Optimal reservoir operation for water supply in dry season the case study of cua dat reservoir in the ma chu river basin

MINISTRY OF AGRICULTURE AND RURAL DEVELOPMENT
WATER RESOURCES UNIVERSITY
**************

OPTIMAL RESERVOIR OPERATION FOR WATER SUPPLY
IN DRY SEASON: THE CASE STUDY OF CUA DAT
RESERVOIR IN THE MA-CHU RIVER BASIN

TRINH XUAN MANH
MSc Thesis
September 2014


Optimal reservoir operation for water supply in dry season: the
case study of Cua Dat reservoir in the Ma – Chu river basin

Master of science thesis

By

Trinh Xuan Manh

Supervisor

Dr. Nguyen Mai Dang (WRU)

Mentors

HA NOI
September 2014
This research is finished for the partial fulfillment of requirements for the Master of science

degree at Water Resources University, Ha Noi, Viet Nam


i
Abstract
Water supply of reservoirs and especially reservoirs used for irrigation,
hydropower, aquaculture, navigation, environment…in the dry season are often
troubled due to increasing water demands according to the economic development and
society, while the flow to the reservoir is limited. In recent years, the depletion of the
river flow during the dry season occurs more frequently and at a more intense level.
This is partly due to forest coverage reduction in the upstream of river basins, and
partly due to the effects of climate change.
Hence, computation of the optimum water supply of reservoir for the water
demands in the dry season is needed. This study presents the initial research on
applying Fuzzy Logic Algorithm for optimal operation of water supply in the dry
season of 2011-2012 of the Cua Dat Reservoir in the Chu River basin, Thanh Hoa
province. The Cua Dat Reservoir is a multi-purpose reservoir for the following tasks:
flood prevention, water supply, irrigation, power generation, and environmental flows.
In addition, MIKE 11 model is also used to simulate the release from the reservoir to
the downstream to evaluate the efficiency of the optimal method.
The research used Fuzzy Logic algorithm based on the rule, the principle of "IF
- THEN" and built the membership functions for the input variables: water level,
inflow to the reservoir, the water demands, and discharge from the reservoir. It is
developed for the Fuzzy operating systems for the Cua Dat Reservoir and is meant to
determine the optimal discharge process in case of shortage of water in the dry season.
Inflows, releases and water levels of the Cua Dat Reservoir were collected from actual
operation of the reservoir. For water demand of stakeholders, the author determined
that the total water demand for whole area was about 4547 Mi.m3. For hydropower
based energy production water is used at the largest rate (67% of total water demand),
while domestic purposes water is obtained smallest rate of water use of the Cua Dat

Reservoir.
Finally, the results from optimal method, the reservoir can meet 80% of water
demand more than actual release throughout the dry season of 2011-2012. The initial
research has been successful and the results showed that this method can be applied
well to the optimal reservoir operation in Vietnam.
Key words: Cua Dat, reservoir operation, optimization, Fuzzy Logic, water
demand, Fuzzy rule, MIKE 11 model.
Trinh Xuan Manh
MSc Thesis


ii

Acknowledgement

First of all, I would like to give a big thank to all people who have supported
and assisted me during the Master Thesis Research. Thanks for their support,
encouragement and guidance that allowed me to complete this study in time.
Especially, I would like to express my appreciation to Dr. Nguyen Mai Dang,
my supervisor, for his unlimited encouragement, guidance, comments and technical
supports on the Fuzzy Logic approach and other models as well as the thesis writing
process from the beginning of the thesis research.
I would like to thank NICHE-VNM-106 project from the Government of the
Netherlands for their financial support during the MSc study in the ThuyLoi
University. I thank to Mrs. Hoang Nguyet Minh and Mrs. Vu Thi Thuy Ngan who
made a linkage between me and NICHE. I also would like to thank Assoc. Prof. Dr.
Nguyen Thu Hien, Dean of the Faculty of Water Resources Engineering, for her help
and comments during the Master study in the ThuyLoi University.
I wish to thank Dr. Ilyas Masih and Ms. Martine Rutten for their feedback,
references and support from the proposal process.

I also wish to thank Mrs. Mariette Van Tilburg, my English teacher, for her
comments and support from the final thesis report.
I also want to thank the ThuyLoi University (TLU), Song Chu Irrigation
Company, National center for Hydro-Meteorological Service (HMS) for providing me
very useful data sets.
Thanks to all of my colleagues at the HaNoi University of Natural Resources
and Environment in Vietnam for your assistance in the last two years. You will always
be in my mind.
Last but not least, I want to take this opportunity to show my appreciation to my
family, my close friends for their inspiration and support throughout my life; this
research is simply impossible without you.

Trinh Xuan Manh
MSc Thesis


iii

Table of Contents
CHAPTER I: INTRODUCTION ................................................................................. 1
I.1. Background ............................................................................................................1
I.2. Problem statement ..................................................................................................2
I.3. Objectives and Research questions ........................................................................3
I.3.1. Objectives of the study .................................................................................. 3
I.3.2. Research Questions ....................................................................................... 3
I.4. Structure of the thesis .............................................................................................3
CHAPTER II: LITERATURE REVIEW ................................................................... 5
II.1. Studies on reservoir operation using optimal theory ............................................5
II.2. Fuzzy logic theory ................................................................................................8
II.3. Overview of hydraulic and hydrological modeling ..............................................9

II.4. MIKE model .......................................................................................................11
CHAPTER III: THE STUDY AREA ........................................................................ 13
III.1. Description of the study area .............................................................................13
III.1.1. Location of the study area ......................................................................... 13
III.1.2. River network ........................................................................................... 14
III.1.3. Topographical characteristics ................................................................... 16
III.1.4. Geological, land and vegetable characteristics ......................................... 18
III.2. Climate and hydrological condition ..................................................................18
III.2.1. Climate condition ...................................................................................... 18
III.2.2. Hydrological condition ............................................................................. 23
III.3. Population and economic characteristics ..........................................................23
III.3.1. Population of the study area ..................................................................... 23
III.3.2. Economic characteristics .......................................................................... 24
III.4. Description of the Cua Dat Reservoir ...............................................................24
CHAPTER IV: DATA AND METHODOLOGY..................................................... 29
IV.1. Data collection ...................................................................................................29
IV.1.1. Meteorological data .................................................................................. 30
IV.1.2. Hydrological data ..................................................................................... 32
IV.1.3. Cua Dat reservoir operation data .............................................................. 34
IV.1.4. Determining total water demand .............................................................. 35
IV.2. Optimal analysis and Fuzzy logic approach for Reservoir operation ...............50
IV.2.1. Methods using in optimal reservoir operation .......................................... 50
IV.2.2. Objective functions and constraints.......................................................... 53
IV.2.3 Using Fuzzy logic technique to optimize the Cua Dat reservoir operation54
IV.3. Hydraulic and hydrological model setup ..........................................................62
IV.3.1. Determination of the model inputs ........................................................... 62
IV.3.2. Model setup .............................................................................................. 63
IV.3.3. Model calibration and validation .............................................................. 65
Trinh Xuan Manh
MSc Thesis

iv
CHAPTER V: RESULTS AND DISCUSSIONS ...................................................... 73
V.1. Optimizing the Cua Dat reservoir operation ......................................................73
V.2. Routing the release to the downstream ...............................................................74
CHAPTER VI: CONCLUSIONS AND RECOMMENDATIONS ......................... 77
VI.1. Conclusions .......................................................................................................77
VI.2. Recommendations .............................................................................................78
REFERENCES ............................................................................................................ 80
APPENDICES .................................................................................................................i

Trinh Xuan Manh
MSc Thesis


v
List of Figures
Figure 2-1: Relationship between the various representations of a model....................10
Figure 3-1: Location of study in the Thanh Hoa province in Viet Nam .......................13
Figure 3-2: Ma – Chu River Network in Viet Nam.......................................................16
Figure 3-3: Digital Elevation Model (DEM) of Thanh Hoa province ..........................17
Figure 3-4: The location of the Cua Dat Reservoir on Ma-Chu river system ...............26
Figure 3-5: The main dam of the Cua Dat Reservoir ....................................................28
Figure 3-6: The spillway of the Cua Dat Reservoir ......................................................28
Figure 3-7: The storage of the Cua Dat Reservoir ........................................................28
Figure 3-8: The intake tower of the Cua Dat Reservoir ................................................28
Figure 3-9: The gate of spillway of the Cua Dat Reservoir ..........................................28
Figure 3-10: The Bai Thuong weir ................................................................................28
Figure 4-1: Distribution of monthly rainfall pattern at Thanh Hoa station ...................30

Figure 4-2: Distribution of monthly air temperature at Thanh Hoa station 31
Figure 4-3: Distribution of monthly average evaporation at Thanh Hoa station in 2011
& 2012 ............................................................................................................................31
Figure 4-4: Distribution of relative humidity at Thanh Hoa station in 2011 & 2012 ...32
Figure 4-5: Annual discharge of the Cam Thuy and Cua Dat station ...........................33
Figure 4-6: Schematization of hydrological station network ........................................34
Figure 4-7: Monthly average discharge of Turbin of hydropower plant in years of
2011, 2012 and 2013 ......................................................................................................35
Figure 4-8: Inflow discharge of the Cua Dat reservoir in 2011 and 2012 .....................35
Figure 4-9: Seasonal period and chart of water requirement of Spring paddy in 2011 39
Figure 4-10: Seasonal period and chart of water requirement of winter paddy in 2011
........................................................................................................................................41
Figure 4-11: Seasonal period and chart of water requirement of sugar cane in 2011 ...42
Figure 4-12: Water use structure of whole downstream area of the Cua Dat reservoir in
2011 ................................................................................................................................48
Figure 4-13: General flow chart of optimal reservoir operation in dry season .............52
Figure 4-14: Fuzzy inference system for Fuzzy Mamdani ............................................56
Figure 4-15: Transformation of input variable to membership value ...........................57
Figure 4-16: Membership function for reservoir level for Fuzzy Mamdani model ......58
Figure 4-17: Membership function for inflow for Fuzzy Mamdani model ...................58
Figure 4-18: Membership function for water demand for Fuzzy Mamdani model.......59
Figure 4-19: Membership function for release for Fuzzy Mamdani model ..................59
Figure 4-20: Fuzzy rules base for operation of Cua Dat reservoir ................................60
Trinh Xuan Manh
MSc Thesis


vi
Figure 4-21: Process of application, implication and aggregation ................................61
Figure 4-22: Hydraulic network of the Ma – Chu river basin .......................................65

Figure 4-23: Observed and simulated hydrograph at Cua Dat station in 2006 .............67
Figure 4-24: Observed and simulated hydrograph at Cua Dat station in 2008 .............68
Figure 4-25: Observed and simulated hydrograph of water level at Ly Nhan Station in
2006 ................................................................................................................................69
Figure 4-26: Observed and simulated hydrograph of water level at Xuan Khanh Station
in 2006 ............................................................................................................................69
Figure 4-27: Observed and simulated hydrograph of water level at Giang Station in
2006 ................................................................................................................................70
Figure 4-28: Observed and simulated hydrograph of water level at Ly Nhan Station in
2008 ................................................................................................................................71
Figure 4-29: Observed and simulated hydrograph of water level at Xuan Khanh Station
in 2006 ............................................................................................................................71
Figure 4-30: Observed and simulated hydrograph of water level at Giang Station in
2006 ................................................................................................................................72
Figure 4-31: Structure of fuzzy system for Cua Dat reservoir ......................................73
Figure 4-32: Comparison of water demand and fuzzy and actual releases ...................74
Figure 5-1: Hydrograph of optimal operation at the Bai Thuong weir .........................75
Figure 5-2: Hydrograph of optimal operation at the Xuan Khanh station ....................75
Figure 5-3: Hydrograph of optimal operation at the Giang station ...............................76

Trinh Xuan Manh
MSc Thesis


vii
List of Tables
Table 3-1: Distribution of natural areas according to provincial border of the Ma river
basin (ha) ........................................................................................................................14
Table 3-2: Characteristics of river shape of some large tributaries ...............................15
Table 3-3: Average annual rainfall for many years at some stations of the Ma river

basin ................................................................................................................................19
Table 3-4: Annual rainfall characteristics ......................................................................20
Table 3-5: Monthly and annual wind speed at some stations of the Ma river basin (m/s)
........................................................................................................................................21
Table 3-6: Average monthly temperature for many years at some stations ..................22
Table 3-7: Monthly average evaporation of some stations of the Ma River Basin .......22
Table 3-8: Some main parameters of the Cua Dat Reservoir ........................................25
Table 4-1: Kinds of data have been used in the study ...................................................29
Table 4-2: Crop distribution of different cultivated area in downstream of the Cua Dat
reservoir ..........................................................................................................................36
Table 4-3: Plant coefficients of paddy ...........................................................................39
Table 4-4: Plant coefficients of other plants ..................................................................39
Table 4-5: Water requirement of Spring paddy in 2011 ................................................40
Table 4-6: Water requirement of winter paddy in 2011 ................................................41
Table 4-7: Water requirement of sugar cane in 2011 ....................................................42
Table 4-8: Monthly water demand of agriculture of whole area in the Cua Dat reservoir
downstream in 2011 .......................................................................................................44
Table 4-9: Water demand of industrial production at downstream of the Cua Dat
reservoir ..........................................................................................................................45
Table 4-10: Domestic water demand of downstream area ............................................46
Table 4-11: Structure of water use of whole area in 2011 .............................................48
Table 4-12: Water demands and inflows in ten-day period in 2011 .............................49
Table 4-13: List of tributary basin on the Ma – Chu river basin ...................................64
Table 4-14: Results of MIKE 11HD model calibration at Ma-Chu river basin in 2006
........................................................................................................................................70
Table 4-15: Results of MIKE 11HD model validation at the Ma-Chu river basin in
2008 ................................................................................................................................72
Table 4-16: The NASH for calculation of alternatives ..................................................74
Table 5-1: Flow characteristics at the Chu River downstream using optimal operation
........................................................................................................................................76

Trinh Xuan Manh
MSc Thesis


1

CHAPTER I
INTRODUCTION
I.1. Background
Reservoirs play an important role in the development of many countries.
Nowadays, there are many reservoirs and dams which were built in many developing
countries for various purposes, for example, water supply, flood control, electric
generation, environment and recreation…However, in 18th Century reservoirs were
built to supply water, flood control and navigation as the main purposes, after that
reservoirs were built for hydropower generation purpose by increasing demand for
energy consumption of human.
As mentioned above, most of reservoirs are used for multiple-purpose. All
those purposes need to be satisfied but the capacity of reservoir is limited. For this
reason some conflicts may happen among the water users who have other interests and
conflicts also may happen in reservoir itself. For hydropower generation, higher
storage of water is needed, on the contrary, much water should be relaesed for
cultivated areas in dry season especially. Besides this, there are also many other
conflicts in user factors such as transportation and hydropower generation, flood
control and environment…etc.
Vietnam has many big river networks with nine major river basins spread along
the country. At present, many multi-purpose reservoirs were built to serve the socioeconomic issues such as Cua Dat, Hoa Binh and Dau Tieng Reservoir...etc. The
management and operation for many purposes are really difficult. On the other hand,
the operation of each reservoir is a challenge for management and operators. Reservoir
operation is needed to balance efficiently interests of water users and satisfy constraint
systems aim to get maximum interests. An optimal policy is necessary to accomplish

the problem objective and rule curve is one of appropriate methods to determine
operation policy of reservoir. Reservoir operation policy specifies the criteria to retain
or release water in or from a reservoir at different times of the year depending upon the
inflows and demands.

Trinh Xuan Manh
MSc Thesis


2
Optimization model used the mathematical programming technique to find the
best possible solution based on a specific performance function and some physical
constraints. Mathematical programming includes several techniques such as dynamic
programming (DP), nonlinear programming (NLP), linear programming (LP), genetic
algorithms (GAs) and optimal control theory (OCT) (Hirad and Ramamurthy 2000).
Within the development of soft computing technique, optimal technique has
been used in number water resources issues. In this thesis, the author will use Fuzzy
technique combine with hydraulic model to develop an operation policy for multipurpose reservoir in an efficient way.
I.2. Problem statement
The Ma river basin is located in the North-West region of Vietnam, it borders
Laos on the West. The upstream basin is located in Vietnam, the middle basin is
located in Laos and the downstream is located in Vietnam. The Ma river basin is an
international basin. The catchment area of Ma river basin is about 31.060 Km2 of
which that in Vietnam is 20.190 Km2 (IWRP 2003). The Chu River is a main tributary
of the Ma River. It is located in the downstream area (IWRP 2003).
Based on potential water resources of this river system, many kinds of reservoir
such as single purpose and multi-purpose were built on the main river of the Ma river
system. The Cua Dat Reservoir is one of the biggest projects related to water resource
projects in Thanh Hoa province. The Cua Dat Reservoir is a multi-purpose reservoir.
Those purposes include as: to reduce flood peak and protect downstream area due to

probability of flood of 0.6% and control water level in downstream area at Xuan Khanh
station on the Chu river (under 13.71m) in high flow season; To supply discharge of
7.715 m3/s for domestic and industrial water demand; To irrigate about 86.862 ha
cultivated area; To generate electricity with capacity of 97 MW; To prevent salt water
intrusion lower than 1‰ at Ham Rong measured station (MARD 2013).
As mentioned above, the Cua Dat Reservoir has purposes are to supply water for
some water users such as hydropower generation, agriculture, industry, domestic and
environment. However, in dry season the increasing water demand of water users is
one of the important problems within water shortage in this river basin due to less
rainfall will enhance the conflicts among all the factors. In order to balance different
Trinh Xuan Manh
MSc Thesis


3
water interests and solve the problems which related to using water, the Cua Dat
reservoir needs to optimize reservoir operation.
I.3. Objectives and Research questions
I.3.1. Objectives of the study
The main objectives of this research are:
- To optimize operation of the Cua Dat Reservoir in dry season, Thanh Hoa province
by using simulation model (MIKE 11 model) and optimal model (Fuzzy Logic
Technique).
- To provide management recommendations or alternatives and suggest appropriate
method of operation of the Cua Dat Reservoir in the Ma – Chu river basin.
I.3.2. Research Questions
1. What is Fuzzy logic theory and how to apply fuzzy logic in reservoir operation?
2. How to balance the water demand and water interests of the stakeholders in
operation of the Cua Dat Reservoir?
3. What are the objective functions and constraints in operation of the Cua Dat

Reservoir?
4. Does the Cua Dat Reservoir supply enough water for all of sectors in downstream
area regarding to current scenarios?
I.4. Structure of the thesis
This thesis structure includes those parts as below:
Chapter 1: This chapter discusses an overview of the study, the problem
statement and the objectives of the study are presented.
Chapter 2: This chapter reviews several researches of optimal reservoir
operation. Overview of hydrological model and optimization formulation are
presented. MIKE 11 model also is briefly introduced in this chapter.
Chapter 3: This chapter presents natural characteristic, natural conditions of the
study as well as population and economic characteristics of the study. Moreover, this
chapter also briefly introduces characteristics of the Cua Dat Reservoir and water
demand of each water user in downstream area.
Chapter 4: This chapter describes all kind of data collection and data analysis
which are used in this study. In this chapter, the author also shows the results of data
Trinh Xuan Manh
MSc Thesis


4
calculation as the input of hydrological modeling and calculating water demand of each
water user in the downstream area. This chapter determines the objective functions and
all of constraint systems in the Cua Dat reservoir as well as using optimization model
to determine optimal rule curve (standard rule curve). Hence, the author also presents
MIKE 11 model set up for calibration and validation model and the results of routing
flow from the Cua Dat Reservoir by MIKE 11 model in this chapter.
Chapter 5: The results of optimal model and simulation model are shown in this
chapter through figures and evaluation tables. The chapter also analyzes the results
from two models in order to achieve the objectives of the study.

Chapter 6: This chapter also focuses on the main performances, conclusions and
recommendations for future studies.

Trinh Xuan Manh
MSc Thesis


5

CHAPTER II
LITERATURE REVIEW
II.1. Studies on reservoir operation using optimal theory
Optimization is scientific field about best choice in some possible alternatives.
Optimal theory has been developed and investigated for many years over the world.
Optimization has been applied to a lot of fields in human life. Especially, in water
resource issues are used optimal theory as one of the effective tools for management
and decision making. Furthermore, optimization techniques have become increasingly
important in management and operations of complex reservoir systems. In reservoir
management, a lot of researchers have developed reservoir optimal operation during
the past four decades using dynamic programming (DP), linear programming (LP),
nonlinear programming (NLP), etc.(Cheng et al. 2008).
Rama and Sharad (2009) have developed operation policy for multi-purpose
reservoir in India using Neuro – Fuzzy technique including Fuzzy Mamdani and
ANFIS (Adaptive Neuro Fuzzy Interactive system). Their research determined
operation policy for monsoon period and non-monsoon period of Ramganga reservoir
and optimum releases against demands for domestic supply, irrigation and hydropower
generation. In other research, Omid et al. (2008) used optimal algorithm (HBMOHoney Bee Mating Optimization) for single and multi-purpose reservoir to minimize
the total present net cost of the system and maximum possible ratio for generate
electricity with installed capacity. In a case study of Hirakud Reservoir in Mahanadi
basin, India, D.Nagesh Kumar et al. (2009) used Folded Dynamic Programming (FDP)

to develop a long -term optimal operation policies for flood control. He showed that
FDP is a new search technique which can take care of all difficulties of other methods
to certain extend faced.
Long N.L et al. (2007) presented successfully a method as a tool for optimizing
operation of reservoir by using a combination of the simulation model and optimal
model. The authors optimized control strategies for the largest reservoir in Vietnam,
Hoa Binh Reservoir, in order to neutralize the conflicts in regulating water between
flood control and hydropower generation. The authors also organized two main
purposes in the flood season. With simulation model, they used MIKE 11 to guide the
Trinh Xuan Manh
MSc Thesis


6
releases of the reservoir system according to the current storage level, the hydrometeorological conditions, and the time of the year. Afterward, the shuffled complex
evolution (SCE) algorithm was chosen as a perfect tool for optimizing the reservoir
operation. Babel et al. (2011) analyzed that the tradeoff between hydropower
production and environmental flow requirements for the hydropower system and the
impact of alternative scenarios of a hydropower system operation on energy production
and natural flow regime in the La Nga river basin in Vietnam. The authors used
different alternative operation policies to simulate the system by the Range of
Variability Approach (RVA) method. Hirad and Ramamurthy (2000) showed a new
composite algorithm as an alternative model to solve the problem related to the size of
reservoir when operating policy of multi-reservoir systems is applied based on
Pontryagin’s minimum principle.
Genetic algorithms have been widely applied in optimization to solve water
resources system. Cheng et al. (2008) used Chaos Genetic Algorithm (CGA) which
based on the Chaos Optimization Algorithm (COA) and Genetic Algorithm (GA) to
apply to the global optimum of the Rosenbrock function, the Schaffer function and the
optimal operation of hydropower station reservoir. M.Habese, Y. Nagayama (2002)

used Neural Network and Fuzzy System to optimize multi-purpose Dam of flood and
non- flood seasons. Base on their results, the fuzzy system is an effective operation
system when the major objective is water use. Besides that Network Fuzzy System is
effective for flood control. In other research, fuzzy mathematical programming was
used in research of Jairaj and Vedula (2001), their study area is a three reservoir system
in the upper Cauvery river basin, south China. As the results illustrated that, use of
fuzzy linear programming in multi-reservoir system optimization presents a potential
alternative to get the steady state solution with less efforts than classical stochastic
dynamic programming (Jairaj and Vedula 2001). Panigrahi and Mujumdar (2000) also
used Fuzzy Logic in their study to reservoir operation modeling, the case study of the
Malaprabha irrigation reservoir in Karnataka, India.
Besides that, there are many researches in reservoir operation in Vietnam. They
also used many optimization and simulation methods. Nghia T.T (2009) used
combination method between optimization and simulation model within advanced tools
Trinh Xuan Manh
MSc Thesis


7
such as hydraulic dynamic model MIKE 11 and optimal technique GAMS. The
research had three major contents as follow: i) Determining water demand of water
users (such as Industry, Agriculture, Navigation and Environment); ii) Determining
upstream constraints of system due to periods; iii) Propose the operation process for
three reservoirs including Thac Ba, Hoa Binh, Tuyen Quang based on optimal
calculation in order to ensure that multi-reservoir can supply enough water for water
requirements in downstream. Hung N.T et al (2010) proposed models for optimal
operation of multiple purpose reservoir. The research proposed three distinct
alternatives including: i) Reservoir has mutual purpose for irrigation and hydropower;
ii) Reservoir’s major purpose is hydropower generation and second is irrigation; iii)
Reservoir has major purpose is irrigation and hydropower generation is second. Based

on models of the authors were built by using Delphi programming language and
applied Dynamic programming. The models were applied on Dinh Binh Reservoir
(Binh Dinh province) and A Vuong Reservoir (Quang Nam province). In other
research, optimization and simulation method were also used in the research of Tuyen
M.H (2009) for supplying water in dry season of reservoir system on Huong River
basin. The research combined GAMS optimization model and MIKE 11 hydraulic
dynamic model to control flow in downstream. The author illustrated that
environmental flow is about 31,5 m3/s in location of Thao Long Weir thoughout GAMS
optimal model. In dry water year with probability of 90%, the reservoir system can still
ensure the lowest discharge into dowstream area is 75m3/s.
Finally, optimization theories have been applied in a number of water resources
issues, especially in reservoir operation. Fuzzy logic technique is one of the useful
optimal tools for supporting reservoir operation and decision making. Using
optimization and simulation models in reservoir operation research are common over
the world. However, the Fuzzy Logic theory has been never applied in any research
about reservoir operation in Vietnam. That is reason in this research the Fuzzy theory
will be used as an optimization tool to optimize operation policy of Cua Dat Reservoir
thoughout objective function and constraints.

Trinh Xuan Manh
MSc Thesis


8
II.2. Fuzzy logic theory
According to Rama and Sharad 2009, the Fuzzy logic is another area of artificial
intelligence. It has been applied successfully in different water resources applications.
The key content about fuzzy logic theory is that it allows for something to be partly this
and partly that, rather than having to be either all this or all that. The degree of
“belongingness” to a set or category can be described numerically by a membership

number between 0 and 1.0 (Rama and Sharad 2009).
In fuzzy logic theory, variables are “fuzzification” through the use of
Membership Function (MF) that defines the membership degree to fuzzy sets. These
variables are called linguistic variables. A fuzzy subset A of a universe of discourse U
is characterized by a membership function μA(x) in the interval [0,1] and represents the
grade of membership in A (Rama and Sharad2009).
The fuzzy objectives and constraints are characterized by their membership
functions. Membership functions are curves that define how each point in the input
space is mapped to a membership value (or degree of membership) between 0 and 1. It
can be of different forms including triangular, trapezium, Gaussian, B-spline, sigmoid
etc. Membership function can be symmetrical or unsymmetrical (Rama and
Sharad2009).
Fuzzy rule base system can be used as a suitable representation of simple and
complex physical systems. The fuzzy rule based model operates on an “IF–THEN”
principle, where the “IF” is a vector of fuzzy explanatory variables or premises and
“THEN” is fuzzy consequence. Fuzzy logic theory allows the user to capture
uncertainties in data. A fuzzy tool is available with the MATLAB software. Two types
of fuzzy inference systems including: Mamdani type and Takagi Sugeno type.
Fuzzy logic theory also has been used widely in modeling of reservoir
operation. According to Panigrahi and Mujumdar 2000 when applying fuzzy theory
need to follow several steps: (a) Fuzzification of inputs, where the crisp inputs such as
the inflow, reservoir storage and release, are transformed into fuzzy variables, (b)
Formulation of the fuzzy rule set, based on an expert knowledge base, (c) Application
of a fuzzy operator, to obtain one number representing the premise of each rule, (d)
Shaping of the consequence of the rule by implication, and (e) Defuzzification.
Trinh Xuan Manh
MSc Thesis


9

Similarly, the Fuzzy logic will be used in this study for the Cua Dat reservoir
operation. And this is the initial research using the Fuzzy Logic in operating reservoir
in Vietnam.
II.3. Overview of hydraulic and hydrological modeling
Any scientific field always need a developed process including monitoring data,
recording and measuring data, simulation and explanation of natural phenomenon.
Hydrology is a science of water on the earth. To understand the hydrological events
can be described in laboratory by physical models. Based on theory and practice,
people have explained clearly the most of hydrological phenomenon such as rainfall,
infiltration, evaporation, and simulated them by hydrological models (hydraulic and
hydrological models).
Accordingly, hydraulic and hydrological models are tools to address the real
hydrological cycle in a simplified way. That kind of models are used for understanding
the hydrological processes as well as making hydrological prediction if there are some
water resources management and utilization activities are implemented (Tuan 2012).
The models are applied several algorithms to provide a quantitative relation between
the input data (e.g rainfall, meteorological data) and output (e.g runoff). The
mathematical models have been developed from 19th century with the simplest
rainfall-run off model by Mulvaney (1851) to more sophisticated models such as MIKE
Package developed by Danish Hydraulic Institute; Soil and Water Assessment Tools
(SWAT), HEC model developed by Hydraulic Engineering Center- USA; SIMONA
2D/3D hydrodynamic models by the Dept of Public Works and Delft 2D/3D by
WL|DELFT HYDRAULICS. Those models are used for simulation of flow, water
quality and sediment transport in estuaries, rivers, irrigation system, channels and
others bodies. They are fundamental to integrated water management as used for
planning and decision making (Tuan 2012).
The figure below shows the relationship between the various representations of
a model (Van Waveren et al 1999).

Trinh Xuan Manh

MSc Thesis


10
The conceptual model is developed on the
basis of knowledge of the system and
serves as the basis for a mathematical
model. This model may be solved either
analytically or numerically. The model
reated is further refined into a model
program and finally into a Computer
Model by entering the proper input data.
(Van Waveren et al 1999).
Figure 2-1: Relationship between the
various representations of a model
Hydrological models have been used frequently in water resources planning and
management such as hydrological forecasting, reservoir operation, water quality,
research on flood, inundation and drought, designing irrigation system, supporting for
the integrated water resources management…etc. Appropriate model selection are
essential for each research or project. These selections should based on study
objectives, considering input data and output data, expected results and solutions.
There are many studies on water field that using model as an effective tool to solve
problems. According to Piman et al (2012), the authors used HEC and SWAT models
to simulate and evaluate flow changes from hydropower development and operation in
rivers: SeKong, Se San and Sepork of the Mekong basin. Long et al (2007) used MIKE
11 simulation model to set up control strategies for Hoa Binh reservoir operation. They
concluded that this model is an effective tool for optimizing complex system.
Bahremand and Smedt (2007) used distributed hydrological modeling (WETSPA) and
sensitivity analysis in Torysa Watershed, Slovakia to predict daily discharge value.
They also presented that a strategy by incorporating a model-independent parameter

estimator PEST for automatic calibration and sensitivity analysis.
In this study, MIKE 11 model will be selected to rout the flow, which is released
from the Cua Dat Reservoir operation to the downstream area in order to evaluate or
test discharge value within constraint system at control points. MIKE 11 model is a

Trinh Xuan Manh
MSc Thesis


11
strong model. This model has been applied widely in Vietnam for many projects,
especially in field of water resources.
II.4. MIKE model
Many kind of hydraulic models have been applied widely in water resources
issues. MIKE 11 model is one of hydraulic models which have used popularly in
Vietnam. The MIKE 11 model has been developed by DHI water and environment
(Danish hydraulic institute). This model is a professional engineering software package
for simulation of flow, water quality and sediment transport in estuaries, rivers,
irrigation system, channels and others bodies (DHI 2011).
The study area has slope topography, short length of river and combine complex
rain regime to make flood regime change complicatedly. In this study, the author
selects the MIKE 11 model for routing the flow on the Ma-Chu river network. To apply
this model for study area, the understanding of model theory plays an important role.
The briefly description of the model theory according to DHI user’s manual as
following (Kmenl 2008):
The most commonly applied Hydro-Dynamic (HD) model is a flood
management tool simulating the unsteady flows in branched and looped river networks
and quasi two-dimensional flows in floodplains. When using a fully dynamic wave
description, MIKE 11 HD module solves the equations of conservation of continuity
and momentum (the ‘Saint Venant’ equations) as bellow:

Continuity equation:
Q A

q
x t

(2-1)

Momentum equation:
 Q2
 
Q
 A

t
x





 gA

QQ
h
g 2
0
x
C AR

Where:
Q: discharge (m3/s).
x: distance along flow direction (m).
g: gravity constant (m3/s)
Trinh Xuan Manh
MSc Thesis

(2-2)


12
q: lateral inflow (m2/s)
R: hydraulic radius (m)
A: flow area (m2)
T: time (s)
H: stage above datum (m)
C: Chezy coefficient.


: momentum distribution coefficient

The MIKE 11 solution of the continuity and momentum equations is based on
an implicit finite difference scheme developed by Abbott and Ionescu (1967). The
scheme is setup to address any form of the Saint Venant equations – such as:
kinematic, diffusive, or dynamic. The water level and flow are calculated at each time
step, by solving the continuity equation and the momentum equation using a 6-point
Abbot scheme with the mass equation centered on h-points and the momentum
equation centered on Q-points. By default, the equations are solved with 2 iterations.
The first iteration starts from the results of the previous time step and the second uses

the centered values from the first iteration. The number of iterations is user specified
(DHI 2011).
Cross sections are specified in both area and longitudinal location through the
user interface. The water level (h points) is calculated at each cross section and at
model interpolated interior points located evenly and specified by the user-entered
maximum distance. The flow (Q) is then calculated at points midway between
neighboring h-points and at structures (DHI 2011).
The hydraulic resistance is based on the friction slope from the empirical
equation, Manning’s or Chezy, with several ways of modifying the roughness to
account for variations throughout the cross-sectional area (DHI 2011).

Trinh Xuan Manh
MSc Thesis


13

CHAPTER III
THE STUDY AREA
III.1. Description of the study area
III.1.1. Location of the study area
The Ma River basin is located in the northwest region of Vietnam, on the
eastern slope of the Truong Son mountain range bordering Laos on the West. The
upstream basin is located in Vietnam, the middle basin is located in Laos and the
downstream is located in Vietnam. Accordingly, the Ma River basin is an international
basin and is the 4th in biggest river basins in Vietnam following The MeKong, Dong
Nai and Red river basin. The whole river spreads from 22037’30’’N to 20037’30’’N
and 103005’10’’E to 106005’10’’E (IWRP 2003).
-

The North borders on Da river basin, Boi river and Vac river basin;

-

The West borders on the Mekong river basin;

-

The South borders on Hieu and Muc river basin;

-

The East borders on the East Sea.

Figure 3-1: Location of study in the Thanh Hoa province in Vietnam
Trinh Xuan Manh
MSc Thesis


14
The total catchment area is about 28,490 Km2, of which area in Vietnam is
17,600 Km2 accounting for 62% of whole area and area in Laos is 10,800 Km2
corresponding to 38% (IWRP 2003). The major tributaries of the Ma river system
originates from the high mountains of Tuan Giao district which belongs to Lai Chau
province, Vietnam. The highest point of the upstream part is 1,500 m, the river flows
through the area of provinces and nation, namely, Son La, Lai Chau, Hoa Binh, Laos
and Thanh Hoa, and flows into the East Sea finally via three river mouths, namely Hoi,
Lach Truong and Lach Sung. Accordingly, Hoi river mouth is a main mouth of the Ma
River. The Ma River has length of 512 Km of which 102 km is located in Laos and in
Viet Nam is about 410 Km.

Table 3-1: Distribution of natural areas according to provincial border
of the Ma river basin (ha)
No.

Provincial units

Natural
area

I
II

Laos
Viet Nam

1.098.751
1.750.249

Possible
Possible
agricultural
Geography
forestry soil
soil
32.962
824.063
High mountain
287.828
1.299.987

1
2
3
4
5

Dien Bien
Son La
Hoa Binh
Nghe An
Thanh Hoa
Total

209.475
477.038
177.836
62.810
823.090
2.849.000

19.649
29.981
38.734
5.000
194.464
320.790

188.452
394.115
83.527

45.000
588.893
2.124.050

High mountain
High mountain
High land
High land
High land- Delta

(Source: Final engineering report of the Cua Dat Reservoir in operation period -2014)

III.1.2. River network
The Ma river flows on the Northwest – Southeast direction, the river direction is
similar to tectonic direction, the length of major river is around 512 Km, originates
from the highland of Tuan Giao district, flows through some of provinces, enter into
Thanh Hoa province at Muong Lat, Quan Hoa location to discharge the East Sea at Hoi
estuary. The basin has river density of 0.66 km/km2, the meandering index is 1,7, the
shape index is 0,17, the average slope of the basin is 17,6 % (IWRP 2003).
The Chu River basin is one of main tributaries of the Ma River in this river
basin. It is located in the downstream area. The catchment of Chu River is about 7,500
km2, of which 65% are located in Laos and 95% of Chu river catchment area is in
Trinh Xuan Manh
MSc Thesis


15
mountains (IWRP 2003). The Chu River joins the Ma River at Giang confluence which
is about at 26 km away from Ma river mouth. The river originates from Sam Nua high
mountain which belongs to Laos, with elevation of 2,000 m, this river flows

meanderingly in dangerous high mountains such as Phu Nam (2,050 m), Phu Bo (1,455
m), entering into Viet Nam at Nghe An province. The main flow has a length of 325
Km of which 100 Km is located in Vietnam.
In Vietnam’s region, the Chu River flows on narrow and slope valleys with a lot
of waterfalls. There are 15 waterfalls from Muong Hinh to Cua Dat location. From the
confluence of the Dat River to downstream, the river networks have risen significantly,
the Chu River has a protected dyke system and many large tributaries such as the Am
River, the Dat River and the Dang River. Among these tributaries, the Am River is a
largest one.
The Buoi River is a second main tributary of the Ma River. This river originates
from Chu Mountain which belongs to Hoa Binh Province. The main river flows
towards North-South direction joining the Ma River at Vinh Khang position. The
length of this river is about 130 km, basin area is 1,790 km2 and average slope is 1.22.
The upstream of the Buoi River includes three major stream, namely Cai, Bin and
Cong Hoa streams.
The Cau Chay River is originated from Den Mountain flowing towards EastWest direction though out delta of Ma River south and Chu River north. The total
length is about 87.5 km and basin area is 551 km2.
The table 3-2 shows detailed characteristics of river shape in the Ma-Chu river
system as below:
Table 3-2: Characteristics of river shape of some large tributaries
No.

River Basin

F
(km2)

1
2
3

4
5
6
7
8
9

Nam Khoai
Nam Thi
Nam Cong
Luong
Lo
Buoi
Cau Chay
Chu
Ma

1.640
705
893
1.580
1.000
1.790
551
7.580
28.400

% of
Aver.
Aver.

Length
area
elevation width
(km)
(%)
(m)
km/km2

Aver.
slope
(%o)

UnsymRiver
Shape
matrical
density
index
(km/km2) index

Mean
dering
index

5,77
2,48
3,14
5,56
3,52
6,30
1,94

26,7
100

18,0
19,3
16,4
19,6
20,4
12,2
5,4
18,3
17,6

0,59
0,47
0,98
0,66

1,45
1,28
1,58
1,27
1,35
1,53
1,62
1,58
1,79

62,5
47,5

52,0
102
76,0
130
87,5
325
512

890
984
1.233
532
615
217
114
790
762

29,7
18,1
19,9
17,6
13,9
16,1
8,0
29,8
68,8

-0,17
-0,57

-0,16
0,19
-0,33
0,16
0,01
-,014
0,32

0,54
0,46
0,22
0,20
0,19
0,14
0,12
0,12
0,17

(Source: Final engineering report of the Cua Dat Reservoir in operation period -2014)

Trinh Xuan Manh
MSc Thesis


16

Figure 3-2: Ma – Chu River Network in Vietnam
III.1.3. Topographical characteristics
The Ma river basin spreads widely many regions between Vietnam and Laos.
This basin ranges from Truong Son Mountain range to Northern Bay, the topography is

strongly fragmented and changing complexly. The main slope directly ranges from
West-North to East-South. The topographical elevation varies from 1.0 m to 2000 m.
which can be divided into three main categories of topography, are described as below:
- High mountain terrain: This area is in the upper part of the river basin: from Ba
Thuoc location to upstream of Ma River, and from Cua Dat location to upstream of
Chu River. The highest elevation of this topography is Phu Lan Mountain with
elevation of 2,275 m. The elevation changes towards North-South direction. The area
of this topography is about 23,228 km2 and it takes 80% in total. The forestry trees are
Trinh Xuan Manh
MSc Thesis