As mention above, the author used MIKE 11 model to simulate the release after
optimal operation from Cua Dat Reservoir and the results illustrated that the minimum
daily discharge at the Bai Thuong weir is around 35 (m3/s) and the Xuan Khanh station
is appropriate 36 (m3/s). The minimum daily discharge of the optimal operation is
higher than minimum daily discharge of requirement and actual release.
Finally, those showed that the optimal operation of the Cua Dat Reservoir using
Fuzzy Logic approach determined efficient release and satisfied the water demands in
downstream area. The optimal operation would decrease water stress and conflicts in
dry season.

111 trang |

Chia sẻ: ngoctoan84 | Ngày: 20/04/2019 | Lượt xem: 376 | Lượt tải: 0
Bạn đang xem trước 20 trang tài liệu **Optimal reservoir operation for water supply in dry season: the case study of Cua Dat reservoir in the Ma – Chu river basin**, để xem tài liệu hoàn chỉnh bạn click vào nút DOWNLOAD ở trên

each level depend on knowledge of
expert. If number of membership function for inputs and output increases, the number
of rules in fuzzy rules base will also increases as well as the error decreases. Regarding
to analysis of observed and calculated data of inputs and output (appendix 3), the
author determined the ranges of each variable and divided into distinct levels.
With reservoir water level has ranges that varied from 70m to 115 m. It was
classified into 7 levels (figure4-16) including ‘very low’, ‘low’, ‘low-medium’,
‘medium’, ‘medium-high’, ‘high’ and ‘very high’. The inflow into the reservoir as
‘very low’, ‘low’, ‘medium’, ‘high’ and ‘very high’ are in range of [0-145] Mi.m3.
Figure 4-16: Membership function for reservoir level for Fuzzy Mamdani model
Figure 4-17: Membership function for inflow for Fuzzy Mamdani model
59
Trinh Xuan Manh
MSc Thesis
In the same way, the demand is divided into class intervals as ‘very low’, ‘low’,
‘meium’, ‘high’ and ‘very high’ in range of [0 -145] M.m3. Finally, the release is
divided into 8 class intervals (figure 4-18) as ‘very low’, ‘low’, ‘medium-low’,
‘medium’, ‘medium-high’, ‘high’ and ‘Very.very.high’.
Figure 4-18: Membership function for water demand for Fuzzy Mamdani model
Figure 4-19: Membership function for release for Fuzzy Mamdani model
IV.2.3.2. Formulation of the Fuzzy rule base
In a Fuzzy system, fuzzy rules base was built due to ‘IF-THEN” principle. The
fuzzy rule set usually is formulated based on knowledge and experience of expert.
60
Trinh Xuan Manh
MSc Thesis
Besides this, the fuzzy rule base also based on the actual observed operation of the
reservoir. According to the collected data and different categories of each input/output
variable in dry season and using Fuzzy Mamdani method, the Fuzzy rules have been
identified as below:
IF (RESERVOIR WATER LEVEL IS LOW) AND (INFLOW IS LOW) AND
(WATER DEMAND IS LOW) THEN (RELEASE IS VERY LOW)
There are 157 fuzzy rules were defined in Fuzzy system of the Cua Dat reservoir
regarding to 03 input variables as reservoir water level, inflow and water demand and
01 output variable as release (Appendix 5). The fuzzy output of release is also in a
fuzzy set to achieve the crisp value which is needed a process of “defuzzification”
method. The figure 4-20 presents the fuzzy rules base for operation of the Cua Dat
reservoir.
Figure 4-20: Fuzzy rules base for operation of Cua Dat reservoir
IV.2.3.3. Application, Implication, Aggregation and Defuzzification
Application, implication, aggregation and defuzzification are main parts to
calculate fuzzy output and determine crisp value for each rule is known in a fuzzy
system. Application is a process that it uses fuzzy operators such as AND or Or to
61
Trinh Xuan Manh
MSc Thesis
obtain value for presenting the result for one rule. Panigrahi and Mujumdar (2000)
have discussed the fuzzy logic operators obey the classical two valued logic. The AND
opertator can be conjunction of the classical logic or it can be the product of the two
parameters involved in it.
Implication is a process after application above. Application process uses fuzzy
operators to calculate a value due to inputs of fuzzy system then Implication process
will apply the value into membership function of outputs in order to obtain a fuzzy set
for each rule. Especially, in implication process is input of it is a number value but the
result is a fuzzy set. For each rule always implement an implication method.
Aggregation is next step after implication step. Aggregation integrates the
outputs which are results of implication method. The inputs of aggregation are fuzzy
set and result of aggregation process is also a fuzzy set for each output variable. The
figure 4-21 presents process of application, implication and aggregation. This figure
was found in Panigrahi and Mujumdar (2000).
Figure 4-21: Process of application, implication and aggregation
The result is optimized from fuzzy system through the processes such as
application, implication and aggregation is a fuzzy set. So that we need a
defuzzification method in order to achieve a crisp value or crisp number. The
defuzzification method is integrated ready in Fuzzy Tool Box of MATLAB software
62
Trinh Xuan Manh
MSc Thesis
including: Centroid; Bisection; Largest of Maximum; Smallest of Maximum, Middle of
Maximum. Panigrahi and Mujumdar (2000) have shown that one of most common
defuzzification method is “Centroid” evaluation. In this thesis, the Centroid method
will be used for defuzzification.
The equation of this method as following:
𝐺 =
∑ 𝑦𝑖 𝑚𝐵(𝑦𝑖)𝑛𝑖=1
∑ 𝑚𝐵(𝑦𝑖)𝑛𝑖=1
(4-12)
Where: G is the centroid of the truncated fuzzy output set B
mB(yi) is the membership value of the element yi in the fuzzy out put set B,
n is the number of elements
IV.3. Hydraulic and hydrological model setup
In this part the author is going to set up a hydraulic model and a hydrological
model for the Ma – Chu river basin. The MIKE 11 HD and MIKE NAM models will
be used as an efficient way as mentioned in the literature review of the thesis. These
models were used to simulate the release from the reservoir to the river in downstream
area as well as the flow at some of control points after routing that to be evaluated
effective optimization.
MIKE NAM is a hydrological model and MIKE 11HD is a hydraulic model,
the combination between 02 models through calibration and validation process to
determine optimal basin parameters. The detailed contents are presented in following
steps.
IV.3.1. Determination of the model inputs
As presented above, the author used two models including the MIKE NAM and
MIKE 11HD model. Each model will need different data as inputs. The inputs of the
MIKE NAM hydrological model include rainfall, evaporation and discharge. The
inputs of MIKE 11HD include discharge and water level at the stations (appendix 4).
To build up hydrological model, the author collected rainfall data from the
meteorological and point stations of two years of 2006 and 2008 of Cua Dat, Bai
Thuong, Cam Thuy, Song Am, Thuong Xuan, Sam Son, Giang stations to simulate
NAM model for Cua Dat catchment. Similarly, discharge data of the Cua Dat station
was collected to calibrate and validate the basin parameters.
63
Trinh Xuan Manh
MSc Thesis
The hydraulic model has the boundary conditions such as upstream boundaries,
downstream boundaries and topographic data as cross sections. The choice of boundary
conditions depends on the physical situation and data available on the basin. In this
study, the author selected the following stations and data itself as boundary conditions.
The upstream boundary is time series of observed discharge at the hydrological
stations as Cua Dat (Chu Rver), Cam Thuy (Ma River), Kim Tan (Buoi River) and Tu
Thon (Bao Van River) stations. Downstream boundary is time series of observed water
level at some of stations in the river mouth as Hoang Tan, Lach Truong and Lach Sung
stations. The time period for calibration and validation of model is in 02 years (2006
and 2008). After data collection for those models the author implemented set up model
network as below:
IV.3.2. Model setup
In this part, the author implemented set up hydraulic schematization with the
boundary conditions for the Ma – Chu river basin. According to topography data and
hydrological data and hydraulic regime of river network, the hydraulic network of the
river basin was chosen as following:
+ Ma River: starting from Cam Thuy station to the estuary (Cua Hoi estuary)
+ Chu River: starting from Cua Dat station to conjunction point of the Ma River (Giang
confluence)
+ Buoi River: starting from Thach Lam station to conjunction point of the Ma River
(Vinh Khanh confluence)
+ Cau River: Starting from Cau Chay station to conjunction point of the Ma River at
Cam Truong confluence.
+ Len River: from Bong confluence to Lach Sung Estuary
+ De Chanel: from separate branch of Len River to conjunction point of Lach Truong
River.
+ Lach Truong River: from Tuan confluence to Lach Truong estuary.
The boundary conditions of this model were distributed:
- Upstream boundary: is time series of daily observed discharge from upstream
stations:
+ Cam Thuy station is located on the Ma River
64
Trinh Xuan Manh
MSc Thesis
+ Cua Dat station is located on the Chu River
+ Kim Tan station is located on the Buoi River
+ Thu Thon station is located on the Bao Van River
- Downstream boundary: is time series of daily observed water level from
downstream stations:
+ Hoang Tan station is located in the Estuary
+ Lach Truong station is located in estuary of Tao Khe River.
+ Thach Sung is located on Len River.
- Tributary boundary: flow which occurs from tributary basin is identified
through MIKE NAM model. List of tributary basins are in below table:
Table 4-13: List of tributary basins on the Ma – Chu river basin
Tributary basins
Name Control areas Area (km2)
Am River Am river basin area 817.5
Dat river Dat river basin area 304
Dang River Dang river basin area 344.7
Cau Chay River Cau Chay river basin area 118
- Control stations: is used to evaluate the efficiency of the model including the
station as following. Those stations are water level stations:
+ Ly Nhan station is located on the Ma River
+ Giang station is located on the Ma river)
+ Xuan Khanh station is located on the Chu River
The hydraulic network of Ma-Chu river basin is presented in the below figure:
65
Trinh Xuan Manh
MSc Thesis
Figure 4-22: Hydraulic network of the Ma – Chu river basin
IV.3.3. Model calibration and validation
This process aims are to obtain the optimal parameters for NAM model and
MIKE 11HD model respectively. To evaluate the efficiency of model using observed
data at some control stations are compared to simulated data from the model. To access
performance of the models the author used Nash- Sutcliffe coefficient (NASH). NASH
has been commonly used in hydrological models to evaluate flow hydrographs. The
NASH is an improvement over the coefficient of determination for stream flow
comparison, because it accounts for model errors in estimating the mean of the
observed datasets. It enables the efficiency of the model to be compared with the initial
variance that is defined by the observed datasets. The observed and simulated
discharge values have been compared to assess the predictive accuracy of the model.
NASH ranges from −∞ to 1. When the values are from 0.9 to 1 indicate that model
performance is perfect. When the values are from 0.8 to 0.9 and 0.6-0.8, the model
performances are good and acceptable, respectively. It is defined as following:
66
Trinh Xuan Manh
MSc Thesis
2
2
)(
)(
1)(
QobsaverQobs
QobsQcal
EINASH
(4-13)
Where:
calQ : Simulated discharge at the time ith
obsQ : Corresponding observed discharge at the time ith
Qobsaver : Average observed discharge
Moreover, it is possible to use other indexes or coefficients to access the results of
model such as: Correlation Coefficient; Peak Error; Volume Error and Peak time Error
etc. The results for both models will be presented in following contents.
IV.3.3.1. NAM model calibration and validation
A- Calibration for NAM model of the Ma –Chu river basin: Calibration is
one of main steps in modeling. Trial and error method was used to stimulate NAM
Model for the Cua Dat Catchment. It was carried out by selecting a possible
parameters, applies it to the model and, if it is not successful, selects another possibility
that is subsequently tried. The process ends when a possibility yields a goodness of fit
between simulated and observed data. The Cua Dat Reservoir is upstream part of Ma-
Chu River Basin. Its area covers 5.708 square kilometers.
To calibrate the hydrological model of the Cua Dat catchment on the Chu River,
the author used daily rainfall of the Cua Dat station and surrounding stations in year of
2006. Average daily evaporation at Thanh Hoa station was also used for the same
period. The observed discharge which was collected from the Cua Dat station was used
to access the performances of the model. The simulated performance for calibration
process is as the Figure 4-23.
67
Trinh Xuan Manh
MSc Thesis
Figure 4-23: Observed and simulated hydrograph at Cua Dat station in 2006
Based on two hydrographs on the above figure, it can be seen that the difference
between them is not quite high. The simulated result is still higher than observed result.
Those caused by number of rainfall stations on the basin is rather limited, it do not
cover the whole basin. However, the shapes of them are quite similar with each other
and NASH is 0.69 at the station indicates an acceptable value. With this result, it is
possible to validate the model in next step as below.
B- Validation for NAM model of Ma –Chu river basin: Model validation is
one of the important steps in modeling to achieve confident model. The model used
only the parameters which were obtained in previous step to simulate runoff from
rainfall. Year of 2008 was used for model validation with the same data and stations in
calibration step. The simulated performance for validation process is as figure 4-24.
68
Trinh Xuan Manh
MSc Thesis
Figure 4-24: Observed and simulated hydrograph at Cua Dat station in 2008
For the validation period, the above figure indicates that between observed and
simulated hydrograph have a good fit and similar pattern. NASH = 0.73 was acceptable
value and indicated well validated model. Therefore, developed NAM model could be
confidently used for the purpose of this study.
After calibration and validation of NAM model, a set of NAM parameters are
determined and they would apply to the Ma-Chu river basin to reveal the flows of
rivers in basin. The objective of application of NAM model was to simulate runoff
from available rainfall data for Ma - Chu River and its branches. However, due to lack
of sufficient discharge data, the application of direct calibration of a rainfall-runoff
model for each sub-basin is impossible. So it were needed to borrow model parameters
from the Cua Dat catchment is to apply for other sub-basins to recovery runoff. The
tributary basins were simulated runoff presented in the Table 4-13.
IV.3.3.2. MIKE 11HD model calibration and validation
In this part, in order to determine the optimal parameter for hydraulic model, the
author used manual method to simulate the flow from upstream station. To do that the
hydrological data was collected for year of 2006 and 2008 respectively. Insides that,
the period of model calibration was chosen in 2006 and year of 2008 was used for
model validation. The performance evaluation of model could be measured by NASH
coefficient as efficient way.
69
Trinh Xuan Manh
MSc Thesis
A- Calibration for MIKE 11HD model
For calibration period, the calibration is implemented by changing roughness
along the river and cross section, time step, etc. The calculated process will be stopped
if the difference between calculated and observed data is less than certain value for
every point. The results for calibration are presented in some figures below:
Figure 4-25: Observed and simulated water level at Ly Nhan Station in 2006
Figure 4-26: Observed and simulated water level at Xuan Khanh Station in 2006
70
Trinh Xuan Manh
MSc Thesis
Figure 4-27: Observed and simulated water level at Giang Station in 2006
Through the results which were showed in the above figures, it is possible to see
that having a good fit between observed and simulated data about pattern and peak
value for the most of the stations. NASH for stations varied from 0.87 to 0.92, the
difference of two peaks is rather low. Therefore, the author also used the Peak Error
and Volume Error to evaluate the results (Table 4-14)
Table 4-14: Results of MIKE 11HD model calibration
at the Ma-Chu river basin in 2006
Name of station NASH Peak Error Volume Error
Ly Nhan 0.97 1.92% 0.33%
Giang 0.89 4.90% 9.50%
Xuan Khanh 0.88 3.98% 6.07%
So that with results of model calibration, the author basically determined the
hydraulic parameter (roughness) for whole river basin, Manning coefficient vary from
0.025 to 0.045 from upstream to downstream area. To evaluate the confidence of the
model the author implemented next step for Validation.
B- Validation for MIKE 11HD model
The results for model validation are presented in some figures below. For the
validation period, it can be seen that there is a good fit between observed and simulated
71
Trinh Xuan Manh
MSc Thesis
data about pattern and peak value for most of the stations. NASH for stations varied
from 0.83 to 0.94, the difference of two peaks is rather low.
Figure 4-28: Observed and simulated water level at Ly Nhan Station in 2008
Figure 4-29: Observed and simulated water level at Xuan Khanh Station in 2006
72
Trinh Xuan Manh
MSc Thesis
Figure 4-30: Observed and simulated water level at Giang Station in 2006
Table 4-15: Results of MIKE 11HD model validation
at the Ma-Chu river basin in 2008
Name of station NASH Peak Error Volume Error
Ly Nhan 0.91 1.60% 3.52%
Giang 0.91 8.25% 6.44%
Xuan Khanh 0.84 4.07% 11.88%
According to the Table 4-15 and Figure 4-28 to 4-30, it can be seen that
observed water level is higher than simulated water level at the beginning of the
hydrograph at Xuan Khanh station. However, this difference is still in the limited
range. With the highest results, developed MIKE 11HD model could be confidently
used for the purpose of this study.
73
Trinh Xuan Manh
MSc Thesis
CHAPTER V
RESULTS AND DISCUSSIONS
V.1. Optimizing the Cua Dat reservoir operation
As presented above, the author sets up a Fuzzy system for operation of the Cua
Dat Reservoir (Figure 4-31). The Fuzzy system includes 03 input variables as 03 input
variables such as reservoir water level, inflow and water demand and 01 output
variable as release. Beside that, the Fuzzy system also includes a fuzzy rules base with
157 rules.
Figure 4-31: Structure of fuzzy system for the Cua Dat reservoir
Through setting Fuzzy system on the Cua Dat Reservoir, the author defined the
graph of optimal release in dry season in 2011 – 2012. The optimal release graph in the
Figure 4-32 presents optimal release that can meet water demand more than actual
release. Especially, in the driest period in year when water demand increased
significantly but the reservoir operated effectively to ensure this demand, while the
actual release was lower significantly and did not supply enough water. To evaluate the
quantification of effect of fuzzy release in the condition of shortage of water in the dry
season, the author used the Nash- Sutcliffe coefficient (NASH) in whole water supply
period in dry season.
The results are shown that it developed for Fuzzy operating systems for the Cua
Dat Reservoir and determine the optimal discharge process in the condition of shortage
74
Trinh Xuan Manh
MSc Thesis
of water in the dry season, and reservoir can still meet appropriate 80% of water
demand throughout the dry season from 2011 to 2012. Actual release was determined
through observed data of the reservoir was only to supply about 73% of actual demand.
Table 4-16: The NASH for calculation of alternatives
Year Releases Nash
2011-2012
Actual release 73%
Optimal release 80%
Optimal operation of the Cua Dat Reservoir using Fuzzy Logic approach
determined efficient release and satisfied the water demands in downstream area more
than the actual release as well as optimal operation would decrease water stress and
conflicts in dry season.
Figure 4-32: Comparison of water demand and fuzzy and actual releases
V.2. Routing the release to the downstream
The flow in dry season of year of 2011-2012 after the optimal operation of Cua
Dat Reservoir is to rout to the downstream by MIKE 11 hydraulic model.
The results of flow characteristics are in some of control points will be showed
in some of figures and tables below:
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
1 3 5 7 9 111315171921232527293133353739414345474951535557596163656769717375
R
es
er
v
o
ir
r
e
le
a
se
/D
e
m
a
n
d
(
M
i.
m
3
)
Ten-day period
Water demand Optimal Release Actual release
75
Trinh Xuan Manh
MSc Thesis
Figure 5-1: Hydrograph of optimal operation at the Bai Thuong weir
Figure 5-2: Hydrograph of optimal operation at the Xuan Khanh station
76
Trinh Xuan Manh
MSc Thesis
Figure 5-3: Hydrograph of optimal operation at the Giang station
Table 5-1: Flow characteristics at the Chu River downstream
using optimal operation
No. Flow characteristics
Q
Optimization
(m3/s)
Q
Requirement
(m3/s)
Q Actual
Release
(m3/s)
1
Qday Baithuong min
(m3/s)
35 30.42 29.3
2
Hday XuanKhanh min
(m)
2.25 - 1.98
3
Qday XuanKhanh min
(m3/s)
35.8 30.42 33.2
4
Hday Giang min
(m)
-0.091 - -1.06
According to the results above, it can be seen that the minimum daily discharge
at the Bai Thuong weir is around 35 (m3/s) and the Xuan Khanh station is appropriate
36 (m3/s). The table 5-1 makes a comparison between the minimum daily discharges of
the optimal operation and minimum daily discharge of requirement in the downstream
area, Hence, the minimum daily discharges of the optimal operation is higher than
minimum discharge of requirement and discharge of actual release. This showed that
the optimal operation of the Cua Dat Reservoir using Fuzzy Logic approach
determined efficient release and satisfied the minimum water demands in downstream
area.
77
Trinh Xuan Manh
MSc Thesis
CHAPTER VI
CONCLUSIONS AND RECOMMENDATIONS
VI.1. Conclusions
The study investigated optimal reservoir operation using Fuzzy Logic algorithm
for the Cua Dat Reservoir, which is located in Thanh Hoa province, middle of
Vietnam. To apply Fuzzy Logic in this study, the author used the Fuzzy Tool Box of
MATLAB software to optimize the release from the reservoir and it also simulated to
downstream river by using MIKE 11 hydraulic model.
In order to have an accuracy and reliable model for the purposes of this study,
Fuzzy Logic algorithm is used with the information as existing rule curve of this
reservoir, inflow into the reservoir, water level of the reservoir, observation discharge
of hydropower plant. The observed data was recorded during the operation of the Cua
Dat reservoir having short period because this reservoir has just operated since 2010.
Moreover, water demand for water users in the downstream were also one of the
important input variables for Fuzzy system, this variable was determined according to
data related to water users such as Agriculture, Industry, Domestic, Hydropower, and
Environment. The author determined that total water demand is about 4547 Mi.m3.
From the results which are optimized from the Fuzzy operating systems for the
Cua Dat Reservoir, the optimal discharge process was determined in the condition of
shortage of water in the dry season, and reservoir can still meet 80% of water demand
throughout the dry season from 2011 to 2012. While that, actual release was
determined through observed data of the reservoir was only to supply about 73% of
actual demand.
To evaluate the efficiency of the optimal operation in the downstream river, the
author has used two models including MIKE NAM and MIKE 11HD as an efficient
way as mentioned in the literature review of the thesis. For model calibration and
validation process, the hydrological and meteorological data was collected such as
rainfall, evaporation and discharge from the stations on Ma- Chu river system in 2006
and 2008. To access performance of the models, the Nash- Sutcliffe coefficient
(NASH) was used. For calibration period of NAM model, the NASH is 0.69 at the
station indicated an acceptable value. For the validation period of NAM model, the
78
Trinh Xuan Manh
MSc Thesis
result indicated that between observed and simulated hydrograph have a good fit and
similar pattern. NASH = 0.73 was acceptable value and indicated well validated model.
Similarly, within MIKE 11 HD model calibration and validation. The results of model
calibration at Ly Nhan, Xuan Khanh and Giang stations showed that having a good fit
between observed and simulated data about pattern and peak value for the most of the
stations. The NASH for stations varied from 0.87 to 0.92, the difference of two peaks is
rather low. The results for model validation of MIKE 11 HD, the NASH for the same
stations varied from 0.83 to 0.94, so that within the highest results, developed MIKE
11HD model could be confidently used for the purpose of this study.
As mention above, the author used MIKE 11 model to simulate the release after
optimal operation from Cua Dat Reservoir and the results illustrated that the minimum
daily discharge at the Bai Thuong weir is around 35 (m3/s) and the Xuan Khanh station
is appropriate 36 (m3/s). The minimum daily discharge of the optimal operation is
higher than minimum daily discharge of requirement and actual release.
Finally, those showed that the optimal operation of the Cua Dat Reservoir using
Fuzzy Logic approach determined efficient release and satisfied the water demands in
downstream area. The optimal operation would decrease water stress and conflicts in
dry season.
VI.2. Recommendations
During the process of this thesis, the author has some following
recommendations and suggestions for the future possible studies: It is very necessary to
collect observed data during the operation of the Cua Dat Reservoir. Because the inputs
for Fuzzy system need a long time period to calculate and design a system exactly.
To determine exactly the water demand for agriculture sector using CROPWAT
model need to collect many information of different crop plants in cultivated area.
Therefore, the information relevant to water use of domestic is also necessary to
determine water demand of system.
One of limitations of this study, the author only used triangular membership
function for the inputs and outputs of Fuzzy system. Moreover, using other data as
rainfall, reservoir storage should pay attention as inputs of the Fuzzy system of Cua
Dat Reservoir. The fuzzy rules in this study were defined by opinion of the author,
79
Trinh Xuan Manh
MSc Thesis
however, it is strongly recommended that the fuzzy rules should derive from a program
or software in order to increase the efficiency of the system.
In future, it is necessary to set up more hydrological and meteorological stations
on Ma-Chu river basin to measure discharge, water level and meteorological factors;
Modeling future events should attend to the dynamic changing of factors driving the
changing of hydrological and hydraulic scheme of the river basin.
For next study, the author really wants to apply the Fuzzy Logic algorithm on
operation of a reservoir system to optimize operation of them and address all issues
relevant to the river basin which has many reservoirs located in.
80
Trinh Xuan Manh
MSc Thesis
REFERENCES
Bahremand A, De Smedt F. (2007). Distributed Hydrological Modeling and Sensitivity
Analysis in Torysa Watershed, Slovakia. Water Resources Management. 22:393–
408
Cheng C.T, Wang W.C, Xu D.M, Chau K. W. (2008). Optimizing Hydropower
Reservoir Operation Using Hybrid Genetic Algorithm and Chaos. Water
Resources Management 22:895–909
DHI software. (2011). MIKE 11 Reference Manual. Dansih Hydraulic Institute.
Hirad Mousavi, A.S. Ramamurthy. (2000). Optimal design of multi-reservoir systems
for water supply. Advances in Water Resources, 23: 613 - 624.
Hydrology Engineering Center (1991). Optimization of Multiple-Purpose Reservoir
System Operations: A Review of Modeling and Analysis Approaches. Research
Document No.34. Retrieved from : www.dtic.mil/dtic/tr/fulltext/u2/a236080.pdf
Habese M, Nagayama Y. (2002). Reservoir operation using Neural Network and fuzzy
system for dam control and operation support. Advances in engineering software
(33) 245 -26.
Institute of Water Resources Planning. (2003). Master Plan for Use and Protection of
Water Resources. 9-56 pp
Jairaj P. G, Vedula S. (2001). Multi-reservoir System Optimization using Fuzzy
Mathematical Programming. Water Resources Management 14: 457–472, 2000.
Kmenl A.H. (2008). Application of Hydraulic MIKE 11 model for the Euphrates river
in Iraq. Slovak journal of civil engineering 2008/2, 1-7 pp.
Khai N.H and Others. (2011). Research on technology to operate reservoir systems
which to prevent flood, regulate flood, operate reservoir safely and appropriately
using water resources in dry season. National Technology – Science Project (in
VietNamese)
Long N. L, Henrik Madsen, Dan Rosbjerg. (2007). Simulation and optimization
modeling approach for operation of the Hoa Binh reservoir, Viet Nam. Journal of
Hydrology, 336, 269 – 281
81
Trinh Xuan Manh
MSc Thesis
Mukand S. B, Chien N. D, Md. Reaz Akter Mullick, Umamahesh V. Nanduri. (2011).
Operation of a hydropower system considering environmental flow requirements:
A case study in La Nga river basin, Vietnam. Journal of Hydro-environment
Research, 6 63 – 73.
Luhandjula M.K., Rangoaga M.J. (2013). An approach for solving a fuzzy multiple
objective programming problem. European Journal of Operational Research,
232, 249 – 255.
Moeini R, Afshar A, Afshar M.H. (2010). Fuzzy rule-based model for hydropower
reservoirs operation. Electrical Power and Energy Systems, 33, 171-178.
MARD (2013). Cua Dat reservoir operation policy. 5-8 pp (In Vietnamese)
MARD (2014). Final engineering report of Cua Dat Reservoir in operation period.
Project of Cua Dat Reservoir.
Nagesh K. D, Falguni Baliarsingh, Srinivasa R.K. (2009). Optimal Reservoir Operation
for Flood Control Using Folded Dynamic Programming. Water Resources
Management (2010) 24:1045–1064.
Omid B. H, Abbas Afshar, Miguel A.M (2008). Design-Operation of Multi-
Hydropower Reservoirs: HBMO Approach. Water Resources Management
(2008) 22:1709–1722.
Piman T, Cochrane T. A, Arias M.E, Green A. and Dat N. D. (2012). Assessment of
Flow Changes from Hydropower Development and Operations in Sekong, Sesan
and Srepok Rivers of the Mekong Basin. Journal of Water Resources Planning
and Management, 1061/(ASCE)WR.1943-5452.0000286
Panigrahi D.P, Mujumdar P. P. (2000). Reservoir Operation Modelling with Fuzzy
Logic. Water Resources Management 14: 89–109, 2000.
Rama Mehta, Sharad K. Jain. (2009). Optimal Operation of a Multi-Purpose Reservoir
Using Neuro-Fuzzy Technique. Water Resources Management (2009) 23:509–
529.
82
Trinh Xuan Manh
MSc Thesis
Tuan L.Q. (2012) Impacts on flow regime caused by dam construction for hydropower
generation in the upper Se San river basin, using SWAT and WAFLEX models,
Msc thesis, UNESCO- IHE, Netherlands, 19 pp.
Vasantharajan S., Optimal Decisions Inc.; R. Al-Hussainy, Amerada Hess Ltd.; and
R.F. Heinemann, Berry Petroleum Co. (2006). Applying Optimization
Technology in Reservoir Management. JPT Distinguished Author Series.
Van Waveren R.H, Groot S, Scholten H, van Geer F.C, . Wösten J.H.M, Koeze R.D,
Noort J.J. (nd). (1999). Good modelling practice handbook. Dutch Dept. of Public
Works, Institute for Inland Water Management and Waste Water Treatment.
Source:
Tuyen, M.H. (2009). Research on optimal operation for reservoir system on Huong
river basin in dry season. Journal of Meteorology and Hydrology.
Hung, N.T, Hung L.N. (2010). Models for optimal operation of multiple-purpose
reservoir. Journal of Meteorology and Hydrology.
Nghia T.T. (2009). Building operation process for multi-reservoir including Hoa Binh,
Thac Ba, Tuyen Quang supplying water in dry season for downstream of Hong-
ThaiBinh river basin. Journal of Hydrology.
i
Trinh Xuan Manh
MSc Thesis
APPENDICES
Appendix 1: Water demand of some of crops in downstream area
Figure A-1: Seasonal period and chart of water requirement of Maize in 2011
Table A-1: Water requirement of Maize in 2011
Months Decades Stage Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Sep 3 Init 0.3 1.08 1.1 6.1 1.1
Oct 1 Init 0.3 1 10 47.5 0
Oct 2 Deve 0.3 0.93 9.3 38.2 0
Oct 3 Deve 0.48 1.42 15.6 27 0
Nov 1 Deve 0.75 2.14 21.4 10.2 11.2
Nov 2 Deve 1.01 2.77 27.7 0 27.7
Nov 3 Mid 1.2 3.03 30.3 3.3 26.9
Dec 1 Mid 1.2 2.79 27.9 11.5 16.4
Dec 2 Mid 1.2 2.53 25.3 15 10.4
Dec 3 Mid 1.2 2.47 27.2 10.2 17
Jan 1 Late 1.1 2.2 22 1.5 20.5
Jan 2 Late 0.82 1.6 16 0 16
Jan 3 Late 0.52 1 11 0.4 10.6
Feb 1 Late 0.35 0.66 0.7 0.1 0.7
Total 245.5 171 158.5
ii
Trinh Xuan Manh
MSc Thesis
Figure A-2: Seasonal period and chart of water requirement of Sweet potatoes in 2011
Table A-2: Water requirement of Sweet potatoes in 2011
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Oct 2 Init 0.5 1.53 1.5 3.8 1.5
Oct 3 Init 0.5 1.48 16.3 27 0
Nov 1 Init 0.5 1.43 14.3 10.2 4.1
Nov 2 Deve 0.56 1.54 15.4 0 15.4
Nov 3 Deve 0.77 1.96 19.6 3.3 16.3
Dec 1 Deve 0.99 2.31 23.1 11.5 11.6
Dec 2 Mid 1.15 2.43 24.3 15 9.3
Dec 3 Mid 1.16 2.38 26.2 10.2 16
Jan 1 Mid 1.16 2.32 23.2 1.5 21.7
Jan 2 Mid 1.16 2.26 22.6 0 22.6
Jan 3 Late 1.15 2.2 24.2 0.4 23.9
Feb 1 Late 1.03 1.95 19.5 1 18.4
Feb 2 Late 0.89 1.66 16.6 1.3 15.3
Feb 3 Late 0.78 1.49 8.9 5 5.6
Total 255.7 90.1 181.7
iii
Trinh Xuan Manh
MSc Thesis
Figure A-3: Seasonal period and chart of water requirement of Spring paddy in 2012
Table A-3: Water requirement of Spring paddy in 2012
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Sep 2 Nurs 1.04 0.38 0.4 5.7 0.4
Sep 3 Nurs/LPr 1.04 0.7 7 53.8 20.4
Oct 1 Nurs/LPr 1.08 3.61 36.1 49.7 0
Oct 2 Init 1.09 3.52 35.2 46.9 70
Oct 3 Init 1.25 3.75 41.3 44.8 0
Nov 1 Deve 1.26 3.49 34.9 43.6 0
Nov 2 Deve 1.44 3.67 36.7 41.9 0
Nov 3 Deve 1.68 4.04 40.4 36.6 3.8
Dec 1 Mid 1.92 4.31 43.1 31.2 11.9
Dec 2 Mid 1.99 4.15 41.5 26.5 15
Dec 3 Mid 1.99 3.84 42.3 20.1 22.1
Jan 1 Mid 1.99 3.53 35.3 12.2 23.2
Jan 2 Late 1.97 3.19 31.9 5.3 26.6
Jan 3 Late 1.68 2.8 30.8 5 25.7
Feb 1 Late 1.32 2.25 22.5 4.6 17.9
Feb 2 Late 1.04 1.83 11 1.9 9.4
490.4 429.9 246.4
iv
Trinh Xuan Manh
MSc Thesis
FigureA-4: Seasonal period and chart of water requirement of winter paddy in 2012
Table A-4: Water requirement of winter paddy in 2012
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Jul 1 Nurs 1.04 0.49 2.5 21.9 0
Jul 2 Nurs/LPr 1.06 2.76 27.6 44 21.2
Jul 3 Nurs/LPr 1.08 4.85 53.3 46.7 76.6
Aug 1 Init 1.18 5.12 51.2 50.1 55.3
Aug 2 Init 1.25 5.2 52 52.7 0
Aug 3 Deve 1.26 5.03 55.3 53.6 1.7
Sep 1 Deve 1.31 4.99 49.9 55.4 0
Sep 2 Mid 1.35 4.88 48.8 57.1 0
Sep 3 Mid 1.35 4.71 47.1 53.8 0
Oct 1 Mid 1.35 4.53 45.3 49.7 0
Oct 2 Late 1.29 4.16 41.6 46.9 0
Oct 3 Late 1.07 3.21 35.3 44.8 0
Nov 1 Late 0.93 2.58 5.2 8.7 5.2
Total 514.9 585.5 159.9
:
v
Trinh Xuan Manh
MSc Thesis
Figure A-5: Seasonal period and chart of water requirement of sugar cane in 2012
Table A-5: Water requirement of sugar cane in 2012
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
May 1 Init 0.79 3.11 3.1 2.9 0
May 2 Init 0.4 1.73 17.3 39.7 0
May 3 Init 0.4 1.8 19.8 40.9 0
Jun 1 Deve 0.4 1.89 18.9 41.5 0
Jun 2 Deve 0.5 2.44 24.4 44 0
Jun 3 Deve 0.64 3.08 30.8 44.3 0
Jul 1 Deve 0.78 3.71 37.1 43.8 0
Jul 2 Deve 0.92 4.32 43.2 44 0
Jul 3 Deve 1.07 4.83 53.1 46.7 6.3
Aug 1 Mid 1.21 5.24 52.4 50.1 2.3
Aug 2 Mid 1.24 5.16 51.6 52.7 0
Aug 3 Mid 1.24 4.93 54.2 53.6 0.6
Sep 1 Mid 1.24 4.71 47.1 55.4 0
Sep 2 Mid 1.24 4.48 44.8 57.1 0
Sep 3 Mid 1.24 4.32 43.2 53.8 0
Oct 1 Mid 1.24 4.15 41.5 49.7 0
Oct 2 Mid 1.24 3.99 39.9 46.9 0
Oct 3 Mid 1.24 3.72 40.9 44.8 0
vi
Trinh Xuan Manh
MSc Thesis
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Nov 1 Mid 1.24 3.44 34.4 43.6 0
Nov 2 Mid 1.24 3.17 31.7 41.9 0
Nov 3 Mid 1.24 2.97 29.7 36.6 0
Dec 1 Mid 1.24 2.78 27.8 31.2 0
Dec 2 Mid 1.24 2.58 25.8 26.5 0
Dec 3 Mid 1.24 2.39 26.3 20.1 6.1
Jan 1 Mid 1.24 2.2 22 12.2 9.8
Jan 2 Mid 1.24 2.01 20.1 5.3 14.8
Jan 3 Mid 1.24 2.06 22.7 5 17.6
Feb 1 Late 1.22 2.09 20.9 4.6 16.3
Feb 2 Late 1.18 2.06 20.6 3.2 17.4
Feb 3 Late 1.13 2.12 16.9 5.9 11
Mar 1 Late 1.09 2.16 21.6 10 11.6
Mar 2 Late 1.04 2.18 21.8 12.8 9.1
Mar 3 Late 0.98 2.44 26.9 11.1 15.8
Apr 1 Late 0.93 2.66 26.6 5.9 20.7
Apr 2 Late 0.88 2.85 28.5 3.2 25.3
Apr 3 Late 0.83 2.99 29.9 14.3 15.6
May 1 Late 0.79 3.11 28 26.1 0
Total
1145.4 1131.5 200.4
vii
Trinh Xuan Manh
MSc Thesis
Figure A-6: Seasonal period and chart of water requirement of Maize in 2012
Table A-6: Water requirement of Maize in 2012
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
May 2 Init 0.3 1.3 7.8 23.8 0
May 3 Init 0.3 1.35 14.8 40.9 0
Jun 1 Deve 0.37 1.73 17.3 41.5 0
Jun 2 Deve 0.62 2.98 29.8 44 0
Jun 3 Deve 0.87 4.16 41.6 44.3 0
Jul 1 Mid 1.12 5.28 52.8 43.8 9
Jul 2 Mid 1.19 5.55 55.5 44 11.5
Jul 3 Mid 1.19 5.35 58.8 46.7 12.1
Aug 1 Mid 1.19 5.14 51.4 50.1 1.3
Aug 2 Late 1.17 4.87 48.7 52.7 0
Aug 3 Late 0.94 3.72 41 53.6 0
Sep 1 Late 0.64 2.44 24.4 55.4 0
Sep 2 Late 0.42 1.52 9.1 34.2 0
Total
453 575.1 33.9
viii
Trinh Xuan Manh
MSc Thesis
Figure A-7: Seasonal period and chart of water requirement of Sweet Potatoes in 2012
Table A-7: Water requirement of Sweet Potatoes in 2012
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Oct 2 Init 0.5 1.61 1.6 4.7 1.6
Oct 3 Init 0.5 1.5 16.5 44.8 0
Nov 1 Init 0.5 1.39 13.9 43.6 0
Nov 2 Deve 0.56 1.44 14.4 41.9 0
Nov 3 Deve 0.77 1.86 18.6 36.6 0
Dec 1 Deve 0.99 2.22 22.2 31.2 0
Dec 2 Mid 1.15 2.4 24 26.5 0
Dec 3 Mid 1.16 2.23 24.5 20.1 4.4
Jan 1 Mid 1.16 2.05 20.5 12.2 8.3
Jan 2 Mid 1.16 1.87 18.7 5.3 13.5
Jan 3 Late 1.14 1.9 20.9 5 15.9
Feb 1 Late 1.03 1.76 17.6 4.6 13
Feb 2 Late 0.9 1.58 15.8 3.2 12.5
ix
Trinh Xuan Manh
MSc Thesis
Months Decades Stages Kc ETc ETc Eff rain Irr. Req.
coeff mm/day mm/dec mm/dec mm/dec
Feb 3 Late 0.79 1.49 8.9 4.4 6
Total
238.2 284.3 75.2
Table A-8: Monthly water demand of agriculture of whole area in the Cua Dat
reservoir downstream in 2012
Water demand of Bac Song Chu cultivated area in 2012
Months 1 2 3 4 5 6 7 8 9 10 11 12 Year
Q(m3/s) 12.7 15.1 16.0 17.1 12.7 26.6 18.9 8.03 4.01 5.19 8.03 33.5 14.8
W(106m3) 34.5 36.9 43.4 44.1 34.5 69.0 51.0 21.9 11.4 14.8 21.2 90.2 473.8
Water demand of Nam Song Chu cultivated area in 2012
Months 1 2 3 4 5 6 7 8 9 10 11 12 Year
Q(m3/s) 42.4 42.7 61.3 57.1 22.3 0.226 16.3 0.117 0.389 0.489 0.231 14.5 21.5
W(106m3) 113.7 103.3 164.2 148.1 59.8 0.6 43.6 0.3 1.0 1.3 0.61 39.0 675.9
Water demand of agriculture of whole area in 2012
Months 1 2 3 4 5 6 7 8 9 10 11 12 Year
Q(m3/s) 55.2 57.8 77.4 74.1 35.0 26.8 35.2 8.15 4.39 5.68 8.27 48.1 36.3
W(106m3) 148.3 140.2 207.7 192.6 94.3 69.5 94.7 22.2 12.4 16.1 21.8 129.2 1149.7
x
Trinh Xuan Manh
MSc Thesis
Appendix 2: Water demand in ten-day period in year of 2012
Table A-9: Water demand in ten-day period in 2012
Months
Period Demand
in ten-day unit in M.m3
Jan 1-10
73.57
Jan 11-20
59.08
Jan 21-31
63.18
Feb 1-10
83.53
Feb 11-20
87.92
Feb 21-28
79.82
Mar 1-10
87.19
Mar 11-20
86.16
Mar 21-31
96.58
Apr 1-10
71.40
Apr 11-20
101.15
Apr 21-30
55.45
May 1-10
50.53
May 11-20
59.67
May 21-31
102.33
Jun 1-10
112.62
Jun 11-20
107.10
Jun 21-30
48.04
Jul 1-10
46.21
Jul 11-20
60.94
Jul 21-31
72.50
Aug 1-10
37.00
Aug 11-20
63.07
Aug 21-30
100.00
Sep 1-10
112.11
Sep 11-20
124.16
Sep 21-31
118.82
xi
Trinh Xuan Manh
MSc Thesis
Months
Period Demand
in ten-day unit in M.m3
Oct 1-10
121.28
Oct 11-20
120.93
Oct 21-31
131.08
Nov 1-10
115.59
Nov 11-20
106.62
Nov 21-30
119.59
xii
Trinh Xuan Manh
MSc Thesis
Appendix 3: Observed and calculated data of inputs and outputs in the Fuzzy system
Figure A-8: Average daily discharge into the Cua Dat Reservoir in 2011 and 2012
Figure A-9: Average daily water level of the Cua Dat Reservoir in 2011
Figure A-10: Average daily water level of the Cua Dat Reservoir in 2012
0
500
1000
1500
2000
2500
3000
0 50 100 150 200 250 300 350
D
is
ch
a
rg
e
(m
3
/s
)
Days2011 2012
75.00
80.00
85.00
90.00
95.00
100.00
105.00
1 51 101 151 201 251 301 351
W
a
te
r
l
ev
el
(
m
)
Days
70
75
80
85
90
95
100
105
1 51 101 151 201 251 301 351
W
a
te
r
l
ev
el
(
m
)
Days
xiii
Trinh Xuan Manh
MSc Thesis
Figure A-11: Average daily turbin discharge of the Cua Dat hydropower plant in 2012
Figure A-12: Average daily turbin discharge of the Cua Dat hydropower plant in 2011
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
1 21 41 61 81 101 121 141 161 181 201 221 241 261 281 301 321 341 361
T
u
rb
in
d
is
ch
a
rg
e
(m
3
/s
)
Days
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
1
1
3
2
5
3
7
4
9
6
1
7
3
8
5
9
7
1
0
9
1
2
1
1
3
3
1
4
5
1
5
7
1
6
9
1
8
1
1
9
3
2
0
5
2
1
7
2
2
9
2
4
1
2
5
3
2
6
5
2
7
7
2
8
9
3
0
1
3
1
3
3
2
5
3
3
7
3
4
9
3
6
1
T
u
rb
in
d
is
c
h
a
r
g
e
(m
3
/s
)
Days
xiv
Trinh Xuan Manh
MSc Thesis
Appendix 4: Inputs of MIKE 11 HD model
Figure A-13: Average daily discharge at the Cam Thuy and Cua Dat station in 2006
Figure A-14: Average daily water level at the Hoang Tan and Kim Tan station in 2006
0
500
1000
1500
2000
2500
3000
1 51 101 151 201 251 301 351
D
is
ch
a
g
re
(
m
3
/s
)
Days
Cam Thuy
Cua Dat
-2
0
2
4
6
8
10
1 51 101 151 201 251 301 351
W
a
te
r
L
ev
el
(m
3
/s
)
Days
Kim Tan
Hoang Tan
xv
Trinh Xuan Manh
MSc Thesis
Figure A-15: Average daily discharge at the Cam Thuy and Cua Dat station in 2008
Figure A-16: Average daily water level at the Hoang Tan and Kim Tan station in 2008
0
500
1000
1500
2000
2500
3000
3500
4000
1 51 101 151 201 251 301 351
D
is
ch
a
rg
e
(m
3
/s
)
Days
Cam Thuy
Cua Dat
-2
0
2
4
6
8
10
12
14
1 51 101 151 201 251 301 351
W
a
te
r
l
ev
el
(m
)
Days
Kim Tan
Hoang Tan
xvi
Trinh Xuan Manh
MSc Thesis
Appendix 5: Fuzzy rules base for the Cua Dat Reservoir
Rules
No. Res.levels (and) Inflows (and) Demands (Then) Releases
1 If Med.High V.Low Low Low
2 If Low.med V.Low low Low
3 If Low.med V.Low low Low.Med
4 If Low.med V.Low Low Medium
5 If Low.med V.Low V.Low Low
6 If Low.med V.Low V.Low V.Low
7 If Low V.Low Very.High Low.med
8 If Medium V.Low V.High Med.High
9 If Low V.Low Low V.Low
10 If Low V.Low Medium Low.Med
11 If Low.Med V.low Medium Medium
12 If Low Low V.Low Low
13 If Medium Low Low Low
14 If Low.Med V.Low Low.Med Medium
15 If V.Low V.Low Medium Low
16 If V.Low V.Low High High
17 If V.Low V.Low Low Low
18 If V.Low V.Low V.Low V.Low
19 If V.Low Low Low Low
20 If V.Low Low Medium Low.Med
21 If V.Low Low Medium Medium
22 If Low Low Medium Medium
23 If Low Low Medium Low.Med
24 If Low Med.High High Med.High
25 If Low Med.High High High
26 If Low.med Med.High High High
27 If Medium Medium High High
28 If Medium Med.High High High
29 If Med.High medium High High
30 If Med.High Med.High High High
31 If Medium Very.High High High
32 If Med.High Very.High High High
33 If Med.High Medium Very.High V.V.High
34 If High Medium Very.High V.V.High
35 If High Med.High High V.V.High
36 If Medium Low Very.High V.V.High
37 If Med.High Low Very.High V.V.High
38 If Medium Very.High Very.High V.V.High
39 If Med.High Very.High Very.High V.V.High
40 If Medium Medium High V.V.High
41 If Medium Medium Very.High V.V.High
xvii
Trinh Xuan Manh
MSc Thesis
Rules
No. Res.levels (and) Inflows (and) Demands (Then) Releases
42 If Medium High High High
43 If Medium High Very.High V.V.High
44 If Med.High Medium High V.V.High
45 If Med.High Medium Very.High V.V.High
46 If Med.High High High V.V.High
47 If Med.High High Very.High V.V.High
48 If Medium Low High High
49 If Medium Low Very.High High
50 If Med.High Low High High
51 If Med.High Low Very.High High
52 If High Low High V.V.High
53 If High Low Very.High V.V.High
54 If Medium Low Medium Medium
55 If Medium Low Medium Med.high
56 If Med.High Low Medium medium
57 If Med.High Low Medium Med.high
58 If Med.High Low High Med.High
59 If Medium Low high High
60 If Medium V.Low Low Low.Med
61 If Medium V.Low Low Med
62 If Medium V.Low Medium Med
63 If Medium V.Low Medium Low.Med
64 If Med.High Low Low Low
65 If Medium Low Low Low
66 If Med.High Low Low Low
67 If Med.High Low Low Medium
68 If High Low Low Medium
69 If Med.High Low Medium Medium
70 If high Low Medium Medium
71 If Med.High V.low Low Low.med
72 If High V.low Low Low.med
73 If Med.High V.low Low Low
74 If Med.High V.low Medium Low
75 If Med.High Low Low Low
76 If Med.High Low Medium Low
77 If High V.low Low Low
78 If High V.low Medium Low
79 If High Low Low Low
80 If High Low Medium Low
81 If Med.High V.low Medium Medium
82 If High V.low Medium Medium
83 If Medium V.low Medium Medium
84 If Medium V.low Medium Med.high
xviii
Trinh Xuan Manh
MSc Thesis
Rules
No. Res.levels (and) Inflows (and) Demands (Then) Releases
85 If Medium V.low High Med.high
86 If Medium V.low High Medium
87 If Medium Low Medium Medium
88 If Medium Low Medium Med.high
89 If Med.High V.low Medium Medium
90 If Med.High V.low Medium Med.high
91 If Med.High V.low High Med.high
92 If Med.High V.low High Medium
93 If Med.High Low Medium Medium
94 If Med.High Low Medium Med.high
95 If Low.med v.low Medium Med.high
96 If Low V.low High Med.high
97 If Low.med V.low High Med.high
98 If Low V.low Low Low
99 If Low V.low Low Low.med
100 If Low.med v.low Low Low
101 If Low.med V.low Low Low.med
102 If V.Low V.low V.low Low.med
103 If V.Low V.low Low Low.med
104 If V.Low V.low Low Medium
105 If Low V.low Low Medium
106 If V.Low Low High Med.high
107 If V.Low Low High High
108 If Low Low High Med.high
109 If Low Low High High
110 If Low V.low High High
111 If Low.Med V.low High High
112 If Low.Med V.low V.high High
113 If Low V.low V.high High
114 If V.Low V.low Low V.low
115 If V.Low V.low Medium V.low
116 If Low V.low Low V.low
117 If Low M.low Medium V.low
118 If Low Med.high V.low Low.med
119 If Low Med.high Low Low.med
120 If Low.med Med.high High Med.high
121 If Low.med High High Med.high
122 If Low.med Med.high High High
123 If Low.med High High High
124 If Medium Med.high High Med.high
125 If Medium High High Med.high
126 If Medium Med.high High High
127 If Medium High High High
xix
Trinh Xuan Manh
MSc Thesis
Rules
No. Res.levels (and) Inflows (and) Demands (Then) Releases
128 If Med.High High High high
129 If Med.High High V.high High
130 If Medium High V.high High
131 If Medium Medium High High
132 If Medium Medium V.high High
133 If Medium Med.high High High
134 If Medium Med.high V.high High
135 If High High V.high High
136 If High Medium High High
137 If High Medium V.high High
138 If High Med.high High High
139 If High Med.high V.high High
140 If Medium Medium High V.v.high
141 If Medium Medium V.high V.v.high
142 If High Medium High V.v.high
143 If High Medium V.high V.v.high
144 If Medium Low High V.v.high
145 If Medium Low V.high V.v.high
146 If High Low High V.v.high
147 If High Low V.high V.v.high
148 If Medium V.low High V.v.high
149 If Medium V.low V.high V.v.high
150 If Med.High V.low High V.v.high
151 If Med.High V.low V.high V.v.high
152 If medium V.low High Med.high
153 If Med.high V.low High Med.high
154 If Low.med V.low High High
155 If Low.med Low High High
156 If Medium V.low High High
157 If Medium Low High High
xx
Trinh Xuan Manh
MSc Thesis
Appendix 6: Existing operation rule curve of the Cua Dat Reservoir
Table A-10: Operation curves of the Cua Dat Reservoir
Periods Crumble curve Limited curve
30/VI 97 73
31/VII 100 82
31/VIII 104 85
30/IX 109 96
31/X 110 105
30/XI 112 106
31/XII 112 106
31/I 112 103
28/II 108 97
31/III 105 90
30/IV 103 83
31/V 99 77
30/VI 97 73
Figure A-17: The map of operation curves of the Cua Dat Reservoir
70
80
90
100
110
120
1/7 1/8 1/9 1/10 1/11 1/12 1/1 1/2 1/3 1/4 1/5 1/6 1/7
B
A
B
C C
D
E
Flood control
increment level=110,00m
Max Water level=119,05m
Dam elevation 121.30 m
3
1
2
Dead storage level=73m
Flood control level=112,00m

Các file đính kèm theo tài liệu này:

- niche_thesis_trinhxuanmanh_2014_4792_2085180.pdf