12-03-2020

WATER SUPPLY ENGINEERING

Dr. MANOJ K TIWARI
SCHOOL OF WATER RESOURCES, IIT KHARAGPUR

Module 08: Water Distribution System

Lecture 42: Basics of Water Distribution System
EL
1
PT
N

 Distribution of Treated Water

 Requirements of a Water Distribution System
 Pressure in Water Distribution Networks
 Types of Water Distribution Systems (Gravity, Pumped and Combined)
 Components of a Water Distribution System

1
 12-03-2020

Components of Water Works for Municipal Water Supply

 Collection Works – Water Withdrawal / Water Intake from source

 Transmission Works – Conveyance of Raw Water from Collection Unit

to Treatment Unit

 Treatment (or Purification) Works – Water Treatment Plant (WTP)

 Transmission Works – Transmission of Treated Water from WTP to

Service Storage

 Distribution Works – Distribution of Treated Water to Consumers

through Pipe Network
EL
3
PT
N

Distribution of Treated Water

 So far, we have discussed withdrawal (intake)
and treatment of water to meet the end-use
requirements.

 After water treatment, the treated water needs

to be transported to the consumers through a
labyrinth of pipes and appurtenances, known as
water distribution system.

 An elaborate arrangement of pipe, valves,

appurtenances, pumps and storage structures
and other supporting infrastructure constitutes
a municipal water distribution system.

Image Source: http://www.pacificwater.org/pages.cfm/water-services/water-demand-

management/what-water-demand-management/the-water-supply-system.html

2
 12-03-2020

Water Distribution System

• Distribution system infrastructure is generally the major
asset of a water utility.

• The American Water Works Association (AWWA, 1974)

defines the water distribution system as “including all
water utility components for the distribution of finished or
potable water by means of gravity storage feed or pumps
though distribution pumping networks to customers or
other users, including distribution equalizing storage.”

• These systems must also be able to provide water for non

potable uses, such as fire suppression and irrigation of
landscaping.

Image Source: https://www.semanticscholar.org/paper/Smart-water-distribution-network-solution-

for-smart-V.Mohanasundaram-Joyce/68fd6115b60cdbc5b82ca0fe4afd885384f25b67/figure/0
EL
5
PT
N

Salient Features of Water Distribution Networks

 Distribution system is a network of pipelines that distribute water to the consumers.

 They are designed to adequately satisfy the water requirement for a combination of:
 Domestic
 Commercial
 Industrial
 Firefighting purposes

A good distribution system should satisfy the followings:

• Water quality should not get deteriorated in the distribution pipes.
• It should be capable of supplying water at all the intended places with sufficient
pressure head.

3
 12-03-2020

Pressure in Water Distribution Networks

 Design pressure at the consumer
end is known as minimum
residual pressure (MRP), or
minimum pressure head (MPH).

 It is important to maintain
adequate (good) pressure head in
water distribution network and at
consumer end, as:

 Supplies at low pressure often fail to meet required demand, and dwellings at higher floors
may not get water without additional pumping arrangements

 Excessively high pressure can be a source of high water losses and wastage, damage of joints,
supply lines, faucets throughout the plumbing system, and can lead to higher water bills.
Image Source: https://www.plumbingsupply.com/residential-water-pressure-explained.html
EL
7
PT
N

Pressure Requirements for Water Distribution Networks

As per CPHEEO, for 24x7 Systems: Minimum water pressure at consumer locations of different countries

 The minimum pressure maintained at any point

in the system varies between urban areas
according to their circumstances. However, a
reasonable objective is to maintain a minimum
pressure of 10 meters (m) at any point –
generally sufficient to reach the second storey of
a domestic dwelling.
 Buildings requiring water at higher elevations
receive the continuous supply into a ground-
level holding tank from which it is usually
pumped to the roof or intermediate level storage
from which the water is supplied by gravity to
the building through internal plumbing.
Sources: Guidance Notes for Continuous Water Supply (24x7 Supply), MoUD, GoI; Strijdom et al (2017). An assessment of sub-standard water pressure in South African potable distribution systems. J. Wat.
Sanitat. Hyg. Develop., 7(4), 557-567.; Jacobs and Strijdom (2009). Evaluation of minimum residual pressure as design criterion for South African water distribution systems. Water SA, 35(2).

4
 12-03-2020

Essential Requirements of a Good Water Distribution Network

 It should be capable of supplying the requisite amount of water at adequate pressure head for
the domestic and industrial uses, and during firefighting.

 The layout should be such that no consumer would be without water supply, during the repair
of any section of the system.

 All the distribution pipes should be preferably laid one metre away or above the sewer lines.

 It should be fairly water-tight as to keep losses due to leakage to the minimum.

 During repairs, it should not cause any obstruction to traffic. The pipelines should not be laid
under highways, carriage ways but below foot-paths.

 The distribution system should be economical and easy to maintain and operate.

EL
9
PT
N

Types of Water Distribution Systems

Depending upon topography of the town, the water may be distributed by:

 Gravity System

 Pumped System

 Combined Gravity and Pumped System (dual system)

10

5
 12-03-2020

Gravity Systems for Water Distribution

 In this system, the gravity force is used to distribute water from source at higher
elevation to lower level zone consumers

 This eliminates the need of pumping altogether.

 The method is economical, reliable and requires less maintenance.

 For proper working of the system the difference of head available between
service reservoir and low level zone should be sufficient to develop enough
pressure at consumers tap.

 This method also minimizes wastage and leakages.

EL
11
PT
N

Gravity Systems for Water Distribution

Advantages Disadvantages
o Simple and economical. o Not suitable if source is not at sufficient elevation
o Less maintenance. to cause flow under gravity.
o Less leakage and wastage. o Cannot provide high pressure for fire demand.
o Reduced pipe sizes.

Image Source: http://www.engineeringenotes.com/essay/water-engineering/essay-on-distribution-system-of-water-water-engineering/15804

12

6
 12-03-2020

Pumped Systems for Water Distribution

 The treated water is directly pumped into distribution mains
without storing in reservoirs. High lift pumps are provided
for forcing water into mains.

 Since the water is fluctuating, the pumps have to operate at

various rates during day. A continuous monitoring is required
at the pumping station to regulate the flow by running only
the required number of pumps out of total pumps installed.

 Pumping system can supply the required quantity of water

for firefighting by running all pumps including standby.
Image Source: http://www.engineeringenotes.com/essay/water-engineering/essay-on-distribution-system-of-water-water-engineering/15804

EL
13
PT
N

Pumped Systems for Water Distribution

Advantages
o Large quantity of water available in case of fire.
o Suitable for any type of topography.

Disadvantages
o Not economical.
o More losses and wastage.

Image Source: http://www.engineeringenotes.com/essay/water-engineering/essay-on-distribution-system-of-water-water-engineering/15804

14

7
 12-03-2020

Combined Gravity and Pumped Systems

 Water is pumped and stored in the elevated
services reservoir (ESR). According to the
supply hours the stored water is supplied to
the public by gravity.
o The system is economical, efficient and
reliable and adopted practically
everywhere.
o Pumping at constant rate increases
efficiency.
o Special supervision is not required.
o Fire demand efficiently met with.
Image Source: http://www.engineeringenotes.com/essay/water-
engineering/essay-on-distribution-system-of-water-water-
o Water is available even during failure of engineering/15804

pump and power.

EL
15
PT
N

Key Components of a Water Distribution System

• Pumping station:
pumps, motors and other electrical and structural components

• Distribution storage:
Mass Balancing Reservoirs, Service Reservoirs, Intermediate Tanks etc.

• Pipes and appurtenances:

Pipes, valves, meters etc.

16

8
 12-03-2020

Storage in Water Distribution Systems

 Water is storage in elevated, underground
or grounded reservoirs in distribution
systems, for ensuring desired rate of supply
in usual hours, with provisions of meeting
emergency demands.

 Usually, the storage reservoirs in

distribution networks include Storage
Reservoirs or Mass Balancing Reservoirs
(MBR), Elevated Service Reservoirs (ESR),
and at times, Intermediate Tanks.

These have already been discussed

in Week 4 (refer to lectures 19-21) Image Source: https://web.stanford.edu/~siegelr/florida/florida2012/IMG_2836%20water%20tower.jpg;
https://mumbaimirror.indiatimes.com/mumbai/other/unclean-elevated-storage-reservoirs-pose-a-
threat/articleshow/16029169.cms

EL
17
PT
N

Pumping in Water Distribution Systems

 Pumps in water distribution systems are
required for lifting water or providing
additional pressure/head in supply lines.

 Different types of pumps are used,

depending on the functional requirement,
capacity needed, and budget available.

The different types of pumps and

their functional approaches have
already been discussed in Week 3
(refer to lecture 16) Image Source: https://steelguru.com/infra/india-s-inaugurates-world-s-largest-underground-water-
pumping-station-in-telangana/546483

18

9
 12-03-2020

EL
19
PT
N

WATER SUPPLY ENGINEERING

Dr. MANOJ K TIWARI
SCHOOL OF WATER RESOURCES, IIT KHARAGPUR

Module 08: Water Distribution System

Lecture 43: Water Distribution Networks

20

10
 12-03-2020

 Layouts of Water Distribution Networks

 Branched and Looped Pipe Configurations
 Pipes in Water Network
 Hydraulics of Pipe Network
 Friction (Major) and Local (Minor) Losses
 Equivalent Pipe for Pipes Connected in Series or Parallel

EL
21
PT
N

Layouts of Water Distribution Networks

 Two Basic Configurations: Branched configuration Looped configuration

 Four principal design approaches of a distribution system layout:

 Dead end or tree system
 Gridiron system
 Circular or ring system
 Radial system
Image Source: https://www.nap.edu/read/11728/chapter/3#20

22

11
 12-03-2020

Dead End or Tree System: Branched Network

 It is similar to the branching of a tree and consists of:
Main (trunk) line, sub-mains and branches.

 Main line is the main source of water supply. There is

no water distribution to consumers from trunk line.

 Sub-mains are connected to the main line and are

laid along the main roads.

 Branches are connected to the sub-mains and they

are along the streets.

 Service connections are given to the consumers from

branches.
Image Source: https://theconstructor.org/environmental-engg/water-distribution-system-layout/21217/
EL
23
PT
N

Dead End or Tree System: Branched Network

Advantages:
• It is a very simple method of water distribution with easy calculations.
• The required dimensions of the pipes are economical.
• This method requires comparatively less number of cut-off valves.

Disadvantages:
• The area receiving water from a pipe under repair is without water until the work is completed.
• In this system, there are large number of dead ends where water does not circulate but remains static.
• Sediments accumulate due to stagnation and can cause bacterial growth at these points.
• It is difficult to maintain chlorine residual at the dead ends of the pipe.
• Water available for fire-fighting will be limited since it is being supplied by only one water main.
• The pressure at the end of the line may become undesirably low as additional areas are connected to the
water supply system. This problem is common in many underdeveloped and developing countries.

24

12
 12-03-2020

Grid-Iron System
 In grid pattern, all the pipes are interconnected with no dead-ends,
and, water can reach any point from more than one direction.

Advantages:
• Stagnation does not occur as readily as in the branching pattern.
• In case of repair or a break down in a pipe, the area connected to that
pipe will continue to receive water from the other side.
• Water reaches all points with minimum head loss.

Disadvantages:
Image Source: https://sswm.info/sswm-university-course/module-
• More length of pipes is required, increasing the cost of the pipe laying 2-centralised-and-decentralised-systems-water-and-sanitation-
1/network-design-and-dimensioning
• More number of valves are required.
• The calculation of pipe sizes are more complicated.
EL
25
PT
N

Circular (or Ring) Distribution System

 This system relies on grid pattern with loops. In such system, the
supply main forms a ring around the distribution area. The branches
are connected cross-wise to the mains and also to each other.

Advantages:
• No stagnation of water
• Repair works can be done without affecting larger network.
• Large quantity of water is available for firefighting.
• Most reliable for a town with well-planned streets and roads.
Disadvantages:
Image Source: https://sswm.info/sswm-university-course/module-
• Longer length and large diameter pipes are required. 2-centralised-and-decentralised-systems-water-and-sanitation-
1/network-design-and-dimensioning
• More number of cutoff valves are necessary.
• Skilled workers are necessary while laying pipes.

26

13
 12-03-2020

Radial System
 The area is divided into different zones. The water is pumped
into the distribution reservoir (usually elevated) kept in the
near middle of each zone. The supply pipes are laid radially
ending towards the periphery. All distribution reservoirs are
connected with main line which is passing through center of
the city. This is suitable for areas with radially designed roads.

Advantages:
• Calculation of pipe sizes is easy.
• The water distributed with high velocity and high pressure.
Image Source: https://sswm.info/sswm-university-course/module-
• Head loss is very small because of quick discharge. 2-centralised-and-decentralised-systems-water-and-sanitation-
1/network-design-and-dimensioning
Disadvantages:
• Increased project cost due to number of individual distribution reservoirs.
EL
27
PT
N

Pipes in Water Network

 Various pipe material have been used in distribution networks in the last 150 years. The most
common pipe materials developed so far are:
 Iron pipes – cast iron, ductile iron
 Asbestos cement pipes
 Unplasticised Polyvinyl Chloride (uPVC) pipes
 Medium density polyethylene (MDPE) pipes

 Some other pipe materials developed for specific uses are:

 Steel pipes with protective epoxy coating – For larger diameter underground pipelines
 Welded stainless steel pipes – For internal pipe work
 Glass reinforced plastic pipes – For large diameter pipes and pipe renovation work
 Pre-stressed concrete pipes – For larger diameter pipes

28

14
 12-03-2020

Hydraulics of Pipe Network Design: Pipe-flow Calculations

Theoretical formulae
 Adequate hydraulic calculations are required to enable the design of water networks.
Development in pipe-flow calculations have been taking place in parallel with water and material
science advancement. The first published formula for the calculation of flow of water in pipes was
Antoine Chezy (1770), known as CHEZY FORMULA

Where, V = average velocity in the pipe, r = hydraulic radius (area of flow/wetted perimeter),
𝑽 = 𝑪 𝒓𝑺 S = slope of the water surface/pipe, C = dimensionless constant representing pipe material friction

 Another formula was proposed by Weisbach in 1845 for pipe flow calculations, known as
WEISBACH FORMULA, which later came to be known as Darcy-Weisbach equation, after the
independent publication of Henry Darcy’s work in 1857, which added to the understanding of the
Weisbach equation.
𝟒𝒇𝒍 𝒗𝟐 Where, f = friction factor (dimensionless), l = length of pipe,
𝒉= D = diameter of pipe, v = average velocity of flow
𝑫 𝟐𝒈
EL
29
PT
N

Hydraulics of Pipe Network Design: Pipe-flow Calculations

Empirical formulae
 A few empirical equations based on extensive experimentation were also devloped, with the
generalized form of:
Where, the values of V, x and y were derived from experiments, with interpolations and
𝑽 = 𝑲𝒓𝒙 𝑺𝒚 extrapolation used to provide missing values

 The two foundation empirical equations for pipe-flow calculations, observed to be accurate over a
wide range of conditions, simple in structure and easily applicable by engineers for practical
design, are:
𝟐 𝟏
 Manning’s formula for open channel (preferred) as well as pipe flow: 𝑽 = 𝑴𝒓𝟑 𝑺𝟑
 Hazen-William’s formula for open channel and pressurized pipe flow (preferred): 𝑽 = 𝑪𝒓𝟎.𝟔𝟑 𝑺𝟎.𝟓𝟒
Where, M = manning’s constant, C = Hazen-William’s constant, r = hydraulic radius
(area of flow/wetted perimeter), S = slope of the pipe

30

15
 12-03-2020

Hydraulics of Pipe Network Design: Head Loss in Pipes

 The head loss due to friction between the moving water and pipe wall is called friction loss or major
loss, which is present all throughout the pipe length.
 Additional head loss due to local disruption of the fluid flow, caused by valves and junctions is called
local loss or minor loss.
 Friction loss is more dominant in long pipes,
whereas minor loss may becomes more
significant in short pipes.
 However, generally in pipe networks, minor
losses do not contribute significantly to
overall losses and can be neglected.
 The most common equations for calculation
of friction losses in pressurized pipes are
the Darcy-Weisbach equation and Hazen-
William’s equation Image Source: http://site.iugaza.edu.ps/halnajar/files/2011/09/Lecture-7.-Hydraulics-of-water-distribution-systems.pdf

EL
31
PT
N

Friction Losses or Major Losses

Darcy-Weisbach (Colebrook-White) equation

 This equation was independently developed by

Henry Darcy and Julius Weisbach, around 1850. It
relates the energy head loss due to friction along
a pipe to the average velocity of the fluid flow.
Where,  = non-dimensional friction factor,
L V2 a function of not only the pipe material but
hf   also the Reynold’s number, which depends
D 2g
on viscosity, density and flow velocity.

 The equation can be used for all pipe flow

categories (laminar, turbulent and transitional)
and as such it is a function of Reynold’s number,
Re and relative pipe roughness, Ɛ/D (Ɛ = absolute
pipe roughness) Image Source: https://analysisofflowinpipesg9uthm.weebly.com/

32

16
 12-03-2020

Friction Losses or Major Losses

Darcy-Weisbach (Colebrook-White) equation
 The functional behavior of  is obtained from
Moody Diagram.
 the value of  for laminar flow conditions is:
 =𝟔𝟒/𝑹𝒆. The laminar flow equation holds for
Re < 2000. For Re > 2000, the flow changes from
laminar to weakly turbulent, and beyond 4000, it
becomes turbulent.
 Over the whole turbulent flow region (Re >
4000), Colebrook-White formula is used to solve
for the Darcy-Weisbach friction factor, 
𝟏 Ɛ 𝟐. 𝟓𝟏 Approximated by: 𝟎. 𝟐𝟓
= −𝟐𝒍𝒐𝒈 + =
 𝟑. 𝟕𝑫 𝑹𝒆  Ɛ 𝟓. 𝟕𝟒
𝒍𝒐𝒈( + )
𝟑. 𝟕𝟏𝑫 𝑹𝒆𝟎.𝟗 Image Source: https://analysisofflowinpipesg9uthm.weebly. com/

EL
33
PT
N

Friction Losses or Major Losses

Hazen-William’s formula
 The Hazen-William’s formula is an empirical equation
used to calculate the head loss along a pipe, based on
the pipe flow and physical pipe properties

𝑸𝟏.𝟖𝟓
𝒉𝒇 = 𝟏𝟎. 𝟔𝟕 𝑳
𝑪𝟏.𝟖𝟓 𝑫𝟒.𝟖𝟕
Where, hf= head loss(m); Q= flow rate(m3/sec); L= length of
pipe(m); d= diameter(m); C= Hazen William’s coefficient

 Hazen-William’s friction coefficient (C) indicates the

roughness of the interior surface of a pipe

 The C value is considered to be a pipe constant. Lower

C values mean higher head losses in the pipe. Image Source: https://uomustansiriyah.edu.iq/media/lectures/5/5_2017_01_22!09_20_01_PM.pdf

34

17
 12-03-2020

Local Losses or Minor Losses

 Local losses are caused by increased turbulence as the
flow passes through valves or other pipe fittings such
as bends, tees and valves.

 The increased turbulence leads to headloss, which is

expressed as the fraction of the initial kinetic energy of
the pipe flow is:
𝒗𝟐
𝒉𝒍 = 𝒌 = 𝑲𝑸𝟐
𝟐𝒈

Where, hl= = local head loss, k = local headloss coefficient, K =

local resistance coefficient, v= characteristic velocity, Q = flow-
rate, g = acceleration due to gravity

 The value of K for various types of pipeline fittings can

be obtained from standard hydraulic textbooks Image Source: Roberson, Hydraulic Engineering

EL
35
PT
N

Local Losses or Minor Losses

Usually, smaller compared to overall friction losses along the pipe.

Image Source: https://analysisofflowinpipesg9uthm.weebly.com/; vanoengineering.wordpress.com

36

18
 12-03-2020

Equivalent Pipes: Pipes in Series

 The pipe links between two nodes may consist of a single uniform
pipe size (diameter) or a combination of pipes in series or parallel.

 For the pipelines made of different lengths of different diameters

connected in series, the loss of head can be computed as:
hL= hL1 + hL2+ hL3, where Q = Q1 = Q2 = Q3
N 2
 Using the Darcy Weisbach equation with constant friction factor f 8 fL Q
hL   2 i 5
and neglecting the minor losses in the pipes, head loss in pipes i 1  gDi
can be calculated as:
8 fQ 2 N
 With equivalent pipe diameter De for N number of pipes in series, hL  2 5
 gDe
L i
the total head loss will be: i 1
𝟎.𝟐
∑𝑵𝒊 𝟏 𝑳𝒊
 Thus, the equivalent dia of pipes in series can be computed as: 𝑫𝒆 = 𝑳
∑𝒊 𝟏 𝒊𝟓
𝑵
𝑫𝒊
EL
37
PT
N

Equivalent Pipes: Pipes in Parallel

 For the pipelines made of different lengths of different diameters
arranged in parallel, the loss of head can be computed as:
hL= hL1 = hL2 = hL3, where Q = Q1 + Q2 + Q3

 The pressure head at nodes A and B remains constant, i.e.

headloss in all the parallel pipes will be the same. Using the Darcy-
𝒈𝑫𝒊𝒉𝑳
Weisbach equation and neglecting minor losses, the discharge Q 𝑸𝒊 = 𝑫𝟐𝒊
𝟖𝒇𝑳𝒊
in pipe i can be calculated as: .
𝑵 𝟎𝟓
𝒈𝑫𝒊𝒉𝑳
 Thus, for N pipes in parallel, the value of Q is: 𝑸=  𝑫𝟐𝒊
𝟖𝒇𝑳𝒊
𝒊 𝟏 .
𝟎𝟓
𝒈𝑫𝒆𝒉𝑳
 The discharge Q flowing in the equivalent pipe would be: 𝑸 = 𝑫𝟐𝒆
𝟖𝒇𝑳
 The length L may be different than any of the pipe lengths L1, L2, L3 𝑵
𝟎.𝟐
𝟎.𝟓
etc. Equating these two equations of discharge 𝑫 =
𝑳
𝑫𝟐.𝟓
𝒆 𝒊
𝑳𝒊
𝒊 𝟏

38

19
 12-03-2020

EL
39
PT
N

WATER SUPPLY ENGINEERING

Dr. MANOJ K TIWARI
SCHOOL OF WATER RESOURCES, IIT KHARAGPUR

Module 08: Water Distribution System

Lecture 44: Analysis of Water Distribution Networks

40

20
 12-03-2020

 Water Distribution Network Analysis

 Energy Heads in Water Distribution Lines
 Analysis of Branched and Looped Networks
 Hardy Cross Method of Pipe Network Analysis
 Newton-Raphson Method and Linear Theory Method of Pipe Network Analysis

EL
41
PT
N

Water Distribution Network Analysis

 Analysis of water distribution network mainly
includes determining flow and head losses in the
various pipe lines, and resulting residual pressures.
 The analysis methods primely include concepts of
flow (mass) balance, and pressure head estimations
using head loss calculation (energy balance).
 Real water distribution networks are usually very
complex and difficult to analyze manually.
 Various soft computing tools (like EPANET,
WaterGEMS, LOOP etc.) are used to analyze water
distribution networks. However, these software
essentially work on basic principles of mass and
energy balance. Image Source: Scarpa et al. (2016). Elementary DMA Design of Looped Water Distribution
Networks with Multiple Sources. Journal of Water Resources Planning and Management, 142 (6).

42

21
 12-03-2020

Energy Heads in Water Distribution Lines

Pressure head = p/γ; Datum head = z; Velocity (kinetic head): v2/2g
Total Energy = p/γ + v2/2g + z (TEL) Hydraulic Head (Gradient) = p/γ + z (HGL)
Energy Balance Equation (Bernoulli's Equation):

Image Source: Clipped from https://www.youtube.com/watch?v=-oecDDrYfyY;

https://www.codecogs.com/library/engineering/fluid_mechanics/pipes/the-hydraulic-gradient.php
EL
43
PT
N

Analysis of Branched Networks

 Branched network, or a Tree Network, are distribution systems having no
loops, and consist of several distribution mains emerging out from a
common input point. These are not common in cities, but sometimes
used for Rural Water Supply.
Summarized Method for Branched Network Analysis:
 Pipe discharges are obtained by adding the nodal discharges and tracing
the path from tail to the input point until all the tail ends are covered.
 Nodal head is found out by proceeding from the input point and adding
head loss and form loss in each link until a tail end is reached.
 The process is repeated until all tails are covered.
Pumping head = maximum head loss + terminal head
Total discharge: QT =  Qoi (i=1 to iL) Image Source: Jolly et al. (2014). Research Database of Water Distribution System Models.
Journal of Water Resources Planning and Management, 140 (4).

44

22
 12-03-2020

Analysis of Looped Networks

 Looped network having one or more closed loops are preferred
from reliability point of view, and thus, are more common. With the
changing demand pattern, both discharge magnitude and flow
directions may change in many links in a looped network.
 In any pipe network, the following two conditions (the laws given
by Kirchhoff) must be satisfied:
o The algebraic sum of pressure drops around a closed loop must
be zero, i.e. there can be no discontinuity in pressure.
o The flow entering a junction must be equal to the flow leaving
that junction; i.e. the law of continuity must be satisfied.
 Based on these two basic principles, the pipe networks are
generally solved by the methods of successive approximation.
Hardy Cross Method, Newton- Raphson Method, and Linear Theory Image Source: Jolly et al. (2014). Research Database of Water
Method are the common methods for looped network analysis. Distribution System Models. J. Wat Resour. Plan. Managem. 140 (4).
EL
45
PT
N

Hardy Cross Method

 The procedure suggested by Hardy and Cross requires that flow in each pipe is assumed by the
designer in such a way that the principle of continuity is satisfied in each junction.
 A correction to these assumed flows is then computed successively for each pipe loop in the network
until correction is reduced to an acceptable magnitude.
 If Qa is the assumed flow and Q is the actual flow in the pipe, then: Q = Qa + , where  is correction
 Now, expressing headloss (HL ) as: HL = K. Qx = K. (Qa + )x = K. [Qax + x Qax-1 ] (neglecting higher order terms)
 Around a closed loop, summation of headloss should be zero:  K. [Qax + x Qax-1 ] = 0
  K. Qax = -  Kx Qax-1 
 As,  is considered same for all the pipes of considered loop, it is taken out of the summation:
 = - K. Qax/ xKQax-1 = - HL /  xHL/Qa (where HL is the headloss for assumed flow Qa)

 Since  is given the same sign (direction) in all pipes of the loop, the denominator is taken as the
absolute sum of the individual items. Hence, = - K. Qax/  xKQax-1  = - HL / [x. HL/Qa]

46

23
 12-03-2020

Hardy Cross Method

 The value of x in Hardy Cross method is usually assumed to be constant, 1.85 for Hazen-Williams
formula and 2 for Darcy-Weiscbach formula.
 Minor losses are generally neglected (can be introduced by substituting an equivalent length of pipe).
Analysis Steps:
 Divided the system into several loops such that each pipe is included in at least one loop.
 Assume any internally consistent distribution of flow. The sum of the flows entering any junction
must equal the sum of the flows leaving the junction.
 Compute the headloss in each pipe by means of an equation or diagram. Conventionally, clockwise
flows are positive and produce positive headloss, and vice versa.
 With due attention to sign, compute the total head loss around each circuit, ie,  Kqaxa, and compute
the sum of  xKQax-1 for the same circuit, without regard to sign. Calculate the corrections .
 Apply the corrections obtained from equations to the flow to each pipe. Pipes common to two loops
will receive both the corrections with due regard to sign.
EL
47
PT
N

Newton-Raphson Method
 The Newton–Raphson method is a powerful numerical method for solving systems of nonlinear equations.
Unlike the hardy cross method, the entire network is analyzed altogether.
 Suppose that there are three nonlinear equations F1 (Q1, Q2, Q3) = 0, F2 (Q1, Q2, Q3) = 0, and F3 (Q1, Q2, Q3) = 0 to
be solved for Q1 , Q2, and Q3. We adopt a starting solution (Q1, Q2, Q3) and consider that (Q1 + Q1, Q2 + 
Q2, Q3 +  Q3) is the solution of the set of equations. Expanding as Taylor’s Series:

Arranging in Matrix form, and solving:

 For large networks, the inverted matrix may be preserved and used for at
least three times to obtain the corrections (as large matrix inversions are
time-consuming). Knowing the corrections, the discharges are improved as:

48

24
 12-03-2020

Newton-Raphson Method: Analysis Steps

 Step 1: Number all the nodes, pipe links, and loops.
 Step 2: Write nodal discharge equations for all nodes, as:
(Qin is the discharge in nth pipe at node j, qj is nodal withdrawal, and jn is the total number of pipes at node j).
 Step 3: Write loop head-loss equations as:
(where Kn is total pipes in kth loop).
 Step 4: Assume initial pipe discharges Q1, Q2, Q3, ... satisfying continuity equations.
 Step 5: Assume friction factors fi (generally 0.02) in all pipe links and compute corresponding Ki.
(where i=pipe link number to be summed up in the loop k).
 Step 6: Find values of partial derivatives ∂Fn= ∂Qi and functions Fn, using the initial pipe discharges Qi and Ki.
 Step 7: Find Qi. The equations generated are of the form Ax = b, which can be solved for  Qi.
 Step 8: Using the obtained  Qi values, the pipe discharges are modified and the
process is repeated again until the calculated  Qi values are very small.
EL
49
PT
N

Linear Theory Method

 The linear theory method, presented by Wood and Charles (1972), and analyzes the entire network is
altogether like the Newton– Raphson method.
 The nodal flow continuity equations are linear, but the looped head-loss equations are nonlinear. However,
in this method, the looped energy equations are modified to be linear for previously known discharges and
solved iteratively. The process is repeated until the two solutions are close to the allowable limits.
 The nodal discharge continuity equations are:
 The nodal discharge equation can be generalized for the entire network:
(where ain is +1 or -1 for positive or negative discharge flows in pipe n, respectively.
ain is 0 if pipe n is not connected to node j. The total pipes in the network are i L).

 The loop head-loss equation for all loops, can be linearized as:
 The equation can be generalized for the entire network, as:
(where bkn = Kn  Qkn for initially known pipe discharges if pipe n is in loop k, or otherwise b kn =0).

 The coefficient bkn is revised with current pipe discharges for the next iteration. This results in a set of
linear equations, which are solved by using any standard method for solving linear equations.

50

25
 12-03-2020

Linear Theory Method: Analysis Steps

 Step 1: Number all the nodes, pipe links, and loops.
 Step 2: Write nodal discharge equations for all nodes, as:
(Qin is the discharge in nth pipe at node j, qj is nodal withdrawal, and jn is the total number of pipes at node j).
 Step 3: Write loop head-loss equations as:
 Step 4: Assume initial pipe discharges Q1, Q2, Q3, ... It is not necessary to satisfy continuity equations.
 Step 5: Assume friction factors fi (generally 0.02) in all pipe links and compute corresponding Ki.
(where i=pipe link number to be summed up in the loop k).
 Step 6: Generalize nodal continuity and loop equations for the entire network.
 Step 7: Calculate pipe discharges. The equation generated is of the form Ax = b,
which can be solved for Qi and recalculate coefficients bkn from the obtained Qi values.
 Step 8: Repeat the process again until the calculated Qi values in two consecutive
iterations are close to predefined limits.
EL
51
PT
N

52

26
 12-03-2020

WATER SUPPLY ENGINEERING

Dr. MANOJ K TIWARI
SCHOOL OF WATER RESOURCES, IIT KHARAGPUR

Module 08: Water Distribution System

Lecture 45: Practice Problems on Pipe Flow and Water Distribution Network Analysis
EL
53
PT
N

 Worked Examples on Flow Through Pipes

 Worked Examples on Equivalent pipes
 Worked Examples on Hardy Cross Method

54

27
 12-03-2020

Practice Problem 1: Reynolds Number

Water at 150C with density 999 kg/m3 and viscosity 1.138×10-3 kg/m.s flowing steadily in 60 mm diameter
horizontal pipe made of stainless steel at a rate of 10 L/s. Suggest if flow is laminar, transitional or turbulent.

Solution: Average flow velocity,

Therefore, Reynolds Number = 1676529.2

Since Re > 4000, the flow is turbulent.

EL
55
PT
N

Practice Problem 2: Frictional Head Loss in Pipe Flow

Water at 150C flows through a 150 m long galvanized steel pipe of diameter 200 mm and at 0.265 m3/s. The
kinematic viscosity of water at 15 0C is 1.14×10-6 m2/s and average surface roughness for galvanized steel=0.15
mm. Determine the loss of head due to friction.

Solution: Average flow velocity,

Therefore, Reynolds Number

Relative Roughness

Friction Factor (from Moody Chart)

Hence, frictional head loss:

= 51.74 m

56

28
 12-03-2020

Practice Problem 3: Minor Head Loss

A 10 cm diameter pipe has a discharge of 600L/min. At a section, the pipe has a sudden expansion to a size of
15 cm diameter. If the pressure just upstream of the expansion is 40 kN/m2, calculate the pressure just after
expansion. Assume pipe to be horizontal at the expansion region.

Solution: Loss of head at sudden expansion:

From Energy Conservation Equation (where z1=z2):

Source: https://analysisofflowinpipesg9uthm.weebly.com/

EL
57
PT
N

Practice Problem 4: Pipes in Series

An arrangement of three pipes in series between tank A and tank B as shown in figure. Assuming the Darcy-
Weisbach friction factor f = 0.02 and neglecting the minor losses. Calculate the equivalent pipe diameter and the
discharge in the equivalent pipe.

Solution:
. .
∑
The equivalent pipe dia, 𝐷 = =
∑ . . .

= 𝟎. 𝟏𝟖𝟓𝒎
× . ×
The equivalent pipe constant 𝐾 = =
. × . × .
= 11450.49 𝑠 /𝑚 Image Source: Swamee, P.K. and Sharma, A.K., 2008. Design of water supply pipe
networks. John Wiley & Sons.

. .
Therefore, discharge in pipe, 𝑄 = =
.
𝟑
= 𝟎. 𝟐𝟖𝟎 𝒎 𝒔

58

29
 12-03-2020

Practice Problem 5: Pipes in Parallel

For a given parallel pipe arrangement in figure. calculate the equivalent pipe diameter and corresponding flow.
Assume Darcy-Weisbach's friction factor f = 0.02 and neglect minor losses. the length of equivalent pipe can be
assumed as 500m.

Solution:
. .
.
.
The equivalent pipe dia, 𝐷 = ∑ 𝐷 = ×
. .
0.25 . + × 0.20 . = 𝟎. 𝟐𝟖𝟑𝒎

Therefore, discharge in pipe,

. Image Source: Swamee, P.K. and Sharma, A.K., 2008. Design of water supply pipe
networks. John Wiley & Sons.
𝑄 = 𝜋𝐷
. × . × .
= 3.14 × 0.283 × 0.283 ×
× . ×
= 𝟎. 𝟐𝟎𝟗 𝒎𝟑⁄
𝒔

EL
59
PT
N

Practice Problem 6: Hardy Cross Method

Calculate the head loss and the corrected flows in various pipes of a distribution network as shown. The
diameter and length of the pipes are shown against each pipe. Compute the corrected flows for two corrections
using Hardy Cross method and William Hazen formula (Assume CH= 100).

Solution:
Assign the magnitudes and
directions of the possible flows in
each pipes considering the law of
continuity at each junction.
Thereafter, the two pipe loops need
to be analyzed for computing the
head loss using Hazen Williams
equation.

HL = (1/0.094) * (Q/CH)1.85 * (L/d4.87)

 HL = (L/470) * (Qa1.85 /d4.87)
 HL = K*Qa1.85 (here, K= L/470d4.87)
Image Source: Garg, S.K., 2010. Water supply engineering. Khanna Publishers.

60

30
 12-03-2020

Practice Problem 6: Hardy Cross Method

Solution: Computations of K value for pipe networks

Pipe L in m d in m K = L/470 d4.87

For Loop ABCD
AB 500 0.30 373
BC 300 0.20 1615
CD 500 0.20 2690
DA 300 0.20 1615
For loop DCFE
DC 500 0.20 2690
CF 300 0.15 6580
FE 500 0.15 10,940
ED 300 0.15 6580
Image Source: Garg, S.K., 2010. Water supply engineering. Khanna Publishers.
EL
61
PT
N

Practice Problem 6: Hardy Cross Method

Solution: Hardy cross procedure for first correction
Pipe Assumed flow K value HL = K*Qa1.85 HL /Qa Corrected
l/s m3/s flows

AB 45 0.045 373 1.12 26 49.9

BC 23 0.023 1615 1.52 66 27.9
CD -20 -0.020 2690 -1.94 97 -7.9
DA -35 -0.035 1615 -3.23 92 -30.1
Σ - 2.53 232
Δ1= ΣHL/x.Σ (HL /Qa) = [(-) -2.53/ 1.85*281]*1000 = 4.91 l/s

Pipe Assumed flow K value HL = K*Qa1.85 HL /Qa Corrected

l/s m3/s flows

DC 20 0.020 2690 1.94 97 7.9

CF 28 0.028 6580 8.80 314 20.8
FE -8 -0.008 10940 -1.47 184 -15.2
ED -5 -0.005 6580 -0.37 74 -12.2
Σ 8.9 669
Δ1’= ΣHL/x.Σ (HL /Qa) = - [8.9/ 1.85*669]*1000 = - 7.21 l/s

62

31
 12-03-2020

Practice Problem 6: Hardy Cross Method

Solution: Hardy cross procedure for second correction
Pipe Assumed flow K value HL = K*Qa1.85 HL /Qa Corrected
l/s m3/s flows

AB 49.9 0.0499 373 1.456 29 48.1

BC 27.9 0.0279 1615 2.150 77 26.1
CD -7.9 -0.0079 2690 -0.347 44 -10.6
DA -30.1 -0.0301 1615 -2.475 82 -31.9
Σ 0.784 281
Δ2= ΣHL/x.Σ (HL /Qa) = [(-) 0.784/ 1.85*232]*1000 = -1.81 l/s

Pipe Assumed flow K value HL = K*Qa1.85 HL /Qa Corrected

l/s m3/s flows

DC 20 0.020 2690 1.94 97 7.9

CF 28 0.028 6580 8.80 314 20.8
FE -8 -0.008 10940 -1.47 184 -15.2
ED -5 -0.005 6580 -0.37 74 -12.2
Σ 8.9 669
Δ2’= ΣHL/x.Σ (HL /Qa) = [(-) 1.197/ 1.85*756]*1000 = 0.9 l/s
EL
63
PT
N

Practice Problem 6: Hardy Cross Method

Solution: Corrected discharge after second correction

Pipe Corrected flow in l/s

AB 48.1
AD 31.9
DC 10.6
DE 11.3
EF 14.3
BC 26.1
CF 21.7

64

32
 12-03-2020

EL
65
PT
N

33