INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Development of Bayesian Network Based Dam Risk Analysis

And Its Application to Rockfill Dam in South Korea

Byoung Han Choi

Rural Research Institute, Korea Rural Community Corporation, South Korea

Hyun-Han Kwon
Department of Civil Engineering, Chonbuk National University, Jeonju-si, South Korea
[email protected]

ABSTRACT: BAYESIAN NETWORK BASED DAM RISK ANALYSIS

Risk assessment is an essential component of dam safety, and statistical analysis plays central role
in dam risk assessment. Dam risk analysis relies on a series of probabilistic analyses of hydrologic,
geotechnical, structural models and their associated inputs. In fact, existing risk analysis problems
try to forecast what risks may be associated with situations that are only hypothetical. In this
regard, there may be little or even no empirical data available for some variables. It may be
necessary to employ subjective information from the expert opinion, and uncertainty about the dam
risk model used may be substantial. In this regard, a Bayesian model might be helpful in
addressing these issues. Bayesian analysis allows better calculation of the probability of a
particular hypothesis based on quantitate and/or qualitative information. Bayesian network
methods natively consider the uncertainty associated with the parameters of a probability model.
An application to a Rockfill Dam in South Korea illustrates how the Bayesian network approach
can lead to potentially realistic estimates of dam safety, while the existing approach provides
rather unreasonable estimates and an assessment of their sensitivity to key features.

Keywords: Bayesian Network, Risk Analysis, Uncertainty, Dam Safety

1. INTRODUCTION

Many countries have experienced frequent floods and these have led to significant social
and economic damages that include destructions in infrastructure, loss of social capital,
large scaled economical expenses, and the loss of life. For this reason, the capacity of
emergency spillways as well as dam safety standards must be enough to meet the safety
needs. Most countries currently accepts the PMF as a standard for dam safety. Generally, it
is considered that a dam spillway that has been designed using PMF represents a fewer
risks and security at least at the time of construction (Kwon and Moon, 2006). However,
the PMF cannot be regarded as a fixed value in a changing climate condition. In addition, a
traditional dam safety approach based on the PMF is not designed to bring into account
uncertainties associated with hydrologic, geotechnical and structural aspects.

V- 1
Numerous studies have been conducted to explore the potential used of the risk-based
analysis in dam safety dams over the last three decades (Bowles, 1987; McCann et al.,
1985; Von Thun, 1987; Resendiz-Carrill and Lave, 1987; Karlsson and Haimes, 1988a, b;
Haimes et al., 1988; Karlsson and Haimes, 1989; Petrakian et al., 1989; Kwon and Moon,
2006; Kuo et al., 2007; Mo et al., 2008; Goodarzi et al., 2012). The risk analysis in dam
safety is generally used to evaluate overtopping probability associated with rare events and
to calculate the relative frequency of extreme events with the PMF (Kwon and Moon,
2006). Kwon and Moon (2006) developed a nonparametric Monte Carlo simulation
technique in conjunction with Latin-hypercube sampling and attempted to address complex
system in the hydrologic dam risk model. Kuo et al. (2007) employed five different
uncertainty analysis techniques (i.e. MFOSM, RPEM, HPEM, LHS, and MCS) to estimate
the overtopping probability for the Feitsui Reservoir. Mo et al. (2008) developed the dam
risk model considering joint effects between flood and wind wave induced loads. Goodarzi
et al. (2012) demonstrated the bivariate flood frequency based dam risk model to estimate
overtopping probability.
As aforementioned, the probabilistic dam risk model would be useful for assessing the
associated uncertainties and risks. The risk-based analysis is a potentially relevant
approach for assessing economic feasibility, social safety, and priority of the dam
rehabilitation. In this regard, a main objective of this study is to develop a fully integrated
dam risk model that can consider possible risk factors, various potential failure modes and
potential damage costs in a statistical framework. Also, we would like to take into account
joint failure events publicly. Major innovations are introduced in this paper to address
these situations. Much focus is placed on a fully integrated probabilistic dam risk analysis
approach using Bayesian networks (BNs). BNs based dam risk model can be very
appropriate for considering the aforementioned issues and associated consequences. In
addition, the use of Bayesian Monte Carlo simulation model for uncertainty analysis of the
rainfall-runoff model and extreme rainfalls are also proposed. The paper is organized to
first introduce concepts and mathematical formulation of BNs for dam safety assessment.
A Bayesian dam risk model for rockfill dam is then presented. Finally, the proposed
approach is applied to the analysis of the safety of an existing dam in South Korea.

2. Bayesian Networks based dam risk Model

In this section, we will discuss an essential concept of BNs and further provide BNs based
dam risk model.

2.1. Concept of BNs

A Bayesian network is a probabilistic graphical model that represents a set of random

variables and their conditional dependencies via a directed acyclic graph (DAG). In BNs,
the nodes represent events and directed edges capture their hierarchy and dependence.

Figure 1. Concept of BNs. A causes B or B is a consequences of A

V- 2
We would say that node A is a parent of B, B is a child of A, that A influences, or causes B,
and B depends on A. This is captured by a directed edge (arrow) that leads from A to B.
The term acyclic in DAG relates to the fact that a closed loop of dependencies is not
allowed. The independence of two nodes in a DAG depends on their relative position in the
graph as well as on the knowledge of other nodes conditioning in the graph.
The influences between nodes in BNs can be either probabilistic or deterministic. Let
denote the parent set of variable E as Par (E ). DAG must be defined together with
conditional probability given its parents. If Par (E ) = 0 then | .

,…, ∏ | (1)

The conditional probability tables of each node given its parent nodes for discrete case
need to be estimated in the BNs. In the case of continuous nodes, one need to discretize the
continuous variables into a number of states, and then analyze in the discrete BNs. The
probabilities in the BN allows to update when new information becomes available.

2.2. BNs Based Dam Risk Model

To begin with, this study reviewed the literature on earth fill dam failures in order to
explore possible risk influence factors, and then 7 factors were included as the factors for
this study. The variables are summarized in Figure 2, along with their descriptions. In this
study, slope stability (SPS) and overtopping (OTP) are considered as a main cause of the
dam failure. This study first assumed that the overtopping failure is affected by the
spillway failure (SPF) which depends on both the spillway functionality (SP) and the
auxiliary spillway functionality (ASP). In addition, the unexpected flood (UEP)
simultaneous influences the risk of overtopping and slope stability, which is the major
driving failure causes in the dam risk analysis. Finally, the crack of the dam (CRC) can
lead to the slope stability failure.

Figure 2. A Bayesian network for the case study. Each node is regarded as a discrete variable with
conditional probability tables (CPT) given parents

V- 3
Assuming that the occurrence probabilities of main gate failure and auxiliary gate failure in
spillway are 0.068 and 0.040, respectively. Based on the uncertainty analysis of
hydrological modeling and the data reported by government, corresponding loading
probabilities and system response probabilities associated with dam failure are estimated
and represented in Figure 2. As shown in the Figure 2, the relationships defining dam
failure mechanisms are strongly inter-dependent process and traditional event tree, which
treats each node as an independent process, will not be appropriate in this setting. In this
sense, this study aims at utilizing the BNs to facilitate such disadvantages with easy.

3. RESULTS

Statistical inferences are analyzed on BNs as represented by Figure 3 to characterize the

overall risk. One can identify the most contributing variable to the overall risk through the
inferences, and the estimated probability of dam breach (BRC) is strongly related to the
spillway functionality (SPF) that is one of the most contributing factor to the overall risk. It
was shown that successful management of the spillway would lead to reducing the overall
breach risk. In other words, securing sufficient discharge capacity along with relevant
mechanical management of the spillway gates would be a significant issue in dam risk
management. Therefore, construction of emergency spillway can be one of the most
effective measure to reduce the probability of dam failure.

Figure 2. Statistical inferences through Bayesian Networks. Condition probabilities given hard
threshold is provided in the right panel.

4. CONCLUSTION

Risk assessment is an essential component of dam safety, and statistical analysis plays
central role in dam risk assessment. Dam risk analysis relies on a series of probabilistic
analyses of hydrologic, geotechnical, structural models and their associated inputs.
However, it may be necessary to employ subjective information from the expert opinion in
estimating loading probabilities and system response probabilities due to a lack of

V- 4
available data for some risk factors. Thus, uncertainty about the dam risk model may be
substantial. This study aimed at developing the Bayesian Network model for dam risk
analysis which might be helpful in addressing these issues. Bayesian network models
natively consider the uncertainty associated with the parameters of a probability model.
The use of Bayesian Networks based dam risk model allowed to better quantify the overall
risk associated with various risk components and to compare the risk reduction measures
among structural and non-structural approaches. In addition, the proposed BNs model has
allowed to identify the most influencing risk factors to the overall dam failure risk.

ACKNOWLEDGEMENT
This research was supported by a grant (11CTIPC02) from Construction Technology Innovation
Program (CTIP) funded by Ministry of Land, Transport and Maritime Affairs of Korean
government.

REFERENCES

Bowles, D.S. (1987): A comparison of methods for integrated risk assessment of dams,
Plate E (eds) Engineering reliability and risk in water resources, Dordrecht,
Netherlands.
Goodarzi, E., Mirzari, M. and Ziari, M (2012): Evaluation of dam overtopping risk based
on univariate and bivariate flood frequency analyses, 39:3, pp. 74-87, Canadian
Journal of Civil Engineering.
Haimes, Y.Y., Petrakian, R., Karlssion, P.O. and Mitsiopoulos, J. (1988): Risk Analysis of
dam failure and extreme floods: application of the partitioned multiobjective risk
method, IWR Report 88-R-1, Environmental Systems Management, Charlottesville,
U.S.A.
Karlsson, P.O. and Haimes, Y.Y. (1988b): Probability distributions and their partitioning,
24:1, pp. 9-20, Water Resources Research.
Karlsson, P.O. and Haimes, Y.Y. (1989): Risk assessment of extreme events: Application,
115:3, pp. 299-320, Water Resources plan and Management.
Kuo, J.Y., Yen, B.C. and Hsu, Y.C. (2007): Risk analysis for dam overtopping Feitsui
reservoir as a case study, ASCE 133:8, pp. 955-963, Journal of Hydraulic
Engineering.
Kwon, H.-H. and Moon, Y.-I. .(2006): Improvement of overtopping risk evaluations using
probabilistic concepts for existing dams, 20:4, pp. 223-237, Stochastic
Environmental Research Risk Assessment.
McCann MW Jr (1985): Proceedings of a Workshop on Current Developments in Dam
Safety Management.
Mo, C.X., Dong, Z.C. and Ma, R.Y. (2008) Application study on overtopping risk analysis
for Guangxi Chengbihe reservoir by integrate-FOSM method, 27:2, pp. 44-49,
Journal of Hydroelectric Engineering.
Petrakian, R., Haimes, Y.Y., Stakhiv, E. and Moser, D. (1989): Risk analysis of dam
failure and extreme floods, risk analysis and management of natural and man-made
hazards, ASCE: 81-121, U.S.A.
Resendiz-Carrillo, D. and Lave, L.B. (1987): Optimizing spillway capacity with an
estimate distribution of floods, 23:11, pp. 2043-2049, Water Resources Research.
Von Thun JL (1987): Use of risk-based analysis in making decisions on dam safety, Plate
E (eds) Engineering reliability and risk in water resources, Dordrecht, Netherlands.

V- 5
 IN TERN ATIO N AL SYM PO SIU M O N

Bali, Indonesia, June 1ST – 6TH , 2014

Spatial Analysis to Identify Sources of Debris (Trees)

hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
Along Hydropower Rivers fffffjfjjfkkfjjj
Case study Pite River, Sweden
2(14pt) A. Söderström
Sweco International AB, Stockholm, Sweden
[email protected]

M. Hansson
Sweco International AB, Stockholm, Sweden

M. Johansson
Skellefteå Kraft AB, Skellefteå, Sweden

V. Carlsson
Skellefteå Kraft AB, Skellefteå, Sweden

ABSTRACT:
Debris in hydropower rivers constitutes a major potential risk for dams, especially during
high flows when the spillway gates have to be available to discharge water through the
spillways without disturbance due to aggregation of debris in front of the gates. One major
source of debris is trees falling into the water from the river shores. In particular trees
growing on steep slopes with instable soil types on areas subject to flooding are a potential
risk for being a source of debris in the form of trees.

In order to gain knowledge of the debris situation along a river, spatial analysis of some
important river characteristics is developed. The spatial analysis is developed to find
sources of debris production along a river and it is based on the following information:
• The slope of the river bank. Based on LiDAR scanning (Light detection and ranging)
of the river valley. Steep river banks increases the risk for trees falling into the
water and thereby being transformed into debris.
• Soil type. Soils with instable behavior in steep ground are important to identify.
• Inundation areas. Areas subject to flooding in combination with steep slopes and
instable soil types increase the risk for trees falling into the water. A hydraulic
model has been setup to calculate water levels at high flows. The results are
presented as flood maps.
• Vegetation types. Tall trees growing on the river shores are the largest risk when
arriving to the dam. For the case study, areas with trees with a length of around 10
meter and larger are identified and analyzed.
• Length of trees. Calculated by LiDAR scanning of the area.

V- 6
All data above is collected as vector objects (polygon or point data). A spatial analysis is
conducted in the Geographical Information System ArcGIS.
The results are compared to field surveys conducted after previous floods and geological
surveys of old landslides.

Keywords: Debris, Trees, Hydropower, Spatial Analysis, ArcGIS.

1. INTRODUCTION

Extreme floods, heavy rains and storms increase the risk for damage on dams. During these
events erosion along rivers increase and trees and other debris are at risk to be moved into
the river.

Floating debris may cause blocking of spillways and cause damage that may lead to
overtopping of dams and thereby initiating a dambreak. Earth-fill dams are at special risk
during an overtopping event but also other dams may be damaged during such an event.

Wind-fallen trees in flood sensitive areas along a river, forest in spillways channels,
sunken timber, swim decks, small huts and floating peat are examples on floating debris
posing a threat to dams. Mitigation measures to avoid damage may be removal of trees on
the river shores and/or fallen trees that can be caught by the stream. There may be
difficulties executing these kind of measures. Forest on river shores are often protected
through logging regulations and vegetation may stabilize instable slopes.

Debris booms installed across the river may be effective for catching floating debris before
it reaches the spillway. During a high flow a debris boom may cause higher water levels
upstream, which has to be evaluated. It may cause flood sensitive areas in the upstream
area.

Other measures are debris traps in tributaries or debris screens in front of spillways.

In order to execute preventive measures some background material is needed to estimate

the potential number of trees falling into the water at extreme situations and where these
trees may enter the river. Thereby a debris management plan can be executed on the right
place in the river. To calculate potential debris yield in terms of trees originating from river
shores a geographical analysis method has been developed and tested.

The analysis is based on the following material:

• High resolution terrain model (built up by LiDAR scanning of the river valley)
• Digital soil maps (vector format for GIS)
• Aerial images for interpretation (orthophotos)
• Hydraulic model for the study area

It is of great benefit if the results from the analysis can be compared with geological
investigations of historical landslides and field studies where the river shores are inspected
after high flows.

2. METHODOLOGY - SPATIAL ANALYSIS OF DEBRIS SOURCES

V- 7
The method to identify potential floating debris from trees along rivers is developed by
using ArcGIS Geographical Information System. The objective with the method is to find
potential risk areas and calculate the number of trees that may fall into the river and
become floating debris. Trees are counted using the complete LiDAR-scanning data set
where vegetation is registered.

As a pilot study Pite River in Northern Sweden was selected as a test area. This was done
by several reasons.

• At Sikfors hydropower plant debris is collected on regular basis, see Figure 1.

• There is only one dam in the river making the study area basically undisturbed
given that the study is limited to the upstream area of the dam.
• A flood mapping study had previously been conducted in the river, enabling a study
of water levels at various high flows.
• A field inventory of the river shores was done prior to the study. In 2010 the spring
flood was larger than the average spring flood which made it possible to study the
slopes of the river shore after the water had been sinking for some week. As the
slopes were saturated from the high water levels, it was possible to study small
landslides along the sandy parts of the river shores, see Figure 2.
• A LiDAR scanning was completed for the river valley in 2013.

Figure 1. Type of debris collected on regular basis at Sikfors hydropower plant.

V- 8
 Figure 2. Fresh landslide after the spring flood 2010.

In Figure 2, note also the fallen trees on the shore to the left of the landslide. These trees
fell due to a smaller landslide in the lower part of the river bank indicating that a large
portion of the river bank may collapse in a near future. The fresh landslide is 50 meter
wide and the whole section of high risk area on both sides of the landslide is 100 m to the
left and 200 meters to the right.

2.1. Spatial Analysis

The chapters below present each step in the spatial analysis. The idea is to start using all
available data in the analysis and then, by applying different filters, establishing a work
process that re-use only the most important information in each step for the debris
inventory.

2.1.1 Slope
An overlay analysis is done with areas that contain both steep river banks (typically slopes
of 20-45%) and soil types with instable behavior (loose sediments and fine materials).

2.1.2 Soil type

By conducting this analysis more stable soils such as compacted till and rock as well as
river banks with mild slopes are excluded from further studies.

V- 9
2.1.3 Vegetation type
This part of the study is done with orthophotos and forest polygons from the national land
register is used as background. Ground without forest is excluded from the analysis.
Orthophotos are used as a visual inspection of the forest polygons which is based on the
national land use register.

2.1.4 Inundation areas

The remaining areas from the analysis in Ch. 2.1.1-2.1.3 are further analyzed with respect
to flood risk. If forest areas with instable soil types and steep slopes are subject to flooding
the risk for debris production increases. A hydraulic model was setup to calculate water
levels along the river. The calculated water levels are interpolated over a digital elevation
model of the river valley, generating a flood map.

2.1.5 Tree identification

After analyzing the potential risk areas tree lengths in these areas are derived from the
LiDAR-scanning. LiDAR data contain unclassified data including vegetation points. Trees
longer than 10 meters are used for counting. The tree length of 10 m is selected due to the
fact that there are three spillway gates in the dam, each with a width of 15 meters. Trees
below 10 meters length may pass the spillway if they arrive individually.

2.1.6 Tree count

The last step in the analysis is to count all trees in the different risk areas. Thereby various
parts of the river can be studied with respect to debris management.

3. ANALYSIS OF PITE RIVER

The background material used for the Pite River study consist of LiDAR scanning of
ground and vegetation, hydraulic model, soil maps, database of historical landslides and
documentation (images and notes) from a field inventory after a high flood.

By weighing information about slope, soil types, flooding areas and vegetation type it was
possible to establish risk areas along the river, see Figure 3.

V - 10
 Figure 3. Risk areas for landslides.

To calculate the potential yield of trees for becoming floating debris tree lengths were
calculated. The number of trees and the tree heights are calculated in ArcGIS, see Figure 4.
The LiDAR scanning have an accuracy of ±0,1 m in vertical direction. From this analysis a
tree count could be done for each of the high risk area.

Figure 4. Tree lengths for trees located in landslide risk areas that are prone to flooding.

V - 11
4. VALIDATION OF METHOD

High risk areas for landslides were localized in the terrain model. These identified areas
were compared to the historical landslide database built up by Swedish Geological Survey.
Aerial images as well as notes and images from a field inventory served as calibration and
validation material.

Validation of tree counting was done by comparing the calculated number of trees in the
LiDAR scanning with some selected test areas where trees could be calculated manually in
aerial images, see Figure 5. The yellow points in Figure 5 (right) are trees identified in the
GIS-analysis of the LiDAR scanning data. The yellow points are displayed on top of an
orthophoto enabling a control of the actual location of trees.

Figure 5. Visual inspection of orthophoto (left) and tree count method (right).

An ocular comparison was done for each of the high risk area, see Figure 6. The
comparison shows that all known landslides are inside the high risk area. Also several
fresh landslides and erosion sites are located within the high risk areas. High risk areas
with no documented historical landslides from the database could be visually inspected in
the digital terrain model. A large number of small and large erosion scars were visible.
 

V - 12
  

Figure 6. Calibration point for calculated risk area and fresh landslide after spring flood 2010.
 
Referring back to Figure 2. In Figure 7 below the landslide is shown as a digital elevation
data (left) and the risk analysis for landslide (right). It is noticeable when analyzing the
slope of this river bank is that the 50 m wide area of landslide has turned from orange to
yellow in the analysis. That can only be interpreted as the slope of the collapsed river bank
is more flat, more stable and it is not likely that this part will slide again (although the area
is now deforested and may be subject to erosion and slide for that reason). However, an
area with a recent landslide history and small erosion scars (see text in Figure 2) on large
parts of the lower river bank needs to be monitored somehow as part of a debris
management plan. It could easily fall some 100 trees in the river at this location. These
would need to be picked up before reaching the dam to prevent risk of damage to the
structure and to ensure usability of downstream located dam.
 

Figure 7. Calibration point for calculated risk area and fresh landslide after spring flood 2010.

V - 13
In Table 1 below a quantification of the calibration results has been done. The resulting
risk areas for landslides and debris yield from the spatial analysis is compared with two
sources of information:
• The historical landslide database from the National Geographic Survey
• A photo series of 155 images taken at a 60 km long part of the river after a high
flow. The 155 images are distributed to 32 sites along the river. Of these 32 sites 5
sites show clear signs of fresh landslides and 4 of these sites are located in areas
with risk level of 4 or 5.

Table 1. Summary of found landslide scares and there location concerning the computed risk areas.

Risk areas 1 2 3 4 5 In risk

areas 4
or 5
By SGU located 0 0 1 5 2 87.5 %
landslides

By Sweco located 0 0 0 2 2 100 %

landslides

5. RESULTS & DISCUSSION

 
The result from the study has been handed over to the dam owner for Sikfors HPP. It will
serve as decision support material for handling of debris in the river. During the spring
flood 2014 a field inventory is planned to compare some hot spots found in the study. Also,
locations for debris booms are to be studied.

From the findings so far it is possible to find risk areas for landslides and combining the
information with a tree count including classification of trees into tree length. The results
from the validation indicates that the analyze method has a high accuracy in finding places
with high risk of landslides. Of the 12 located larger landslides along Pite River 11 are
located in the areas were the GIS-analysis show high risk of landslide (4 or 5 (5)).

The conclusion is that a river with limited access to the shores can be analyzed only by
gathering digital information, processing it in GIS to gain knowledge of the river which is
impossible only by visual inspection.

REFERENCES (12PT BOLD AND ALL CAPS)

Elforsk, 2011. Drivgods vid extrema tillfällen. Nulägesbeskrivning. Rapport 11:32

Marklund, J., 2011. Drivgodsinventering, riskanalys och åtgärdsförslag för Granfors och
Krångfors vattenkraftverk i Skellefteälven. Examensarbete vid Luleå Tekniska Universitet.

V - 14
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Development Of The Sediment Removal Suction Pipe By

Laboratory And Field Experiments

M. Miyakawa & N. Hakoishi

Public Works Research Institute , Ibaraki , Japan
[email protected]

T. Sakurai
National Institute for Land and Infrastructure Management , Ibaraki , Japan

ABSTRACT:
The construction of a dam can interrupt the movement of sediment through the river. Decreased
sediment downstream causes environmental problems related to the riverbed such as degradation,
armoring, and fewer opportunities to renew the riverbed material. Furthermore, sedimentation
causes reduction in reservoir storage capacity. Therefore, measures are required for sediment
supply from the reservoir. Thinking of the conditions and time variation of the downstream
riverbed environment, it is desirable to be able to control the timing of sediment supply and the
quantity and quality (mainly particle size) of sediment.
The authors have been working to develop a new sediment supply measure to solve reservoir
sedimentation problems and downstream riverbed environmental problems. As a result of earlier
studies, we proposed a “burrowing type sediment removal suction pipe method”, using the water
head energy differential between the upstream and downstream areas of a dam. We have carried
out laboratory experiments and field tests to examine hydraulic characteristics and applicability of
the pipe. We compared the results of the laboratory experiments (with pipe diameter: 60 mm, 100
mm and 200 mm) and the results of the field tests (with pipe diameter: 200mm) at an actual small
reservoir located in a mountainous area. As a result, the hydraulic characteristics such as the
relationship between velocity in the pipe and the sediment concentration, water head energy loss in
the pipe , etc. were found. It was confirmed that this kind of pipe could be applied to remove non-
cohesive debris-less sediment material in a small reservoir.

Keywords: Reservoir sedimentation, Burrowing type sediment removal suction pipe, Hydraulic
model test, Sediment discharge test, Field test.

1. INTRODUCTION

The construction of a dam can interrupt the movement of sediment through the river.
Decreased sediment downstream causes environmental problems related to the riverbed
such as degradation, armoring, and fewer opportunities to renew the riverbed material.
Furthermore, sedimentation causes reduction in the reservoir storage capacity. Therefore,
measures are required for sediment supply from the reservoir. Thinking of the conditions

V - 15
and time variation of the downstream riverbed environment, it is desirable to be able to
control the timing of sediment supply and the quantity and quality (mainly particle size) of
sediment.

So far, besides traditional measures such as excavating and dredging, sediment flushing
with water level drawdown and sediment bypassing have been developed and used in
Japan. However, the conditions for applying these measures are restricted and it is difficult
to control the exact quantity and quality of the discharging sediment by these methods.
Then, the authors have been working to develop a new sediment supply measure. We set
following objectives of development. (1) A change of reservoir operation is not required.
(2) It is able to control a sediment discharge rate according to a water discharge rate. (3)
Size of facility is small and economical. As a result of earlier studies, we proposed the
“burrowing type sediment removal suction pipe method” using the differential water head
between the upstream and downstream areas of a dam (e.g., Sakurai and Hakoishi 2011,
2012).

In this study, we have carried out the laboratory experiments and the field tests to examine
the hydraulic characteristics and the applicability of the burrowing type sediment removal
suction pipe.

2. BURROWING TYPE SEDIMENT REMOVAL SUCTION PIPE

Fig.1 illustrates the representative shape of the burrowing type sediment removal suction
pipe that the authors had proposed. It is a U-bend flexible pipe that has a water intake at
the upstream end, an impermeable sheet, and sediment suction holes at the bent part and
the upstream part of the pipe. For further detail was described in Sakurai and Hakoishi
2011.

Bottom sediment suction hole

Discharge to downstream

Impermeable sheet Plan view

Water intake

Side sediment suction hole

Profile view

Figure 1. Outline of Burrowing Type Sediment Removal Suction Pipe.

A sediment discharge procedure of the burrowing type sediment removal suction pipe is
considered as follows. The pipe is initially set on the surface of deposited sediment. After
the start of discharge, the pipe is expected to suck up sediment through the sediment
suction holes at the bent part and gradually burrow into the sediment using the differential
water head.

V - 16
3. EXPERIMENTAL METHOD

In order to develop the suction pipe method, the authors conducted many experiments
(about 100 tests) using various types of the suction pipe. Table 1 shows the specification of
five representative pipes used for experiments. Fig.2 indicates the plan view of the five
pipe shapes.

The laboratory experiments were carried out using two experimental facilities. The field
experiments were implemented at the small reservoir located in the mountainous area
managed by Disaster Prevention Research Institute Kyoto University. The plan view and
vertical view of the experimental facilities are shown in Fig.3.

Table 1. Specification of the Pipes Used for Experiments.

Laboratory Laboratory Field
Diameter Length water sediment sediment
Pipe Pipe material *) Bent part structure Diameter of suction hole
(mm) (m) discharge discharge discharge
test test test
Upstream bottom: 3cm
PVC Polyvinyl chloride pipe
Pipe-1 60.5 Bent part bottom: 3cm 3.0 Examined Examined -
AP: 0.04MPa with sheet
Bent part side: 2cm
Upstream bottom: 4.5cm
PVC Same of the pipe
Pipe-2 100.0 Bent part bottom: 4.5cm 5.0 Examined Examined -
AP: 0.03MPa with sheet
Bent part side: 3.3cm
Upstream bottom: 9cm
PVC Same of the pipe
Pipe-3 200.0 Bent part bottom: 9cm 4.0 Examined - -
AP: 0.01MPa with sheet
Bent part side: 6.6cm
Fiber
Upstream bottom: 9cm
reinforcement Same of the pipe
Pipe-4 200.0 Bent part bottom: 9cm 4.5 - Examined -
PVC with sheet
Bent part side: 6.6cm
AP: 0.02MPa
PVC Steel Upstream bottom: 9cm
Pipe-5 200.0 5.0 Examined Examined Examined
AP: 0.15MPa without sheet Bent part bottom: 10cm
*) PVC: Polyvinyl chloride resin, AP: Allowable pressure

Pipe-1

Pipe-2

Pipe-3

Pipe-4

Pipe-5

Figure 2. Plan View of the Suction Pipe Shapes.

V - 17
The laboratory small water tank was 4.5 m long, 2.5 m wide and 1.3 m high. The tank has
a rectangular weir to maintain the water level. An outlet pipe is installed at the downstream
wall with a discharge control gate at the end of the pipe. The laboratory large water tank
was 7.5 m long, 7.5 m wide and 3.5 m high. The small dam reservoir was 14 m long, 6.55
m wide and 4.65 m high.

7.50m
14.00m
Plan view

4.50m

6.55m
7.50m
2.50m

Section view
1.30m

4.65m
3.50m

Dam

Laboratory small tank Laboratory large tank Small dam (Field test)

Figure 3. Experimental Facilities.

In order to understand the basic characteristics of the pressure loss of the pipes, we carried
out water discharge tests using the laboratory experimental facilities. The experimental
conditions of the water discharge tests are summarized in Table 2.

The sediment discharge tests were conducted using the all pipes at the laboratory tanks.
However, the pipe-3 was broken by its lack of pipe strength during sediment discharge test.
The field sediment discharge tests were carried out using the pipe-5. The experimental
conditions of sediment discharge tests are shown in Table 3.

Table 2. Experimental Conditions of Water Discharge Tests.

Diamiter Experimental Discharge rate Velocity of pipe flow
Case Pipe Reynolds number
(mm) facility (L/s) (m/s)
W-1 Pipe-1 60.5 Small tank 2.43 - 4.82 0.85 - 1.68 51,100 - 101,400
W-2 Pipe-2 100.0 Large tank 6.00 - 19.60 0.76 - 2.50 76,400 - 249,600
W-3 Pipe-3 200.0 Large tank 18.10 - 63.50 0.58 - 2.02 115,200 - 404,300
W-5 Pipe-5 200.0 Large tank 19.02 - 69.60 0.61 - 2.22 121,100 - 443,100

Table 3. Experimental Conditions of Sediment Discharge Tests.

Diamiter Experimental Discharge rate Velocity of pipe flow
Case Pipe Reynolds number
(mm) facility (L/s) (m/s)
S-1 Pipe-1 60.5 Small tank 0.79 - 4.79 0.27 - 1.67 16,600 - 100,800
S-2 Pipe-2 100.0 Large tank 1.86 - 15.44 0.24 - 1.97 23,700 - 196,600
S-4 Pipe-4 200.0 Large tank 20.90 - 40.90 0.67 - 1.30 133,100 - 260,400
S-5-1 Pipe-5 200.0 Large tank 52.40 1.67 333,589
S-5-2 Pipe-5 200.0 Small dam 113.40 3.61 721,954
S-5-3 Pipe-5 200.0 Small dam 116.60 3.71 742,299

As the experimental sediment material, we used sand of mixed particle size. The grain size
distributions and the photos of the sediment materials for the laboratory test and the field

V - 18
test are shown in Fig.4. The grain size distribution of the field test was larger than that of
the laboratory test. The authors think that the suction pipe is effective to discharge non-
cohesive sediment such as sand and gravel. But it is difficult to apply the suction pipe to
cohesive sediment, because cohesive sediment does not collapse easily in the water.

100
90
80
70
60
P (%)

50
40
30 Material of laboratory test
20 Laboratory test
10 Field test
0
0.01 0.1 1 10 100

d (mm)
d: Diameter (mm), P: Percent finer by weight (%).
Material of field test

Figure 4. the Grain Size Distribution and the Photos of Sediment Materials.

The experimental procedure of laboratory sediment discharge test was as follows: 1)

Sediment was placed at a height of 0.6 m in the small tank or a height of 2 m in the large
tank. 2) The burrowing type sediment removal suction pipe was set up on the sediment. 3)
Water was pumped into the tank at a constant discharge rate. The water level in the tank
was kept almost constant by overflowing from the weir. 4) Sediment discharge was started
by opening the gate at the end of the pipe. 5) We observed the sediment discharge situation
and measured the water level in the tank, pressure head in the pipe using piezometers,
discharge rate and sediment discharge rate. And the experimental procedure of field
sediment discharge test was almost same as the laboratory test. But inflow water was
supplied by the upstream small mountain stream. Because of small inflow discharge rate of
the stream (about 10 L/s), we conducted the test by repeating sediment discharge and
keeping water operation. In the cases of S-1 and S-2, we changed the gate opening during
the test to examine the different discharge rate conditions.

4. EXPERIMENTAL RESULTS

4.1. Results of water discharge tests

In order to explain how to calculate a pressure loss coefficient from piezometric head data,
one example of the piezometric head profile is shown in Fig.5. The piezometric head
decreased with increasing distance along the pipe. The pressure loss coefficient (similar
equation as friction loss coefficient) of the pipe flow was calculated using Eq. 1.

hL D 2 g
f  (1)
L V2

V - 19
f: pressure loss coefficient, hL: pressure head loss (m), D: pipe diameter (m), g: gravity
acceleration (m/s2), L: pipe length (m), and V: cross sectional average pipe flow velocity
(m/s). The pressure head loss and the pipe length of the upstream pipe were obtained from
measurement data of piezometers installed the upstream pipe. Those of the downstream
pipe were obtained from piezometers installed the downstream pipe as shown in Fig.5.

The pressure loss coefficient of the bent part was calculated using Eq. 2.

2g
f B  hLB (2)
V2

fB: pressure loss coefficient of the bent part, hLB: pressure head loss of the bent part (m).
The pressure head loss of the bent part was obtained from the difference between just
upstream and downstream of the bent part piezometric heads as shown in Fig.5. In the
cases using Pipe-5, the velocity was calculated using rectangular cross section shape of the
bent part (with 200mm height and 200mm width).

⑤
⑥ ⑦ ⑧

③ ② ①
④
0
① ② ③ ④
-20

-40
Calculated bent part
-60 pressure loss
Calculated upstream pipe pressure loss
Hp (cm）

-80

-100
⑤
-120

-140
⑥
⑦ ⑧
-160
Calculated downstream pipe pressure loss
-180
0 100 200 300 400 500 600 700 800
L (cm)

L: Distance from the first piezo position (cm), Hp: Piezometric head (cm).

Figure 5. Example of Piezometric Head Profile.

Fig.6 shows the relationship between Reynolds number and pressure loss coefficient of the
results of the water discharge tests. There is a low correlation between Reynolds number
and pressure loss coefficient as a whole. The reason for a great deal of scatter of the data,
especially Case W-1, W-2 and W-3 is considered that there is roughness at the inner
surface of the pipe due to the weakness of the pipe. As for pipe flow, pressure loss
coefficients of upstream pipe are larger than those of downstream pipe except Case W-5.
Pressure loss coefficients of bent part of Case W-5 are smaller than other cases.

V - 20
 W-1, Upstream W-1, Downstream W-1
0.16 5.0
W-2, Upstream W-2, Downstream W-2
0.14 4.5 W-3
W-3, Upstream W-3, Downstream
4.0 W-5
0.12 W-5, Upstream W-5, Downstream
3.5
0.10 3.0

fB
0.08 2.5
f

0.06 2.0
1.5
0.04
1.0
0.02
0.5
0.00 0.0
0 100,000 200,000 300,000 400,000 500,000 0 100,000 200,000 300,000 400,000 500,000
Re Re
Re: Reynolds number, f: Pressure loss coefficient (Pipe), fB: Pressure loss coefficient (Bent part).

Figure 6. Relationship between Reynolds Number and Pressure Loss Coefficient of Water
Discharge Tests.

4.2. Results of sediment discharge tests

The time series of discharge rate and sediment concentration of the sediment discharge
tests are shown in Fig.7. Fig.8 shows the relationship between velocity and sediment
concentration.

The sediment concentration was obtained by analysis of sampled water. The sediment
concentration is a volume concentration and is estimated by “sediment volume / (water
volume + sediment volume)”. In addition, sediment volume doesn’t include void volume.

In the Case S-1 and S-2, as the discharge rate was increased, the sediment concentration
became larger. However, the rate of increase in sediment concentration in the second
discharge rate increase was smaller than that in the first change. The sediment
concentration of the Case S-5-1, S-5-2 and S-5-3 are smaller than the other cases.
120 S-1
100 S-2
80 S-4
Q (L/s)

S-5-1
60
S-5-2
40
S-5-3
20
0
0 50 100 150 200 250 300 350 400 450 500 550 600
T (min)

18 S-1
16 S-2
14 S-4
12 S-5-1
C (%)

10 S-5-2
8 S-5-3
6
4
2
0
0 50 100 150 200 250 300 350 400 450 500 550 600
T (min)
T:
Time (min), Q: Discharge rate (L/s), C: Sediment concentration (%).

Figure 7. Time Series of Discharge Rate and Sediment Concentration.

V - 21
 18
S-1
16
S-2
14
S-4
12
S-5

C (%)
10
8
6
4
2
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

V (m/s)
V: Velocity (m/s), C: Sediment concentration (%).

Figure 8. Relationship between Velocity and Sediment Concentration.

In Fig.8, as the velocity increases, the sediment concentration increases for Case S-1 and
S-4. It is considered that we can control the sediment concentration to some extent by
adjusting the velocity.

Experimental Results of sediment discharge tests were summarized in Table 4. A removed

sediment volume with void was obtained by a surveying sedimentation surface shape
before and after sediment discharge. For example, sedimentation situation of Case S-5-3 is
shown in Fig.9.

Table 4. Experimental Results of Sediment Discharge Tests.

Average
Removed Removed
Discharge Total discharge sediment
Diamiter Discharge rate Velocity sediment volume sediment volume
Case Pipe time water volume concentration
(m) (L/s) (m/s) 3
without void with void
(min) (m ) 3 3 without void
(m ) (m )
(%)
S-1 Pipe-1 0.0605 0.79 - 4.79 0.27 - 1.67 130 24.4 0.72 1.20 2.95
S-2 Pipe-2 0.1000 1.86 - 15.44 0.24 - 1.97 600 387.6 18.70 31.16 4.82
S-4 Pipe-4 0.2000 20.90 - 40.90 0.67 - 1.30 110 246.8 18.75 31.25 7.60
S-5-1 Pipe-5 0.2000 52.40 1.67 225 707.4 11.58 19.30 1.64
S-5-2 Pipe-5 0.2000 113.40 3.61 57 387.8 2.07 3.45 0.53
S-5-3 Pipe-5 0.2000 116.60 3.71 78 543.0 10.44 17.40 1.92

Before sediment discharge After sediment discharge

Figure 9. Sedimentation Situation of Field Sediment Discharge Test (Case S-5-3).

V - 22
At present, we set the target sediment concentration to from 2 to 5 % considering actual
situation of Japanese reservoir. The average sediment concentrations without void in Table
4 are larger than or equal to about 2 % except Case S-5-2.

Because of the simple setting method that the pipe has been placed on the sediment surface,
the actual setting and removing the pipe in the field tests were easy.

Fig.10 shows the relationship between Reynolds number and pressure loss coefficient of
the results of the sediment discharge tests. Fig.11 shows the relationship between sediment
concentration and pressure loss coefficient. Pressure loss coefficient data in Fig.10 and 11
were calculated by the same method as the water discharge tests.
S-1, Upstream
2.0 16
S-1, Downstream
1.8 S-1
S-2, Upstream 14
S-2
1.6 S-2, Downstream
S-4, Upstream 12 S-4
1.4
S-4, Downstream S-5
1.2 10
S-5, Upstream
1.0
f

S-5, Downstream
fB
8
0.8
6
0.6
0.4 4
0.2
2
0.0
0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 0
0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000

Re Re
Re: Reynolds number, f: Pressure loss coefficient (Pipe), fB: Pressure loss coefficient (Bent part).

Figure 10. Relationship between Reynolds Number and Pressure Loss Coefficient of Sediment
Discharge Tests.

2.0 16
S-1
1.8 14
S-2
1.6 S-1, Upstream
12 S-4
1.4 S-1, Downstream S-5
S-2, Upstream 10
1.2
S-2, Downstream
fB

1.0 8
f

S-4, Upstream
0.8 S-4, Downstream 6
0.6 S-5, Upstream
S-5, Downstream 4
0.4
0.2 2

0.0 0
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
C (%) C (%)
C: Sediment concentration (%), f, fB: Same as Figure 10.

Figure 11. Relationship between Sediment Concentration and Pressure Loss Coefficient.

As a whole, the pressure loss coefficients of the sediment discharge tests are larger than
those of the water discharge tests. Both of the pipe flow and bent part pressure loss
coefficients of Case S-5 are smaller than those of the other sediment discharge test cases.
These quantitative data of pressure loss coefficient would be useful for design of the actual
suction pipe facility.

V - 23
4. CONCLUSION

(1) We proposed the “burrowing type sediment removal suction pipe method” for sediment
supply from reservoirs and carried out laboratory and field tests. As a result, it was
confirmed that three diameter suction pipes (with pipe diameter 60.5mm, 100mm, 200mm)
can discharge sediment at almost the expected performance for sand material.

(2) As a result of experiments, we obtained quantitative data of the pressure loss for
various pipe diameters and bent part shapes. Those are useful to design of the actual
suction pipe facility.

(3) In order to achieve practical use of the “burrowing type sediment removal suction pipe
method”, we must conduct larger scale pipe experiments. It is also necessary to solve
problems such as blocking by debris or driftwood and discharging cohesive sediment.

ACKNOWLEDGEMENT
The authors are deeply grateful to Professor Masaharu Fujita, Dr. Daizo Tsutsumi, Dr. Shusuke
Miyata of Disaster Prevention Research Institute Kyoto University, and Mr. Shuuich Maeda of
Electric Power Development Co.,Ltd for their cooperation with our field experiment research.

REFERENCES

Sakurai, T. & Hakoishi, N. (2011): Burrowing-type sediment removal suction pipe for a
sediment supply from reservoirs, Dams and Reservoirs under changing Challenges
"79th Annual Meeting of ICOLD 2011", pp.371-378, International Commission on
Large Dams , Lucerne,Switzerland,

Sakurai, T. & Hakoishi, N. (2012): Present conditions and future prospects of sediment
supply from reservoirs to downstream rivers, Twenty-fourth International Congress
on Large Dams, Question 92, R.34:pp.496-517, International Commission on Large
Dams , Kyoto, Japan.

Sakurai, T. & Hakoishi, N. (2012): Hydraulic Characteristics of the Burrowing Type

Sediment Removal Suction Pipe, International Symposium on Dams for A Changing
World, No.281, International Commission on Large Dams, Kyoto, Japan.

V - 24
 INTERNATIONAL SYMPOSIUM ON
DAMs IN A GLOBAL ENVIRONMENTAL CHALLENGES
Bali, Indonesia, June 1ST – 6TH , 2014

Analysis on variation of soil erosion and sediment yield in the Three Gorges
Reservoir reach
Paper Title Line 2 (14pt)
Xu Tao
Three Gorges Hydropower Complex Cascade Dispatching Communication Center, Yichang, China
[email protected]
2(14pt)
Zhu Jun
Three Gorges Hydropower Complex Cascade Dispatching Communication Center, Yichang, China

Zhou Man
China Three Gorges Corporation, Yichang,, China

ABSTRACT:
The Three Gorges Reservoir reach is the potential area for water and soil loss with huge
topographical amplitude and heavy precipitation, so sediment generated by the Three Gorges
Reservoir reach may take a significant proportion of the inbound sediment of the Three Gorges
Reservoir. However, this part of sediment is always ignored because of its difficulty to quantify
since the preliminary design. In this article, as an important basis for quantitative estimation of
sediment from the Three Gorges Reservoir reach, the amount of soil erosion is calculated in three
typical years (2000, 2007 and 2010) by GIS spatial analysis technology and RUSLE model based
on statistical method. According to the model calculation, the average soil erosion modulus is
2474.26t/(km2.a), 2126.02t/(km2.a), 1844.24t/(km2.a), so the amount of soil erosion in three typical
years can be estimated as 145 million tons, 124 million tons and 108 million tons, respectively.

Keywords: Sediment; soil erosion; RUSLE model; the Three Gorges Reservoir Reach

1. INSTRUCTIONS
The sediment in the Three Gorges Reservoir reach mainly originates from slope-face water
and soil loss. As a result of huge topographical amplitude and heavy precipitation, the area
is prone to water and soil loss. With the construction of the Three Gorges Reservoir, the
issue of water and soil loss in the Three Gorges Reservoir reach has again become the
focus of concern. As the basic unit of sediment resulting from soil erosion, the slope face
controls the process of sediment segregation, transport and settlement. Sediment resulting
from slope-face soil erosion interferes with normal operations of the hydroelectric facilities
after entering the reservoir area through the waterway. Thus, it’s crucial to conduct in-
depth research on the temporal and spatial variation of water and soil loss in the Three
Gorges Reservoir reach, so that effective measures can be adopted to reduce the amount of
soil erosion in the reach, thereby decreasing or slowing down sediment deposition in the
reservoir, as this will have vital significance for extending the service life of the reservoir.
Judging from the current state of the research on soil erosion in the area, there are
primarily two methods for calculating the amount of soil erosion in the area, depending on

V - 25
the application of different measures and technical characteristics. One of the methods is
quantitative assessment of soil erosion based on remote sensing. This particular method is
the primary one used in general surveys of soil erosion in China. Based on the standards
issued by Ministry of Water Resources, this method involves the extraction of information
on three factors - type of land utilization, extent of vegetation cover and terrain slope –
from remote sensing images and some basic geographic information maps and drawings;
the intensity of soil erosion at each pattern spot is then determined on the basis of
experience, and spatial distribution maps showing the intensity of soil erosion are
generated to obtain straightforward data on the degree of soil erosion in different areas.
The other method is based on a common international calculation model for soil erosion
calculation (such as USLE and RUSLE based on statistical methods, WEPP and SWAT
based on distributed thought and other models). As the quantitative description of some of
the most basic physical processes, such as erosion, deposition and transport, in the current
soil erosion research remains inadequate, the distributed erosion-generated sediment model
is mainly used in calculations for small or medium-sized drainage basins, and is not well-
suited to large-sized basins or areas. As such, this paper uses the RUSLE model based on
statistical methods to calculate the amount of soil erosion in the Three Gorges Reservoir
area.

2. RESEARCH AREA AND METHODS

2.1. Research Area
The area under research is the Three Gorges Reservoir reach between Chongqing and
Yichang, with the mainstream waterway being 650km from east to west and less than
100km from south to north, covering nearly 60,000km2. The area is part of a mountainous
drainage basin with a high level of topographic complexity. The tributaries on both banks
have a short flow path, steep slopes, and rapid confluence. Research suggests that high-
intensity rainstorms are the main culprit of the area’s soil erosion.

2.2. Research Methods

The RUSLE model (the Revised Universal Soil Loss Equation model) was officially
adopted in 1997 by the Natural Resources Conservation Service (NRCS) of the U.S.
Department of Agriculture. It has been promoted and implemented worldwide [1-4].
The basic form of the common soil erosion equation[1-4]:
A  R  K  LS  C  P (1)

Where A is the average amount of soil loss (t·hm-2· a-1) over time and space in the unit area,
mainly referring to the amount of soil erosion in the rills and between the rills on the slope
created by rainfall and its runoff;
R is the rainfall erosivity factor (MJ· hm-2·mm·h-1·
a-1), represented by the result of the
kinetic energy of rainfall multiplied by the maximum rain intensity of 30 minutes;
hm-2·
K is the soil erodibility factor (t· MJ-1· hm2·mm-1· h), representing the amount of soil
erosion on the unit area created by the unit rainfall erosivity under the standard small area;

V - 26
The remaining four variables are dimensionless factors, representing the impact of slope
length, slope gradient, biological measures, and engineering measures and tillage practices
on soil erosion;

Slope length factor L is the ratio of the amount of slope-face soil erosion on a slope length
(projection) to the amount of soil erosion on a 22.13m slope length (projection), other
conditions being equal;
Slope gradient factor S is the ratio of the amount of soil erosion on a slope face with a
particular gradient to the amount of soil erosion on a slope face with a gradient of 9%
(5.14°), other conditions being equal;
Cover-management factor C is the ratio of the amount of soil erosion on a slope of a
particular length to the amount of soil erosion in the area under the plough-continuous
idleness at a certain level of cover and management;
P is the water conservation measure factor, representing the ratio of the amount of soil
erosion with water and soil conservation measures in place to the amount of soil erosion
caused by direct tillage along the slope. [1-4]

3. DATA SOURCE
3.1. Rainfall
The rainfall data selected are the data on the annual rainfall from 1991 to 2010 collected by
23 meteorological stations in the Three Gorges Reservoir reach and its vicinity. Spatial
distribution maps for the annual rainfall in 2000, 2007 and 2010 were generated using the
Gauss-Krieger interpolation method.

3.2. Soil
The soil maps used are the 1:1,000,000 digital soil maps compiled by the Nanjing Soil
Research Institute of the Chinese Academy of Sciences, and the soil data are sourced from
the soil species records of Hubei Province and Chongqing City. Data on mechanical
composition and organic matters are extracted from the soil species records on the basis of
the calculation methods for soil erodibility. Data on soil structural factors and soil
penetration levels are determined according to the soil structures described in the soil
species records.

3.3. DEM
DEM data required for the calculations are sourced from the United State Geological
Survey (USGS). The DEM used is an approximately 50m-grid DEM (digital elevation
model).

3.4. Land Utilization

V - 27
 Figure 1. Basic data in the Three Gorges reach

The land utilization map is based on data from 2000 and has a scale of 1:100,000. To meet
calculation requirements, the types of land utilization are classified into dry land, paddy
field, woodland, grassland, water body, urban and rural, mining, residential land and
unutilized land.

4. RESULTS AND ANALYSIS

4.1. Rainfall Erosivity (R) Factor

V - 28
 Figure 2. Distribution of rainfall erosivity in the Three Gorges Reservoir area in 2000, 2007 and
2010

The distribution of rainfall erosivity in the Three Gorges Reservoir area is generated by
means of spatial interpolation according to the calculations of the annual rainfall erosivity
at 23 local meteorological stations in 2000, 2007 and 2010. Figure 2 shows that the spatial
distribution of rainfall erosivity over the three years is significantly different, exhibiting
roughly two patterns of change: In 2000, in the Three Gorges Reservoir area, the annual
rainfall erosivity was small on both the eastern and western ends, and was significant in the
central area. Rainfall erosivity exceeded 8,000MJ·mm/(ha· a) in the central area and
declined progressively towards the east and west, falling in the range of 4,000~5,000
MJ·mm/(ha· a). In 2007, the annual rainfall erosivity followed an overall pattern of
progressive decline from the west and south to the east, dropping from 8,000~9,000
MJ·mm/(ha· a) to 5,000~6,000 MJ· mm/(ha· a). In 2010, the annual rainfall erosivity
followed an overall pattern of progressive decline from the west to the east. Except in some
areas in the southeast, south and north where the annual rainfall erosivity exceeded
6,000MJ·mm/(ha· a), in all other areas, the annual rainfall erosivity increased progressively
from both the eastern and western ends to 6,000MJ·mm/(ha· a) and decreased gradually
from the south to north.

3.2. Soil Erodibility (K) Factor

The soil erodibility factor formula and the established soil indicator database are used to
calculate the value of soil erodibility factor for each soil genus; the results are then
combined with the soil map to obtain a spatial distribution map of the soil erodibility factor.
According to the calculations, the K-factor value in the Three Gorges Reservoir area
averages 0.06 and generally falls within the range of 0.04~0.08. A K factor value of
0.06~0.08 has the most extensive distribution, covering a total area of approximately
34,349.82 km2, accounting for 58.39% of the total land area. A K factor value of more than
0.08 is mostly found in the eastern part of the Three Gorges Reservoir area, in Yingchang
and Zigui. A K factor value of less than 0.02 is largely found in the northern area.

3.3. Slope Gradient and Length (LS) Factor

V - 29
 Figure 3. Distribution of soil erodibility factors K value in the Three Gorges Reservoir area

Figure 4. Distribution of terrain factor LS value in the Three Gorges Reservoir area

A spatial distribution map of the value of the L· S factor is obtained using the ARCGIS
spatial analysis functions on the basis of the existing DEM (digital elevation model), as
shown in Figure 4. According to the calculations, the LS factor value in the Three Gorges
Reservoir area averages 1.22 and generally falls within the range of 0-3. In the western
parallel ridge and gorge areas and the eastern head area, the LS factor typically ranges
from 0 to 1, while in the central part of the reservoir area, it generally ranges from 1 to 3.

3.4. Soil Conservation (CP) Factor

V - 30
 Figure 5. Distribution of vegetation and crop management factor in 2000, 2007 and 2010

The land cover-management factor and the soil conservation measure factor in different
years are calculated according to the vegetation cover and soil conservation measures in
the Three Gorges Reservoir area in different years. The CP factor value maps for 2000,
2007 and 2010 are drawn, as shown in Figure 5. The CP factor value followed a pattern of
declining noticeably from the east to the west in the said three years, particularly in 2010.
Statistics show that in these three years, the CP factor value averaged 0.07, 0.05, and 0.06,
with the highest in 2000, the lowest in 2007 and the medium level in 2010.

3.5. Soil Erosion Amount

Figure 6. Distribution of soil erosion in the Three Gorges Reservoir area in 2000, 2007 and 2010

With the spatial distribution map of each factor, all of the above factors are superimposed
in the ARCGIS for calculations, thus obtaining a distribution map of the water and soil loss
in the Three Gorges Reservoir area. Through a statistical analysis, the average value of the
soil erosion modulus is derived, which stands, in descending order, at 2,474.26t/(km2· a),

V - 31
2,126.02t/(km2·a), and 1,844.24t/(km2· a) in 2000, 2007 and 2010, respectively. Thus, the
amount of soil erosion in each of the three years can be estimated as 1.45×10 8t, 1.24×108t,
and 1.08×108t respectively.

Table 1. Statistics on the erosion intensity area in different grades in 2000, 2007 and 2010
Grade of soil 2000 2007 2010
erosion Percentage Percentage Percentage
intensity Area (km2) Area (km2) Area (km2)
(%) (%) (%)
Light 6127 10.49 7034 12.04 10052 17.21
Medium 15018.4 25.72 11406.31 19.53 8850.57 15.16
Strong 8104.73 13.88 9041.93 15.48 4401.94 7.54
Very strong 3011.3 5.16 2831.97 4.85 2552.64 4.37
Severe 1025 1.76 757 1.30 1421 2.43
Total 33286.43 57.00 31071.21 53.20 27278.15 46.71

The analytical data (see Table 1 and Figure 6) indicate that from 2000 to 2010, the area of
water and soil loss dropped from 33,286.43km2 to 27,278.15 km2, and its share of the total
reservoir land area fell from 57% to 46.71%, but the area of slight erosion exhibited an
upward trend, rising from 6,127km2 to 10,052 km2; the area of medium erosion and the
area of strong erosion shown a downward trend, with the area of medium erosion dropping
from 15,018.4 km2 to 8,850.57 km2 and the area of strong erosion decreasing from
8,104.73 km2 to 4,401.94 km2; the area of very strong erosion largely remained unchanged;
and the area of severe erosion expanded moderately from 1,025 km2 to 1,421 km2.

6. ANALYSIS OF RATIONALITY OF MODEL CALCULATION RESULTS

Table 2 . Comparison between RESLE model calculation value and bulletin value
Area of Area of different grades of erosion intensity (km2)
water and
Data source Very
Year soil loss Light Medium Severe
2 Strong strong
(km )
Bulletin data 29559 5185 13590 7085 2812 887
Model
2000 33286.43 6127 15018.4 8104.73 3011.3 1025
calculations
Error (%) 12.61 18.17 10.51 14.39 7.09 15.56
Bulletin data 28246.61 6341.57 10133.9 8039.87 3039.68 691.59
Model
2007 31071.21 7034 11406.31 9041.93 2831.97 757
calculations
Error (%) 10.00 10.92 12.56 12.46 -6.83 9.46
Bulletin data 23413.18 8286.55 7300.4 3703.6 2831.28 1291.35
Model
2010 27278.14 10052 8850.57 4401.94 2552.64 1421
calculations
Error (%) 16.51 21.31 21.23 18.86 -9.84 10.04

To test the consistency between the data on soil loss in the Three Gorges Reservoir area
which have been calculated in this study on the basis of the RUSLE model and the data on
soil loss which have been published in the Water and Soil Conservation Bulletin of the

V - 32
above four issues, the data on soil loss in 2000, 2007 and 2010 are compared against the
data in the Bulletin (see Table 2).
Data in Table 2 show that the amount of soil loss calculated in this study using the RUSLE
model is largely consistent with the data in the Water and Soil Conservation Bulletin. In
particular, the data on the area of water and soil loss and on erosion of each grade of
intensity in 2000 and 2007 are significantly consistent with the data in the Bulletin, with a
difference of less than 20%, but there is a considerable difference for data in 2010, with the
difference in the area of light and medium erosion exceeding 20%; this is possibly because
the land utilization map used in this research is based on data from 2000, and over the
decade, interference from human activities has resulted in a significant change in land
utilization conditions.

7. CONCLUSION AND OUTLOOK

The amount of soil loss in the Three Gorges Reservoir area in 2000, 2007 and 2010 is
calculated using the modified Universal Soil Loss Equation (RUSLE). The results show
that from 2000 to 2010, the area of water and soil loss dropped from 33,286.43km2 to
27,278.15 km2, and its share of the total reservoir land area fell from 57% to 46.71%, but
the area of slight erosion exhibited an upward trend, rising from 6,127km2 to 10,052 km2 ;
the area of medium erosion and the area of strong erosion shown a downward trend, with
the area of medium erosion dropping from 15,018.4 km2 to 8,850.57 km2 and the area of
strong erosion decreasing from 8,104.73 km2 to 4,401.94 km2; the area of very strong
erosion largely remained unchanged; and the area of severe erosion expanded moderately
from 1,025 km2 to 1,421 km2.

Using the above data, the average soil erosion modulus of the Three Gorges Reservoir area
is calculated and descends in the sequence of 2000, 2007 and 2010, standing at
2,474.26t/(km2· a), 2,126.02t/(km2· a), and 1,844.24t/(km2·
a), respectively. Thus, the amount
of soil erosion in the Three Gorges Reservoir area in each of the three years is estimated at
1.45×108t, 1.24×108t, and 1.08×108t. The results of the above model calculation are
compared against the data published in the nation’s Water and Soil Conservation Bulletin;
it shows that the model calculation results are largely consistent with the data in the
Bulletin, indicating that the model is viable for use in the estimation of the amount of soil
loss in the reservoir area in the year of the Bulletin.
Research on land utilization and soil erosion has been a hot frontier subject in recent years.
Because of influence of the diversity of land utilization variables, the complexity of soil
erosion mechanisms, and scale variation, many subjects in the research of land utilization
changes and soil erosion call for further exploration and increased attention. For example,
the data on the types of land utilization used in this paper are exclusively from around 2000;
this causes some deviation in the calculations of the amount of soil erosion in 2010, thus
affecting the results of the calculations of soil erosion to a certain degree.

REFERENCES

V - 33
CHEN,Y M. LIU, G. B. and ZHENG, F. L. (2004): Proceeding and Application on Soil
Erosion Model of RUSLE, Research of Soil and Water Conservation, 11:4, pp.80-83.
Shanxi, China
Yoder Daniel, (1995): The future of RUSLE: Inside new Revised Universal Soil Loess
Equation, J. Soil and Water Conservation, 50:5 pp484-489.USA
Kautza T J , Schertz D L ,(1995): Lessons learned in RUSLE technology transfer and
implementation, J. Soil and Water Conservation,50:5 pp490-492.USA
ZHANG X.Q, CHEN, Y. Q. and CHEN, H.K (2013): Evaluation on Soil Erosion of
Huanggang River Basin in Eastern Guangdong Based on GIS and RUSLE, Soil and
Water Conservation in China,2:371 pp 34-37. Zhengzhou, China

V - 34
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Decision Support System for water resources planning

in Karun river basin

A. Heidari & E. Bozorgzadeh

Iran water and power resources development Co (IWPC), Tehran, Iran
[email protected]

ABSTRACT:
A Decision Support System (DSS) was developed for planning of water resources management
because of dynamic nature of the basins in terms of hydrological changes and human
manipulations. The developed DSS is a tailor made system that uses many software such as
ArcGIS, HEC series, MODSIM, ARSP, and TOPSIS along with a series of developed modules for
modeling of conjunctive use, water quality, flood inundation maps, flood damage estimation,
optimization, and economical evaluation. In addition to simulation and optimization models, a
multi criteria decision making module was developed to screen the development scenarios based
on defined indexes and stakeholders interests. The system was applied for the Dez and the Karun
river basins where accommodate more than 20 large dams, 16 Hydro Power Plants (HPPs), and
20 irrigation systems at different stages of operation, construction and investigation. The DSS is
now in the final stage of development and, some of the results summarized in this paper, have been
used to screen planning scenarios for hydropower schemes such as runoff river and storage HPPs.
The DSS helped the decision makers to select the runoff river scheme of the Karun2 HPP based on
different criteria such as economic indexes, drought mitigation, social impacts, water quality and
environmental impacts, and technical and construction aspects. Moreover, the result of
optimization module was used to determine optimum features of dams, hydropower plants, and
water conveyance systems. The results showed that active volume of under investigation reservoirs
should basically increase at integrated planning in comparison with the results of individual
studies. The main feature of hydro projects was finalized in the Karun hydro-system in terms of
inflow, water withdrawal in upstream catchments, and design parameters of HPPs after debating
with the stakeholders and decision makers.

Keywords: DSS, Hydropower, Karun, Iran

1. INSTRUCTION

Water resources planning and management is usually carried out by mathematical models
and an information-technology framework is needed to link the models in convenient
platform to make the results tangible. The diverse nature of water resources modeling, file
formats and data structures along with enormous number of required data, make the
process a complex task which can only be accomplished through new technological
advances. The systemization of computer-aided collaborative processes in this area has
been achieved by many researchers such as Thiessen and Loucks (1992) and
Koutsoyiannis el al. (2003). Thiessen and Loucks (1992) have reviewed the application of

V - 35
models in water resources planning, discussed the challenges and presented the criteria for
evaluating success. They restructured the dynamics and procedures of the negotiation
process in order to make salient the possible gains from integrative bargaining.
Koutsoyiannis el al. (2003) presented a DSS for water resource system in Athens including
information systems that perform data acquisition and models that perform simulation and
optimization for operation of the existing hydro-system. The mathematical framework of
their methodology performed the allocation of the water resources to the different system
components, keeping the number of control variables small.

A solution for reaching the required integration is to use a spatial data analysis framework
as a common background for connecting the models. Accordingly, a geographical
integrated modeling system has been proposed and implemented for decision making in
many case studies. For instance, Mysiak et al. (2000) introduced MULINO-DSS to meet
the needs of European water management authorities through the integration of socio-
economic and environmental modeling techniques with Geographical Information System
(GIS) capabilities and MCDM aids. Zhang el al. (2003) developed a framework of GIS-
based DSS for water resources management at the Flint river basin. The framework
included surface and ground water modeling system, database, GIS-based and user-
friendly graphical interface, analysis tools, and alternative analysis and evaluation
modules. Pallottino et al. (2005) presented a scenario-based DSS to perform water system
planning and management under climatic and hydrological uncertainty. They used
independent hydrological scenarios and exploited the inner structure of temporal evolution
to obtain a decision policy statistically, so that the risk of wrong decisions is minimized.
Enormous numbers of spatial data are involved in decision making process and decision
support systems need to embed GIS tools to increase analytical capabilities. On the other
hand, computer models provide a very rapid and standardized means of evaluating spatial
data and multiple criteria likely to concern conflicting stakeholders interests and
considering social and environmental impacts. The need for robust DSS has clearly
emerged as a result of the increasing complexity of the decision process caused by
conflicts, diversity of stakeholders, and spatial variation of parameters.

The common challenge of the water resources management is how to evaluate and
compare the different scenarios regardless of purpose of management. In this research, a
scenario-based DSS was developed in order to address the stakeholders' interests and meet
the minimum requirements of the hydro-systems in decision making process. The main
purpose of the DSS was to plan and review the design feature of the projects based on
overall net benefits and taking into account the environmental and social impacts.
Additionally, the DSS is able to screen the proposed scenarios based on predefined criteria
and evaluate different schemes as well as to determine optimum features of the projects.
The system was applied and validated for the Dez, Karun (Big Karun) river basins where
accommodate many large dams, Hydro Power Plants (HPPs), and irrigation systems at
different stages of operation, construction, and investigation. The DSS includes many
modules to simulate hydro-systems in order to quantify the tangible and intangible indexes.
Finally, MCDM evaluate the scenarios based on the different indexes and depicts the
results for per-defined areas such as prefectures or for individual projects, separately.

2. THE DSS FEATURES

V - 36
A sort of components was needed to achieve an operational end-to-end integrated and
dynamic modeling system for water resources planning and management under GIS
framework. Integration of analytical modules and GIS in the DSS has made it a useful tool
with many capabilities for water resources management.

The DSS has been developed by utilizing many software such as ArcGIS, HEC series,
ARSP, and TOPSIS along with a series of modules for simulating conjunctive use and
water quality as well as optimization and evaluation of the planning scenarios. The
capabilities of the free domain models have been enhanced in order to customize as much
as possible with the requirements of the hydro-systems. In addition, the DSS has capability
to add new elements of the river system such as dams and HPPs and to generate the
required data for modeling along with new schematic of the hydro-system.

2.1. River basin simulation module

River basin was simulated by upgrading Acres Reservoir Simulation Program (ASRP) for
the Karun basins. The model is stationary so that operation policies, river system network,
and requirements do not change for time steps of simulation. The module estimates
peak/off peak and firm/non-firm annual power generation of HPPs based on design criteria
in addition to predefined capabilities of the original models.
The ARSP model determines optimum policy for water allocation at each single time
period by out of kilter algorithm (OKA) in a penalty structure framework. C++
programming language was used to upgrade the model's capabilities and resolve
computational restrictions such as curse of dimensionally and number of dams. In the new
version, stream network consists of a set of nodes (i) which connected by arcs (j). The flow
of each arc (Xij) is confined to upper and lower limits and a cost (Cij) is assigned for each
arc and then following optimization scheme is carried out by OKA:

Minimize Z= Σ for i, j (Cij*Xij) (1)

Subject to:

Lij<Xij<Uij ; upper and lower limits defined by user (2)

Xji-Xij=0; mass balance of flow for the nodes at each time step (3)

Physical components of river system such as channels, confluences, and reservoirs were
defined by flow/demand channels and junction nodes with and without reservoirs. Flow
channels were categorized to uncontrolled flow such as spill, water loss and controlled
flow. Similarly, demand channels included power demand, water abstraction demand, and
minimum in stream flow requirements. Power channels were used to define install
capacity, power generation, minimum power requirement (firm energy), and minimum
required release for the HPPs. In addition to nodes and arcs, other inputs were also defined
for the module, e.g., zonal penalty structure, reservoirs priority for evacuation, operating
levels, dead storage, and rule curves. Conservation volume of each reservoir cab be divided
to maximum 15 zones depends on reservoirs purposes and a penalty can be assigned to

V - 37
each zone by user. Figure 1 presents main menu of the simulation module for the Karun
basin.

Figure 1. The Karun river system simulation module

2.2. Optimization modules

Particle Swam Optimization (PSO) and algorithm was utilized to develop optimization
module. PSO algorithm is a population-based stochastic optimization technique which
developed by Kennedy and Eberhart (1995). The algorithm is inspired by social behavior
of bird flocking or fish schooling and shares many similarities with evolutionary
computation techniques such as Genetic Algorithms. PSO is initialized with a population
of random generation and searches for optimal global solution.
Total net benefit (NB) value was considered as objective function of optimization
algorithms taking into account the design criteria and restrictions of the river basin as
follows:

Maximize F= (B-C)/Ps (4)

Where:

B: annual benefit; C: total annual cost; Ps: penalty value for violating minimum
requirements of the river system.
Ps will be set to 1, if all the requirements of the system are satisfied, otherwise it will be set
to higher than 1. The penalty of the minimum requirement was divided to four parts
including 1) power network (Ppow), 2) irrigation (Pir), 3) domestic and industrial (Pdd),
and last but not least environmental (Penv). Therefore Ps was defined by following eq.:

Ps=Ppow*Pir*Pdd*Penv (5)

Ppow was determined by reliably and deficit for firm energy of HPPs as follows:

V - 38
 (6)

Where:

Ψpow: coefficients of penalty pertaining to permissible reliability of firm energy

μpow : coefficients of penalty pertaining to permissible deficit of firm energy

Reliability and deficit coefficients were used in the same way to determine Pir, Pdd and
Penv, however, the permissible coefficients were different depend on design criteria for
each water user sector. In the validation of the DSS for the big Karun basin, the design
criteria for [reliability; deficit] were assumed [90%; 5%], [80%; 20%], [95%; 5%], [90%;
10%] for hydropower, irrigation, domestic and industrial, and environmental demands,
respectively.

2.3. MCDM module

The module uses tangible and intangible indexes of decision making for comparing the
scenarios by extending the main indexes to subgroup-indexes. Subgroups-indexes were
determined in a dual weighting mechanism after preprocessing the indexes. Then, the
subgroup-indexes were unified to obtain ultimate decision indexes.

3. CASE STUDY

The DSS was utilized for integrated water resources planning in the largest basin of Iran
namely; Karun. As shown in Figure 2, six large dams and HPPs of Karun1, Masjed-E-
Solayman, Karun3, Karun4, Gotvand , and Dez are under operation and Khersan 3,
Bakhtiary and Rudbar are under construction at the big Karun basin. The rest of the dams
out of 16 large dams are under investigation in this basin. The Karun basin contributes
78% of hydro electricity generation of the country in full development situation.
Additionally, many irrigation networks are under development with total cultivation area
of 0.8 million hectare including existing and under planning networks as shown in Figure
2.

The Karun river basin is concerned with water quantity challenges such as water resources
scarcity, priority of hydropower and irrigation demands, and inter-basin water transfer
projects. The DSS modules were applied for the Karun water resources planning as
follows:

3.1. Sediment module application

The sediment module were applied for the Karun river basin to estimate sediment in
different locations of the basin based on observed sediments of the Karun1 and Karun3
reservoirs that are under operation since 1977 and 2005, respectively. More than 40 years
sediment deposition of Karun1 which surveyed two times during operating period was
used to calibrate the soil erosion models and Discharge (Q)- Suspended Sediment load
(QS) relationships of gage stations. Sediment profile and area-volume-height of Karun1
reservoir are shown in figure 3 for initial and during the operating periods.

V - 39
 Existing Inter Basin water transfer Existing dams
Under construction Inter Basin water transfer Under construction dams

Under Study Inter Basin water transfer Under Study dams

CheshmeLangan Tunnel

Qomrud Tunnel

Kohrang First

Kphrang Second
Tunnel

Kphrang hird
Tunne

Tunnel
Beheshtabad Tunnel

Marbor Tunnel

Figure 2. Location of large dams, conveyance tunnels, and downstream irrigation networks at the
Dez and the Karun rivers basins

The relationships between the erosion models results and recorded sediment in the gage
station are shown in figure 4 for the Modify Pacific Southwest Interagency Committee
(MPSIAC) and Erosion Potential Methods (EPM) models, separately after calibrating with
the trapped sediment of the reservoirs. The modified relationship of EPM model as the
best model of soil erosion was obtained as follows:

Y=31.106*X0.4255 R=0.806 (12)

Where:

X: EPM sediment estimation per ton/year/km2

Y: the observed sediment in gage station per ton/year/km2
R: regression coefficient

The sediment estimation in upper catchments of Karun basin has increased at least 20%
based on the modified formula of the sediment in comparison with conventional methods,
however, the major features of the HPPs projects did not change based on new estimation
of sediment due to high minimum operation level of the reservoirs.

V - 40
 Reservoir area (km2)
Reservoir water level (masl)

Hydrography of 2005
Hydrography of 1977
Initial capacity Maximum water level
Capacity after operating for 20 years Normal water level
Capacity after operating for 48 years Min. operation level
Initial area
Area after operating for 20 years
Area after operating for 48 years

Reservoir capacity (mcm)

Distance from dam (m)

Figure 3. Karun1 reservoir characteristics during the operation period

A. MPSIAC
B. EPM
Special sediment erosion model

380 1000
360 Observed 900 Observed
340 Linear Linear
(Ton/Km2/year)

800
Logarithmic Logarithmic
320
Pow er
700 Pow er
300 Expon
Expon 600
280
500
260
400
240
220 300

200 200
0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
Special sediment by gage data Special sediment
‫گ‬
by gage data
(Ton/Km2/year) (Ton/Km2/year)

Figure 4. Relationship between erosion models and recorded sediment at gage station (A.
MPSIAC, B.EPM)

3.2. Simulation module application

Simulation of the reservoirs in an integrated river system was carried out to evaluate
different statues of development for the Karun basin. Two different schemes of the Karun2
dam site were assessed by the DSS. The Karun2 dam site is located in the reach between
the existing dams of Karun3 and Karun1 and two schemes of storage and runoff river
HPPs were under investigation at feasibility stage. Location and specifications of the
runoff river and storage schemes is shown in figure 5. In spite of social impacts of the
storage scheme, the advantages of 3.2 billion m3 reservoirs capacity was one of the criteria
to evaluate the schemes.

V - 41
 Legend:
Residential area
Existing dam
Under planning dam
River/reservoir

Figure 5. Location and specifications of storage and runoff river scheme of the Karun2 HPP

The module simulates multi-reservoir multi-purpose river system for different scenarios of
development horizon in terms of fulfilling rate for irrigation and electricity demands as
well as meeting design criteria and the requirements of the hydro-system. The design
criteria and minimum requirements are automatically met by changing the state variables
such as irrigation areas and HPPs install capacity. Figure 6 display irrigation supply
(regulation of agricultural runoff) and cultivation areas for different scenarios of
developing reservoirs in the river system. In order to assess the schemes of Karun2 HPP,
the figure shows the results for two separate cases of existing reservoir of the Dez basin
(the Dez dam in parallel to Karun reservoirs) and full development of the Dez basin as
sensitivity analysis. In the figure, the scenarios of developing the Karun reservoirs are the
same for the two cases, therefore the legends are not shown in the figure for full
development situation. Percentage of irrigation supply varies from 45% to 83% of annual
runoff for without dams as a virtual status and full development situations, respectively
based on the same design criteria of 80% reliability and 20% permissible annual deficit for
the demands. In the methodology of the simulation, release priority was assumed equal for
irrigation and hydropower, that means the reservoirs release was determined by maximum
value of irrigation and energy demands at each time step.

V - 42
 95.0 800

700
85.0

600

75.0
500

Downstream Irrigation area (1000 hr)

Regulation of agricultural runoff (%)

65.0 under 
400
under  operation  under 
3 reservoirs  operation  and  operation 
under 
2 reserrvoirs  (Karun1& and  constuction ,  and  300
Karun3 & operation 
55.0 (Karun1 & constuction ,  and  constuction , 
and 
Dez dam Dez dam) and  investigation  and 
4  constuction 
investigation 
dams of  investigation  dams   of  200
reservoir  dams  of  dams   of 
Karun  &  Karun 
(Karun1& Karun 
45.0 Karun3& Dez Dam  Karun (runoff  (Storage Karun2  (Storage Karun2 
river Karun 2) MOL of 650) & 
without  Karun4&  Dez Dam  MOL of 635) &  100
& 
dam  Dez dam) Dez Dam 
Dez Dam 

35.0 0
0 2000 4000 6000 8000 10000 12000 14000 16000
Active volume of reservoirs (mcm)

regulation of annual runoff (full development of Dez basin) regulation of annual runoff (existing situation of Dez)

irrigation area (full development of Dez basin) Irrigation area (existing situation of Dez)

Figure 6. Regulation of agricultural runoff and irrigation areas for different features of the Karun
reservoirs

The Karun2 storage scheme only increases irrigation supply 1.9% corresponding to 1.4%
of irrigation area in comparison with runoff river HPP scheme assuming exist situation of
the Dez basin. However, these rates decrease to 1% for both factors considering full
development of the Dez basin. The difference for irrigation supply increases to 5.5 % for
high priority of agriculture demand in comparison with equal priority, however, the
contribution of the Karun2 storage HPP scheme is only 0.9%.

3.3. MCDM module application

The MCDM module was used in subject-based framework such as screening schemes of
projects e.g. the Karun2 HPP storage and runoff river schemes. The DSS helped the
decision makers to select the runoff river scheme of the Karun2 HPP based on different
criteria such as economic indexes, drought mitigation, social impacts, water quality and
environmental impacts, and technical and construction aspects. MCDM was used to rank
the indexes for each scheme based on two cases of the preference. In case 1, the indexes
with high importance had relative preference with respect to indexes with medium and low
importance and the indexes with high importance had high preference with respect to low
important indexes. In case 2, the indexes with high importance had high preference with
respect to indexes with medium and low importance and indexes with high importance had
extremely high preference with respect to low important indexes. Table 1 shows that the
runoff river scheme of Karun2 HPP is the best scheme in the two cases—f preferences.

V - 43
Table 1. Average value of decision indexes for the Karun2 schemes by MCDM module
Capital costs
schemes decision Index Drought Social and Feasibility Construction
Economic and
mitigation environmental uncertainty facilities and Average
index investment
index impact index duration value
fund index
Weight Weight Weight Weight Weight Weight
Case
Case1 Case2 Case2 Case1 Case2 Case1 Case2 Case1 Case2 Case1 Case2
1
Case1 Case2
32% 35% 32% 35% 13% 11% 13% 11% 5% 4% 5% 4%
Run off

6 5 9 6 6 6 0.75 0.72
River
Storage

5 6 1 5 5 5 0.25 0.28

4. CONCLUSIONS

This paper presents the tailor-made DSS and the application results for the Karun river
basins in south-west of Iran. Integrated water resource planning is one of the fundamental
elements of sustainable development particularly for the complex hydro-systems with
diverse interests of the stakeholders. In this research, scenario-based DSS was developed in
order to overcome the diversity of stakeholders. The aims of the DSS were to store and
retrieve the data and maps in a Geo-database as well as to evaluate the scenarios by
integrated water resources planning. Although the DSS has been designed for specific
usage in the Karun hydro-system, a great effort was done to make the methodologies and
the software as much as flexible in order to be adaptable to the future changes in the
system. The DSS has many capabilities for water resources modeling nevertheless, the
input data and models parameters were assumed deterministic and uncertainty analysis was
ignored in this research. However, the DSS is being improved by adding more capabilities
and uncertainty analysis which likely is one of the future tasks for the research. The
deterministic DSS was successfully applied in screening the runoff river and storage
schemes for the Karun2 HPP. Moreover, the main feature of hydro projects was finalized
in the Karun hydro-system in terms of inflow, water withdrawal in upstream catchments,
and design parameters of HPPs after debating with the stakeholders and decision makers.

ACKNOWLEDGEMENT

This paper has been prepared based on the results of integrated water resources studies that carried
out by IWPC with collaboration of Mahab Ghodss and Dezab consultanting engineers. I appreciate
all colleagues who had remarkable contribution for managing, supporting and developing the DSS.

REFERENCES

Caro G. D., & Dorigo M. (1998). AntNet: Distributed Stigmergetic Control for
Communications Networks. Journal of Artificial Intelligence Research, 9:317-365.

V - 44
Durbin E.P. , & Kroenke D.M. (1967). The out-of-Kiler Algoritm: A Primer Memorandum
RM-5472-pr. the Rand Corporation California.
Ford L.R. , & Fulkerson D.R. (1962). Flow in Network. Princeton University Press. ISBN
0691079625, 9780691079622, 194 pp.
Heidari A. (2014). Flood vulnerability of the Karun river system and short-term mitigation
measures. Journal of flood risk management, Volume 7, Issue 1, pages 65–80, DOI:
10.1111/jfr3.12032
Jensen P.A. , & Barnes J.W. (1980). Network Flow Programming. John Wiley and Sons
Inc., NY, 408 pp.
Kennedy J ., & Eberhart R.C. (1995). Particle swarm optimization. Proc. IEEE Int'l. Conf.
on Neural Networks, IV, 1942–1948. Piscataway, NJ: IEEE Service Center.
Koutsoyiannis D., Karavokiris G., Efstratiadis A., Mamassis N., Koukouvinos A.., &
Christofides A. (2003). A decision support system for the management of the water
resource system of Athens. Physics and Chemistry of the Earth, 28 (14-15), 599–609.
doi:10.1016/S1474-7065(03)00106-2.
Labadie J.W. ,& Gat T.K. (2011). Spatial Decision Support System for Integrated Water
Quantity/Quality Management in the Lower Arkansas River Basin. Colorado water,
28(1), 2-7.
Labadie J.W. ,& Roger L. (2007). MODSIM 8.1: River Basin Management Decision
Support System, User Manual and Documentation. Department of Civil and
Environmental Engineering, Colorado State University.
Mysiak J., Giupponi C.,& Fassio A. (2000). Decision Support for Water Resource
Management: An Application Example of the MULINO DSS. Integrated
Assessment and Decision Support, Proceedings of the 1st Biennial Meeting of the
IEMSs.
O'Connor D.J. , & Dobbins W.E. (1958). Mechanism of reaeration in natural streams.
Transactions of the American Society of Civil Engineers, 123, 641-684.
Pallottino S., Sechi G.M., & Zuddas P. (2005). A DSS for Water Resources Management
under Uncertainty by scenario analysis. Journal of Environmental Modeling &
Software, Elsevier Science Publishers B. V., 20(8), 1031–1042.
Thiessen E.M. ,& Loucks D.P. (1992). Computer assisted negotiation of multi objective
water resources conflicts. Journal of the American Water Resources Association
(JAWRA), 28(1), 163-177.
Zhang Y., Hawkins D., Zheng W., & Wen M. (2005). The Framework of GIS-based
Decision Support Systems (DSS) for Water Resources Management at the Flint
River Basin. Proceedings of Georgia Water Resources Conference, Athens, Georgia.

V - 45
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Inter-basin Water Resources Development and Integrated Reservoirs System

Operation for the Bandung
Bandung Metropolitan,
N Indonesia
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf fffffjfjjfkkfjjj
Ick Hwan KO
Vice President, Yooshin Engineering Corporation, Seoul, Korea
[email protected]

Adang Saf Ahmad

Director, BBWS Citarum, Ministry of Public Works, Indonesia

Byoung-Seub CHOI
Vice President of Water Resources Division, K-water, Daejeon, Korea

Arie Moerwanto
Director, Directorate of Water Resources Management, DGWR, Ministry of Public Works, Indonesia

Donny Azdan
Director, Directorate of Water Resources and Irrigation, BAPPENAS, Indonesia

Basuki Hadimuljono
Director General of Spatial Planning and Development, Ministry of Public Works, Indonesia

ABSTRACT:
The Citarum River Basin (CRB) is located in the West Java Province of Indonesia and covers
13,000 km2 lying immediately to the east of Jabodetabek conurbation. The CRB, as an important
national strategic river basin, is the heart of the rapid and sustainable economic growth of the
nation toward ‘Green Everlasting Indonesia’. However, rapid increase in population and
industrialization in the Bandung Metropolitan and surrounding upstream area of the CRB has been
accelerating serious water scarcity, river pollution, and flood damage.
In order to solve these complex and complicated water management issues in the rapidly
developing mega city, intensive and integrated efforts for reshaping urban and basin water
management model are urgently required. As part of these efforts, a cooperative project between
Indonesia and Korea has recently initiated to develop new dam(s) and integrated water
infrastructure management system in the upper CRB.
This paper introduces the main concept and approach of this combined structural (new dams) and
non-structural (integrated dams operating system of existing and new dams) measures with on-
going development practice. The project will contribute not only for securing necessary water and
hydro energy for the people in Bandung mega city but also for adapting and mitigating the
increasing risk from global warming and climate change in the region.

Keywords: Integration, Structural (New Dam), Non-structural Measures (Operating System)

V - 46
1. INTRODUCTION

Water resources planners and managers in the 21st century are facing serious challenges
due to population increase, urbanization, and industrialization. On top of these challenges,
global warming and climate change are adding risk especially for water management in the
rapidly developing river basins and megacities in most developing countries.

The Citarum River Basin (CRB) is located in the West Java Province of Indonesia and
covers 13,000 km2 lying immediately to the east of Jabodetabek conurbation. The Citarum,
as a national strategic river basin in Indonesia, is the heart of the rapid economic growth of
the nation. The basin is essential for urban and industrial development in the heavily
urbanized Jabodetabek1 and Bandung areas, especially for export industries, agricultural
production through major irrigation systems, rural water supplies, electricity generation,
and fisheries. While water resources of the basin are relatively abundant, competition for
these resources has increased significantly over the past 20 years leading to a situation of
acute water and environmental stress in the region. Especially, increasing population and
rapid industrialization in the Bandung Metropolitan and surrounding area in the upstream
of the CRB have been accelerating serious water pollution, water scarcity, and flood
damage.

In order to solve these complex and complicated water management issues in the rapidly
developing mega city, integrated efforts for reshaping urban and basin water management
model are urgently required. As a first step, a national framework was enforced by
adopting the principles of Integrated Water Resources Management (IWRM) and
sustainability of the water resources in the 2004 Water Law. Then, the Government of
Indonesia (GOI) and Asian Development Bank (ADB) have initiated a long-term
Integrated Citarum Water Resources Management Investment Program (ICWRMIP) under
the umbrella of a Roadmap towards IWRM in the basin.

However, successful IWRM adoption and implementation hinge upon effective

information management and its sharing and dissemination among stakeholders of river
basin organizations (RBOs). As part of these efforts, a cooperative project between
Indonesia and Korea has recently initiated to develop new dam and integrated water
infrastructure management system in the upper CRB.

2. PROJECT OUTLINE

The Project titled ‘Feasibility Study and Basic Engineering Design for Water Conveyance
and Integrated Dam Operating System for the Bandung Area’ is to provide planning for the
practical solution on the serious water shortage problem of Bandung metropolitan through
inter-basin water transfer development with integrated dams operating system. The project
is implemented by Korea International Cooperation Agency (KOICA) and Indonesian
Ministry of Public Works (MPW). Under the technical supervision of K-water, Yooshin
Engineering Corporation Korea is carrying out the project in close collaboration with the
basin water management agency, Balai Besar Wilayah Sungai Citarum (BBWSC) and
other key stakeholder water management agencies in the CRB in Indonesia.

1 The greater Jakarta metropolitan area known as (Jakarta, Bogor, Depok, Tangerang, Bekasi) and its surrounding area

V - 47
Figure 1 shows the map of the CRB and the Cisangquy sub-basin which covers the project
area in the south of the Bandung metropolitan. Main project work scope includes
reviewing engineering design of new water supply dam (Santosa), providing detailed
engineering design for the conveyance transmission pipe line from Cikalong intake to
Bandung, and establishing integrated dams operating system for the existing hydropower
dams (Cipanunjang and Cileunca) and newly developing Santosa water supply dam in the
upper Citarum basin.

Figure 1. Map of CRB and Cisangquy Project Area in the Upper CRB

3. HYDROLOGY AND WATER INFRASTRUCTURES IN THE CISANGQUY BASIN

3.1. Land Use

As shown in the Figure 1, the Cisangkuy river is located in the upstream of the Citarum
river. The catchment area of the river is 283.7 km2,, The average watershed altitude is El.
1465.2 m, and the average basin slope is 26.82 %. Water flows from south to north in the
project area. Among the total cropping land with 154.3 km2 in the basin, 28 km2 is for dry
field (18.1 %), 19.7 km2 is for rice paddy (12.76%), and 70.4 km2 is for forests (45.61%).

3.2. Hydrometeological Characteristics

According to the recorded meteorological data at Bandung station operated by the BMKG
(Badan Meteorologi, Klimatologi Dan Geofisika), monthly average temperature is 23.27°C,
(maximum 32.3°C and minimum 16.7°C), and the percentage of sunshine is 58.1% in the
region.

Two flow measurement stations (Pataruman and Kamazan) are operated by BBWSC in the
Cisangkuy basin. Flow regime analysis was made at Pataruman gage with recent 10 years

V - 48
historic flow data (2003~2012). The average drought flow is 3.18 ㎥/s, and 95% flow is
3.37 ㎥/s.

3.3. Water Infrastructures

Cisankuy river, a tributary of the upper CRB, is one of the main water sources to Bandung.
In this project area, two existing hydropower dams and reservoirs (Cipanunjang, Cileunca),
three micro hydropower plants (Plengan, Lamajan, Cialong), two tunnels (Plengan,
Lamajan, Cialong) and one water intake (Cikalong) are operated. In addition, a new dam
construction is under planning as part of the project to increase water supply capacity to
Bandung region.

3.3.1. Dams and Reservoirs

Cipanunjang and Cileunca dams and reservoirs managed by Indonesian Power were
constructed for hydropower generation purpose. Natural overflow through Morning-glory
type spillway is practiced during flood control operation. Maximum 5.5 ㎥/s of water is
released from Cipanunjang to Cileunca reservoir through gate installed in the water intake.
From Cileunca reservoir, maximum 6 ㎥/s is supplied to the three micro hydropower plants
(Plengan, Lamajan, Cikalong) in downstream which are operated by PLTA. Table 1
summarizes main features of the two hydropower reservoirs. Figures 2 and 3 display
historic variations of reservoir water level of the two dams, respectively.

Table 1. Main Features of Existing Hydropower Reservoirs

Year of NHWL DWL Effective Storage
Classification
Completion (EL. m) (EL. m) (m3)
Cipanunjang 1930 1,446.0 1,432.6 16,900,000
Cileunca 1924 1,418.5 1,408.2 11,230,000

Figure 2. Reservoir Water Level Variation of the Cipanunjang Dam (2002~2012)

V - 49
 Figure 3. Reservoir Water Level Variation of the Cileunca Dam(2002~2012)

3.3.2. Micro Hydropower Plants

Table 2 summarizes main features of the three micro hydropower plants in the project area.
Figure 4 displays historic variation of the monthly average power discharge at the plants.

Table 2. Main Features of the Three Micro Hydropower Plants

Classification Plengan Lamajan Cikalong

Number of turbine 5 3 3

Type Francis Horizontal Francis Vertical Francis Horizontal

3×1500, 1×2900,
Capacity 3×9000 3×9000
1×2280

rpm 750~600 600 750

Qpower(㎥/s) 3×1.6, 1×2.85, 1×2 3.90 5.50

Head(m) 90 216 140

Year of
1922, 1962, 1996 1925, 1934 1961
Completion

V - 50
 Plengan Lamajan Cikalong

Figure 4. Variation of Monthly Hydropower Discharge (2007~2012)

3.3.3. Water Intake

Cikalong Water Intake is located at the downstream of the Cikalong Micro Hydropower
Plant. The capacity of current water intake is 1.9 ㎥/s (1.4 ㎥/s for raw water, 0.5 ㎥/s for
treated water). It will be increased through a new dam construction by 1.4 ㎥/s.

4. WATER BUDGET ANALYSIS

For stable water supply planning in the future, appropriate sizing of the planned dam is
necessary to be confirmed through water budget analysis (WBA). Also, the basis of
comprehensive WBA is essential for integrated and coordinated water infrastructures in the
project area considering new dam (Santosa), existing hydropower reservoirs (Cipanunjang,
Cileunca), micro hydropower plants (Plengan, Lamajan, Cikalong), intake (Cikalong),
irrigation water demand, and environmental flow for the downstream of the Cisangkuy
river.

KModSim, a Generalized river basin network model jointly developed by K-water Institute
and Colorado State University, was applied for this WBA study. Simulation network was
established by dividing the project basin into 30 sub-basins reflecting main water
infrastructures, facilities, and tributary inflow with reference to the Pataruman gage as
shown in Figure 5.

Based on the network, water supply safety for the main control points were analyzed with
current and planned water intake conditions at the Cikalong. In addition, the effect of new
dam construction was reviewed considering the increased inflow scenario.

4.1. Water Budget Analysis based on Current Water Supply System

Table 3 summarizes the results of WBA based on current water supply system with the
scenarios of present and planned intake capacity. In case of applying the planned intake
(3.3 ㎥/s) under current water supply system, water supply safety was decreased by 8.29%
at the Cikalong Intake and decreased by 12.45% at the Pataruman gage in the Cisangkuy
river (Q95% condition) respectively in comparison with present intake condition (1.9 ㎥/s).

V - 51
The analysis results indicate that additional water facilities are necessary to meet the
increased water demand.

Figure 5. Schematic Diagram of the KModSim Network Simulation Model

Table 3. Results of Water Budget Analysis based on Current Water Supply System
Planned Intake Condition
Current Condition (1.9 ㎥/s)
(3.3 ㎥/s)
Classification Qpower Water Supply Water Supply
Simulation Shortage Simulation
Safety Safety
(Days) (Days) (Days)
(%) (%)
Cikalong Intake 309 92.3 642 84.0
5년 4,018
Pataruman Q355 953 76.3 1,409 64.9
Gage Average
Q95% 993 75.3 1,493 62.8
Note: Q355(Drought Flow) : 3.18 ㎥/s, Q95% : 3.37 ㎥/s

4.2 Water Budget Analysis Reflecting New Dam Development Condition

In order to review the effect of additional water supply from the Santosa dam construction,
WBA study for 96 scenarios with different water supply from the new dam in combination
with different set of micro hydropower plant operation rules.

V - 52
As shown in Table 4, water supply safety was increased with the scenario of additional
dam water supply more than 1.5 ㎥/s and 60% operation during wet season, and 40%
operation during dry season in comparison with current water supply system.

Table 4. Results of Water Budget Analysis reflecting New Dam Development Condition
Simulation Condition Water Supply Safety (%)
Power Plant
Santosa Dam Operation Pataruman
Scenario Qintake Cikalong
Water Supply (%)
(㎥/s) Intake
(㎥/s) Wet Dry
Q95% Q355
Season Season
- - 60 40 1.9 92.14 73.69 74.42
50 1.5 50 40 3.3 95.99 82.85 86.36
54 1.5 60 40 3.3 94.90 79.99 83.20
73 2 50 50 3.3 96.34 83.08 83.55
74 2 50 40 3.3 97.83 88.05 91.74
78 2 60 40 3.3 97.83 84.79 88.63
82 2 70 40 3.3 97.04 82.75 86.11

5. FRAMEWORK FOR DEVELOPING INTEGRATED DAMS OPERATING SYSTEM

5.1 Development Concept and Direction

Various river management factors such as water supply and allocation, hydro-energy
generation, and water quality in the Cisangkuy basin are affected by reservoirs and other
water infrastructures operation in the project area. In addition, new dam constructions are
under consideration to overcome serious and chronic water shortage problem in Bandung
region as shown in Figure 6. In conjunction with these structural measures, developing
integrated water infrastructure operating system is increasingly necessary to maximize
water management efficiency and sustainability.

Figure 7 shows basic the concept and direction for integrated operation system
development. ‘Smart Operation’ means advanced ICT based intelligent water management,
‘Inter-connected Facility’ emphasizes the need for integrated and coordinated operation of
water infrastructures in the basin. ‘Seamless Data’ is to improve the accuracy of the
integrated operation system through continuous and consistent hydrologic data with high
reliability.

V - 53
 Figure 6. Bandung Bulk Water Supply Development Plan

Figure 7. Basic Direction for Integrated Operation System Development

5.2 System Configuration Scheme

Direction of integrated dams operating system development is toward an intelligent

decision support system (DSS) which integrates the whole process of data acquisition and
analysis, watershed runoff analysis, optimal reservoirs system operation, water budget
analysis, water quality analysis, and the assessment of operation effect. As displayed in
Figure 8, the whole system configuration consists of three boxes such as data box, analysis
box, and assessment box. Each box is formed with unit module for carrying out specific
function.

V - 54
 Figure 8. System Configuration

6. CONCLUSITION

This research reviewed the introduction of integrated dams operating system as non-
structural measure in combination with the structural measure of new dam development.

Development and application of the innovative basin water management toolkit will
contribute to knowledge sharing, better decision making in water resources system
planning, operation and environmental protection, conflict resolution on water issues, and
even raising public awareness and consensus building for the greener water management
movement for upper CRB.

ACKNOWLEDGEMENT
This research was supported from the technical assistance funded by Korea International
Cooperation Agency (KOICA).

REFERENCES

BBWSC (2011): Final Design Report for Santosa Dam, Bandung, Indonesia
Kwater (2012): Decision Support System for Integrated Water Resources Management in
the Citarum River Basin, Indonesia
Yooshin Engineering Corporation (2013), Feasibility Study and Basic Engineering Design for
Water Conveyance and Integrated Dam Operating System for the Bandung Area, Inception
Report
Ick Hwan KO (2911): IWRM Practice in Rapidly Developing River Basins, AWRA
Summer Specialty Conference, USA

V - 55
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
Implementing a Sediment Transit Gate at Rizzanese fffffjfjjfkkfjjj
dam

P. Carlioz
EDF-CIH-Le Bourget du Lac-France
[email protected]

& V. Peloutier
EDF-CIH-Mulhouse. France.

ABSTRACT
The Rizzanese Project is a 40m-high RCC dam completed by EDF in 2012 and located in southern
Corsica (France).
In 2008, late changes in the environmental legislation meant that EDF had to adapt the design of
the diversion tunnel in such a way so that it can be used as a sediment-flushing bottom outlet once
the dam is in operation. The target was to ensure that sediment deposited in the reservoir could be
flushed out during flood events.
Both physical and numerical model studies were carried out. This was to assess project feasibility
and sediment flushing-out effectiveness through the bottom outlet.
Main design modifications consisted of (i) adding an Intake tower (with stop logs) upstream from
the diversion tunnel, (ii) lining the tunnel with steel, (iii) installing 2 gates downstream.
In order to minimize impact on the work schedule, construction of the sediment-flushing gate was
divided into two main phases:
- The 1st one, completed by the end of 2009,included structures upstream from the
dam/tunnel grout curtain: Intake tower with stoplogs and 60 m of tunnel steel-lining.
- The 2nd one, completed mid 2013, included 60 m of remaining tunnel steel lining and
downstream radial and roller gates. This 2nd phase was completed after the filling of the
reservoir.
Today, the sediment transit gate works are over, the steel lining conduit was filled with water and
the gates tested. The Rizzanese scheme is under operation. Future significant flood will provide the
opportunity to check the sediment gate effectiveness to flush out sediment.

Keywords : Rizzanese – Bottom Outlet – Sediment – Physical Model.

1. INTRODUCTION

The integration of modern environmental issues in the Rizzanese hydroelectric scheme

(southern Corsica) led EDF to design and build a structure capable of transporting
sediment deposited in the reservoir through the diversion tunnel.
The main features of this structure, referred to as a Sediment Transit Gate (STG), are
described below, along with overall design key phases.
Information is herein provided regarding:
- The feedback from a similar scheme.
- The hydraulic and hydro-sedimentary physical model implemented.
- Numerical modeling of sediment transport used to estimate the STG effectiveness.

V - 56
 - Final design selection
An overview of the STG works main phases is also included.
2. BACKGROUND

The Rizzanese dam project (55 MW) was designed to increase the existing capacity in
Corsica by 10%. Its commissioning was completed in 2013. The Project Owner was EDF-
SEI (Système Electrique Insulaire). The Head Contractor was EDF-CIH (Hydro
Engineering Centre). The Joint-Venture in charge of building the dam was composed of 3
main Civil Contractors (Vinci, Bec and Razel).

Diversion Tunnel

Figure 1. Overview of the dam and location of the Diversion Tunnel

Intake Tower

Diversion Tunnel - STG

Outlet Gates

Figure 2. Overview of the Sediment Transit Gate (STG)

V - 57
The Rizzanese project includes the following:
- A RCC dam, 40 m above foundation and 140 m long at the crest.
- A 5,800 m long TBM-bored headrace tunnel of 3.5/3.0 m in diameter.
- A 89-m high Surge Shaft.
- A 1,245 m long Penstock of 1.75 m in diameter.
- A Powerhouse equipped with 2 vertical-shaft Pelton units, with a rated capacity of
2 X 27.5 MW.
The design flow is 15 m3/s (7.5 m3/s per unit), for a head of 418 m. The net capacity of the
reservoir is 1 hm3. The Full Supply Level (FSL) is 541.0m a.s.l. and the Maximum Water
Level is 545.7m a.s.l. The dam is equipped with a mid-level outlet gate and an overflow
Spillway set at the FSL (Creager-type).

Square/round junction

2 vent pipes

4 orifices

Stoplogs

Figure 3. Vertical section, horizontal

section and front view of the Intake.

Preliminary works were carried out between the summers of 2007 and 2008 (access roads
and main platforms). The main construction works of the dam were completed by the end
of 2012.
The integration of environmental issues, as well as recent changes in the legislation (“Law
on Water and the Aquatic Environment”), led the Owner to decide to install a system that
could offer the largest possible effectiveness with respect to river sediment transport. A

V - 58
study was then launched to convert the temporary diversion tunnel used during the
construction phase into a permanent structure used for sediment-flushing operations during
floods. This measure, which was not initially planned, was okayed by the Project’s
Steering Committee in May 2008. The works, which started in May 2009, were divided
into two phases:
- The first one (May to December 2009) included the excavation of the diversion tunnel,
the construction of the upstream Intake for the stoplogs, and 60 m of tunnel steel lining
from the upstream Intake to the dam grout curtain.
- The second one, completed in 2012-2013, included the installation of the downstream
section of the 60 m tunnel lining, as well as the two downstream gates with associated
Civil structures. A riparian release unit was also installed.

3. DESIGN PRINCIPLES

The STG includes:

- A rough Intake (Fig. 3 above), on the upstream side, that houses (i) a stoplog to isolate
the conduit from the reservoir, and (ii) four orifices of approximate dimensions 3.25 m (H)
x 2.3 m (L) to prevent large floating debris from entering. The integrated stoplog can be
controlled from a platform located at level 529.5m a.s.l. (i.e. 50 cm above the minimum
water level) by means of a lifting beam and a mobile crane.
- A circular steel lining, inside the tunnel, with an inner diameter of 4.2 m and a length of
133 m, which can withstand the abrasion of sediments. Upstream from the conduit, two
vent pipes 600 mm diameter supply the tunnel with air and thereby facilitate the transitions
between pressurized and free surface flows.
- A fixed-wheel gate (dimensions 3.60 m x 3.60 m), and a radial gate (dimensions 3.60 m x
3.80 m) on the downstream side (Fig. 4 & 5) ensuring control of the released flow rate.

Radial gate

Fixed-wheel gate

Figure 4. Vertical cross-section of the outlet

V - 59
 Figure 5. Overview of the Outlet

Figure 5. Plan view of the Outlet.

The initial configuration of the temporary diversion tunnel, as designed in the 1997 Basic
Design, was significantly modified to house the STG. In addition to main design changes
(Intake with stoplogs and vent pipes, steel ling of the tunnel, downstream gates), the
overall layout of the tunnel was modified (excavation section was not changed) mainly in
order to avoid scouring of a potentially unstable rock mass at the outlet, on the left bank of
the river. Tunnel outlet position was moved approximately 30 m downstream and its level
was changed too. In addition, the tunnel slope, initially sub-horizontal over the first 30
meters, was homogenized from the square/round junction down to the outlet, resulting in a
constant slope of 6.9 % over this section.

4. PRINCIPLE OF THE STG OPERATING

The principle of the STG operating is to gradually lower the reservoir level at the
beginning of a flood event, if possible several hours before the flood peak. The aim is to
allow discharges of peak flows at low reservoir level in supercritical flow conditions, the
effect of which is to flush out deposited sediment through the STG, and then downstream
from the dam.

5. FEEDBACK FROM THE JOTTY DAM

One of the early key stages of the design consisted of searching for similar existing
structures and collecting information. The experience gained by EDF Operator of the Jotty
dam over the last forty years provided useful information on how such schemes can be
efficiently operated.

The Jotty dam is part of the Bioge hydroelectric scheme, which is located in northern
French Alps. It is a 50 m high and 3-m thick arch dam.
As soon as it was built, in the late 1940s, the temporary diversion was equipped with two
roller gates positioned parallel (L = 2.1 m x H = 3.7 m), separated by a central wall, with
an individual maximum discharge of 70 m3/s under the FSL. The 118m-long temporary
diversion tunnel is nearly horizontal (0.6 %), with a 16 m2 horseshoe-type section.
Between 1949 (its commissioning) and 1970, the temporary diversion tunnel was not used
to flush out sediment, and a total of 400,000 m3 of material gradually accumulated in the
reservoir, i.e. an average of 20,000 m3/year.
As from 1970, regular sediment flushing operations were carried out, every two or three
years on average, as flood events occurred.

V - 60
After a difficult start (the first flushing operation deposited too much sediment
downstream, causing several complaints), EDF gradually fine-tuned the flood operation
procedure which satisfied all local partners: administrations, local authorities, fishing
associations and tourist infrastructures.

Today the Jotty current operating mode is as follows:

- In response to flood alerts, preliminary information is sent to all local partners at the
latest 4 hours before starting the sediment flushing operation.
- An environmental monitoring lab is installed downstream: real-time measurements of
physicochemical parameters (such as suspended solids, pH, temperature, NH4, dissolved
O2) are made at different points, along the river downstream from the dam. The lab
controls the parameters but can also give instructions to the Operator during gates
opening/closing.
- Unless contrary instructions are given, the sediment-flushing operation is launched when
a given incoming flow threshold is reached (~ 40 m3/s). The reservoir is gradually lowered
below the minimum operating level. A maximum flow of 30 m3/s above the flow entering
the reservoir is kept.
- Total outflow is maintained below a given maximum discharge rate (90 m3/s),
corresponding to the limit required to avoid flooding a recreational site located
downstream.

The aim is to empty the reservoir before the occurrence of the flood peak. Should a flood
peak occur before the reservoir level has been entirely lowered, the sediment flushing
operation is then interrupted by the Operator. Previous experience showed that sediment
flushing loses most of its efficiency if the reservoir level isn’t lowered before the flood
peak.
In case the reservoir is emptied before the occurrence of the flood peak, the sediment-
flushing operation is run. At the end of the flushing operation, the gates are progressively
shut down before the end of the flood (below 25 m3/s). This is to re-fill the reservoir and
"rinse" the river downstream from the dam by discharging clear water over the spillway for
a few hours.
The Jotty temporary diversion tunnel has previously been lined with steel along the
sidewalls, but a large portion remains unlined and is being locally eroded. Repairs are
performed on the tunnel concrete walls every 10 years on average. The latest repairs were
performed in 2000 with silica-fume concrete.
According to the Operator, the main difficulty consists of making the decision to launch
the sediment flushing operation, due firstly to flash-flood type events (requiring only 2 and
3 hours to reach their peak), and secondly to the 4-hour authorization request constraints.
However, only one out of the six sediment-flushing operations performed in the last 12
years was a failure. During these flushing operations, the majority of the sediment
accumulated in the reservoir was satisfactorily flushed out downstream.

6. PHYSICAL MODELING

A 1/30 scale physical model was carried out for the Rizzanese, designed and constructed
by the University of Liège (Belgium). It was divided into two stages: (1) the study of the
hydraulic behavior; (2) the study of the hydro-sedimentary behavior.

V - 61
6.1. Hydraulic Study
The hydraulic studies showed that depending on the reservoir level and the opening of the
radial control gate, there are either pressurized flows, or free surface flows, or mixed flows.
When the downstream gate is fully open, at low reservoir levels, free-surface flows
develop along the entire length of the structure (see Figure 6).
Gate fully open with reservoir level higher than the Intake upper slab (~ 520 m asl),
pressure flows start to develop in the upstream part of the tunnel (between the Inlet
structure and the downstream end of the square/round junction). Further downstream in the
tunnel, free-surface-flow conditions are observed.

Figure 6. Flow at low reservoir levels (tunnel and Inlet)

Gate fully open with the reservoir above 530m asl, mixed water-air flow conditions cease
and the tunnel is completely pressurized.
Starting from free surface flow conditions, when the downstream gate closure starts,
pressure flows develop in the downstream part of the tunnel, with a hydraulic jump.

V - 62
 Reservoir at 526 NGF

Reservoir at 514 NGF

Figure 7. Water level longitudinal profiles along the tunnel, control gate fully opened.

Figure 8. Stage-discharge law against percentage of gate opening

The maximum flow rate through the STG is 280 m3/s at the MWL and 265 m3/s at the FSL
with the gate was fully open (see Fig. 8 above). The maximum velocities measured in the
tunnel are close to 20 m/s for pressure flows and 15 m/s for free surface flows.
The quantity of air supplied into the tunnel by the vent pipes is at a maximum for flow
discharge rates corresponding to the pressurizing limit of the tunnel (reservoir level close
to 520.0 m a.s.l. with gate fully opened). Measured air velocity reach up to 60 m/s.

6.2. Hydro-sedimentary Study

The hydro-sedimentary study was carried out using two 0-2 mm and 0-6 mm crushed
limestone sands, modeling sediment transported by the river. Two types of tests were

V - 63
carried out: stationary tests to check the transport capacity of the tunnel at a given flow rate
with low gate openings (pressure flow), and transient tests to assess the STG during
flushing operations in the flood recession phase.
This study confirmed the ability of the STG to transport sediment from the reservoir
downstream from the dam.
Sediment transport through the STG occurs with reservoir level close to 519 m a.s.l., i.e.
the level of the upper slab of the Intake. The maximum flow velocity in the tunnel at this
level is close to 11 m/s. Above 519 m a.s.l., there is almost no sediment taken through the
STG because of pressure flow conditions in the upstream part of the tunnel. These
observations show that reservoir sediment flushing is effective provided that free surface
flow conditions take place all along the tunnel.
The transient tests showed that it is preferable to carry out sediment flushing operations
once flow discharge rates are sufficiently high. They also showed it is necessary to open
the gates to the maximum. This is to prevent releasing large concentration of sediment
downstream, as well as leaving residual sediment within the tunnel.

Figure 9. Residual deposits at the end of one of the hydro-sedimentary test

7. NUMERICAL MODELING

Numerical simulations of sediment transport were conducted by ETRM, a consultant firm

specialized in mountain river studies. The simulations focused on riverbed sediment
transport to compute the total sediment discharge rate through the STG, as riverbed
sediment is mainly composed of medium/coarse sand and gravel.
The study particularly focused on the evolution of the river longitudinal profile. Flow
hydrographs of the main floods recorded over the past 25 years were applied successively
to the model, which enabled to simulate the evolution of the longitudinal riverbed profile
that could be expected.

The assumptions used for the simulations are as follows:

- Initial level of the reservoir at 530 NGF, i.e. a level close to the minimum operating level
(529.0 NGF).
- No flow turbined during flood events.
- During the flood rising stage, as soon as the incoming flow exceeded 30 m³/s, the
outgoing flow is increased by either 30 or 50 m³/s compared to incoming flow (two cases

9
V - 64
analyzed), and this increase is applied when the downstream flow rate remains below 100
m³/s.
- Gradual opening of gates (mid-level outlet gate is opened first, followed by the STG)
with a gradient of 3 m³/s/minute.
- Closing of the STG if the incoming flow exceeds 250 m³/s (to avoid any risk of
obstruction by floating debris) or if it is below 40 m³/s (flood recession).
- Using the spillway at the end to allow transit of floating debris.

Figure 10. Modeling of bed level variations in the Rizzanese and its two tributaries against
distance upstream from the STG Inlet

The results of the simulations indicate no significant differences between simulations with
an excess flow at 30 or at 50 m3/s. Alluviation in the reservoir remains moderate, i.e. below
3 m in general, whatever the flood hydrograph applied. In particular, the maximum
alluviation measured at the power Intake remains 2 meters below the Intake sill elevation.
The phenomenon usually observed is a sediment deposit at the upstream end of the
reservoir during flood recession and a subsequent recovery of this sediment during the next
flushing operation. The overall slope and the paved character of the riverbed are
conserved. The material concentration on the downstream side of the dam remains below
2%, i.e. at reasonable levels with respect to fish life.

8. OVERVIEW OF THE WORKS

The first phase of the works was performed between May and December 2009.
It consisted of excavating the tunnel, setting the steel-lining along the upstream part of the
tunnel (60 m), and building the upstream Inlet structure.

10
V - 65
 Fig.11. Upstream Fig.12. Tunnel steel lining Fig. 13. Upstream Inlet structure
end of the tunnel excavated 90% completed (temporary
diversion is operational)

The second phase of the works started after filling of the reservoir. It was carried out under
the protection of the STG upstream stoplog, between March 2012 and November 2013. It
involved installing the remaining 60 meters of lining in the tunnel, the two downstream
gates with associated Civil works structures, as well as a riparian release turbine (200 kW).

Fig. 14. Steel lining introduction from Outlet Fig. 15. Outlet Civil works complete.

Fig. 16. Radial gate erected.

11
V - 66
9. CONCLUSION

Design studies for the transformation of the initial diversion tunnel into a Sediment Transit
Gate were produced quickly enough to allow completion of the 1st-phase works in 2009,
before diverting water into the diversion tunnel. Second-phase works took place during
filling of the reservoir, without any significant difficulty and no impact on the Project’s
schedule.
By the end of 2013, the sediment transit gate was fully completed and tested.
In the next following months, once sediment has started accumulating in the reservoir,
significant incoming flood flows will produce an opportunity to check the STG
effectiveness to flush out sediment.

THE AUTHORS

P. Carlioz graduated in civil engineering at INSA Lyon in 1980, has been working for EDF since 1982.
After working on several hydro-projects in France (Haute-Durance, St Egrève, La Bravone, etc.), he was
involved as a resident engineer in a number of large hydro-schemes including Petit Saut in French Guyana
and Nam Theun 2 (1 070 MW) in Laos, and also in the construction of Ling Ao Nuclear Plant in China. He
was then the Design Manager of the Rizzanese hydro-project from 2006 to 2012. He is now technical
supervisor at EDF Hydro-Engineering Center (CIH) at Bourget-du-Lac.
V. Peloutier graduated in civil engineering from ESTP Paris in 1994, completed postgraduate studied in
hydraulics at Glasgow University in 1998. After working for Halcrow and Sogreah, he joined EDF in 2007
where he was involved in detailed design studies of dams (Rizzanese, etc.) and was then Design Manager of
Gavet hydro-project. He is now CIH Civil Branch Head at CIH-Mulhouse.

REFERENCES
[1] Morris, G.L. & Fan, J. (1997): Reservoir sedimentation Handbook, Mc Graw-Hill, USA.
[2] Garcia, M. (2008): Sedimentation engineering, processes, measurements, modeling and
practice, ASCE, USA.
[3] Graf, W. & Altinakar, M. (1993): Hydraulique fluviale Tome 2, écoulements non permanents et
phénomènes de transport, Presses polytechniques et universitaires romandes, Suisse.
[4] Orvain, F & Laperrousaz, E. (2012): Laboratory measurements and sand/mud sediment
erodability in the Longefan EDF retention basin, 6th International Conference on Scour and
Erosion, Paris.
[5] Partheniades, E. (1965): Erosion and Deposition of Cohesive Soils. Journal of Hydraulices
Division 91:105-139, USA.
[6] Malavoi, J.R. & Bravard, J.P. (2011): Elements d’hydromorphologie fluviale, ONEMA, France.
[7] Sogreah (2008) : Rizzanese, expertise sur l’efficacité de la vanne sédimentaire. France.
[8] Koulinski, V. (2009) : Rizzanese, détermination des apports solides. ETRM. France.
[9] Koulinski, V. (2009) : Rizzanese, étude des chasses solides de la retenue. ETRM. France.

12
V - 67
 Conservation Action Plan to Extend Life-Time
of the Djuanfa Dam and HEPP
Herman Idrus1, Anton Mardiyono2 dan Elyawati Siregar3
Jasa Tirta II Public Corporation, Jatiluhur, Indonesia

ABSTRACT

Djuanda multi-purpose dam is one of the Indonesian vital areas. The biggest reservoir in country
with impounding capacity of 3x109 m3 has unique structure of cyclus deical with ogee type
spillway called the Morning Glory tower. The tower acts as spillway, intake for turbine and
irrigation, penstock, and conveying the water to under-water power house which is located in the
upstream of the dam.The multi-purpose dam is intended to meet the domestic, municipal, and
industry water requirements, rice-field irrigation area of 240,000 Ha, flood control, fishery
cultivation, installed capacity of 187.5 MW of hydro electric power plant (HEPP), and tourism. As
a multi-purpose dam, the life-time of dam influenced not only by the volume effective of the
reservoir, but also affected by continuity of the HEPP infrastructures function.
The important issue in the management of Djuanda dam and its under-water HEPP is the
environmental problems (quality and quantity) that arise due to the interaction between local
economic activity and the limited of environmental carrying capacity of reservoir. The quality of
the aquatic environment of reservoirs is strongly influenced by the level of stability of the waters,
carrying capacity, the sustainability of reservoir and function, as well as activities and
environmental conditions of local people.
Jasa Tirta-2 Public Corporation (PJT-2) has a government mandate to manage Djuanda dam and
HEPP, including conserving reservoir environment to extend the dam life-time. This strategic
policies and action plan is devised within Integrated Djuanda Dam Spatial Planning (IDDSP)
that sets up the platform of Masterplan and Roadmap of management and development of Djuanda
dam and reservoir.

Key Words : life-time of dam and HEPP, environmental problems, IDDSP

1
President Director of Jasa Tirta-2 Public Corporation
2
WR R&D Dept.Head of Jasa Tirta-2 Public Corporation
3
Profesional Staff of Jasa Tirta-2 Public Corporation

V - 68
I. BACKGROUND

A Government Regulation number 7 year 2010 on Jasa Tirta-2 Public Corporation (PJT-2)
assign a duty and responsibility in water resources management, by putting forward a
social, environmental and economic functions, for public benefits and adequate to the
fulfillment of livelihood, include the provision of surface water for basic needs; providing
water irrigation; hydro electric power plant (HEPP); the monitoring and evaluation of the
quantity and quality of water sources; and conservation of natural resources.
Djuanda dam as the national vital area and multi-purpose dam, such as technical irrigation
of 240.000 ha, raw water for drinking water to Jakarta city, district/city/municipalities and
industry, and the cultivation of fisheries, the flood control, HEPP 187,5 MW, tourism and
transportation, should be preserved because of the vital function.
One important issues in the management of Djuanda dam is environmental problems (both
quality and quantity) arising due to the interaction between economic activity and the
environmental carrying capacity of reservoir is limited. The quality of the aquatic
environment of reservoirs was influenced by the level of stability and the carrying capacity
of aquatic, sustainability functions of dam and reservoirs, tourism, as well as activity and
environmental conditions of the local community.
A decrease of water and air quality in the area of Djuanda reservoir as a result of increased
element of the waters derived from the activities of floating net fishery (FNF) cultivation,
considering the water quality from outlet of Cirata dam, inlet and outlet of Djuanda dam
with decreasing, signifying the intervention in reservoir, considering no other sources of
pollution. And also sedimentation from other rivers (Cilalawi and Cisomang) affecting the
volume effective of reservoir. The decrease of water and air quality result the construction
of dam and infrastructure of HEPP became rotten (concrete) and corrosion (metal) since
the 2000’s until today. It is feared will lead to a decrease in function and will significantly
reduce the life-time of the dam and HEPP infrastructure.
Djuanda dam as the national vital area so that the efforts to the aquatic environmental
conservation should be implemented properly, to environ the aquatic and dam buffer itself.
The integrity of land as the reservoir buffer area should be preserved from insistence on
development so that is aimed, in accordance not only consider economic aspect, but also
supports the aquatic environmental conservation.

II. DISCUSSION

In the management of Djuanda Dam, there are problems resulting in decreased function,
life-time of the dam and HEPP infrastructure, this happens due to external and internal
influences, which among other things caused by economic activities and environment
carrying capacity of reservoir is limited.

2.1. External Influence

2.1.1. Sedimentation of Cilalawi and Cisomang River

Incoming water source to Djuanda dam comes from Cirata dam, Cisomang and Cilalawi
rivers, as well as the stream of the river that are around dam and rainfall. Water from the
Cirata dam contains relatively small, while the sediments from the rivers high causes
occurred the basic surface of Djuanda dam thus reducing the capacity of reservoirs.

V - 69
The bathymetri study of Djuanda dam results in year 2009 when compared to the map
contours at year 2000.The basic contours of reservoirs around the outlet Cirata dam and the
estuary of Cisomang river showed the abration of reservoirs basic surface ± 5 meters. The
basic contours of reservoirs around the estuary of Cilalawi river showed the abration of
reservoirs basic surface 10-15 meters.

2.1.2. The development Insistence in the Surrounding Area of Djuanda Dam

At this time there was insistence on development around the land of Djuanda dam that
causes many utilization of PJT-2 land in illegally by outside parties and society. Land is
used for house, factory, shop, office, etc. Used of the land without permission and make a
permanent construction. PJT-2 is difficult to conducting surveillance, due to the resistance
from local community who felt disturbed land business.

2.1.3. Regulated Spatial Planning

In Purwakarta District Spatial Planning year 2011-2031, which is related to the Djuanda
dam area, among others contains a spatial strategy, spatial planning system, network
system infrastructure and the pattern of spatial plan. As a strategic area in districts spatial,
with the aim of realizing the development center of industry, agriculture, integrated
tourism, competitive and environmentally empowered.
Djuanda dam has an important position in the development of Purwakarta District because
the problems in the region become strategic issues to be resolved through the suspected
plans development, such as:
a. the decrease in the Djuanda dam water quality because the rise of the activities of the
FNF cultivation.
b. the subsidence decrease dam water level impact on the water supply and irrigation.
c. there is any potential natural disasters such as floods in case of overflow from
reservoirs.

Accomplish a purpose of spatial planning and to answer the strategic issues in

development of Purwakarta District, have been set by spatial planning strategy. The related
strategy and directions:
a. organize and develop inland fisheries activities.
b. developed the water and adventures tourist areas, tours and transport routes to tourist
areas, and building a showcase of results of agricultural production and small industry
flagship in tourist and urban areas.
c. develop protected and/or the cultivation area not built up around dam as a buffer zone,
should be planned selectively, so it can maintain the function, and keeping the assets.
d. improved land use order of land around greenbelt.
e. developing the dam area in accordance with a system of national and provincial and
integrated development of industrial activity, with a development center of agriculture,
and tourism.

Djuanda dam with a broad 8,300 Ha, located in 4 sub-district, (Jatiluhur, Sukasari,
Tegalwaru, and Sukatani), served 2 region development as government, tourism, services
and trade, and small industry and agriculture based on conservation watershed. A system
of infrastructure network plans related to the development plan network system of the
freight transportation covering: 1) stipulation grooves of lake shipping and crossings,
2) rehabilitation of the pier, 3) improving the quality and quantity of freight transport.

V - 70
Other infrastructure network system plan was the water resources network system plan, the
dam management is one part of the development plan, management, and water resources
conservation network, and on the plan of water damaged power also mentioned the
optimization of Djuanda dam.
The development of tourism in Djuanda dam should be integrated with the spatial planning
and the development plan of Purwakarta District that placing it upon a strategic position
with ways to improve synergy between Purwakarta District and West Java Province. On
the map spatial plans in Purwakarta District Spatial Planning year 2011-2031, has set a
Djuanda dam spatial planning, such as contains forest areas, the cultivation, settlement,
industry, tourism, and security.

2.2. Internal Influence

2.2.1. Water quality decrease
a. Water quality standard
The condition of the water quality is not in accordance with aimed based on
Government Regulation number 82 year 2001 about water quality management and
water pollution control, for the first class of water quality classification that can be used
for drinking water, raw water and or designation others equal water quality to the
usefulness.
b. Self purification decrease

Figure 1. BOD5 and COD contents in Cirata dam outlet, HEPP inlet
and Djuanda dam outlet, year 2002–2011
In Figure 1 illustrates the water quality Djuanda reservoir for the last 10 years (2002-
2012) for COD and BOD5 parameters in Cirata dam outlet, HEPP inlet and Djuanda
dam outlet:
1) in the 2002-2006 function self purification still running, seen from the Cirata dam
outlet is improving at HEPP inlet.
2) since year 2006-2011 the self purification function has been no longer, seen from
the water quality in HEPP inlet getting worse than Cirata dam outlet. It indicates
the intervention in reservoir, considering the no pollution source, and indicated that
pollution derived from the FNF activities.
c. Trophic status
The trophic status level of pollution of Djuanda reservoir based on levels of
chlorophyll from Water Resources Research and Development Center data in 2007
(Ecoterra Multiplan, 2010) are on the status of eutrof. This needs to be alerted because
of the content of the waste deposited in the reservoir will affect water acidity is

V - 71
 corrosive. An indication of the high sulphide substances/methane from high
chlorophyll in the bottom of the lake due to the rampant growth of algae bloom that
causes odor and corrosive, so the potential to damage the turbine (Macbub, 2011).
d. Indication of the high H2S
The water quality of H2S year 2006-2011 in Cirata dam outlet, HEPP inlet and Djuanda
dam outlet started in 2007 increase the levels of H2S on HEPP inlet and Djuanda dam
outlet than Cirata dam outlet. That indicated from the high chlorophyll content in the
bottom of the lake due to the growth of algae that comes from fish feeding (Figure 2).

Figure 2. H2S contens in Cirata dam outlet, HEPP inlet and Djuanda dam outlet

2.2.2. The number of floating net fishery (FNF)

The difficulty of controlled the location and the number of FNF cultivation. The current
FNF which is 20,545 patchs, the maximum limit was 2,100 patchs according of the
Purwakarta District Decree. The result of two phase FNF study in year 2006 and 2007, the
activities of FNF cultivation identified as causes of the decrease in the water and air
quality.
Causes and the negative impacts due to the cultivation FNF, such as:
a. Waste from FNF activity
1) FNF cultivation
- Remaining feed not consumed (25-42% feed equal to N=1,420 ton/year and P=13
ton/year)
- fish metabolism
- mass fish deaths
- un operated and broken floating net
- remaining fuel and oil from water transportation
2) FNF unit activity (± 5,000 guard daily activity and domestic waste)
b. Mass fish deaths caused by turn-over (the
mass reversal), a herpes virus, the descent of
the water reservoir elevation, weather and the
decrease in water quality (Figure 3).

Figure 3. Mass fish deaths

V - 72
c. Blooming Algae
Phosphate compounds are not toxic, but in
excessive levels of N nutrient elements
together will accelerate the growth of algae in
the aquatic, in explosive phenomena will arise
blooming algae that occurs almost every year
with the longer periods, often and widely as
well as thick as Figure 4.

Figure 4. Blooming algae

d. Water quality decrease
Water quality analysis tended to decrease the quality of starting from Cirata dam outlet,
Djuanda dam inlet & outlet has been subjected to eutrophication.
e. Decreased air quality, high content of H2S gas at around HEPP.
f. Heavy metals contents
The study shows that fish, feed, and mud around FNF contents heavy metal: lead (Pb),
mercury (Hg), arsenic (As), selenium (Se), zinc (Zn), copper (Cu) and chrome (Cr).

Control efforts extremely difficult to do and many find obstacles faced by management,
such as: 1) required a large cost, 2) extent of the regional management, 3) pressures from
the farmer groups to increase the number of FNF, 4) the work plan of controlling and the
reduction of FNF that has been socialized got considerable pressure from fish farmers,
5) farmers are not willing to move the zone set, 6) businessman and owner of the FNF
much from outside the area of Purwakarta District.

2.2.3. The Rotten and corrosion of HEPP construction and infrastructure

Increasing the number of part of HEPP construction and building subjected to rotten
(concrete) and corrosion (metal) identified as water nutrient element increase that derived
from the remaining of the feeds and the metabolism of fish and human activities. It was
feared that will result in a decrease of HEPP building function significantly.
The influence of environmental conditions on the reduction of HEPP construction and
building as seen in Figure 5 and 6: 1) cooling systems (heat-exchangers and systems of
pipes), 2) modules for electronic controller, 3) groundings, terminals, etc., 4) civil
structures: corrosion in concrete (CO2-aggressive), 5) addition of power failure (faults), 6)
the intensity of the replacement of components are getting higher, 7) rising cost of
maintenance.

Figure 5. The Corrosion of HEPP equipment Figure 6. The rotten and corrosion of HEPP
construction and infrastructure

V - 73
2.2.4. Tourism developmen and activity not suitable with the master plan

The master plan of the Djuanda tourism area are currently is the product of year 2001, is a
plan which development has not been fully implemented because it is not in accordance
with the condition of the field and has not been well managed properly. From study of
various master plan which drawn up in general having good potential in an increase of
tourism. Programs planned, especially with a big investment, should be sharpened in the
criteria and the direction of its development. Some of these programs need to pay attention:
how the management, organizational structure, standardization, provision of facilities, as
well as the infrastructure that will be developed in an effort to maintain consistency of
tourism. Conservation efforts not separated from the development of the tourism,
considering the importance role of Djuanda dam.

Tourism management policies required the formulation of and planning, by developing

activities more actively, not only economically, but also supports the conservation land. It
requires the development of tourism activities that are well planned and realistic in current
condition.

2.2.5. Greenbelt area

On the Government Regulation number 37 year 2010 about Dam, the utilization space of
reservoir including the inundation and greenbelt of reservoirs. Greenbelt areas is certain
areas around reservoir bounded by greenbelt lines. On President Decree number 32 year
1990 on the management of protected areas and Public Work Ministry Regulation number
63 year1993 about Greenbelt, the benefits, and the former river, established that the line
boundary of reservoir (greenbelt) set at least 50 meters from the highest tide point towards
the ground. In this area are not allowed to do the logging of trees and land management.

Djuanda dam not have the buffer area (greenbelt) that serves as buffer zones reservoir area
and to protect of sedimentation and an avalanche including in greenbelt area. Greenbelt
area (up with elevation + 112.5 meters above sea level) also not well-functioned optimally
because not to 50 meters wide at the average level of the tide highest 111.6 meters above
sea level.

2.3. Proposal of Djuanda dam aquatic spatial planning

2.3.1. Policy

The regulation of water resources conservation on reservoir aimed at keeping: 1) the

sustainability of the existence of reservoir, 2) the reservoir carrying capacity, 3) the
capacity of the reservoir, and 4) a water resources function on reservoir. To achieve the
water resources conservation established three major dam activities: 1) the reservoir
protection and preservation, 2) water conservation, and 3) the management of water quality
and water pollution control.

The study of dam utilization policy focused on upcoming events which has relation
directly with dam administering and business development. In Government Regulation
number 37 year 2010 about Dam explained that dam administering done if there is any
change of space due to sediment and/or the utilization of water and water resources for
other purposes. Dam administering is done by assigning dam utilization zones and water
designation.

V - 74
Dam utilization zones includes dam space up to greenbelt lines, that are set up as a
conservation and cultivation functions, by taking into account: 1) water fluctuations that
affected by season, 2) interests of various types of utilization, 3) the role of the surrounding
community and other parties concerned, 4) the region and dam functions, and 5) dam
safety include support buildings.

2.3.2. The proposal of aquatic spatial as the completion of spatial planning of

Purwakarta District

PJT-2 has been proposed the aquatic spatial according to the latest condition to the local
governments in the draft spatial planning of Purwakarta District year 2011-2031 to be
integrated. For these purposes, PJT-2 has done a study “Review of Djuanda Dam Potential
Tourism and Tourism Area Management – Conservation based compiled by Consultant PT
LAPI ITB in 2011.

Referring to the preparation of the study are: 1) evaluates and presents recommendations
for the management and entrepreneurial Djuanda reservoir environment based on
conservation, acquire spatial concept, governance, and the best in terms of the suitability of
the land, climate, vegetation, and the fulfillment of the technical interests, operational,
social and economic, and 2) compiled a action plan of structuring and developing dam area
with conservation based are integrated and thorough.

While the purpose is: 1) defines spatial of aquatic dams and reservoirs in current condition,
and 2) formulate the development of the dam with a comprehensive and integrated
conservation based as a tourist area, greenbelt, and FNF.

Water resources management policies become a reference for the development area of
Djuanda dam. Based on the study, especially those related to the dam, the development
area of Djuanda dam must consider:
a. increased water resources conservation by developing activities that capable to maintain
and protect the function of water catchment areas, the existence of dam as a source of
water as well as control of water sources utilization in order to improve the availability
of water.
b. dam utilization for the justice and public welfare by developing activities that may
involve public in the management and control.
c. increased efforts to water quality management and water pollution control through the
development of environmentally friendly activity, and support efforts to control the
cultivation of FNF.
Based on the dam conservation policy study is concluded:
a. the development of the Djuanda dam should consider: 1) land use that has been
established for the catchment area, 2) utilization area has been set at the inundation of
reservoirs, 3) the utilization of area of land management in upstream, 4) arrangement of
greenbelt areas, 5) the stipulation of water quality required.
b. development of Djuanda dam to be directed to contribute on efforts: 1) dam protection
and preservation of reservoirs, 2) preservation of water, and 3) water quality
management and tackling water pollution, especially control program of damage dam
and handling water pollution in reservoir.

V - 75
c. The development of Djuanda dam should be integrated with the land use planned and
activities have been directed to catchment conservation efforts, the protection of
reservoirs (border area, including greenbelt), and the inundation of water reservoirs.

Recommendations and action plan priority for the development of Integrated Djuanda Dam
Spatial Planning (IDDSP) in appropriate functionality and zoning Figure 8, with an
explanation as follows:
a. Conservation
Djuanda dam region function as a conservation area to preserve the sustainability of the
function of reservoirs and could be developed with limitations and specific directives
do not interfere with the primary function of the area. The importance protection of
water catchment areas function, maintenance and control of reservoirs as water sources
that development activity should utilize of protective/environmental conservation area
of inundation and water catchment areas around it.
b. Morning Glory
Zone around the morning glory area must be secured, along 200 m from morning glory
toward. The mounting with construction of safety buoy protect.
c. Floating net fishery
Zoning and restrictions on the number of FNF maximum 3,257 patch of appropriate
carrying capacity of reservoir environment in FNF study, in an effort to control of
water pollution and restore the function of reservoirs self purification.
FNF arrangement with the management and supervision of dam conservation
integrated: 1) the application of the District Decree which contains statutes zone and
limits the number of FNF, by means of: the zone, stop a new license and reduce the
amount gradually, 2) the demolition/withdrawal without license, broken and expired,
3) making post supervision in every zone, 4) patrolling and inspection routinely,
5) control feed on fish which indicated containing a heavy metals, 6) socialization to
the farmers and society to participate in monitoring and prevention addition of FNF,
7) legal sanction against parties who violates established rules.
d. Greenbelt
Reorganize and restore the function of greenbelt as a buffer zone the line of coast/
border and protection zones of the area as part of the mainland the shore of the dam.
Through the buffer improvements and provision gradually: 1) establishment of buffer
area, 2) ownership of buffer land by PJT-2, 3) improvement of the buffer conditions by
planting, 4) become conservation areas 5) supervision and cooperation with Forest
Public Corporation (Perhutani).
e. Tourism
Development of tourism activities is utilizing the available resources, with as little as
possible to change characteristic of land. Activity which has been running and regarded
start to threaten the function of conservation should be curtailed and directed according
to a planning area development.
f. Industry / market / shop
The management of the zone and control the locations of industry/market/shop
according to land use. For each sector investment, need to apply the Environmental
Impact Study (AMDAL) as the prelude study, and referring to spatial had agreed.

V - 76
III. CONCLUSION AND RECOMENDATION

a. The spatial plan recommendations for the arrangement of spatial planning to the
District of Purwakarta that presents the recommendations for Djuanda dam
environmental management, with the best concept of management in terms of
conformity land, climate, vegetation and the fulfillment of the technical interests,
operational and social economy, and as well as a reference in the preparation of spatial
plan with the concept of an integrated and comprehensive conservation as an Integrated
Djuanda Dam Spatial Planning (IDDSP).
b. The development of Djuanda dam should be integrated with the land use planned and
activities have been directed to catchment conservation efforts, the protection of
reservoirs (area border of reservoirs, including greenbelt), and the inundation of water
reservoirs.
c. Conservation efforts and the proposals of spatial planning as efforts to dam
management to restore the function and extend the life-time of dam and HEPP
infrastructure.
d. Controlling the number of floating net fishery activities should be done as soon as
possible remembering the dam self purification would require a prolonged period of
time to restore to a former condition.

IV. REFERENCE
PJT-2. (2011): Review Potensi Wisata dan Penataan Kawasan Wisata Waduk Ir.H.
Djuanda - Jatiluhur Berbasis Konservasi, PT LAPI ITB, Bandung, Indonesia.
Laboratorium Kualitas Air. (2002 – 2011): Rekapitulasi dan Evaluasi Data Hasil
Pengujian Kualitas Air Tahun 2002 – 2011, PJT-2, Purwakarta, Indonesia.
Biro Bina Operasi dan Konservasi. (2008):Program Pemantauan Kualitas Air Waduk Ir.H.
Djuanda, PJT-2, Purwakarta, Indonesia.
PJT-2. (2006-2007): Studi KJA Waduk Ir.H. Djuanda Tahap I&II, Consultant PT Dua Ribu
Satu Pangripta, Bandung, Indonesia.
Biro Penelitian dan Pengembangan. (2002): Minimasi Kandungan Gas H2S di Lingkungan
PLTA Ir.H. Djuanda PJT II, PJT-2, Purwakarta, Indonesia.
Biro Penelitian dan Pengembangan. (2001): Penelitian Kualitas Air dan Udara di Dalam
dan di Lingkungan PLTA - Tindak Lanjut Meminimasi Bau gas H2S di Dalam
PLTA, PJT-2, Purwakarta, Indonesia.

V - 77
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Sedimentation Management In The Cimanuk Watershed

To Reduce Jatigede Dam Sedimentation

Adi Prasetyo, James Zulfan, Yiniarti Eka Kumala

Research Center For Water Resources
Jl. Ir.H Juanda No .193, Bandung, Indonesia
[email protected]

ABSTRACT
Dam sedimentation problems occurred not only in Indonesia but also all over the world. The most
threatening issue of sedimentation is the shrinkage of dam service-age, shallowness and storage-
capacity declining. There are many dams in Indonesia which have the same problem, one of which
is Jatigede Dam in Sumedang, West Java Province. It is predicted that the rate of sedimentation of
this dam reached approximately ± 7,7 milion m3/year. If this condition happens continuously, it will
cause fresh water and irrigation crisis. To face this challenge, in 2011 Research Center for Water
Resources (Pusair) has conduct some researches related to this issues. One of the researches which
has been done was the prototype of Sediment Control Structure in the Cikamiri river which is one of
the tributaries river in Cimanuk watershed that flows into Jatigede Dam. The objectives is to capture
the sediment in the upstream river to reduce the sedimentation rate in Cikamiri river in order to
reduce sedimentation rate of Cimanuk watershed. Based on the research, this structure can capture
sediment ± 615 m3/year. The composition of sediment consisted of gravels, sand, clays and silts,
while sand is the most dominant material (>60%). To maintain the river morphology, it is important
to consider the aspect of sediment transport equilibrium in downstream area. The differences
between this sediment control structure with other structures (such as check dam and sabo dam) are
the captured sediment in the pond can be taken and managed directly by the local communities,
therefore the operation and maintenance aspect of this structure running continuously without
relying on government budget. Due to this condition, the advantages of this structure is not only to
decrease the sedimentation in Jatigede Dam but also to improve the local economic growth.

Keyword : sedimentation, dam, sediment control structure, tributary river, local communities

V - 78
INTRODUCTION
General Background
River is one of the natural resources which have tremendous advantage for human life. One of its
benefit is the source of fresh water for human needs. Water resources utilization of the river can be
done by build a dam. The existence of a dam is always followed by a sedimentation problem.
Sedimentation can be caused by the deposition of sediments transported by rivers into the reservoir.
The characteristic of the river in Indonesia generally carry sediment material, both bed load and
suspended load. Sediment which transported by the river is generally derived from each catchment
area and will deposited in the reservoir. Sedimentation problems as described above will also be
faced by Jatigede Dam which has been built on the Cimanuk river, West Java Province, Indonesia.
The deposition sediment in the reservoir will shorten the dam service life and will interfere its
function. There are many methods taken to reduce dam sedimentation, such as dredging and flushing
with heavy equipment, but if there is no comprehensive improvement in the upstream area then all
this methods will become ineffective.
Another issues related to sedimentation problem in this river is the illegal sand mining. Many local
people dredge the sand along the river and then sell it without considering the environmental impacts
that would damage the river morphology. Therefore, Research Center For Water Resources
(PUSAIR) built a prototype of Sediment Control Structure in Cikamiri river which is one of the
tributaries in Cimanuk watershed that flow into Jatigede Dam. The location of the structure can be
seen in Figure 1 below.

Location of the sediment

control structure

Figure 1. Location Of The Sedimen Control Structure

Source : Pusair, 2011
The structure was built to control and capture the sediment load in the rivers which flowing from the
upstream to reduce the sedimentation rate in Cimanuk river. So, it is expected that the sedimentation
in Jatigede dam will be reduced. The purpose of this study was to obtain the effectiveness
performance of sediment control structure that has been built on the Cikamiri river in order to reduce
sedimentation rate of Cimanuk River.

LITERATURE STUDY
The construction of dams and reservoirs on a river can have significant impacts on its flow of water
and sediment, and can cause long term morphological changes of the river. Reservoir’s effective
flood control volume, and in some severe cases can cause overtopping during floods. Sediment

V - 79
deposition against a dam can reduce its stability, and affect the operation of low level outlet works,
gates, and valves. The abrasive action of sediment particles can roughen the surface of release
facilities and cause cavitation and vibration. Erosion and degradation downstream of a dam can
undermine the foundation and affect the stability of a dam. These adverse effects could jeopardize a
dam’s integrity and safety. Sedimentation can also affect a reservoir’s water quality and reduce its
flood control, water supply, hydro power and recreation benefits. Consequently, it is important to
take sedimentation into consideration in the planning, design, operation, and maintenance of a dam
and reservoir. (Chih Ted Yang, 1996)

Aggradations and degradation of riverbed.

The balance between annual water discharge Q, yearly bed material discharge Qs, channel slope S,
and the bed material size d for such streams is expressed qualitatively by Lane’s balance analogy
(Garde et. al, 2006) (Figure 2). This equilibrium can be disturbed by natural causes or man-made
changes, and then the channel adjusts to the new conditions by either increasing the slope over a
reach (known as aggradations), or by decreasing the slope over a reach (known as degradation). Thus,
if Qs increased keeping Q and d the same, the slope will increase by sediment deposition so that with
the increased slope and unaltered Q and d, the stream can carry the increased sediment load. In a
similar manner, if Q is increased keeping Qs and d the same, a smaller slope will be required to carry
this sediment load; this is achieved by lowering of the bed levels resulting in reduction in slope and
thus degradation results.

Figure 2. Lane’s balance analogy (Garde et.al 2006)

If the rate at which sediment entering a given reach of the stream is less than that at which it is going
out, the excess sediment will be picked up from the bed and banks, and there will be lowering of bed
level unless the bed is non-erodible; this is known as degradation.

On the other hand, if the rate at which sediment enters a given reach of a stream is greater than the
rate at which it goes out, the channel bed experiences deposition of sediment; this is known as
aggradations. When sediment is removed from the channel bed for construction activity, the sediment
transported by the stream will get deposited in the depression created by removal of material, and
hence the flow downstream will have less sediment load compared to its capacity. As a result,
degradation occurs in the downstream reach.
Sediment Surveys
Knowledge of sediment transport, in terms of both quantity and quality, is essential for design of
river engineering works on alluvial stream. The primary sediment problems on alluvial problems are

V - 80
related to deposition in reservoirs and reservoir life; degradation below dams as related to
establishing stilling basin and draft tube elevation, and so on; bank erosion; bed scour; and deposition
in crossings. To address these problems, one must know the relationship of streamflow to suspended
sediment load and to bed load as well as composition of bed and bank materials, historical
information on rate of deposition and location of deposits in existing reservoirs, and degradation
downstream of existing dams. To estimate the volume of deposition in reservoirs, one needs to know
the unit weight of deposited material, size composition, age of deposits, and the extent to which they
are alternately submerged and exposed. To estimate degradation downstream of major dams one
needs to know the natural sediment transporting capacity of the stream, detention time in the
reservoir, magnitude of the reservoir releases, composition of bed material to a considerable depth,
and the nature and stability of stream banks (Peterson, 1986).
The basic sediment data collection program includes sampling such as the following :
- Suspended sediment samples to be analized to determine concentration as a function of
discharge and grain size composition,
- Bed load samples to be analyzed to determine concentration as a function of discharge
and type and grain size of material moving,
- Samples of surface material on the bed to determine grain size composition.

METHODOLOGY OF STUDY
This study was conducted by the following stages : 1) Literature study from the previous studies
related to sedimentation in Jatigede Dam, 2) Data collection by measuring discharge, water level,
and flow velocity in the Cikamiri river, sediment sampling then take the samples to the laboratory
for further study to determine the type and composition, 3) Analyzing the data, and 4) Monitoring
the sediment pond to determine the volume of sediment captured.

Figure 3. Field Measurement Activity

RESULTS AND DISSCUSIONS

Sediment control structures such as check dam and sabo dam was usually built across the river in
order to restrain and depositing the sediment. The operation and maintenance activities on this
structures were usually performed with conventionally ways using heavy equipment or manpower to
dredge the sediment. But sometimes this activities were hampered by several factors, including the
difficulty to access the location, weather conditions and material handling post-dredging. In the other
hand, many cases show that if the maintenance activities were not carried out continuously then the
sediment will captured and settles permanently in the structure. This condition will cause the

V - 81
structure to lose its function. Therefore Cikamiri Sediment Control Structure was designed by
integrating the weir components and intake channel to drain the sediment into the sediment pond.
This sediment pond is located off stream to accomodate the maintenance and dredging activities so
the dredging activities can be done continuously without being distracted by the weather or river
conditions at that time. The design of Cikamiri Sedimen Control Structure can be seen in Figure 4
and Figure 5 below.

Weir
River flow and
Sediment transport

Sediment pond

Figure 4. Illustration of Cikamiri Sediment Control Sructure

Figure 5. Design of Sediment Control Sructure in Cikamiri River

The maintenance activities of this structure can be performed any time because maintenance
activities carried out in the settling basin area outside the river channel. This is essential because
when flood happens, it is usually carry a large volume of sediment material so we need to perform
the operation and maintenance activities continuously.
Before the construction of structure, river modeling was conducted to determine the side effect of
the structure to the upstream and downstream area. Based on the modeling results in Figure 7 below,
we can see that with the peak discharge 50 m3/s the presence of the sediment control sediment was
not raising the water level in the upstream of the river. So the upstream area will not experience
flooding if the structure built on the Cikamiri river.

V - 82
 Elevation (m)

Distance (m)
Figure 7. Hydraulic Modelling of Cikamiri River

The sediment which captured at the sediment pond can be used as a reference to determine the
effectiveness of the structures. Therefore, monitoring was done in year period to see the volume of
sediment that can be deposited. Sedimentation volume was calculated by measuring the surface
elevation of sediment in the basin before and after dredgin g in a given period of time. Based on the
monitoring results, the total sediment captured by this structure is 615,39 m3/year with the average
sedimentation rate approximately 1,95x10-5 m3/s. The volume of sediment which captured in the
sediment pond can be seen in Figure 8 below.
Volume (m3)

Month (2012)
Figure 8. Volume of sediment captured
The sediment material which captured consists of gravel and sand. The composition of the sediment
grain are 29,17% of gravel, 61,75% of sand, 3.6% of silt and clay at 5.48%. Based on the laboratory
results, it shows that the sediment captured mostly is sand as can be seen in Figure 9 below. That is
why there are many illegal sand mining along the river Cikamiri.

V - 83
 Percentage (%)

Figure 9. Sediment Test Result

Therefore, Research Center For Water Resources (PUSAIR) cooperate with local community in this
area to manage this structure. At this time Cikamiri managed by the local community under the
supervision of the Research Center For Water Resources (PUSAIR). The local community dredge
the sediment manually from the sediment pond to maintain the capacity of the pond. The benefit
obtained by the local community is that they can sell the sand, because the sand in the Cikamiri river
have good quality. This condition helped to revive the local economy growth and can reduce the
illegal sand mining activities because now they can easily take the sand from sediment pond.
Therefore the maintenance costs become lower because of maintenance activities carried out by the
society continously. Collaboration between government institutions and local communities are
appropriately developed to support the sustainability of the project.

Figure 10. Sediment Dredging Activities By The Local Communities (1)

Figure 11. Sediment Dredging Activities By The Local Communities (2)

V - 84
CONCLUSION
The differences between this sediment control structure with other structures (such as check dam and
sabo dam) are the captured sediment in the pond can be taken and managed directly by the local
communities. Therefore the operation and maintenance aspect of this structure continue running
without relying on government budget. Based on the monitoring results, Sediment Control Structure
in Cikamiri river is able to reduce sedimentation to 615 m3/year. This results comes from 1 structure,
if we want to reduce the sedimentation rate in Cimanuk watershed we can implement this structure
in another tributaries river which have high sedimentation rate. Reducing the sedimentation rate in
Cimanuk watershed means reducing sedimentation in Jatigede dam. It is also important to consider
the aspect of sediment transport equilibrium in downstream area to maintain the river morphology
changes so the environmental aspect will also maintained. However the sedimentation problems must
be solved properly, not only focus on the structural solution but also non structural/social solution.
The collaboration with the local communities give us benefit, not only to decrease the sedimentation
in Jatigede Dam but also to improve the local economic growth.

ACKNOWLEDGEMENT
First and foremost, the authors would like to thank to Research Center for Water Resources
(PUSAIR) and Dinas PSDA Garut for the valuable guidance, and finally also to our families and
friends for their support on us in completing this paper.

REFERENCES
Garde, RJ. 2006. River Morphology, New Age International (P). Ltd., New Delhi.
Pusat Penelitian dan Pengembangan Sumber Daya Air. 2011. Laporan Akhir Pengembangan
Teknologi Bangunan Pengendali Sedimen dan Lahan Basah Buatan, Puslitbang SDA,
Bandung.
Pusat Penelitian dan Pengembangan Sumber Daya Air. 2012. Laporan Akhir Monitoring dan
Informasi Bangunan Pengendali Sedimen di Sungai Cibuah dan Cikamiri, Puslitbang SDA,
Bandung.
Pusat Penelitian dan Pengembangan Sumber Daya Air. 2013. Execituve Summary Bangunan
Pengendali Sedimen, Puslitbang SDA, Bandung.
Petersen, M. 1986. River Engineering. Prentice-Hall, Englewood Cliffs, New Jersey.
Yang, C.T. 1996. Sediment Transport: Theory and Practice, The McGraw-Hill Companies, Inc,
U.S.A.
Yiniarti, Angkutan Sedimen. 2011. Program Pendidikan SP 1-PSDA Teknik Sipil, ITB, Bandung.

V - 85
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Intake Vortex Flow Effect on Sediment Evacuation of Dam Reservior

using Physical Modelling

A. Zabihollah Zadeh
Jondishapour University of Technology,Dezfoul,Iran
[email protected]

S.M. Kashefipour
Professor of Shahid Chamran University,Ahwaz,Iran

ABSTRACT
Vortex flows are featured by some complicated hydraulic components,thus it requires to conduct an
independent study to characterize each individual intake type and orientations.Sediment motion, as
we know has very compound behaviour as well. Such intricate disposition has made scientists to
propose empirical relationships to predict particle behaviour. Therefor it is inevitable to arrange
an autonomous research to study vorticity-sediment interaction for different strutural and
hydraulic conditions.In this study vortex flow effect on sediment evacuation of reservoir has been
physically modelled for semi-inclined intakes.Geometric and hydraulic parameters involved have
been identified using dimensional analysis.They have been measured to evaluate vortex flow
intensity and to characterize its interaction with sediment motion.
Increase in Circulation Number as well as reduction in Relative Depth of Sediments( )caused
volumetric consistency of vacated water( ) to arise as an indicator of particle vacation
magnitude. Sediment transportation index for vortex free condition( =initial volumetric
consistency) is measured(max=10 )and vortex effect on its amplification has been detected as
well(280~425%).
If circulation number(  ) could be limited up to 0.5, then ( ) would be lowered than 3× 10 .
But at the condition of > 3, ( ) values of downstream water will drop to zero for each vortex
flow intensities . This means that in order to keep downstream water quality safe of sediments
moved out by votices,  should be confined up to0.5 or to keep dreg surface low enough to
maintain the condition of ℎ > 3 . That way we can trap catchment sediments in reservoir if it is
favorable. In the contrary we could taking advantage of the hypothesis by providing adverse trend
of changing parameters to amplify sediment evacuation of reservoir under the influence of swirling
flow intrusion.

Keywords:sediment evacuation,vortex flow, circulation number, consistency

1.INTROCDUCTION

V - 86
Several scientists had devoted their time eliciting equations and charts to classify and
prognosticate vortex flow attributes such as critical relative submergence for various intake
types,air entrainment,changes imposed on intake discharge coeficient,etc.For instance
Inviscid theory were developed at 1950 by Binnie and Harris to study discharge coeficient
for nozzels. Fluid viscosity effect on vortex flow has been studied by Einstein and
Li(1955), that resulted in Eddy and effective viscosities recognition. Polikovski and
Perelman(1956)had studied vortex core depth.Equal- velocity criterion has been proposed
by Danny and Young(1957)as a result of their research on geometric, dynamic, and
kinematic simulation.
Sediment-vortex compound behaviours have been investigated as well via numerous
experiments throughout the world. However as it is mentioned before, delicate governing
theorems, makes it necessary to probe each individual structural,hydraulic and geometric
circumstances for every specific phenomenon seperately.
In this research we are about to see what effect vortices have to amplify particle movement
into semi-inclined intaks at typical proportions of southwestern Iran dam reservoirs.
Regulated water can approach intakes through a glassy flume over sediment
blanket,spreaded along the apron,that makes phenomena to be observed. Hydraulic and
sediment hydraulic dimensionless parameters involved have been evoked of dimensional
analysis, including Circulation Number(N =  ), particle density Froud Number(
V/√(gD(Gs-1)))
as incipient motion parameter for particles, Relative Depth of Sediments ( ) beneath the
inlet, and volumetric consistencyof vacated water(Cv). Sediment depths were increased
step by step. At each step dimensionless parameters have been measured for different
vorticity strengths provided by varying hydraulic conditions.Whole measurements have
been repeated for three intake diameters including 2,1.5, and 1.25 inches to monitor
probable scale effect available.
It was expected, consistency of discharged water to decrease when sediment depth
increases or circulation number declines.Our major purposes is to determine what ranges
of particle consistencies in vacated water could be expected at certain relative depth of
sediments and vortex strength. What percentages vorticities can influence sediment motion
into intakes?
Which particle relative depth or circulation number available makes sediment evacuation
eliminated, thoroughly?

2.METHODS
In order to provide practical results of phenomena, typical ofsouthwestern Iran dam
reservoirs withtwo semi-inclined intakes was physically modelled at the scale of 1/200. A
glassy flume of about 5m×1m×0.5m with fine silts spread along the apron, was built on a 0.8
meter height platform.
To determine the most important dimensionless parameters involved in our study, we
conduct a dimensional analysis at the first place.That way we could taking advantage of
elicited results universally.
2.1.Dimensional Analysis

V - 87
Sediment-vortex flow effective parameters can be divided into geometric,kinematic, and
dynamic parameters.To illustrate geometric variables in our model an image has been
shown here,see Figure1.
2.1.1.Geometric Variables

The most essential geometric variables are intake diameter(D),particle size index(ds∽ d50),
intake submergence(h), and deposit surface depth beneath the intake inlet.

Figure1.Constructed physical model

2.1.2.Kinematic Variables

Intake discharge(Q), Intake velocity(V), circulation imposed upon inlet streams( ), and
vacated sediment consistency(C).
2.1.3.Dynamic Variables

Dynamic variables that phenomena could be influenced with are, weight force(Fg) by
gravity acceleration index(g), viscosity force(Fv) by dynamic viscosity index(µ), elastic
force(Fe)by eddy viscosity index(ε), surface tension force(F )by surface tension index( ),
pressure force(Fp)by pressure index(p), inersial force(Fi) by density(ρ) as index, and
differential density of particles(ρs-ρ). To make vacated water consistency dimensionless,
we can making use of volumetric
consistency(C = ,
V = sediment volume, V = water sample volume, and (V + V ) =
total volume of sample) .
General function is as below:

F1(D,h,hs,V,Q, ,c,g,µ,ε, ,p,ρ,ρs-ρ)=0 (1)

Because of incompressible flow in open channels, we could condone elasticity(ε). Pressure

foce could be eliminated of resultant effective stream forces using one freedom degree as
well. Ignorable amount of viscose and surface tension forces, as well as their meagre
changes at laboratory ambient tempretures makes it possible to cross out( ) and ( ) In
particle movement studies. We are taking into account intake velocity and diameter
parameters, then intake discharge could be cancelled as well. Thus, it results in this
relation:

V - 88
 F2(D,h,hs,ds,V, ,Cv,g,ρ,ρs-ρ)=0 (2)

Probation of ten above variables is shown,everythree major dimensions L,M, and T are
incluced. If we cosider D,V, and ρ as repeated variables, then to evoke dimensionless
parameters from unrepeated variables such as g,it can be written( bakinghom theorem):

1=V D ρ g=(LT ) (L) (LT )=M L T (3)

It can be concluded:
1+ 1+ 1=0

- 1-2=0⤇( 1=-2 , 1=+1 , 1=0 ) ⤇ 1= V D g= (4)

1=0

The whole process has been performed for each variable and resulted in:

π2 = =N (5)

3=C (6)

π4 = (7)

5= = − 1 = (G − 1) (8)

6= (9)

7= (10)

If we adverse and take a radical of 1and 5, they come out to( ) and ( ).
Multiplying these two parameters will resulted in:

8= . = (11)
( )

We can taking advantage of this parameter that is compound of intake froud number and
particle density indicator as kind of particle incipient motion parameter.Thus it is
concluded, ourfinal parameters involved are as below:

F3( , , , −1, , , )=0 (12)

( )

2.2.Laboratory Water Supply and Measurement

This research has been conducted at Shahid chamran university water laboratory.In order
to provide a constant hydraulic head during experiments,a 6 meters height head tank is in
circuit. Water is pumped of an underground water reservoir into the head tank.

V - 89
Subsequently water flows through operating hydraulic instruments in a closed circuit to
make measurements more accurate,see Figure2.
2.3.Discharge Measurement

Since discharge is of paramount importance in most hydraulic researchs, it is necessary to

measure its magnitude accurately. In this hydraulic model input discharge is evaluated via
a flow meter installed online into watersupply pipe and is controlled by a valve.
outputflow, is meseaured by WSC flume that can be checked by a cutthroat flume.

Figure2.Constant head-Head tank

2.4.Model Features

We have directed input water appraoching the reservoir through underneath layers at very
low velocities to simulate the real.Water supply pipe discharges water into the inlet
masonry box through a vertically upward transition to dissipate its energy mostly.Then
water flows through three underside openings into masonry transition and glassy flum
afterwards. Water sweeps deposit surface at the length of 3.5meters toward intakes, where
odservations have been performed.
We constructed flume walls at the inward slop of 1/0.4, thus intake orientations are 21.8 ᵒ
with horizen. At the first step we installed two 2” pipes at the distance of 12cm and 30cm
from each other and end wall respectively.
2.5.Measurement Techniques and Data Record

- At each experimentvortex type occurrence is determined using visual criterion presented

by Durgine&Hecker(1978). You can see illustrations for each 6 vortex types, monitored in
our experiments,see Figure3.
-We read intake submergence(h) using two scale gauges, installed on wall, one near the
intakes and another at reservoir upstream. WSC flume measures total vacated water
discharge for two intakes(Q ), thus output water for each intake is(Q= ),that is
remeasured using a cut throat flume. Pipe velocity for each intake diameter(D)
is⤇V= ⤇F (submergence froud number)= .
- Since Circulation magnitude( ) is constant along free vortex region(upper half of vortex
tube), we measured vortex core air tube diameter(2r) at a point near its middle length using
a T-Ruler on flume wall.

V - 90
- Then we have measured tangential velocity(V )at the above mentioned point,tangential to
vortex air core by applying a Micro pitot tube,see Figure 4.
- Because tangential velocity(V ) for whole length of each rotating orbit on vortex air core
is almost constant, we can conclude that:

=∑ V . Δs (Δs =Length element on vortex tube rotating orbit)⤇ =V . S = V . 2πr

(13)

Vortex type 1 Vortex type 2 Vortex type 3

Figure3. Vortex type 4 Vortextype 5 Vortex type 6

- Thus we can calculate Circulation Number(N ), as the indicator of vortex flow strength,
Eq.14.

N =Γ (h=Intake submergence,Q=Intake discharge) (14)

- Gradation curve of sediments is drawn and its particle gravity is measured(G = 2.63)

Figure 4. Pitot tube-Tangential velocity measurement

-We have used three intake pipe diameters(2”,1.5”,1.25”) to explore probable scale effect
on phenomena.Whole experiments is performed for each pipe size.

V - 91
-Sediments have been spreaded beneath intake inlet in different relative depths of intake
diameter ( = , , 1, 1.5, 2, … ).
Then we take samples of intake downstream water for each vortex type occurance at each
deposit relative depths to determine a relative depth with no particle evacuation at every
vorticity power, see Figure 6.
- To meter sample’s volumetric consisteny(C = , V = Solids volume, V =
Water volume), total volume of sample(V + V ) is measured in an scaled container and
weigh(M + M ) using an accurate balance.Then we evaporate water(in an autoclave at
80 ⃘ C to avoid boiling) and weigh it again(M ), naturally container’s weight is
contemplated. Differential weight is meant as water mass(M ). Since laboratory water
density at ambient tempreture is 1.0, (ρ = 1.0)⤇M (gr)=V (cm ),Thus if we subtract
V of (V + V ), it resulted in sample solids’ volume(V ). That way C is computable for
each sample,see Figure 5.

Figure 5. Volumetric consistency measurement for vacated water

-Incepient motion dimensionless parameter( ,V&D=Intake velocity and diameter )

( )
Has been calculated for each experiment.

3.RESULTS
As it is mentioned previously, we increase depth of dregs’ surface to explore vortex flow
effect on sediment evacuation and to determine the relative depth in which, no particle is
vacated even at the presence of the most powerful vortex types. N and C are the most
fundamental measured variables at this study. Resulted graphs for 1.25” intake pipe is
shown as below(Y axis demonstrates N values and X axis shows their corresponding
values for C × 10 ,see Figure 6.
We can elicit some invaluable results of the records:
-There is an initial volumetric consistency(C ) for near zero circulation numbers or in
other words at the absence of every vortices (Volumetric consistency of vortex free
stream).Thus total vacated sediment can be stated as Eq.15.

V - 92
V - 93
 C =C + C (C =Vortex originated volumetric consistency) (15)
-Particle vacation increases (C values) with vorticity amplification( when N grows).
- Vortex free stream consistency( C )declines when sediment surface elevation
lowers(C reduces to zero when reaches 1.5).
-C significantly mitigated by sediment relative depth augmentation
(C scales down to zero when reaches 3 for all vorticity powers).
In another data detection, relationship among N , particle density froud number
( ) , ,and C has been checked for each intake pipe,see Figures7,8.
( )

V - 94
Graphs show, declines when N increases and C goes down, when rises for
( )
all tree intake sizes .
-Curves location for different intake diameters replace each other at varying sediment
depths with no perceived order, thus illustrates nearly identical behaviour for various
intake proportion scales.
-It is found that C is below 0.3× 10 when N is lower than 0.5 for every intake sizes.
-At =3 , C extenuates to zero for all intake proportions(D) and every N values.
During particle movement studies toward intakes a remarkable phenomenon was observed:
-For relative submergences( )lower than 2, sediments move toward inlets as bed load.
-For relative submergences between 2 and 4.5, particles partially suspend in reservoir and
appraoch intakes afterward.
-For relative submergences greater than 4.5, sediments move upstream reservoir in an
spiral shape manner, suspended to the nearly surface of water, and approach intakes as
suspended load,see Figures 9,10.

Figure 10. Sediment locomotion at various relative submergences

4.DISCUSSION AND CONCLUSION

- Curves’slope changes on Figure 6 depict upward trend of C gradient by N

augmentation.This means as it was expected more severe vortices cause more sediment
vacation.
-To eliminate particle entrance into intakes or to trap dregs at reservoir, it is of paramount
importance to keep sediment surface depth(h ) more than 1.5D in case of vortex free
hydraulic conditions and more than 3D for vortex involved situations, even for the most
intense types.

V - 95
-To evaluate vortex flow relative effect on particle vacation, differential C could be
described using Figure 6 graphs data,Eq.16,17.(For C values, quetient 10 could be
omitted of following divisions)
For( = ):
.
(C =1)&(C =3.8)⤇(DifferentialC )= = = =280% (16)
For( = 1 ):
. .
(C =0.4)&(C =2.1)⤇(DifferentialC )= = = .
=425% (17)
-This means, despite C losses by (2.8 for = to 1.5 for =1.5), Nevertheless
vortex flow relative effect on particle transportation is more appreciable for less elevated
sediment surfaces.
- Curves depicted in Figures 7,8 verify mentioned above evoked conclusions as well and,
-No regular scale effect is detected via testing three intake sizes at the scale of , ,
and .
-Observed three different particle motion behaviour(Figures 9,10)respective to , requires
different attitudes for our sediment-related plans, whether if to trap them or moving them
out of reservoir.
ACKNOWLEDGEMENT
I appreciate my professors’ guidance and my wife’s kind helps. I also tender my highest respects to
my deceased supervisor,Dr. Mahmoud Bina.

REFERENCES
Durgine, W.W.& Hecker,G.E.(1978): The Modeling of Vortices at Intake Structures,ASCE
Joint Symposium Desigh and Operation of Fluid Mechinery,Csu Fort Collins.
Yildirim,N& Kocabas,F.(2002): Prediction of Critical Submergence For an Intake
Pipe,Journal of Hydraulic Research, Vol, 40, No4.
Munro,R.J.& Bethke,N. & Dalziel,S.B.(2009): Sediment resuspension and Erosion by
Vortex Rings,Physics of Fluids(Impact Factor 1.94).01/2009;21.Dol:10.1063/1.3083318.
Mattilio,M.&Alsina,J.M.&…(2012): Experimental Investigation of the Nearbed Dynamics
Around a Submarine Pipeline Laying …,Advances in Water Resources,Volume 48,P.31-46
Stastna,M. & Lamb,K.G.(2002):Vortex Shedding and Sediment Resuspension Associated
With…,Geophysical Research Letters,Volume 29,Issue 11,Pages7-1-7-3,June 2002.
Saffman,P.G. & Baker,G.R.(1979): Vortex Interactions, Annual Riview of Fluid
Mechanics,Vol.11:95-121. Dol:10.1146/annurev.fl.11.010179.000523.
Munro,R.J.(2010): Experimental Study of The Critical Criteria for Incipient Sediment
Movement Using a Simple Model Eddy,International Congress on Environment's
Sake,Fifth Biennial Meeting,Ottawa,Canada.

V - 96
 Spatial Modeling of Cimuntur Catchment Area for
Comprehensive and Integrated Watershed Management
By :
Ajeng Aprilia*, Kuntho Wibisono* , Dr. Ir. Suharyanto**, MSc., Ir. Hari
Nugroho**, MT
*) Undergraduate Student at Civil Engineering Dept, Engineering Faculty, Diponegoro University
**) Lecture at Civil Engineering Dept, Engineering Faculty, Diponegoro University
Email : [email protected], [email protected]
[email protected], [email protected]

ABSTRACT
Cimuntur catchment area is one of catchment areas which contribute to high
sedimentation to Segara Anakan lagoon at Citanduy River Basin. This high sedimentation
from Cimuntur catchment area is pre-dominantly due to watershed degradation, causing
increase erosion rate and sediment material transported downstream. At year 2006, the
estimated erosion rate reaches 89 ton/ha/year which is equivalent to sheet erosion 0.9
mm/year. This paper present the GIS modeling on predicting the erosion and
sedimentation rate resulted from Cimuntur catchment area using USLE approach. The
development of this GIS modeling is calibrated by field sediment measurement. The
developed GIS model will then used to plan the land conservation measures required to
reduce soil erosion rate at Cimuntur catchment area. The model shows that in year 2013,
the erosion rate has risen up significantly. To reduce the erosion and sediment material
from Cimuntur catchment area, it requires urgently a more comprehensive and
progressive measures combining some physical, vegetal, and no-physical measures. It
shows that implementation of such measures can reduce the erosion and sediment
material down by 60-75%, thus reducing the sedimentation at Segara Anakan as well.

Key words : Watershed management, GIS Modeling, Erotion-Sedimentation.

INTRODUCTION
Cimuntur Sub-watershed with area ±58.845,85 ha is part of Cintanduy watershed.
With wide area, making Cimuntur Sub-watershed is one major of contributor to sediment.
Land degradation that occurs in the upstream Cimuntur Sub watershed resulted in
a high rate of erosion and sedimentation. It can be seen from deteriorating water quality.
Indication of deteriorating water quality can be seen from the river water turbidity with
reddish brown color which indicates that the rivers flow carrying large sediment loads.
To investigate the erosion problems, it is necessary to perform this study to assess
the extent of erosion and sedimentation in Cimuntur Sub Watershed using GIS spatial
modeling.

PURPOSE AND OBJECTIVE

The intent of this study is to predict the rate of sheet erosion and sedimentation at
Cimuntur Sub-watershed through GIS modeling using ArcGIS software.
The purpose of this study is to plan for the management and conservation
techniques in an integrated way at Cimuntur sub-watershed.

V - 97
STUDY LOCATION
Cimuntur Sub-watershed geographically positioned at east longitude 108°16 '-
108 ° 32'40 "south latitude 7 ° 4' - 7 ° 21'45". Boundary Cimuntur sub watershed is as
follows:
1. At East : Cijolang Sub Watershed
2. At South : Ciseel Sub Watershed
3. At West : Citanduy Hulu Sub Watershed
4. At North : Citarum Watershed and Cimanuk-Cisanggarung Watershed

Figure 1. Cimuntur Sub-Watershed Location Map

LITERATURE REVIEW
Conservation is an effort to preserve the environment, but still consider the
benefits that can be obtained at any time by sustaining the presence of each component of
the environment for future benefits. These conservation efforts are necessary partly due to
persistent degradation to the soil in the study area which promote erosion and
sedimentation to increase.
Erosion is a process where soil particle detachment from its main body and then
the detached soil particles carried by water flow or by wind movement. Flowing
downstream, the transported material will then deposited somewhere.
The rate of erosion and sedimentation at Cimuntur Sub-watershed is anlysed
using USLE method which define the soil erosion rate is influenced by several factors
such as : rain erosivity (R), soil erodibility (K), slope and slope length (LS), land cover
(C), and practical conservation (P).
Rain erosivity shows the degree of a rain drop to cause soil erosion. The higher
the number, the potential of soil erosion to take place is higher. Rain erosivity is the
driving force that causes detachment of soil particles and thus allowing soil particles to be
carried to a lower/downstream place.
The erosion is also influence by a factor call soil erodibility (K) which indicates
the degree easiness of soil particles to be detached from its source. The higher the

V - 98
number, shows that that soil is easily detached. By definition, soil erodibility (K) is the
amount of soil lost each year on average per unit of rainfall erosion index on a piece of
land without crops (bald), without the prevention of erosion, with land slope 9% (= 5 °) ,
and a length of 22 m (standard plot).

Table 1. The value of K for some types of soil in Indonesia

No. Type of Soil K Value
1. Latosol (Inceptisol,Oxic subgroup) Darmaga 0.02
2. Mediterania Merah Kuning (Alfisol) Cicalengka 0.05
3. Mediterania (Alfisol) Wonosari 0.21
4. Podsolik Merah Kuning (Ultisol) Sentolo 0.15
5. Regosol (Inceptisol) Sentolo 0.11
6. Grumusol (Vertisol) Blitar 0.24

(Source : Arsyad 1989 in Suripin 2002)

The slope and slope length can be determined from topographic map. Both slope
length (L) and steep slopes (S) affects the amount of soil lost due to erosion. LS factor is
the ratio of soil lost from a plot with a specific length and steep slopes compared with that
in a standard plot.

Table 2. The Value of Slope Index (LS)

Number of Contour
No Steepness Class Slope Range LS Value
line per cm
1. Flat < 8% <2 0.4
2. Gently Sloping 8-15% 2-3 1.4
3. Slightly Steep 15-25% 3-5 3.1
4. Steep 25-40% 5-8 6.8
5. Very Steep >40% >8 9.5
(Source : Hamer,1980)

Land cover is the ratio of specific soil and plants with bare soil. On bare
soil (standard plot) value of C = 1.0. Meanwhile, to get the value of the annual C-
note a change in the land use change each year.

Table 3. The Value of C for Some Land Use in Indonesia

Number Cropping Types C

1 Fallow land but processed periodically 1.00
2 Irrigated rice field 0.01
3 Rainfed rice field 0.05
4 Plants cultivated (not specified) 0.7
5 Cassava 0.8
6 Corn 0.7
7 Beans 0.6
8 Potato 0.4
9 Peanuts 0.2
10 Paddy 0.5
11 Cane 0.2
12 Bananas 0.5
13 Fragrant lemongrass 0.6
14 Natural forests;

V - 99
 - Full of litter 0.001
- Litter slightly 0.005
15 Shrub / grass 0.3
16 Once reeds burned each year
0.1
17 Vacant lot, not processed 0.95
18 Vacant land cultivated 1
19 Shifting cultivation 0.4
20 Plantation crops with good cover 0.1
21 Plantation crops with bad cover 0.5
22 Undisturbed shrub 0.001
23 Undisturbed forest, a bit litter 0.005
24 Undisturbed forest, full of litter 0.001
(Source : Arsyad,1989 in Suripin 2002)

Practical conservation is the ratio of the lost soil when the soil conservation
efforts carried out (terras, plants in contour and so on) compared with no soil
conservation efforts.

Table 4. P Values for Some Types of Practical Conservation

No. Conservation Practices P
1. Terrace:
- Standard design and in good condition 0.04
- Standard design and in average condition 0.15
- Standard design and in poor condition 0.35
2. Traditional terrace 0.04
3. Land Management and Cropping following contour at a slope :
- 0 – 8% 0.50
- 9 – 20% 0.75
- > 20 % 0.90
4. Grass cultivation (Bahlia) in a strip :
- Standard design and in good condition 0.04
- Standard design and Not in good condition 0.40
5. Crotaliria cultivation and rotation 0.60
6. Use of mulsa (jerami 6 ton/ha/year) 0.30
Use of mulsa (jerami 3 ton/ha/year) 0.50
Use of mulsa (jerami 1 ton/ha/year) 0.80
7. Planting with low cover vegetation in a yard :
- High density 0.10
- Average density 0.50
(Source : Arsyad,1989 in Suripin 2002)

Foster and Meyer (1977) explains that sedimentation-erosion process which is

caused by water mainly include the release process (detachment), washout
(transportation), and deposition of the soil particles. The ratio of the transported soil
erosion to the deposited eroded material in a stream is called Sediment Delivery Ratio or
SDR (Wischmeier and Smith, 1978).
Conservation efforts must be performed to reduce erotion and sedimentation
problems. Conservation efforts can be done by vegetative conservation and/or mechanical
conservation.
Vegetative conservation is an effort to conserve soil and water through the
provision of land with vegetation cover and land management that protected the soil from
direct raindrop.
Mechanical conservation is the conservation effort done by constructing sediment
control structures (check dams). The structures serves to catch the sediment material
transported by water, such that does not cause further erosion.

V - 100
METODOLOGY
The study was started by collecting data both primary data and secondary data.
This data type consists of rainfall data, topographical maps, soil type, and land use.
From the data that has been collected, the transformation of point data into spatial
data and inputted into GIS Map is performed. The analysis on the soil erosion spatially is
performed by overlying maps of rain erosivity, soil erodibility, LS factor, and CP factors.
From the analysis, it can be determined that conservation efforts need to be done in each
area.

ANALYSIS OF EROSION AND SEDIMENTATION

Analysis of erosion and sedimentation in the sub watershed Cimuntur using
USLE method with the following steps:
1. Calculating Rain erosivity factor
Rainfall erosivity factor is calculated from the daily rainfall data from
five rainfall stations at Cimuntur sub watershed. Then the results of the rainfall
erosivity is plotting to the map using ArcGIS software.

Figure 2. Rainfall Erosivity at Cimuntur Sub Watershed.

2. Calculating Soil Erodibility Factor

Different soil types at Cimuntur sub watershed involved different
absorbing water or different permeability that have K are different. K values to
each types of soil at Cimuntur sub watershed can be seen in Table 5.
Table 5. Soil Erodibility (K) Values for Each Soil Types at Cimuntur
Percentage of
No. Type of Soil Area (ha) K
the Area (%)
1. Aluvial 6,632.42 0.200 11.27
2. Latosol 6,045.59 0.175 10.27
3. Organosol 22,765.74 0.200 38.68
4. Podsolik 23,411.10 0.150 39.78
Total 58,854.85 100.00
(Source : Balai Besar Wilayah Sungai Citanduy 2011)

V - 101
 Figure 3. Soil Types and K Values Cimuntur Sub Watershed.
3. Calculating Slope and Length Slope Factor

By using ArcGIS software, it can be described Cimuntur sub watershed

slope map in Figure 4.

Figure 4. Slope Map Cimuntur Watershed

4. Calculating Land Cover Factor

Sub-watershed land use at Cimuntur grouped into several groups among
land use bodies of water, dry land forests, plantations, settlements, plantations,
rice fields, and shrubs. Here is a land use and values of the C's at Cimuntur sub
watershed can be seen in Table 6.

V - 102
 Tabel 6. Tata Guna Lahan dan Nilai C
Percentage of
No. Type of Land Use Area (ha) C
Area (%)
1. Water Body 50.53 0.100 0.09
Forrest at Dry
2. 37,667.38 0.005 64.00
land
3. Forrest 4,879.93 0.100 8.29
4. Housing 2,513.26 0.800 4.27
5. Yard 1,237.09 0.500 2.10
6. Paddy Field 9,116.66 0.500 15.49
7. Bush 3,390.00 0.300 5.76
Total 58,854.85 100.00
(Source : Balai Besar Wilayah Sungai Citanduy 2011)

Figure 5. Landuse and C Values Cimuntur Sub Watershed.

5. Calculating Factors Practical Conservation

Existing practical conservation is 1. It is assumed there had never
conservation efforts.
Using software Arc GIS, overlay the five of map so that it will get the value of
soil erosion that occurs in each area.

V - 103
 Figure 6. Land Erosion Map Cimuntur Sub Watershed

Overlay the results obtained from the average soil erosion that occurs in the sub-
watershed Cimuntur (Ea) amounted to 963.747 tonnes / ha / year and sediment
transport in the sub watershed Cimuntur is equal to 14.649 tons / ha / yr. It is
considered a very large erosion. Therefore, conservation efforts are needed to
repair damaged lands and to minimize erosion and sedimentation. Conservation
efforts can be divided into 2 types of vegetative conservation and conservation
mechanically.
1. Vegetative Conservation
Vegetative conservation efforts carried out by means of practical
conservation is to improve land management in Sub-watershed Cimuntur.
Table 7. Type of Conservation And the P Values Used in the Study
No. P Value P Value
Type of Conservation (existing) (Conserved)
1. Stripe planting with contour 1 0.200
2. Grass Stripe Planting 1 0.400
3. Without Conservation 1 1.000
4. Terraces with good condition 1 0.040
Using Arc GIS software, the maps obtained practical conservation at
Cimuntur sub watershed.

V - 104
 Figure 7. Practical Conservation and P Values Cimuntur Watershed
Because of improvements in its management, the value of C from land-
use changes, as well. It can be seen the type of plant and its C values in Table 8.

Table 8. Land Use Types After Conservation and C Values

No. Type of Land use C.existing C.conservation
1. Water Body 0.100 0.100
2. Forrest at Dry land 0.005 0.001
3. Forrest 0.100 0.001
4. Housing 0.800 0.100
5. Yard 0.500 0.100
6. Paddy Field 0.500 0.010
7. Bush 0.300 0.001
By using Arc GIS software, five new maps overlays that will get the
value of land erosion after conservation.

V - 105
 Figure 8. Land Erosion After Conservation Cimuntur Sub Watershed Map
After conservation, the average soil erosion that occurs in Cimuntur sub
watershed is decreased to 237.415 tons/ha/yr and sediment transport by 3.61
tons/ha/yr. The result of combined Vegetative conservation in reducing the
erosion and sedimentation is shown in Table 9.

Table 9. Comparison on the Influence of Conservation

Percentage
Area After
Existing Percentage of Area Change After
Practiced
No. Erosion Class Area of existing After Conservation
Conservation
(km2) Area (%) Conservatio (%)
(km2)
n (%)
1 Tolerable 106.558 336.366 18.105 57.152 39.05
2 Light 276.091 127.829 46.910 21.719 -25.19
3 Average 3.817 66.0498 0.649 11.222 10.57
4 Crtical 57.453 28.885 9.762 4.908 -4.85
5 Very Critical 144.63 29.4195 24.574 4.999 -19.58
Total 588.549 588.549 100.000 100.000

2. Physical Conservation
To handle the transported sediment, it is necessary to Technically handle by
constructing check-dam for sediment control. To achieve the conservation and to
control the sedimentation, it needs minimum 34 check dams with minimum
capacity of 4000 m3. The check dam locations can be seen in Figure 9.

Figure 9. Check Dam Location Map

CONCLUSSION
The conclusions of this study are:
1. Cimuntur sub watershed is one of Citanduy watershed that has large erosion and
sedimentation.

V - 106
2. From the analysis of erosion and sedimentation obtained turns out that the average
soil erosion loss at Cimuntur sub watershed is 963.747 tons/ha/year and sediment
transport 14.649 tons/ha/yr. After some vegetative conservation by doing practical
conservation, the erosion and sedimentation in Cimuntur sub watershed is decreased
to 237.415 tons/ha/yr and the sediment transport 3.61 tons/ha/yr.
3. To decrease the rate of sedimentation further downstream, it needs technical
conservation by constructing sediment control such as check dams at the upstream
steep sites prone to erosion and at the erosion rate at the category of large and very
large. The amount of the check dam needs to be built is 34 with a minimum capacity
of 4000 m3
4. The reduction on sheet erosion at watershed must be comprehensively and
consistently performed at appropriate pace and with intents participation calls from
societies and stakeholders. Meanwhile, the technical measures such as construction
check dams are also very crucial and important to prevent further river bed
degradation and erosion.
5. The land conservation upstream at Cimuntur sub-watershed as well as at Citanduy
watershed is very urgent to reduce sedimentation at Segara Anakan Lagoon.

REFERENCES
Kamiana, I Made. 2011. Teknik Perhitungan Debit Rencana Bangunan Air. Yogyakarta :
Graha Ilmu;
Kuntho, D. U. dan Ajeng A., 2013. Perencanaan Konservasi Erosi-Sedimentasi di Sub-
DAS Cimuntur, Ciamis.Tugas Akhir di Jurusan Teknik Sipil, FT Undip.
Soewarno. 1995.; Sosrodarsono,S. dan Tominaga, M. 1985. Perbaikan dan Pengaturan
Sungai. Terjemahan oleh Gayo, M.Y. Jakarta : Pradnya Paramita;
Suresh, R. 2000. Soil And Water Conservation Engineering. Delhi : A.K. Jain;
Suripin.1998. The Effects of Land Use Alteration And Conservation Measures On
Sediment Yield With Reference To Reservoirs In Tropical Areas.;
Suripin, 2002. Pelestarian Sumber Daya Tanah dan Air. Yogyakarta : Andi;
Suripin. 2004. Konservasi Lahan.; Suripin. 2004.;
Wesley, LD.1977.Mekanika Tanah. Jakarta : Badan Penerbit Pekerjaan Umum;
________, 1983. Design of Sabo Facilities. JICA;
________, 1985. Perencanaan Bangunan Pengendali Sedimen. JICA

V - 107
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

THE PERCENTAGE OF REDUCTION IN EROSION POTENTIAL

UUSINGUSING HORIZONTAL
USING HORIZONTAL
DRAINAGE DRAINAGE
HORIZONTAL DRAINAGEaper Title Line 1 (14pt)
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
Akhmad Azis fffffjfjjfkkfjjj
Civil Engineering, Polytechnic State Of Ujung Pandang, Makassar, Indonesia
Akhmad_azis @yahoo.com

Hamzah Yusuf
Civil Engineering, Polytechnic State Of Ujung Pandang, Makassar, Indonesia

ABSTRACT:
Erosion that occurs in nature to cause damage to the land, not only upstream but also downstream.
Cause ground damage downstream sedimentation in the watershed which would reduce the function
and life of the dam. This study aims to determine the percentage of reduction in erosion potential using
horizontal drainage on sand clay. The study was conducted in the laboratory using a raifall simulator
by holding variation of rain intensity of 45 mm / hr, 65 mm / hr and 85 mm / hr, while the slope is 5o,
15o and 30o. The amount of drainage that is 1 pieces, 2 pieces and 3 pieces using bonding fibers. The
results show, when using drainage as much as 1 pieces a decline of 16.59%, 2 pieces drainage for
drainage 33,98% and 53,3% for 3 pieces. The results showed the more horizontal drainage, the
greater the percentage decline so it is feasible to use to address the problem of erosion that would
threaten the sustainability of the dam.

Keywords: erosion, dam, sand clay, horizontal drainage.

I. Introduction
Erosion that occurs in nature to cause damage to the land , not only in the place of erosion
( usually in the upstream ) , but also in the downstream. Upstream land degradation, among
others, the loss of a surface layer of soil nutrients led to decreased crop quality and reduced
productivity of dry land . While damage in the downstream as a result of receiving the results
of erosion in the form chemicals sediment in the soil causing decreased water quality and
sedimentation. When watersheds received sediment submissions will reduce the functionality
and life of reservoirs and river morphology. Erosion not only in Indonesia but also in various

V - 108
countries such as the United States, so the Land Research Institute in the country actively
doing research on soil erosion that began in the 1920s until today where it has generated a
basic knowledge of erosion and parameters affect. Asdak chay (1995 ).
The relationship between the magnitude of the degree of slope is a parameter that is
investigated by Zings in 1940 where it has been demonstrated that the greater the angle of
slope, erosion is increasing. Meanwhile, according to Musgrave, 1947, stated that there is a
relationship between rainfall characteristics and the amount of eroded soil ( Kusumastuti.I.D,
1994).
Some conservation methods developed so far in order to prevent / reduce the magnitude of the
rate of erosion on a field , broadly divided into three ways: agronomically, mechanically and
chemically. Conservation agronomically or vegetatively by using plants or plant and crop
residues before reaching the ground. Mechanically conservation efforts are carried out above
ground level with a view to reducing the water flowing on the surface of the ground , whether
it is to slow down the speed of the flow, to accommodate the flow, increase the power of water
infiltration into the soil and others. While chemically conservation efforts are being made to
improve the soil structure by giving chemical materials. ( Suripin, 2011).

2. LITERATURE REVIEW
2.1. Erosion
Erosion is an event missing of soil or parts of soil are transported from one place to another
place by the force of the wind , rain or snow flow ( Suharyadi, 2006). Meanwhile, according
Suripin ( 2001 ), in general, erosion and sedimentation is a detached process of soil particle
and transported from a place to another place by water or wind movement, followed by
deposition of material.
The process begins from detached of soil particle due to the rain fall from the atmosphere
which pound the ground surface of causes soil particles peeled. If the soil is saturated water ,
there will be a runoff carrying soil particles that have separated them. On the soil surface
rather flat, the soil particles are evenly distributed in all directions, but at ground level sloping,
there is the sedimen transport, from of the slope leading down to the river basin
(reservoirs and rivers) that may cause sedimentation.

2.2. Determinants of Erosion

Erosion process that occurs in nature does not only occur because of the sensitivity factors of
rain and soil, but also influenced by vegetation, slope and humans so by Utomo (1994),
erosion is expressed in the formula :
E = f (i,r,v,t,h) (1)
Where : E = erosion
i = climate
t = topography
v = vegetation
h = human

V - 109
1. Climate
The main factors affecting erosion in tropical regions such as Indonesia is raining. Some
influence on the rain erosion, among others :
a. Rain Intensity :
The intensity of rain is short- term rainfall intensity expressed in per - hour. According
Sosrodarsono. S ( 1987), the rainfall varies depending on the duration of rainfall, the
geographical location , the frequency of occurrence and others. Arsyad (1989 ) in Hardiyatmo
(2006 ) classify rain as shown in Table 1.
Table 1 . Classification of rainfall intensity

Rainfall Intensity (mm/hr) Clasification

0-5 Very low

5-10 Low
11-25 Moderate
26-50 Somewhat higher
51-75 High
 75 Very high

In this study, rainfall intensity used is artificial rain produced by rainfall simulation tool
(Rainfall Simulator).
The formula used to calculate the rainfall intensity of rainfall simulation apparatus according
Kusumastuti, et al (1994 ) are as follows :
Q
I= x 600 (2)
A.t
Where : Q = volume of rain water in the container settings ( ml )
A = surface area of the container ( cm2 )
t = time ( min )
I = Intensity ( mm/h )

b. Number of rain
The amount of rain that likely does not cause severe erosion when the intensity is low, so is
the high intensity in a short time will not cause substantial erosion ( Kusumastuti , et al , 1994)
c . The size and shape of a grain of rain.
Grains of rain have different sizes , ranging from grains slightly larger than the maximum
grain diameter of fog until about 7 mm. Most of the rain that falls grain size of 1 to 4 mm.
Kowal and Kassam (1976 ) found that the average grain diameter of rain in the tropics is about
3 mm and a diameter of 4.4 mm. The relationship between the velocity of raindrops falling to

V - 110
the size of a grain of rain is falling velocity of raindrops increases with increasing grain
diameter of rain ( Kusumastuti, et al, 1994).
d . Grains fall speed of rain
Speed grain rain fall increases with the size of a grain of rain. According to Schwab, et al
speed of falling grains of rain influenced the destruction of the land and cause erosion
( Kusumastuti, et al, 1994)
e . Kinetic energy of rain .
Kinetic energy is the energy that rain occurs at the time of the fall of grains of rain from the
atmosphere onto the surface of the ground. The kinetic energy of the rain is the principal cause
of the destruction of soil grains (Kusumastuti, et al, 1994) .
By Anonymous (1992 ) The kinetic energy of rain can be calculated with the basic formula :
Ek = 0,5 m. v2 (3)
If rainfall intensity is known, then the energy can be calculated by the formula :
Ek = 11,87 + 8,73 log I (4)
Meanwhile, according to Hudson (1965 ), to the tropics can use the formula :
Ek = 29,8 – 127,5 / I (5)
2. Topography
In this topographic factors there are two things that can cause erosion in the soil surfaces
namely slope and length of slope factor because both of these are very influential on the the
flow velocity of soil surface and volume of water flow. If the slope is getting steeper will
increase runoff, while the slope more lenght, volume of water flowing to be large. Index
factors of topographic is L and S, each representing the influence of the length and slope
against of the magnitude of erosion ( Asdak chay, 1995) .
Slope length factor ( L ) is defined mathematically as follows :
L = (1/22,21)m (5)

Where : L = length of slope ( m )

M = number exponent which varies dependent slope
namely : 0.5 for slopes above 5 %
0.4 for slopes from 3.5 to 4.9 %
0.3 for slopes less than 3.5 %
3. Vegetation Cover Land
Vegetation cover affects the amount of soil erosion due to :
a. Protect the soil surface from the impact of rain water
b . Lowering runoff
c . Restrain soil particles
d . Increase the absorption of water into the soil
4. Human
Humans have a role in accelerating and reducing the rate of erosion. The rate of erosion can be
accelerated when humans exploit nature by way of logging , farming the wrong way and so
on.

V - 111
2.3. Horizontal drainage
The method of making the placement of horizontal drainage in the direction of the contour
lines and transverse direction of the slope, is intended to shorten the length of the slope and
increase power of infiltration so that the volume of water and surface runoff can be minimized
With reduced water volume and flow rate, erosion is expected can be controlled .
According to Gray DH et al (1982 ) Horizontal drainage function in controlling erosion is
1 . Stabilize the surface layer of the soil slope against wind, gravity, and collisions water.
Stability is caused by the effect of fibers bonding grown in drainage.
2 . Energy release obtained by slowing the movement of water. The length of the slope to be
several short slopes carrying soil material down slope with slowly to each of the contour of
horizontal drainage.
3 . Capture sediment that occurs due to the embankment as well as bonding fibers that filter
out any material that passes.
4 . Increase the infiltration because surface runoff is running slow due to the presence of
embankment from bonding fibers on the horizontal drainage.

Gambar 1. Horizontal Drainage

(Gray D.H dkk, 1982)
In this method there are three main components to be prepared that is making trenches,
bonding fibers, and wood stakes.
1. Making trench.
Trench made in the transverse direction of the slope, where the use of human labor to dig. ½
trench shape is a circle. Made fairly shallow depth of which is equal to 20 cm. Early
excavations began on the very bottom of the slope to the top of the slope where the distance of
the trench with the other trench of 1-4 m.
2. Bonding fibers
Fibers that exist in the universe is made of palm trees, where it has the properties of a
permeable material and is also very good to be used as a filter. Fibers tied into a single unit
with a diameter of 40 cm so that the party of the fibers bond will appear on the surface slope,
shaped like a dike that would put a halt to the flow surface.
3. Wooden stakes

V - 112
Wooden stakes to be used on horizontal drainage is intended to refute the fibers bond to be
stronger position. Stakes can be made from several types of wood. The length of wooden
stakes is 60 cm where some stakes appear above of the slope surface for refute the bonding
fibers that appear on the surface of the soil.

3. METHODS
To carry out the research, then do the following steps:
a. Preparation and examination of the ground.
The soil used in this study first examined to determine the type of soil.
b. Rainfall intensity measurements, to determine the rainfall using rainfall simulator.
c. Model spaced drainage
The study consisted of two models, namely the condition of the ground surface without
drainage and the use of horizontal drainage. To detail can be seen in Figure 2.

Soil
Soil

fiber fiber

a. Drainage space L b. Drainage space 1/2L c. Drainage space 1/3L

Figure 2. Drainage space model
In conditions of use drainage consists of three variations, namely :
- Spacing L drainage. In this space drainage amount used is 1 piece fiber.
- Spacing drainage 1/2L. In this space drainage amount used is 2 pieces fiber.
- Spacing drainage 1/3L. In this space drainage amount used is 3 pieces fiber.
d . Collected of data
Data were collected by means of observation (Observational) directly.

4 . RESULTS AND DISCUSSION

4.1. Result
1. Testing of Soil
Based on the results of the soil as the table above and the appropriate attachment USC soil
classification system , obtained is soil type sand clay = SC because the percentage of 37.33 %
fine fraction ( > 12 % ) and 12.66 % plasticity index ( > 7 % )

2. Measurement of Potential Erosion

Potential erosion measured in this study is that erosion occurs when water-saturated soil which
began on the condition of surface runoff and infiltration of rain water into the soil is constant
(maximum). Potential measurements performed on the soil surface erosion without horizontal

V - 113
drainage, drainage ratio using L (1 drainage), 1/2L (2 drainage), and 1/3L (3 drainage) can be
seen in Table 2, 3 and 4.

Table 2. Potential erosion at rainfall intensity of 45 mm/hour

Erosion Potensial

Slope (0) Horizontal Drainage Ratio

Zero L 1/2L 1/3L
Drainage (1 pieces) (2 pieces) (3 pieces)
5 0,3 0,2 0,1 0

15 0,9 0,8 0,7 0,3

30 5,3 4,2 3,8 2,3

Figure 3. The relationship between the potential erosion with slope at I = 45 mm/jam

Table 3. Potential erosion at rainfall intensity of 65 mm/jam

Erosion Potensial

Slope (0) Horizontal Drainage Ratio

Zero L 1/2L 1/3L
Drainage (1 pieces) (2 pieces) (3 pieces)
5 0,8 0,5 0,2 0,1

15 2 1,5 1,2 0,6

30 7,8 6,7 5,3 4,7

V - 114
 Figure 4. The relationship between the potential erosion with slope at I = 65 mm/jam

Table 4. Potential erosion at rainfall intensity of 85 mm/jam

Erosion Potensial

Slope (0) Horizontal Drainage Ratio

Zero L 1/2L 1/3L
Drainage (1 pieces) (2 pieces) (3 pieces)
5 1,05 0,8 0,7 0,2

15 3,2 2,3 2 1,4

30 9,7 8,9 6,5 4,9

Figure 5. The relationship between the potential erosion with slope at I = 85 mm/jam

4.2. Discussion
From tables and figures the above, it is seen that the magnitude of the potential erosion is
influenced by the presence of horizontal drainage. Average potential erosion that occurs on the

V - 115
surface condition without drainage of 3.45 gr, the ratio of space L amounted to 2.88 gr, the
ratio of 1/2L space amounted to 2.28 gr and at the ratio of 1/3L spaces amounted to 1.61 gr
or a decline of 16.59 % when using the ratio of spacing L, and 33.98 % for 1/2L space ratio,
and 53.3 % in the ratio of space 1/3L.
From the calculation of the percentage reduction in the potential for erosion is seen that the
use of spacing ratio 1/3L is more effective than the ratio 1/2L and L. This is due to the greater
number of horizontal drainage , the slope of the shorter length so that the smaller the surface
runoff , increased infiltration capacity , and reduce the potential for erosion. This is according
to the equation 1 that states the amount of erosion is directly proportional to the length of the
slope. The shorter length of the slope, the less erosion.
Based on the results of the study showed that the higher the rainfall intensity, the higher the
potential for erosion . This suggests that high intensity rainfall can transport soil particle size
greater than the rainfall intensity is lower , so that the resulting number of potential erosion
will be greater . In addition to the higher intensity of rainfall can cause increased runoff
volume so that the power dorongnyapun increased. Thus the amount of soil particles
transported will increase as well . From Table 2,3, and 4, visible surface soils has given
drainage harizontal spacing ratio L, 1/2L, and 1/3L, permanent erosion potential increases
with increasing intensity of rain even though fewer in number compared with no use of
horizontal drainage. Thus it can be said the intensity of rainfall directly proportional to erosion
potential.
Similarly, tilt the slope, which looks increasingly steep a slope, then the potential erosion
caused greater. This is due to the steep slopes of the soil particles will be more easily
transported because the drag force becomes larger. The results of the study showed an increase
in the potential erosion of the average slope of 50 to 300 amounted 36.19 times. For
comparison , the results of research conducted by Martono (2004 ), an increase in the potential
for erosion average from tilt slope of 2.50 to 300 amounted 47.51 times.

5 . CONCLUSIONS
From the discussion of the results of the research can be given some conclusions as follows :
1 . Rainfall intensity and slope influence on erosion potential . At intensity of rainfall fixed,
slope the greater, the erosion potential is enlarged. Similarly, if the slope remains, rainfall
intensity magnified the potential for erosion is also greater, so it can be said that the intensity
of rainfall and slope is directly proportional to the potential for erosion.
2 . The use of horizontal drainage, well spaced ratios L, 1/2L and 1/3L greatly affect the
potential for erosion. The more drainage installed diminishing the potential for erosion.
Effective use of horizontal drainage that is , if use a space L ratio decreased by 16.59 %, the
ratio of space 1/2L amounted to 33.98 %, and the ratio of 53.3 % 1/3L spaces.

6. RECOMMENDATIONS
1 . Based on the results of this study, to reduce erosion is suggested that the ground surface
have a higher slope and high intensity of rainfall should use spaces ratios more larger.
2 . The results of this study should be continued with directly in the field of research, because
the research laboratory variables affecting erosion is very limited.

V - 116
REFERENCES
1. Barker. D.H., 1991, Geotextile - Reinforced Vegetated Barriers, Geotecnical fabrics
Report pp. 16-21.
2. Kusumastuti. D . I. , 1994, Physical Model Synthesis Materials Utilization In The Land of
Sand Slope, PAU Engineering UGM.
3. Hermawan. Y., Ir, 1984, Hydrology for Engineers, Erlangga Publisher, Jakarta.
4. Instruction Manual Rainfall Simulator, January 1992, Armfield Ltd., Hampsire.
5. Kirkby. M. J., and Morgan .R.P.C., 1980, Soil Erosion, John Wiley and Sons, Ltd.
6. Robert. M. K., Designing With Geosynthetics, 1990, Prentice Hall.
7. Liong, G,T., 1990, Geosynthetic and Its Application in Civil Engineering, seminars
Lustrum ITB, Bandung.
8. Martono., 2004. Effect of Rainfall Intensity and Slopes on Soil Loss rate on Regosol
Grey. Thesis. Semarang : Diponegoro University.
9. Soemarto.C.D., 1995 , Engineering Hydrology , publisher Jakarta.
10. Sosrodarsono.S.,1987, Hydrology for irrigation, PT. Pradnya Paramita .
11. Suryolelono. K.B., 2000, Geosynthetic, Publisher Naviri, Yogyakarta.
12. Utomo , Wani. H., 1994, Soil Erosion and Conservation, Malang : Teachers' Training
College Publisher, Malang.

V - 117
 IMPLEMENTATION CONCEPT OF BIO-LANDSCAPE
MANAGEMENT THROUGH DEVELOPMENT OF BIO-VILLAGE
AT THE UPPER WATHERSHED OF DAM 1

Sahid Susanto
Professor at the Faculty of Agricultural Technology, Gadjah Mada University
[email protected]

Sigid Santoso
Chairman of Division of Irrigation System Development
Main Office of Serayu-Opak River Basin Development, Yogyakarta, Indonesia
[email protected]

ABSTRACT

In order to restore degraded upper watershed of dam, bio-landscape management concept

were used to develop implementation strategy to ameliorate the degradation. The concept
basically oriented to design an optimal land use pattern at a watershed. The concept was
applied at some upper watersheds of dam in Cenral Java and worked well. In this study, the
aim of study was oriented to build strategy for implementing the concept through developing
bio-village approach. The upper watershed of Sempor Dam system located at Serayu-Opak
River Basin, Central Java was used as a study. Bio-landscape of the watershed of dam is
dominated by open agricultural cultivation.

Developing strategy was started with socio-economic analysis at the watershed to find the best
probability in applying various land use at the watershed for producing agricultural raw
materials. Then, development bio-village model was proposed at the upper watershed of dam.
Projection of the development was focused on improving quality of environment management
economic prosperity of people living at the upper watershed. Empowerment for the farmer
society in the bio-village model to be an agent of various functions was needed. In order to
integrate social, economic and environmental aspect between the upper watershed and the
lower of the dam, three sub-district of the lower part of the Dam, namely Kebumen, Purworejo
and Kutoharjo were plotted as central economic growth. The excising small and medium scale
of agricultural product processing business in these sub-districts was plotted as the receiver of
the agricultural raw materials from the upper watershed. Policy actions from the angle of
political economy were discussed.

Keywords: bio-village, upper-lower area of dam, social, economic and environment

1 This paper is to be presented at the 82th Annual Meeting of International Commissions on Large Dams
(ICOLD) on June 1-6, 2014 at Bali Nusa Dua Conference Center, Bali

V - 118
 1. INTRODUCTION

The present environmental of watershed in Indonesia, particularly Java, are degrading. Bio-
landscape management concept was proposed to develop implementation strategy to
ameliorate the degradation (Emin and Glenwood, 1996). Bio-landscape management concept
is basically oriented to design an optimal land use pattern at a watershed (Susanto, 2013).
Using developing some realistic scenarios, the optimal land use pattern can be found from
manipulation of permanent land cover index and production land cover index. A quantitative
assessment was used for evaluating the scenarios. The concept was applied at some upper
watersheds of dam at Central Java and worked well.

The aim of study was directed to build strategy for implementing the concept in the upper
watershed of dam through developing bio-village approach.

2. METHODOLOGY

In developing strategy for implementing the concept of bio-landscape management at the

upper watershed of dam, a bio-village approach was applied. The bio-village was developed as
a centre of agricultural raw materials production. Developing strategy was started with socio-
economic analysis at the watershed to find the best probability in applying various land use at
the watershed for producing agricultural raw materials. Rapid Rural Appraisal (RRA) and
Participatory Rural Appraisal (PRA) were used in this analysis. Projection of the development
was focused on improving quality of environment management economic prosperity of people
living at the upper watershed (Martín, et all, 2005; Märiten and Therese, 2012) (see Figure.
1). Principle of watershed management with land use pattern formatted in the form of quadrant
based landscaping was considered in applying the bio-village development (Figure 2).

A qualitative analysis of integrating social, economic and environmental aspect between the
upper watershed and the lower area of dam was applied to build bio-village model in the upper
watershed of dam. Consideration for developing toward an empowerment and policy action
were applied for analysis.

3. RESULT AND DISCUSSION

3.1. Bio-village model

The Sempor Dam system located at Serayu-Opak River Basin, Central Java was used as a
study. Total area of the upper watershed of dam is 43.27 km2. Figure 3 shows that bio-
landscape of this watershed is dominated mostly by many types of agricultural upland crops
cultivation (90%), settlement (5%) and others (5%). This area is densely populated with

V - 119
agricultural land owner is less than 0.25 ha per family. Subsistent farming with upland crops
such as corn, peanut, coconut sugar are main products of this area.

Present DEVELOPMENT OF BIO-VILLAGE APPROACH Projection

management and social-economic condition

Human Resources Natural Resources Human Life Index

The Present Status of natural resources

Quality of environment management

Development Restoration and
Management

economic prosperity
Improving Economic Growth Through Integration of Agricultural

BIO-VILLAGE
Cutivation and Post Harvest Technology

Rural Infrastructure Application of Soil and Central of Bio-

Development water conservation mass Production
technology

Figure. 1. Flow of thought for developing Bio-Village at the

upper watershed of Sempor Dam.

PRODUCTION SERVICE
Interrelationship between LOCATION LOCATION
upper and lower area of
watershed UPPER AREA LOWER AREA
VEGETATION MARKET
TYPE I II

SERVICES NON-MARKET
TYPE III IV

Quadran I : food crops, vegetable crops, grass and livestock

Quadran II : drinking water, irrigation, flood and sedimentation control
Quadran III : wild life protection, aquatic ecosystem protection, land
slide control, natural vegetation preservation
Quadran IV : lower area ecosystem protection, water quaity of
recreation area, aestetic
Figure 2. Watershed management approach
Figure 2. Watershed management approach

Surveying results from the aspect of social-economy-environment gave the anatomy of

poverty cycles and its relationship with land degradation, see Figure 4. Their income from

V - 120
agricultural production is very limited, and it is only enough for survival. It is typically a
traditional subsistence farming society.

In general, this reality happens in

almost upper watershed, espe-
cially in Java Island. About 40%
of total population of Indonesia
(around 230 million in 2012) is
poor and 17% on it lives at upper
watershed.

Results from socio-economic

analysis at the upper watershed
showed that the agricultural
society was very dependent to
agricultural land. In imple-
menting strategy of the concept
of bio-landscape, the principles
of watershed management was
applied to develop longterm
integrated program of the bio-
Figure 3. Existing condition of bio-landscape of the
upper watershed of Sempor Dam village.

Anatomy of poverty cycles and

land degradation
Land
degradation The implementing strategy was
Conventional
agricultural directed in two areas: a) at river,
production PENURUNAN FUNGSI
EKOSISTEM which was oriented in
Land Poor acute constructing sediment controls,
pressure and (b) at agricultural land,
UPPER
WATERSHED Low level of skill which was focused on
Land
scarcity OF SEMPOR and education introducing cultivation with
DAM principles of soil and water
Low Weak conservation.
income competition
Small land
Difficulties of
Some small scale programs of
Agricultural
cultivation
work access bio-villages were selected to be
as a village model. Production
incentives-based policy was
Figure 4. Anatomy of poverty cycles and applied in the village model.
its relationship with land degradation at the upper
watershed of Sempor Dam 4

V - 121
 Integrating socio-economic and environmental development In order to integrate of social,
between the upper watershed and the lower part of economic and environment
Sempor Dam between upper watershed and lower
area of the dam, three sub-districts
Central Economic Growth
of Kebumen Sub District
of the lower part of the Dam,
namely Kebumen, Purworejo and
Kutoharjo were plotted as a central
economic growth (Figure. 5).
BIO-VILLAGE
MODEL
There are some excising mostly
UPPER WATERSHED OF SEMPOR DAM small and conventional scale food
BIO-VILLAGE BIO-VILLAGE
industries in these sub-districts.
MODEL MODEL However, only some agricultural
raw material of upland crop
produced from the upper watershed
can be absorbed for these
industries. Strengthening the small
scale food industries to become
absorbed agent of raw material
Figure 5. Integrating socio-economic-environment from the upper watershed of the
between the upper watershed and lower area dam is needed.
of Sempor Dam

3.2. Toward an empowerment

As mentioned above that typical characteristic of society in the upper watershed is subsistence
farming with economically is very weak. If the situation is linked to the national economic
development, high economic growth achieved by government of Indonesia does not give
assurance for the society in adequate to generate the surpluses necessarily to address poverty
comprehensively and sustainability on the sole basis of transfers and social welfare programs.
It is therefore indispensable to generate a structural attack on poverty with a strategy that
assists the poor in developing a self-help capacity, through mobilizing their own productive
potential.

In the context of empowerment, the three main tenets of the International Fund for
Agricultural Development -- IFAD -- paradigm may be suitable to be adopted (Alamgir and
Elhaut, 1994). Those are:

i. The poor need to be approached as small producers, contributing to sustainability

growth

V - 122
 ii. Therefore their access to factors of production, productive land, capital and labour
markets must be secured
iii. This is only effective if a production incentives-based policy framework is in place,
enforce by participatory institution

Considering to the paradigm, the following empowerment for the farmer society must be put
as part of implementation strategy in developing bio-village model to realize the integrating
social, economic and environmental aspect between the upper watershed and the lower area.

a) Farmer society as a producer agent

The society as mentioned above however can be considered as a social capital. This traditional
farmer society must be empowered. The empowerment must be directed to make the society
become a producer agent. Economic growth policies therefore must consider the producer
agent developed from the society as a small holder in agricultural sector, and micro
entrepreneur. High quality growth depends on the efficiency with which the decision-makers
mobilize the production potential of small producers. The active participation of the poor in
this growth process, as producers and not just as consumers or recipients of social services,
will make growth sustainable and turn into authentic socio-economic development.

b) Farmer society as an agent to access to resource markets

In order to assure the poor farmer to access to investment and working capital, within the
sustainable framework for financial intermediation must also be assured. Local resources,
which are generally low cost, are an ideal basis for credit programs. External resources best
fulfil the role of a growth supplement to local resources, enhancing the investment,
production, income savings and reinvestment cycle. In this context, the traditional society
can play role as a credit institution and also as an agent of marketing agricultural product, as
well. Some modification of the concept of Village Unit Cooperation (Ind. KUD) seems
reasonable to be applied. Farmers as member of the KUD are motivated to become
shareholder of a saving and credit institution enhances the sustainability of that institution.
Another consideration is protection the poor from such market inefficiencies, efforts must be
made to enhance the bargaining power in determining agricultural products.

c) Farmer society as an agent to access to labour markets

Farmers have some idle time during cultivating their agricultural lands. Therefore, having part
timer job for them is important to increase their income from outside of agricultural work.
However, there is a need of information access to inter to the labour market in order to bring
them into the process of growth. The farmers in this upper watershed of the dam must be
supported to inter to the small scale food industries system at the lower area of the dam is
possible to be as a good model. A comprehensive upgrading skill that includes an effective
entrepreneurship, access to credit, management know-how and appropriate technology is

V - 123
important to enable for the farmers to better access to labour markets. In this context, the KUD
must be improved to become a professional agent for labour markets.

d) Farmer society as an agent for developing indigenous ideas

There is a fact that the farmer society is economically poor, limited skill in planning and
articulation, as well as their supposed limited insight hoe their daily lives interacted with wider
environment. However, they have thousands of indigenous knowledge based on their
experience in living with nature and agricultural cultivation. Development of bio-village
model is closely with the dynamic of nature. It is reasonable that the development should be
designed with the active participation of the beneficiaries taking into consideration their
perception of their own local needs and possibilities. It is necessary, therefore, to look at the
ideas and expectations of the people concerned with regard to development.

The farmer society can be used as a grass root agent to understand and adopt the farmer's
indigenous knowledge and idea effectively and then used them to develop policy action in
various levels.

3.3. Development policy action

There is the fact that the farmer in the upper watershed of dam is a group of poor population.
With this condition any development introduced for them seems to be unchanged. The key
word is political economy. Therefore the development should be started from the political
economic angle. Considering to this angle, Fig. 6 shows the two factors must be considered in
developing the traditional farmer society to become various agents as mentioned above. Those
are sufficient conditions and necessary conditions.

In order to play as various agents, the institution needs to be given an appropriate set of market
signal and incentives from government. Those are:
(i) balanced financial allocation of development sectors,
(ii) regional and national policy reform, and
(iii)sufficient profit margin. For raising the probabilities to success, the sufficient
conditions should also be met:
a. application of appropriate irrigated agricultural technology,
b. integrated planning approach with farmers as a subject of development using
culturally-based consideration, and
c. targeted planning and implementation program with full participation of farmer
and local community.

There is a need, therefore, for a strong government policy in developing traditional farmer
society comprehensively in the form of a national plan of political economy, with clear
priorities in term of manpower and financial resources.

V - 124
In order to protect the poor from such market inefficiencies, efforts must be made to:
(i) enhance their bargaining power and productivity,
(ii) make transparent and reliable economic and agricultural policies for the farmer society,
and the huge population of farmers as a whole,
(iii)create favourable climate for investment,
(iv)the agricultural price-cost ratio must be arranged in such way that investment pay for
themselves, risks are covered and sufficient income is earned,
(v) set up marketing facilities.

With this context, agricultural private enterprise in various levels should be built as a starting
point to develop a basic condition.

National monetary International

reform monetary reform

Balanced financial Regional and national Sufficient profit

allocation of policy reform margin of irrigated
development sectors agriculture

NECESSARY CONDITIONS

TRADITIONAL FARMER SOCIETY

AT THE UPPER WATERSHED OF DAM

SUFFICIENT CONDITIONS

Application of Integrated planning Targeted planning and

appropriate upland approach with farmer as implementation program with
agriculture technology a subject of development full participation of farmer and
local community

Culturally based soil and water

resources management for
upland agriculture

Figure 6. TowardFig.solution
6. Toward ofabio-village
solution of model of
bio-village model
the upperatwatershed of Sempor Dam
the upper watershed of dam

V - 125
IV. CONCLUSION AND REMARK IMPLICATION

1. The upper watershed of Sempor Dam has degrading significantly. This tendency is also
happened in almost upper watersheds of Dam at Java. From the aspect of social-economy-
environment gives the anatomy of poverty cycles and its relationship with land
degradation. The problem is complex and convoluted, being bound in many facets of the
problem situation. Using the bio-landscape management concept, implementing strategy
is built and development of bio-village model is proposed.

2. In order to realize the bio-village model, the empowerment of the existing traditional
farmer society has to be strong considered to be placed as various agents. Those are: (i) as
a producer agent, (ii) as an agent to access to resource markets, (iii) as an agent to access
to labor markets, and (iv) as an agent for developing indigenous ideas.

3. Integration of social, economic and environment between upper watershed of dam and
lower area of dam was proposed. The upper watershed was directed to be as up-land
agricultural raw material production and the lower area oriented to be as a center of
economic growth with small scale food industries as absorbed agents.

4. The strategy has implication to the government to build a policy action from the angle of
political economy.

ACKNOWLEDMENT

This paper could not be realized without financial support from the Office of Serayu-Opak
River Basin Development, Yogyakarta (2005) in cooperation with Dept.of Agric. Engineering,
Fac.of Agric. Technology, Gadjah Mada University. For this reason, sincerely thank is
addressed.

REFERENCES

Alamgir M. and Elhaut, T., 1994. Empowering the rural poor for self-help: IFAD's primary
objective. Agricultural Rural Development. Vol. 1, no. 1/1994.
Emin Z. Baskent, Glenwood A, 1996. Jordan Designing forest management to control spatial
structure of landscapes. Original Research Article Landscape and Urban Planning,
Volume 34, Issue 1, January 1996, Pages 55-74
Martín Alfonso Mendoza, Juan José Fajardo, Jesús Zepeta, 2005. Landscape based forest
management, a real world case study from Mexico. Original Research Article. Forest
Ecology and Management, Volume 209, Issues 1–2, 18 April 2005, Pages. 19-26.
Märiten Jansson, Therese Lindgr, 2012. A review of the concept ‘management’ in relation to
urban landscapes and green spaces: Toward a holistic understanding. Review
Article. Urban Forestry & Urban Greening, Volume 11, Issue 2, 2012, Pages 139-145
Susanto, S., 2013: Concept of Bio-landscape Management for Water Conservation of
Watershed. Proceeding of the International Symposium on Agricultural and Bio-
system Engineering (ISABE), Yogyakarta, 27-28 August, 2013.

V - 126
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Construction of a flood retention basin by using slight erodible loess

C. Boley
University of the Federal Armed Forces,
Department for Soil Mechanics and Geotechnical Engineering, Munich, Germany
[email protected]

C. Meier
Boley Geotechnik – Consulting Engineers, Munich, Germany

M. Rosport
Wald + Corbe – Consulting Engineers, Huegelsheim, Germany

ABSTRACT:
The civilian airfield of Mazar-e-Sharif (MeS) is located in the draining area of several wadis that
run off from the Marmal mountain range in the south in the direction of the utilised area. Nearly
annually, the run-off from the spring snowmelt and precipitation that merge into the wadis have
repeatedly flooded and inflicted damage on airfield's installations and the runway/tarmac area in
particular. In order to mitigate the effects, a flood retention basin was constructed in 2012. The
rectangular, horseshoe/U-shape design, consisting of three dam walls, incorporates the
natural slope of the terrain, so the basin becomes gradually shallower toward the "open" side. This
also negates the necessity of pumping the water out of the basin, since the outlet is located at the
basin's lower-most point. Due to the scarce availability of construction materials in the region, it
was opted to utilize local soils, mainly aeolian and fluvial silt (loess), which is deemed to be slight
erodible. Due to their structure and high salt content, these soils tend to rapidly destabilize with the
ingress of water, which, given the nature of the edifice, posed several geotechnical challenges.
In order to counter subsidence and reduce erodability, the soil had to be removed, upgraded with a
certain cement and water content and refilled and compacted, one layer at a time. An extensive and
special research program was developed, to find out how to use erodible soil for a dam.
the construction of an impermeable dam core, using soil material in situ posed a particular
challenge. For this the dam wall's cores were made up of several layers of Hesco-barriers. Parts of
the basin's interior were also covered in a layer of impermeable, plastic geomembranes to prevent
ground swell.

Keywords: loess, erodible, aeolian, fluvial, sediments

V - 127
INTRODUCTION

Afghanistan's northern loess soils are highly susceptible to subsidence and collapse after
saturation, resulting in extensive structural damage to buildings in the area around Mazar-
e-Sharif after heavy rain events. With this type of soil, the addition of water can result in
subsidence ranging up to 10 cm in buildings. These incidents, as well as similar problems
during the construction of the high speed railway in West China showed that an extensive
research program for such kind of geotechnical specialties was necessary for the successful
construction of flood retention basin in the region.
The loess was analyzed in regards to composition, chemical properties and in particular,
how addition of water affects the soil's behaviour under static and dynamic loads.
Furthermore, special consideration had to be given to the design of the retention pond, due
to having to rely on local or in-situ soil as the main construction material.

LOESS PROPERTIES AND FAILURE PROCESSES

To improve the understanding of the complex mechanical properties of this kind of soil,
initial laboratory tests (index tests) were carried out. These tests are necessary to get an
impression of the density of soil grain, the density of soil, natural water content, grain size,
liquid and plastic limit. To get the optimum water content with the associated density,
compaction tests were conducted. The knowledge about the optimum water content bear
relevance to the collapse behaviour under static, as well as under dynamic loads (Meier et.
al. 2008).

Index tests

The density ρs of soil particles of the loess soils from the three trial pits were determined
according to Germany Standard DIN 18124. The test results are listed in Table 1 (MP =
point of measuring and taking samples). The average value of the particle density is s =
2.72 g/cm³, which is the typical value of clay containing silts. The natural water contents w
of the samples were measured in laboratory. The values of natural water contents are listed
in Table 1 as well. The values of the natural water content w decreased with increasing
depth. The average value of the natural water content is w = 4.77 %, which is a typical
value of loess soils in dry area. The wet densities ρ of the undisturbed soil samples with
natural water content were measured in laboratory. The dry densities ρ d of the soil samples
are calculated by wet density ρ and natural water content w. Their values are also listed in
Table 1. The dry density with the average value of ρ d = 1.46 g/cm3 is relatively low. The
calculated values of void ratio e of the most samples are higher than 0.8, which is relatively
high for silts. The calculated degree of saturation Sr ranges between 10 % and 30 %. The
liquid limit wL and plastic limit wP of the loess soils were determined in the laboratory
according to Germany Standard DIN 18122. The liquid limit wL ranges between 26 % and
28 %, the plastic limit wP between 13 % and 19 %, thus the plasticity index IP ranges
between 8 % and 14 %. With the measured natural water content w, liquid limit wL and
plastic limit wP the consistency index IC was calculated. According to Germany Standard
DIN 18122 the loess soils are in a “hard state”. The grain size fractions of the loess soils
were determined according to Germany Standard DIN 18123. The grain size distribution
curves of the loess soils are very similar. The fraction of silt grains is for all of the loess
soils ranges between 56 % and 70 %. The sand grain fraction is above 16 %. The rest is the

V - 128
clay fraction. The grain size fractions are the typical compositions of loess soils in central
Asia (Boley & Zou 2007).

Table 1. Optimum water content wopt and maximum dry density ρ dmax
_____________________________________________________________________________________________________________

MP1* MP2** MP3*

_____________________________________________________________________________________________________________

density of soil grain

ρs [g/cm³] 2,73 2,72 2,71

water content
w [%] 5,00 8,13 4,41

wet density
ρ [g/cm³] 1,59 1,56 1,47

dry density
ρd [g/cm³] 1,51 1,45 1,41

void ratio
e [-] 0,80 0,88 0,94

degree of saturation
Sr [%] 17,93 22,95 11,50

liquid limit
wL [%] 27,30 26,20 27,00

plastic limit
wP [%] 13,50 14,50 18,20

plasticity index
IP [%] 13,80 11,70 8,80

consistency index
IC [-] 1,63 1,43 2,55

grain fractions:
sand [%] 16,50 70,00 13,50
silt [%] 25,00 56,00 19,00
clay [%] 17,00 69,00 14,00
_____________________________________________________________________________________________________________

*average values of 3-sample-testing

**average values of 2-sample-testing
_______________________________________________________________________________________

Chemical Analysis

In order to study the binding effects between silt grains, the lime contents VCa of the loess
soils were determined by laboratory tests according to Germany Standard DIN 18129. The
lime contents VCa in the loess soils are listed in Table 3. Their values are above 23 %,
which is very high (Meier et al. 2008)

V - 129
 Table 2. Lime content
_____________________________________________________________________________________________________________

MP1* MP2** MP3*

_____________________________________________________________________________________________________________

lime content
Vca [%] 23,90 24,70 34,60
_____________________________________________________________________________________________________________

*average values of 3-sample-testing

**average values of 2-sample-testing
_______________________________________________________________________________________

Modified compression tests, so-called collapse tests, were carried out, to identify the
compaction behaviour as a function of changes in moisture content and load level.
To get an impression of what happens to the microstructure of the soil, electron
microscope pictures were taken before and after the collapse.

Collapse tests

In order to study the sensitivity of the loess soils in Mazar-e-Sharif to water, collapse tests
were performed in laboratory according to Chinese Standard GBJ 123-88. For the tests,
undisturbed loess samples with natural water content were compressed at first in an
oedometer under different vertical pressures p. Then the samples were watered, and at the
same time the vertical deformations of the samples were measured. After watering the
saturated samples were further compressed. The relative deformations of the samples
before, during and after the watering in relation to the vertical pressures p were noted.
Figure 1 shows the relative deformations c (collapse deformation) during the watering in
relation to the vertical pressure p for three different samples (MP1, MP2 and MP3). The
collapse deformations c increase with rising vertical pressure p. The collapse deformations
c of the three samples for p = 200 kN/m² is larger than 7 %. Thus, the loess soils belong to
“high collapsible loess soils” according to the Chinese Standard. The in-situ overburden
pressure of the samples is p = 16.0 kN/m² approximately. The collapse deformation c of
the sample MP3 for p = 16.0 kN/m² is smaller than 1.5 %, thus the sample is “not
collapsible loess under overburden pressure” according to the Chinese Standard. Because
the collapse deformation c of the samples MP1 and MP2 for p = 16 kN/m² is larger than
1.5 %, the two samples belong to “collapsible loess under overburden pressure” (Boley &
Zou 2007).

V - 130
 Figure 1: Collapse deformation of loess samples

RETENTION POND DESIGN

Local conditions

The civilian airport of Mazar-e-Sharif is located at the foothills of the Marmal mountain
range, an offshoot of the Hindukush, in an area that several wadis drain into. These wadis
are usually only active during the rainy season which last from December to May. Since
the mountains serve as their main catchment area, the runoff is usually increased by melt
water in spring. In combination with a heavy precipitation event, this can lead to massive
flooding in the airport's immediate vicinity. The existing flood protection measures were
limited to a drainage ditch, which runs to the east in the direction of the small town of Gūr-
e-Mar and ends in a plain that has no natural runoff. Flood discharges therefore lead to a
backwater extending into the area of the airfield. These backwater phenomena can be best
controlled through flood retention measures. The construction of a flood retention basin
just north of the runway was deemed to be the most effective protection concept for the
airport terminal and the aircraft movement areas.

General Outlay

The Institute of Hydraulic Engineering at the University of Stuttgart was tasked with
developing a flood protection concept involving the creation of a water reservoir with a
volume of 270,000 m³. This concept, designed to be sufficient for a 20-year rain event,
based on a hydro-geological study of the local run-off conditions, taking into account the
expected service life, the protection and safety requirements, and the economic necessities.
For the event of heavier than predicted rainfalls, the retention pond required a spillway in

V - 131
order for the structure to be protected against exceptional load cases. The spillway was
designed for a 100-year flood event.
The retention pond was designed in a rectangular U-shape or horseshoe-like fashion: It
consists of three perpendicular berms and an "open" segment at the southern end. The
"middle" berm runs roughly East to West and contains the pond's outlet and the spillway.
In order to negate the need of pumps for drainage, the retention pond was built into to the
natural slope of the area. Therefore, the pond becomes gradually shallower and the height
of the dams decreases towards the open end of the reservoir.
An outlet with 1m diameter was installed in the northern section of the dam. A drainge
ditch, which connects to the already existing drainage ditch, was constructed.

Soil upgrading

Locally available materials had to be used in the construction of the dam structure. As we
have shown, these materials are extremely unsuitable for dam construction in their natural
state. Therefore, additional measures were necessary:
Firstly, the soil had to be upgraded with the addition of a Portland-type cement called “
CEM I 35,5 R” to increase erosion-resistance. A series of modified pinhole-tests according
to the American ASTM D4747 was conducted to determine the effects of adding cement to
the soil. Cement contents of 0 %, 3 %, 6 % and 9 % were conducted, the content referring
to percent in weight and a degree of compaction greater than > Dpr95 %. A hydraulic
gradient of i = 27 was applied. The tests were conducted for a period of approximately 200
hours. After completion of the tests, the samples were cut open along their longitudinal
axis and the resulting erosion channel was analysed. While soil samples without a cement
admixture showed immediate and distinct erosion soil with 3 % cement content showed
some degree of erosion after a certain period of time and the 6 % and 9 % cement-
upgraded soil showed virtually no change in the erosion-channel at all.

A series of standard proctor test according to the German DIN 18127 were conducted to
determine maximum achievable compaction depending on cement content. The tests were
also conducted for soil with 0 %. 3 %, 6 % and 9 % cement content.
Table 3. Optimum water content wopt and maximum dry density ρ dmax
_____________________________________________________________________________________________________________

Aliquot 1 2 3 4
_____________________________________________________________________________________________________________

Cement content
[%] 0 3 6 9

Optimum water content

wPr [%] 15,21 15,65 16,09 16,12

proctor density
ρ pr [g/cm³] 1,822 1,772 1,764 1,771

water content for DPr

w [g/cm³] 13-17 13-17 13-17 13-17

Erosion resistant
no no yes yes
_____________________________________________________________________________________________________________

V - 132
With the addition of at least 6 % CEM I 35,5 R, the soil can be considered erosion resistant
under the said test conditions (Meier 2011).

Preparation and Construction of the Dam Footing

To prevent settlements in the dam footing, the soil had to be upgraded. The soil in the area
of the dam footing had to be excavated to a depth of 2,5m. After an admixture of 6 %
(weight) cement and water was added, the soil was reinstalled. In order to reach optimal
proctor density, the soil was poured out in layers of 25cm and then compacted with
vibration rollers.
Furthermore, the individual layers were "interlocked". The layer surfaces were roughened
with a sheepfoot roller to minimize the risk of water channels forming in the smooth joints
between two layers.

Construction of the dam core

Since resorting to technical upgrading measures such as admixing lime or cement cannot
make the in-situ soils erosionally stable, a structurally and hydraulically effective dam core
wall was constructed from several layers of gabions.
The gabions are made up of steel mesh cage of about 1x1x1m and lined with geo-textile.
They were arranged in a pyramid like fashion, with up to 4 layers in the "deeper" northern
segment of the dam and 1 layer in the tips the southward running segments. The lowest
layer consists of 4 gabions side-by-side and ends with a single gabion in the top layer.
Individually, the gabions had to be connected with metal links without perforating the
geotextile. The gabions were filled with 3 % cement- and water-upgraded soil. The soil
was filled in in layers and subsequently compacted. When one layer of gabions was
completed, the tops of the gabions were covered with geo-textile.
Several settlement measuring gauges were also installed.

V - 133
 Figure 2. Cross-section of the dam core

Construction of the dam embankment

After completion of the dam core, construction of the dam embankment could begin. 3%
cement- and water-upgraded soil was installed in layers on either side of the gabion core
and subsequently compacted via vibration rollers. Similar to the dam footing, the layer
surfaces were roughened up by a sheepfoot roller to "interlock" the individual layers. Once
all layers were applied, a 1:4 slope was formed by scraping away excess material with
excavators.

Sealing of the pond's upstream area

Synthetic sealing sheets were used to seal off the upstream dam embankments and parts of
the reservoir area of the flood retention pond. A level subgrade was necessary first and
special care was taken to ensure that it was free of rocks and other objects that might
perforate the sealing sheets. The individual sheets were heat-sealed to together and
subsequently checked for air-tightness. The soil in the reservoir area excavated to a depth
of 0.5 m beneath the future ground line. As with the dam footing and embankments, the
soil was upgraded and reinstalled and compacted. At the location in the reservoir where the
synthetic sealing sheets end, a with trench approx. 1.0 to 1.50 m deep was excavated, into
which the synthetic sealing sheet was lowered, to prevent water from undermining the
sealing sheets.

V - 134
 Figure 3. Synthetic sealing sheets on the dam embankment

The sealing sheets c over the entire downstream portion of the embankments, the only
exception being the segment of the spillway, where it extends beyond the crest of the dam.
When the synthetic sealing sheets were completed, they were covered with another layer of
upgraded soil to protect against mechanical damage.
The embankments on both sides were then covered in a geo-textile with the geo-textile
robustness class 5, which was then covered in a 20cm layer of coarse stones (d > 50 mm).
to protect against erosion.
An asphalt road on the crest of the dam enables easy access for inspection along the entire
length of the dam.

Construction of the spillway

The spillway was constructed as asphalt, peak-discharge overflow embankment section.

Asphalt was chosen due to its superior ductility compared to concrete. The synthetic
sealing sheets extend over the crest of the berm in this segment and was covered in geo-
textile with a robustness class of 5. A drainage-layer constructed from 0/32 round-grain
gravel was established on top of the geo-textile and then stabilized with a 0/32 asphalt
layer. The asphalt had to be laid in 2 layers of 10cm, with good bonding between them and
offset seams to prevent water conductivity.

V - 135
A 30 cm deep trough was constructed at the foot of the spillway for better energy
dispersion. A layer of round-grain gravel and a drainage pipe were installed beneath the
trough.

Construction of the outlet structure

The outlet structure consists of steel tubing set in a reinforced concrete slab. The concrete
is intended to stabilize the pipe and simultaneously prevent any seepage from the pipe.
The steel pipe also had to be welded, to further prevent any potential seepage from gasket-
related leaks, which could prove catastrophic in this type of soil.

REFERENCES

Meier C., Boley C. (2012): Method for the empirical prediction of the collapse
deformation of loess soils, Bautechnik 89 9/2012: pp. 579 -584; Ernst&Sohn Press,
Berlin, Germany
Meier C. (2011): Untersuchungen zur Mikrostruktur und zum Setzungsverhalten von
Lössböden, Doctor's Thesis, Institute for Soil Mechanics and Geotechnical
Engineering, University of the Bundeswehr Munich, Germany
Meier C. (2011): Report on complementary laboratory experiments on cement-upgraded
soil, Boley Geotechnik, unpublished; Munich, Germany
Meier C., Boley C. and Zou Y. (2008): Practical relevance of collapse behaviour and
microstructure of loess soils in Afghanistan, Institute for Soil Mechanics and
Geotechnical Engineering, University of the Bundeswehr Munich, unpublished;
Munich, Germany
Boley C., Zou Y. (2007): Geotechnical Report: Geotechnical Study on the Properties of
the Loess Soils in Mazar-e-Sharif, Institute for Soil Mechanics and Geotechnical

V - 136
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Wonogiri Reservoir Sedimentation as Influenced by

Change of Catchment Characteristics
[Blank line 11 pt]
2(14pt) D.A. Wulandari & S. Darsono
Department of Civil Engineering, Universitas Diponegoro, Semarang, Indonesia
[email protected]
[Blank line 10 pt]
D. Legono
Department of Civil and Environmental Engineering, Universitas Gadjah Mada, Yogyakarta, Indonesia
[Blank line 10 pt]
[Blank line 10 pt]
[Blank line 10 pt]
[Blank line 10 pt]
ABSTRACT:

The rate of erosion and sedimentation of Wonogiri Reservoir is higher than planned, so that the
reservoir storage capacity decreases rapidly. The sedimentation problem of Wonogiri Reservoir
has reached the critical condition where reducing the sediment entering the reservoir is no longer
effective that caused sedimentation had reached the intake structure. This paper presents the result
of studies on the dynamics of the Wonogiri Reservoir sedimentation. Series of data on monitoring
reservoir bed dynamics have been used to analyze the sedimentation rate and series of bathymetry
map have been used to discuss the deposition pattern. Evacuation of the sediment from the
reservoir has been carried out in 2003 by means of dredging. The dynamics of the reservoir
sedimentation during 1993, 2004, 2005, and 2011 are performed, and the results show that in 2011
the total reservoir capacity remains 67% whereas the dead storage capacity remains 50% of the
initial capacity. The comparison of the aforesaid dynamics with the change in catchment
characteristic is discussed and suggests necessity of integrated countermasures on both catchment
and reservoir management.
[Blank line 10 pt]
Keywords: reservoir sedimentation, deposition pattern, integrated countermasures.
[Blank line 10 pt]
[Blank line 10 pt]
1. INTRODUCTION
[Blank line 10 pt]
The Wonogiri Reservoir is located in the Southeast part of the Central Java province of
Indonesia (Figure 1). Location of the reservoir is on a mountainous area, since it is the
Upper part of the Bengawan Solo river basin. Area of reservoir watershed is 1,305 km2
consists of eight sub watersheds. The Wonogiri dam was constructed in 1980 and
Impoundment December 1981. The reservoir began operation in November 1982. The
purposes of reservoir are for flood control, water supply for industrial and drinking water,
irrigation, hydropower, fisheries and tourism and maintenance of minimum flow the
Bengawan Solo River. Reservoir design life is 100 years, with original total storage of 730
MMC. 220 MMC of the reservoir storage at the elevation +135.3 m up to +138.3 m is
allocated for flood control and 440 MMC of the reservoir water for water supply at the
elevation of +127m up to +136 m. 120 MMC of the reservoir storage is purposed as the
dead storage at elevation below +127 m.

V - 137
 Figure 1. Map of the Wonogiri Reservoir (JICA, 2007a)
[Blank line 10 pt]
Almost all parts of the Indonesian are considerable rapid inpopulations’growth, and the
lands are utilized very intensively. Agricultural and non-agricultural land uses such as
settlements, transportations, industries, are two types of land utilizations. Similar with
other reservoirs, erosion at catchment area of the Wonogiri Reservoir caused sedimentation
problem (Mardjono and Sutadi, 1987). Most of reservoirs in Java Island have
sedimentation rate higher than estimated sediment rate, so that the reservoir storage
capacity decrease faster than planed capacity. The Wonogiri Reservoir sedimentation rate
of 15.2 MMC/year, it is equal to 1.4 %/year of the reduction rate for reservoir capacity. It
is a high rate for a reservoir sedimentation rate (Hastowo, 2003). The Wonogiri Reservoir
problem is not only rapid decreasing in the reservoir capacity, but also spatial
sedimentation problems such as sedimentation in the front of reservoir intake.
Sedimentation leads to reduction of the reservoir’s economic life, disruption of operations
due to intake clogging from The Keduang River debris and a decrease in the water
reservation ability, especially during the dry season (JICA, 2007a). According to sediment
measurements in the year 2011 by Perum Jasa Tirta I (PJT I) shows the remaining effective
capacity of the reservoir is only 70 % from the initial design reservoir capacity (Wulandari
et al., 2012).
Blank line 10 pt]
Reservoir sedimentation in the reservoir dead storage is one aspect for determining
economic life of reservoir. If the capacity of dead storage is full then the efficiency of
reservoir operation will be disrupted and eventually ended reservoir function. These items
are three aspects may influence the rate in reducing reservoir capacity: (a) the quantity of
sediment inflow, (b) the percentage of sediment trap, and (c) the deposit sediment density
(Loebis et al., 1987). Regular sediment monitoring or measurements need to be performed
for determining and estimating the rate of sedimentation. An evaluation of reservoir
sediment inflow and spatial reservoir sedimentation is required when the sedimentation
rate exceeds the rate of sedimentation plans.

V - 138
[Blank line 10 pt]
Based on monitoring of reservoir bottom elevation, the dynamics and deposition pattern of
Wonogiri Reservoir sedimentation presents and discusses. The comparison of the
dynamics with the change in catchment characteristic is discussed and suggests necessity
of integrated countermasures on both catchment and reservoir management.
[Blank line 10 pt]
[Blank line 10 pt]
2. DYNAMICS SEDIMENTATION PROCESSES

The reservoir sedimentation monitoring was conducted since 1981. In the period 1981 –
1988, the reservoir sediment inflow is estimated based on measurements of sediment in the
six major tributaries. The total sediment volume into the Wonogiri reservoir in 1981-1988
was estimated 9.63 million m3 . Thus the annual average sediment inflow was estimated
around 1.2 million m3/year. In 1985 and 1990, the reservoir bathymetry measurements
measured by echosounding. The estimated sediment deposits (loss of storage volume
below El. 138 m) were around 86.2 million m3 in 1981-1985 and 156.4million m3 in 1980-
1990. The estimated annual average sediment deposition is around 15.6 million m3/year in
1981-1990 (JICA, 2007b).

Subsequent sedimentation monitoring held in 1993 – 2011, the reservoir bathymetry

measurements using echosounding. Figure 2 shows the H-V curve and the H-A curve of
Wonogiri Reservoir in periods 1980 - 2011. Based on Fig. 2 it can be seen the storage and
the reservoir surface area decreases from time to time. Table 1 shows sedimentation rate
varies from time to time with the average sedimentation rate 7.74 MMC/ year. The highest
sedimentation rate of 35.62 MMC / year occurred during 2005 to 2008. The sediment
inflow are not only settle in the dead storage, the sediment inflow settles on the entire
surface of the reservoir bed. Sediment that settles on effective storage and flood storage
will reduce capacity of the effective storage and the flood storage, respectively. The
reservoir sedimentation rate is higher than planned so that at the age of 31, half of dead
storage capacity has been filled with sediment. If the sedimentation rate remained in the
range of 7.74 MMC/ year, total storage will be full on a half within the next seventeen
years. Figure 3 shows decreasing the Wonogiri Reservoir storage capacity. Up to the year
2011 remaining of total reservoir capacity is only 67% and remaining of dead storage
capacity is only 50%. Effective storage capacity and flood storage capacity also decreased,
their remaining capacity are 69 % and 57 %, respectively. Figure 4 shows the
sedimentation in the front of the intake. This was a critical condition since the
sedimentation had reached the intake structure in such that release through the intake could
not be carried out. Horizontal intake bell mouth easily clogged sediment because it located
in the dead storage zone.

Figure 5 shows the sediment deposition in each zone. Negative values indicate that in the
zone occurs bed surface erosion. Positive values indicate that in the zone occurs sediment
deposition. During 1993 - 2005, in the flood storage zone sediment deposition does not
occur. Sediment that settles in the dead storage zone decreased, this is due to dredging
around the intake. Sediment that settles in the effective storage zone increased. During
2005 - 2011, sediment deposition occurs in all zones. In the dead storage zone, deposition
of sediment decreased. In the flood storage zone and effective storage zones, deposition of
sediment increased. The increase in deposition of sediment that settles in flood storage
zone and effective storage zone will reduce reservoir function. Generally, deposition of
sediment in Wonogiri Reservoir increase. Dredging is done not greatly affect returns dead

V - 139
storage capacity, because the volume of dredged is small and dredging only done around
the intake.

Reservoir surface area (Km2)

90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
140 140

135 135

130 130

Capacity year 1980

Elevation (m)

Elevation (m)
125 Capacity year 1993 125

Capacity year 2004

Capacity year 2008
120 120
Capacity year 2011
Area year 1993
Area Year 2004
115 115
Area year 2008
Area year 2011

110
Capacity year 2005 110
Area year 1980

105 105
0 100 200 300 400 500 600 700 800

Storage Capacity (MMC)

[Blank line 10 pt]

Figure 2. H-V Curve and H-A Curve of Wonogiri Reservoir
(BBWS Bengawan Solo and Perum Jasa Tirta I)
[Blank line 10 pt]
Table 1. Sedimentation rate of Wonogiri Reservoir
Year
Description Unit 1980 1993 2004 2005 2008 2011
Sediment below El. 138.3 m MMC 0.00 80.00 32.00 2.00 106.55 19.33
Sediment dredged volume MMC 0.00 0.00 0.35 0.10 0.30 0.30
Kumulatif sediment inflow MMC 0.00 80 112.35 114.45 221.30 240.93
Sedimentation rate MMC/year 0.00 6.15 2.94 2.10 35.62 6.54
Average sedimentation rate MMC/year 7.74
[Blank line 10 pt]

V - 140
 100.00
90.00
80.00
Storage Capacity (%)

70.00
60.00
50.00
40.00
dead storage (below el. 127 m)
30.00
efective storage (between el. 127 m and el. 136 m)
20.00
flood storage (between el. 136 m and el. 138.3 m)
10.00
total storage (below el. 138.3)
0.00
1980 1993 2004 2005 2008 2011
Year

Figure 3. Decrease of the Wonogiri Reservoir storage capacity

[Blank line 10 pt]

Figure 4. Sedimentation in front of the intake (JICA, 2007b)

Figure 6 shows the bathymetry map of Wonogiri Reservoir year 1993, 2004, 2005 and
2011 base on measurement by Balai Besar Wilayah Sungai Bengawan Solo, JICA and
Perum Jasa Tirta I. The darker color indicates the deeper of reservoir depth. In 1993 the
pattern of sediment deposition in the longitudinal direction forms puddles, the deepest part
in the middle. Around the intake has occurred sedimentation, the depth became shallower.
In 2004 there has been a change in the pattern of sediment deposition. Around the intakes
look deeper due to the dredging of sediment that began in 2003. From the Keduang river
mouth, it is seen groove toward the intake. In the longitudinal direction, the pattern of
sediment deposition has the deeper part in the middle. In 2005, the depth around the intake
until Keduang river mouth still deep. In the longitudinal direction, the pattern of sediment

V - 141
deposition has the deeper part in the middle and toward the dam forming a groove. The
around intake becomes shallower. In 2011 deposition occurs uniformly across the surface
of reservoir zone. Although dredging around the intake is done, the depth around the intake
still shallow. This suggests that the dredging is done can not offset sedimentation rate. In
the longitudinal direction, the pattern of sediment deposition form puddles with the deepest
part in the middle. In tributary sediment deposition occurs toward the upstream direction as
a result of backwater. Finer sediments are carried out further into the reservoir. Due to the
wider part of the river mouth, velocity decreases and sediment settles easily. So this section
becomes shallow. At the Keduang river mouth, this condition becomes critical due to the
intake clogging. In general, the pattern of sediment deposition have a tendency to form
puddles in the longitudinal direction, this is due to the reservoir consists of several
tributaries (dendritic). Sediment that settles on flood storage zone and effective storage
zone may occur due to reservoirs operated at high water level during floods. So that the
sediment settles in the dead zone storage, the reservoir should be operated at low water
levels during floods. This operations need to be reviewed further because it will affect the
function of reservoir as a flood control.

70.00
60.00
50.00
40.00
Sedimentation (%)

30.00
20.00 flood storage
efective storage
10.00
dead storage
0.00
-10.00 1993 2004 2005 2008 2011

-20.00
-30.00
Year

Figure 5. Sediment deposition in each zone year 1993 - 2011

Key questions in the assessment of degraded tropical watersheds and the potential for
control of sediments can be posed : 1). What is the natural background rate of erosion and
sediment movement which can be expected? 2). What has been the sediment contribution
of human activities and how capable are we of controlling them ? 3).What is the potential
for storage within the basin of eroded material, whether of natural or human origins?
(Nagle et al, 1999). Wonogiri Reservoir was built on watershed which was basically had a
high erosion. According to Notohadiprawiro et al. (1981) the western part of Wonogiri
watershed was eroded almost two times more weight than the eastern part of Wonogiri
watershed. It is caused by two factors : the more erosive rainfall and soil erosion are more
sensitive. In the western region, the erosion was 412.2 T/ha/year. In the eastern region, the
erosion was 216.1 T/ha/year. The average of erosion was 314.2 T/ha/year.

V - 142
[Blank line 10 pt]
A). Year 1993 B). Year 2004
[Blank line 10 pt]

C). Year 2005 D). Year 2011

Figure 6. Bathymetry Map of the Wonogiri Reservoir
[Blank line 10 pt]

V - 143
According to Puslitbang Pengairan (1986) Tirtomoyo sub watershed and Bengawan Solo
sub watershed are sub watershed with heavy erosion. Severe erosion also occurred in the
Wuryantoro sub watershed but this does not significantly affect the depth of Wonogiri
Reservoir because the sediment was trapped by Parangjoho reservoir. Base on Abdurrosyid
and Santosa (2004) erosion in Wonogiri Reservoir watershed is mainly influenced by
vegetation cover and land conservation. Furthermore, the amount of erosion that occurs on
each sub-watershed is affected by sub-watershed area. Increasingly the area of watershed
causes higher erosion. Sedimentation rate between the years 1990 - 2000 had a significant
decrease, as the positive impact of soil conservation and land rehabilitation carried out in
1987-1994. But the reservoir sedimentation rate of 2.63 mm/year remained above the
threshold by 1.2 mm/year. Base on JICA (2007a) Erosion that occurs in sub-watershed
ranged between 5.86 – 19.50 mm/year, exceeding the allowable erosion of 1 mm/year. The
highest erosion occurred in Tirtomoyo sub-watershed, followed by Bengawan Solo sub-
watershed, Temon sub-watershed, Keduang sub-watershed, Remnant sub-watershed,
Ngunggahan sub-watershed, Wuryantoro sub watershed and Alang sub-watershed. Though
erosion is not the highest in Keduang watershed but Keduang watershed supply the highest
sediment yield, it is due to the largest area of sub watershed.

Conversion of sedimenting reservoir into sustainable resources which generate long term
benefits requires fundamental changes in the way they are designed and operated. It
requires that the concept of a reservoir life limited by sedimentation be replaced by a
concept of managing both water and sediment to sustain reservoir function. Sustainable use
is achieved by applying the following basic sediment control strategies are reduce sediment
inflow, route sediments, sediment removal, provide large storage volume and sediment
placement (Morris and Fan, 1997). Reservoirs built on the river that each has different
sediment behavior so the handling of the sedimentation also different. To obtain maximum
results, several methods can be combined (Liu et al, 2002; Wang and Hu, 2009).

The Wonogiri Reservoir sedimentation management which has been and will be
implemented has been to combine several methods. There needs to be an evaluation of the
sedimentation management program that has been done before, as the basis for moving the
program of a new sedimentation management. To deal with the most important reservoir
sedimentation is addressing a source of sediment. To handle a source of sediment can be
done with conservation, both technically or vegetation. Conservation of vegetation takes a
long time so that while conservation of vegetation can not function technically addressed
first. And this conservation should involve and benefit local communities so ensure the
sustainability of the program. The construction of small dams in the upstream sub-
watershed is necessary to trap the sediment. Reservoir operation need to be evaluated to
improve the pattern of sediment deposition and further more to handle sedimentation.
[Blank line 10 pt]
4. CONCLUSIONS
[Blank line 10 pt]
The reservoir sedimentation rate faster compare to design rate, the remaining reservoir
storage in the 2011 is only 67% and the remaining capacity of dead storage is only 50%.
Excavation of the reservoir sediment cannot produce a great result for maintaining the
reservoir storage capacity. Wonogiri Reservoir is a unique reservoir, because the reservoir
water inflow came from river mouths that located around the reservoir. Therefore,
sediment input will deposit at the river mouth around the reservoir. As a result 73% of the
sediment settles in the effective storage zone and the flood storage zone, and the sediment

V - 144
settles in the dead storage zone only 27%. Up to now, sedimentation management of the
Wonogiri Reservoir is not success to preserve the reservoir capacity. The concept of a
reservoir life limited by sedimentation must be replaced by a concept of managing both
water and sediment to sustain reservoir function. A special mathematical model is required
to map spatial dynamic sedimentation pattern for designing capacity of a new reservoir
with similar topographical condition and river inflow with the Wonogiri Reservoir.
[Blank line 10 pt]
[Blank line 10 pt]
ACKNOWLEDGEMENT
Thanks to the Balai Besar Wilayah Sungai Bengawan Solo and Perum Jasa Tirta I, which has
allowed the data used in the study.
[Blank line 9 pt]
[Blank line 9 pt]
REFERENCES
[Blank line 9 pt]
Abdurrosyid, J. and Santosa, H. (2004) : Kajian erosi lahan dan pengaruhnya terhadap
sedimentasi Waduk Gajah Mungkur di Kabupaten Wonogiri Jawa Tengah, Jurnal
dan Pengembangan Keairan, no. 1 year 11, pp.19-31.
JICA. (2007a) : Volume II : Main Report The Study On Countermeasures For
Sedimentation in The Wonogiri Multipurpose Dam Reservoir In The Republic Of
Indonesia, Dirjen SDA Kementerian Pekerjaan Umum , Indonesia.
JICA. (2007b) : Volume III : Supporting Report I The Study On Countermeasures For
Sedimentation in The Wonogiri Multipurpose Dam Reservoir In The Republic Of
Indonesia, Dirjen SDA Kementerian Pekerjaan Umum, Indonesia.
Legono, D. (2005) : Research on sedimentation issues and countermeasures in some
reservoirs in Central Java and DIY, Second Workshop on Countermeasures for
Sedimentation in the Wonogiri Multipurpose Dam, Solo, Indonesia.
Liu, J., Liu, B. And Ashida, K. (2002) : Reservoir Sedimentation Management in Asia,
ftp://ftp.hamburg.baw.de/pub/Kfki/Bib/2002-ICHE/ARTICLES/PDF/128C4-SD.pdf
didownload 1 Nov 2010
Loebis, J., Soeharto, Mahbub, B., and Ilyas, M.A. (1987) : Review of the problem of
reservoir sedimentation in Indonesia, Workshop on Reservoir Sedimentation,
Bandung, Indonesia.
Mardjono, S., and Sutadi, G. (1987) : Case study on planning and design of the Wonogiri
Dam, Workshop on Reservoir Sedimentation, Bandung, Indonesia.
Morris, G.L., & Fan J. (1997) : Design and Management of Dams, Reservoirs, and
Watersheds for Sustainable Use, McGraw Hill, USA.
Nagle, G.N., Fahey, T.J. and Lassoie, J.P. (1999) : Management of Sedimentation in
Tropical Watershed, Environmental Management, 23(4), pp.441-452.
Notohadiprawiro, T., Sukodarmodjo, S., and Dradjad, M. (1981) : Beberapa fakta dan
angka tentang lingkungan fisik Waduk Wonogiri dan kepentingannya sebagai dasar
pengelolaan, Lokakarya Pengembangan dan Pelestarian Wilayah Waduk Wonogiri,
Tawangmangu-Surakarta, Indonesia.
PJT I. (2011) : Pengerukan Sedimentasi, www.jasatirta1.co.id di download 04-11-2013.
Puslitbang Pengairan. (1986) : Executive Summary : Monitoring erosi/sedimentasi berupa
pengukuran, perencanaan dan penelitian erosi/sedimentasi di catchment area Waduk
Wonogiri, Balitbang PU, Departemen PU, Indonesia.
Wang, Zhao-yin and HU, Chunhong. (2009) : Strategies for Managing Reservoir
Sedimentation, International Journal of Sediment Research, 24(4), pp.369-384.

V - 145
Wulandari, D.A., Darsono, S., and Legono, D. (2012) : Optimasi Pemanfaatan Air Waduk
Wonogiri Dengan Program Dinamik, Proceeding PIT HATHI XXIX, Bandung,
Indonesia.

V - 146
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

Sustainable sediment management engineering solutions for solving

safety, performance
performance and environmental sediment-related issues
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
at Electricité de Francefffffjfjjfkkfjjj
(EDF) hydropower installations

2(14pt)
D. Aelbrecht, E. Valette, J. Pralong, A. Clutier
Electricité de France (EDF) – Hydro Engineering Center (CIH),
[email protected]

ABSTRACT

Quantitative management of sediments in reservoirs usually becomes a tricky problem after years
or decades of operation. It may affect safety, performance or environmental conditions of
hydropower schemes, requiring complex and often costly solutions, even more when these issues
have been underestimated or neglected in early design phases of projects development. The present
paper delivers 3 examples of sediment management solutions developed by Electricité de France
(EDF) at 3 existing hydropower schemes, covering three different situations of issues: safety
(flooding risk increase induced by siltation in a run-of-river reservoir), performance (managing
high siltation rate in a pumped storage upstream reservoir), and environment (morphodynamic and
sediment transport restoration of a bypassed river reach through bank erosion). These examples
illustrate the methodological approach and related skills and tools developed by EDF Engineering
and R&D teams for solving issues, which systematically comprise the following steps : measuring
the processes, understanding the processes (through models), developing adapted solutions
(through multi-disciplinary approach), and thinking long.

Keywords: sediment management, safety, environment, performance

1. INTRODUCTION: SEDIMENT ISSUES AND ENGINEERING APPROACH

DEVELOPPED BY ELECTRICITE DE FRANCE (EDF)

Quantitative management of sediments in reservoirs usually becomes a difficult but

inevitable issue after years or decades of operation. It may affect safety, performance or
environmental conditions of hydropower schemes operation, requiring complex and often
costly solutions, even more important when these issues have been underestimated or
neglected in early design phases of projects development. Electricité de France (EDF)
Engineering and R&D teams have developed a general comprehensive approach for
solving sediment-related issues, which systematically comprise the following steps:

Measuring the processes: there is a strong need to enlarge knowledge of sediment

processes through enhanced data collection, when existing databases are usually poor
(compared to hydrology or hydraulics data): bathymetry monitoring, suspended load,
siltation rate. Innovation in measurements techniques is also a key element: EDF and

V - 147
scientific partners develop new technologies such as turbidimeters, pit-tagged gravels
tracking techniques, bed-load transport measurement, etc…

Understanding the processes and identifying the issues: numerical and physical modelings
have to be consistently both used to improve our capacity to predict evolution of sediment
processes in the long term. EDF develops the 1D/2D/3D MASCARET-TELEMAC open
source hydroinformatics suite (www.opentelemac.org), which comprises coupled
hydrodynamics and sediment transport modules, to help understand processes at different
space/time scales, assess efficiency of possible solutions and optimize final solution before
implementation. EDF R&D also operates its own hydraulics laboratory in Chatou (France).

Developing adapted solutions through multi-disciplinary approach: once issues have been
clearly identified, solutions to be implemented strongly depend on (1) the hydraulics
conditions of the system, and particularly the capacity of flow to erode and transport
sediment material; (2) the existing works compatibility with sediment passage (weirs and
gates, turbines sensitivity to sediment loading,…). Multidisciplinary approach means
develop optimized solutions that require hydraulics engineers to work closely with
mechanical engineers (e.g. when sediment passage through turbines is considered), with
biologists (e.g. to understand sediment-ecosystems interactions), with economists and
stakeholders to find optimized ways to limit sediment economical impacts, or make use of
dredged material for other industrial usages when possible (e.g. road construction, …).

Thinking long: one of the main difficulties in solving sediment issues is associated to very
long time scales that are characteristics of sediment processes. Solutions must be robust
and potentially adaptive though continuous monitoring of sediment processes evolution.

The present paper delivers 3 examples of sediment management solutions developed by

EDF at 3 existing hydropower schemes, covering three different situations of issues:
- safety issue: siltation-driven flooding risk in a run-of-river reservoir (see § 2)
- performance issue: managing high siltation rate in a pumped-storage upstream
reservoir (see § 3)
- and environment issue: morphodynamics and sediment transport restoration of a
bypassed river reach through bank erosion (see § 4).

2. EXAMPLE 1 (SAFETY): FLOODING RISK INCREASE INDUCED BY

SILTATION IN A RUN-OF-RIVER RESERVOIR

2.1. Saint-Egrève reservoir case and safety issues

This example is a focus on flood risk management evaluation in the Saint-Egrève run-of-
river plant reservoir, where siltation potentially exacerbates inundation risk in inflow
design flood conditions. It will highlight how 1D hydraulics and morphodynamics
numerical modeling helps define optimal reservoir cross-profiles to be maintained in the
long term through optimized flushing or other measures to control flooding risk and
guarantee the minimum freeboard.
The St-Egrève reservoir (France) was built in the late 80’s in France, and operation started
in 1991. It is located in the Grenoble urban area in the French Alps, as shown on Figure 1,
downstream the confluence of the Isère and Drac rivers (catchment area : 9270 km²). The
Saint-Egrève dam and power house constitute a run-of-river power system, with a

V - 148
maximum turbine discharge of 540 m3/s. The dam comprises 5 identical gates with
overflow flaps, and a 25-meter wide tainted gate with a 6 m lifting height and a weir at
elevation 196.50 m NGF. The normal reservoir level (FSL) during operation is 205.50 m
NGF. The capacity of the reservoir in 1992 was 3.86 hm3, just at operation start. For safety
reasons, a 1 meter freeboard must be guaranteed for a flood of 3000 m3/s.

Saint-Egreve
reservoir

Isère
river
France

Grenoble
Drac
city area
river

Figure 1. Location of Saint-Egrève hydro plant reservoir (France)

The Saint-Egrève reservoir is affected by sediment accumulation on the left bank that
continues to build up (see Figure 2). In 2010, bank elevation was 204.5 m NGF on average,
i.e. only 1 m below FSL. If this bar continues to silt up, bank volume will reach 1.45 hm3.
The remaining channel has variable topography for each cross section: its minimum area in
the absence of flushing can be estimated to 250 m², i.e. a volume of about 0.6 hm3 for a
2500 m length. The channel is deepened during floods with flushing operations, and the
maximum volume that can be reached is estimated at 3 hm3, as shown on Figure 2.

Figure 2. Saint-Egrève reservoir left bank during flushing, and reservoir capacity evolution

2.2. 1D-morphodynamics simulations: calibration, validation and application

The COURLIS 1D sediment transport model, coupled to the 1D hydrodynamics open

source MASCARET software (www.opentelemac.org), was used to determine bottom
evolution kinetics during floods and flushing operations. COURLIS software computes the

Page 3
V - 149
bottom evolutions in a channel section as a function of the bed shear distribution in the
cross-wise direction of the flow (Marot et al. 2005). This numerical code can handle
cohesive and non-cohesive sediment physics.

The calibration data corresponds to May 2008 flushing operation. This flushing operation
was preceded and followed by two bathymetric surveys, one in April and the other one in
August 2008. The grid was based on cross-section profiles every 100 meters, derived from
bathymetric data. Strickler coefficient was taken equal to 45 m1/3.s-1. Model calibration
revealed the necessity to model three distinct sediment layers: (a) a top layer represents the
slightly consolidated sediment (easily remobilized); (b) a second layer constitutes the
recently deposited sediment (few years), and (c) a third sediment layer stands as the most
consolidated sediment.

The flushing parameters (discharge through gates, reservoir level) and the morphodynamic
simulation results are summarized on Figure 3. Simulation and measurements are
compared for following parameters: erosion sediment flux (in tons/s); cumulated erosion
(in tons). Main parameters to be calibrated are: hydraulics and skin friction factors
(Strickler coefficients); critical erosion and deposit stresses for each sediment layers.
Simulation results are compared against measurements of (i) erosion sediment fluxes
(based on suspended load concentrations measured with 1h-sampling rate turbidimeters),
and (ii) net cumulated erosion (derived from the difference between ingoing and outgoing
fluxes).

The following phases can be observed:

- During lowering of the water level (phase 1), one observe an erosion peak  of the
upper slightly consolidated silt layer;
- The main erosion peak  corresponds to the end of phase 1, when the water level
has reached its minimum level. It is pretty well represented by the 1D model
- The increased erosion  is due to passage of flood peak (phase 3). This third peak
is represented though underestimated somehow by the model.

The cumulative calculated mass of eroded sediment (1.13 Mt) is in good agreement with
the measured mass (1.14 Mt). At the time of the flood peak on May 30 at 1 pm, over a
million tons of sediment had already been eroded. It must also be reported that an
important accumulation of sediment in the reservoir occurred during the period between
the end of flushing (e.g. end of simulation) and the bathymetric survey period. Indeed, the
high flow episode of June 2008 lasted after the flushing, causing solid matter inflows and
significant settling in the reservoir of an order of magnitude of 300,000 tons (i.e. an
average of 1 meter sediment deposit in the channel over three months).
The model has then be used and validated against June 2010 flushing data. The correlation
between the calculated mass of eroded sediment (0.48 Mt) and the measured one (0.52 Mt)
is satisfactory. The evolution of the eroded mass over time is also well represented (not
presented here but pretty similar to patterns shown for the calibration against 2008 event).

2.3. Conclusion for Saint-Egrève reservoir “safety” case

Calibration and validation of the 1D coupled hydrodynamic/sediment transport model in

the Saint-Egrève reservoir, against two recent flushing/flood events (2008, 2010), show
that the erosion sediment fluxes and cumulated erosion are appropriately predicted by

Page 4
V - 150
MASCARET-COURLIS model. The model demonstrates that it is possible to expect
sufficient erosion in early phase of flushing procedure, to avoid reaching critical water
levels and thus to ensure a minimum freeboard in extreme design flood conditions, despite
unavoidable progress of left bank siltation. In conjunction with morphological analysis,
including effect of bed load transport (not simulated here), these simulations help define a
“target state” of reservoir cross profiles to be maintained in the long term. Flushing
procedures are now under review on the basis of these results. Details can be found in
Valette et al. (2014).




Figure 3. Model calibration through comparison against 2008 flushing measurements

3. EXAMPLE 2 (PERFORMANCE): PUMPED STORAGE PLANT

PERFORMANCE CHALLENGED BY RESERVOIR SILTATION

This example illustrates a cost-effective sediment diversion/bypassing solution, studied at

basic design stage so far, to ensure the long term efficiency and performance of a pumped-
storage system with high siltation rate in the upstream reservoir, while not impacting water
quality downstream.

3.1. Flumet reservoir case and performance issues

Flumet-Cheylas pumped storage plant is part of a whole hydropower system located along
the Arc and Isère rivers in France (Figure 4). Flumet upper reservoir had a 1.8 hm3
sediment deposit volume in 2013, out of 5.1 hm3 of total original capacity, and 4.7 hm3 of
usable capacity. An estimated residual non avoidable average flux of 50 000 m3 a year of

V - 151
new incoming sediment continues building up siltation process. Note that this residual
incoming flux has already been reduced over time and “optimized” through additional
measures taken at upstream facilities of the hydropower system.

Figure 4. Situation of Flumet reservoir in the Arc-Isère hydropower system

Technical and economical performance require to find sustainable solutions for (1)
extracting the 1.8 hm3 existing deposited material, and (2) periodically remove from the
reservoir new additional sediment yield that will continue over years.

3.2. Analysis of solutions for extracting sediments from Flumet reservoir

Two main categories of solutions may be envisioned: (1) Extracting and managing
sediment “out of water”; or (2) extracting and managing sediment “underwater” (dredging
and release of sediments in rivers). Category 1 solutions are not possible for some reasons:
- It would generate about 200 000 trucks rotations to transport extracted sediments in
urbanized areas, and about 25 000 trucks laps every 5 years for removing new
sediment load;
- Sediments would require specific treatment (drainage, drying, …) before any
transport and/or use;
- No available space was found at reasonable distance for long term storage.

Solutions have to be found in category 2. Different sub-solutions have been explored:

- Solution 2.A: dredging and releasing sediments in the Breda “small” river, right
upstream Flumet reservoir: Breda river flow is too low for diluting sediment and
getting sufficiently low and acceptable concentrations; furthermore, it would have
certainly been difficult to get a permit for releasing sediments in another catchment;
- Solution 2.B: dredging and releasing sediments through a local bypass of existing
power units:
 In this case, it is not possible to turbine water during extractions of sediment
(high loss of generation);
 To avoid risk of clogging in the existing penstock, high velocities are
necessary, thus imposing a huge – and not realistic – design of the bypass;
 The sediment problem is only “shifted” from the upper reservoir to the
lower one: sediment from the lower reservoir has to be ultimately extracted
anyway.

V - 152
 - Solution 2.C: dredging and releasing sediments directly through power units
 So far, maximum concentration acceptable for units is not precisely known,
despite some existing new guidance in the turbine industry (IEC, 2013)
 Potential impacts on units are: runners erosion, clogging of unit filters,…
 A precautionary maximum concentration value of 0.05 g/l has been
prescribed by mechanical engineers; this would impose a 20 years +
procedure, which is unrealistic.
- Solution 2.D: dredging and releasing sediments further downstream in the Isère
river that “naturally” receives those sediments, through a new dedicated pipe or
conduit which entirely bypasses existing penstock and powerhouse. This solution
seems to remain the only one that could be technically and hopefully economically
feasible.

3.3. Feasibility and optimization of a sediment release conduit

A solution of conduit installed in the existing penstock has been rejected, due to generation
loss induced, and risk of perturbation on generation in the long term. Thus, conduit path
must be found out of the existing penstock. The feasibility and optimization rationale runs
from downstream to upstream:
- Step 1: what are acceptable release conditions in the river downstream ?
- Step 2: design dredging methods that meet these release conditions
- Step 3: design a sediment release conduit appropriate to downstream release
conditions and upstream reservoir dredging solutions. Design of the conduit is
based on the following optimization parameters and factors:
 Conduit path: property issues; social, environmental and economical
impacts of works; crossing singularities (railroad, roads)
 Conduit diameter: required (minimum) and acceptable (maximum) flow
velocity, standardized diameter available, cost
 Conduit material: resistance to abrasion and pressure, double coated conduit
 Position of downstream release outlet and release procedure: 24h average
and maximum concentration authorized in the Isere river downstream,
seasonality of release,…
 Total cost and cost-effectiveness of the project (the project must avoid long
term generation loss due to reservoir capacity reduction)
 Regulatory procedure compliance.
The study is presently converging towards a conduit design with following characteristics:
6 600 m long, with a progressively decreasing diameter 800-600 mm, allowing a head flow
of 2 m3/s. Conduit will be cast iron made (with or without protective inner coat); a 24-h
average sediment concentration of 0.5 g/l after dilution in the Isere river downstream will
be ensured, and April-August season is preferred for dredging and release operations.

4. EXAMPLE 3 (ENVIRONMENT): MORPHODYNAMIC RIVER

RESTORATION THROUGH BANK EROSION

4.1. Old Rhine river case and environmental issues

This example describes an innovative concept of controlled bank erosion to restore alluvial
dynamics and sediment yield and transport, and ultimately improve habitats conditions for
aquatic ecosystems and biodiversity in the Old Rhine river. River morphological

V - 153
alterations issues are increasingly addressed, and there is a societal demand for finding
relevant and sustainable restoration solutions, which will depend on processes and sources
of perturbations. Habersack & Piegay (2008) give some interesting principles and
examples of restoration measures implemented in Alpine rivers, which may act on water
fluxes, solid fluxes or river bed structure.

The area of interest here is located along the French left bank of a 45 km long bypassed
reach of the Rhine river (named as “Old Rhine”) near the Swiss-German-French borders,
parallel to the “Grand Canal d’Alsace” (GCA) that diverts main part of river flow from
Kembs dam for hydropower generation (Kembs, Ottmarsheim, Fessenheim, Volgelgrün
hydro plants) – see Figure 5.

Over the two last centuries, this reach was subject to channelization for flood protection
and navigation (19th century), and then damming for hydropower generation (since about
1925), with reduction of flow (e.g. diversion to the GCA). Successive engineering works
induced morphological simplification and stabilization of the channel pattern from a
formerly braided form to a single incised channel, presenting a bottom armour layer, and
generating ecological alterations, as illustrated by maps reported in Arnaud (2012).

EDF commitments within Kembs hydropower station licensing renewal comprise a set of
pro-active hydro-environmental measures (Clutier et al., 2012):
− increase of minimum flow conditions in the Old Rhine bypassed reach, seasonally
variable to adapt to ecological needs, to ;
− creation of a naturalized new channel connecting upstream Rhine and the Old
Rhine, allowing new passage for fauna;
− and an innovative morphodynamic restoration program through gravel injections,
and controlled (left) bank erosion, which is the topic of interest here.

4.2. Bank erosion concept and optimization on a physical model

The concept of controlled left bank erosion is quite simple: as almost no sediment material
can be supplied naturally from upstream, sediment supply must be searched from riparian
areas. The idea is to weakening the bank and let “usual” floods (Q1-yr, Q10-yr) erode the
bank and transport former Rhine gravel sediment that compose the embankment back to
the river, and generate new morphological patterns favorable for ecosystems habitats. This
erosion process must also be controlled to preclude any safety concern for adjacent
hydraulic structures (bridge piers, the Grand Canal, hydrometric stations,…) in extreme
conditions (typically Q100-yr) - see details of this project in Aelbrecht et al. (2014).

Among all bank sites screened potentially favorable for such controlled erosion, a pilot site
O3 (initials refer to municipality name initials sites are located in) has been chosen to study
and optimize the way erosion can be initiated, through physical modeling. Physical model
testing resulted in an optimized rearrangement of the bank configuration:
- Total removal of riprap is required (e.g on a 400m length)
- Groynes must be detached from the bank to allow flow to generate a eroding flow
at the toe of the bank that can undermine the bank and then generate sliding of the
bank
- Groyne spacing must be increased (180 m ~ 10 times groyne length), to avoid
masking and bank protective effect of one groyne to the next one.

V - 154
 Figure 5. Location of Old Rhine river
each of interest, between Kembs and
Breisach dams.

France GCA Old

Germany Rhine

flow
Switzerland

The resulting optimized configuration for O3 site is described on Figure 6. Details of

scientific basis and results of physical model testing can be found in El-Kadi Abderrezak et
al. (2013).

Riprap totally
removed (400 m)
Groyne 3
detached
Groyne 2
removed

Groyne 1
detached

Figure 6. O3 site optimized riprap and groynes modification scenario

4.3. Field implementation and first results following June 2013 flood

Implementation works on site O3 have been completed during March-May 2013. Just after
O3 site works completion, a timely significant flood occurred in early June 2013 (flood
peak occurred on June 2nd). Maximum flow in the Old Rhine was estimated about 2400
m3/s (e.g. 3850 m3/s at Basel upstream station, Switzerland), corresponding to a Q20-yr
statistical flood approximately. So far, only qualitative observations of the effect of June
2013 effect on erosion of O3 bank site are available. They are depicted on Figure 7 (a, b,
c), where comparison can be made between (a) end of work, (b) end of June 2013 flood
event, and (c) after-flood situations. The position of a tree pointed on Figure 7 helps
evaluate erosion development.

V - 155
Long term physical (topography survey with 3D scan laser, tracking of 1050 pit-tagged
pebbles) and ecological (fish population characteristics) monitoring are planned and will
allow to evaluate the efficiency of morphodynamics evolution of the site and its ecological
response. Based on positive June 2013 flood effect on O3, duplication of this erosion
concept is already planned for 3 other sites. These works should be implemented in 2015.

(a) (b)

Tree
Tree

Groyne 1

(c)

Tree

Figure 7. O3 bank site (a) end of works (May 27, 2013) ; (b) end of Flood event (June 6, 2013) ;
(c) low flow conditions (end of August 2013).

REFERENCES

Aelbrecht D., Clutier A., El-Kadi Abderrezzak K., Barillier A., Pinte K., Garnier A., Die-
Moran A., Lebert F. (2014). Morphodynamics restoration of the Old Rhine through
controlled bank erosion: concept, laboratory modeling, field testing and first results
on a pilot site. Submitted to River Flow 2014 conference, Lausanne, Switzerland.
Arnaud F. (2012). Historical and experimental geomorphology for the process-based
restoration of a modified fluvial system: the Old Rhine between Kembs and Breisach.
PhD thesis, University of Lyon II, Dec. 2012, pp. 280.

V - 156
Clutier A., Aelbrecht D., El-Kadi Abderrezzak K., Die Moran D., Pinte K., Barillier A.
(2012). Restauration du transport sédimentaire dans le Vieux Rhin par érosion
maîtrisée des berges. I.S.Rivers. Recherches et actions au service des fleuves et
grandes rivières, Lyon, France, 26-28 Juin, 2012.
El Kadi Abderrezzak K., Die Moran A., Mosselman E., Bouchard J. P., Habersack H.,
Lebert F., Aelbrecht D. (2013). A physical, movable-bed model for non-uniform
sediment transport, fluvial erosion and bank failure in rivers. Journal of Hydro-
environment Research. Online Oct. 2013: http://dx.doi.org/10.1016/j.jher.2013.09.004.
Habersack H, Piegay H. (2008). River restoration in the Alps and their surroundings: past
experience and future challenges. In Rinaldi M. Habersack H., Piegay H. edits.
Gravel-bed rivers 6: From process understanding to the restoration of mountain
rivers, Elsevier, Amsterdam, pp. 703-737.
IEC 62364/Ed1 (2013). Hydraulic machines – Guide for dealing with hydro-abrasive
erosion in Kaplan, Francis, and Pelton turbines. Final draft international standard.
Marot D., Bouchard J.P., Alexis A. (2005). Reservoir bank deformation modeling:
application to Grangent reservoir. Journal of Hydr. Engineering 131 (7), 586-595.
Valette E., Tassi P., Jodeau M., Villaret C. (2014). St-Egrève reservoir – Modelling of
flushing and evolution of the channel bed. Submitted to River Flow 2014 conference,
Lausanne, Switzerland.

V - 157
 CATCHMENT AREA MANAGEMENT FOR SUSTAINABLE
PASELLORENG IRRIGATION DAM IN CONNECTION WITH THE
GLOBAL CLIMATE CHANGE
Haryanto1, Rahayu1, Eka Rahendra1, Subandi1, Andi Muhammad Ratmiadi1
& Arif Paputungan2
1)
The Pompengan Jeneberang Large River Basin Organization, Makassar, Indonesia
2)
Water Resources Consultant of Pompengan Large River Basin Organization , Makassar, Indonesia
[email protected]

ABSTRACT
In connection with the global climate change, a serious problem of Paselloreng district is (1) no
environmental development (2) no sustainable water for existing environmental development (3) The
land acquisition related with land used for conservation development. The Paselloreng dam is earth
fill dam will solve serious problem in the catchment area to be better. The catchment area
management for sustainable Paselloreng irrigation dam in connection with the global climate change
must be solved under land use planning in the dam, sediment control, assessment of catchment
erosion, public participation, land and water conservation Theoretically, that human activities are the
principal cause of climate change, this region is experiencing heat waves, severe droughts and wild
fires while other regions are facing unusually strong monsoons, widespread flooding, and rain-
induced landslides. Caused of these extreme events, people around the region are facing some form of
climate related crisis with increasing frequency. Therefore, to solve the problem, efforts to advise the
region on how to reduce their greenhouse gas emissions and cope with climate change are ongoing.
National strategies for sustainable development are being implemented by many regions as well as
programs to monitor and mitigate greenhouse gas emissions and the global climate change will be a
long term challenge which will likely require substantial changes in the behavioral patterns of
society. Based on survey that no development of conservation in this region, conservation
improvement works must be prior implemented to secure water for the reservoir. The result expected
that the Paselloreng reservoir will store 132 million m3 water to irrigate 7.000 Ha Gilirang
Irrigation and increasing of existing environmental in the catchment area. In 2014, about 20 kms of
access road will be constructed to the dam site not only for dam construction but also for
environmental development including for the region economic and communities development.

Keywords: Catchment area management, Sustainable paselloreng dam, Gilirang irrigation.

1. INTRODUCTION

The location of the Paselloreng irrigation dam and the Gilirang irrigation scheme as shown in
figure 1, is located in Paselloreng village, Wajo regency, South Sulawesi, Indonesia. The
Paselloreng reservoir will store 132 million m3 water to irrigate 7.000 Ha Gilirang Irrigation
and increasing of existing environment in the catchment area. In 2014, about 20 kms of
access road will be constructed to the dam site not only for dam construction but also for
environmental development including for the region economic and communities
development. Paselloreng irrigation dam and Gilirang irrigation scheme is one among the
projects prioritized within in South Sulawesi which will be managed the major objectives of
irrigation sub-sector under the former of the Pompengan Jeneberang large river basin
organization, Directorate General Water Resources is directed to meet the following : (1) To
expand the irrigation schemes in order get a more balanced water use and ensure water
conservation, protect production area and prevent devastation by flooding and drought to

V - 158
support the utilization of new agricultural fields. (2) To increase irrigated agriculture so as to
maintain the water resources and irrigation schemes (3) To promote a more a more
harmonized development in related with other sectors such as settlement, industries, river
maintenance, hydroelectric power, tourism, strengthen maintenance and rehabilitation of
irrigation schemes through strengthening the farmers’ capacity and participation.

The Paselloreng irrigation dam is particularly essential from regional perspective because (1)
The Paselloreng irrigation dam to the Gilirang large rice granaries plays ab important role in
supplying rice to another deficit region (2) In the Gilirang project area is paddy cropping
under rainfall conditions to be widely agriculture practice, the crop productivity is slowly,
automatically without Paselloreng dam and Gilirang irrigation development, the income level
and living standard of its inhabitants are lower than the dam and irrigation establish. To solve
the problem mentioned above the GOI will construct the Paselloreng Irrigation dam and
improve the existing Gilirang irrigation to increase the existing environment, the 7.000
hectares irrigation improvement including irrigation and drainage system. During the high
water level at 53.5 m above sea level, the Paselloreng reservoir will cover or inundate the
whole area of the Paselloreng village about 11 km2. The consequences of this the Paselloreng
irrigation dam construction such as severe and permanent economic, social and psychological
risk to the displaced population, for example; lost of property among the displaced population
in related with houses, land, tree, crop and other assets, grazing land, the social and
psychological cost for displaced population even the whole communities to the resettlement
site in Bekkae village. The environmental cost in terms of pressure over the existing forest
land, grazing land at the resettlement site which eventually affect the sustainability of the
Paselloreng dam. Some of the consequences mentioned above can be quantified. Streaming
from such a wide range of consequences towards the Paselloreng people and their natural
resources. It is essential to monitor all stages of preparation of human resettlement,

V - 159
relocation and development in related with the project stages, for example: From before
construction up to post construction stage. The objective of monitoring human resettlement of
the dam project will help minimize something related to constraints towards the project
construction as well as the sacrifice of the project affected population, in particular
Paselloreng people.

The goals of monitoring human resettlement are directed (1) To facilitate pertinent proses of
land acquisition of Paselloreng people and the resettlement site of the Bekkae village for the
construction of the project facilities in connection with Paselloreng dam (2) To avoid extreme
sacrifice among the Paselloreng people for advantage people in resettlement site. (3) To
encourage an integrated development activities for a better life of the project affected
population in the resettlement site during the dam construction. (4) To result the Paselloreng
reservoir will store 132 million m3 water to irrigate 7.000 Ha Gilirang Irrigation and
increasing of existing environment in the catchment area including for the regional economic
and community development.

2. CATCHMENT AREA MANAGEMENT FOR SUSTAINABLE DAM

The catchment area management for sustainable Paselloreng irrigation dam in connection
with the global climate change under land use planning in the dam, sediment control,
assessment of catchment erosion, public participation, land and water conservation will be
explained briefly:

2.1. Land Use Planning in the Paselloreng Irrigation Dam

According to the detailed design, the resettlement plan is normally prepared by the
transmigration and the forest agencies, as the displaced 310 households in the Bekkae village

V - 160
as a resettlement. Theoretically, the resettlement site is ready to be occupied continued by
mobilization and resettlement. When the re-settlers have arrived at a resettlement site will be
continued by construction of resettlement facilities such as school, market, small hospital
and other relevant public services. According to the detailed design that a household will be
given 2 ha land acquisition by the government. It will be used 0.5 ha of compound activities,
0.5 ha of the first plot and 1 ha of the second plot of farm activities. However, in special
situation, some exceptional treatment. A composition of 1.118 ha land use for resettlement
consist of (1) Opened Land of 325.30 ha = 29.13% (2) Unopened Land of 792.70 ha =
70.90%. Detailed Percentage of land use as follows: (1) Opened Land will be used for
13.42% of compound area, 13,42% of the first Farm, 0.56% of public facilities, 0,18% of test
farm, 0.18% of seed farm, 0.38% of main road, 0.46% of village road, 0.38% of village land,
and 0.18% of the cemetery, (2) Unopened Land will be used for 26.84% of the second farm,
0.30% of farm road, 10.68% of conservation, 32.23% of reserved land and 0.81% of grazing
land.

2.2. Sediment Control

Sediment inflow to the Paselloreng reservoir come from the surrounding Gilirang river, many
factors influence the sediment transport rate, the sediment load must be analyzed for sediment
control. The obvious effect of the sediment inflow to the reservoir will increase bed elevation
in the reservoir over time, this situation will be the largest inflow of sediment outflow of
sedimentation from the reservoir therefore to anticipate the sediment flow to the reservoir
must be constructed several sediment control dams or sabo dam like check dam, sabo dam,
sand pockets in upstream of the Paselloreng reservoir. Land slope is an important factor in
catchment erosion. In this case, 4 factors besides land slope influence to the soil erosion
determination such factor as (1) rainfall factor or erosion (R) will depend on the monthly

V - 161
rainfall intensity, number of rainy days in one month and 24 hours maximum rainfall in a
month. (2) Soil factor or erosion (K) depends on the soil characteristics and the soil
parameters influence the erosion like Percentage of silt, fine sand, top soil depth, organic
matter content and permeability (3) topography factor (LS) is a combination of slope length
factor (m or km) and slope steepness factor (% or degree) (4) cropping and management
factor (C) is reduction effect factor on erosion due to vegetation cover. The better area area
will result in a smaller and unexploited woodland in natural condition which will has a factor
between 0.001 and 0.004. Conservation practice (P) is the soil loss reduction factor obtained
by more appropriate practiced cultivation.
The best reduction factor is at bench terrace type with excellent terraced which will gain P
factor of 0.04. The four factors mentioned above influenced by the slope except the rainfall
factor. The total amount of sediment yield in the basin can be predicted with the total soil loss
in Universal Soil Loss Equation (USLE). Not all the total soil removed from the land of the
basin is carried down to the Gilirang river, but a large part of trapped or deposited along its
way. The ratio between the sediment yield measured at a gauging station and annual soil loss
in its basin to be called by delivery ratio. The other sediment source called by mass
movement which detachment of bulk of soil or rock from original position due to
modification of the slope equilibrium condition. The main objective of sediment transport
equation will predict whether equilibrium condition, erosion, scouring, deposited sediment,
silting will occur and determine the quantity of sediment. The rate of sediment transport,
expressed as mass, weight or volume per unit time can be determined from measurement of
calculation. The main reason of the empirical character of sediment transport is the
complexity of the sediment transport process. The interaction of turbulent flow, the
characteristics of which are inky known by empiric and boundary consisting of loose
sediment cannot be described by simple equitation. Related with the sediment, there terms
must be known, it is (1) Initial of sediment motion, (2) Sediment transport mechanism (3)
Bed material sediment transport. (1) Initial of sediment motion, the equilibrium of a particle
on the bed as the stream will be if the resulting effects of disturbing forces like drag force,
lift and various forces on the particle surface which become greater than stabilizing forces as
gravity an cohesion. (2) Sediment transport mechanism, for turbulent flow over a rigid bed, a
description of the flow structure could be given by empirical methods. Bottom shear stress,
water depth and bed roughness are the most important parameters. Description of particle
motion under the action of the flow is also empirical so that not difficult to be understood that
the relation between flow and sediment transport. Based on the sediment transport
mechanism, two major models will be distinguished between Bed Load and Suspended Load.
Bed Load is the movement of particles in contact with the bed by rolling sliding and jumping.
Suspended load is the movement of particles is continuously compensated by the diffusive
action of the turbulent flow field. (3) Bed material sediment transport is the transport of
particles without or with small quantities in the bed. The material is supplied by erosion and
no direct relationship with local condition exists, generally, fine material is lest than 50 um.

2.3. Assessment of Catchment Erosion

Land degradation and subsequent soil erosion and sedimentation play a significant role in
damaging water resources within watersheds and basins. Using conventional methods to
assess soil erosion risk is expensive and time consuming. Geographic Information Systems
(GIS), coupled with the use of an empirical model to assess risk, can identify and assess soil
erosion potential and estimate the value of soil loss. The objectives of this study are to: (1)
assess soil erosion risk within a GIS and the Revised Universal Soil Loss Equation (RUSLE);
(2) establish spatial information of water soil erosion for Gilirang Catchment Area. It is

V - 162
located about 35 Km west of Bone Bay and covers an area of approximately 295 km² of the
southern West Bank A set of factors as identified in the RUSLE are studied and reviewed.
These include rainfall erosion factor (R-factor), soil erosion factor (K-factor), slope and slope
length factor (LS-factor), vegetative cover factor (C-factor) and conservation practice factor
(P-factor). Each of the above mentioned RUSLE factors, with associated attribute data, is
digitally encoded in a GIS database to eventually create five raster layers. Simultaneous
overlay operation on these five layers produces a resultant raster layer. The Final Erosion
Risk Map has classified into five Soil Erosion Classes, these are Very low, Low, Moderate,
high, and Severe areas. The study shows that 3.1 % of the catchment area are located in the
severe class and consists of about 9.1 Km² of the study area. Soil erosion is the removal of
surface material by wind or water, which is very dynamic and spatial phenomenon. In the
process of passing a good farmland to the next generation, it is necessary to have sound and
sustainable conservation practice not to allow soil erosion at a rate, which depletes soil
resources faster than they can be renewed. To achieve this goal, there is necessitating for
modeling the spatial distribution pattern of soil erosion. The erosion components like soil
infiltration characteristics and rainfall are spatially and time variable and have a strong
impact on water erosion process. All the factors of Revised Universal Soil Loss Equation
(RUSLE) are geographic in nature, which means they can be referenced to a particular
location. The Geographic Information System (GIS) techniques are two valuable tools for
spatial data analysis and modeling. It is quite obvious as it assumes the use of various data
like soil, vegetation, relief, climate data, and other factors that can be effectively processed in
GIS environment.
Soil erosion may affect agricultural productivity and land use changes, quantitative estimates
on the relationship between soil erosion and crop productivity are vital. The overall objective
of this study is to serve the catchment area with an erosion risk model using (GIS) to reduce
the risk causes by erosion, by giving an overview about the future situation of the soil. More
specifically the main objectives of this study are (1). Development of a soil erosion intensity
map using Revised Universal Soil Loss Equation (RUSLE) with the aid of remotely sensed
data in a GIS environment. (2). Development a database for the Hebron catchment area about
soil erosion. Gilirang catchment area, catchment in Wajo district, is located about 35 Km
west of Bone Bay and covers an area of approximately 295 km². It's mostly hilly, rocky and
its slope gradient is also increase to the south. The climate of study area tends to be cold in
winter and warm in summer and relatively humid. The average maximum temperature is
about 34 C° in the dry season and the average minimum temperature is 9 C° in rainy season
and the average humidity varies between 25-80%.
The Universal Soil Loss Equation is an empirical equation designed for the computation of
average soil loss in agricultural fields. This equation was developed for detachment capacity
limited erosion in fields with negligible curvature and no deposition and represents soil loss
averaged over time and total area. The equation has the following form (Wischmeier and
Smith 1978) is E = RKLSCP. Where E [ton/(acre.year)] is the average soil loss, R [hundreds
of ft.tonsf.in/acre.hr.year] (in SI: [MJ.mm/ha.hr.year], R[SI]=17.02R[EU]) is the rainfall
intensity factor, K[tons/acre/unit 3R] = [tons.acre.hr/hundreds.acre.ft.tonsf.in] is the soil
factor, LS[dimensionless] is the topographic (length-slope) factor, C[dimensionless] is the
cover factor and P[dimensionless] is the prevention practices factor. Revised USLE or
RUSLE uses the same empirical principles as a USLE, however, it includes numerous
improvements, such as annual factors for rainfall. For this study improved empirical
equations were used for the computation of rainfall erosion (R), topographic (LS) factor and
crop management (C) factor (Wischmeier and Smith, 1978; Novotny and Olem, 1994). For
Rainfall-Runoff Erosion (R) Factor, since rainfall intensity is not available for the region, the
R factor can be approximated using annual mean precipitation data. The rainfall erosion R-

V - 163
factor was derived from a 36-year (1960-1996). The R factor values, are computed from
equation (1) below (Renard and Freimund, 1994):

R = 3.85 + 0.35(P) (1)

Where, R = rainfall-runoff, erosion (hundreds of ft.tonsf.in/acre.hr.year), P = Average annual

rainfall (mm/yr). For Soil-Erosion (K) Factor, The only soil map available for the study area
is the soil association map with scale 1:250,000, that conducted by Ravikovitch, 1992. This
soil association map was digitized using Arc View GIS software. An intensive field sampling
effort of different soil association was made in order to collect soil samples to identify the
textures, structures, and moisture content of each soil association. Plotting the field data on
the monograph, the K factor for each soil association was calculated. As a result the K factor
map was prepared from the soil map, using the values given. For modified LS factor that the
topographic factor, LS, is divided into two components: S is the slope grade (drop in
elevation/slope length) expressed as a %; and L is the length of the slope. Slope grade affects
mainly the speed of runoff. Slope length affects mainly the amount of runoff, since the runoff
is cumulative. In assessing the effects of slope length it is necessary to take into account the
total length of the slope over which runoff occurs, not just the length of the field in question.
To calculate the LS factor A digital contour shape file with 5 meter contour interval is
obtained from the GIS Lab. This contour shape file is used to create a Digital Elevation
Model (DEM) with 30 meters grid size using Arc View 3.2 spatial analyst extension. A slope
grid map with 30 meters grid size is created from this DEM. Then LS values are computed
using equation (2) (Morgan, 1986):

LS = L(0.065 + 0.045* S + 0.0065* S2)0.5/(22.13) (2)

Where, L is slope length in meter and S is slope steepness in percent (%). In this study, a
supervised classification was employed in Landsat image of 2005 to prepare the land use or
land cover map of the Gilirang catchment area. The catchment, is classified into five land
use/land cover classes, namely crop land, forest land, pasture land , urban land, and bare rock
In this study, the Cover (C) factor values were drawn from several sources. Prevention
Practices (P) Factor, the P factor map was prepared from land use/land cover map, which was
prepared from supervised classification of Landsat images, using the values given. The P
factor values were chosen based on Agricultural Handbook. By multiplying all the five
factors, using Arc View 3.2 GIS software, an integrated composite map of erosion intensity is
generated. The Final Erosion Risk Map is classified into five Soil Erosion Risk Classes, for
example : Very low, Low, Moderate, high, and Severe areas. It's important to note that only
3.1% of the study area are located in the severe class and consists of about 9.1 Km² of the
study area. Therefore, the Gilirang riverbank and the groynes must be constructed to protect
the Gilirang catchment area from erosion. Erosion will be controlled by check dam or the
sabo dam or erosion control dam located in the upstream of Paselloreng reservoir. Gilirang
riverbed erosion or the river bed degradation can be solved by the construction of the
ground-sill.

2.4. Public Participation

Based on review, analysis on the potential level of Gilirang farmers’ participation that despite
the establishment of such a large number of WUAs mentioned previously, it is evident that
the effectiveness of their operation had been very poor. This had been suspected by the
impacts of small land holding condition, which brought about farm incomes which are far

V - 164
from adequate for the farmers to fully participate in the irrigated farming activities. It is it
was concluded that the maximum size of land holding for irrigated paddies in Gilirang
irrigation that could be performed by relying the family labor only marginal subsistence
farming. Any size smaller than this figure potentially suffered from the risk of negative
income, and hence not likely possible to contribute adequate financial or labor resources for
securing sustainable O&M of irrigation schemes. In fact, the land holding category up to 0.50
ha per farm household – which dominates the irrigated land areas of the country at 48.60% –
is considered to be marginal subsistence farming, and hardly expected to participate sharing
any contribution for sustainable O&M. The land category of >0.5< 1.00 ha and of >1.00
<2.00 ha area currently stood at about 22.2% and 16.80 % of the total agricultural land of the
country respectively. From the analysis of financial return, the same analysis concluded that
the land holding category of >0.5 ha; 1.0 ha and 2.00 ha produce the net value of production
of US$ 91.6; US$ 463.51; and US$ 1,119.53 respectively. These figures have been based on
irrigated paddy at 1.30 cropping intensity, after deducting indirect costs such as materials and
labor, and indirect costs such as taxes, home consumption, and yet, without imposing any
irrigation service fees.
Potential Capacity for Farmer’s Participation that a review of the potential level of farmers’
participation is further scrutinized by using some assumptions, including the basis for full
participation in the land holding rounded to 2.00 ha per farm household. The size of the
commanding area for the WUA of 100 ha/WUA, while the average commanding area for
water users’ association federation (WUAF) at 1,000 ha per WUAF. The estimated potential
level of participation for sustainable O&M have been based on farm budget analysis and
empirical estimate (expressed in terms of magnitude between 10 and 100), at the magnitude
of 20 for the land holding category of smaller or equal to 0.5 ha; the magnitude of 30 for the
land holding category of >0.5<1.00 ha; the magnitude of 80 for the land holding category of
>1.00<2.00 ha; and the magnitude of 95 for the land holding of >2.00<3.00 ha. The
remaining capacity to participate in an irrigated farming activity must be dedicated to non
agricultural employment in the urban areas (seasonal urbanization). it is apparent that the
WUA’s institution as amongst the important prerequisites for implementation of PIM suffers
from a number of nontechnical constraints among others: A lot of farmers are involved as the
member in the WUA under the small land holding condition. For illustration, a WUA with an
average land holding of 0.50 ha would compose of 200 farmer households working an area
of 100 ha; in addition to the average capacity to participate at the magnitude of 20 out of 100,
since they have to seek seasonal employment in the urban areas. For the national average of
land holding at 0.83 ha/farm household, a single WUA of 100 ha command area, would
involve about 120 farm households, with the capacity to participate at only about 30 out of
100, for they have to share their time for earning non farming extra income in the urban area.
During their absences, their participation (in person) in the routine irrigation management are
hardly possible, a size of irrigated farming organization to be diversified socioeconomic
conditions, with low level of potential participation, which is far from manageable. If we take
the optimum size of land owner (2.00 ha/farm household) as the determinant parameter for
establishing WUA, the number of members would be 50 farmers, which is reasonably
manageable, however, such as the optimum size of land holding only represent about 15% of
the total national irrigated agricultural land. Rationale of the Low Level of Participation that
the analysis results presented above, it is evident that the farmer’s participation in O&M of
irrigation is not merely the question of technical and economic but far from those matters,
there remains a complicated constrain on sociology-cultural as well as organizational
predicaments. The rationale of the current low participation of the farmer is not only because
of the farmers are unwilling participate, but it is quite a logic explanation that the farmer,
under the extremely small land ownership, would naturally set up his own priority in mind,

V - 165
whether to participate partially or seeking non-farm extra-incomes elsewhere. Public
participation from beginning until post Paselloreng reservoir construction will be
implemented conservation improvement work, reforestation and soil conservation practice for
anticipate next erosion and landslide disaster in the catchment area for (1) To facilitate
pertinent proses of land acquisition of Paselloreng people and resettlement site of Bekkae
village for construction of the project facilities in connection with Paselloreng dam (2) To
sacrifice among the Paselloreng people for advantage people in resettlement site. (3) To
encourage an integrated development activities for a better life of the project affected
population in the resettlement site during the dam construction. (4) To result the Paselloreng
reservoir will store 132 million m3 water to irrigate 7.000 Ha Gilirang Irrigation and
increasing of existing environment in the catchment area including for the regional economic
and community development.

2.5. Land and Water Conservation

The basic concept of land and water conservation is to treat the land and water in order to
yield optimum advantages for human life in the Gilirang river surrounding with minimum or
without negative impact on the river environment. Some basic requirement for human life
provided with fresh water for drinking, water for agriculture & fisheries, river navigation,
swimming, bathing & washing cloths, discharge of excess and waste water. Some negative
aspects of rivers to human life, for example, flooding and inundation, erosion and
sedimentation, hampering the sediment transport across the river. Furthermore, the land and
water conservation was is the promotion of the land and water advantage and minimize the
negative aspects barrier. In the Gilirang estuary where saline water penetrates up to 25 km
upstream, the construction of rubber dam or a tidal barrage should be constructed as a
desalination protection. Navigation and fisheries on the Gilirang lower reaches is proposed to
construct a barrage like Tempe barrage will be facilitating a navigation lock and fish ladder.
Gilirang flooding and inundation may be eliminated with some kind of structures, depend on
the river stretches and topographic condition as described that In the low land and flat area
flooding can be eliminated by widening of Gilirang river channels with the dykes, cut off
construction. It is a method to decrease in extent the flood discharge downstream of the
Paselloreng dam.

3. CONCLUSION AND RECOMMENDATION

In 2014, access road for preparation works of The Paselloreng irrigation dam is being
constructed. The reservoir will store water to provide for the development of 7.000 ha
gilirang irrigation. The project will contribute to strengthen the South Sulawesi to supply rice
to deficit region. The project will increase crop productivity and increasing the standard of
farmer’s living in Maningpajo, Sajoanging and Majjauleng Sub District of Wajo Regency
including to increase existing environmental in connection with the global climate change. It
is recommended that (1) the monitoring program for human resettlement and existing
environmental development will include surveys for establishment of benchmark information
as a basic information for evaluation made by the local government. (2) Implementing the
proposed development works should involved the displaced population to participate in
determination of the planning process, program implementation, in monitoring and evaluation
(3) the Land acquisition and resettlement plan must be implemented before the dam
construction starting (4) Environmental improvement work and conservation development
must be done urgently (5) Catchment area management for sustainable paselloreng irrigation
dam must be implemented to increase cropping in existing Gilirang irrigation.

V - 166
ACKNOWLEDMENT

Thanks to committee of the 82nd ICOLD Annual Meeting and symposium to approve and let this
paper to be presented in the meeting for transferring and sharing the dams in global environmental
challenges knowledge among participants.

REFERENCES:

A. Hafied A. Gany, (2007): Problems And Perspectives of Participatory Irrigation

Management Under The Small Land Holding Condition: With A Special Reference To
Indonesian Practice, ICID Publisher, Tehran, Iran
Anonym, (2012): No. 7 of 2004 Indonesia Law on Water Resources, DGWR Publisher,
Jakarta, Indonesia
DGWR, Yayasan Air Adi Eka and JICA, (2013): River Management in Indonesia,
Pompengan Jeneberang Large River Basin Organization, DGWR Publisher, Jakarta,
Indonesia
G.R. Hancock, (2013); A catchment scale assessment of increased rainfall and storm
intensity on erosion and sediment transport in Northern Australia, Elsevier
International Journal Publisher, San Diego, CA 92101, USA
Maher Owaiwi, and Mahmoud Whadeen, (2013), Assessment of Soil Erosion Risk within the
Hebron Catchment Area using GIS and RUSLE, Elsevier International Journal
Publisher, San Diego, CA 92101, USA
Nippon Koei Co., Ltd in associated with Local Consultant, (1999): Consulting Engineering
Services for Gilirang Irrigation Project, Engineering Report on Gilirang Irrigation
Project, Monitoring Program for Human Restlement, Nippon Koei Publisher,
Makassar, Indonesia
______, (2001): Consulting Engineering Services for Gilirang Irrigation Project, Design
Report on Gilirang Irrigation Project, Paselloreng Dam, Nippon Koei Publisher,
Makassar, Indonesia
______, (2003): Consulting Engineering Services for Final Report of Master Plan Study on
Integrated Development and Management of Walanae Cenranae River Basin, Nippon
Koei Publisher, Makassar, Indonesia
Titu Anghel and Sandu Todica, (2013); Quantitative Assessment of Soil Erosion using GIS
Empirical Methods. A Comparative Study between The Motru Mining Area And The
uceviła Catchment, Elsevier International Journal Publisher, San Diego, CA 92101,
USA

V - 167
 INTERNATIONAL SYMPOSIUM ON

Bali, Indonesia, June 1ST – 6TH , 2014

The basin management in the water resources development and construction

in Jiaoxi River basin of China
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
fffffjfjjfkkfjjj
Ye Shouren, Luo Jian, Zhang Yuanming & Wu Yongnian
2(14pt)
Taihu Basin Authority of Ministry of Water Resources, Shanghai,China
[email protected]

ABSTRACT:
Jiaoxi River basin is located in the border area of Zhejiang Province and Fujian Province where is
the typical subtropical monsoon climate area, and there is rich hydropower resources with
abundant rainfall in summer, so that the needs of hydropower resources development and water
supply condition improvement are pressing. For the rational development, efficient utilization as
well as effective protection of water resources, Taihu Basin Authority of Ministry of Water
Resources has formulated the Complicated Plan of Water Resources on Jiaoxi River Watershed.
The Plan, holding the scientific outlook on development and the discipline of harmony between
human and water, co-ordination and ecological protection, properly handles the relations between
basin and region, development and protection, upstream and downstream. The Plan considers a
variety of requirements such as flood control, water supply, electrogenerating, interbasin water
diversion and ecological environmental protection etc., and makes the rational distribution
program for hydropower cascade development and construction. The paper, taking the Plan as the
case, discusses the basin management issues during the hydropower development and construction
in Jiaoxi River watershed, so that it can contribute to the regulation or development of other cross-
provincial rivers.

Keywords: Jiaoxi river basin, dam, planning

1. GENERAL SITUATION OF JIAOXI RIVER BASIN

Jiaoxi River originates from Zhejiang Province, mid-down stream belongs to Fujian
Province. The total area of the basin is 5635 ㎞². There are two main branch called Dongxi
River and Xixi River upstream of Jiaoxi River basin. Upwards form Jiaoxidu, Dongxi
River dispart to two branches again, one is called Zhetai River, another is called Shoutai
River. These two branches are both trans-boundary river of Zhejiang and Fujian. The main
stream of Jiaoxi River passes through Fu’an city from north to south, and joins into sea at
Baimamen. Jiaoxi River basin is comparatively depressed area, the population is less, the
economy is underdeveloped relatively. Upwards from Dongkou of Dongxi River, the main
involved districts are Taishun County of Zhejiang Province, Fu’an, Zherong, Shouning,
and Fuding county of Fujian Province.

V - 168
The precipitation is flush in Jiaoxi River Basin, the multi-year average quantity of it is
about 1748mm. The surface flow comes from precipitation mainly, and the multi-year
average quantity of it is about 4.1 billion m3. Dominated by subtropical monsoon climate,
the yearly distribution of the runoff has distinct change. The amount of runoff in flood
season(Apr.-Sept.) accounts for 75% of entire year, in low water season(Oct.-Mar.)
accounts for 25%. The landform in upstream area of Jiaoxi River Basin is mountains and
canyons. Affected by the landform, the precipitation increases from south to north-west,
and the upstream of Dongxi River is the rainstorm center, the maximum multi-year average
quantity of precipitation is more than 2100mm. Big rainstorm occurs frequently in flood
season every year. Rainstorm is the main reason forming flood.

Jiaoxi River Basin has plenty of water power resources. But because of having different
opinion on how to fathering and exploiting the river between Zhejiang Province and Fujian
Province, the water resources developing level is relative low on the river, especially on
the trans-boundary reach. The main challenges in Jiaoxi are as following.

1.1. Flood threat in downstream

Because of Affected by typhoon, flood occurs frequently in Jiaoxi River Basin. Fuan
strengthened to build the flood control works in recent years. The flood control dike of
Fuan city has been built, and the flood control capability has been increased greatly. But,
because of dense rainstorm and high peak flood runoff, only depend on the dike to defend
from flood can’t fulfil the demand of flood control of Fuan city. An upstream flood control
reservoir is needed for adjusting and storing the flood to cut the peak runoff of the flood.
Jiaoxi Programming needs to study suitable position and size of the reservoir.

1.2. Water shortage in the near-sea area

Fuding, located near the sea, its economy grows very fast these years. At the same time,
because of absence of landform condition to constructing reservoirs, the available water
resources in this area is relatively less. This situation restricts the economy developing, so
Fuding asks to transfer proper amount of water from water-rich area around it to fulfil the
increasing water demand of civil life, agriculture and industry. Guanyangxi River, the
upstream branch of Jiaoxi River, has about 110km2 catchment belongs to Fuding County.
So Fuding asks to transfer water from Guanyangxi River. But cross-basin water transfer is
a complicated and sensitive, the project can’t be put in practice because of the dispute from
Zhejiang. Jiaoxi Programming needs to study the transferring scale deeply.

1.3. Improper exploitation and low efficiency use of hydropower resources

Because of lack of united programming, the water energy exploitation on the trans-
boundary reach of Jiaoxi River is out of order, and large-scale exploitation can’t be
implemented. Plenty of water power resources loss along with the time going. Jiaoxi
Programming needs to study the reasonable layout of the dam to guarantee the resources

V - 169
taken full advantage, and the resources allotted fairly to avoid the quarrel between two
sides of the river.

2. DAM CONSTRUCTION SCHEME OF JIAOXI PROGRAMMING

Based on the characters and the actual conditions of the reach of Jiaoxi River and this
region, the programming takes the combination of seeking benefit and avoiding disaster as
the principle. Under the precondition of insuring the security of water supply and flood
control downstream, the programming makes the most use of water power resources,
exploits water power plants which have preferable regulating capability to improve the
electric energy quality and benefit of water power plants.

A step exploitation scheme with water plants named Guihu, Youyi, Jiajiadu, Xiangxi,
Shangbaishi, Caihong, Donghu, Dongkou and the project of ‘Guanyangxi cross-basin
water transfer’(Xitou reservoir) has been fixed. Among these projects, Shangbaishi water
conservancy hinge exerts the function of flood control and energy generation, Youyi water
conservancy hinge exerts the function of water supply and enery generation, Guihu,
Jiajiadu, Xaingxi, Caihong, Donghu, Dongkou exert enery generation mainly. The total
amount of capacity of all the reservoirs is 550 million m3. The total hydrodynamic load is
35
problem of Fuan city and water supply problem of Taishun and Fuding meanwhile.

Shangbaishi water conservancy hinge locates in the joint point of Zhetai River and Shoutai
River. It has 314 million m3 reservoir capacity, can control the flood discharge from
rainstorm center in upstream, and be in charge of flood regulation task. The main measure
is: setting up special flood control capacity to regulating the runoff, according to the
forecast of runoff from interzone from Shangbaishi down to Fu’an city, regulating the
discharge runoff, considering the transmit time from Shangbaishi to Fu’an city, keeps the
total runoff run through Fu’an city under 11300 m3/s. Compare to the natural situation,
22% of the maximum discharge at the dam site of Shangbaishi can be cut. The multi-year
average flood control benefit is about 21.5 million Yuan.

The programming adopts the exploitation means of reach partition, Fuajian exploits
Shangbaishi, Zhejiang exploits Guihu, Xiangxi, and Jiajiadu, and takes the normal water
level of Shangbaishi reservoir as the resources division point between Zhejiang and Fujian
province. Via fixing on rights of two provinces according to the catchment area of each
province, making an integrated consideration on the quantity of electricity, guaranteed
hydrodynamic load, and the investment, the programming fixes the normal water level of
Shangbaishi reservoir.

The project ‘Guanyangxi cross-basin water transfer’ will construct an adjust reservoir
named Xitou reservoir with 11 million m3 capacity. With the precondition of discharging

V - 170
the base flow to guarantee the usage of civil life, production, and ecosystem, remainder
water can be transferred to Wenxi village through the 15.5km length canel to produce
water energy, the amount of it is 95 million kWh. The water after power generation is by
way of water supply sources of Fuding County, and exerts multi-benefit.

3. BASIN MANAGEMENT IN THE WATER RESOURCES DEVELOPMENT AND

CONSTRUCTION IN JIAOXI BASIN

3.1. The basin management organ has responsible for mediation of cross-boundary
river conflict.

The basin management organ is the resident agency of water administrative department of
center government, and exerts managerial responsibilities in key rivers and lakes.
Mediation for the water conservancy conflicts is one of the important duties. The basin
management organ, without the own benefit, could consider the development, the
regulation and the protection as a whole to mediate the profits of all parties with justice and
equity. Over the years, the organ has solved a large number of water disputes effectively,
and promoted the orderly inter-provincial water resources development.

3.2. The water resources programming of cross-boundary rivers guarantees the

scientific and organized development.

Before 2002, for the development and utilization in Jiaoxi basin Zhejiang and Fujian have
prepared the respective plans, while cannot reach an agreement. Since TBA organized to
formulate the water resources programming of cross-boundary rivers, TBA had carried
through much works such as analysis, demonstration, consultation and coordination. The
achievement of the programming has gain the general consent from Zhejiang and Fujian,
thinking that the programming is compliance with the principles of legislation, science and
democracy, and the achievements are accord with the common benefits of people of the
two provinces. The programming well balances the relations of basin and region, risk
avoiding and interests pursuing, development and protection, and also balance the multi
function of flood control, water supply, power generation and eco-environment protection.

3.2.1. Priority principle of flood control

Shangbaishi water control project has the dedicated flood control capacity for control the
flood from upstream.

3.2.2 Principle of efficient

Optimization of Engineering project layout has enhanced the adjusting capacity for flood
and enlarged the volume of water resource, and guaranteed the efficient usage of water
resource and maximize of multi-benefits.

V - 171
3.2.3 Principle of justice and reasonable
Clear-cut division of decollation point with the sectionalized developing pattern has
distributed hydropower resource fairly.

3.2.4 Principle of optimizing configuration

Guanyangxi cross-basin water diversion project has preferably solved the water shortage in
Fuding coastal region, in the precondition of not destroy the water ecology in Jiaoxi basin.

3.2.5 Principle of eco environment protection

Analysis have been made for the influences against environment which are caused by
measures of water resources protection and programming implment, and discussions have
been made for the response strategies.

Special item of Guanyangxi cross-basin water diversion project has been found to research
the relevant topics such as the influences to the water allocation, the necessary of cross-
basin water diversion, the proper scale of water diversion, the influences to the water usage
in the downstream and to the hydropower development, and the compensatory approaches.
According with the approved programming, the measures are been pushed forward in good
order in Zhejiang and Fujian.

3.3. Supervision and administration duty should be performed by basin management

organ.

Because the most sub-projects in Jiaoxi basin are cross-provincial projects, so the
implementing situation should be supervised by basin management organ. Therefore, as
stipulated in the files approved by MWR and the provincial governments of Zhejiang and
Fujian, TBA has to strengthen the guidance, the coordination and the supervision in the
programming implement. The water engineering project planning agreement, the justifying
of water resources and the water drawing licensing involving with cross provincial and
cross-basin water diversion should be approved by TBA. Especially for enhance the
surveillance on ecological basic flow discharging, TBA is responsible for the review on
basic flow discharging facility designing scheme, and also for the supervision of the
construction and operation.

3.4. Strong and deeper technique analysis is the foundation of the cross-provincial
basin management.

Authority of basin management organ is also shown that with the steady technique analysis
to convince the both sides, and mediate scientifically. During the programming
formulation, TBA insists to take technique analysis before action, and take deep research
on the existing files and the proposals by the both local governments. To check on the key
technique index, TBA invites the relevant consultants to join to the research work. TBA is
emboldened on the base of technique works, so that the coordination and the programming
are smoothly implemented.

V - 172
 APPLICATION OF SIMPLE HYDROLOGIC MODEL
FOR PREDICTING THE EFFECT OF WATER CONSERVATION
MEASURES AT THE UPPER WATERSHED OF DAM 1
Chandra Setyawan
Faculty of Agricultural Technology, Gadjah Mada University, Yogyakarta.
[email protected]

Sahid Susanto
Faculty of Agricultural Technology, Gadjah Mada University, Yogyakarta.
[email protected]

ABSTRACT

The recent condition of degraded land at the upper watershed of dam in Java is increasing significantly.
It implies on sedimentation problem flowing to the dam. The aim of study was directed to apply a simple
method for simulating the effects of water resource conservation measures. Hydrologic model of Mock
was proposed for the study. The model is basically rainfall-run-off relationship containing three tanks
arranged in vertical position. Finding optimal six parameters in the model were conducted by trial and
error. Two degraded upper watersheds of dam at Central Java were used as the study. This area study
falls into tropical monsoon region. The results show that the model is sensitive enough to predict
hydrologic regime in the upper watershed of dams studied. The model was then used to simulate the
effects of water resource conservation measures applied in the upper watersheds. The simulations were
directed to design optimal vegetation landscape in the form some realistic scenarios of land use pattern.
Comparing with the existing condition, the simulation results showed that they give significantly effect
in reducing surface run off coefficient (from 10% to 20%), sedimentation (15%-20%), increasing water
resources availability (5%-15%).

Key words: hydrologic model, water resource conservation measures, realistic scenarios of land use
pattern

1. INTRODUCTION

The condition of degraded land at upper watershed of the dams in Java is increasing
significantly. Climate at Java Island fall into tropical monsoon region. Land degradation form
of erosion and sedimentation caused by excessive farming activity at the upper of watersheds
without conservation principles. Its gives effect on erosion and sedimentation flowing to the
dam and decreasing its function in providing water for various needs and electric source of
power plants. The aim of study was directed to apply a simple hydrologic model for simulating
water flowing to the dam as implication of the effect of water resource conservation measures
in these two upper watershed of dam.

1
This paper is to be presented at the 82th Annual Meeting of International Commissions on Large Dams
(ICOLD) on June 1-6, 2014 at Bali Nusa Dua Conference Center, Bali

V - 173
2. METHOD

A simple of hydrologic model of Mock was chosen in this study. Basically, the model is rainfall-
runoff relationship analogized by tanks concept (Figure 1). The model contains three artificial
tanks arranged in vertical position (Brooks, 2003 and Mock, 1973). There are six parameters in
the model. Trial and error technique is used to find optimal parameters of the model. Some
statistic and graphical criteria is apply to achieve the best performance of the model.
P AET = CF x Eto GWS = 0.5x(1+k)xI+(kxIGWS)
ER = P - AET IGWS = GWSI-1
WS = ER - ∆SM BF = I - ∆S
AET ISM = SMI-1 ∆S = GWS - IGWS
I = IC x WS QTot = DRO + BF

Where:
P = rainfall (mm)
CF = crop factor
Eto = potential evapotranspiration (mm)
WS DRO (WS-I) AET = actual evapotranspiration (mm)
∆SM ER = excess rainfall (mm)
SMC DRO = direct runoff (mm)
ISM ∆S = change of ground water volume
∆SM = change of soil moisture (mm)
SMC = soil moisture capacity (mm)
ISM = initial soil moisture (mm)
WS = water storage (mm)
I IGWS = initial ground water storage
∆S GWS = ground water storage (mm)
GWS IC = infiltration coeficient
IGWS BF = base flow
BF=I-∆S I = infiltration

3. RESULT Figure 1. Simple Hydrologic Model of of Mock

The hydrologic model is then applied to simulate the effects of hydrologic regime due to water
resource conservation activities at the upper watershed of dams. The simulation is conducted by
manipulation of land use pattern in the form of some realistic scenarios. Principally the scenario
is oriented to convert open agricultural cultivation land at the upper watershed of dam to become
forest area, mainly at high slope (more than 45%) and high erosion area. Three parameters are
applied in the simulation. Those are: (a) change in hydrologic regime in the form of run off
coefficient and water availability, (b) erosion and, (c) sedimentation.

3. LOCATION OF STUDY

Two degraded upper watersheds of dam at Central Java were used as the study area. Those are
upper watershed of Sempor dam and upper watershed of Wadaslintang dam having an area of
44.15 km2 and 192.53 km2, respectively. Water capacity of Sempor dam and Wadaslintang dam is
52 million m3 and 443 million m3, respectively. The irrigation common areas serving by the
Sempor dam is 6.485 ha and for Wadaslintang dam is 32.064 ha (Figure 2). Besides supplying
water for irrigation, this two dams have also play role in flood control, fisheries and electric
source of power plants. Landscape at this two location of study are dominated by open
agricultural cultivated land with 70%-80% of total area, even at the high slope land (Figure 3a
and 3b).

V - 174
 Farmers who lives at this upper
Sempor Wadaslintang watershed area of dam are very
Dam Dam
dependents to agricultural cultivation
land for upland crops. The farmers
cultivate upland crops such as corn,
soybean, peanuts as well as cassava.
This upland crops are also planted
even in the area with slope more 45%.
It implies in increasing erosion at
Central Java surface land even until high risk
erosion. Further implication is
source: bbws.so increasing sedimentation flowing to
Figure 2. Location and land use of area study the dams.

The upper
Landscape watershed of
Sempor dam

Open 3,718 ha
cultivated (84.2%)
upland crop
Forest land 164 ha (3.7%)
Sempor Dam
Settlement 278 ha (6.3%)
Water 256 ha (5.8%)
storage

Figure 3a. Landscape at the upper watershed Sempor dam

The upper
watershed of
Landscape
Wadaslintang
dam
Open
cultivated 13,936 ha
upland crop (72.4%)

Forest land 1,524 ha (7.9%)

Settlement 2,398 ha
Wadaslintang Dam (12.5%)
Water 1,395 ha (7.2%)
storage

Figure 3b. Landscape at the upper watershed Wadaslintang dam

4. RESULTS AND DISCUSSION

V - 175
4.1. Results

Data at the year of 2000 and 2001 both at the upper watershed of Sempor dam and Wadaslintang
dam was use as calibration and verification process of the hydrologic model, respectively.
Optimal parameters result in calibration process is tabulated at Table 1, both for the two location
study. With the optimal parameters, statistically the model gives the best performance with value
of correlation coefficient (R) and volumetric error (VE) for upper watershed of Sempor dam is
0.98 and 34%, respectively. For the upper watershed of Wadaslintang dam gives 0.93 (R) and
20% (VE), respectively. Graphically, using scatter and time series diagram show that observed
and predicted discharge is relatively matched. Both for the two location study. Figure 4a and
4b shows in scatter diagram for the upper watershed of Sempor dam and Wadaslintang dam,
respectively. In the form of time series diagram is presented in Figure 5a and 5b for the upper
watershed of Sempor dam and Wadaslintang dam, respectively.

Table 1. Optimal parameter of calibration process

Parameter unit Simbol Sempor Wadaslintang

1. Area of watershed km2 A 44.15 192.53
2. Infiltration coefficient in rainy season - WIC 0.5 0.1
3. Infiltration coefficient in dry season - DIC 0.5 0.4
4. Initial soil moisture (mm) ISM 200 100
5. Soil moisture capacity (mm) SMC 200 250
6. Initial groundwater storage (mm) IGWS 1500 1500
7. Groundwater recession constant - K 0.99 0.94

a) Upper watershed of Sempor dam, Year 2000 b) Upper watershed Wadaslintang dam, Year 2000

8 50
Q Calculated (m3/s)

7
Q Calculated (m3/s)

40
6
5 30
4
20
3
2 10
1
0 0
0 2 4 6 8 10 12 0 10 20 30 40 50
Q Observed (m3/s) Q Observed (m3/s)

Figure 4. Scatter diagram in calibration process, Year 2000: (a) at the upper watershed of
Sempor dam, (b) at the upper watershed of Wadaslintang dam

V - 176
 (a) Upper watershed of Sempor dam, Year 2000 (b) Upper watershed of Wadaslintang dam, Year 2000
50
12
10 Q cal Q Obs 40 Qcal Qobs

Discharge, Q (m3/s)
Discharge, Q (m3/s)

8 30
6
20
4
2
10

0 0
Jan IFeb Mar
I Apr
I M
I ei IJun IJul IAgt ISep IOkt N
I ov D
I es I Jan IFeb M
I ar A
I pr M
I ay JI un I Jul IAug ISep IOct INov D
I ec I

Figure 5. Time series diagram in calibration process, Year 2000: (a) at the upper watershed of
Sempor dam, (b) at the upper watershed of Wadaslintang dam

In the model verification process, discharge was simulated by the optimal parameters model
obtained from the calibration process. The hydrologic data of the year 2001 was used in the
verification. Verification process was also conducted for both at the two location studies. The
result show that statistically the model gives the best performance with value of correlation
coefficient (R) and volumetric error (VE) for the upper watershed of Sempor dam is 0.91 and
29%, respectively. For the upper watershed of Wadaslintang dam gives 0.81 (R) and 29% (VE),
respectively.

In the form of scatter diagrams are presented in Figure 6a and 6b for the two location of studies.
Figure 7a and 7b show in the form of time series diagram. From these calibration and
verification process, the results prove the model is sensitive enough to simulate discharge in
tropical monsoon region.

Figure 6. Scatter diagram in verification process, Year 2001: (a) at the upper watershed of
Sempor dam, (b) at the upper watershed of Wadaslintang dam
Application of hydrologic model in executing the realistic scenarios at the two upper watershed
studies gave results, as follows (Table 2):
a) reducing surface run off coefficient (from 10% to 20%),
b) reducing sedimentation flowing to the dams (15%-20%), and

V - 177
c) increasing water resources availability (5%-15%).

Figure 7. Time series diagram in verification process, Year 2001: (a) at the upper
watershed of Sempor dam, (b) at the upper watershed of Wadaslintang dam

Table 2. Simulation result with realistic scenarios of landscape in the form of landuse pattern
at the upper of watershed of Wadaslintang and Sempor dam

Realistic Scenarios of Land Hydrology, Erosion Sedimentation Water availability

Use Pattern (run off coef.) (mm/thn) (mm/thn)
Existing condition 0.27-0.50 19.5-21.0 5.5-6.0 Surplus

Scenario I 0.24-0.25 9.85-12.5 2.85-3.5 Surplus

70-80% LcDi & 20-30% LcPi
Scenario II 0.18-0.21 7.34-10.5 1.86-2.5 Surplus
60-70% LcDi & 30-40% LcPi
Scenario III 0.13-0.15 4.77-5.1 0.94-1.2 Surplus
50-60% LcDi & 40-50% LcPi
Notes: LcDi= Land Covering Production Index; LcPi= Land Covering Permanent Index

4.2. Discussion

From the results as mentioned above prove that the simple hydrologic model can be applied to
predict the effect of water conservation meaures at the two upper watershed of dams in the form
of convertion of vegetation landscape in three realistic scenarios. To realized the scenario
depends on social and economic charateristics that is different from region to region. However,
the social and economic charateristics generally are in similar conditions at the degraded upper
watershed of dam. The conditions are characterized by poor economic and simple social way of
life with low level education.

As metioned above that landscape at the two upper watersheds studied is dominated (more than
70% of total upper watershed area) by open area of agricultural cultivation. This is due to
densely populated area with more or less 700 people per square km. The population mainly
works in subsistence agricultural cultivation with small land ownership. It gives an

V - 178
environmental stress. Generally it happens at mountainous area of upper watershed at Java
Island.

Implementation of the best realistic scenarios should be considered not only from the aspect of
land and water conservation principles but more from the economic and social characteristics
of people who live and strongly dependent to land for agricultural cultivation in the upper
watershed of dam.

4. CONCLUSSION

a) The Mock simple hydrologic model has been applied at watershed in tropical monsoon
region. The upper watershed of Sempor and Wadaslintang dam which are located at Central
Java was used as location study. The result shows that the model gives good performance
and sensitive enough to simulate discharge at the two location study.

b) The model then was applied to simulate discharge flowing to the dam as the response of
water conservation measures at the two upper watershed of dams. The water conservation
measure was simplified by realistic scenarios with modification of vegetation landscape in
the form of land use pattern in the upper watershed of dams. The scenarios were directed to
convert open agricultural cultivation land at the upper watershed of dam to become forest
area. Using three parameters of water regime, i.e: (a) change in hydrologic regime in the
form of run off coefficient and water availability, (b) erosion and, (c) sedimentation, the
significant effects of water conservation measure can be evaluated. Comparing with the
existing condition, the simulation results showed that they give significantly effect in reducing
surface run off coefficient (from 10% to 20%), sedimentation (15%-20%), increasing water
resources availability (5%-15%).

c) Implementation of the best realistic scenarios should be considered not only from the aspect
of land and water conservation principles but also from the economic and social
characteristics of people who live and strongly dependent to land for agricultural cultivation
in the upper watershed of dams.

ACKNOWLEDMENT

This paper could not be realized without financial support from Main Office of Serayu-Opak River Basin
Development, Yogyakarta. During the research has been supported by our coleague, namely Sukirno
and some undergraduate students of Dept.of Agric. Engineering Fac.of Agric. Technology UGM. For
that reason, sincerely thank is addressed.

REFERENCE
Brooks, K.N. 2003. Hydrology and the Management of Watershed. Iowa State University Press.
Mock, F.J., 1973. Land Capability Appraisal Indonesia. Water Availability Appraisal. Report
Prepared for the Land Capability Appraisal Project. Bogor-Indonesia,1973.

V - 179