Module - 8 Lecture Notes - 2 Water Distribution Systems

Water Resources Systems Planning and Management: Water Resources Systems Modeling Water
Distribution Systems
MODULE - 8 LECTURE NOTES 2

WATER DISTRIBUTION SYSTEMS
INTRODUCTION
The main purpose of water distribution network is to supply water to the users according to
their demand with adequate pressure. Water distribution systems are composed of three major
components: pumping stations, storage tanks and distribution piping. These systems are
designed according to the loading conditions i.e., pressure and demand at nodal points. The
loading conditions may include fire demands, peak daily demands or critical demands when
the pipes are broken. A reliable design should consider all the loading conditions including
the critical conditions. In this lecture we will discuss the simulation and optimization models
for the design and analysis of water distribution networks.
COMPONENTS OF WATER DISTRIBUTION SYSTEMS

Various components of water distribution systems are:
(i)
Pipes: These are the principal elements in the system. The flow or velocity is usually
described using Hazen Williams equation
1.318CHW R0.63S 0.54

f
(1)
where V is the average flow velocity. CHW is the Hazen Williams roughness
coefficient, R is the hydraulic radius and Sf is the slope.
In terms of headloss hL, the above equation can be expressed as,
hL
KLQ1.852
1.852 4.87
CHW
D
K p Q1.852
(2)
where L is the length of the pipe, D is the diameter and Q is the flow rate.
Headloss can also be determined using Darcy Weisbach equation as
hL
L V2
D 2g
8 fL
Q2
2
gD 2
K pQ 2
(3)
where f is the friction factor (determined from Moodys diagram) and g is the
acceleration due to gravity.
D Nagesh Kumar, IISc, Bangalore
M8L2
(ii)
Node: Junction nodes are connections of pipes to transfer the water. The diameter of
pipe is changed at these nodes. Fixed grade nodes is where pressure and elevation are
fixed i.e., reservoirs, tanks etc.
(iii) Valves: These are used to vary the head loss or to control the flow.
(iv) Tanks: It stores water and acts as a buffer by storing water at low demands and
releasing at high demands.
(v)
Pumps: Used to increase the energy
SIMULATION OF NETWORK
The flow distribution through a network should satisfy the conservation of mass and
conservation of energy. Consider the network structure in Figure 1 with 6 pipes and 5 nodes.
750
200
1
Loop 1
Loop 2
350
200
Fig. 1
Conservation of mass: Flow at each junction nodes must be conserved
Q in
out
Qext
(4)
where Qin and Qout are the flows in and out of the node respectively and Qext is the external
supply or demand.
Conservation of energy: For each loop, energy must be conserved i.e., sum of head losses
should be zero.
hL i . j
H pump
(5)
M8L2
where hLi,j is the head loss in the pipe connecting nodes i and j and Hpump is the energy added
by the pump (if any). hLi,j can be determined using either eqn. 2 or 3.
Energy must be conserved between the fixed grade nodes which are points of known head
(elevation plus pressure head).
EF
hLi , j
H pump
(6)
If the number of pipes in the network is NL, number of junction nodes is NJ and number of
fixed grade nodes is NF, then total number of equations will be NL + NJ + (NF -1).
The set of equations obtained can be solved by any iterative techniques like Hardy-Cross
method, linear theory method and Newton - Raphson method.
Hardy-Cross method
In this method, the loop equation (eqn. 5) in terms of flow is used. The loop equations are
transformed into so called Q equations in the form
K p ,i , j Qi , j
Qi , j
(7)
i,j
Here head loss is determined from equation 2 or 3.

Eqn. 7 is rewritten to account the direction of flow as
hL ,i , j
K p ,i , j Qi , j
Qi , j sign Qi , j
Qi , j
(8)
i,j
This equation can be finally expressed as
K p ,i , j Qi , j
i, j
Qp
nK p ,i , j Qi , j
n 1
(9)
i, j
First a flow distribution is assumed across the network. Then the correction Q as given in
eqn. 9 is applied in a particular loop p. The numerator in eqn. 9 is the algebraic sum of
headloss in loop p taking care of the sign of the flow. If clockwise flows are taken positive,
then the corresponding headlosses are positive. The same is applicable while applying
M8L2
correction also i.e., Qp is added to flows in the clockwise direction and subtracted from
flows in counterclockwise direction.
Example:
Consider the pipe network shown below. The friction factor = 0.2. Determine the flow rate in
each pipe.
L = 1500 m; D = 0.6 m
L = 1200 m;
D = 0.7 m
L = 1200 m;
D = 0.7 m
L = 1500 m; D = 0.8 m
1 m3/s
E L = 1500 m; D = 0.7 m
L = 1500 m; D = 0.7 m
0.5 m3/s
L = 1200 m;
D = 0.7 m
2.5 m3/s
1 m3/s
Solution:
Step 1: Determine the K values in eqn. 3
hL
8 fL
Q2
2
2
gD
KAB = 3.88;
KQ 2 where K
8 fL
2
gD 2
KBC = 6.89; KFE = KED = 5.06; KAF = KBE = KCD = 4.05
Step 2; Assume initial flows in each pipe as shown below
Loop 1
1 m3/s
1 m3/s
Loop 2
0.5 m3/s
0.5 m3/s
1 m3/s
1.5 m3/s
0.5 m3/s
2.5 m3/s
0.5 m /s
1 m3/s
D
1 m3/s
M8L2
Step 3: Consider loop 1 and calculate Q according to eqn. 9; n = 2. Consider anticlockwise

flows positive.
Q1
2
1 K AF Q AF
2 K AF Q AF
2
K FE QFE
K FE QFE
2
K BE QBE
K BE QBE
2
K AB Q AB
K AB Q AB
2
2
2
2
1 4.05 1 5.06 1 4.05 0.5 3.88 1.5
2 4.05 1 5.06 1 4.05 0.5 3.88 1.5
0.0187
Step 4: Consider loop 2 and calculate Q
Q2
2
2
2
2
1 4.05 0.5 5.06 0.5 4.05 0.5 6.89 1
2 4.05 0.5 5.06 0.5 4.05 0.5 6.89 1
0.2088
Step 5: Flows for next iteration

QAF = 1 + 0.0187 = 1.0187
QFE = 1 + 0.0187 = 1.0187
QBE = 0.5 0.0187 + 0.2088 = 0.6901
QAB = 1.5 - 0.0187 = 1.4813
QED = 0.5 + 0.2088 = 0.7088
QCD = 0.5 0.2088 = 0.2912
QBC = 1 0.2088 = 0.7912
Repeat steps 2 to 5 with new flows till Q is insignificant.
Linear theory method
Linear theory is more efficient when compared to Hardy-Cross method. Here we will
demonstrate through the example network given below.
M8L2
750
200
1
Loop 1
3
5
Loop 2
2
3
350
200
The equations for linear theory are:

Conservation of mass:
Node 1: Q1 + Q6 = 750
Node 2: Q1 - Q2 = 200
Node 3: Q2 + Q3 Q4 = 350
Node 4: Q4 + Q5 = 200
Node 5: Q6 - Q3 Q5 = 0
Among these 5 eqns. only 4 need to be used to avoid redundancy
Conservation of energy:
Loop 1: K12 Q122 + K23 Q232 K35 Q352 K51 Q512 = 0
Loop 2: K53 Q532 + K34 Q342 - K45 Q452 = 0
Linearising the above eqns. using k = Kp Q, the eqns. can be written as
Loop 1: k12 Q12 + k23 Q23 k35 Q35 k51 Q51 = 0
Loop 2: k53 Q53 + k34 Q34 - k45 Q45 = 0
M8L2
These 2 eqns. along with the 4 mass conservation eqns. can be solved to obtain 5 unknown
discharges.
OPTIMIZATION OF WATER DISTRIBUTION SYSTEMS
Simulation of distribution networks as discussed above helps to determine the hydraulic
parameters such as pressure heads, tank levels etc. These models are unable to determine the
optimal or minimum cost system. In addition to the cost minimization, the typical goals of
water distribution systems problem in designing pipe system can be:
A) Meeting the household demands.
B) Meeting the required water pressure at all nodes of the distribution system.
C) Optimal positioning of valves.
Therefore, designing water distribution system is a multiobjective problem, which is also
characterized by nonlinearity resulting from the simulation model.
Since the main purpose of a water distribution system is to supply according to the demands
with adequate pressure, a typical optimization problem will be to minimize the systems cost
while meeting the demands at required pressures. Hence optimization problem can be stated
as:
Minimize : Total cost (Capital cost + Energy cost for pumping water throughout the system)
Subject to:
(i)
Hydraulic constraints
(ii)
Water demand constraints
(iii) Pressure requirements
M8L2

Module - 8 Lecture Notes - 2 Water Distribution Systems

Uploaded by

Copyright:

Available Formats

Module - 8 Lecture Notes - 2 Water Distribution Systems

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Module - 8 Lecture Notes - 2 Water Distribution Systems

Uploaded by

Copyright:

Available Formats

Water Resources Systems Planning and Management: Water Resources Systems Modeling Water

MODULE - 8 LECTURE NOTES 2

COMPONENTS OF WATER DISTRIBUTION SYSTEMS

1.318CHW R0.63S 0.54

D Nagesh Kumar, IISc, Bangalore

Pumps: Used to increase the energy

Here head loss is determined from equation 2 or 3.

This equation can be finally expressed as

D Nagesh Kumar, IISc, Bangalore

KBC = 6.89; KFE = KED = 5.06; KAF = KBE = KCD = 4.05

Step 2; Assume initial flows in each pipe as shown below

D Nagesh Kumar, IISc, Bangalore

Step 3: Consider loop 1 and calculate Q according to eqn. 9; n = 2. Consider anticlockwise

Step 4: Consider loop 2 and calculate Q

Step 5: Flows for next iteration

D Nagesh Kumar, IISc, Bangalore

The equations for linear theory are:

D Nagesh Kumar, IISc, Bangalore

Water demand constraints

(iii) Pressure requirements

D Nagesh Kumar, IISc, Bangalore

You might also like