International Journal of Advances in Engineering and Management (IJAEM)

Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Evaluating Hydraulic Performance Of

Water Supply Distribution Network: A
Case of Asella Town, Ethiopia
Shambel Belachewa, Tadele Shiferawb, Dr.Dawud
a
Temamc,and Habtamu Diribad
School of Graduate Studies, Jimma Institute of Technology, School of Civil and Environmental
Engineering,Hydraulic Engineering studies, Jimma University, Oromia, Ethiopia
b
Jimma University, School of Civil and Environmental Engineering, Chair of Hydrology and Hydraulic
Engineering,Oromia, Ethiopia
c
Jimma University, School of Civil and Environmental Engineering, Chair of Hydrology and Hydraulic
Engineering,Oromia, Ethiopia
d
College of Environmental Science and Engineering, State Key Laboratory of Pollution Control and Resource
Reuse Study, Tongji University, Siping Rd 1239, Shanghai, 200092, PR China

---------------------------------------------------------------------------------------------------------------------------------------
Submitted: 15-10-2021 Revised: 28-10-2021 Accepted: 30-10-2021
---------------------------------------------------------------------------------------------------------------------------------------

ABSTRACT and it indicates the critical point showing that

Population growth in the town usually exerts needs a modification. The amount of water which
enormous pressure on existing water supply actually reached the consumers on average from
systems. The continuous and repeated deficiency in 2016 G.C to 2020 G.C is 64.76% of the total
the performance of the water network becomes one annual water production. It is recommended that,
of the most critical issues in the water supply sector the water utility hasadded new water source to
that requires immediate action. Asella town water deliver adequate water and add parallel pipes or
supply system has problems related to water supply increasing its diameter to deliver water with the
coverage, water quantity, velocity, and system required pressure.
pressure. The main objective of this study is to Key Words: Asella Town; Hydraulic Model;
evaluate the hydraulic performance of Asella Water Distribution Network; Water GEMS
town’s existing water supply distribution system Corresponding author: E-mail:
with respect to pressure and velocity using Bentley [email protected] Tel:+251-
Water GEMS v8i software. The average daily per 926201181.
capita water consumption and water supply
coverage of the town in 2020 G.C is 35.31 l/p/d I. INTRODUCTION
and 42.249% respectively.The simulated result for Water is one of the most essential
extended period simulation at peak hour commodities of every living being in the
consumption showed that the performance of world.Globally, the population using piped
distribution system related to pressure 47.08% for drinking water supplies between 2000 and 2017
pressure value (<15m), 32.92% for pressure value year is increased from 3.5 billion to 4.8 billion, this
(15-60m) and 20% for pressure value (>60m) head equates to an average of 85,000 people per day
and the pressure at minimum consumption hour is over a 17-year period [12]. While over the same
10% for pressure value (<15m), 45.85% for period, the population using non-piped drinking
pressure value (15-60m) and 44.15% for pressure water supplies increased from 1.6 billion to 2.2
value (>60m. The velocity of pipe flow at peak billion. In Ethiopia accessibility is equally low in
hour consumption showed that 79.56% ofthe rural (5%) and urban (72%) there is a 67
velocity (<0.6m/s), 14.09% ofthe velocity range percentage point gap between rural and urban areas
(0.6-2m/s) and 6.35% ofthe velocity (>2m/s). From [12].
the total 650 nodes in the model, 306 nodes receive According to second Growth and
water with less than 15m pressure head of water Transformation Plan (GTP-2) the goal were set to

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1418
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

provide rural water supply access with minimum pressure zone. The town has a geodetic difference
service level of 25 l/c/day within a distance of 1 km of about 550m, from south to north; necessitating
from the water delivery point for 85% of the rural for creation of pressure zoning of the distribution
population of which 20% are provided with rural network and hence resulted in unbalanced supply
piped systems and to provide urban water supply from the existing distribution network. Due to the
access minimum service level of 100 l/c/day for low pressure of water in the distribution network,
category-1 towns/cities, 80 l/c/day for category 2 consumers at relatively higher spots and expansion
towns/cities, 60 l/c/d for category-3 towns/cities, areas of the town cannot get water.
50 l/c/day for category-4 towns/cities, up to the Models are used to predict pressures under
premises and 40 l/c/day for category-5 towns/cities specific demand conditions and under a wide
within a distance of 250 m with piped system for variety of scenarios to identify low pressures and to
75% of the population [9]. The total average daily select infrastructure that will improve flow or less
per capita water consumption of the Asella town is pressure deficiency [4]. Hydraulic modeling
35.31 l/c/day which is very low as compared to the simplifies the analysis of water distribution system
value set by MoWIE (2019) for GTP-2which 60 and it helps to predict uncertainties in present and
l/c/d for category-3 town. future demands of existing distribution systems
Water distribution systems can be either [11]. In Assella town damaged water pipe and
looped or branched. Looped systems are generally topography of the area are the major problems
preferable to branching systems because pipe which can cause low water pressure and
breaks can be isolated and repaired with minimal uncertainty of water demand in existing water
impact on users outside the immediate area of the supply distribution system. So, to increase the
looped system. In a branched system, on the other sustainability, evaluation of hydraulic performance
hand, all users downstream from the break will in the distribution system is significant.
have their water supply cut off until the repairs are
completed.[10]. Water distribution systems are II. MATERIAL AND METHODS
required to supply water to domestic, commercial, 2.1. Description of the Study Area
and industrial entities above or at a threshold Asella town is situated in the Arsi zone,
pressure with consumer demands that vary being the zonal capital, at a road distance of 175km
throughout the day, weak, season and year. The from Addis Ababa or 75km from Adama town. It is
minimum pressure that should be observed at accessed through asphalt road running from Addis
junctions throughout the system varies depending Ababa via Adama to Bale Robe. The town is
on the type of water consumption [8]. bounded by geographical coordinates between
According to the report of Asella town UTM 3907’0”E to 3909’0”E longitude and
water supply service enterprise the existing Asella 7054’30”N to 7058’30”N latitudes.
town water supply system is defined as one

Figure 1.Map of the study area (adopted from ArcGIS).

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1419
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

2.2. Materials used capita consumption using the projected total

2.2.1. Equipment population figure during (2020 G.C). The
The Global Positioning System instrument following formula was applied for the
was used to collect the required elevation data determination of per capita consumption
during pressure reading. Pressure readings were (liter/person/day) [5].
done using pressure gauge which is commonly Per capital water consumption =
l
taken in the selected points of the distribution Annual consuption (m ∗3 ∗1000 3
m
system. ……….………
Population ∗365
2.2.2. Software: Water GEMS (Eq.3.1)
The model is something that represents
things in the real world and computer model uses The water supply coverage of the town has been
mathematical equations to explain and predict evaluated based on annual water production and
physical events. Modeling of water distribution annual water demand as follows:-
systems can allow determining system pressure and (annual production ∗100 %)
Water supply coverage =
flowing rate under a variety of different conditions Annual demand
without having to go out and physically monitor ………………...………… (Eq.3.2)
the system [2].
Bentley water GEMS V8i is selected for 2.4.1.2. Water Loss Analysis
this study because of the following reason:- The water loss analysis of Asella town
Graphical user interferences as compared to was evaluated in aggregated form in numerical as
Epanet software, integration with external software, well as a percentage of the non-revenue water,
like Auto CAD and ArcGIS and requires less effort which was obtained from the total water production
and shorter time to build a model than others. and water consumption. Unaccounted-for-water,
2.3. Data Collection expressed as a percentage, is calculated as the
The data collection process was performed amount of water produced by the public water
using both primary and secondary data collection system minus the metered customer use divided by
techniques to get the required information. Water the amount of water produced multiplied by 100 as
production and consumption data used to evaluate follow [7].
water losses and water supply coverage of the Total water loss (%) =
Total water produced −Total water billed ∗100
town. Survey, design data, and the town existing …………
Total water produced
water distribution network layout were used to ……. (Eq.3.3)
construct the model using Water GEMS v8i 2.4.2. Model Representation
software and the pressure for ten sample nodes was The model is constructed using Water
measured by using a pressure gauge to calibrate the GEMS software by giving all the necessary inputs
model. collected from Asella town water supply
2.4. Data Analysis distribution network layout, Pipe data, such as pipe
Primary and secondary data can be diameter (mm), C-value and length (m) are
analyzed both qualitatively and quantitatively. assigned to the network. Input nodes are elevation
Qualitatively, the data’s are analyzed with the help (m), water demand (LPs) and time pattern. Pump
of tables, charts or in words and quantitative data head (m) and flow (LPs) are required data for the
was analyzed with the help of Bentley Water construction of the pump curve. Figure 2 shows the
GEMS v8i software. GPS and Arc map 10.1 is used constructed model of the water supply network
to collect data and to generate maps of the study from source to the WTP and WTP to service
area respectively. Qualitative data were interpreted reservoir.
with the help of Microsoft Excel. The volume of
water consumed for domestic purpose is estimated
by converting the annual consumption data to
average daily per capita consumption using the
projected total population figure during (2020
G.C).
2.4.1. Water Supply Coverage Analysis and
Water Loss Analysis
2.4.1.1. Water Supply Coverage Analysis
The volume of water consumed for
domestic purpose is estimated by converting the
annual consumption data to average daily per

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1420
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Figure 2. Layout of the Existing water system of Asella town.

2.4.3. Hydraulic Modeling of water supply extrapolation over a five-year period makes it
Distribution Network using Bentley Water suitable. It is a hybrid of the geometric and
GEMS V8i arithmetic methods and corrects the anomalies of
Analysis of water distribution network the methods [1].
provides the basis for the design of new systems, Pn =
the extensions, and control of existing systems. The p0 E rn ………………………………………………
flow and pressure distributions across a network ……………………. (Eq.3.4)
are affected by the arrangement and sizes of the Where: Pn=population at n decades or year,
pipes and the distribution of the demand flows. PO=initial population (from census), are=growth
2.4.3.1. Assigning Base Water Demands to Each rate, and =decade or year, e=constant exponential
Node value (2.718).
To assign base demand to each supply
node, it is necessary to estimate base demand of Step Two: Identification of a number of houses
each node in the distribution network by following around each supply node: -In ArcMAP, the
the steps below: orthomap of Asella town was opened and the town
Step One: Population Forecasting: -In order to water supply distribution network constructed in
avoid over or underestimation of the future Water GEMS by using model builder was exported
population 2007 CSA population projection using into the AutoCAD DXF file and imported into
1994 medium variant growth rate set for Oromia ArcGIS was overlapped with it.
region was used. The exponential population Total current population
Average people per house =
forecasting method is used to forecast the current Total number of houses
………………………… (Eq.3.5)
Asella town population.This method is useful for
projections on a short term basis hence

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1421
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Figure 3. Distribution network overlapped on the map of the Asella town.

Step Three: Determination number of people following roughness coefficients are suggested for
per single-family residence each supply node: - existing pipes, depending on age and the material
Currently, the population of the town is about and the remaining pipe sections are adjusted for
97,118 peoples. The total number of houses their C-values.
identified was 20,475, giving an average count of 2.4.3.3. Assigning Demand Patterns
4.74 people per house. To calculate the population The type of simulation used for this
served to each node multiply assigned by that node modeling is the extended period simulation to
by the average number of people in each house as evaluate system performance over time. For such
follows: type of simulation, the demand patterns of the town
Number of people for supply = for each node should be identified and the demand
Number of houses assigned by variation of each pattern has to be clearly set as
that node ∗ average number of people Eq.3.6) well. The major demand patterns of the town are:
in each house Residential, Commercial, Public, and Industrial are
Step Four: Determination of average day water the major ones.
demand of Asella Town: -Average water demand 2.5. Model Calibration
of the town was calculated by multiplying the For model calibration and validation effort
average per capital demand with the estimated data were collected from field selected sample
number of populations as follows locations. This involves making minor adjustments
ADD = Per capital water consumption to the input data, then the model accurately
∗ total population simulated the pressure rate in the system. Pressures
Step Five: Determination base water demand in are measured throughout the water distribution
each supply node: -After the average daily water system using pressure gage instrument to use the
demand of the system was determined, base water data for model calibration.
demand for the particular supplied node were 2.5.1. Pressure Measurement
calculated by using equation 3.7 and finally Pressures are measured throughout the
assigning into the node manually. water distribution system to monitor the level of
Base demand for supply node = service and to collect data for use in model
The population served by node
∗ ADD ………… (Eq.3.7) calibration. In this study the pressure measurements
Total population
were taken at a direct connection to the water main
2.4.3.2. Assigning Roughness Coefficients to nodes and nearer to the supply main nodes at
Pipelines homes faucet as shown in figure 4 below.
Hazen-William roughness factors were
used to incorporate frictional losses and the

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1422
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Figure 4. Pressure measurement at different location.

2.5.2. Network Simulation The average domestic water supply

Extended period simulations are used to coverage of the town is 35.31 l/c/day. This average
evaluate system performance over time. Modeling per capita consumption is very low as compared to
tanks filling and draining, regulating valves the value set by MoWIE (2019) for GTP-2which
opening and closing, and pressures and flow rates 60 l/c/d for category-3 towns/cities within a
changing throughout the system in response to distance of 250 m. The annual water demand for
variable demand circumstances and automatic the year 2020 is 13427.49 m3/day * 365 days which
control strategies developed by the modeler are all is 4901033.85m3 and the annual water production
possible with this type of analysis. of the town in 2020 G.C is 2070653m3. So the
water supply coverage is the ratio of annual water
III. RESULT AND DISCUSSIONS production and annual water demand.
2070653 m3
2.6. Water Supply Coverage Analysis and Water Supply Coverage % = ∗
4901033 .85m3
Water Loss Analysis 100 = 42.249%
2.6.1. Water Supply Coverage Analysis 2.6.2. Water Losses Analysis
2.6.1.1. Average Per Capita Demand Coverage The total water loss for the town was
The water supply coverage of the town is calculated using the total annual water produced
evaluated based on the average per capita and distributed to the distribution system, as well as
consumption. The average water consumption per the water billed from individual customer meter
capita was derived from the town’s annual readings.The water loss is usually expressed in
consumption, which was aggregated from the terms of percentage unaccounted for water (UFW),
individual water meter and the public tap. Thus, the loss per kilometer length of main pipes and loss per
annual water consumption data are converted to properties or number of connections. The total
average daily per capita consumption using the water loss has been evaluated based on the three
population data of the town.Average daily per measurement approaches as explained here below.
capita water consumption of the town in 2020 G.C 2.6.2.1. Total Water Loss Expressed as
was calculated from the total annual recorded Percentage (UFW)
consumption of the town by using equation 3.1. The total annual water produced and
Per capita consumption LCD distributed to the system within the specified year
(1,251,806.00m3 ∗ 1000l/m3) of 2020 G.C is 2,070,653 cubic meters and the
= = 35.31 lcd
97118 ∗ 365 day annual total water loss is 818,847 cubic meters that

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1423
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

accounts for 39.55 % of the total water production. Asella town accounts for only 64.76% of the total
As depicted in figure 5 below the total annual water produced. As it is shownfigure 5 below, non-
water loss of the water supply system is 28.39% in revenue water from the system is varied from year
2016, 31.83% in 2017, 39.28% in 2018, 37.16% in to year due to the aging of pipe that leads to
2019 and 39.55% in 2020 G.C.The average amount leakage, pipe bursting, installation (extension of
of water, which actually reached the consumers in network in new area) and illegal connection.

50.00
39.28 37.16 39.55
40.00 31.83
Percentage of water loss( %)

28.39
30.00

20.00

10.00

0.00
2015 2016 2017 2018 2019 2020 2021

year

Percent of losses %
Figure 5.Annual water losses of the town.

2.6.2.2. Water Loss Expressed as Per Number 2.6.2.4. Possible Reasons of High Water Loss
of Connection Regarding the system efficiency of the
Water loss expressed as a percentage is an existing distribution system, the data obtained from
appropriate means to show the extent of the loss the water supply service enterprise bill data has
within a given environment, but it is not a good been used to estimate the water loss within the
indicator for comparing the losses from one area to system. As depicted in figure 5 above the losses of
another. Taking the total number of connections in water within the system don’t have a uniform trend
the town as 18,200 the water loss per connection of increase or decrease instead it undulate from
for the similar duration was derived as,Water loss = year to year.
818847 x1000÷ (18200×365) = 123.26
liter/connection /day. This figure shows as litters
per service connection per day increase water 2.7. Model Calibration and Validation
losses also increases. 2.7.1. Calibration of Hydraulic Network
2.6.2.3. Water Loss Expressed as Per Length of Model
Pipes Calibration is the process of comparing
Water loss expressed as per kilometer the model results to field observations and, if
length of main pipes is also used as indicator to necessary, adjusting the data describing the system
compare water loss. This indicator is usually until model-predicted performance reasonably
recommended for non- densely populated areas. agrees with measured system performance over a
The total length of pipes of greater or equal to wide range of operating conditions. Ten data sets
50mm diameter have been used to evaluate total were selected from field observation and from
water loss of the entire town is 98.021km. Using simulated results for calibrating the model.
total pipe length of the entire town, the water loss 2.7.1.1. Pressure Calibration
per kilometer length of main pipes was derived to The degree of accuracy varies depending
be 818847 ÷ (98.021 km×365days) = on the size of the system and the amount of field
22.8871m3/km/day. This shows that as the length data and testing available to the modeler. Bentley
of the pipe increases the amount of water losses per (2008), states that the average difference of ±1.5m
day increases. to a maximum of ± 5.0m for a good data set and ±

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1424
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

3.0 to ± 10m for a bad data set would be a reasonable target.

Table 1. Data arrangement for pressure calibration and time series with pressure networks.
Sample Location

Measured Time
Location

pressure error
Pressure (m)
pints

Difference
x (m) y (m) Elevation

Computed
Observed

Scenario
Pressure
S.NO

1 J-199 514382.24 875689.38 2542.42 10 15.65 5.65 6:30

2 J-208 513759.03 879858.54 2362.24 1 0.26 -0.74 7:15
3 J-171 513621.56 875785.56 2521.1 23 26.67 3.67 7:45
4 J-388 514393.67 878118.78 2451.26 16 18.42 2.42 8:30
5 J-75 515792.26 879088.09 2462.73 12 9.01 -2.99 9:00
6 J-233 513925.7 879367.74 2371.79 64 63.12 -0.88 10:00
7 J-107 514094.97 877521.32 2488.24 40 45.08 5.08 10:45
8 J-140 516035.76 878974.14 2485.12 35 37.7 2.7 11:30

Base scenario
9 J-169 515469.79 877807.04 2514.32 8 9.27 1.27 12:00
10 J-582 514749.45 877787.97 2458 9.5 11.01 1.51 1:00
Average Error 1.769

As shown in table 1 above, computed under average level (average +1.5m to the
values are within an average error of 1.769m maximum +5m). The agreement between the
pressure simulated observed values. Hence, the observed field data and the model result graphically
model is acceptable calibrated which is satisfied the sketched to show the overall relationship in
setting pressure calibration and validation criteria between the two data sets as follows.

Variation of actual and simulated pressure at

sample nodes
70
Observed and simulated pressure

60
50
Observed
40 pressure(m)
30 Simulated
(m)

20 Pressure(m)
10
0
J-75
J-171

J-107
J-199
J-208

J-388

J-233

J-140
J-169
J-582

Junction
Figure6.Actual and simulated pressure at samples node.

Pressures were measured in the field in observed pressures versus calculated pressures at
order to compare with the results of the distribution various distribution lines and taps throughout the
system. Figure 7 below is a comparison plot of system.

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1425
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Observed and Simulated Pressure corolation plot(m)

70
y = 1.006x + 1.622
60
R² = 0.980
50

40
pressure value (m)

30
Observed

20

10

0
0 10 20 30 40 50 60 70

Simulated Pressure value (m)

Figure 7.Correlation between observed and simulated pressure parameters.

The diagonal line on the plot represents minimum pressure adopted for this study is 15m of
the line of perfect correlation in figure 7 above. water head. As depicted in Table 2 below about
Ideally, all the points should align themselves on 306 out of 650 Nodes are below the minimum
this line; meaning that all observed pressures are adopted system pressure. This indicates that the
equal to the computed pressures, giving a pressure within the distribution system is 47.08%
correlation coefficient of 1 that is the best of nodes are below the minimum desirable
correlation between observed and simulated. The pressures during peak hour demand and these
linear correlation coefficient (R2) of observed nodes are not capable of supplying the necessary
versus computed pressures is 0.9805. The demand to consumers and 20.00% of nodes are
coefficient of determination (R2) value was exceeded to maximum allowable pressures of 60m
98.05%, it indicates that observed and simulated at normal condition as described in table 5 under
relation is strongly as values tend to 1(the the methodology part. While 32.92% of nodes are
computed pressures are within the acceptable within the permissible pressure ranges of minimum
limit). 15m and maximum 60m pressure head. At this
2.8. Model Analysis peak hour level the water consumption demand
The system conditions have been expected to more over all the hour demands.
computed over twenty-four hours with a specified There are some reasons that are why the
time increment of one hour and starting model run negative pressure is occurring in the water supply
time at 12:00 PM. The software simulates non- distribution system is as a result of the following:
steady-State hydraulic calculation based on mass elevations difference, high demands, pipes of
and energy conservation principle. The model is inadequate capacity (too small diameter), rough
simulated for every one-hour time setup in the pipes (e.g. corroding iron pipes or pipes with a
twenty-four hour duration. However, for the build-up of sediment), and equipment failures (e.g.
analysis the peak and minimum hour demand are Pumps and valves).
simulated to identify the current problems of the The low pressure nodes are normally those
system and to locate the critical points in existing nodes which are locatedrelatively at high elevations
water supply distribution network. and far from the supply points. Low pressure can
2.8.1. Hydraulic Parameter on Existing Water cause reduction of quantities of water supplied to
Supply Distribution Network the consumer and entry of a contaminant or self-
2.8.1.1. Pressure deterioration of water quality within the network
In this study, the model run from the input itself a severe damage to public health.
of existing data a total node of 650 was reported As described in figure 8 below the area highlighted
from the project inventory dialog box. The by green color is indicate lower pressure (negative

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1426
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

pressure) below 15m of water head, the area the area highlighted by red color are indicate
highlighted by aqua color is indicate permissible pressure above maximum allowable operating
pressure range between15m to 60m water head and pressures 60m of water head at normal condition

Figure 8. Pressure contour map of nodes at maximum consumption hours.

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1427
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Table 2. Distribution of pressure at minimum and maximum consumption hours.

Pressure (mH2O) pressure at minimum consumption pressure at maximum consumption
hours hours
Node number Percentage (%) Node number Percentage (%)
<10 51 7.85 290 44.62
10-15 14 2.15 16 2.46
15-20 32 4.92 23 3.54
21-30 51 7.85 36 5.54
31-40 67 10.31 52 8.00
41-50 63 9.69 49 7.54
51-60 85 13.08 54 8.31
61-70 83 12.77 52 8.00
>70 204 31.38 78 12.00
650 100.00 650 100.00

During low flow typically at mid-night only 32.92% nodes are with permissible pressure
distribution system of the case study is marked by due to excessive demand.
excessive pressure. As shown in table 2 above, As described in figure 9 below the area
Figure 9 below and detailed in appendix D, 10% highlighted by green color is indicate lower
and 44.15% of nodes below minimum and exceed pressure (negative pressure) below 15m of water
maximum allowable operating pressures in the head, the area highlighted by aqua color is indicate
distribution network respectively. Minimum permissible pressure range between 15m to 60m of
pressure is also observed during low consumption water head and the area highlighted by red color
period. Only 45.85% of nodes are received the are indicate pressure above maximum allowable
water of optimum pressure at the low consumption operating pressures 60m of water head at normal
hour. As compared to distribution of pressure at condition.
maximum consumption hour table 2 above, shows

Scenario: Base

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1428
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

HYDRAULIC MODELING OF AS Bentley Systems, Inc. Haestad Me Bentley WaterGEMS V8

ELLA WATER thods Solution Center i (SELECTseries 5)
SUPPLY DISTRIBUTION NETW 27 Siemon Company Drive, [08.11.05.61]
ORK -BY Suite 200 W Page 1 of 1
SHAMBEL BELACHEW.Wtg Watertown, CT 06795 USA +1-
12/2/2020 203-755-1666
Figure 9. Pressure contour map of nodes at maximum consumption hours.

According to this study output for Asella distance and elevation in pressure distribution of
town, pressure zones (for elevated area, lower area selected nodes is given figure 10below.
and commercial or institutional area) may be better Figure 10 below shows that, effects of
to see for modification. Because of during intermit distance and elevation in pressure distribution for
supply pressure become above simulated pressure selected junctions. The first (junction 199) and the
head. This also affects the hydraulic performance last (junction 193) have an elevator and pressure
of the network. head of 2542.42m a.s.l with 15.65m pressure and
Households located at higher elevations 2534.93m with 10.39m pressure respectively.
and close to reservoir site have got water at low When the elevation decrease from junction 199 to a
water pressure. Variations of pressure during day lower point, pressure increases to that point and
and night can create operational problems, resulting after the lowest point (at junction 142) elevation
in increased leakage and malfunctioning of water starts to increase and pressure starts to decrease and
appliances. Reducing the pressure fluctuations in continue up to the last junction. At junction 176
the system is therefore required [3]. The effect of elevationsstartwithdrops, but pressure suddenly
increases.

Figure 10. Profiles of pressure vs. Elevation of selected nodes showing distance from junction 199 to the
farthest point.

The high value of pressures affects lower location. Also produce low velocities which
adversely the hydraulic performance of the accelerate the deterioration and corrosion of the
distribution network at night time during low pipes in the distribution system and leakage rate are
consumption period, the pressure in the system expected to be high because at this time no water
becomeshigher and it causes a pipe burst at the flow occurred in the distribution.

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1429
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

less than 2m/sec, in the distribution system and not

2.8.1.2. Velocity more than 2.5 m/s in a transmission system. A
The velocity of water flow in a pipe is one minimum velocity of 0.6 m/sec had taken. At the
of the important parameters in the hydraulic peak hour demand the values are different as
modeling, performance evaluation of the efficiency compared to the minimum consumption hour. The
of water supply distribution and transmission line. water supply system network velocity during peak
Velocity distribution also varies with demand hour demand is summarized in the table 3 below.
pattern changes. Water velocity should maintain at

Table 3.Velocity Distribution in Pipeat peak hour demand.

Velocity range (m/s) Count Count (%) Effect
≤0.6 576 79.56 Sedimentation problem

0.6-2 102 14.09 Normal

≥2 46 6.35 Erosion and high head loss occurred

As depicted in table 3 above, during the Low velocities are undesirable because
peak hour demand situations about 6.35% of the they lead to lower pipe flows, since discharge is a
pipes are failing to satisfy the permissible velocity function of velocity. Also low velocities are
or maximum velocity in distribution and undesirable for reasons of hygiene and
transmission line (>2 m/s), in addition to that, sedimentation problem. In the opposite way, higher
79.56% of the pipes also below the minimum velocity, not more than 2.0 m/s and 2.5 m/s in
velocity in a distribution line (<0.6 m/s). While, distribution system and transmission system
only 14.09% of the pipes are in the permissible respectively to prevent erosion and high head
velocity ranges. Velocity has also a great impact on losses.
water quality as turbidity and the likes.

Scenario: Base

HYDRAULIC MODELING OF ASE Bentley Systems, Inc. Haestad Method Bentley WaterGEM
LLA WATER s Solution S V8i (SELECTseri
SUPPLY DISTRIBUTION NETWO Center es 5)
RK -BY 27 Siemon Company Drive, [08.11.05.61]
SHAMBEL BELACHEW.Wtg Suite 200 W Page 1 of 1
12/8/2020 Watertown, CT 06795 USA +1-203-
755-1666

Figure 11.Velocity Distribution in Pipeat peak hour demand.

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1430
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

In general the study area of water 663, p-664, p-665, p-666, p-667, p-668, p-669, p-
distribution system has some problems with respect 675, p690, p-723, p-724 are very high.
to hydraulic network modeling. These are low Generally, undersized pipes would lead to
pressure, high pressure, high velocity, and low increased head losses due to increased friction.
velocity due to undersized and oversized service However, over sizing pipes beyond reasonable
pipe diameter and inadequate water supply. The limits would increase the contracting cost. As the
Low pressure problem is due to high elevation and length of the network increases and the number of
undersized pipe diameter and high pressures are pipes, valves, fittings and other obstructions in the
usually caused by low elevation and oversized system increase, both major and minor losses
service pipe diameter. increases.

2.8.1.2.1. Possible Reasons of Low 2.8.1.3. Identification of the Location of Critical

Velocity in Pipe System Points
Since discharge is a function of velocity If a pipe is too small, it may become a
and velocity is a function of pipe size, the results of problem only during high flow conditions; the best
discharge and velocity is used for the judgment in time for diagnosing problems is the model
solving the distribution network problems related simulation during peak hour flow. A color coding
to pipe size. Inadequate water supply, oversized is specified for several ranges of pressure heads at
service pipe diameter and topography is the major the junctions and was helped to understand the
problem which causes low velocity of water in the difference in the pressure range at various
pipe system. Topography of Asella town is junctions.
characterized as rugged and inclined. At low There are total 650 nodes in the model,
elevation and oversized service pipe diameter, the out of them 306 nodes receive water with less than
velocity of water is low and the pressure is high. two 15m pressure head of water, which is
2.8.1.2.2. Head Losses in the Pipes inadequate and 52 nodes receive water with greater
The modeled head losses enable to judge than to 60m pressure head of water and those nodes
whether booster stations are needed or not to boost denotes as critical points. It is observed that the
the water pressure and add energy to let the flow pipe capacity is insufficient and oversized diameter
continue. The model simulated shows that head to deliver water with the required pressure. To
loss in p-269, p-270, p-271, p-272, p-273, p-274, p- maintain the pressure head at those nodes, it is
275, p-279, p-280, p-285, p-286, p-288, p-488, p- better to add parallel pipes in the distribution
489, p-498, p-499, p-500, p-501, p-502, p-503, p- network. Use pressure sustaining valves to control
506, p-507, p-508, p-509, p-510, p-511, p-512, p- the occurrences of minimum pressures and pressure
513, p-514, p-515, p-516, p-660, p-661, p-662, p- reducing valve to control occurrences of maximum
pressures for parts of the high elevation network.

Scenario: Base

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1431
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

HYDRAULIC MODELING OF A Bentley Systems, Inc. Haestad Me Bentley WaterGEM

SELLA WATER thods Solution S V8i (SELECTseri
SUPPLY DISTRIBUTION NETW Center es 5)
ORK -BY 27 Siemon Company Drive, [08.11.05.61]
SHAMBEL BELACHEW.Wtg Suite 200 W Page 1 of 1
12/8/2020 Watertown, CT 06795 USA +1-
203-755-1666

Figure 12. Identification of critical points in the distribution system.

In the figure 12 above the green color circumstances risk of obtaining no water is
nodes have a pressure head below 15m of water observed because of the pressure in the distribution
head, red color nodes have a pressure head above system is below the permissible minimum
60m of water head and they denote as the critical requirement. Nodes below minimum and exceed
points in the distribution network while the nodes maximum allowable operating pressures in the
in the blue have a pressure head between 15m-60m distribution network represent the critical points.
of water head. The town water supply distribution system service
2.8.2. To Cope-up the Above Problems coverage was also evaluated using the water
Asella town water supply and sewerage demand and water production having 42.249 %
enterprise must redesign the water distribution coverage for the year 2020G.C.
system at peak hour with maximum day demand.
By examining what is going on the system as a V. ACKNOWLEDGEMENT
result of peak hour, solutions are given to the I would like to give thanks and praise God
problems faced (pressures and velocities out of the for his provision, grace and mercy in all the paths
design limit) within the network. Take a of my life who, favoring me with health,
modification to the problems by creating new knowledge, wisdom and courage to finalize my
alternatives and scenarios. At peak hour demand study and do this research. I am grateful to my
the velocities out of the design range are modified Advisor Dr. Dawud Temam for his efforts, useful
by resizing pipe diameters and pressures at junction suggestions, and encouragements, which provided
of lower portion were high, reduction to the desired valuable guidance for me starting from the proposal
pressure has been made by using pressure reducer up to the whole thesis work.I am very grateful to
valves and pressures at junction of higher portion express my sincere thanks to my co-advisor Mr.
were low, uses pressure sustaining valves to Tadele Shiferaw (MSc)for his-constructive
control the occurrences of minimum pressures. comments, his guidance helped me in all the time
writing of my thesis and guiding me in the right
IV. CONCLUSIONS direction to- accomplish this work successfully.I
Asella town is suffering from the would like to express my gratitude to the staff of
discontinuous supply of water in the distribution Asella Town water supply office that helped me a
systems. Pressures in distribution system fail at the lot to get all required data and finally I would like
maximum consumption hour and high during night to express my deepest gratitude to all my friends
time as consumption decreases. The deficiency of for those share supportive idea in doing the
hydraulic parameter (flow velocity and pressure) research.
occurred due to random connection (placement) for Conflict of interest:The authors declare no conflict
nodes and pipe without any scientific of interest.
method/mathematical calculation of flow and
pressure. Therefore, this study is used to evaluate REFERENCES
the hydraulic performance of a water supply [1]. Mekuriaw, T. Analysis of Current and
distribution system of Asella town by using Future Water Demand Scenario in Yejube.
Bentley water GEMS V8i. The total annual water International Journal of Advanced Science
loss of the water supply system is 28.39% in 2016, and Engineering 2019, 6(2),1291-1304.
31.83% in 2017, 39.28% in 2018, 37.16% in 2019 [2]. Dr. R. M. Damgir; Supriya, P. Hydraulic
and 39.55% in 2020 G.C. Modelling of Water Distribution System
In extended period simulation of peak with Reference to Technical Sustainability
hour consumption, parts of the distribution system by using Water Gems.International Journal
receive water with low pressure and under any of Innovative Research in Science,
DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1432
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 3, Issue 10 Oct 2021, pp: 1418-1433 www.ijaem.net ISSN: 2395-5252

Engineering and Technology 2017,6(6), distribution system using flow distribution.

10832–10838. MSc Thesis. California Polytechnic State
[3]. Ermias,M. Assessment of urban water University, San Luis Obispo 2012.
supply and sanitation: the case of Bedesa [9]. MoWIE (2019). ‘Water Supply & Sanitation
Town, Damot Woyde woreda of Wolaita Sector National ODF Campaign Framework
Zone, Southern Ethiopia. Journal of Biology, .
Agriculture and Healthcare 2016,6,60-64. [10]. MD. ATIQUZZAMAN. Water
[4]. American Water Works Association. distribution network modeling:
Computer Modeling of Water Distribution Hydroinformatics approach.
Systems. Manual of Water Supply ScholarBank@NUS Repository 2004.
Practices—M32,2017. [11]. Udhane, D.; Kataria, N.; Sankhe, S.; Pore, T.
[5]. Desalegn, W. B. Water Supply Coverage And Motegaonkar, S. Hydraulic Modeling
and Water Loss in Distribution Systems The and Simulation of Smart Water Distribution
case of Addis Ababa.Procedia Engineering Network. International Research Journal of
2005, 70(1) 958-966. Engineering and Technology 2018, 5(4)
[6]. Bentley Water GEMs .Water Distribution 3636-3639.
Design and Modeling, Full Version V8i. [12]. WHO/UNICEF. Progress on household
2008. drinking water, sanitation and hygiene I
[7]. EPA (2010) Control and mitigation of 2000-2017:World Health Organization and
drinking water losses in distribution United Nations Children’s Fund Joint
systems’. Monitoring Program for Water Supply and
[8]. Hopkins,M. Critical node analysis of water Sanitation 2017.

DOI: 10.35629/5252-031014181433 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 1433