ELECTRONIC OFFPRINT

Use of this pdf is subject to the terms described below

This paper was originally published by IWA Publishing. The author’s right to reuse
and post their work published by IWA Publishing is defined by IWA Publishing’s
copyright policy.

If the copyright has been transferred to IWA Publishing, the publisher recognizes the
retention of the right by the author(s) to photocopy or make single electronic copies
of the paper for their own personal use, including for their own classroom use, or the
personal use of colleagues, provided the copies are not offered for sale and are not
distributed in a systematic way outside of their employing institution. Please note
that you are not permitted to post the IWA Publishing PDF version of your
paper on your own website or your institution’s website or repository.

If the paper has been published “Open Access”, the terms of its use and distribution
are defined by the Creative Commons licence selected by the author.

Full details can be found here: http://iwaponline.com/content/rights-permissions

Please direct any queries regarding use or permissions to [email protected]

 287 © IWA Publishing 2017 Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Evaluating the pressure-leakage behaviour of leaks in

water pipes
J. E. van Zyl and R. Malde

ABSTRACT

Much progress has been made over the last decade in understanding the behaviour of flow through J. E. van Zyl (corresponding author)
Department of Civil Engineering,
leak openings with changes in water mains pressure. In particular it has been established that University of Cape Town,
Cape Town,
variations in leak areas with pressure is the main factor responsible for the range of leakage South Africa
E-mail: [email protected]
exponents observed in practice, and several numerical and experimental studies have investigated
R. Malde
this behaviour. This paper provides an overview of the advances in leakage modelling over the last Weld-Con Ltd,
Mombasa,
decade and then presents the results of a new experimental study of various leak types (round holes
Kenya
and longitudinal, spiral and circumferential cracks) in different pipe materials (unplasticised
polyvinylchloride, modified polyvinylchloride, high density polyethylene and steel). The experimental
results are evaluated in light of the latest theoretical advances and recommendations are made for
further experimental studies.
Key words | Fixed and Variable Area Discharges (FAVAD), leakage, N1, pressure, water losses

INTRODUCTION

The International Water Association (IWA) has been at the of the pressure head differential h over the leak opening as
forefront of water loss management for several decades. described by:
Their efforts were formalised in 1995 with the formation pffiffiffiffiffiffiffiffi
of the Water Loss Task Force, which has made extensive Q ¼ Cd A 2gh (1)
contributions to the field, including the internationally
where Cd is the discharge coefficient, A leak area and g
used IWA Water Balance, Infrastructure Leakage Index
acceleration due to gravity.
and several other benchmarks and guidelines (Lambert
However, since the results of numerous international
).
field studies have shown that the orifice equation often
One of the important activities of the IWA Water Loss
does not fit the measured pressure-leakage response, a
Task Force (now the Water Loss Specialist Group) has
power equation (known as the N1 power equation) has
been to investigate why leak flow rates are often significantly
been adopted by the IWA and has become widely used
more sensitive to pressure than predicted by the orifice
(Gebhardt ; Ogura ; Hiki ; Lambert ;
equation (Lambert , ; Lambert et al. ).
Farley & Trow ; Al-Ghamdi ):
Hydraulically, pipe leaks are orifices and thus can
reasonably be expected to comply with the orifice equation, QL ¼ ChN1 (2)
which is derived from the principle of conservation of
energy. According to the orifice equation, the flow rate Q where QL is the power equation leakage rate, C is the leak-
through a leak should be proportional to the square root age coefficient and N1 the leakage exponent. It should be

doi: 10.2166/aqua.2017.136
 288 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

noted that making the exponent of the leakage equation a The next section provides an overview of the current
variable severs it from its fluid mechanics foundations and understanding of leak area behaviour and its implications for
turns it into a purely empirical equation. Thus the power leakage modelling and management. This is followed by the
equation approach is not standard practice in orifice hydrau- description of a proposed standard experimental procedure
lic theory (Idelchick ; Franchini & Lanza ). and data analysis of different leak types (round holes and longi-
While the value of N1 should be 0.5 to comply with the tudinal, spiral and circumferential cracks) in different pipe
orifice equation, in field studies it has been found to range materials (unplasticised polyvinylchloride (uPVC), modified
between 0.36 and 2.95 (see Schwaller & Van Zyl () for PVC (mPVC), high density polyethylene (HDPE) and steel).
a summary of the ranges for N1 found in different countries).
It is important to understand the causes of the observed
high leakage exponent values as this may allow engineers to LEAK AREA VARIATION AND ITS IMPLICATIONS
better predict the response of systems to changes in pressure, FOR LEAKAGE MODELLING
and ultimately improve the management of leakage. Van Zyl
& Clayton () proposed a number of potential causes of Pressure and pipe wall stresses
the observed range of N1 values, including leak hydraulics,
soil hydraulics and variations in leak area with pressure. Pressure in a water pipe causes stresses to develop in the pipe
Other factors have also been shown to play a role, including walls, resulting in material strain and subsequently changes in
errors in field test assessment of average pressure and leakage the areas of leak openings in pipes. It is possible to derive
estimates, the spatial distribution of leaks (Schwaller & Van theoretical equations for the circumferential and longitudinal
Zyl ; Schwaller et al. ), soils surrounding the pipe stresses in a closed cylindrical container resulting from water
(Walski et al. ; Van Zyl et al. ) and changes in axial pressure (for instance, see Cassa et al. ()). These
momentum (Ferrante et al. a, b). It is now widely equations show that both circumferential and longitudinal
accepted that changes in leak area are the most important stresses are linear functions of the pressure, but that the cir-
causative factor of the observed behaviour. cumferential stresses are double the size of the longitudinal
Greyvenstein & van Zyl () published the results of stresses.
an exploratory experimental study in Journal of Water However, unlike pressure vessels, pipes are not closed at
Supply: Research and Technology – AQUA, showing that their ends and are generally supported by thrust blocks at
the leakage exponents measured in the field are not unrealis- bends and junctions that transfer the longitudinal forces to
tic, and can be reproduced in the laboratory. They found the soil. Thus it is the circumferential rather than longitudi-
leakage exponents between 0.41 and 2.30 and agreed that nal pipe wall stresses that will vary as a result of changes in
variation in leak area is the most likely cause for the devi- pressure.
ation from the orifice equation. It should be noted that several external factors, such as
Significant progress on understanding the behaviour of the weight of soil, external loads, soil movements and ther-
leaks areas in pipes has been made in the last decade. mal expansions also influence the pipe wall stresses.
Thus the aims of this paper are two-fold: to provide a However, unlike the circumferential stresses induced by
review and synthesis of developments in understanding fluid pressure, these stresses are independent of water
leakage behaviour over the last decade and to present the pressure and thus are likely to have a constant impact on
results of a recent experimental study in the light of these leak area variations.
findings. It is hoped that the proposed experimental
method and data analysis will provide a uniform basis for Linearity of the pressure-area relationship
the further experimental work required to establish the
characteristics of the various leak types in different pipe The mechanisms through which pipe material can deform in
materials and sections, including service connections and response to changes in pressure are elastic, viscoelastic and
leaks at joints and fittings. plastic deformation and fracture.
 289 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Over the last decade, several experimental and modelling • The areas of round holes in all materials are stable and
studies have been carried out under elastic deformation con- vary very little with pressure. In practice this means
ditions on a large range of pipe materials, section properties, that their head-area slopes may be assumed to be zero.
leak types and loading conditions. These studies concluded • The head-area slopes of all types of leaks in steel pipes are
that all leak areas vary as a linear function of pressure (Buckley very small and may also be assumed to be zero. However,
; Cassa & Van Zyl ; Van Zyl & Cassa ). corrosion failures in metal pipes have not been studied in
Plastics such as PVC and PE are viscoelastic materials, detail and further work is required to determine whether
which means that they display both elastic and viscous reduced wall thicknesses due to corrosion will have a sig-
deformation behaviour. Several researchers have investi- nificant effect on the head-area slope.
gated the resultant time-dependent response of leak areas • The head-area slopes of circumferential cracks are gener-
to changes in pressure (Ferrante et al. , a, b; Fer- ally small and often negative, meaning that the crack area
rante ; Massari et al. ; De Marchis et al. ; Fox reduces with increasing pressure (a result of the circum-
et al. a, b). Ssozi et al. () showed that at any ferential stresses elongating the crack, pulling it closed
given time after loading, plastic pipes also have a linear due to Poisson’s ratio effect).
pressure-area relationship, but that the slope of the relation- • Longitudinal cracks have the largest head-area slopes of
ship increases with time until it stabilises at the relaxation all leak types. Cassa & Van Zyl () found that the
time of the material. crack width and Poisson’s ratio of the pipe material had
Unlike the elastic and viscoelastic cases, leak areas in a negligible effect on the head-area slope. They proposed
pipes undergoing plastic (permanent) deformation or frac- an equation for predicting the head-area slope of longi-
ture cannot be assumed to have linear pressure-area tudinal cracks as a function of the pipe diameter d,
relationships. However, both these processes are non-revers- crack length Lc, elasticity modulus of the pipe material
ible and can thus only continue for a limited time before E and pipe wall thickness t (ρ is the density of water
they will either stabilise (i.e. become elastic or viscoelastic) and g acceleration due to gravity):
or result in catastrophic fracture (Buckley ). In addition,
they can only occur when the water pressure is increased 2:93157d0:3379 L4:80 100:5997( log Lc ) ρg
2

(i.e. pipe wall stresses are increased) and not when the m¼ c
1:746
(4)
Et
pressure is decreased. Since zonal pressure management vir-
tually always involves lowering of pressures, the effects of
Van Zyl & Cassa () showed that this equation pro-
plastic deformation and fracture may be ignored in analys-
vides good results based on data from several different
ing pressure management zone data.
laboratory studies.
The observed relationship between pressure and leak
area can now be described with the following function:
Implications for leak hydraulics

A ¼ A0 þ mh (3)
To understand the impact of the observed linear head-area
relationship on leak hydraulics, Equation (3) is replaced in
where A0 is the initial area (the area of the leak opening at
Equation (1) to obtain:
zero head differential) and m the head-area slope.
The linearity of the pressure-leakage relationship means pffiffiffiffiffi
that it is only necessary to know a leak opening’s initial area Q ¼ Cd 2g(A0 h0:5 þ mh1:5 ) (5)
and head-area slope to fully characterise its area and thus its
hydraulic behaviour using Equation (1). The following The form of this equation was earlier proposed by Ledo-
points summarise the main findings on the head-area slope chowski () and particularly May () whose paper
of pipe leaks (Greyvenstein & van Zyl ; Cassa et al. became highly influential as the FAVAD (Fixed and Variable
; Cassa & van Zyl ; Van Zyl & Cassa ): Area Discharges) equation.
 290 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

The first term of Equation (5) is the orifice equation and pressure range, but may result in significant errors if
describes the flow through a fixed initial area of the leak. used to extrapolate beyond this range.
The second term in the equation describes the flow through • The leakage exponent of a system (with a given set of
the expanded area of the leak. leaks) is not constant, but varies with system pressure.
It should be noted that while the discharge coefficient Higher pressures in the same system will result in
Cd is an unknown in Equation (5), it can be eliminated by higher N1 values, while lower pressures will result in
combining it with the initial area and head-area slope. In lower N1 values.
this
0
arrangement, A0 is called the effective area • While the head-area slope is not affected by the width of a
(A ¼ Cd A), A00 the effective initial area (A00 ¼ Cd A0 ) and crack, the leakage number (Equation (8)) is significantly
0 0
m the effective head-area slope (m ¼ Cd m). Equations (3) affected due to the change in the initial area. From
and (5) now become: Equation (9) it can now be shown that the same crack
will have substantially higher leakage exponents at smal-
A0 ¼ A00 þ m0 h (6) ler crack widths. The implication of this observation is
that the leakage exponents determined in laboratory
and tests, where a slit is normally machined into a pipe, will
tend to underestimate the leakage exponent of the same
pffiffiffiffiffi 0 0:5
Q¼ 2g(A0 h þ m0 h1:5 ) (7) crack length that forms in the field without removal of
pipe material (and thus with a smaller width).

Van Zyl & Cassa () proposed a dimensionless leak-

• Finally, it can be shown from Equations (6)–(9) that leak
openings with negligibly small head area slopes (i.e. fixed
age number LN defined as the ratio of the flow through the
leak openings) will result in N1 ¼ 0.5 and leak openings
expanded to the initial leak areas, and this is given by:
with initial areas close to zero (such as cracks closing
mh under zero pressure conditions) will result in N1 ¼ 1.5.
LN ¼ (8)
A0
Implications for DMA leakage modelling
They then showed that there is a direct relationship
between N1 and the leakage number described by the
Leakage in water distribution systems is normally assessed
equation:
at DMA (district metered area) level through minimum
night flow and water balance analyses. While larger leaks
1:5LN þ 0:5
N1 ¼ (9) can mostly be detected and repaired, little is known about
LN þ 1
the number and distribution of smaller background leaks.
The behaviour of systems with many leaks adhering to
While Equation (5) or (7) seems to predict a leakage
the power leakage equation was studied by Ferrante et al.
exponent between 0.5 and 1.5, it can be seen from Equation
(), who found that the DMA leakage exponent is
(9) that the leakage exponent can adopt a wider range. In
larger than the mean individual leakage exponent due to
particular, the leakage exponent will approach infinity
the spatial variability of leaks. Schwaller & Van Zyl ()
when the leakage number approaches minus one.
found that individual leaks adhering to the FAVAD equation
The implications for leakage modelling can be sum-
can result in a wide range of DMA leakage exponents, simi-
marised as follows (Van Zyl & Cassa ; Van Zyl et al.
lar to that observed in practice.
):
The linearity of the leak area-pressure relationship has
important implications for the understanding of DMA leak-
• The N1 power equation is an empirical equation not age behaviour. In particular it means that the total DMA
based on fundamental fluid mechanics principles. It can leakage, which is the sum of the individual leak behaviours,
provide reliable results when used within its calibration will also be a linear function of average zone pressure and
 291 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

that Equations (3) and (5)–(8) can thus be applied to DMAs EXPERIMENTAL STUDY
with many leaks.
In practice this means that if the DMA leakage at the Introduction
time of steady minimum night flow is known at two different
Average Zone Night Pressures, which are currently used to An experimental study was conducted to investigate the be-
determine N1 using Equation (2), the same data may be used haviour of various types of leaks (round holes and
in Equation (7) to estimate the sum of effective initial areas longitudinal, spiral and circumferential cracks) in 100 mm
and head area slopes for all the leaks – both detectable and nominal diameter pipes of different materials predominantly
non-detectable – in the DMA. used for distribution mains. The pipe materials that were
The resulting DMA initial area provides an estimate for included in the study were uPVC, mPVC, HDPE and steel.
the total leak area under zero pressure conditions and the mPVC is a pipe material designed to have additional duct-
DMA effective head-area slope can be used in combination ility and more stable long-term characteristics than uPVC
with the known values for different leak types to estimate (DPI ).
the dominant leak type in the DMA. In addition the N1 for The properties and dimensions of the leaks and pipes
the DMA can be estimated at different pressures from its used in this study are summarised in Table 1. The elasticity
initial area and head-area slope using Equations (8) and (9). modulus of the materials were not measured, but estimated
Schwaller et al. () used a spreadsheet model with from literature. All cracks had a width of 1 mm.
stochastic leak distributions to confirm that the initial area
and head-area slope of a DMA provide good estimates of Experimental setup and procedure
the sum of the initial areas and head-area slopes of individ-
ual leaks in the system respectively. Unrealistic values of the The main component of the experimental setup is shown in
system leakage parameters may be used to diagnose poten- Figure 1 and consisted of two removable end sections fitted
tial problems, such as measurement errors or leaking to a sample pipe section with a failure using flexible
boundary valves (this is already the case in the expert soft- couplings.
ware PresCalcs (Lambert )). The system was held together with three 20 mm diam-
Finally, Kabaasha et al. () incorporated the FAVAD eter stainless steel rods secured to the end sections. The
approach into hydraulic network models and showed that steel rods took up the longitudinal forces exerted by the
significant errors in the leakage demand at nodes can be water pressure and thus prevented longitudinal stresses
made when the current power equation approach is used. being induced in the sample pipe walls. Test sections were
The next section of the paper reports on the results of an all 800 mm long, leaving a minimum distance of 350 mm
experimental study of the pressure-leakage behaviour of between the leak opening and pipe section end.
pipe leaks, and analyses these results in light of the theory One end section was connected to a pumped water
discussed in this section. supply from an underground sump through a 25 mm

Table 1 | Properties and dimensions of the leaks and pipes investigated in this study

Main dimensions of leaks tested (mm)

Cracks
Internal diameter Wall thickness Elasticity modulus
Pipe material (mm) (mm) E (MPa) Round holes Longitudinal Spiral Circumferential

uPVC 105.5 4.54 3 12 50, 100 50, 100 50, 100

mPVC 106.6 3.44 3 12 50 50 50
HDPE 103.6 6.38 1 12 73, 100 50, 78 54, 80
Steel 105.1 4.92 203 12 50, 100 50, 105 53
 292 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Figure 1 | Schematic layout and photograph of the main components of the experimental setup.

calibrated magnetic flow meter with an accuracy of ±0.5%. and 1,000 readings of pressure and flow rate. A typical
The downstream end was fitted with a calibrated pressure raw data set is shown in Figure 2.
transducer with a working range of 0–20 bar and accuracy Experimental data points were obtained by identifying
of ±0.5%. Water was supplied by a variable speed submers- stable sections of the flow and pressure graphs and then
ible centrifugal pump capable of supplying a flow of 5 L/s at taking the average values over each of the ranges. These
a head of 100 m. (typically 25) average values were then plotted and analysed
At the start of an experimental run, a small flow was to determine the leakage characteristics for each
introduced and the setup tilted towards an opening on the experiment.
downstream end to remove all trapped air. The setup was It was observed that the first raising leg of tests on cracks
then placed horizontally with the pressure transducer and in plastic pipes were often distinctly different from the rest of
leak on the same level. The leak discharged into the atmos- the data, possibly due to initial plastic deformation before
phere and no flow existed in the system apart from the the leak opening stabilised. In such cases the first few
leakage. (between one and six) data points were omitted from the
Flow and pressure were increased and then decreased in results.
about five steps by varying the pump speed. Each step lasted
about 30 seconds and was long enough to ensure that both
flow and pressure readings stabilised. This procedure was RESULTS
repeated three times in succession before downloading
and analysing the logged data. Data were collected at 1 The results of the experiments are summarised in Table 2.
second intervals and generally consisted of between 900 The analyses conducted to obtain these results are first
 293 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Figure 2 | Logged flow and pressure data for a typical experimental run.

explained, followed by a discussion of significant obser- parameter are given by:

vations and their implications for leakage modelling.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The leakage exponent N1 (column 3 in Table 2) for each
β^i ± tn2,0:025 s (X0 X)1 ii
experiment was obtained by fitting a power curve to the flow
against head data as shown for round holes in Figure 3.
where tn2,0:025 is the upper 2.5th percentile of a t-distri-
The effective leak area at each pressure was then calcu-
bution with n  2 degrees of freedom.
lated from Equation (1) and plotted against the pressure
However, since neither parameter can be assumed
head as shown for longitudinal slits in Figure 4.
known in many cases, a more realistic estimate is obtained
A linear model was fitted to the data points for each
by considering the uncertainties in the two parameter
experiment to determine the model parameters and the
values simultaneously, resulting in a two-dimensional 95%
95% simultaneous confidence intervals for each parameter.
confidence region. This region for the simultaneous confi-
The statistical inference theory used to obtain the confi-
dence limits for β may be determined using the equation:
dence intervals can be obtained from standard statistics
text books such as Johnson & Wichern (). The model
{βj(β^  β)0 (X0 X)(β^  β)  2s2 F2,n2,0:05 }
specifies a linear relationship of the form (compare to
Equation (6)):
where F2,n2,0:05 is the upper 5th percentile of an F-distri-
A0i ¼ A00 þ m0i hi þ εi bution with 2 and n  2 degrees of freedom; X is the
n-by-2 design matrix (first column contains 1 s, second
where hi and A0i are the pressure head and effective area column contains the hi values); and s2 is the estimated
respectively for observation i (i ¼ 1,2,…, n), and εi , which value of σ 2 , and equals the sum of squared residuals (differ-
captures noise, follows a normal distribution with a mean ences between observed A0i values and fitted expected
of zero and variance σ 2 , and is independently drawn for values) divided by n2.
each observation. The simultaneous confidence interval widths for the
The intercept and slope of the linear relationship, con- experiments were found to be on average 27% (varying
^0 m
tained in β ¼ [A0 m0 ] , are estimated as β^ ¼ [A
0
b 0 ]0 so as to between 24.6 and 29.3%) larger than the single parameter
0 0
maximise the likelihood function. confidence intervals.
Assuming that one model parameter is known, the stan- The effective initial area and its 95% simultaneous con-
dard confidence interval limits for the other model fidence interval are given in columns 4 and 5 of Table 2
 294 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Table 2 | Summary of experimental results

Pipe material Length/diameter (mm) N1 A00 (mm2) 95% SCI for A00 (mm) m0 (mm2/m) 95% SCI for m0 (mm2/m) p for m0 (%) Cd

Round holes
uPVC 12 0.499 68.3 ±0.11 0.00141 ±0.00230 12.5 0.603
mPVC 12 0.500 68.3 ±0.17 0.00021 ±0.00360 87.9 0.604
HDPE 12 0.501 70.3 ±0.16 0.00180 ±0.00345 4.5 0.621
Steel 12 0.497 67.4 ±0.18 0.00486 ±0.00365 0.2 0.596
Longitudinal slits (width ¼ 1 mm)
uPVC 50 0.886 27.4 ±2.08 0.28658 ±0.0193 <0.1 0.547
uPVC 100 1.041 52.9 ±2.59 2.51200 ±0.129 <0.1 0.529
mPVC 50 0.989 37.5 ±6.00 0.86711 ±0.139 <0.1 0.749
HDPE 73 0.835 62.8 ±11.7 0.97499 ±0.348 <0.1 0.860
HDPE 100 0.798 94.5 ±12.0 2.03031 ±0.578 <0.1 0.945
Steel 50 0.500 22.0 ±0.22 0.00011 ±0.00275 10.3 0.440
Steel 100 0.529 44.7 ±0.13 0.02166 ±0.00205 <0.1 0.447
Spiral slits (width ¼ 1 mm)
uPVC 50 0.657 36.6 ±0.53 0.10802 ±0.00800 <0.1 0.732
uPVC 100 0.797 78.4 ±1.31 1.24352 ±0.0476 <0.1 0.784
mPVC 50 0.798 43.2 ±1.47 0.37375 ±0.0288 <0.1 0.865
HDPE 50 0.656 33.0 ±3.08 0.09377 ±0.0451 <0.1 0.660
HDPE 78 0.703 69.0 ±4.68 0.48148 ±0.123 <0.1 0.884
Steel 50 0.490 22.8 ±0.08 0.00267 ±0.00095 6.4 0.457
Steel 105 0.523 44.2 ±0.17 0.01615 ±0.00265 1.6 0.421
Circumferential slits (width ¼ 1 mm)
uPVC 50 0.455 34.8 ±0.26 0.02170 ±0.00380 <0.1 0.696
uPVC 100 0.327 50.2 ±1.28 0.11423 ±0.0189 <0.1 0.502
mPVC 50 0.433 37.2 ±0.27 0.03403 ±0.00380 <0.1 0.743
HDPE 54 0.185 20.6 ±1.74 0.06212 ±0.0213 <0.1 0.382
HDPE 80 0.262 37.4 ±5.14 0.20706 ±0.0646 <0.1 0.467
Steel 53 0.499 22.6 ±0.37 0.00008 ±0.00485 33.3 0.427

SCI, simultaneous confidence interval.

respectively. Similarly, the effective head-area slope and its area of the leak openings can be estimated from the
95% simultaneous confidence interval are given in columns pipe samples, it is possible to estimate the discharge coeffi-
6 and 7. cient by dividing the effective initial area by the actual
The p-value in column 8 of Table 2 is the probability of initial area.
observing the effective head-area slope or a more extreme
estimate, where the true value of the head-area slope is
equal to zero (null hypothesis: m0i ¼ 0). A small p-value pro- DISCUSSION
vides evidence against this null.
The final column of Table 2 gives an estimate of the The results in Table 2 and Figure 1 show that the areas
discharge coefficient of the leak openings. Since the initial of 12 mm round holes varied very little with pressure
 295 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Figure 3 | Flow rate against pressure head for 12 mm diameter round holes in different pipe materials.

for all pipe materials tested. Even at the largest absolute and thus to adhere to the orifice equation. For experimental
95% interval, the effective head-area slopes were all studies, 12 mm round holes can be used as a benchmark for
below 0.01 mm2/m. All the measured leakage exponent the accuracy and consistency of the results.
values were equal to the theoretical value of 0.50 at two Steel pipes also displayed small head-area slopes with
significant digits. leakage exponents at or close to 0.50 for the range of
This means that in practice 12 mm round holes (and leaks tested. Longer longitudinal and spiral slits displayed
likely a wider range of diameters) can be assumed fixed larger head-area slopes than shorter slit lengths (the same

Figure 4 | Effective leak area against pressure head for longitudinal slits.
 296 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

trend was observed for other pipe materials tested). These It has been shown that the leakage exponents of most
results support the practice of assuming N1 values of 0.5 leaks are not fixed, but vary with fluid pressure (Van Zyl
for networks with steel or cast iron pipes. However, it & Cassa ). While the experimental study only estimates
should be noted that pipes with extensive corrosion a single overall N1 for each experiment, it is possible to cal-
damage may have larger N1 values as shown in a limited culate the actual range of N1 values for each experiment
study using very low pressures by Greyvenstein & van Zyl from the effective initial area, head area slope and the
(). range of experimental test pressures. This is done by first cal-
All circumferential slits tested were found to display culating the leakage number for each data point using
negative head-area slopes, meaning that the areas of these Equation (8) and then the leakage exponent at this pressure
slits decreased with increasing pressure resulting in leakage using Equation (9).
exponents below 0.5. Longer slits displayed more negative The range of leakage exponent values thus determined
head-area slopes than shorter slits. The 80 mm long circum- for each experiment is shown against the single value
ferential slit in HDPE had the most negative head-area slope obtained by fitting a power equation to the data in Figure 6.
and a negative N1 value, which means that not only the slit The figure shows significant variation in the actual leakage
area, but also the flow rate through the slit decreased with exponent, except at the value of 0.5 where both methods
increasing pressure. give only theoretical value. The 80 mm circumferential slit
The head-area slopes of longitudinal slits were found to in an HDPE pipe has a particularly large leakage exponent
be the largest of all the leak types tested, followed by spiral range, which can be explained by its leakage number range
slits, particularly for longer slit lengths. Equation (4) was (0.28 to 0.63) starting to approach the asymptote at
used to predict the head-area slopes for the longitudinal minus one in Equation (9).
slits tested based on the values in Table 1 and assuming a Finally, the flow predictions of the N1 power and
discharge coefficient of 0.6 (Fox et al. a) to calculate FAVAD equations are compared for 100 mm longitudinal
the effective head-area slope. The predicted effective head- slits in uPVC, HDPE and steel for a pressure range of zero
area slopes are compared with the experimental values in to 100 m in Figure 7. The figure shows that both models
Figure 5 showing good performance of the equation except fit the data well, but that there are large differences between
for the 50 mm slit in a steel pipe, which for practical pur- the two equations for the uPVC and HDPE pipes outside the
poses had a fixed leak opening. measured pressure range.

Figure 5 | Comparison of the effective head area slope predicted using Equation (4) and a discharge coefficient of 0.6 with the experimentally determined values.
 297 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Figure 6 | The range of N1 values for each experiment against the single N1 determined experimentally.

The FAVAD equation is based on a fundamental not allow data to be collected at higher pressures,
fluid mechanics theory, incorporating the linear head- and thus further work is required to verify this
area slope demonstrated in Figure 4. Thus it can be assumption.
assumed to describe the true behaviour, while the Both equations performed well on the steel pipe due to
empirical N1 equation results in significant modelling the practically zero head-area slope, and thus pure orifice
errors outside the measured pressure range. The capacity flow that is equally well described with the FAVAD and
of the pumping system used in the experiment did N1 power equations.

Figure 7 | Comparison of the N1 power and FAVAD equation flow predictions for 100 mm slits in uPVC, HDPE and steel pipes.
 298 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

CONCLUSIONS Cassa, A. M. & Van Zyl, J. E.  Predicting the pressure-leakage

slope of cracks in pipes subject to elastic deformations.
J. Water Supply Res. Technol. AQUA 62 (4), 214–223.
The aims of this paper were to provide a review of the devel- Cassa, A. M., Van Zyl, J. E. & Laubscher, R. F.  A numerical
opments in understanding the behaviour of leaks over the investigation into the effect of pressure on holes and cracks in
last decade and use these to interpret the results of a water supply pipes. Urban Water J. 7 (2), 109–120.
De Marchis, M., Fontanazza, C. M., Freni, G., Notaro, V. & Puleo,
recent experimental study on the pressure-leakage behav-
V.  Experimental evidence of leaks in elastic pipes. Water
iour of various leak types in different pipe materials. Resour. Manage. 30, 2005–2019.
Recent research has established that the areas of all DPI  Design Guidelines for PVC Pressure Pipe Systems for
leaks types vary as linear functions of pressure under both Trenched Civil Applications. DPI Plastics. Available from:
www.dpiplastics.co.za/Documents/PVC%20Pipeline%
elastic and viscoelastic conditions. This behaviour can be
20Design%20Guidelines%20Manual.pdf (accessed 8
characterised by determining the initial area and head-area December 2016).
slope of a leak opening. Farley, M. & Trow, S.  Losses in Water Distribution Network.
IWA Publishing, London.
It has been established and confirmed by this study that
Ferrante, M.  Experimental investigation of the effects of pipe
round holes have very small head-area slopes that can nor- material on the leak head-discharge relationship. J. Hydr.
mally be assumed as zero, circumferential slits have small Eng. ASCE 138 (8), 736–743.
head-area slopes that are often negative and longitudinal Ferrante, M., Massari, C., Brunone, B. & Meniconi, S. 
Experimental evidence of hysteresis in the head-discharge
slits have the largest head-area slopes. An equation for pre-
relationship for a leak in a polyethylene pipe. J. Hydr. Eng.
dicting the head-area slope of longitudinal slits (Cassa & 137 (7), 775–780.
Van Zyl ) was shown to perform reasonably well in Ferrante, M., Massari, C., Brunone, B. & Meniconi, S. a Leak
this study. behaviour in pressurized PVC pipes. Water Sci. Technol.
Water Supply 13 (4), 987–992.
Finally, the study showed that the N1 power equation
Ferrante, M., Todini, E., Massari, C., Brunone, B. & Meniconi, S.
produces good results when used within its calibrated b Experimental investigation of leak hydraulics.
range, but can result in significant errors when used outside J. Hydroinform. 15 (3), 666–675.
Ferrante, M., Meniconi, S. & Brunone, B.  Local and global
this range.
leak laws. Water Resour. Manage. 28 (11), 3761–3782.
Fox, S., Collins, R. & Boxall, J. a Experimental study exploring
the interaction of structural and leakage dynamics. J. Hydr.
Eng. 143 (2), 1–12.
ACKNOWLEDGEMENTS Fox, S., Collins, R. & Boxall, J. b Physical investigation into the
significance of ground conditions on dynamic leakage
The authors would like to acknowledge the assistance of behaviour. J. Water Supply Res. Technol. AQUA 65 (2), 103–115.
Franchini, M. & Lanza, L.  Leakages in pipes: generalizing
Reshma Kassanjee of the Statistical Consulting Service at
Torricelli’s equation to deal with different elastic materials,
the Department of Statistical Sciences, University of diameters and orifice shape and dimensions. Urban Water J.
Cape Town for the statistical analysis, as well as Allan 11 (8), 678–695.
Gebhardt, D. S.  The effects of pressure on domestic water
Lambert for constructive comments on the manuscript.
supply including observations on the effect of limited water-
gardening restrictions during a period of high demand.
WaterSA 1 (1), 3–8.
Greyvenstein, B. & van Zyl, J. E.  An experimental
REFERENCES
investigation into the pressure-leakage relationship of some
failed water pipes. J. Water Supply Res. Technol. AQUA
Al-Ghamdi, A. S.  Leakage-pressure relationship and leakage 56 (2), 117–124.
detection in intermittent water distribution systems. J. Water Hiki, S.  Relationship between leakage and pressure.
Supply Res. Technol. AQUA 60 (3), 178–183. Jpn Waterworks Assoc. J. 50–54.
Buckley, R. S.  Theoretical Investigation and Experimentation Idelchick, I.  Handbook of Hydraulic Resistance. Begell
into the Expansion of Round Holes and Cracks within House, Redding, CT.
Pressurised Pipes. M.Eng. dissertation, Department of Civil Johnson, R. A. & Wichern, D. W.  Applied Multivariate
Engineering, University of Johannesburg. Statistical Analysis, 6th edn. Pearson, New York, 800 pp.
 299 J. E. van Zyl & R. Malde | Evaluating the pressure-leakage behaviour of leaks in water pipes Journal of Water Supply: Research and Technology—AQUA | 66.5 | 2017

Kabaasha, A. M., Van Zyl, J. E. & Piller, O.  Modelling www.leakssuite.com/wp-content/uploads/2016/10/JOHN-

pressure–leakage response in water distribution systems MAY-SEMINAL-1994-ARTICLE-4.pdf.
considering leak area variation. In: 14th International Ogura, L.  Experiment on the relationship between leakage
Conference on Computing and Control for the Water Industry, and pressure. Jpn Water Works Assoc. 1 (192), 50–54.
CCWI2106, 7–9 November, Amsterdam, The Netherlands. Schwaller, J. & Van Zyl, J. E.  Modeling the pressure-leakage
Lambert, A.  Pressure Management: Leakage Relationships. response of water distribution systems based on individual
Theory, Concepts and Practical Application. In: Paper leak behavior. J. Hydr. Eng. 141 (2), 1–8.
presented at IQPC Seminar, London. Schwaller, J., Van Zyl, J. E. & Kabaasha, A. M.  Characterising the
Lambert, A.  What do we know about pressure:leakage pressure-leakage response of pipe networks using the FAVAD
relationship in distribution system? In: System Approach to equation. Water Sci. Technol. Water Supply 15 (6), 1373–1382.
Leakage Control and Water Distribution Systems Ssozi, E. N., Reddy, B. D. & Van Zyl, J. E.  Numerical investigation
Management, Brno, Czech Republic. of the influence of viscoelastic deformation on the pressure-
Lambert, A.  Water Leakage and Pressure Management: leakage behavior of plastic pipes. J. Hydr. Eng. 142 (3), 1–9.
Welcome. Available from: www.leakssuite.com (accessed 15 Van Zyl, J. E. & Cassa, A. M.  Modeling elastically deforming
December 2016). leaks in water distribution pipes. J. Hydr. Eng. 140 (2), 182–189.
Lambert, A.  Personal communication. Van Zyl, J. E. & Clayton, C. R. I.  The effect of pressure
Lambert, A., Fantozzi, M. & Thornton, J.  Practical on leakage in water distribution systems. Water Manage.
approaches to modeling leakage and pressure management in 160 (WM2), 109–114.
distribution systems – Progress since 2005. In: International Van Zyl, J. E., Alsaydalani, M. O. A., Clayton, C. R. I., Bird, T. &
Conference on Computing and Control for the Water Industry, Dennis, A.  Soil fluidisation outside leaks in water
CCWI2005, Exeter, UK. Available from: www.leakssuite.com/ distribution pipes – preliminary observations. Proceedings of
wp-content/uploads/2012/11/CCWI_Sep2013 paper_Pressure- the Institution of Civil Engineers: Water Manage. 166 (WM10),
burstsALMFJT-1-2003-2013K1.pdf. 546–555.
Ledochowski, W.  An analytic method of locating leaks in Van Zyl, J. E., Lambert, A. O. & Collins, R.  Realistic modeling
pressure pipe-lines. J. S. Afr. Inst. Civil Eng. 6 (12), 341–344. of leakage and intrusion flows through leak openings in
Massari, C., Ferrante, M., Brunone, B. & Meniconi, S.  Is the pipes. J. Hydr. Eng. 143 (9), DOI: 10.1061/(ASCE)HY.1943-
leak head-discharge relationship in polyethylene pipes a 7900.0001346.
bijective function? J. Hydr. Res. 50, 409–417. Walski, T., Bazts, W., Posluszny, E. T., Weir, M. & Whitman, B. E.
May, J. H.  Pressure Dependent Leakage. World Water and  Modeling leakage reduction through pressure control.
Environmental Engineering, October 1994. Available from: J. Am. Water Works Assoc. 98 (4), 147–155.

First received 18 December 2016; accepted in revised form 19 April 2017. Available online 20 May 2017