Chapter 20

WATER TRANSMISSION
AND DISTRIBUTION

IlntroductionlClassification of transmission and

distribution systems/J?istribution network!
Design considerations/Service connections

20.1 INTRODUCTION

Water transmission is the transport of water from one location to another by means of
conduits, canals, aqueducts or tunnels. Water may be required to be transported from
source to treatment plant or treatment plant to consumers. Water distribution system
refers to that part of the water supply system which receives water from a pumping
station, elevated reservoir or from conduits and delivers it to consumers distributed
over an area. In piped water supply systems, water transmission is usually achie ed
by a single conduit where distribution is done by a network of pipelines . The part of
the pipeline which coveys bulk amounts of water through a single pipe is called ' tran s-
mission main' and the part in which pipes are arranged to distribute th e con ve ed
water to different consumption points is called 'distribution system'. The transmis-
sion and distribution systems include storage reservoir, pipes, valves, hydrants, house
connections, standpipes, etc.

The main purposes of the construction of water transmission and distributi on pipe-
lines are:
• to make water available in close proximity to the consumers ;
• to supply water in adequate quantities according to the demand of the con-
sumers;
• to supply water with adequate pressure;
• . to regulate water supply as per requirement.

389
390 PART III: WATER SUPPLY

In small water supplies, water transmission and distribution are not required to
be dealt with separately. The design considerations are similar and hydraul i
design can be done by applying the same equations, charts, nomographs , etc.

20.2 CLASSIFICATION OF TRANSMISSION AND DISTRIBUTION

SYSTEMS
Water transmission and distribution may be classified depending on the meth ods
involved as:

• gravity-flow system,
• system with direct pumping and
• system with pumping and storage.

A gravity system is adopted when the source of supply is at a sufficient elevation

with respect to the consumption points. In the gravity-flow system shown in
Figure 20.1 water is conveyed by gravity and no energy is required to operate
it. In the system with pumping, water is pumped into the transmission main or
distribution system as shown in Figure 20.2. The energy thus imparted by
pumping into the system causes flow and provides residual pressure required at
the consumption points. A system with pumping and storage is called a direct-
indirect or dual system as shown in Figure 20.3. When the demand rate exceeds
the rate of pumping, the flow into the distribution system is caused by pumping
and by the elevated reservoir. When the pumping rate exceeds the consumption
rate, the excess water is stored in the reservoir. The elevated reservoir can feed
the distribution network when the pump is not in operation.

Water source/storage

Water consumption

Figure 20.1 Gravity-flow water supply system

 CHAPTER 20 : WATER TRANSMISSION AND DISTRIBUTIO

The advantages and disadvantages of the three systems xare presented In

table 20.1.

Table 20.1 : Advantage and disadvantages of transmission and distribution

systems

System Advantagesl Disadvantages

Gravity-flow Advantages
system 1. Requires no energy to operate as water is
conveyed by gravity
2. No pump is required and there are very few
moving parts
3. Construction, operation and maintenance are
simple
Disadvantages
1. Not applicable in flat countries where an
elevated source of water supply is not available
2. Water loss by leakage and wastage is
comparatively higher as the system remains
under constant pressure.
Direct pumping Advantages
system 1. Water can be pumped only when required,
2. Low water loss due to system leakage.
Disadvantages
1. Direct pumping at a uniform rate is not able to
meet varying water demand and maintain
required pressure under varying rates of
consumption ,
2. A power failure means breakdown of the
system ,
3. Maintenance and operation costs are high,
4 . Inflow of water through leaks may cause water
contamination durinq non-pumpinq hours.
Pumping with Advantages
storage system 1 . The system is more reliable and can cope with
fluctuation of water demand,
2. The pumps can be operated at rated capacity,
resulting in higher efficiency and economy 0
operation ,
3. Reasonable pressure can be maintained ith
varying water demand and there is no possibility
of inflow of polluted wate r in the system.
Disadvantages
1 . Relatively higher initial cost.
2. Comparatively highe r loss due to leakage and
wastaqe .
392 PART III: WATER SUPPLY

Pump

Transmission main

Water distribution

Figure 20.2 System with direct pumping

Pum P ~ater ResE rvoir

............ ;:.........;
o Transmission main
'-r"" ...........
Water reservoir
(alternative

Water distribution network

Figure 20.3 System with pumping and storage

20.3 DISTRIBUTION NETWORK

There are basically two main types of distribution networks:

• branched distribution network and

• looped distribution network.

The distribution pipe networks mostly follow the road network or access roads .
The looped network can also be arranged into :
• grid iron systems,
• ring system and
• radial system
depending on the requirements and the road network system.
 CHAPTER 20: WATER TRANSMISSION A D DI

The branched network, also called the dead-end system, compriss a transmission
supply main starting from the service reservoir or source and laid along the
main road with sub-mains branching off from the main at an angle in both
directions along other roads joining the main road as shown in figure 2004. The
minor distributors or branches may also come out of the sub-mains to connect
small clusters of consumers. The network is suitable for developing areas with
an irregular pattern of development. The looped distribution network is an
improvement over the dead-end system. Here the ends of mains or sub-mains
are connected to avoid dead-ends as shown in Figure 2004. The network is
suitable for a well-planned developed area with a definite pattern of road
network.

Source

NOdes/

-
Branched distribution system Looped network distribution system

Figure 20.4 Branched and looped distribution networks

The relative advantages and disadvantages of the t\\O. networks are given in Table
20.2.
394 PART III: WATER SUPPLY

Table 20.2: Advantages and disadvantages of branched and looped

distribution networks

Distribution network Advantages and disadvantages

Branched network Advantages

1. Relatively cheap as length of pipe required is less
2. Easy for hydraulic design and determining dischar e
and pressure at any point
3. Can be easily expanded to provide coverage to newty
developed areas
Disadvantages
1. Stagnant water, particularly at dead-ends, promotes
Sedimentation and water contamination,
2. Frequent blow-off or flushing is needed to keep the
system clean
3. Repair work in mains and sub-mains cuts off water
supply downstream
Looped network Advantage
1. No stagnation of water, consumption of water at an y
point activates flow in the whole network
2. Continuity of water supply anywhere in the system
despite any repair work to a main or sub-main
3. The system provides very good control over flow of
water.
Disadvantages
1. Relatively high initial cost
2. A large number of valves is needed, if control of flow
in the system is desired

20.4 DESIGN CONSIDERATIONS

Water demand and peak factor

The demand for water depends on many factors and varies hourl y, dail y a nd
seasonally. The per capita water con sumpti on is the total water consumed by a
community in a year di vided by the popul ation of the community and numbe r
of days in the year. Water consumption on hot summer days is higher than that
on a cold winter day. The peak daily demand can be estimated by adding 30 to
50% to the average daily demand over a year. Thus the peak factor for dail y
demand (J,) may be assumed to be between 1.3 and 1.5.

The hourly variation in water demand is much greater. Usually the max imu m
demand occurs in th e mornin g and min imum late at night. A typical vari ation
of water demand during a day is shown in Figure 20 .5. The peak hourl y water
demand can be estimated by mUltipl ying the average hourly demand by hourl y
peak factor (j2)' The peak factor is dependent on the size and inherent character
of the community. The hourl y peak factor for a small er community tends to be
high and it is usually less for larger co mmunity and large urban centres. Th e
 CHAPTER 20: WATER TRANSMISSION AND DI STRI

hourly peak factor (fJ ranges from 1.5 to 2. A water distribution system .
usually designed to cater for the maximum hourly demand, and the combined
peak factor (j) is computed as:

c:
o 2.0 r----------------------------------------------------,
li Peak consumption
E
~
C/J
c:
o
o
>-
"§

~ 1.°1.~='C=~.
~
~
Q)
Average consumption

cti
~

~ 0.00~~~--~--~----~--+---4---~--~--~----~---L--~
4 8 12 16 20 24
Hour

Figure 20.5 A typical variation of water demand during a day

The average per capita water demand varies between rural and urban centres
and again amongst the urban centres depending on the size of the urban centre.
In Bangladesh the average per capita per day water consumption used for
design purposes is given in Table 20.3 .

Table 20.3: Water consumption in Bangladesh

Areas Water consumption Peak factor, k

(Ipcd) (recommended)
Rural areas 50 3
Up_azila towns 100 2
Zilla towns 120 2
City corporations 180 1.5

The recommended peak factor can be used for the design of pipelines . If water
tanks are available at roof level in each household or at individual consumption
points the peak factor may be assumed to be l.5.

Design flow
The water supply distribution system is designed for peak water demand of a
future population. The design flow can be computed by using the equation:
396 PART III: WATER SUPPLY

Q =f q Pj / ( 1 - 0.01 w) (20.1 )

Where Q = peak design flow per day , Pj = design population, f = peak factor, q
= average water consumption per capita per day and w = loss and wastage in
percent. The future design population PI. can be estimated by applying any of
the population projection methods descnbed in section 17.1, depending on the
availability of population data. The peak design flow adopting the empirical
method of population estimation as shown in equation 17.1 becomes:

Q =fq PI' ( 1 + r )"/( l-O.Olw) (20.2)

PI' = present population, 11 = number of years (design period), and r = rate of

population growth per year.

Storage reservoir
Storage reservoirs are constructed in the distribution system to provide storage
to meet the fluctuation of water demand, stabilize pressure and provide storage
for emergency requirements. In the absence of a water reservoir in 'a direct
pumping system, pumping has to continue for all hours of supply and be
regulated to meet the fluctuation in demand. This is generally not economical
and sometimes not technically feasible. The storage reservoir should be large
enough to accommodate the cumulative difference between water supply and
demand .

The storage volume can be determined by plotting the cumulative water demand
and supply at different hours of a peak day as a percentage of total demand for
the day. The algebraic sum of the highest differences in demand and supply is
the capacity of the storage reservoir. Figure 20.6 shows the cumulative sum of
water demand from midnight to midnight over a 24-hour period .

The 24-hour and 12-hour supply by pumping at a constaM rate are also shown
in the figure . The volume of the water reservoir in case of a 24-hour supply is
aa' plus bb', which is about 22 % of the daily demand. If the water is supplied
over 12 hours at a constant rate during a high demand period, the volume of the
reservoir is cc' plus dd' , which is 28% of the demand for the day. A service
reservoir with a storage capacity of 20 to 40% of the peak water demand would
generally be adequate. A larger reservoir may be constructed where disruption
of water supply would be critical and where water for 'e mergency demand such
as fire fighting would be essential. The reservoir should preferably be located at
the centre of the water distribution network or as close to the distribution area as
possible.
 CHAPTER 20 : WATER TRANSMISSION AND DISTRIBUTIO

d
100
-
;,e
~
Cl
Cumulative water
consumption
c::
.c..
E
:::l

--
c...
c::
0
aE 24 Hours pumping
:::l 50 at constant rate
en
c::
0
()
~

.E:l
C\J
3
Q)
.:2:
co
"S
E
:::l
()
00
0 4 8 12 16 20 24
Hours

Figure 20.6 Cumulative water .consumption and supply by pumping

Distribution system design

The Hazen Williams equation can conveniently be used for the hydraulic
design of both branched or looped networks of pipes. The Hazen Williams
equation can be written in the form:

Q = 3.7 X 10"6 C D263 (HIL/ 54

(20.3)

where Q = flows, Ips

C = roughness coefficient (100-140 for rough to smooth pipes)
D = diameter, mm
H = head loss, m
L = length of pipe, m

For a definite value of C, the equation (20.3) can be written as:

HIL = 1.39 x 106 Q 185/ (D 4.87 ) [ For C = 130 ] (20.4)

HIL = 1.59 x 106 Q 1 85/ (D 4 87 ) [ For C = 120 ] (20 .5)

The process in volved in the design is to make a pipe layout, assume the pipe size
and then work out the terminal pressure head which could be made available at
398 PART III : WATER SUPPLY

the end of each pipe section when discharging the peak flow . The a ail
pressure heads are checked to see if they correspond to permissible re i
pressure heads. If not, the pipe size is changed and the system is rein vestiga
until satisfactory conditions are obtained.

Branched network: The following design procedure may be adopted for

branched network:
( 1) Collect/prepare a map of the area to be served with roads, streets and other
features and make a layout of mains, sub-mains and branches includ in g
location of valves and other appurtenances.
(2) Estimate the peak flow at different points and determine the quanti t
flowing through each section of the pipe. Peak flow = average daily flo w x
peak factor.
(3) Assume pipe sizes of all the pipes in the network (to calculate approximate
pipe size the velocity may be assumed to be around 1 rnIs).
(4) Calculate frictional head loss per unit length of pipe and then multiply b y
pipe length of the pipe to find the total head loss .
(5) Determine the terminal pressure head taking the change in the elevation of
the pipe into account.
(6) In case of a difference between the computed terminal pressure and the
permissible pressure head, revise the pipe size.

Looped network: The determination of probable flow in each pipe in a pipe

network requires complicated and tedious trial and error solutions. In any
looped pipe network two conditions must be satisfied:
• The flow entering a junction must equal the flow leaving it
• The algebraic sum of the pressure drop(head loss) around any closed loop·
must be zero

Pipe network problems in water distribution systems are usually solved by

methods of successive approximation, since any analytical solution requires the
use of many simultaneous equations, some of which are nonlinear. The equation
(20.4) can be written for a particular pipe as :

H= kQ' (20.6)

In which k is the constant depending on the length, diameter and roughness of

the pipe as well as the fluid property , x is the component equal to 1.85 for
Hazen Williams' equation and 2 for the Manning equation.

Hardy Cross developed a method of successive approximation in which the

circuits are balanced, distribution of flow is determined and the above two
conditions of flow are satisfied. The solutions for pipe network problems
suggesteq. by Hardy. Cross requires that the flow in each pipe be assumed so that
the principle of continuity is satisfied at each junction. A correction , ~, to the
assumed flow is computed successively for each pipe loop in the network until
the correction is reduced to an acceptable value. If Qa is the assumed flow and Q
is the actual flow, then the correction is Q - Qa i.e.
 CHAPTER 20 : WATER TRANSMISSION AND DISTRIBUTIO :m

(20.7 )

Inserting the value of Q from equation 20.7 and applying the condition that the
head loss around any closed loop is zero the head loss equation 20.6 gives:

(20.8)

Expanding equation 20 .8,

(20 .9)

If L1 is small compared to Q, the third and all successive terms of the equation
(20.9) may be neglected, hence,

IkQa' + !:J.xlkQa ,.J = 0 (20.10)

Solving equation 20.10 for !:J. :

!:J. = -lkQaX1(xlkQ/·J) or

!:J. = - IHI( xlHI QJ (20.11 )

Consider the simple loop shown in Figure 20.7, the arrowheads showing th e
assumed directi on of flow and head loss, the correction !:J. can be computed by :

Where L1 = flow correction, Qabc = assumed flow in clockwise direction , Q adc =

assumed flow in the counterclockwise direction, Hab = head loss in pipe ab, Hbc
= head loss in pipe be, Had = head loss in pipe ad, Hdc = head loss in pipe de and
x = exponent = 1.85 for Hazen Williams' equation.

a b
Qabc
Q --~------------~------------~ .

Qadc Qabc

Qadc
dL-----------~------------~c~ Q

Figure 20.7 A looped network for flow analysis

400 PART III : WATER SUPPLY

The procedure for the Hardy Cross method of pipe network analysis is given
below:
(1) Carefully examine the network and assume reasonable rates of flow in each
pipe such that inflow equals outflow at each junction
(2) In each loop, determine the head loss, Hand HIQ for all pipes
(3) With due attention to sign, compute the total head loss around each circuits
(4) Compute 'IHIQ for the same circuit without giving any consideration to
SIgn
(5) The correction , ~ is computed for each loop by the equation (20.11). The
minus sign may be disregarded if the correction so obtained is made b y
inspection
(6) Apply the correction to each pipe in each loop. When the sign of ~ is
negative (-), decrease the clockwise flows and increase the counterclockwis.e
flows. When the sign is positive (+), increase clockwise flows and decrease
counterclockwise flows. Pipes that are common to two loops require double
correction -
(7) With adjusted flow s, repeat the procedure for the second approximation. The
procedure is continued until the desired accuracy is attained

Computation of head loss in each pipe in a network using equation 20.4 or

20.5 is a tedious exercise. Head loss in a pipe can be easily obtained from the
head loss determination graph shown in Figure 20.8.

Pipe
diameter

0.5

0.2

0.1

0.05
~--I 0.02
I 0.01 1---++--+----,<-+---I-f__--itL--v-~-+-+-~-+.___I_+__+,'"_f__+_+_++__,H_,'"__A___+_+

0.005

0.002

0.001

0.0005

0.0002 -
104 ~~~__~~U-~L-~~-L~~L--WL_~~~~~~~~_ _~~

0.01 0.02 0.05 0.1 0.2 0.5 2 5 10 20 50 100 200 500 1000 2000 500010000
Flow Rate, a, Ips
Figure 20.8 Head loss determination diagram (for C = 120)
 CHAPTER 20: WATER TRANSMISSION AND DISTRIBUTION 401

20.5 SERVICE CONNECTIONS

The water from the distribution system is finally distributed to the consumers by
service connections. Service connections may be of three types:

• house connection,
• yard connection or
• public standpipe.

A house connection is a water service pipe connected to the in-house plumbing

system and having one or more taps. The service pipe is connected to the
distribution network in th e street by a T-piece for small diameter pipes, or by a
ferrule or saddle for large di ameter pipes. A yard connection is similar to a
house connection , the onl y diffe rence is that the taps are placed in the yard
outside the house. This type of connection is provided for low-income groups.

A public standpipe is the low-cost water distribution system for the population
who cannot afford to have a house or yard connection and is suitable for low-
income communities and urban slums. The standpipes should be located at a
suitable point in the commun ity within short walking distance. A single-tap
standpipe can comfortab ly erve 40-70 people; a multiple-tap standpipe may
provide a reasonable service for 250-300 persons . Sectional elevations of single-
tap and double-tap standpipes are shown in Figure 20.9.

Tap

Platform

Double-tap standpipe Single-tap standpipe

Figure 20.9 Single-tap and double-tap standpipes

Wastage of water from standpipes, especially when users do not tum the tap off
after use, can be a serious problem. In order to avoid excessive loss and wastage
402 PART III: WATER SUPPLY

public standpipes should operate at low pressure. Water collected fro

public standpipe is carried home in a container for consumption . The qu ali 0 -
water, particularly the bacteriological quality, may deteriorate during coli rio
transportation, storage and consumption. The water consumption from
standpipe is generally low, between 20 and 30 litres, because the water u e 0
purposes other than drinking, cooking, etc. are curtailed depending on tb
distance and the number of users of a standpipe. In spite of many shortco mi ne "
the public standpipe is the most suitable low-cost option for water suppl fo
large numbers of people who cannot afford to have house or yard connection

20.6 WORKED EXAMPLE

Example - 1
Calculate the peak water demand of a rural community having a present
population of 500 for a design period of 5 years. The average per capita water
consumption is 50 Ipcd with a peak factor of 3, the population growth rate is 2 %
( r=0.02) and the loss and wastage is 20%.

Solution :
Applying equation (20.2) Q=jqPI'( 1 +r)"/( 1-0.01w)
Q = 3 x 50 x 500 ( 1 + 0 .02 )5 / ( 1 - 0.01 x 20 )
or Q = 75,000 ( 1.02 ) 5/ ( 0.80 )
or Q = 103,508 Ipd = 103.508 m 3/day

Example - 2
Calculate the flow in each of the pipes in the following looped pipe network:

A B
45 Ips-. -=-'12 Ips
ctI
'0
800 m - 200 mm dia
E E
E E
0 0
L1")
L1")

E E
0 0
0 0
~
... 0 ~
Co..
4 Ips ~ ... 12 Ips
800 m - 150 mm dia
E
E E
E 0
0 0
L1")

E
0
E 0
0 ~

5 Ips
..E
....
0
~ F
r- 12 Ips
800 m - 100 mm dia
 CHAPTER 20: WATER TRANSMISSION AND DISTRIB

A flow of 45 Ips enters at joint A from a source. This water is drawn at node B,
C, D, E and F, at the rate of 12, 12, 4, 5, and 12 Ips respectivel for
consumption as shown.

Solution: First trial with assumed flow, Q"

Circuit z Length, m Dia. mm 00 Hr/L. Ho HoiOo 6. 01

Lps m/m M Ips Ips
AS 800 200 +25 0.0038 +3.04 0.1 216 +0.54 +25.54
BC 400 150 +1 3 0.0046 +1.84 0.1415 +0.54 +13.54
I AD 400 150 -20 0.0102 - 4.08 0.2040 +0.54 - 19.46
DC 800 150 - 8 0.0018 - 1.44 0.1800 +0.54- - 5.57
+1.89-

Sum : -0.64 0.6471

Correction 6. = - I HI( xIH/O ) = - ( - 0.64)/ ( 1.85 x 0.6471) = + 0.54
\ - 0.54-
DC 800 150 +8 0.0018 +1.44 0.1800 - 1.89- +5.57
II CF 400 100 +9 0.01 70 +6.80 0.7556 - 1.89 +7. 11
DE 400 150 - 8 0.0018 - 0.72 0.0900 -1 .89 - 9.89
EF 800 100 - 3 0.0022 - 1.76 0.5867 - 1.89 -4. 89

Sum: 5.76 1.6453

Correction 6. = - I H/( xIHlO ) = - ( +5.76)/(1.85 x 1.6453) = - 1.87

Common pipe DC required correction from both the circuits

Second trial with cor rected flo w, Q/

Corcuit Pipe Length, Dia. 01 Hl L. HI Hl 0 1 6. O2

m mm Lps m/m m Ips Ips
AS 800 200 +25.54 0.0040 +3.20 0.1253 -0.46 +25.08
Be 400 150 +13.54 0.0050 +2.00 0.1477 - 0.46 +13.08
I AD 400 150 - 19.46 0.0097 - 3.88 0.1994 - 0.46 - 19.92
DC 800 150 - 5.57 0.0010 - 0.80 0.1436 - 0.46- - 6.1 4
- 0.11-

Sum : +0.52 0.6160

Correction 6. = - I H/( xIHIO) = - (+0.52 )/ (1.85 x 0.6160) = - 0.46
+0.46*
DC 800 150 +5.57 0.0010 +QED 0.1436 + 0.11 - + 6.14
II CF 400 100 +7.1 1 0.0109 + 4.33 0.6132 + 0.11 +7.22
DE 400 150 - 9.89 0.0028 - 1.12 0.1133 + 0.11 - 9.78
EF 800 100 -4.89 0.0055 -440 0.8998 + 0.11 - 4.78

Sum : - 0.36 1.7699

Correction 6. = - I H/( xIHIO) = - ( - 0.36 )/ (1.85 x 1.7699) = + 0.11

*Common pipes require correction fro m both the circuits

404 PART III: WATER SUPPLY

The corrected flow after the second trial IS Q2. More trials will gIve more
accurate flows in the pipe network.

Question
l. What are the main purposes of transmission and distribution systems? What are the
different water supply systems? Write the advantages and disadvantages of each water
transmission and distribution system.

2. What are the main types of distribution networks? What are their relative advantages
and disadvantages?

3. How can water demand of a community be estimated? Mention the water consumption
data and the corresponding recommended peak factor for the design of water supply
systems in Bangladesh.

4. Calculate the peak water demand for a design period of 10 years of a rural village having
a present population of 1,500. The average per capita water consumption is 50 lpcd
with a peak factor of 3, the population growth rate is 1.5% and the loss and wastage is
30%.

5. State systematically the process of design of branched as well as looped network of

distribution systems.

6. How can you calculate the flow in a looped network using the Hardy Cross method?

7. Calculate the flow in the following pipe network for the inflow and outflow shown in
the diagram using the Hardy Cross method.

30 __-.~_______4_0_0_m_-_1_50__m_m_d_ia________-r_l~
Ips Ips
150 m 100 mm dia.

- - t - - " 8Ips

150 m: 100mm dia.

6 Ips
 CHAPTER 20: WATER TRANSMISSION AND DISTRIBUTION 405

8. Find out the approximate flow in the pipes in the network shown below:

50 Ips

5 Ips ...
400mm - 200mm dia • 300mm - 150mm dia

cti
-.10 Ips

ctl cti is
is is E
E E E
E E C>
C>
C> C>
U")
U")
I
I E
E E E
C> C> C>
C> C> C>
co co co

•
20 Ips+- ---.5 Ips
400m - 150mm dia 300m - 100mm dia
10 Ips

9. What are the different types of service connections? Draw a neat sketch of a yard
connection.