Data
As stated earlier, silting patterns for each combination of inflow rate and storage stage were simulated. The results for each computational grid were then synthesized for the number of days as in Table 3, by multiplying the number to the results from a 24 hour simulation. The synthetic silting depths at each location for the life time are summarized in Figure 8.
Synthetic silting depths at the reservoir were compared with the sediment survey data of Figure 9. Unfortunately, the survey was made a few years after dredging of the reservoir at the upper part. This prevented direct comparisons between the two. The overall picture may be cross checked between the two. Firstly, the synthetic results show lighter silting depths in general, which is due to the failure in using heavier incoming sediment concentrations of high inflows. Seceondly, the results are in general qualitatively similar to the surveyed, except the portions of shorelines. The differences may be
- 200 - due to the wave-induced shoreline erosion and land erosion directly slided to the bottoms of the reservoir. This may lead to a conclusion that the simulation is apt to give a qualitative description of the reservoir silting patterns where shoreline erosion contributions are limited.
SUMMARY AND CONCLUSIONS
An attempt was made to simulate the silting patterns of a tested reservoir of 21 years old, using hydrologic and hydrodynamic sediment transport models. The hydrological behaviours of a reservoir were simulated using a daily reservoir operational model. Variations of inflow rates to and water levels at the reservoir during a specific period were simulated and the results were grouped into several combinations using a state variable concept. For each combination of inflow and water level, a depth- averaged, two-dimensional hydrodynamic sediment transport model was applied to depict the velocity fields and sediment deposition pattern. The results were then used to synthesize the silting depths for the period.
The synthetic silting pattern was compared to the sediment survey data. The results were found to be qualitatively comparative to the survey data within the objectives of the study. For accurate simulation, incoming sediment rates and shoreline erosion components need to be considered.
Figure 6. Distribution of sediment concentration at stage 1 x state 1(after 24 hours)
- 201 -
Figure 7. Distribution of sediment deposited on the bed at stage 1 x state 1(after 24 hours)
Figure 8. Result of sediment transport simulation at Banwoel reservoir for 21 years (1966 - 1986)
Figure 9. Deposition of sediment in Banwoel reservoir for 21 years (1966 - 1986) ; Surveyed at 1986
- 202 -
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
HUAI KHO RESERVOIR SEDIMENTATION
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Hydrologist, Sediment Branch, Hydrologr 참여연구원 : Division, Bankok 10300, Thailand Somkuan Yimsricharoenkit Hydrologist, Research And Applied Hydrolgy 참여연구원 : Branch, Hydroloar Division, Bangkok 10300, Thailand Rungsan Sirayayon
Summary
Most areas in Northeastern part of Thailand are quite dry, especially in the middle plateau, which calls
- 203 - Tung Kula Rong Hai, is extreamly drought nearly through the year. Huai Kho reservoir, Maha Sarakham Province, has been impounded there since 1968, with capacity 31.5 mcm, it's irrigated areas about 3,360 hectares. Sediment deposition was surveyed and resurveyed in 1978 and 1984 the Huai Kho's capacity were 30.0 mcm and 25.0 mcm, respectively. The rapidly increasing of sediment inflow, 0.09 mcm/yr (1968-78) and 0.93 mcm/yr (1978-84), has been induced many practical soil ＆ water conservation plans. If the rate of deposition can not be alleviated, the operation's life of the reservoir may be 20 years. The more time the sediment problems take to be happended, the more time erased the solution can be solved. Even God may be cried, if he has to cope with these kind of problems.
1. INTRODUCTION
Huai Kho reservoir is located at Lattitude 15˚ - 49′- 23″ North and Longtitude103˚ - 02′- 48″East, which is Na Chuak district, Maha Sarakam province, Northeastern part of Thailand.
(Fig.1)
It has been operated since 1968 and served more than 3,000 hectares of irrigation areas. It was a earth-dam which had some important figures as Table 1.
More than 20 years, Huai Kho has been irrigated. It's capacity was lost 20％. That means it can neither serve the whole project area nor extend any futher activities. There are intensive enlargement of agricultural land to supply more sufficient products because of population's growth and economic's growth. Unfortunately many rural villagers have been unforeseenably shifting their crop's cultivation depended on the price rate of that crop. Sedimentation in Huai Kho reservoir induces by this activity as well as cultivates on the reservoir edges and riparian areas in it's catchment.
Table 1 : Characteristics of Huai Kho Reservoir
- 204 -
2. WATERSHED CHARACTERISTICS
The tropographic of the drainage area is hilly rolling with 1:431 average slope, the land elevation ranges 156 - 170 m (msl). The basin's weather is influenced by south-west monsoon with average annual rainfall depth of 1,146 ㎜, about 80％ is from May to September. The average annual evaporation is 1,459 ㎜, temperature is 27℃, relative humidity is 71％. The average annual inflow is 61.7 mom, white the run-off yield of the basin is 8.43 ℓ/sec. The parent materials of most soil's texture, in the basin, are siltstone, sandstone and sandy shale. In Thai's soil classified series, there are Nam Pong (Ng＝40％), Roi-et (Re＝30％) and Ubolratchthani (Ub＝12％) which are sandy soil or silty soil and low organic matter soil (＜0.5). The land uses in the basin are paddy field (50％), farm cropping and pasture (40％).
3. SEDIMENTATION SURVEY
There are a number of methods which can be carried out reservoir deposition survey. Huai Kho reservoir was done in 1978 and 1984 by contour method, sets of survey's lines paralleled to dam site, each line is 200 meters apart which every 40 meters point leveling, the reservoir contour's map was made up from these point-levels. The decreasing of reservoir capacity, by comparision the result from 1968, 1978 and 1984 contour's map is due to sedimentation.
4. HUAl KHO SITRATION
The results from Huai Kho reservoir survey in 1968, 1978 and 1984 indicated that the capacity was 31.42 mcm, 30.62 mcm and 24.96 mom respectively. The deposition rate of period 1968-1978 was 0.08 mcm/yr while period 1978-1984 was 0.94 mcm/yr, on the other hand the reservoir storage was reduced by 2.5％ and 18.5％ respectively. In 1984, the areas and capacities above 162.00 m (msl) were decreased, aggregation, while below that were increased degradation Table 2. The reservoir bed's elevation was risen 11 ㎝/yr in average as in Fig. 2. The majority of deposition was occurred at the top of water levels, Table 3 and Fig.3. The degradation at the elevations below the dead-storage elevation, may be due to retrogressive erosion which may be induced by bottom drain outlet or pumping in dry season.
Table 2 : Reservoir Area and Capacity of Huai Kho Reservoir
- 205 -
Table 3 : Change of Reservoir Area and Capacity at Huai Kho Reservoir
5. SEDIMENTATION CONTROL
Generally, there are three critera to control the sedimentation ; minimizing, maximizing and recovery. Sediment Problems may be sloved by one or two combinations or all of these approaches which are depended on local situation. Huai Kho reservoir should be mitigated by two combinations, minimizing and maximizing. First of all, in watershed development cultivation with soil ＆ water conservation strategie should be applied. Later, the bottom drain outlet should be appropriately operated which need much more data information on sediment inflow as well as sediment outflow.
- 206 -
6. CONCLUSION
Many watershed development projects have been faced a number of unavoidable problems such as erosion, deposition, flood, shortage, etc. for Huai Kho reservoir, there are soil erosion in the upstream basin area, and channel degradation downstream, white sedimentation occured in the reservoir. The rate of sedimentation was quite high (0.93 mcm/yr), from the past records (1978-84), that should be the alarm indicator not only for Thailand but also everywhere especially developing or under developed country.
Fig. 1
- 207 -
Fig. 2
Fig. 3
- 208 -
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
PROBLEM OF EROSION AND DEPOSITION IN THE MAJOR RIVERS OF BANGLADESH
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Dept. of Water Resources Engineering, Bangladesh University of Engineering at 참여연구원 : Technology, BUET, Dhaka-1000, Bangladesh M. Monowar Hossain
Summary
An attempt has been made to estimate the erosion and deposition and variation of mean bed level in three major transboundary sediments carrying rivers of Bangladesh, i. e. , the Ganges, Brahmaputra and Teesta for a reach within the geographical boundary of Bangladesh using field data.
- 209 -
A 125 ㎞ reach of the Ganges extending from few kilometers downstream of geographical boundary of Bangladesh to Ganges-Brahmaputra confluence between the period 1967-68 and 1955-86 was considered. For the Brahmaputra about 230 ㎞ reach from the Ganges Brahmaputra confluence towards upstream was considered between the years 1965-66 and 1985-86. For the Teesta a reach of about 10 ㎞ long extending from a location where the Teesta barrage is being constructed towards upstream between the years 1970 and 1987 is considered.
Analysis of field data showed that aggradation was more than that of degradation in the study reaches of the selected rivers during the study period. The study also revealed that there was an increase in the slope of the mean bed levels by about 35 percent and 19 percent respectively for the study reaches of the Ganges and Brahmaputra between the study period. On the other hand the mean bed slope of the Teesta was found to decrease slightly by about 3.5 percent from 1970 to 1987. The increase of mean bed slope in the Ganges and Brahmaputra was mainly attributed to deposition of more sediments in the upper portion and scouring in the lower portion of the study reaches. But for the case of Teesta river mean bed level change was minimum due to existence of quasi-balance between water and sediment discharge.
The present study has further indicated on the possible causes of deposition of higher sediments in the beds of the major rivers of Bangladesh and its impacts on navigation and flood.
INTRODUCTION
Bangladesh is a flood prone country. It has within her geographical boundary the confluence of the rivers - the Ganges, the Brahmaputra, the Meghna and the Teesta. Most part of the country is the flood plain formed by these rivers and their tributaries and distributaries. Every year, a large part of the country goes under water due to high stages in the major rivers. Severe floods occurred in the years 1954, 55, 70, 74. 84, 87 and 88. Floods of 1987 and 88 were the most devastating in the history of Bangladesh. In 1987 about 40 percent of country's total area was inundated.
(BWDB, 1987)
In 1988 about 60 percent area of the country was flooded.
(BWDB, 1988)
It may be mentioned that total area of Bangladesh is only about 144000 sq. ㎞, where more than 115 million people live. The major rivers including their approximate catchment area totalling about 1.6 million sq.㎞ are shown in Fig.1. Vast quantity of sediment are carried by the major river systems of Bangladesh mostly during the flood seasons. These international rivers finally discharges into the bay of Bengal and receives the sediment from the erosion of the large catchments spreading over China, India, Nepal and Bangladesh. It has been estimated that the quantity of sediment carried by these rivers through Bangladesh vary from some 1.1 to 1.6 billion tons annually
- 210 - (Hossain, 1990)
Part of these sediment are carried to the deeper zone of the Bay of Bengal , and rest are deposited in flood plain including the river bed. Siltation of the major rivers of Bangladesh is considered to be a vital factor for increasing the flood magnitude in recent times.
REVIEW
Description of the sedimentary and hydraulic characteristics of the Brahmaputra has been provided by Coleman(1969). Latif(1969) dealt with the shifting character of the lower Brahmaputra. Later Galay(1980) studied the shifting of hank lines of both the Ganges and the Brahmaputra. Meander travel of the lower Ganges in recent time was investigated by Hossain(1988). Hossain inferred that imbalance of deposition had contributed significantly to the erratic trend in the bend migration. Sediment transport aspects for the lower Brahmaputra and Ganges were studied by Alam and Hossain(1988) and Hossain(1989a) respectively. Fluvial geomorphology of the Teesta basin has been dealt by Mukhapadhyay(1982) and sediment transport aspects by Hossain et. at. (1991). Habibullah (1987) determined the bed level variations of the Jamuna using field data for a reach within Bangladesh utilising a limited number of data. The present study takes into account more cross-section and recent data than the previous studies.
DATA AND METHODOLOGY
For present study, cross-sectional data collected by Bangladesh Water Development Board (BWDB) were utilised. The bed level was determined at different sections from the cross-sectional maps of various years. For computation of mean bed level, a normalised cross-section was assumed. The net erosion and deposition below and above the mean bed level for the normalised section remained unchanged for a hydrological cycle. This creteria was accomplished with the help of a planimeter by trial and error.
(Hossain 1989a)
Mean bed levels at 34 section of Brahmaputra for a reach of about 230 ㎞ extending from Ganges- Brahmaputra confluence towards upstream between the years 1965-66 and 1985-86 were considered. These sections were denoted by J1 to J17. Similarly 22 sections of the Ganges from Ganges- Brahmaputra confluence towards upstream for a reach of about 125 ㎞ denoted by G01 to G18 were utilised between the years 1967-68 and 1985-86 to compare mean bed levels. For the case of Teesta river, 13 section were considered between the period 1970 and 1987 for reach of about 10 ㎞ extending from a place where Teesta barrage is being constructed towards upstream. It may be worth while to mention that these cross-sectional data were collected during the lean flow seasons of the study period.
- 211 -
MEAN BED LEVEL OF THE BRAHMAPUTRA
Mean bed levels at 34 different sections for the years 1965-66 and 1965-86 are shown in Fig.2. The consecutive points at each section were connected by straight lines. On the basis of this figure it may be inferred that bed level shows an aggradation in the upper reaches and degradation in the lower reaches between the study periods. In most of the upper reaches, that is between section J6 and J17, sedimentation took place except at J9 and J10-1. The maximum rise of bed level in 1985-86 was found to be at section J16-1 by about 4.5m (15.0 ft) compared to bed level of 1965-'66. Mean bed level was also found to rise between J13 and J10-1 and between J8-1 to J5-1. In the lower reaches, the analysed section from J1 to J6 has a general trend of degraded mean bed level except at J3-1. J4 and J1-1 have suffered the highest decrease in levels respectively by about 3.5m (11.0 ft) and 3.7m (12.0 ft) lower than the mean level of 1965-'66.
MEAN BED LEVEL OF THE GANGES
Mean bed levels at 22 sections of the Ganges of the years 1967-68 and from G01 to G18 and the net change in mean bed levels from 1967-'68 to 1985-86 are shown in Fig.3. It may be observed from this figure that the upper reach (reach extending from G10 to G18) suffers aggradation while the lower reach suffers degradation. The maximum rise of bed level was observed at G15 by about 3m (10 ft.) above its mean level in 1967-68. In the lower reaches, the analysed sections from G01 to G9 has a general trend of fall except at G2 and G1 near the confluence with the Brahmaputra. G7 and G01 near the confluence has suffered the highest decrease in mean bed level during the study period. The decreases are respectively 2m (6.5 ft.) and 3m (10 ft.) from that of 1967-68. Bed degradation near the confluence i. e., at G01 may be due to a variety of reasons and is erratic in nature. Thus, for accepting the highest fall of mean bed level section G7 is considered.
MEAN BED LEVEL OF THE TEESTA
Mean bed levels for 13 sections for a 10 ㎞ reach of the Teesta were computed between the years 1970 and 1987 and are plotted in Fig. 4. The cross-sectional identifications are shown in the abscissa. It may be observed that the mean bed levels at various sections show an erratic trend of fall and rise. But. on an average, the mean bed profiles for 1970 and 1987 do not show significant trend of change over the study reach. The highest fall and rise of mean bed levels in the study reach are found to occur at CSB and CS5 with a fall and rise of 0.4m and 0.7m respectively. The figure shows that aggradation were more than that the degradation in the lower portion of the study reach. It may also be observed from figure that the slope of the mean bed profile of 1987 was slightly lower than that of
- 212 - 1987.
DISCUSSION
Rise and fall of mean bed levels indicate deposition and scouring trend respectively. But as the widths at each sections in different years are not same, a rise in bed level of a section do not necessarily mean a decrease in its cross-sectional area. This point may be seen in Figure 5, where variation of cross-sectional area at different section from G01 to G18 have been shown. Consider section G15 where the maximum mean bed level rise has occurred from 1967-68 to 1985-86. Here cross-sectional area did not decrease. instead an increase of area can be noticed. This implies that width has increased due to the bank erosion. Changes in cross-section at other areas may be seen from this figure. Similar character may also be found for the case of the Brahmaputra and the Teesta.
It is interesting to note that the mean gradient of the bed has increased for both the Ganges and the Jamuna due to deposition in the upper portion and scouring in the lower portion. Mean line for the bed level data used in Figure 2 and 3 was drawn using regression techniques. for the Ganges the mean gradient of the bed was 5.18 ㎝/㎞ ［0.28 ㎝/㎞ (0.355 ft./mile) in 1905-66 to 7.99 ㎝/㎞ (0,42 ft./mile) in 1985-86］. For the Jamuna the mean gradient of the bed increased from 6.72 ㎝/㎞ (0.355 ft/mile) in 1965-66 to 7.99 ㎝/㎞ (0.42 ft/mile) in 1985-86. Mean bed profile of the Teesta shows slight decrease in bed profile in 1987 compared to that in 1970. In 1970 mean bed profile was found to be about 0.503 m/㎞ (2.97 ft./mile) while that in 1987 it was about 0.561 m/㎞ (2.95 ft./mile).
It is possible that some morphological changes in the Ganges-Jamuna-Teesta system were the result of seismic disturbance in the Himalayas and Assam Hills. The massive land slides may cause to supply huge quantity of sediments into these rivers, The seismic disturbance may also cause partial clogging of the rivers and this may result in increased erosion and deposition. This increase sediment load in the rivers and accumulation of sediment takes place when it goes beyond the carrying capacity. In general, the Increased sediment load continues for several years. Increased supply of sediment into these rivers may also be due to massive development works undertaken and deforestation due to a variety of reasons in the basin of these rivers. The increase of flood magnitude in recent years may largely be attributed to the silting up of major rivers and mouth of their tributaries in Bangladesh. This silting also cause severe navigational problem during lean flow season.
Apparent changes in mean bed level (MBL) for the Ganges Brahmaputra, and Teesta as mentioned earlier were due to an imbalance between erosion and deposition. It may be revealed from the figures 2 and 3 that the amount of materials eroded away and carried out of the reach. Assuming linear variation of changes in cross-sectional area between two consecutive sections, volume of materials eroded and deposited in different reaches have been calculated by integrating the area bounded by the profile below and above the zero reference line. The net deposition approximately equals to 4.26 × 10^{8} ㎥ during 1967-68 to 1985-86 for the Ganges and 3.4 × 10^{8} ㎥ during 1965-66 to 1985- 86 for the Brahmaputra were found respectively. For the case of Teesta river the net deposition was about 2.5 × 10^{5} cubic meter during the study period in the selected reach.
- 213 - CONCLUDING REMARKS
Analysis of mean bed levels shows that erosion and deposition in the study reaches is very erratic and deposition is more than that of erosion. The highest rise and fall of mean bed level occurred to section Jl6-1 and J4 respectively for the Brahmaputra. For the Ganges the highest rise and fall was found to occur at G15 and at G7. Mean bed gradient of the Brahmaputra had increased by 19 percent while mean bed gradient of the Ganges has increased by about 35 percent. For the Teesta upper portion of the study reach suffers slight degradation while the lower portion suffers aggradation and mean bed gradient is reduced by about 3.5 percent. Increased supply of sediments into these rivers may be due to seismic disturbances as well as due to development works undertaken at upstream reaches. The rise in bed level may be held responsible for more flooded area in Bangladesh in recent time and create navigational problem during lean flow season.
Net deposition occurred within the study reaches. The net total volume of material deposited in the reach of 125 ㎞ for the Ganges in 18 years was 4.26 × 10^{8}㎥, for the Brahmaputra in 20 years for a reach of 230 ㎞ was 3.4 × 10^{8}㎥ and for the Teesta net deposition was about 2.5 × 10^{5} cubic meters
FIG. 1 THE GANGES-BRAHMAPUTRA-TEESTA BASIN (SOURCE:ZAMAN et. al 1983)
- 214 -
FIG. 2 VARIATION OF MEAN BED LEVEL FOR THE JAMUNA FROM 1965-66 TO 1985-86
FIG. 3 VARIATION OF MEANBED LEVEL (MBL) WITH DlSTANCE FOR THE GANGES ( 1967 - 68 TO 1985-86)
- 215 -
FIG. 4 VARIATION OF MEAN BED LEVEL FOR THE TEESTA BETWEEN 1970 AND 1987
FIG. 5. VARIATION OF CROSS-SECTIONAL AREA WlITH DISTAHCE FOR THE CANOES (1961-68 TO 1905-86)
"REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF
- 216 - INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
RIVER PROCESSES DOWNSTREAM OF A RESERVOIR
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Head. Isotope Hydrology Division, INST, AERE, 참여연구원 : Savar, P.O. Box 3787. Dhaka, Barlgladesh Md. Delwar Hossain Sikder
Summary
The Karnafuli river is the principal drainage of the Chittagong Hill Tracts and is one of the important rivers of Bangladesh. It rises in the Lushai Hill in India and flows generally south-western course through the Chittagong Hill Tracts and the alluvial coastal plain in the eastern part of Bangladesh. Near Chittagong, it enters the Bay of Bengal. It flows for above 195 ㎞ through Bangladesh and is joined by several tributaries and has a total catchment area of about 14,000 sq. ㎞ of which 11,000 sq.㎞ of it is controlled by the Kaptai dam. which is situated about 77 ㎞ upstream. The Kaptai dam has impounded a huge reservoir together with the Karnafuli hydroelectric plant. The river is tidal all the way up to the dam.
Since 1961, operation of the Kaptai Reservoir has permitted regulation of the river. This programme has got about radical changes in the regime below the dam. The training works e. g. river bank revetment were completed in the lower Karnafuli i. e. in the port area, there has been a marked improvement in the width and depth of the navigation channel of the port, which appears to be
- 217 - stabilised now but for seasonal fluctuation. Chittagong port is facing sediment problem at the outfall of the Karnafuli. In 1990, tracer studies were successfully carried out on the movement of bed sediment at the outfall.
1. INTRODUCTION
Bangladesh is a land of rivers. The Karnafuli river is the principal drainage of the Chittagong Hill Tracts and is one of the most important rivers of Bangladesh.
(Fig-1)
Chittagong, the main Sea Port of Bangladesh is situated on the right bank of the river and is 13 ㎞ away from the Bay of Bengal.
Geologically the area represents a complex zone of Bengal Basin. The area is covered by recent deposits of stream, in places rocks are predominant deposits. These rocks are laminated and stratified silty shale and sandy shale. At some locations, these shale layers lie on or close to the surface of the river bed and at other places at a much greater depth.
In 1961, the Kaptai dam was constructed on the river Karnafuli and as a result a huge reservoir was formed in the upstream of the dam.
(Fig-2)
This is the only reservoir in Bangladesh and there is also situated the Karnafuli hydro-power plant. At the hydro-power plant, there are 5 unite installed with a capacity of 230 MW. The generators of the hydro-power plant operate within the head of minimum 21 metre and maximum 33 metre above mean sea level. If the water level rises above the maximum level during the monsoon then the excess water is released through the spill way of the dam. In the monsoon period, the hydro-power plant releases turbid water whereas in the dry season it releases clean water. The total discharge of five generators varies from 3,110 cubic metre per second to 3,255 cubic metre per second. The Karnafuli river processes downstream of Kaptai reservoir is a complex system. It is very active and gradual change is going on due to the consturction of many industries and projects on its banks.
2. THE RIVER CATCHMENT AND ITS CHARACTERISTICS
The Karnafuli is a meandering and alluvial river with pronounced tidal effect in its tower 48 ㎞ reach.
(Fig-3)
- 218 -
It rises in the Lushai Hill in India, about 2,440 metre above mean sea level and flows generally south- western course throuel the Chittagong Hill Tracts and the alluvial coastal plain in the eastern part of Bangladesh. Near Chittagong, it enters the Bay of Bengal through the Patenga outfall in a south- westernly direction roughly at right angle to the numerous outfalls in the Deltas of the Ganges, the Brahmaputra and the Meghna rivers. The Karnafuli which flows for about 195 ㎞ through Bangladesh is joined by several tributaries namely the Tuichong, the Sogalui and the Halda which has a total catchment area of about 14.000 sq. ㎞. At present, approximately 11,000 sq.㎞ of it is controlled by the Kaptai Dam, which is situated about 77 ㎞ upstream.
Excluding the northern extremity, the entire Karnafuli catchment lies within the tropics and has monsoon climate. The average annual rainfall in the catchment area varies between about 2,520 ㎜ on the alluvial coastal plain and about 3,780 ㎜ in the Hill Tracts. About 60 percent of it falls in June, July and August during the south-western monsoon.
The lower Karnafuli shows the typical features of an estuary. The most noticeable phenomena in the area are the strong currents caused by in and out going tidal flows, and the high river discharges during the monsoon.
The nature of the sediment in the catchment area and the estuary is such that large quantities can be transported by these flows. Density currents, caused by difference in density (salinity) between sea water and fresh water, influence the transport pattern of sediment in the estuary erosion and deposition. It varies over the season, and causes gradual changes of the estuary.
3. HISTORY OF MORPHOLOGICAL CHANGES
In the past, the Karnafuli was subject to considerable changes. The process of morphologic changes in the upper reach is continuing. Efforts were made by the Port Authority to stabilize the river by constructing training works since 1905. Between 1905 and 1912 the earliest major training works constructed which consisted of 4.8 ㎞ of revetment to stop the erosion of the right bank downstream of Chittagong. Later on, between 1930 and 1943, the lower part of the estuary was improved by the construction of the Patenga, the Juldia, the Gupta and the Danger char training walls.
(Fig-4)
Further improvement, so far stability is concerned, was achieved since 1954 by training of the upper estuary. The training works in this part included;
3.1 Closure of Dalur channel.
3.2 Removal of an indurated clay bar just upstream of Sadarghat,
3.3 Cessation of spoil dumping opposite the jetties,
- 219 -
3.4 Removal of a 915 metre long stone training spur on the inside of the Kolagaon Bend and.
3.5 Closure of the right bank channel at Halda char by a 2 ㎞ revetted longitudinal embankment and cross bunds.
After completion of a new bridge on the Karnafuli in 1989 at about 18 ㎞ upstream there is a tendency of forming shoal in the upstream of the bridge and is gradually incresing.
Since the training works and Kaptai Dam were completed, there has been a marked improvement in the width and depth of the navigation channel in the lower Karnafuli, which appears to be stabilised now but for seasonal fluctuations.
The water depths in the area are subject to seasonal changes. The river section between the Kalurghat bridge and the fishery harbour showed considerable changes in the season of 1976/77. The river stretch downstream of the fishery harbour appears to be rather stable with a general tendency to shoal in the rainy season and to deepen in the next dry season. This tendency is reflected in the seasonal depth fluctuations at the Outer Bar. A considerable change in depth has been found at the Outer Anchorage. There the depth pockets present in the dry season appears to be filled up atmost completely with soft mud in the next rainy season. Apparently, huge amounts of sediment, approximately 2 billion tons per year being discharged by the Ganges, the Brahmaputra and the Meghna rivers reach the Outer Anchorage of Chittagong port. Obviously, the silt is removed during the transition period in the next dry season, but the dry season soundings of this area do not show systematic changes over the last ten years.
In the dry season, i. e. from November to March huge water is being withdrawn from the Karnafuli for four irrigation projects namely Ichamati Unit, Sylok Unit, Boalkhali Unit and Halda Unit.
(Fig-2) as well as for industrial, domestic and sewerage uses. As a result, upstream discharge reduces to a minimum and consequently tidal wave reaches further in the upstream.
4. TIDES
The tidal wave, situated in the north-eastern corner of the Bay of Bengal, approaches the Karnafuli Outfall from the south and continues its path in northern direction to Sandip Channel. As the water depth decreases in northern direction, the propagation speed of the tidal wave decreases while its height increases. In the upstream, the height of the tidal wave gradually decreases; at Kaptai Dam, it is still just noticeable at low discharge rates of the generators. The tidal ranges at Patenga, on the average, vary between 1.65 m (mean neap tide) and 3.8 m (mean spring tide).
Measurement of the velocity profile were made by the Netherland Economic Institute at Patenga and Kalurghat bridge during a tidal cycle in 1977. From the measurement, it was found that the tidal volume varied from 40.1 × 10^{6} ㎥ to 60.1 × 10^{6} ㎥ at Kalurghat bridge and 80.1 × 10^{6} ㎥ to
- 220 - 140.1 × 10^{6} ㎥ at Pantenga. Reversal of the flow due to tidal wave will occur as long as the river discharge does not exceed the maximum discharge rate due to tidal motion.
5. CURRENTS
In the Karnafuli estuary, the flow velocities depend on the characteristic dimensions of the estuary(storage area, crosssectional area and hydraulic resistance), the variations of the sea level at Patenga( tides ＆ wind) and the river discharges into the estuary. During the dry season, the currents in the estuary are almost completely governed by the tidal motion. At Patenga, the maximum velocities then vary from about 1.0 m/s at neap tide to about 1.5 m/s at spring tide. The corresponding velocities are somewhat higher in the Gupta bend and lower at Kalurghat. Along the whole river, the flood velocities are lower than the ebb velocities.
6. SEDIMENT TRANSPORT
The strong currents in the estuary and the adjacent sea area transport large quantities of sediment. The bottom profiles in the area considered here, are more of less, in a state of dynamic equilibrium, they change continuously, but within certain limits. To understand the morphology of the area, the phenomena involved in erosion and deposition are of great importance. The most important factor governing sediment movement is the shear stress between the flowing water and the river bed. The shear stress generates turbulent motion in the water and thus upward transportation of sediment by turbulent diffusion.
6.1 Suspended Sediment Transport.
The rate of suspended sediment transport varies strongly with the tidal range. Large quantities of suspended sediment are transported up and down of the river by flood and ebb respectively. These quantities vary between about 15,000 ㎏/m (low neap tide) and about 500,000 ㎏/m (high spring tide). On the average, this quantity is estimated at 150,000 ㎏/m respectively, a mean transport rate of suspended sediment is of 6/7 ㎏/ms. This coincides with an overall mean sediment concentration of about 800 ㎎/1 and a tide slightly above the average tide with a maximum velocity of 1.3 m/s at Patenga and 1.3 to 1.5 m/s at the Outer Bar. The sediment particles are found to be coarser than about 20 micron but smaller than about 70 micron.
6.2 Sediment Transport along the Bottom.
The material on the top of the river bed consists of sand with only small percentage of silt. The samples taken along the river axis in July 1977 showed that the percentage of silt was 0.1 to 4.2 with an average of 0.74 percent. The diameter of the sand fractions ranged from 152 to 380 micron with an average of 280 micron.
- 221 -
In the dry season i.e. January, 1977 say, the percentage of silt content was somewhat higher i. e. of average about 16 ％. The average median diameter of the sand in the lower Karnafuli at that time was about 235 micron. So, in the river, the characteristics of the bed materials do not appear to be subject to significant seasonal changes. The sediment of the river bed is distinctly coarser than the suspended material and will mainly be transported along the bottom.
In the coastal area, in front of the Karnafuli entrance, a distinct seasonal effect on the characteristics of the bed material was found in February and July, 1977. In the rainy season, a layer of soft mud containing less than 5％ of sand (very fine) covers the sea bottom approximately beyond the 10 metre-depth contour and the 8 metre-depth contour near Norman point. In the next dry season, the layer of soft mud disapears and a bottom of more consolidated mud remains.
The results of 6 bottom samples collected in 1982 and 52 samples collected in 1987 at the river mouth showed almost pure fine to medium sand i.e. D_{50} ＝ 225 to 328 micron.
Chittagong port is facing sedimentation problem in the river bed as well as at the outfall of the Karnafuli in the Bay of Bengal. In 1990, radioactive scandium glass was used successfully to study the bed sediment movement at the mouth of the river to find out the source and direction of the sediment deposited there. Along with the radioactive tracer, coloured sands were used to study the dynamic layer of the bed. One more tracer study is being planned for the study of bed sediment movement in the Karnafuli river near to the port area.
7. CONCLUSION
The Karnafuli river processes downstream of the Kaptai reservoir is a complex one. The most important sea-port "Chittagong" is situated on the right bank of the river. As a result, river navigation is very important for the port. To maintain smooth navigation in the river, Chittagong Port Authority has been trying its best to stabilize the river by consturcting training works since 1905. Port jetties, bridges, many irrigation projects, industries and many other projects are being grown up on the banks of the river and on the river itself. These projects have some far reaching impacts on the river system. So, the changes on the river bed as well as on its bank will continue in the years to come. But to maintain navigational depth, river bed is being dredged and some modern techniques are being used to study the sediment movement in the river bed as well as at its mouth in the Bay of Bengal.
- 222 -
Fig-1
- 223 -
Fig-2
- 224 -
Fig-3
- 225 -
Fig-4
- 226 - "REGIONAL TRAINING PROGRAMME ON EROSION AND SEDIMENTATION FOR ASIA" (RAS/88/026) PRICEEDINGS OF INTERNATIONAL SYMPOSIUM ON SPECIAL PROBLEMS OF ALLUVIAL RIVERS INCUCING THOSE OF INTERNATIONAL RIVERS
SEDIMENT CHARACTERISTICS OF THE UPPER INDUS AND SEDIMENATION OF TARBELA RESERVOIR
1991
주관 KOREA INSTITUTE OF CONSTRUCTION 수 행 기 관 : TECHNOLOGY(KICT), KOREAN ASSOCIATION OF HYDROLOGICAL SCIENCES(KAHS) Project Director, Surface Water Hydroloar 참여연구원 : Project, WAPDA. 31-A. Masson Road, Lahore- Pakistan Muhamnad Salim Wars
1. THE RIVER AND CATCHMENT
Above Kalabagh dam site the total catchment of Indus River is 110,500 sq. miles, 70,400 squares miles of which is above the elevation of 15000 ft. The topography of the catchment is mountainous at the upstream and hilly at the downstream end; and includes some high peaks, big glaciers and broad flat valleys. Only an area of 40,100 sq. miles is subject to rain precipitation out of which only 400 sq. miles are located above Tarbela Dam. Natural vegetation is sparse in the lower areas and increase somewhat at the upper reaches with some forests. Location map showing Tarbela Dam and Proposed Kalabagh and Basha dams is given as Fig. 1.
Snow-melt contribution from about 14,000 sq. miles glacier catchment of Indus River has been observed to have a predominant effect in formulation of water and sediment regime of the river.
- 227 -
2. SEDIMENT TRANSPORT
With respect to sediment transport and soil erosion the region drained by Indus above Kalabagh can be broadly divided into the following sub basins :
1. Upper Indus Basin.
2. Chitral-Kabul-Swat Bara Basins.
3. Siran-Haro-Soan Basins.
4. Kohat-Toi, Kurram-Tank Zam-Gomal Basins.
The catchment area upstream of Kalabagh excluding the catchments of left and right bank tributaries below Tarbela may be called the Upper Indus Catchment.
Geologic erosion is predominant in the Upper Indus catchment along the youthful Himalayas where there is practically no plant cover. Large alluvial deposits, with deep gravel and sand bars, are observed in the upper reaches of the Indus and along its tributaries.
It is considered that the high sediment yield of the Indus is due to poor vegetal cover, steep slopes, inappropriate means of soil conservation and due to the fact that the soil and rocks of the Indus valley are geologically young and easily erodable.
2.1 Sediment producing characteristics of glaciers.
When looking at the history of major floods in the Indus, one always traces their origin in the upper mountainous parts of this great river. These floods are connected with exceptional glacier advances and natrual damming up of the rivers in the mountains resulting in dam burst floods which bring with them a very large volume of water and sediments.
2.1.1 Glacier surges
Glacier surges in the upper Indus are associated with erosional effects of the huge ice masses producing large quantities of mud flows. The phenomenon which produces large floods occurs frequently and have significant erosional effects.
An interesting event of a heavy mud-rock flow, connected with advance of Balt Bare glacier in Karakoram Mountains, is occured in 1974 and described here under :
- 228 -
" On April 12, 1974 a rarely heavy mud-rock flow bursted out from the Balt Bare Valley by the Shashikat village, where the Karakoram Highway, Pakistan passes through. The mud rock flow blocked Hunza river forming a debris-rock dam which created a reservoir 10 miles long and 160 ft. deep. It flooded a big bridge of 394 ft. in length and a section of highway about 2.50 miles; so the communication was forced to be held up".
The increased snow-melt is, however, more important than the glacier ice movement at the flowing water picks up thick deposits of the arid proglacial valley and carries them downstream of the river.
2.1.2 Natural damming and major dam burst floods.
More than 32 major dams burst floods have occured in the upper Indus since1826. At least three major floods, caused by breaching of natural dams, are well known in history which are associated with heavy sediment erosion and its transportation in the upper Indus region as given below:
3. SEDIMENT YIELD OF THE INDUS RIVER AND SILTING OF RESERVOIRS.
Suspended sediment load of the Indus at various hydromet stations established in its upper mountainous reaches indicates wide variation from year to year identifying most severe sediment yielding characteristics of some of its tributaries.
(FIG-2)
Due to large amount of sediments carried by the Indus any reservoir built on it has got a short life. The following are some of the results of sedimentation of Tarbela reservoir.
- 229 - (FIG-3)
3.1 Sediment accumulation in Tarbela reservoir.
Most of the sediments carried by the river Indus are trapped in the reservoir and deposited in the upper reaches. During the period of reservoir's depletion sediments deposited in the upper reaches are reworked and carried downstream within the reservoir. A small portion of these sediments passes out of the reservoir through the outlets while bulk of the sediments get accumulated in the form of major delta. Total sediment load computed from mean inflow at Tarbela during the past 17 years (1974-1990) using rating curves derived from hydrometric data, flow duration and monthly suspended sediment rating curves in 3,781 MST. The sediment load as given below estimated from hydrographic survey for the same period is 3,231 MST.
Nearly 42％ of this sediment has partially filled the dead storage of reservoir below EL. 1,300 in the form of a delta.
(FIG-4) which is being built by the incoming sediments every year and advancing towards the Main Dam at an average rate of about half a mile per annum. During the year October 1989 - September 1990 under a report by Tarbeta Dam the minimum reservoir level remained high at EL. 1,381.38 and the Hydrographic Survey showed no delta advancement. Fig.5 shows reduction in reservoir capacity (upto Oct. 1990) during a period of 16 years of its operation.
- 230 -
3.2 Delta advancement.
Delta advancement, showing period of minimum reservoir level below EL.1,320, the annual sediment load computed from hydrographic survey and the river inflows 1979 to 1990 are given hereunder:
- 231 -
It is evident that in 1985 the longer duration for which the reservoir remained below EL. 1,320 had caused maximum advancement of delta towards the dam, and the shorter duration in 1990 has resulted in no advance of delta.
3.3 Trap efficiency of the reservoir during Oct. 1989 to Sept. 1990 ranged between 91.8％ to 99.9％ for most of the period and depended upon seasonal inflow and reservoir operations.
4. SEDIMENT INFLOWS TO TARBELA RESERVOIR AS OBSERVED AT BESHAM QILA.
Besham Qila hydrometric data after adjustment upto Darband (now submerged) , a stream gauging station just upstream of the periphery of Tarbela reservoir, gives the following sediment inflows :
1. Mean suspended sediment load (post Tarbela )
1974-87(14years) 237MST/Year.
2. Mean suspended load(Pre-Tarbela)
1960-73(14years) 316MST/Year.
Prior to construction of the reservoir long term sediment inflows to Tarbela reservoir was estimated as 440 MST/Year which incorporated exceptional sediment inflows associated with rare floods experienced in the past. However, as seen above, during recent decades the sediment inflows to Tarbela reservoir have been moderated to low. The huge quantity of sediments being carried by the river with a large proportion of course sediments, require devising ways and means to adopt sediment reduction measures to increase life of reservoirs on Indus.
- 232 -
5. SEDIMENT REDUCTION MEASURES.
The following sediment control measures are usualIy adopted:
1. Watershed management practices,
2. Sediment removal from the reservoirs by using mechanical or hydraulic methods,
3. Creation of off-channel reservoirs,
4. Using favourable reservoir operation so that sediment laden waters could be passed out of the reservoir when the sediments are still in suspension.
Although Watershed management measures, adopted in the catchment of Mangla Dam, have proved quite beneficial and increased the life of Mangla reservoir; these measures could only be fruitful on a limited rainfall portion of the lower reaches of the catchment whereas in the arid upper Indus valleys and devoid of vegetation high mountains, these measures may not be so promising. Of all the other measures to reduce sediments, creation of off channel reservoirs and using sediment sluicing through the low lying and medium height outlets are the possible solutions of this problem.
Sluicing of sediments have been used in China and other countries for sediments control but there about 80 to 90％ of the annual sediments load is discharged by the river in July-August whereas only 25 to 50％ of the annual runoff occurs in the same period. In case of Sansmenzia reservoir in China, the sediment laden waters during July-August are sluiced through the low lying outlets of the dam by drawing down reservoir level and filling of the reservoir is done subsequently during the non-flood season. This has substantially reduced accumulation of sediments in that reservoir. But in case of Tarbela and other dams to be built on Indus, the filling of the reservoirs has to be done during June- October as seen from Fig.6. The sediment concentration is also maximum during flood season. However, since sand and silt contents are predominant in its flood flows, these sediment could be sluiced out partially during the period when the reservoir is drawn down enabling reworking of the accumulated sediments which may get passed out of the suitably located outlets. This could also be done by passing sand and course silt of the highly sediment laden flood flows during initial rising water period. But the success of this operation is then dependent on the climatological conditions which govern glacier and snow melt during summer season with a risk of partial filling of the reservoirs during the non-sluicing period of the reservoir.
Under these conditions on long term basis we could adopt some typical reservoir operation techniques in addition to sluicing of sediments with multiple reservoir operations built upstream of Tarbela reservoir to increase life of Tarbela or proposed Kalabagh and other dams is to be examined in depth. However, quick sedimentations of reservoirs built on Indus and possibilities of climate change are factors to consider building of carry over storages on the Indus for their long term utility in water resources development of Pakistan.
- 233 -
FIG - 1
- 234 -
FIG - 2
FIG - 3
- 235 -
FIG - 4
- 236 -
FIG - 5
- 237 - FIG - 6
- 238 -
- 239 - 참 고 문 헌
분야 : VARIATION OF FROW AND R0UGHNESS CHRACITRISTICS IN FLUVIAL RIVIRS Kuroki, M. , Kishi, T. , Study of regime criteria of bed configurations of fluvial rivers with bar , 26th Japanese Conference of Hydraulics 1982 in Japanese
Kuroki, M. , Kishi, T. , Analytical study of regime criteria of bed configurations of meso-scale , Trans. No. 342 , JSCE , 1984 in Japanese
Yamaguchi, H. , Kuroki, M. , Types of bar in fluvial rivers and their regime criteria , Proc. 18th Symp. of Natural Disaster Science 1981 in Japanese
Kishi, T. , Kuroki, M. , Bed forms and resistance to flow of fluvial rivers , Proc. 3rd Japan-China Bi-lateral Conference on River and Dam Engineering 1986 in Japanese
Kishi, T. , Kuroki, M. , Bed forms and resistance to flow of fluvial streams(Ⅰ) , Rep. Faculty of Eng. No. 67 , Hokkaido Univ. , 1973 in Japanese
Kishi, T. , Kuroki, M. , Bed configurations and flow resistance in alluvial streams , Rep. Control of flood flows in fluvial rivers and improvement of safety of river channels during flood Gov. Grant- in-aid , 1988 in Japanese
Kishi, T , Kuroki, M. , Study of resistance to flow of fluvial rivers , Rep.Basic studies on the functions of rivers Hydraulics and Hydrologsy Laboratories Hokkaido Univ. , 1980 in Japanese
Fukuoka, S. , Finite amplitude development of alternate bars , River Meandering Water Resources Monograh 12 A. G. U. , 1989
분야 : FORECASTING IMPACTS OF HYCRAULIC STRUCTURES ON AN ALLUVIAL RIVIR Dou, G. et al , Studies on the sedimentation problem in the varying backwater zone of the Three Gorges Project by sediment laden flow model , Proceedings, 4th International Symposium on River Sedimentation p.1-14 , Beijins, China , 1989
Han, Q. W. , Preliminary study of nonequilibrium transport of sediment in a reservoer , 24pp , Changjiang Water Conservancy and Hydroelectric Power Research Institute , 1972
Report in Chinese
- 240 - Han, Q. W. , M. He , A new mathematical model for reservoir sedimentation and fluvial process , International Journal of Sediment Research vol. 5 , no.2 , 1990
Lin, B. , Huang, J. , Li, X. , Unsteady transport of suspended load at small concentrations , J. of Hydraulic Engineering vol. 109 , no. 1 , ASCE , 1983
Proc.
Lin, B. , shen, H. , 2-D flow with sediment by characteristic method , J. of Hydraulic Engineering Vol. 110 , no.5 , ASCE , 1984
Proc
Anonymous , Sedimentation and river process of the Yangtze River in the fluctuating backwater region of the Three Gorges Project and their impacts on flood control and navigation , Chinese Ministry of Water Resources , 1990
Internal Report, in Chinese
Tang, R. (Ed.) , , Sediment Research, Engineering Monograph of Gezhouba Project 223 pp. , Beijing , Water Resource ＆ Electric Power Press , 1990 in Chinese
Lin, B. , Recent Chinese Progress in Hydraulics , Proc. 6th Regional Congress of IAHR- APD Vol. 5 , Kyoto , 1988
Chanjiang Research Institute , The design of the movable-bed model of the Yiufanggou Reach in the fluctuating backwater region of Danjiangkou , 33pp. , Wuhan , 1990
Internal Report
Zhou, J. , Balancing point method for the satisfaction of conditions at irregular boundaries , no.7 , Shuili Xuebao , 1988 in Chinese
Zhou, J. , B. Lin , Wang, L. , 2D computations of flow in a reach , no.5 , Shuili Xuebao , 1991 in Chinese
Dou, G. et al , Study on two-dimensional total sediment movable-bed mathematical model , 4th International Symposium on River Deimentation Beijing, China , 1989
Proceedings
Zhou, J. , On numerical modeling of unsteady flow in two space variables and morphological changes of the bed , Dr. Eng. dissertation 89pp , IWHR , 1988 in Chinese
Institute of Water Conservancy Research , Chengdu University of Technolegy , Study of 2D model for bed load in the port of Chongqing , 46pp , 1990
- 241 - in Chinese
Han, Q. , M. He , A mathematical model for reservoir sedimentation and fluvial pocessess , International Journal of Sediment Research Vol. 5 , no.2 , 1990
분야 : THE STATE AND THE PROSPECTS OF THE DIFECT SEDIMENT REMOVAL METHODS FROM RESERVOIRS Ackers, P. , Thompson, G. , Reservoir Sedimentation and Influence of Flushing , Sediment Transport in Gravel -Bed Rivers London , John Wiley ＆ Sons , 1987
Chao, P. , Ahmed, S. , A Mathematical Model for Reservoir Sedimentation Planning , Water Power ＆ Dam Construction 1985
Farhoodi, J. , Preliminary Evaluation of Flushing Operation in Sefid-Rud Reservoir , Iranian National Committee on Large Dams Technical Committee on Sedimentation , 1985
International Commissionon Large Dams , Guidelines on Sedimentation of Reservoirs , Bulletin No.67 1989
Ministry of Energy , Iran, Sediment Flushing at the Sefid-Rud Reservoir , Technical Bureau, Water Affairs , 1984
Scheuerlein, H. , Reservoir Sedimentation - A Vital Prolblem of the Future and A Challenge to Dredging Technology , ⅩⅠth World Dredging Congress Brighten, UK , 1986
Scheuerlein, H. , Sedimentation of Reservoirs - Methods of Preventing Techniques of Rehabilitation , First Iranian Symposium on Dam Engineering Tehran, Iran , 1987
Stigter, C. , Meulblok, M. , Ottevanger, G. , Van Drimmelen, N. J. , Scheuerlein, H. , Sediment Control Practices, Dredging and Flushing , ICOLD Committee on Sedimentation of Reservoirs 1990
White, W. , Bettess, R. , the Feasibility of Flushing Sediments through Reservoirs , Challenges in African Hydrology and Water Resources 1984
IAHS Publication No.144.
Stigter, C. , Korver, J. N. , Dredging of Reservoirs with Limited Water Consumption , ⅩⅠⅠ World Dredging Congress Orlando, Florida, USA , 1989
분야 : J.C. STEVENS AND THE SILT PROBLEM A REVIEW Brune, G. M. , Trap efficiency of reservoirs , v.34 , p.409-418. , Trans. Am. Geophys. Union , 1953
Collar, P. D. , Guzman-Rios, S. , Sedimentation rates and capacity restoration, Lago Loiza, Puerto Rico , Proc. 5th Inter. Sed. Conf. p.15-17~15-25 , Las Vegas, NV , 1991
Einstein, H. A. , The bed-load function for sediment transportation in open channel flows , U.S. Dept. Agriculture Tech. Bulletin 1026 p.71 , 1950
Gorback, C. A. , Baird, D. C. , Case history: channel aggradation above a reservoir: Proc. , 5th Fed. Inter. Sediment. Conf. p.4-113 to 4-122 , Las Vegas, NV , 1991
IRTCES , Lecture notes of the training course on reservoir sedimentation , International Research and Training Center on Erosion and Sedimentation Beijing China , 1985
- 242 - Lee, Wing H. , , 1991 personal communication
Long Yuqian , Zhang Qishun , Rebuilding and operation of Sanmenxia Project , in Selected papers of researches on the Yellow River and present practices p.157-165 , Yellow River Conservancy Commission , Zhengzhou, China , 1987
0'Brien, M. P. , Review of the theory of turbulent flow and its relation to sediment transportation , p.487-491 , Trans. Am. Geophys. Union , 1933April 27-29,1933
Rouse, H. , Modern conceptions of the mechanics of fluid turbulence , Trans. Am. Soc. Civil Engineers v.102 , p.463-563 , 1937
Stevens, J. C. , The Silt problem , Trans. Am. Soc. Civil Engineers V.101 , p.207-288 , 1936 with discussion by Messrs.Harry G. Nickle,E.W.Lane, Frank E. Boner, Morrough P. O'Brien, Harry F. Blaney, W. W. Waggoner, N. C.Grover. and J .C. Stevens
Strand, R. , Pemberton, E. , Reservoir Sedimentation , p.48 , U.S. Dept. of the Interior, Bureau of Reclamation , Denver, Colorado , 1982
Stroebe, G. G. , Soms hydrologic facts concerning the Yangtze River , Jour. Assoc. of Chinese and Amer. Eng. v. 6 , no. 1 , p.21-31 , 1925 plus 25 Plates
Vanoni, V. A. , Transportation of suspended sediment by water , Trans. Am. Soc. Civil Engineers v. 111 , p.67-133 , 1946
Williams, D. T. , Sedimentation problems and solutions , Rosieres dam andreservoir, Sudan p.411-416. , Hyd. Engr. Amer. Soc. of Civil Engineers , 1991
Proc. 1991 Natl. Conf.
Xia Zhonghuan , Han Qiwei , Jiao Enze , The long-term capacity of reservoirs , Proc. 1st Intern. Symp. on River Sed. p.753-762 , Beijing, China , Guanhua Press , 1980
분야 : SEDIMENTATION ASPECTS OF FLOOD - PLAIN MANAGEMENT IN BANGLADESH Linsley and Franzini , Water Resource Engineering ,
Bangladesh Country Report on the Study on the Causes and Consequences of Natural Disasters and Protection and Preservation of the Environment , Bangladesh National Committee Constituted for the said purpose ,
Annual Flood Report of Bangladesh 1990 , BWDB ,
Statistical Year Book of Bangladesh 1990 , Bangladesh Bureau of Statistics ,
James M. Coleman , Sedimentary Geoloar ,
G. R. Chowdhury , Ex-Chairman , Management of Sediment in Bangladesh , BWDB ,
Amjad Hossain Khan , Ex-chairman , Morphological Survey and Sediment Sampling in Bangladesh , BWDB ,
- 243 - M. A. Matin Director. BWDB
Methods and Problems of Flood control in Asia and the Far East , Flood Control Series No. 2 United Nations ,
Flood Report in Bangladesh , 1987
Ven Te Chow , Applied Hydrology ,
National Flood Protection program , Ministry of Irrigation Water Development ＆ Flood Control GOB , 1988
BWDB Achievement , Directorate of Monitoring BWDB ,
분야 : ON FAILURE OF RIVER TRAINING WORKS WITH REFERENCE TO AFNIKO HIGHWAY Kuck Andreas (. . .) , Bioengineering in River Training , Department of Irrigation , Nepal ,
River Training Project
The Rising Nepal , , Daily Newspaper (3 July 1987) Publisher Gorkhapatra Sansthan , 1987
WECS , Study of Glacier Lake Outburst Floods in the Nepal Himalaya , 1987
Phase Ⅰ Interim Report, Report No. 4/1/200587/1/1, Seq.No.251
분야 : SEMIMENTATION ASPECTS OF FLOODPLAIN MANAGEMENT FOCUSED ON THE 1990 HAN RIVER FLOOD Engelund, F , Hansen, E , A Monograph on Sediment Transport in Alluvial Streams , Teknisk Vorlag, Copenhagen, Denmark , 1967
Karim, M. F. , Kennedy, J. F. , Computer-Based Predictors for Sediment Discharge and Friction Factor of Alluvial Streams , Iowa Institute of Hydraulic Research, Report No. 242 , Univ. of Iowa , 1981
Movable-Bed Model Test of the Lower Han River , Ministry of Construction in Korean , 1983
Van Rijn, L. C. , Sediment Transport, Part II: Suspended Load Transport , J, of Hyd, Div. Vol. 110 , ASCE , 1984
HY11
White, W. R. , Milli, H. , Crabbe, A. D. , Sediment Transport: An apraisal Methods, Performance of Theoretical Methods, When Applied to Flume and Field Data , Hydraulic Research Station Vol. 2 , Report IT119 , Wallingford, U.K. , 1973
Woo, H. , Yu, K. , Development of a Guidline for the Selection of Sediment Transport Formulas , Korea Institute of Construction Technology Report No. KICT 89-WR-113 , in Korean , 1989
분야 : IMPACTS OF HYDRAULIC STRUCTURES ON ALLUVIAL RIVERS : RIVER PROCESSES OF TIDAL REACHES Anwar, J. , Geology of coastal area of Bangladesh and recomendation for resources development and management , CIRDMA workshop Dhaka , October, 1988
Nisat, A. , Review of present activities and state of art of the coastal area of Bangladesh , CIRDMA workshop Dhaka , October. 1988
- 244 - Brammer, H. , Monitoring the evidence of the green house effect and its impact on Bangladesh , CIRDMA workshop Dhaka , March. 1989.
ADC , Feasibility Report, Package-Ⅰ, Khulna Coastal Embankment Rehabilitation Project , Vol.Ⅰ , 1986
Main Report
Hossain, M. , The green house effect and the coastal area of Bangladesh: its people and economy , CIRDMA workshop Dhaka , March, 1989
Shah, A.H. , Measurement of suspended load through a cross-section of the old Brahmaputra river in Bangladesh , Workshop of UNDP/UNESCO Regional Project RAS/88/026 Thailand , Dec.17-21, 1990
Rahman, A. A, , Bangladesh coastal environment and management , CARDMA workshop Dhaka , October, 1988.
Chowdhury, G. R. , Khan, T. A. , Developing the Gange Basin , National symposium on River Basin Development Dhaka , December, 1981
LDL , Coastal Embankment Project , Engineering and Economic Evaluation Vol - Ⅰ , 1968
MPO , Second Interim Report , Vol.Ⅷ , 1984
Rahman, M. T. , Upakulia Samassay Varshymahin Paribasha Bangladesh , Pani Parikkrama May-June 1990
7th edition, 6th year
MPO , National Water Plan Vol. Ⅰ Sector Analysis , 1986
CKC , Completion Report of Khulna Coastal Embankment Rekabilitation Project , 1991
Shahsjahan, M , Hossain, A.H.M.A. , Beel Dakatia: a review of environmental situation , IEB technical session March. 1991
35th convention
Khan, H.R. , Regional symposium , Dhaka , 1985
Barua, D.K. , Some consideration on the selection of the height of empoldering level in the newly accreted south eastern deltaic region of Bangladesh , May, 1990
IECO , Master plan , Volume-Ⅱ , 1968
분야 : SEDIMENT TRANSPORT PATTERN OF ALLUVIAL RIVERS IN KOREA Lee, Soontak and et al. , Progress Report of International Hydrological Program in Korea , Ministry of Construction , Korea , 1982-1988
Lee, Soontak and et al , Sediment Transport Pattern in Nakdong River, Korea , Proceedings of the 4th International Symposium on River Sedimentation Vol.3 , Beijins, China , 1989
- 245 - Cho, K.K. , S.W. Park , Soil Erosion on Upland Slope , Journal of the Korean Society of Arricultural Engineer Vol .23 , No. 2 , 1981
Ahn, S.Y. , B.h.Min , A study on Sediment Load in the Milyang River , Journal of the Korean Society of Agricultural Engineers Vol .22 , No.4 , 1980
Kim, H.J. and et al. , A study on Sediment Load in River(Ⅰ) , Journal of the Korean Association of Hydrological Sciences Vol. 14 , No. 3 , 1981
Kim, H.J. and et al. , A study on Sediment Load in River(Ⅱ) , Journal of the Korean Association of Hydrological Sciences Vol. 15 , No.2 , 1982
Nahm, S.W , A Study on the Sediment and the River Bed Variation , Journal of the Korean Association of Hydrological Sciences Vol. 11 , No. 1 , 1978
ISWACO , Makdong Estuary Barrage and Reclamation Project , Daejeon, Korea , 1982
Final Design Report
Lee, J.K. , Reservoir Sedimentations of the Enclosure of Estuary Barrage in Gumgang Basin , Journal of the Korean Assoication of Hydroloaical Sciences Vol. 9 , No. 1 , 1976
KIST and ODI , A Study on Recent Sedimentation in the Nakdong River Estuary , Report PE 00011-20-5 Seoul . Korea , 1978
Bruun, P. , The Stability of Tidal Inlets , Theory and Engineering 500 pp. , Amsterdam, Oxford, New York , Elsevier Sci.Publ. Co. , 1978
분야 : SEDIMENT TRANSPORT PATTERN OF ALLUVIAL RIVERS IN KOREA Chow, V. T. , New trends in hydrology , Nature and Resources 3 , 4-9 , 1967
Duck, R. W. , The effect of road construction on sediment deposition in Loch Earn. Scotland , Earth Surface Processes and Landforms 10 , 401-406 , 1985
Gibbs, R. J. , A settling tube system for sand-size analysis , J. Sed. Petrol . 44 , 538-588 , 1974
Gibbs, R. J. , Matthews, M. D. , Link, D. A. , The relationship between sphere size and settling velocity , J. Sed. Petrol. 41 , 7-18 , 1971
Mizuyama, T. , Measurement of wash load in mountain rivers , Civil Engineering J. 22 , 46-51 , 1980
Nippes, K .R. , A new method of computation of the suspended sediment load , In: Mathematical Models in Hydrology no. 101 , 659-666 , IAHS Publ , 1974
Park, J. K. , Suspended sediment yield at Yamaguchi River basin in Mt. Tsukuba , Bulletin of the Environmental Research Center no. 14 , 99-108 , The University of Tsukuba , 1990
Park, J. K. , Characteristics of sediment discharge in a mountain stream viewed from the variation pattern of suspended sediment concentration , Trans 12-1 , 51-67 , Japanese Geomorphological Union , 1991
Walling, D. E. , Gregory, K. J. , The measurement of the effect of building construction on drainage basin dynamics , J. Hydrology 11 , 129-144 , 1970
- 246 - Wood, P. A. , Controls of variation in suspended sediment concentration in the River Rother , 24 , 437-445 , West Sussex, England , 1977
Sedimentology
분야 : MATHEMATICAL SIMULATION OF NAVICATION CHANNEL CHANGES IN THE FLUCTUATING BACKWATER REGION OF THE THREEGORGES RESERVOIR Dou Guoren , Zhao Shiqing , Huang Yifen , Study on two-dimensional total sediment movable bed mathematical model , Preceeding of 4th International Symposium on River Sedimentation Beijing , 1989
Han Qiwei , Theoretical basis of a general mathematical model for reservoir deposition and channel evolution , Institute of Water Resources and Hydroelectric Power Research Report 1986
Li Yitian , Xie Jianhen , Mathematical modeling of two-dimensional flow in alluvial rivers , Journal of Hydraulic Engineering No. 11 , 1986
Li Yitian , A primary study on the calculation of two-dimensional river bed deformation in alluvial rivers , Journal of Sediment Research No. 1 , 1988
Li Yitian , Yin Xialoing , Influence of unsteady flow on the steady sediment mathematical model simulation result , Wuhan University of Hydraulic and Electric Engineering Report 1990
Wu Weiming , Li Yitian , On the connecting problems of one- and two-dimensional model , Wuhan University of Hydraulic and Electric Engineerirg Report 1991
YVPO , Comprehensive analysis of computations for sediment deposition in the Three Gorges Porject , Compendium of research reports on sedimentation problems of the Three Gorges Project published by the department of Science and Technology, Chineses Ministry of Water Resources and Electic Power , 1988
Zhang Ruijing , , River mechanics Industry Press , 1961
분야 : A COMPUTER-AIDED GIDELINE FOR THE SELECTION OF SEDIMENT TRENSPORT FORMILAS Ackers, P. , White, W. R. , Sediment Transport : New Approach and Analysis , J.of Hyd. Div. Vol .99 , HY11 , ASCE , 1973
Brownlie, W. R. , , Prediction of Flow Depth and Sediment Discharge in Open Channels W. M. Keck Laboratory of Hydraulics and Water ResourceCalifornia Inst. of Technolog , Pasadena, California , Nov., 1981a.
Report . No. KH-R-43A
Brownlie, W. R. , , Compilation of Alluvial Channel Data: Laboratory and Field W. M.California , Pasadena, California , Nov., 1981b.
Report No. KH-R-43B
CoIby, B. R. , Discharge of Sands and Mean Velocity Relationships in Sand-Bed , U.S. Geological Survey 1964
Professional Paper 462-A
DuBoys, P. , Le Rhone et les Rivieres, a Lit Channels , Annales des Ponts et Vol. 18 , 1879
- 247 - Einstein, H. A. , The Bed Load Function for Sediment Transportation in Onen Channel Flows , Technical Bulletin 1026 U.S. Department of Agriculture, Soil Conservation Service , 1950
Engelund, F. , Hansen, E. , , A Monograph on Sediment Transport in Alluvial Streams Copenhagen, Denmark , Teknisk Vorlag , 1967
Nordin, C. F. Jr. , Beverage, J. F. , Sediment Transport in the Rio Grande New Mexico , U.S. Geological Survey 1965
Professional Paper 462-F
Ranga Raju, K. G. , Garde, R. J. , Bhardwaj, R. C. , Total Load Transport in Alluvial Channels , J. of Hyd. Div. Vol .107 , HY2 , ASCE , 1981
Rijn, L. C. van , Sediment Transport, Part Ⅰ : Bed Load Transport , J. of Hyd.Div. Vol.110 , HY10 , ASCE , 1984
Rijn, L. C. van , Sediment Transport, Part Ⅱ : Suspended Load Transport , J. of Hyd. Div. Vol.110 , HY11 , ASCE , 1984
Shen, H. W. , Hung, C. S. , An Engineering Approach to Total Bed Material Load by Regression Analysis , Proc. of Sedimentation Symposium Berkeley. California , 1971
Shen, H.W. , Wash Load and Bed Load, Suspended Load, Total Sediment Load , River Mechanics Vol .1 , Fort Collins, Colo. , 1971
H.S. Shen edited
Stevens, H.H. Jr. , Yang, C.T. , Summary and Use of Selected Fluvial Sedi-ment Discharge Formulas , U.S. Geological Survey 1989
Water Resources Investigation Report 89-40 26
Toffaleti, F. B. , Definitive Computations of Sand Discharge in Rivers , J. of Hyd. Div. Vol.95 , HY1 , ASCE , 1968
Whites, W. R. , Milli, H. , Crabble, A.D. , Sediment Transport : AnAppraisal Method, Vol.2, Performance of Theoretical Methods. When Applied to Flume and Field Data , Hydraulics Research Station , Wallingford, United Kingdom , Nov., 1973
Report No. IT.119
Yang, C. T. , Incipient Motion and Sediment Transport , J of Hrd. Div. Vol .99 , HY10 , pp.1679- 1704 , ASCE , 1973
분야 : PRELIMINARY STUDY ON SEDIMENT PROBLEM AT HOABINH RESERVOIR Long Yuqian , Measuring techniques of reservoir sedimentation , Beijing, China , 1985
G. W. Annandale , Reservoir Sedimentation , Elsevier, Amsterdam, Oxford, New York, Tokyo , 1987
Fan Jiahua , Methods of preservion reservoir capacity , Beijing, China , 1985
Ding Lianzhen , Computation of deposition and scoring in reservoir , Beijing, China , 1985
- 248 - 분야 : RESERVOIR SEDIMENTATION AND WATER QUALITY IN KOREA Bostrom, B. , Ⅰ. Ahlgren , R. Bell , Internal Nutrient Loading in a EutrophicLake. Reflected in Seasonal Variations of Some Sediment Parameters , Verein, Limnol No.22 , pp.3335-3339 , 1985
Fillos, J. , W. R. Swanson , The Release Rate of Nutrients from River and Lake Sediments , J. of Water Pollution Control Fed. Vol.47 , No.5 , pp.1032-1042 , 1975
Hosomi, M. , R. Sudo , Characteristics of Phosphorous Distribution in Sediments of Lake Kasumigaura , NIES report No.22 , pp.45-54 , 1981
Jun, S.H. , Fractional Composition of Sediment Phosphorous and Potential for Water Pollution in Artificial Reservoirs in Chunchon Area, Korea , J. of Korea Water Pollution Res. Contr. Vol .4 , No.2 , pp.49-57 , 1988
Jun, S. H. , Forus and Mobility of Pollutions Retained in the Sediments , Korean J. of Limnology Vol.23 , No.5 , 1990
Jun, S. H. , Y. A. Park , Forms and Mobility of Sediment Phosphorous in Lake Soyang , Korean J. of Limnology Vol. 22 , No.2 , pp.261-271 , 1989
KOWACO , Report on the Sediment Survey of the Andong Multipurpose Reservoir , Korea Water Resources Corporation , 1983a
KOWACO , Report on the Sediment Survey of the Soyang Multipurpose Reservoir , Korea Water Resources Corporation , 1983b
KOWACO , Report on the Sediment Survey of the Namgang Multipurpose Reservoir , Korea Water Resources Corporation , 1983c
KOWACO , Report on the Sediment Survey of the Sumjin Multipurpose Reservoir , Korea Water Resources Corporation , 1983d
KOWACO , Report on the Long Term Planning for the Water Resources in Korea , Korea Water Resources Corporation , 1990a
KOWACO , Water Quality Conservation Using a Ecosystem Model for Daechong Reservoir , Korea Water Resources Corporation , 1990b
KOWACO , A Study on the Effect of Inland Fish-farm on the Reservoir Water Quality , Korea Water Resources Corporation , 1991
Moniwa, T. , Study on the Effect of Bottom Sediments in River , Japanese J. of Service Water and Waste Water Vol. 1 , No. 1 , pp.63-69 , 1978
Tchobanologlous, G. , E.D. Schroeder , , Water Quality Addison Wesley Publishing CO. , 1985
Vollenweider, R. A. , Input-Output Model with Special Reference to the Phosphorous Loading Concept in Limnology , Schweiz Z. Hydro. 37 , pp.53-84 , 1975
분야 : RESERVOIR SILTING PATTERS FROM A TWO-DIMENSIONAL HYDRODYNAMIC MODEL Abbott, M. B. , , Computational Hydraulics London , Pitman Publishing Ltd. , 1979
ASCE , Fundamentals of sediment transportation , 97 , HY12 , p.1979-2023 , ASCE , 1971
Proc.
- 249 - Kim, H. Y. , S. W. Park , Simulating daily inflow and release rates for irrigation reservoirs (3) - model application to daily reservoir operations , Journal of the Korean Society of Agricultural Engineers Vol. 30 , No. 3 , p. 95-105 , 1988
Lee, B. H. , Predietion of Reservoir Sedimentation Patterns Using a Two-Dimensional Transport Model , Seoul , 1990 thesis for the degree of Master of Science presented to Seoul National University
Pais-Cuddon , I. C. dos M. , N. C. Rawal , Sedimentation of reservoir , Journal of the Irrigation and Drainage Division IR3 , p. 415-429 , ASCE , 1969
Proc.
Park, C. E. , S. W. Park , Application of depth-averaged two-dimensional mathematical model to tidal computations in the estuary near Gunsan port , Journal of the Korean Society of Agricultural Engineers, Vol. 28 , No. 1 , p. 60-67 , 1986
Thomas E. Croley 2 , K. N. Raja Rao , Fazle Karim , Reservoir sedimentation model with continuing distribution, compaction, and sedimentation slump , Iowa Institute of Hydraulic Research, The University of Iows , Iowa , 1978
Report No. 198
USBR , Design of Small Dams , p. 529-563 , Denver, Colorad , A Water Resources Technical Publicatior , 1987
Third Edition
Vanoni, V. A. , Sedimentation Engineering , ASCE , New York , 1975
분야 : PROBLEM OF EROSION AND DEPOSITION IN THE MAJOR RIVERS OF BANGLADESH Alam, M. K. , Hossain, M. M. , Sediment transport in the river Jamuna , Journal of IEB Vol. 16 , No. 2 , pp. 17-26 , Dhaka , 1988
BWDB , Flood in Bangladesh , Ministry of Irrigation, Water Development and Flood Control , Govt. of Barigladesh , December, 1987
BWDB , Damage due to flood in 1988 , Ministry of Irrigation, Water Development and Flood Control , Govt. of Bangladesh , December, 1988
Coleman, J. M. , Brahmaprtra river : channel process and sedimentation , Sedimentary Geology Vol. 3 , No.2/3 , pp. 129-239 , 1969
Galay, V. J. , River channel shifting on large rivers in Bangladesh , Proceedings of the International Symposium on River Sedimentation Vol. 1 , pp. 543-562 , 1980
Ed. by Chinese Society of Hydraulic Engineering
Habibullah, M. , Computer modelling of river channel changes in alluvial condition , IFCDR, BUET , Dhaka , 1967
First interim report, R02/87
- 250 - Hossain, M.M. , Geomorphic Characteristics of the Ganges upto Brahmaputra confluence , Final report, R02/89 pp. 158 , IFCDR. BUET , Dhaka , 1989a
Hossaimn, M.M. , Siltation of river bed and flood problem , Flood in Bangladesh Dhaka , Published by Community Development Library , 1989bpp. 77-83
Edited by Mohiuddin Ahmed
Hossain, M.M. , Meander travel of the Ganges within Bangladesh during 1973-1983 , Proceedings of the seminar on Technolog of self reliance and development Dept. of Water Resources Engineering, BUET and Water resources Division , Bangkok, Dhaka , AIT , November, 1988
Hossain, M.M. , Total Sediment Transport of the lower Ganges and Jamuna , Proceedings of the Seminar on Costal Environment and Development Bangladesh National Geographical Association , Khulna , May, 1990
Hossain, M.M. , Shahjahan, M. , Ahmed, E. , Assessment of sediment transport in the Teesta river , Bangladesh , 1991
Paper accepted for publication in the Journal of the Institution of Engineers
Latif, A. , Investigation of Brahmaputra river , Journal of Hydraulics Division Vol. 95 , No. Hy5 , pp. 1687-1698 , ASCE , 1969
Mukhopadhyay, S. C. , The Teesta basin, A study in fluvial geomorphology , Calcutta , K. P. Bagchi and Company , 1982
Biswas, A.K. , Khan, A.H. , NIshat, A. Ed. , River basin development , Water Resources Series vol 4 , Dhaka , TycolIy International Publishing Ltd. , 1983
분야 : RIVER PROCESSES DOWNSTREAM OF A RESERVOIR Report for Construction of Karnafuli River Bridge at Chittagong , Govt. of Jasan. , East Pakistan , by Overseas Technical Co- operation Agency , March, 1966
EPWAPDA , Karnafuli Irrigation Project , Feasibility Study Volume 1 , Part 1,Halda Unit , Consulting Engineers , Dacca. East Pakistan , May 1968
Prepared by Justin-Courtney-Hohlweg-Watts
Chittagong Port Entrance Study , Annex Ⅳ , Parts A ＆ B , NEI , The Netherlands , Rotterdam , 1978
Draft Final Report
Dhanmondi R. A. , Technical aspects , Techno-economic Study for Construction of Karnafuli Bridge at Chittagong Vol. 1 , Consulting Engineers and Architects , Dhaka , by Bangladesh Consultants Ltd. , Sept. 1985
Studies on Sand and Silt Movement in Chittagong Harbour by Radioactive Tracer Techniques - Sikder, M. D. H. , Final Draft Report 1991 INST, AERE, Savar , Dhaka ,
- 251 -