Water Supply Distribution Systems
Introduction
Prof A Taigbenu
School of Civil & Environmental Engineering
University of the Witwatersrand
Water Distribution systems
Water supply distribution systems are pressurized interconnected pipes and
associated appurtenances which supply water of the required quantity, quality
and pressure to consumers.
They have been vital systems for the public health and development of many
early civilizations, and they play the same role in urban and rural
communities in our current civilization.
In South Africa, the planning and design guidelines for these systems are
discussed in Section J of The Neighbourhood Planning and Design Guide
(Red Book): Creating Sustainable Human Settlements, July 2019.
2
Water Distribution systems (II)
With a good design of the distribution system, the network of interconnected
pipes, storage tanks, pumps and regulating valves is capable of supplying
adequate quantity of water of suitable quality and pressure throughout the
network.
The water quality is considered adequate when the water is fit for the intended
purpose by meeting the required quality standards relevant to that specific use
- domestic use (potable and non-potable), industrial, mining, commercial,
agricultural purposes, or the maintenance of ecosystems.
Ideally, the reliability of water supply service (adequate quantity, quality and
pressure) should be 100%. But this is not always possible because
interruptions are inevitable but they should last no longer than an aggregated
48 hours over one year period. This can be achieved when the infrastructure
is well designed, constructed and maintained. “Reliable water services may
increase consumers’ willingness to pay for services and subsequently may
improve the revenue base of the water supply service provider.” (Red Book,
2019)
3
Water Demand Estimation
Estimation of the water demand is one of the critical initial steps to carry out
in the design of water distribution infrastructure. In this regard, reference is
made to the Red Book (2019) which provides various methods of estimating
the demand on the basis of landuse type, number and sizes of stands,
population density, etc.
It is advisable to design for the water demand at the end of the economic life
of the project or to design for phased increase in demand where that is
possible.
Water demand should take into account the various categories of consumers –
domestic, industrial, commercial, public and private institutions, etc.
Each consumption category has its unique consumption pattern that varies
hourly, daily, weekly and seasonally. For instance, an area with high tourism
activities will experience peak consumption during the tourism season. While
a predominantly residential area will have a consumption pattern during the
working week that is different from the weekend.
4
Demand Patterns
Diurnal curves are patterns which relate to the changes in demand over the course
of the day, reflecting times when people use water more than or less than the
average demand.
The demand at any time can be expressed as:
Qt = ft × Qaverage
The typical diurnal curve for a residential area during the work week is below. The
demand variation is met by a combination of operation of reservoirs and pumps.
5
Water Supply Infrastructure
The water supply infrastructure comprises pipes, joints, fittings, valves,
reservoirs and pumps. The complexity of the infrastructure depends on various
factors such as the landuse, extent of the development, number of households,
population density, etc.
The pipes in the infrastructure can be classified as:
Trunk main carries water from the water treatment plant to the distribution
area of the demand. The pipes are usually the largest in the water supply
infrastructure.
Secondary mains serve the basic skeleton of the distribution system. The
secondary pipes link the pumps, reservoirs and other main components of the
distribution system, thereby ensuring its smooth operation, including under
emergency conditions like fire and major pipe burst. Only large consumers are
connected to the secondary pipes.
Distribution mains convey water from the secondary mains to the consumers.
6
Types of Distribution schemes
Gravity scheme makes use of the topography to distribute water to
consumers.
Advantages: no energy costs, simple operation, low maintenance cost,
marginal pressure variability.
Disadvantages: limit on future expansion of scheme; requires large pipe sizes
to minimize head losses so the water can be available to consumers farther
away.
Direct pumping scheme: Distribution system is driven by pumps that have
to be well operated.
Advantages: Can meet variable demands at required pressures, appropriate
for future expansion of scheme;
Disadvantages: high operating and maintenance costs; more frequent
occurrence of pipe bursts.
Combined scheme uses a combination of pumping stations and demand
balancing reservoirs. It enjoys the advantages of the gravity and direct
pumping schemes.
7
Distribution network configurations
Serial network has a source, end and intermediate nodes that are linked by
pipes. Advantages: suited for very small (rural) distribution; cheap.
Disadvantages: Low reliability with the whole system being shut down when
there is a burst pipe; water quality problems because of dead end.
8
Distribution network configurations (II)
Branched network could have multiples sources. It has many ends, which
branch out. Advantage: suitable for small communities; Disadvantages: low
reliability; water quality problems from the dead ends; demand variability
manifests as high pressure oscillations in the network.
9
Distribution network configurations (III)
Loop or Grid network allows flow in different directions to and from a
node. It eliminates the disadvantages of the branch network; it allows for
isolation of a part of the network during maintenance of pipe bursts; demand
variability produces less impact on the pressure fluctuations.
10
Hydraulics of pipe distribution systems
The hydraulic analysis of pipe distribution systems is generally easier to carry
out compared to open channel hydraulics primarily because the traverse
dimension of the flow for the former is known (the pipe size). For open
channel hydraulics, the transverse dimension is the water depth which is an
unknown that has to be calculated.
TEL: Total Energy line given
by: p V2
z
2g
Elevation head + pressure
head + dynamic or velocity
head
p
HGL: z
hL is the frictional loss over
the pipe length, L.
11
Energy Losses
Energy losses in pressurized pipe systems broadly consist of frictional losses
along the length of the pipe and localized or minor losses which occur in the
vicinity of valves, fittings, junctions, sudden and gradual expansions and
contractions, etc.
The frictional loss is accounted for by either the Darcy-Weishbach (D-W)
equation or the Hazen-Williams (H-W) equation.
L V2
The Darcy-Weishbach equation: hL f
D 2g
Where f is the friction factor. For a circular pipe, the D-W equation can be
expressed as
8 fL
hL KQ 2
where K
2 D5
The Hazen-Williams is hL CQ n
Where C is the H-W coefficient and the exponent n lies between 1.85 and 2.
12
Expressions for the friction factor
The friction factor, f, which depends on the Reynolds number and the pipe
roughness characteristics, can be obtained from the Moody diagram which
is found in many Hydraulics texts. However, in this era of computer
modelling, the Moody chart has fallen out of vogue.
Expressions for the friction factor, obtained from extensive experimental
investigations, are presented for various pipe flow regimes.
By laminar flow, we mean that the streamlines of the flow are aligned with
the predominant flow direction as the transverse flow of mass and
momentum is at molecular scale and is unable to upset the predominant
flow direction.
By turbulent flow, we mean that the transport of mass and momentum in
the transverse direction is at a scale comparable to the transverse length of
the flow (the pipe diameter, D), such that the streamlines in the predominant
flow direction can no longer maintain their orderly form, giving rise to
eddies which constantly mix the flow in a chaotic fashion.
13
Expressions for the friction factor (II)
The pipe flow regimes are captured by the dimensionless parameter, known
as the Reynolds number, Re, which is the ratio of the inertial force to the
frictional force.
When the flow is laminar, Re ≤ 2000, where Re = VD/ν and ν is the kinematic
viscosity [m2/s], the expression for the friction factor is
64
f
Re
When the flow is turbulent, Re ≥ 4000. The flow regime when the Reynolds
number lies between 2000 and 4000 is referred to as transitional flow regime.
The expressions for the friction when the flow is turbulent have come from
the experimental work of Nikurade on artificially roughened pipes and of
Blasius.
Blasius’ work: valid of Re < 105, 0.316
f
Re0.25
14
Expressions for the friction factor (III)
For turbulent flow, it is expected that the friction factor will depend on the
Reynolds number and the roughness of the pipe, ε. If the viscous sublayer, on
the inner circumference of the pipe wall where the velocity is small, has a
thickness δ>>ε, then the turbulent flow regime is said to be the smooth
turbulent flow. This is because the main turbulent body of flow has the
viscous sublayer as its boundary. The expression for the friction factor is
given by the von karman – Prandtl equation
1 Re f
2 log10
2.51
f
15
Expressions for the friction factor (IV)
With increased turbulence, the viscous sublayer thins out, and the pipe
roughness becomes exposed to the turbulent region. This is the rough turbulent
flow regime.
Between the smooth and rough turbulent flow regime, there is a transitional
turbulent flow regime.
The Colebrook-White equation applies to the entire turbulent flow regime and
it is given by:
1 2.51
2 log10
3.7 D Re f
f
One disadvantage with the Colebrook-White equation is that it is implicit. An
explicit equation that gives comparable values for f is the Barr (1975) equation
1 5.13
2 log10 0.89
f 3.7 D Re
The von karman – Prandtl equation for the rough turbulent flow regime is
1 3.7 D
2 log10
f
16
Minor Losses
Minor or local losses are caused by localized turbulence due to installed
valves, fittings, bends, reducers, etc in the water distribution system.
Although the effect of the turbulence is spread over a short distance, the minor
losses are for simplicity taken to occur at a section of the pipe where the
device is located.
A few different forms of expressions are used to depict the minor losses, one
of which is
hm K mQ 2
Where Km is the minor loss coefficient. For long, extensive distribution
systems, the minor losses are usually neglected. Another way to account for
these losses is to use the concept of equivalent pipe lengths.
Self study
The experimental investigation on the pipe head loss equipment will enable
you to obtained the friction factor of the pipes and minor loss coefficients of
some fittings. The experimental values should be compared with the values in
the literature.
17
End of Presentation
18