Lecture Notes ECE 2410: Hydrology II by Dr. J.

K Mwangi

FLOOD ROUTING
Flood routing is the process of determining a flood hydrograph at a location downstream of a
reservoir / channel section from knowledge of the upstream inflow hydrograph. The net effect of
the reservoir / channel section is to modify the flood hydrograph. If interest is to know the flood
hydrograph from an existing / proposed reservoir it is called reservoir routing while if it is to
know the effect of the channel section to an inflow hydrograph at a D/S location as it propagates
it is called channel routing.

Routing helps to predict changing magnitude, speed and shape of a flood wave with time (flow
hydrograph) at one / more points along a water course. During routing, modifications do occur
to the inflow hydrograph for example:

a) The peak of the outflow hydrograph is lower than that of the inflow hydrograph
(attenuation).
b) The time base of the hydrograph increases due to combined effects of storage and
channel friction.
c) The peak of the outflow hydrograph occurs sometime after that of the inflow hydrograph
(lag of peak) due to the travel time of the flood wave in the reservoir / channel section.

The flood is therefore said to be moderated as it passes through a reservoir / channel section.
Outflow from a reservoir is directly related to the head over the spillway (for ungated spillways).
Flood routing impacts peak discharge, time of peak discharge, depth and extent of flooding and
environmental factors e.g. stream bank erosion, floodplain scour, sediment transport, and
deposition.

Flood routing is classified as either lumped or distributed. In lumped (hydrologic) routing, flow
is computed as a function of time at one location along a water course and employs the
continuity equation. In distributed (hydraulic), routing flow is computed as a function of time
simultaneously at several cross sections along a water course and employs both continuity and
momentum equations. Once a flood hydrograph has been generated at a site, interest is then to
predict what happens to the flood as it moves downstream with respect to:

1) Size of the flood peak (Qp)

2) When the flood peak reaches a designated point (tp)
1
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

To address these issues the flood is routed from an upstream point to some downstream point.
Flood routing is used to:

i. Establish the height of a flood peak at a downstream location thus estimating the
protection that would result from construction of a reservoir.
ii. Determine the required height of embankment for flood protection works.
iii. Determine adequacy of reservoir spillways.
iv. Predict the behaviour of a river after change in channel conditions.
v. Schedule H.E power operations according to the predicted progress of a flood wave.
vi. Route hypothetical floods through river systems in order to determine the effects of
proposed flood-control projects.

HYDROLOGIC RESERVOIR ROUTING

When a flood wave enters a reservoir, water level in the reservoir rises. Excess water from the
reservoir is then released through spillways and under sluices. In reservoirs, the head over the
spillway (elevation / stage) is related to its storage and discharge (outflow). Therefore, for any
elevation of the reservoir, storage S and outflow Q can be read from a plot of elevation-storage
and elevation-discharge relationships.

Reservoir routing is based on the following assumptions:

(1) Reservoir water surface is horizontal throughout its length.

(2) Outflow (discharge) is a unique function of storage volume.
(3) Outflow rate varies linearly with time during each time period (∆t). Because of this
Assumption, reservoir routing is known as level pool routing.

The basic equation in reservoir routing is the continuity equation. In reservoirs, inflow I (t)
comes from the river which is known. Outflow Q (t) comes from the reservoir and is controlled
though spillway gates (also known). Both storage S (t) and outflow Q (t) vary with time which
causes variation in elevation H (t). When Inflow (I) is known, determination of storage S (t),
outflow Q (t) and elevation H (t) is known as routing. During reservoir routing, the flood peak is
attenuated and time base broadened due to storage effects. The peak of outflow hydrograph is
also lagged and its time base broadened (figure below):

2
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Data required for reservoir routing includes:

 Storage volume (S) versus elevation (E) for the reservoir
 Water surface elevation versus outflow and hence storage versus outflow discharge
 Inflow hydrograph I = I (t) taken at the upstream river section.
 Initial values of S, I and Q at t = 0.
 Routing increment, ∆t

Flood routing problems are solved by applying the continuity equation I = Q + ∆S

Where I = inflow
Q = outflow
∆S = change in storage
In differential form the continuity equation for the reservoir is given by:
dS
 I  Q Where I is inflow and Q is outflow
dt
A numerical form of this mass balance equation is given as:
 I1  I 2   Q  Q2 
 t   1 t  S 2  S1
 2   2 
Several level pool routing techniques have been proposed, many of them graphical / semi-
graphical such as Goodrich’s and Pul’s methods with the latter being more common.

Modified Pul’s method

This method when applied to reservoirs consists of a repetitive solution of the continuity
equation. It assumes that the reservoir water surface remains horizontal hence outflow is a
unique function of reservoir storage. It uses graphs and tables to solve the storage equation:

It  Qt  S (1)

Considering values at the beginning / end of time interval through suffixes 1 and 2, we have:

 I1  I 2   Q  Q2 
 t   1 t  S 2  S1 (2)
 2   2 

The time interval ∆t should be sufficiently short so that inflow and outflow hydrographs can be
assumed as straight lines during the time interval. ∆t must also be shorter than the transit time of
the flood wave through the reach.

Assuming that the inflow hydrograph is known for all t and that the initial outflow Q1 and initial
storage S1 , are known at time t1 , then equation (2) contains two unknowns: Q2 and S 2

Equation (2) is therefore re-arranged as follows:

 I1  I 2   Q1 t   Q 2 t 
 2  t +  S1  2  =  S 2  2  (3)
     

In equation 3, all terms on the LHS are known at the start of  t hence the value of the function
 Q 2 t  on the RHS can be calculated at the end of  using equation (3).
S 2  2  t
 

3
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Since the storage-elevation S = S (h) and discharge-elevation Q = Q (h) are known, S  Qt 
 
 2 2
enables determination of reservoir elevation and hence discharge at the end of  t . This
procedure is repeated until the entire inflow hydrograph is routed.

For practical use in hand computation, the following semi-graphical method is convenient.

i. Select a time step  t for routing, such that the peak of the hydrograph is not missed.
ii. From the known storage-elevation and discharge-elevation data, draw a curve of
 Qt  versus elevation (figure 1). Here  is the chosen time interval, approximately
S  2  t
 
20 to 40% of the time of rise of the inflow hydrograph.

iii. On the same graph draw a curve of outflow discharge versus elevation (figure 1).

iv. Storage, elevation and outflow discharge at the start of routing (in yellow) are known.

Figure 1: Modified PUL’S Method of storage routing

I  I   Q1 t 
v. For the first-time interval  t ,  1 2  Δt and  S1  2  are known and using equation
 2   
(3) the term  S 2  Q2 t  can be determined.
 2 

vi. Water-surface elevation corresponding to  S 2  Q2 t  is found from the plot of step ii

 2 
above. Outflow discharge Q2 at the end of the time step  t is found from the plot of
step iii.
vii. Deducting Q2 Δt from S 2  Q2 t  gives S  Qt  for the beginning of the next time
   
 2   2 1
step.
4
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

viii. The above procedure is repeated until the entire inflow hydrograph is routed and
ordinates of the outflow hydrograph obtained.

Example1:
A reservoir has the following elevation, storage and discharge relationships.
Table 1:
Elevation (m) Storage (x 106 m3) Outflow discharge (m3/s)
100.00 3.350 0
100.50 3.472 10
101.00 3.880 26
101.50 4.383 46
102.00 4.882 72
102.50 5.370 100
102.75 5.527 116
103.00 5.856 130

When the water level was at 100.50m, the following flood hydrograph entered the reservoir
Time (hrs.) 0 6 12 18 24 30 36 42 48 54 60 66 72
3
Discharge (m /s) 10 20 55 80 73 58 46 36 55 20 15 13 11

Route the above flood and determine:

i. The outflow hydrograph
ii. Reservoir elevation versus time curve during passage of the flood wave.

Solution:
A time interval of 6 hours is selected and from the available data, the elevation-discharge and
 Qt 
S  2 
table is prepared. Δt = 6 hrs. = 60 x 60 x 6 = 0.0216 x 106s. Q is the given outflow
 
discharges while S is as given from the storage / discharge relationships (table 1).
Table 2.
Elevation (m) 100.00 100.50 101.00 101.50 102.00 102.50 102.75 103.00
Storage (S) (Mm3) 3.350 3.472 3.880 4.383 4.882 5.370 5.527 5.856
Discharge (Q) (m3/s) 0 10 26 46 72 100 116 130
Qt 0 0.108 0.2808 0.4968 0.7776 1.0800 1.2528 1.404
2
Qt 3.35 3.58 4.16 4.88 5.66 6.45 6.78 7.26
S
2
A graph of Q versus elevation and  S  Qt  versus elevation is prepared from table 2 (fig. 1).
 2 
Qt
At the start of routing, reservoir elevation = 100.50m, Q = 10.00 m3/s and  S  
 = 3.364
 2 
Qt 
Mm3. Starting from this value of  S  Qt  equation 3 is used to obtain  S  at the end of
 2   2 
t
the first time step of 6 hours as  S  Qt  = I1  I 2  * + S  Qt  = (10 + 20) *
0.0216
+
   
 2 2 2  2 1 2
(3.364) = 0.324 + 3.364 = 3.688 Mm3. From figure 1, the water-surface elevation corresponding
to this value of  S  Qt  = 3.688 Mm3 is 100.62m (lower red circle) and the corresponding
 2 

outflow discharge is 13 m3/s (top red circle).

5
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

For the next time step, initial value of  S  Qt  =  S  Qt  of the previous time step less
 2   2 
 
Qt = 3.688  (13x0.0216) = 3.407 Mm3.
Qt
Starting from this value of  S  Qt  equation 3 is used to obtain  S  
at the end of 12
 2   2 
0.0216
hours as S  Qt  = I1  I 2  * t + S  Qt  = (20+55)* + 3.407 = 4.217 Mm3.
 2  2  2  1 2
 2 
Water surface elevation corresponding to this value is 101.04m and outflow is 27m3/s (fig.1).
For the next time step, initial value of  S  Qt  =  S  Qt  of the previous time step less Qt
 2   2 
 
= 4.217  (27x0.0216) = 3.634 Mm3. Starting from this value of  S  Qt  equation 3 is used to
 2 
Qt 
obtain  S   at the end of 18 hours as
 Qt  = I1  I 2  * t + S  Qt  = (55+80)*
 2  S  2  

 2 1
 2 2
0.0216
+ 3.634 = 5.092 Mm3. Water surface elevation corresponding to this value is 101.64m
2
and outflow is 53 m3/s (fig.1). For the next time step, initial value of  S  Qt  =  S  Qt  of
 2   2 
 
the previous time step less Qt = 5.092  (53x0.0216) = 3.947 Mm . Starting from this value of
3

Qt  equation 3 is used to obtain  Qt 

 S  2  at the end of 24 hours as  S  2  = I1  I 2  *
  Qt 
S  2   
   2
t +  Qt  = (80+73 )* 0.0216 + 3.947 = 5.599 Mm3. Water surface elevation corresponding
 S
2  2  1 2
to this value is 101.96m and outflow is 69 m3/s (fig.1). For the next time step, initial value of
 Qt  =  Qt  of the previous time step less Qt = 5.599  (69x0.0216) = 4.109 Mm3.
 S S
 2  
 2 
Qt
Starting from this value of  S  Qt  equation 3 is used to obtain  S  
at the end of 30
 2   2 
0.0216
hours as S  Qt  = I1  I 2  * t + S  Qt  = (73+58 )* + 4.109 = 5.524 Mm3. Water
    2
 2 2 2  2 1
surface elevation corresponding to this value is 101.91m and outflow is 66 m3/s (fig.1).

For the next time step, initial value of  S  Qt  =  S  Qt  of the previous time step less Qt =
   
 2   2 

5.524  (66x0.0216) = 4.098 Mm3. Starting from this value of  Qt  equation 3 is used to
S  2 
 
Qt 
obtain  S   at the end of 36 hours as
 Qt  = I1  I 2  * t + S  Qt  = (58+46)*
 2  S  2  

 2 1
 2 2
0.0216
+ 4.098 = 5.22 Mm3. Water surface elevation corresponding to this value is 101.72m
2
and outflow is 57 m3/s (fig.1). For the next time step, initial value of  S  Qt  =  S  Qt  of the
 2   2 
 
previous time step less Qt = 5.22  (57x0.0216) = 3.988 Mm . Starting from this value of
3

Qt  equation 3 is used to obtain  Qt 

 S  2  at the end of 42 hours as  S  2  = I1  I 2  *
  Qt 
S  2   
   2

6
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

t +  Qt  = (46+36)* 0.0216 + 3.988 = 4.874 Mm3. Water surface elevation corresponding
 S
2  2  1 2
to this value is 101.48m and outflow is 48 m3/s. (fig.1). This procedure is repeated for the entire
duration of the inflow hydrograph as summarised in table 3.

Table 3: Flood routing using Modified PUL’S method

Qt 
 (Mm3)  S  Qt  (Mm3)
Time Inflow Ῑ ῙΔt Elevation Q
S
(hrs.) (m3/s) (m3/s) (Mm3)  2 
 2  (m) (m 3
/s)
1 2 3 4 5 6 7 8
0 10 100.50 10
15.00 0.324 3.364 3.688
6 20 100.62 13
37.50 0.810 3.407 4.217
12 55 101.04 27
67.50 1.458 3.634 5.092
18 80 101.64 53
76.50 1.652 3.945 5.597
24 73 101.96 69
65.50 1.415 4.107 5.522
30 58 101.91 66
52.00 1.123 4.096 5.219
36 46 101.72 57
41.00 0.886 3.988 4.874
42 36 101.48 48
31.75 0.686 3.902 4.588
48 27.5 101.30 37
23.75 0.513 3.789 4.302
54 20 100.10 25
17.50 0.378 3.676 4.054
60 15 100.93 23
14.00 0.302 3.557 3.859
66 13 100.77 18
12.00 0.259 3.470 3.729
72 11 100.65 14
3.427
The data highlighted in yellow is obtained from figure 1 and is the one to be plotted.

Using data in columns 1 (time), 8 (outflow discharge Q) and 7(elevation) the outflow
hydrograph (figure 2) and a graph showing the variation of reservoir elevation with time (figure
3) are drawn. Sometimes a graph of  S  Qt  versus elevation prepared from known data is
 2 

plotted in figure 1 to aid in calculating items in column 5. The above calculations are
sequential in nature and an error at any stage is carried forward thus affecting the entire
results. Accuracy of the method depends on the value of t selected with smaller t values
giving better results.

7
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Figure 2: Variation of inflow and outflow discharges.

Figure 3: Variation of reservoir elevation with time

GOODRICH METHOD
This reservoir routing method uses equation (3) page 3 re-arranged as per equation 1:
2S 2S
I1  I 2  Q1  Q2 = 2  1 (1)
t t
Suffixes 1 and 2 represent values at the beginning and end of a time step respectively.

Putting known and initial values together,

I1  I 2  +  2S1  Q1  =  2S 2  Q2  (2)

 t   t 

8
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

 2S 
For a given time step, the LHS of equation 2 is known and the term   Q  is determined
 t 2
 2S 
from equation (2). From the known storage-elevation-discharge data, the function   Q  is
 t 2
known as a function of elevation, hence the discharge, elevation and storage at the end of the
 
time step are obtained. For the next time step  2S  Q   2Q2  of the previous time step =
 t 2 
 2S 
  Q  for use as the initial values. The procedure is best illustrated using example 2 below:
 t 1

Example 2:
Route the flood hydrograph below through the reservoir of example 1 using Goodrich method.

Time (hrs) 0 6 12 18 24 30 36 42 48 54 60 66
3
Inflow (m /s) 10 30 85 140 125 96 75 60 46 35 25 20
Initial conditions are: at t = 0, the reservoir elevation is 100.60m

Solution:
A time increment t of 6 hrs = 0.0216 x 106s is selected. Using the given storage-elevation-
discharge data, table 1 below is prepared. A graph showing Q versus elevation and  2 S  Q 
 t 
versus elevation is prepared from this data (figure 4).
Table 1:
Elevation (m) 100.00 100.50 101.00 101.50 102.00 102.50 102.75 103.00
Outflow (m3/s) 0 10 26 46 72 100 116 130
Storage (m3/s) 3.350 3.472 3.880 4.383 4.882 5.370 5.527 5.856
 2S  m3/s 310.2 331.5 385.3 451.8 524.0 597.2 627.8 672.2
 t  Q 
 

Given initial conditions, when t = 0, elevation = 100.60m. From figure 4 when elevation =
 2S   2S 
100.60m, Q = 12m3/s and   Q  = 340 m3/s. Since Q = 12m3/s,   = 340-12 = 328 and
 t   t 
 2S 
hence   Q  = 328-12 = 316 m3/s. For the first time interval of 6h, I1 = 10, I2 = 30, Q1 = 12
 t 1
 2S 
and   Q  = (10+30) + 316 = 356 m3/s. From figure 4, reservoir elevation corresponding to
 t 2
this value of  2S  Q  i.e. 356m3/s is 100.74 m and the corresponding outflow discharge is
 t 2
17m3/s. For the next time increment  2S  Q = 356 – (2 x 17) = 322 m3/s……………………
 t 
  1

The procedure is repeated in a tabular form (table 2) until the entire flood is routed. Using data
in columns 1, 7 and 6, the outflow hydrograph and a graph showing the variation of reservoir
elevation with time (figure 5) are plotted. Like in the PUL’S method above, accuracy depends
on the chosen value of t with smaller values of t giving better accuracy is more involving.

9
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Figure 4: Goodrich Method of reservoir routing

Table 2:
Time (h) I (m3/s) I1  I 2  2S  3  2S  3 Elevation Discharge
 t  Q  m /s  t  Q  m /s (m) Q (m3/s)
   
1 2 3 4 5 6 7
0 10 (340) 100.60 12
40 316 356
6 30 100.74 17
115 322 437
12 85 101.38 40
225 357 582
18 140 102.50 95
265 392 657
24 125 102.92 127
221 403 624
30 96 102.70 112
171 400 571
36 75 102.32 90
135 391 526
42 60 102.02 73
106 380 486
48 46 101.74 57
81 372 453
54 35 101.51 46
60 361 421
60 25 101.28 37
45 347 392
66 20 101.02 27
335

10
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Figure 5: Result of reservoir routing for example 2

CHANNEL ROUTING
The length of stream channel between upstream section where a hydrograph is known and
downstream section where the hydrograph is to be determined is known as channel reach.
Hydrograph at the upstream end of the reach is the inflow hydrograph while that at the
downstream end is called outflow hydrograph. Lateral contribution to the channel consists of
tributary inflows joining the reach at different points and / or contributions from ground water.

In reservoir routing storage is a unique function of outflow discharge S = f (Q) but in channel
routing storage is function of both outflow and inflow discharges and therefore a different
routing method is applied. Flow in a river during a flood is a gradually varied unsteady flow and
the water surface in the channel is not only parallel to the channel bottom but also varies with
time. In a channel reach with a flood flow, total volume in storage is considered to consist of:

Prism and wedge storage (figure 6 a and b)

Prism storage is the volume of water that would exist if uniform flow occurred at the
downstream depth i.e. volume formed by an imaginary plane parallel to the channel bottom
drawn at the outflow section water surface (figure 6 and b).

Wedge storage is the wedge like volume of water formed between the actual water surface
profile and the top surface of prism storage (figure 6 and b).

11
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Figure 6: Storage in a channel reach

Assuming cross sectional area of flood flow to be directly proportional to discharge at the
section, volume of prism storage is computed as outflow (Q) times the travel time through the
reach (K) i.e. KQ. Wedge storage on the other hand is computed as the difference between
inflow and outflow (I-Q) times a weighting coefficient X and travel time K i.e. KX (I-Q). The
coefficient K corresponds to the travel time of the flood wave through the reach. Parameter X is
a dimensionless constant that expresses a weighting of the relative effects of inflow and outflow
on storage within the reach. Muskingum method defines storage in the reach as a linear
function of weighted inflow and outflow:

Prism storage = KQ
Wedge storage = KX (I - Q)

In the initial stages when I > Q, wedge storage is positive but at later stages when I < Q a
negative wedge is formed and volume of wedge storage is then equal to KX (I-Q) where X is a
dimensionless weighting factor with a range of 0 to 0.5. For most natural streams it lies between
0.1 and 0.3 with a mean of about 0.2. Parameter K depends on the length of reach and
roughness characteristics of the channel and has dimensions of time. Values of K and X for a
reach are determined from a pair of observed inflow and outflow hydrographs.

S is total storage in the reach; Q is rate of outflow from the reach while I is inflow to the reach.

An X value of “0.0” produces maximum attenuation, while “0.5” produces pure translation.

Total storage in the river reach is equal to the sum of the two components (prism and wedge
storages) i.e. S = KQ + KX (I-Q) which when re-arranged gives the storage function of the
Muskingum method:

The Muskingum Equation

Consider a time interval ∆t. Let the storage inflow and outflow at the beginning of ∆t be S1, I1
and Q1 and at the end of ∆t be S2, I2 and Q2, respectively.

Applying total storage equation S = KQ + KX (I-Q) or S = K [XI+ (1-X) Q] for the period ∆t,

12
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

S1  K I1 X  1  X Q1 
S 2  K I 2 X  1  X Q2 
Subtracting S1 from S2 we get

S 2  S1  K X I 2  I1   1  X Q2  Q1 

The continuity equation can be re-written as Inflow (I) - Outflow (Q) = change in storage (∆S)

I1  I 2 Q1  Q2 S 2  S1 I  I2 Q  Q2
  or S 2  S1  1 t  1 t
2 2 t 2 2
Combining these two equations and solving for Q2 we get Q2  C0 I 2  C1 I 1  C 2 Q1 which is
known as the MUSKINGUM ROUTING EQUATION.

C 0 , C1 , and C2 are called routing coefficients and are given by:

0.5t  KX 0.5t  KX K  KX  0.5t
C0  , C1  , C2 
K  KX  0.5t K  KX  0.5t K  KX  0.5t

C 0  C1  C 2 = 1 (unity) which can be used as an arithmetical check on the computed values of

the routing co-efficients. In computing the co-efficients, units of K and t must be the same
(hrs). Given an inflow hydrograph, initial outflow, a chosen time interval ( t ) and routing
parameters K and X, the routing co-efficients can be calculated from the above equations and
the routing hydrograph determined from the equation:

Q2  C0 I 2  C1 I1  C2Q1 [Muskingum routing equation]

The equation can also be written in the general form for the nth time step as:

Qn  C0 I n  C1I n 1  C2Qn 1
The Muskingum routing equation provides a simple linear equation for channel routing and has
been found to give best results for routing interval Δt when K > Δt > 2KX. The weighting factor
X should always be less than 0.5 as for values greater than 0.5, Q2 becomes negative. Although
t is chosen arbitrarily, smaller values give better results.

Determination of the outflow Q2 at the end of any time interval using the above equation
requires the value of Q1 (outflow at the end of the previous time step) which is obtained from an
earlier iteration. To use Muskingum equation to route a given inflow hydrograph through a
reach, values of K and X are required and the procedure of determining them is as follows:

 Knowing K and X select an appropriate value of Δt

 Calculate C0, C1 and C2
 Starting from the initial condition I1, Q1 and known I2 at the end of the first time step Δt
calculate Q2 from the equation Q2 = C0 I2 + C1 I1 + C2 Q1
 The outflow calculated in the above step becomes the known initial outflow for the next
time step. The calculations are repeated for the entire inflow hydrograph.
These calculations are best done in a tabular form as shown in the example below:

13
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Example
Route the flood hydrograph given below through the channel reach and derive the outflow
hydrograph. Take X and K as 0.278 and 12 hours respectively.

Time hrs. 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56
Flowm3/s 42 68 116 164 194 200 192 170 150 128 106 88 74 62 54

Solution:
K= 12 hrs. x = 0.278 and t = 4 hrs.

0.5 x 4  (12x0.278) 2  3.336  1.336

C0 = = =  0.125
12  (12x0.278)  0.5 x 4 12  3.336  2 10.664
0.5 x 4  12x0.278 5.336
C1 = =  0 .5
12  (12x0.278)  0.5 x 4 10.664
12  (12x0.278)  (0.5 x 4) 6.664
C2 = =  0.625
12  (12x0.278)  0.5 x 4 10.664

Arithmetical check C0+C1+C2 = -0.125 + 0.5 + 0.625 = 1.0 OK

Routing equation Q2 = C0I2+C1I1+C2Q1 or Q2 = - 0.125 I2 + 0.5 I1 + 0.625Q1

To start with for the first time interval 0 - 4 hours, the initial outflow Q1 is assumed to be the
same as I1 if not given. Thus we have I1 = 42 m3/s, I2 = 68 m3/s and Q1 = 42 m3/s
Q2 = - 0.125*68 + 0.5*42 + 0.625*42 = 38.75 m3/s

This Q2 becomes Q1 for the next time interval.

For the second time interval 4-8 hours we have
I1 = 68 m3/s, I2 = 116 m3/s and Q1 = 38.75 m3/s
Q2 = - 0.125 *116 + 0.25*68 + 0.625* 38.75 m3/s = 43.75 m3/s
Subsequent ordinates of the outflow hydrograph are obtained successfully following the
procedure in the table below:
 Assumed value

14
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Inflow and Outflow hydrographs

Selection of routing time
The continuity equation assumes that the average rate of flow during the interval t is given by
 I1  I 2 
 2  t implying the hydrograph is a straight line during the time interval. The controlling
 
factor in selecting the routing period is that it must be sufficiently short for this assumption to
be valid. Time of travel within reach may be taken as the time lag between the peak of the
inflow hydrograph and the peak of the corresponding outflow hydrograph. The time of travel
can be obtained from a pair of observed inflow and outflow hydrographs. The routing period
should never exceed the travel time through the reach because if it does then, it would be
possible for a flood peak to pass completely through the reach during a routing period in which
case it will not be captured.

Estimation of K and X values for use with the Muskingum routing equation:

1. Collect flood data of the channel reach from flood records of previous years
2. Assume values of X such that (0 < X < 0.3)
3. Calculate values of the term [XI + (I-X) Q] with the chosen value of X (say X = 0.1, 0.2,
0.3 etc.).
4. Estimate the storage S at different times from known values of inflows I and outflows Q.
5. Plot S versus [XI + (I-X) Q] as shown in figure MM below for different assumed values
of X (say X = 0.3, 0.25, 0.20, 0.15, 0.10 etc.)
6. It is clear from the plot that for values of X equal to 0.3, 0.25, 0.20, 0.15 … forms a loop.
At some particular value of X, the curve rises and traces back almost the same path (see
OB in figure MM). The value of X at which this occurs, is the value of X required.
7. Let us say that in our case this occurs when X = 0.1. Extending OB to point A, the slope
of line OA is the estimated value of K. It is seen that the unit of K is time (hrs. / days)
and is approximately equal to the travel time of the flood wave through the reach.
15
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

Figure MM: Graphical estimation of Muskingum X and K values

Example:
The following inflow and outflow hydrographs were observed in a river reach.
Estimate the values of K and X applicable to this reach for use in the Muskingum equation.

Time (hrs.) 0 6 12 18 24 30 36 42 48 54 60 66
Inflow (m3/s) 5 20 50 50 32 22 15 10 7 5 5 5
Outflow (m3/s) 5 6 12 29 38 35 29 23 17 13 9 7
Solution
Using a time increment of t = 6hours, calculations are performed in a tabular form as shown in
table XX below. The incremental storage S and storage S are calculated in columns 6 and 7
respectively. It is advantageous to use the units [m3/s.h] for storage terms. As a first trial x =
0.35 is selected and the value of [XI + (I-X) Q] evaluated (column 8) and plotted against S as in
figure YY. Since a looped curve is obtained, further trails are formed with X = 0.30 and X =
0.25. From figure YY it is seen that for x = 0.25 the data nearly describes a straight line and
hence X = 0.25 is selected as the appropriate value for the reach. The gradient of the plot
S gives the value of K in hours.
XI  ( I  X )Q
Table XX
Time (I m3/s) Q (m3/s) I-Q Mean S S=∑∆S
(hrs.) (m3/s) (I-Q) 3
m /s. h m3/s. h [XI+I-X)Q] m3/s

X = 0.35 X = 0.30 X = 0.25

1 2 3 4 5 6 7 8 9 10
0 5 5 0 0 5.0 5.0 5.0
7.0 42
6 20 6 14 42 10.9 10.2 9.5
26.0 156
12 50 12 38 198 25.3 23.4 21.5
29.5 177
18 50 29 21 375 36.4 35.3 34.3
7.5 45
16
Dr. J.K. Mwangi Dr. J.K. Mwangi
 Lecture Notes ECE 2410: Hydrology II by Dr. J.K Mwangi

24 32 38 -6 420 35.9 36.2 36.5

-9.5 -57
30 22 35 -13 363 30.5 31.1 31.8
-13.5 -81
36 15 29 -14 282 24.1 24.8 25.5
-13.5 -81
42 10 23 -13 201 18.5 19.1 19.8
-11.5 -69
48 7 17 -10 132 13.5 14.0 14.5
-9 -54
54 5 13 -8 78 10.2 10.6 11.0
-6 -36
60 5 9 -4 42 7.6 7.8 8.0
-3 -18
66 5 7 -2 24 6.3 6.4 6.5

Figure YY: Determination of K and X for a river reach.

NB: The plot of x = 0.25 gives a better straight line relationship compared to those of x = 0.30
and x = 0.35. Therefore, X is selected as X = 0.25 and the value of K estimated from the invert
gradient of the resultant line as K = 400/30 i.e. 13.3 hours.

17
Dr. J.K. Mwangi Dr. J.K. Mwangi