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

net/publication/325092401

Optimization of Water Supply System Using Software EPANET 2.0

Chapter · January 2019

DOI: 10.1007/978-3-319-90893-9_52

CITATIONS READS

0 458

4 authors, including:

Dejan Romic Goran Orasanin

ViK University of East Sarajevo
3 PUBLICATIONS   0 CITATIONS    19 PUBLICATIONS   6 CITATIONS   

SEE PROFILE SEE PROFILE

Jovana Blagojevic
University of East Sarajevo
5 PUBLICATIONS   1 CITATION   

SEE PROFILE

All content following this page was uploaded by Goran Orasanin on 25 February 2021.

The user has requested enhancement of the downloaded file.

 Optimization of Water Supply System
Using Software EPANET 2.0

Aleksandar Košarac ✉ , Dejan Romić, Goran Orašanin, and Jovana Blagojević

( )

University of East Sarajevo, 71123 East Sarajevo, Bosnia and Herzegovina

kosarac88@hotmail.com, dejan.romic@gmail.com,
goranorasanin@yahoo.com, jovana.pajkic@gmail.com

Abstract. Concepts of planning water supply networks in countries in transition

were based at the beginning and during 20th Century, when most of the existing
objects were built. In that period main aim of water supply system was to provide
sufficient amount of water for the population and economy, so the water objects
were planned based on the input available at that time. Upgrades of water supply
systems were done in different scale, without systematic approach to an integral
overview to a weakness in whole system. New concept of Water Utilities, which
has appeared in countries in transition, changes approach where in old systems
water utilities have to satisfy demand driven consumption towards the system
where utilities start to managing with demand managed consumption. First task
to use new concept is introduction of efficiency in management and maintenance
of water supply system.
In this paper there is presented the process of water supply system optimiza‐
tion by using software package Epanet 2.0 on real example of DMA Pavlovac
which is part of water supply system under the authority of the UC “Vodovodi‐
kanalizacija” East Sarajevo.

Keywords: District metered area · Optimization · Water supply system

1 Introduction

Water utilities have a difficult task to provide fresh and clean water for normal func‐
tioning of cities with increasing urbanization. As The United Nations Commission on
Sustainable Development (CSD) stated, the amount of water per person decrease as
population grows, increasing possibility for water shortages [1]. Therefore, it is neces‐
sary to invest significant efforts to ensure enough drinking water with the aim of meeting
the needs of future generations and preserving natural resources.
The premise in work of water utilities so far was to provide a continuous supply of
water with maintaining the water quality at an acceptable level for all consumers. This
especially applies to the developing countries, where it is often considered that water is
a social category and that there is no market value. In order to satisfy water consumption,
water utilities must gradually change their priorities, rather than the permanent expan‐
sion and the opening of new sources, they have to turn towards inner reserves, to reduce
losses from the network and to reduce irrational use of water of their customers [2].

© Springer International Publishing AG, part of Springer Nature 2019

I. Karabegović (Ed.): NT 2018, LNNS 42, pp. 443–451, 2019.
https://doi.org/10.1007/978-3-319-90893-9_52
444 A. Košarac et al.

New priorities for water utilities, instead of permanent seek for new sources, should
be rehabilitation of old systems, reducing of water losses, increasing economic efficiency
as well as raising awareness of consumers about the value of water, thereby reducing of
water consumption. With the new priorities system of water supply will be satisfactorily
kept under control by reducing losses in water systems or with optimization of water
supply systems.
Through this paper there will be presented the optimization process on real example
of water supply zone Pavlovac which is part of the water supply system under the
authority of the UC “Vodovodikanalizacija” East Sarajevo.

2 Influence of Pressure on Water Losses

In optimizing water system it is needed to particularly consider at what point the loss
reduction is economically justified, which includes monitoring of all data, from the input
data, costs of optimization, output as well selection of optimization methods. Control
of pressure in the network can directly influence the degree of water losses. If assumption
is that the water is leaking from pipes through the cross section A, then the outbound
flow is equal to [2]:
√ ( )
Q = CQ A 2gH m3 ∕s (1)

where:

A – Area of cross-section of orifice (crack) (m2),

CQ – Flow coefficient (depends on the shape of the orifice) (-),
g – Gravitational acceleration (m/s2),
H – Pressure at the orifice location (m).

√ given form it is obviously that with double pressure increasing, the loss increases
In the
for 2 = 1.41 times. This is true if the area of cross-section of the leak is constant.
However, with increasing pressure the area of crack through which water leaks is
increasing too, so resulting relationship between flow of leaking water and pressure is [3]:

Q = CQ × HN1 (2)

Where:

N1 – leak coefficient, depends on the type of materials, hydraulics and flow consump‐
tion.

3 Optimization Process of the Water Supply

Optimization of the water supply system is implemented in several related steps:

– collecting of input for mathematical model of the system,
 Optimization of Water Supply System Using Software EPANET 2.0 445

– development of mathematical models based on the input,

– model calibration,
– simulation of different operating conditions on the mathematical model,
– analysis of simulation results and identifying possible activities in order to improve
the water supply system.
The input for mathematical model of the water supply system can be classified into four
main groups [4]:
– Geographic information consist of topographical maps, cadastral data, aerial photo‐
graphs and other supporting information that are useful for determining the physical
location of the system.
– Objects data should contain all information about pipes, pumps, valves, reservoirs
and other physical elements of water system.
– Operating data are important for the establishment of a boundary condition in math‐
ematical model of the system. They consist of the values of pressures, flow rate, water
level in the reservoirs, adjusted valves.
– Data on needs, or water consumption should be available from a database of
consumers.
The collection of input for model is not a one time job, it’s a permanent process. The
model must be constantly updated in order to attain the correct results of the simulation.
Model calibration is done so that the model be in more realistic condition. The basis
for the calibration are measurement of consumption during some periods of day so that
on the basis of these measurements there can be possible to determine the unevenly of
consumption and performs the calibration.
With simulation of various operating conditions the information is obtained about
the behavior of system which can determine further steps to improve the system.
After the simulation, faults of the system are recorded and solutions are being found
for their mitigation or removal. Also in this process mathematical model plays an essen‐
tial role, because they can model objects or elements in the water supply system, or
setting up the parameters can be made by estimated measures of optimization.
By repeated simulation of models with improved infrastructure and configuration
there can be determined level of enhancements designed to optimize measures and based
on the results the most advantageous solutions are adopted.
In this paper the EPANET software package is applied developed by the U.S. Envi‐
ronmental Protection Agency (EPA-Environmental Protection Agency). The program
does not require special licence (it is free of charge) so it’s affordable for all utility
companies in countries in transition. With application of mathematical model EPANET
it is possible to display real water flow and pressure in the water supply system. More‐
over, it provides the possibility of simulation of different steady states and dynamic
states of water systems. EPANET is easy to use and the results are obtained in the form
of tables and charts.
446 A. Košarac et al.

4 Calibration of the Model of the District Metered Area (DMA)

Pavlovac

Description of the existing situation and view of basic characteristics of the system is
extremely important if we want to define the conditions for good quality implementation
of the optimization process of water supply system.
Selected DMA Pavlovac (Fig. 1) is located in city of East Sarajevo. DMA covers
the settlements Pavlovac and the central part of the municipality East New Sarajevo.
Reservoir with volume of 180 m3 is located at 617 m above sea level. The reservoir
Pavlovac is supplied from the main reservoir of II height zone of East Sarajevo water
supply system.

Fig. 1. DMA Pavlovac

Water network consists of PEHD pipes with a diameter from PE63 to PE110, and
pipes of cast iron with a diameter from DN50 mm to DN150 mm. Total length of the
pipeline is around 11 km. Over 4400 inhabitants are supplied in this DMA with 1450
home water meters and also 165 business entities.
On the basis of collected data the model of hydraulic system is made. It is necessary
to point out that the first model is not calibrated so this model illustrates the theoretical
hydraulic condition of Pavlovac water supply system.
 Optimization of Water Supply System Using Software EPANET 2.0 447

Consumption in the water supply system is not constant but it changes during the
day. The smallest consumption is during night hours, and maximum is in the afternoon.
This uneven consumption in a hydraulic model is displayed using the time pattern. The
basis for the time pattern is data of the company “Vodovodikanalizacija” East Sarajevo
obtained from measurements of flow at the input and output of reservoir Pavlovac i.e.
flow and water consumption monitored during 24 h on June 22nd 2017. The interval
between measurements is one minute.
It is necessary to point out that there is a smaller variation of the intake, and increased
water consumption during the day. Consumption is reduced during the night hours, and
it reaches maximum value in the afternoon, while the input in a reservoir is constant.
Time pattern multipliers are calculated as follows [5]:

Average hour demand

Time pattern multiplier = (3)
Average daily demand

With value of daily and hourly average demand for June 22nd 2017 and with use of time
pattern multipliers we are obtaining time pattern of a 24-h uneven consumption. Figure 2
shows the window for setting the time pattern in EPANET for model of DMA Pavlovac.

Fig. 2. The adjusted time pattern of DMA Pavlovac model

After the calibration of consumption the model of the actual state of the water supply
system is obtained, the next step is the implementation of simulations and analysis of
the existing situation.
448 A. Košarac et al.

5 Simulation and Analysis of Water Supply System

Key parameters for analysis of the existing situation and system optimization, obtained
as a result of the conducted simulation, are the value of pressures in nodes, speed of
flow through pipeline network. Also, an important parameter is the water consumption
in the system and the speed of charging and discharging the reservoir on the basis of
which it is being determined whether the reservoir capacity is optimal.
For better insight into DMA Pavlovac, apart from the calibrated model simulation
(of the existing actual state) there was also conducted model simulation with a fire
protection demand lasting for 2 h.
Pressure analysis – Lower critical points of pressure in water supply systems are
2.5 bars and higher critical point is pressure more than 5 bars. The analysis found out
that there are no nodes with pressure lower than 2.5 bars, while in the network there are
43 nodes with the pressure higher than 5 bars.
Velocity analysis – Recommended flow velocities in water pipes are within the
limits 0.2–2 m/s. Higher and lower velocity of these limits are not recommended, and
optimal velocity is around 0.8 m/s. After the model tests it’s found out that the flow
velocities in the distribution grid pipes are in range of 0.2–0.7 m/s. There are no sections
with flow velocity above 2 m/s in system but there is the total of 112 sections with flow
velocity less than 0.2 m/s. These data suggest that the pipeline network from hydraulic
point of view is oversized and the velocity of water flow can lead to decreasing of water
quality, however there should be kept in mind that these oversized pipes are to a certain
extent in correlation with fire protection regulations that require certain pipe diameter
for hydrant connections.
Analysis of water consumption and capacity of the reservoir – The average daily
consumption is around 10 l/s which corresponds to the data from the company, and inlet
is approximately constant with around 11.5 l/s. Based on the preceding considerations
it is concluded that current inlet flow corresponds to the average consumption, and the
reservoir can meet current needs, given that in the night hours when consumption is the
lowest there is full water charging and spillover from the reservoir.
In order to determine the response of the system to unforeseen demand we include
in the model in two nodes a consumption for the case of fire extinguishing in the amount
of 5 l/s. Consumption is given to the most unfavorable hydraulic nodes in the system in
the time period from 15:00 to 17:00 h when there is greatest consumption in DMA.
After repeated simulations it’s found out that there are no great changes in values of
pressures in pipe network, except in the nodes where there is assigned fire protection
consumption and another node where pressure in the time period is below a critical value.
Also there has been no big change in flow velocity in pipes, and there has been no
occurrence of critically high velocities during testing of fire duration. During the testing
of the fire duration, consumption has risen to around 20 l/s. Due to the increased
consumption the level in the reservoir is decreasing, and a longer period of time is
required to refill the reservoir. Accordingly, it is concluded that the reservoir at higher
demand cannot respond to the needs of consumers.
 Optimization of Water Supply System Using Software EPANET 2.0 449

6 System Optimization

After the execution of simulations and determining the values of key parameters for the
system operation the activities for upgrade of those parameters commence i.e. system
optimization. This paper deals with optimization by decreasing the critical values of
pressure and optimization of reservoir capacity.

6.1 Pressure Optimization in System

The principle of optimization by decreasing the pressure is based on the installation of

pressure reduction valve (PRV) on a specific place in the water supply system. PRV
decreases the variable input pressure to a constant output pressure.
As previously stated the pressure on several sections exceeds the critical value of 5 bars.
For the place of the PRV installation setting at 3 bars one location in the model is
identified. After another simulation the number of nodes with pressure higher than crit‐
ical pressure was reduced from 43 to 6 and average pressure reduced by 2.5 bars.
Figure 3 shows the values of pressures after installation of the PRV.

Fig. 3. Pressure values on places with critical pressure after installation of PRV
450 A. Košarac et al.

6.2 Optimization of Reservoir Volume

As stated above that the present volume of the reservoir cannot adequately respond to
unplanned demand increase, that is why it is necessary to determine the optimal reservoir
volume. Calculation of the optimal reservoir volume is defined as [6]:

V = V1 + V2 + V3 (4)

where,

V – total reservoir volume (m3),

V1 – storage for several hours equalizing (m3),
V2 – storage needed for fire suppression (m3),
V3 – reserve volume (m3).

The volume for several hours equalizing V1 is obtained as the product of the coefficient
for several hours equalizing “a” and the maximum daily consumption Qmax,dn (m3/day).
In this case it’s taken eight hour equalizing so the value of coefficient a = 0.3, maximum
daily consumption Qmax,dn (m3/day) is obtained as the product of the average daily
consumption Qsr,dn (m3/day) and the highest coefficient of uneven daily consumption (Kd):
Average daily consumption for the system that is considered is Qsr,dn = 864 m3/day,
and the highest coefficient of uneven daily consumption is in the period between 19:00
and 20:00 h and it amounts to Kd = 1.44, so maximum daily consumption is
Qmax,dn = 1244 m3/day, therefore V1 amounts to 375 m3.
Storage for fire suppression is calculated as the amount of water of 5 l/s needed for
the fire duration of 2 h and it amounts to V2 = 40 m3, and reserve volume is V3 = 65 m3.
Based on the previously determined volumes the optimal reservoir volume is
V = 480 m3.
After modelling the reservoir with the calculated volume and after conducting new
simulation lasting for 72 h, with fire protection demand of 5 l/s for 2 h per day, the
response of the reservoir level has been obtained on 2 unfavorable hydraulic nodes in
the system as shown on Fig. 4.

Fig. 4. Water level in reservoir during 72 h simulation period

 Optimization of Water Supply System Using Software EPANET 2.0 451

Based on the reservoir response it can be seen that longer time is needed to fully
charge the reservoir, and after its full charging the level fluctuations are relatively small
even during the fire demand time, so it is concluded that the capacity of the reservoir is
optimal and consumers’ needs can be satisfied as well as an unplanned demand which
may occur in DMA.

7 Conclusion

With application of the software package EPANET 2.0 for hydraulic modelling and
simulation of the system operation one obtains data which help to determine critical
points regarding water losses. The critical points in system are determined based on
critical pressure values which can lead to higher leakages or even to pipe break and
damage. Also the sections of pipes with critical velocity can be determined. In the DMA
Pavlovac the zones with high pressures are located and it is determined through the
analysis that reservoir capacity is not optimal. Moreover, with application of EPANET
possible activities for removing those shortages in the system are defined, and that is
installation of pressure reduction valve (PRV) and calculation of optimal reservoir
volume.
The main advantage of EPANET is relatively fast and easy model building and
simulation of the system operation, as well as fast simulation of potential problem solu‐
tions and analysis of their results without financial investments. Precondition for
achieving good results with EPANET optimizations are field measurements which are
input for system modelling and quality and accuracy of model depend on them. All of
the above stated lead to a conclusion that software EPANET 2.0 and similar software
packages for modelling and simulation of hydraulic systems will have significant role
in planning, managing and optimizing of water supply systems.

References

1. Vučijak, B., Ćerić, A., Silajdžić, I., MidžićKurtagić, S.: Water for life, basic of integrated water
management (Vodazaživot, osnoveintegralnogupravljanjavodnimresursima), Sarajevo (2011)
2. Prodanović, D.: Rational use of water as an alternative to new sources
(Racionalnokorišćenjevodekaoalternativanovimizvorištima), predavanje, Univerzitet u
Beogradu (2002). http://hikom.grf.bg.ac.rs
3. Lambert, A.O., Hirner, W.: Losses from Water Supply Systems: Standard Terminology and
Recommended Performance Measures. International Water Association (2000)
4. AWWA: Computer Modeling of Water Distribution Systems Manual of Water Supply
Practices
5. Rossman, L.A.: EPANET 2-Users Manual, EPA/600/R-00/057, Water Supply and Water
Resources Division, National Risk Management Research Laboratory, Office of research and
development, U.S. Environmental Protection Agency, Cincinnati
6. Milojević, M.: Water supply and sewer (Snabdevanjevodomikanalisanjenaselja),
Naučnaknjiga Beograd (1995)

View publication stats