See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332049265

Analysis of Flood Routing in Channels of Ban Sagar Project: A Case Study

Article in SSRN Electronic Journal · January 2019

DOI: 10.2139/ssrn.3353099

CITATIONS READS

0 1,679

3 authors, including:

Lalitesh Sinha Sandeep Narulkar

IPS Academy Institute of Engineering and Science Shri Govindram Seksaria Institute of Technology and Science
9 PUBLICATIONS 2 CITATIONS 15 PUBLICATIONS 181 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hydrological Study of Lakes in Urban and Peri Urban Areas: A case study on Khajrana Lake View project

Experimental study and modeling of infiltration in black cotton soil View project

All content following this page was uploaded by Lalitesh Sinha on 16 May 2019.

The user has requested enhancement of the downloaded file.

 Available online at http://ssrn.com
SSRN-ELSEVIER (2018-2019)

International conference on “Recent Advances in Interdisciplinary Trends in

Engineering & Applications”

Analysis of Flood Routing in Channels of Ban

Sagar Project: A Case Study
Lalitesh Sinha a,*, Sandeep M. Narulkar b, Archana K.Chowdharya,
a
Department of Civil Engineering, IPS Academy Institute of Engineering and Science, Indore,452012,India.
b
Professor, Department of Civil Engineering & Applied Mechanics, S.G.S.I.T.S, Indore, 452003, India.

Abstract
The precision of flood routing is an important in hydrology and hydraulics. Exact information
of the flood peak attenuation and the duration of the high water levels obtained by channel
routing are of most significance in flood forecasting operations and flood protection works. This
study implements Muskingham method to estimate the inflow and outflow discharge at the Sone
river bank, Sirmor, Rewa-District. The parameters of Muskingum method are determined using
two methods i.e. graphical method and least square method and assessment of results are
compared by using these two methodologies. The result of the study will serve useful purposes in
design of flood control and predicting flood events in similar basins.
Keywords: Flood Routing; Hydrological Method; Parameter Estimation; Muskingham Method; Graphical
Method; Least Square Method.

1. Introduction
During rainfall, some portion of water is retained by the soil depending on the degree
of saturation, some are absorbed by vegetation, some evaporate and the reminder,
which reaches stream channels, is called runoff. Flood is a high stage of river, which is
caused by high tides, heavy rainfall and then run off in stream channels and constructed
reservoir. In the past, India has witnessed many of the largest floods, causing
permanent damage to people’s livelihood, property and infrastructure in which state
Madhya Pradesh struggles with heavy monsoon rain and ensures flood. Due to heavy
rainfall in the region Madhya Pradesh various flood routing methods are used to control
flood. Flood routing is a tool to creating hydrograph at a specific downstream area of a
waterway stream by the use of past information. The types of flood routing are
reservoir routing and channel routing. A variety of routing methods are available and
they can be grouped in to hydrologic routing and hydraulic routing.

Nomenclature

S reaches storage (m3)

K storage constant, (sec)
X weighing factor, (dimensionless)
I inflow discharge, (m3/sec)
Q outflow discharge, (m3/sec)
∆t routing interval, (hours)

Corresponding author. Tel.: +91-9926410889.

E-mail address: [email protected]

For flood routing various methods are available in literature such as (Ranaee et. al, 2010)
described a procedure to study the sensitivity of flood routing models for different time steps of

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

input discharge statistics and they used MIKE11 software for flood routing and used artificial
neural networks to compute sensitivity of river flood routing. (Kumar et. al, 2011) extended
Muskingum method for routing of floods they derived various types of coefficient using linear
programming instead of time lag and prismoidal flow parameters. (Doiphode and Oak, 2012)
studied the phenomenon of unsteady and non uniform complicated flow in river channel due to
movement of flood wave and estimated the changes in the hydrographs under the estimated
worst scenarios. (Soleymani and Delphi, 2012) compared some flood routing models including
wave models and numerical models in the real field. (Mahessar et. al., 2013) used flood routing
to determine the level discharge and flood hydrograph at different reaches of channels. The
results of numerical model had been compared to analyze solution of observed data in Indus
River. (Ogbonna et. al, 2017) focused on the application of Flood Routing Models for Flood
Mitigation in Orashi River, South-East Nigeria. Three flood Routing models were comparatively
applied including Muskingum model, Level Pool model and Modified Pul’s model for reservoir.
Recently (Pandey et.al, 2018) analyze the parameters of Muskingham method are determined by
using three methods i.e. graphical method, least square method and regression analysis
approach and the assessment of results are compared. Present study aims at channel routing
during flood in Ban Sagar project on river Sone in India.

2. Description of Study Area and Data Acquired

To demonstrate the application of Muskingum Method for flood routing of channel Ban Sagar
project is chosen as a case study. It is a multipurpose river valley project on river Sone in the
Ganges basin in Madhya Pradesh, India (Location 24°-11'-30''-N and 81°-17'-15''-E). The Ban
Sagar Dam constructed across the Sone River near the Deoland village in the Shahdol district. It
is surrounded by Katni, Rewa, Satna and Shahdol district. . The project comprises of a 67m high
and 1020m long Masonry/Earthen dam. The Ban Sagar Canal Project started in 1978 was
handed over to the nation by PM Narendra Modi on 15-July-2018 and utilize water from the
Ban Sagar Reservoir. For modelling of channel routing Sihawal Canal System is selected during
flood from 9-Aug-2018 to 11-Aug-2018 and collected flow data from Water Resources
Department (WRD), Madhya Pradesh. For the present study inflow data during flood was
acquired as shown in Table 1 and Table 2. To assess the accuracy and reliability of collected
data and used for flood routing.

3. Methodology

The study focused on the comparison of results of flood routing inflow and outflow discharge
using Muskingham methods were applied. The parameters of Muskingham method are
determined using two methods namely graphical method and least square method. The
reliability of above two methods depends on data availability and the time for calculations. The
graphical method is the oldest method to calculate the parameters of Muskingum method but the
least square method is more fast and accurate in comparison to graphical method.

3.1 Muskingum Method Parameters Determination by Graphical Method:

In the basic Muskingum method, if inflow and outflow data of the reach is available the value of
‘K’ and ‘X’ can be estimated graphically of the given reach. But if ‘S’ is plotted against
{XI + (1-X) O}, a straight line with a slope of ‘K’ should outcome. Numerous values of ‘X’; are
tried; the value that gives the narrowest loop in the plotted relationship is taken as the correct
‘X’ value and the slope of the plotted relationship is taken as the ‘K’ value. When X=0.35 then
the value of ‘K’ is straight line of the narrowest loop. The Muskingum method assumes a single-
discharge relationship and this assumption causes an effect known as hysteresis, which may
introduce errors into the storage calculation. The hysteresis effect between reach storage and
discharge is due to the dissimilar flood wave speeds during the rising and falling of the

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

hydrograph and for the same river stage, the wave of flood moves faster during the rising limb
of hydrograph. After number of studies has been carried out to analyze that ‘K’ can be easily
estimated from the observed inflow and outflow data. Cunge (1967) has received widespread
acceptance to estimate the model parameters without the observed hydrograph. While in the
absence of flow data, the Muskingum-Cunge method may be used for parameter estimation.
This concept is applied when measured flow data are available, but with significant degree of
uncertainty.
The Muskingum-Cunge parameters are computed based on the flow and the channel
characteristics. This method involves the use of a finite difference scheme for solving the
Muskingum equation, where parameters are determined based on the grid spacing for finite
difference scheme and the channel geometry characteristics.

3.1.1 Formulas used generally for estimating storage are:

For a given channel reach by selecting a routing interval ∆t and using the Muskingum Equation,
storage is shown in Eq. (1).

S=K [XI+ (1-X) Q] (1)

Table 1. Parameter determination using Graphical approach for the below data.

Outflow Storage Weighted Average Flow

Inflow (I) Average
3 (Q) (I-Q) S=∑∆S [XI+(1-X)Q] (m3/sec)
(m /sec) (I-Q)
(m3/sec) 3
{(m /sec).h} X=0.25 X=0.35 X=0.10 X=0.15
53 48 5 45 49.25 49.75 48.5 48.75
77 61 16 10.5 76.92 65 66.6 62.6 63.4
85 72 13 14.5 121.00 75.25 76.55 73.3 73.95
92 83 9 11 154.44 85.25 86.15 83.9 84.35
102 90 12 10.5 186.36 93 94.2 91.2 91.8
116 98 18 15 231.96 102.5 104.3 99.8 100.7
128 110 18 18 286.68 114.5 116.3 111.8 112.7
134 121 13 15.5 333.80 124.25 125.55 122.3 122.95
135 128 7 10 364.20 129.75 130.45 128.7 129.05
148 140 8 7.5 387.00 142 142.8 140.8 141.2
156 156 0 4 399.16 156 156 156 156
144 152 -8 -4 387.00 150 149.2 151.2 150.8
131 141 -10 -9 359.64 138.5 137.5 140 139.5
125 134 -9 -9.5 330.76 131.75 130.85 133.1 132.65
118 124 -6 -7.5 307.96 122.5 121.9 123.4 123.1
111 119 -8 -7 286.68 117 116.2 118.2 117.8
88 115 -27 -17.5 233.48 108.25 105.55 112.3 110.95
79 96 -17 -22 166.60 91.75 90.05 94.3 93.45
58 79 -21 -19 108.84 73.75 71.65 76.9 75.85
47 57 -10 -15.5 -47.12 61.72 54.5 53.5 56

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

Fig. 1. (a) Storage vs Weighted Average Flow for determination of K and X.

Fig. 1. (b) Storage vs Weighted Average Flow for determination of K and X.

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

Fig. 1. (c) Storage vs Weighted Average Flow for determination of K and X.

Fig. 1. (d) Storage vs Weighted Average Flow for determination of K and X.

From the above Fig.1(a),Fig.1(b),Fig.1(c) and Fig.1(d) as shown, the best fit line for least square
is for X=0.35. Slope of the line for X =0.35 is K & given by K=2.30h as shown in Fig.1(c).

3.2 Muskingum Method Parameters Determination by Least Square Method:

It consisting in generating graph of [XI+ (1-X) Q] vs Storage (S) for dissimilar value of X
arbitrarily selected such that 0<X<0.5.the best value of X is selected as that which generate
narrowest and straightest loop graph of [XI+ (1-X) Q] vs X. the slope of the least square fit to
resulting point is the estimate of ‘K’.
3.2.1 Formulas used generally for estimating storage are:

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

For estimating change in storage, inflow and outflow with respect to ∆t and i=1, 2…n as shown
in Eq. (2).
Si+1=Si + {∆t/ (Ii+1-Ii)} - {∆t/2(Qi+1+Qi)} (2)

3.2.2Formulas for least Square Approach:

$ # $ $ $ $ # $ $ $
" !" . " &" '" - " !" *" . " &" '" " !" . " &" '" - " '" *" . " &" !"
A= $ ' # . $ ! # -[ $ ! * ]# B= $ ! # . $ ' # -[ $ ' * ]#
" " " " " " " " " " " " " "
Now using least square method we find out the value of ‘K’ and ‘X’.

Table 2. Parameter determination using Least Square approach for the below data.

Inflow (I) Outflow (Q) Storage (S) Q2 I2 QI SQ SI

(m3/sec) (m3/sec) (m3/sec) (m3/sec)2 (m3/sec)2 (m3/sec)2 (m6/sec) (m6/sec)

53 48 48.00 2304 2809 2544 2304 2544

77 61 77.34 3721 5929 4697 4717.7179 5955.152
85 72 121.77 5184 7225 6120 8767.6481 10350.7
92 83 155.63 6889 8464 7636 12917.5 14318.19
102 90 186.36 8100 10404 9180 16772.4 19008.72
116 98 232.14 9604 13456 11368 22749.323 26927.77
128 110 286.68 12100 16384 14080 31534.8 36695.04
134 121 333.80 14641 17956 16214 40389.8 44729.2
135 128 364.20 16384 18225 17280 46617.6 49167
148 140 387.00 19600 21904 20720 54180 57276
156 156 399.16 24336 24336 24336 62268.96 62268.96
144 152 387.00 23104 20736 21888 58824 55728
131 141 359.64 19881 17161 18471 50709.24 47112.84
125 134 330.76 17956 15625 16750 44321.84 41345
118 124 307.96 15376 13924 14632 38187.04 36339.28
111 119 286.68 14161 12321 13209 34114.92 31821.48
88 115 233.48 13225 7744 10120 26850.2 20546.24
79 96 166.60 9216 6241 7584 15993.6 13161.4
58 79 108.84 6241 3364 4582 8598.36 6312.72
47 57 61.72 3249 2209 2679 3518.04 2900.84
2
∑Q= ∑ ∑ QI= ∑ SQ= ∑ SI=
245272 I2=246417 244090 583995.52 584121.4

After determination of ‘A’and ‘B’ value we can determine the parameter ‘K’ as shown in Eq.(3).

K=A+B (3)
A=0.844942sec. B=1.553131sec
K=A+B
K=2.398073/3600
K=2.3hours
From the value ‘A’and ‘B’ we can calculated the parameter ‘X’ as shown in Eq. (4).
X= A/ (A + B) (4)
X= A/ (A + B) = 0.35

Electronic copy available at: https://ssrn.com/abstract=3353099

 Lalitesh Sinha/ SSRN-ELSEVIER (2018-19) 000–000

4. Conclusion

Based on the result of flood routing obtained from the Ban Sagar Project of Sone River the
following conclusions can be made.
• In the above study, we have found that all the two methods are equally liable for exact
calculation of parameters of Muskingum method. Thereby we can use anyone of the
method for the calculation of Muskingum method.
• In graphical method, we have found that linear shape of the graph is coming from the
value of X=0.35 and slope of the line for X= 0.35 gives value of K as K=2.30 hours.
• Also in least square method, we have found that the value of ‘X’ for which the best
relationship is found is for X =0.35, value of K=2.30 hours.
• For further study we analyzed flood routing of channel as well as reservoir of Ban
Sagar Canal project.

Acknowledgement

The authors wish to thank Water Resources Department (WRD), Rewa (Madhya Pradesh) for
providing a data for study and research. Authors are also thankful to Dr. S.M.Narulkar,
Professor, Civil Engineering Department, S.G.S.I.T.S, Indore, for his continuous guidance,
patient advice and encouragement at all stages of this work.

References

Doiphode, S.L. and Oak, R.K.(2012). Dynamic flood routing and unsteady flow modelling: A
case study of upper Krishna river, International Journal of Advances in Engineering and
Technology, 3(3), 55-59
Kumar, D. N., Baliarsingh, F., & Raju, K. S. (2011). Extended Muskingum method for flood
routing. Journal of Hydro-environment Research, 5(2), 127-135.
Mahessar, A. A., Qureshi, A. L., & Baloch, A. (2013). Numerical study on flood routing in
Indus river. International Water Technology Journal, 3(1), 3-12.
Ogbonna, D., Okoro, B. C., & Osuagwu, J. C. (2017). Application of Flood Routing Model for
Flood Mitigation in Orashi River, South-East Nigeria. Journal of Geoscience and Environment
Protection, 5(03), 31.
Pandey, R., Kumar, V. and Pandey, R.D. (2018) .Analysis of flood routing in channels.
International Journal of Engineering Development and Research, 6(1),11-18
Ranaee, E., Mahmoodian, M., & Rahati, S. (2010). Sensitivity Analysis of River Flood Routing
Model to Input Data. International Journal of Environmental Science and Development, 1(3),
256.
Soleymani, M., & Delphi, M. (2012). Comparison of flood routing models (case study: Maroon
River, Iran). World Appl. Sci. J, 16(5), 769-775.
.

View publication stats Electronic copy available at: https://ssrn.com/abstract=3353099