Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue IV, April 2018 | ISSN 2321–2705

Design of Water Distribution Network for a Small


Rural Area Using EPANET
Harshan K G#, Keerthana L Madhu*, Anjali A*
#
Professor, Dept. of Civil Engineering, *Student M TECH (Environmental Engineering), M DIT Ulliyeri, India

Abstract- With the tremendous growth of population, the demand 3) Tanks- Storage tanks used to store water. The input
for water also increases. In order to supply sufficient quantity of properties include bottom elevation, diameter, and
water, a good water distribution network is necessary. This initial and maximum water level. The output data
current work utilizes software named EPANET to design a water computed are hydraulic head.
distribution network for a small area in Vettom Panchayat and it
4) Pipes- Links that carry water from one network to
is validated using Hardy Cross Method.
another. The input parameters include start and end
Keywords-Epanet, Simulation, Network, Elevation, Nodes nodes, length, diameter and roughness coefficient.
Computed output parameters include head loss,
I. INTRODUCTION velocity, flow rate and Darcy-Weisbach friction
factor.
D ay by day water demand increases. If proper water
distribution networks are not laid, it will affect the water
supply of the whole areas served by the system. To design a
5) Pump- Links that impart energy to a fluid thereby
raising its hydraulic head. The primary input data
includes pump curve and start and end nodes. The
water distribution system a thorough study of the nearby
principal output data includes flow and head gain.
available water resources, existing distribution network, water
Pump offers unidirectional flow only.
demand and required discharge etc are necessary. Since
manual design becomes difficult, software assisted analysis B. Preliminary Data Collection
are nowadays used. EPANET is a water simulating software.
Before starting with EPANET, the data regarding area should
A distribution network consists of reservoirs or storage tanks,
be collected. This includes
pumps, pipes, nodes and valves. EPANET determines water
flow in each pipe, water level in storage tank and also helps to 1) Fixing nodal points
calculate the pressure at each node. In this paper a new water 2) Data regarding population, elevation and demand at
distribution network is designed for a small rural area in each node
Vettom Panchayath in Malappuram district using water 3) Length between nodes and
simulating software EPANET. The same network is analysed 4) Rough outline of distribution network
manually using Hardy Cross method.
The nodal points are fixed in such a way that the point covers
maximum population. Elevation is determined using
II. METHODOLOGY
theodolite survey and population from field survey. By
EPANET is a computer program that performs extended Geometrical increase method the population after 3 decades is
period simulation of hydraulic and water quality behaviour calculated as
within pressurized pipes. In a distribution network links are
Present population of the area (P) = 15000
connected to nodes. Links indicates pipes, pumps and valves
whereas a node indicates junctions and reservoirs. Decennial population growth (Ig) = 13.39%
A. Physical Components rg = Ig/100= 0.1339
1) Junctions- are points where pipes join together and it Design period (n) = 3 decades
is the place where water enters and leaves. The basic
input data for junctions include elevation, water Future population (Pn) = P (1+rg) n = 21869
demand and initial water quality. The output data The distribution network is designed for a population of
computed includes hydraulic head, Pressure and 21869. Water demand at each node can be calculated as a
water quality. product of per capita demand and population at that node. The
2) Reservoirs- These are the nodes that represents water length of pipe is determined by chain survey. The
sources like lakes, rivers and ground water aquifers. observations are tabulated below
The initial input properties for a reservoir includes
hydraulic head and initial water quality. It has no
computed output parameters.

www.rsisinternational.org Page 118


International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue IV, April 2018 | ISSN 2321–2705

TABLE I
DEMAND AT EACH NODE Pipe 4 395
Elevation Base Demand Pipe 5 343
Node ID
m LPS
Pipe 6 463
1 12.04 0.65
Pipe 7 50
2 12 0.49
Pipe 8 675
3 12.06 0.48
Pipe 9 510
4 12.76 1.41
Pipe 10 502
5 15 0.58
Pipe 11 454
6 11.42 1.4
Pipe 12 178
7 10.92 0.6
Pipe 13 486
8 10.01 0.57
Pipe 14 239
9 12.06 0.80
Pipe 15 389
10 12.86 1.15
Pipe 16 965
11 12.86 1.04
Pipe 17 400
12 13.36 0.80
Pipe 18 400
13 17.86 0.57
Pipe 19 160
14 12.42 0.52
Pipe 20 590
15 12.02 0.80
Pipe 21 315
16 11.42 1.54
Pipe 22 260
17 12.22 0.66
Pipe 23 299
18 11.91 0.84
Pipe 24 476
19 12.6 0.98
Pipe 25 574
20 13.16 1.00
Pipe 26 314
21 13.16 0.8
Pipe 27 123
22 12.01 0.83
Pipe 28 250
23 12.42 0.79
Pipe 29 50
24 12.02 0.35

25 11.82 0.3

TABLE II

LENGTH BETWEEN NODES

Link ID Length
m

Pipe 1 320

Pipe 2 201

Pipe 3 403
Fig. 1 Rough outline of distribution network

www.rsisinternational.org Page 119


International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue IV, April 2018 | ISSN 2321–2705

C. Adding object properties The output parameters of links and nodes are tabulated below.
The computed output parameters for links include flow,
The properties of pumps, reservoirs, pipes, junctions and velocity, head loss and friction factor.
storage tanks are added. Data needed for reservoirs is their
head and for pumps it is the pump curve. It is plotted between TABLE III
flows (in LPS) versus head in meters. Elevation in meters and LINK OUTPUT PARAMETERS
base demand (LPS) are the mandatory inputs in the case of
junctions. Link Id Length Diameter Roughness Flow
m mm LPS
In case of pipes length, diameter and roughness coefficient are
the input parameters. Diameter is the design parameter and is Pipe 1 320 400 140 10.46
given in millimetres. The diameter which gives required flow
and pressure are chosen by trial and error method. Cast iron Pipe2 201 300 140 9.97
pipe is used as main distribution pipe and the Hazen Williams
Pipe3 403 300 140 9.49
coefficient value is taken as 140. For tanks required input
parameters are elevation, initial level, minimum level, Pipe4 395 102.6 140 1.26
maximum level and diameter. After inputting parameters next
step is to run the project. If any input errors occur then run Pipe5 343 102.6 140 0.68
will not be a success. Pipe6 463 102.6 140 -1.49

Pipe7 50 400 140 -8.27

Pipe8 675 400 140 -8.84

Pipe9 510 300 140 6.82

Pipe10 502 300 140 6.02

Pipe11 454 300 140 4.87

Pipe12 178 300 140 3.83

Pipe13 486 186.4 140 3.03

Pipe14 239 186.4 140 2.46


Fig. 2 Pump curve
Pipe15 389 186.4 140 1.94
D. Hardy Cross method
Pipe16 965 102.6 140 0.59
It is an iterative method for determining the flow through each
Pipe17 400 102.6 140 -0.78
pipe where the inputs and outputs are known. In Hardy Cross
method flow is assumed in such a way that principle of Pipe 18 400 102.6 140 0.54
continuity is satisfied at each junction. For validating
EPANET one loop among the four are considered for Hardy Pipe19 160 102.6 140 -0.12
Cross analysis. Pipe20 590 102.6 140 0.17

Pipe21 315 102.6 140 -1.13

Pipe22 260 250 140 -2.11

Pipe23 299 250 140 -3.11

Pipe24 476 186.4 140 -3.91

Pipe25 574 186.4 140 -4.74

Pipe26 314 250 140 -5.53

Pipe27 123 250 140 -5.88


Fig. 3 Problem definition of Hardy Cross method
Pipe28 250 250 140 -6.18
III. RESULTS AND DISCUSSION
Pipe29 50 400 140 19.95

www.rsisinternational.org Page 120


International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue IV, April 2018 | ISSN 2321–2705

TABLE IV TABLE V
LINK OUTPUT PARAMETERS NODE OUTPUT PARAMETERS
Link Id Velocity Unit Head Friction
m/s loss Factor Node ID Elevation Base Demand Pressure Head
m/km m Demand LPS m
LPS
Pipe 1 0.08 0.02 0.024

Pipe2 0.14 0.08 0.023 Junc 1 12.04 0.65 0.65 37.13 49.17

Pipe3 0.13 0.07 0.023 Junc 2 12 0.49 0.49 37.16 49.16

Pipe4 0.15 0.32 0.027 Junc 3 12.06 0.48 0.48 37.08 49.14

Pipe5 0.08 0.10 0.030 Junc 4 12.76 1.41 1.41 36.36 49.12
Pipe6 0.18 0.43 0.027
Junc 5 15 0.58 0.58 33.99 48.99
Pipe7 0.07 0.01 0.025
Junc 6 11.42 1.4 1.40 37.53 48.95
Pipe8 0.07 0.02 0.024
Junc 7 10.92 0.6 0.60 38.24 49.16
Pipe9 0.10 0.04 0.025
Junc 8 10.01 0.57 0.57 39.15 49.16
Pipe10 0.09 0.31 0.254

Pipe11 0.07 0.02 0.026 Junc 9 12.06 0.80 0.80 37.04 49.10

Pipe12 0.05 0.01 0.027 Junc 12.86 1.15 1.15 36.08 48.94
10
Pipe13 0.11 0.09 0.026
Junc 11 12.86 1.04 1.04 36.07 48.93
Pipe14 0.09 0.06 0.027
Junc 12 13.36 0.80 0.80 35.57 48.93
Pipe15 0.07 0.04 0.028
Junc 13 17.86 0.57 0.57 31.02 48.88
Pipe16 0.07 0.08 0.031

Pipe17 0.09 0.13 0.029 Junc 14 12.42 0.52 0.52 36.45 48.87

Pipe18 0.07 0.07 0.031 Junc 15 12.02 0.80 0.80 36.84 48.86

Pipe19 0.01 0.00 0.039


Junc 16 11.42 1.54 1.54 37.41 48.83
Pipe20 0.02 0.01 0.037
Junc 17 12.22 0.66 0.66 36.61 48.83
Pipe21 0.14 0.26 0.028
Junc 18 11.91 0.84 0.84 36.92 48.83
Pipe22 0.04 0.01 0.029

Pipe23 0.06 0.02 0.027 Junc 19 12.6 0.98 0.98 36.32 48.92

Pipe24 0.14 0.14 0.025 Junc 20 13.16 1.00 1.00 35.76 48.92

Pipe25 0.17 0.20 0.024 Junc 21 13.16 0.8 0.80 35.77 48.93
Pipe26 0.11 0.06 0.025
Junc 22 12.01 0.83 0.83 36.98 48.99
Pipe27 0.12 0.07 0.024
Junc 23 12.42 0.79 0.79 36.69 49.11
Pipe28 0.13 0.08 0.024
Junc 24 12.02 0.35 0.35 37.11 49.13
Pipe29 0.16 0.07 0.022
Junc 25 11.82 0.3 0.30 37.32 49.14

www.rsisinternational.org Page 121


International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue IV, April 2018 | ISSN 2321–2705

E. Comparison of results centrifugal pump from underground water source and then
they are lifted up to the overhead water tanks and from there
The flow obtained by using both methods are tabulated below
with the help of gravity system water is transferred to the
main rising pipe. The distribution layout used here is loop
TABLE VI system which is according to the layout of the area. The above
COMPARISON OF RESULTS
results were checked for accuracy using Hardy-Cross method
for one loop. In manual calculation using Hardy-Cross
method, for a single loop it took about forty iterations. It is
Pipe EPANET Hardy-Cross Method time consuming and the chance of causing error is high. For a
(LPS) (LPS) large area, it includes larger number of loops and hence the
calculation and design part itself may take many years. The
number of professionals needed for the completion of work
1-2 10.46 10.4603
will also be high. Future studies are also recommended for
more areas for validation of results. The EPANET software is
2-4 9.97 9.9703 a simple tool for the design of water distribution network.

REFERENCES
4-6 1.94 1.8
[1]. Arunkumar M. and Nethaji Mariappan V.E. (2011): ‘Water
Demand Analysis of Municipal Water Supply Using Epanet
6-7 -1.49 -1.4897 Software’, International Journal on Applied Bioengineering, Vol.
5, No.1, pp. 8-19.
[2]. Asmelash Zewdu (2014): ‘Assessing Water Supply Coverage and
7-8 -8.27 -8.2697 Water Losses from Distribution System for Planning Water Loss
Reduction Strategies (Case Study on Axum town, North
Ethiopia)’, Civil and Environmental Research Journal, Vol. 6, No.
8-1 -8.84 -8.8397 8, pp. 82-87.
[3]. Cross, H. (1936): ‘Analysis of flow in networks of conduits or
Conductors’. Engineering Experiment Station, University of
Illinois, Bulletin No. 286.
The results obtained from the above calculations shows that, [4]. Epp R., and A.G. Fowler,(1970): ‘Efficient code for steady-state
both the values obtained using EPANET and Hardy Cross flows in networks’, J. Hydraulics Di v., Proc. Amer. Soc Civil
methods are almost same, that is validity of EPANET has Engineers, 96 (HYl), 43-56.
been checked. For Hardy Cross method it took about forty [5]. Harry E. Hickey (2008): ‘Water Supply System and Evaluation
Methods’, Water Supply System, Vol. 1, No.1, pp. 1-146.
iterations for one single loop. Small variations obtained may [6]. Jeppson, R. W., and A. L. Davis, (1976): ‘Pressure Reducing
be due to errors in manual calculations. It can be minimized Valves in pipe Network Analysis’, ASCE Journal of the Hydraulic
with more iteration. Division, 102(HY7):987.
[7]. Lewis A. Rossman (2000): ‘EPANET 2 User‘s Manual’, U.S.
Environmental Protection Agency, Edition 1.
IV. CONCLUSION [8]. Martin, D.W. and Peters, G. (1963): ‘The application of Newton‘s
method to network analysis by digital computer’.Journal of the
Water distribution network for Vettom Panchayath has been Institute of Water Engineers, 17, 115-129.
designed with the help of EPANET. The method of [9]. Mc Corale and denial (1960): ‘Pipeline network flow analysis’
distribution used here is combined gravity and pumping J, Mer water works association 41, 422 – 428 1949.
system as firstly the water is pumped with the help of

www.rsisinternational.org Page 122

You might also like