See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/44055423

Pipe Leakage - Future Challenges and Solutions

Article · January 1999

CITATIONS READS

15 3,078

6 authors, including:

S. Burn
The Commonwealth Scientific and Industrial Research Organisation
149 PUBLICATIONS   1,920 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Desalination in Agriculture View project

Asset Management View project

All content following this page was uploaded by S. Burn on 02 June 2014.

The user has requested enhancement of the downloaded file.

 Pipes Wagga Wagga, Australia 1999

PIPE LEAKAGE – FUTURE CHALLENGES

AND SOLUTIONS
Stewart Burn1, Dhammika DeSilva1, Matthias Eiswirth2, Osama Hunaidi3,
Andrew Speers1 and Julian Thornton4
1
CSIRO Building, Construction & Engineering, Highett, Australia
2
AGK, Karlsruhe Universitat, Germany
3
NRC, Ottawa, Canada
4
BBL Limitada, Sao Paulo, Brazil

ABSTRACT

Pipe leakage in Australia is perceived to be a major problem by many water authorities, both from
an environmental point of view, as well as the associated costs that are incurred due to overdesign
of our sewerage systems (to cope with wet weather loads) and the treatment of additional potable
water that is lost due to leakage. This paper discusses the state of our reticulation systems, design
and environmental concepts to allow us to categorise the problem, methods of leakage detection,
solution concepts to solve leakage and finally methods to allow planning to occur to prioritise asset
management solutions.

Key words: Pipes, leakage, environmental, pressure control, rehabilitation, asset management.

INTRODUCTION

Major advances have been made in recent years in pipeline technology. However, typical sewerage
and water distribution systems in many cities worldwide contain a large percentage of older pipes.
For example, in many German systems over 54% of the pipes are older than 25 years and 24% are
older than 50 years (Fuhrmann n.d.). In Australia, systems are generally newer, and a typical system
has only 47% of pipes that exceed 25 years in age and 13% older than 50 years (D. Santamaria,
South East Water, pers. com.). As pipes age, the problems of infiltration and exfiltration increase in
sewer pipes causing potential environmental problems, although in many cases these problems have
yet to be quantified (Fuhrmann n.d.). Also, in water reticulation pipes the failure levels increase
with age (WSAA Facts 1998) and consequently the levels of unaccounted for water and the
associated lost revenue from water reticulation mains can also increase due to these failures and the
associated water leakage.

In many water distribution systems, a significant percentage of water is lost while in transit from
treatment plants to consumers. According to an inquiry made in 1991 by the International Water
Supply Association (IWSA), the amount of lost or “unaccounted for water” (UFW) is typically in
the range of 20–30% of production (Cheong 1991). In the case of some systems, mostly older ones,
the percentage of lost water could be as high as 50% (AWWA 1987). In Australia and Canada the
problem is not quite as severe due to relatively newer systems. For example, water authorities in
Australia report UFW levels varying between 8 and 28% with the average being 15% in 1997/98
(WSAA Facts 1998).

UFW is usually attributed to several causes including leakage, metering errors and theft. According
to the IWSA survey, however, leakage is the major cause. Water leakage is a costly problem, not
only in terms of wasting a precious natural resource but also in economic terms. The primary
economic loss due to leakage is the cost of raw water, its treatment and transportation. Leakage
 Pipes Wagga Wagga, Australia 1999

inevitably also results in secondary economic loss in the form of damage to the pipe network itself,
e.g. erosion of pipe bedding and major pipe breaks, and in the form of damage to foundations of
roads and buildings. Diminution of supply security as a result of a reduction in water stored per
capita may also represent a cost if such diminution requires augmentation of supply to maintain
security. Besides the environmental and economic losses caused by leakage, leaky pipes create a
public health risk, as every leak is a potential entry point for contaminants if a pressure drop occurs
in the system.

In sewer systems, infiltration and exfiltration raises a number of problems such as increased cost
and environmental impact. This is due to increased sewerage transportation and treatment costs
associated with wet weather water ingress, or potential leaks and overflows and their associated
environmental problems. In Australia wet weather flows are 3–8 times dry weather flows
(D. Santamaria, South East Water, pers. com.). Whilst these increased flows cause additional
operating costs due to increased volumes of sewerage being treated, they can also cause significant
operating problems for sewerage treatment plants, especially in areas with significant salt water
ingress. If the system is not designed to cope with these extra loads (peak day to peak system load
factors in Australia vary from 0.4 to 1.9 (WSAA Facts 1998)), then overflows can occur, with their
associated bad publicity, environmental impacts and health concerns.

CURRENT STATUS OF THE AUSTRALIAN INFRASTRUCTURE

Water Supply System

The supply system for a capital city in Australia typically consists of large supply lines, referred to
as transfer mains, delivering water from a catchment or treatment plant to a network of distribution
mains that deliver to reticulation mains. These mains are generally 900 mm (diameter) and above,
although smaller mains of 300–825 mm pipelines make up the total system. Although these
pipelines may have originally been in cast iron, in nearly all cities they have since been replaced
with mild steel (enamel or cement lined) pipelines. From about 1925, installations were almost
exclusively in lined mild steel. The transfer pipelines supplying Australian cities are in generally
good condition with nearly 90% of the pipes with a residual life up to 50 years or more. Figures 1
and 2 show the typical pipe profiles in a 1000 km system and their projected residual life.

400 36.6%
350 1000
Total Lengtk (km)

Total Length (km)

300 (82.5%)
800
250
200 17.6% 600
150 14.5%
400
10.6% 10.8%
9.96%
100
200
50 (9.8%) (6.5%)
(0.7%) (0.6%)
0 0
600- 900- 1150- 1350- 1700- 2100-& Cond. 1 Cond. 2 Cond. 3 Cond. 4 Cond. 5
825 1000 1300 1600 2000 above (> 50 yrs) (20-50 yrs) (5-20 yrs) (1-5 yrs) (<1 yr)
Size (mm)

Figure 1. Profile of transmission pipe Figure 2. Projected residual life of pipes

sizes in a typical system. in a transmission pipe systems.

In contrast to the transfer mains, the distribution and reticulation systems typically reflect the full
spectrum of pipeline types installed from the very early days, i.e. cast iron, galvanised wrought iron,
asbestos cement, ductile iron, uPVC and polyethylene. High proportions of these pipelines are in
the 100–300 mm size (approx. 80%). However, the size range extends from 50 to 825 mm. The
smaller pipes (50–63 mm) are mainly in cul-de-sacs (courts) and sub-50 mm pipes are generally in
 Pipes Wagga Wagga, Australia 1999

supply lines to properties (commonly referred to as property connections). It is in these pipelines

where small leaks are likely to go undetected, especially in sandy soils. Figure 3 shows the typical
profile of the growth of distribution pipelines from the 1850s to 1997.

The failures encountered in distribution and maintenance pipes are typically described as “bursts”
and not differentiated into joint or pipe failure. High proportions of these failures are in the older
cast iron pipes (WSAA Facts 1998). However, asbestos cement pipes also have high failure rates,
especially for pipes in acidic soils or subject to movement. Figure 4 shows the typical failure profile
in a network.

Sewer System

Vitreous clay was the principal pipe type for sewer systems until the late 1970s. uPVC pipes have
penetrated the market since the 1970s and these pipes are now the principal pipe material used in
sewer systems. Figure 5 shows the typical growth of sewer networks in Australian capital cities.
The predominant problem with sewer systems is thought to be infiltration/exfiltration through
joints, however, although most water authorities have a good idea of their wet weather versus dry
weather flows, generally the relative contribution of the various components, including joints, in the
sewer systems is not known.

Pipeline Lengths by Pipe Type per 1000 km Other

(installed ~1850 to 1997) Polyethylene: HDPE and MDPE
1200
UPVC Thin Walled

UPVC – Thick Walled Pres.Pipe

1000
Ductile Iron Cement Lined (CL)

800 Mild Steel Enamel Lined

Mild Steel CL
km

600 Mild Steel (unlined)

Galvanised Wrought Iron

400
Galvanised Wrought Iron CL

RF Cement
200
Asbestos Cement

0 Cast Iron CL

Grey Cast Iron

59

69

79

89

99

09

19

29

39

49

59

69

79

89

97
18

18

18

18

18

19

19

19

19

19

19

19

19

19

19

Figure 3. Profile of the growth of distribution pipelines in a 100 km distribution network.

80
70
Breaks/100 km*/year

60
50
40
30
20
10
0
*100 km of that AC CI/ CICL DICL GWI MS UPVC PE/ MDPE CU
* 100 km of of that
particular pipe type Pipe m aterial
particular pipe type

Figure 4. Typical failure frequency in pipe types.

 Pipes Wagga Wagga, Australia 1999

1600

1400

PE
1200
PVC

1000 DICL
AC

Km
800 MS
CI
600 RC
CM
400 VC

200

0
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 1997

Years

Figure 5. Profile of the growth of sewer pipelines to service 100,000 properties in 1997.

DESIGN AND ENVIRONMENTAL EFFECTS OF LEAKAGE

Little attention has been given to the environmental aspects of leakage from water mains and the
main focus has been on damage to property and the infrastructure. In the area of environmental
effects, most attention has been applied to sewerage pipes. Characterisation of sewerage system
performance requires recognition of the impacts of leakage into and out of the pipe network. While
overflows do cause environmental impacts, these vary in significance from location to location.
Thus a balanced understanding is required. Of equal significance for sewer system managers and
the customers they serve, may be the cost incurred in designing pipes to take into account leakage,
particularly inflow during heavy weather. Once again, these costs vary according to circumstances.

Leakage occurs at both designed overflow points and from joints and cracks in pipelines. The
purpose of designed overflow structures is to relieve pressure in pipes at controlled locations, i.e.
adjacent to stormwater drains, so that overflows do not occur at private residences or businesses or
environmentally sensitive locations. They function when pipe capacity is exceeded due to
infiltration or blockage during storms. Except in the most poorly designed systems, they do not
function in dry weather when no blockage exists. Cracks, on the other hand, may function at any
time, releasing sewerage into local waterways or soils (exfiltration) or allowing storm-flow to enter
sewers (infiltration), leading to pressure build up and possible overflows during heavy weather.

The environmental impacts of leakage include impacts on ecosystems, aesthetic impacts and human
health risk. These impacts, however, need to be considered in context. The Sydney Water
Corporation has recently completed a series of Environmental Impact Statements (EISs) looking, in
part, at the environmental effect of overflows and sewer system leakage (Sydney Water Corporation
1998). These documents listed the potential environmental effects of overflows as being:
• Eutrophication as a result of nutrient-rich sewage reaching receiving waters.
• Toxicant impacts (especially chlorpyrifos and dieldren (respectively, organophosphate and
organochlorine pesticides) copper and ammonia).
• Faecal coliforms; oxygen reduction in receiving waters (which may lead to fish kills and other
impacts).
• Increased turbidity and increased sediment loads (and litter).

Each of these has potentially serious impacts. Their actual environmental effect depends, however,
on the volume and concentration discharged and the receiving water environment. So too must the
general environmental conditions at the time of discharge be considered (i.e. rain or dry weather).

The wealth of data included in the Sydney Water EISs cannot be reproduced here, but by way of
illustration, Figures 6 and 7 show the relative contribution of faecal coliforms and nutrients of
overflows from Sydney Water’s systems at different locations.
 Pipes Wagga Wagga, Australia 1999

100 100

80 80
Diffuse Diffuse
60 60
Overflow Overflow
STP 40 STP
40
20
20
0
0
Sydney Upper Nepean
Sydney Harbour Upper Nepean
Harbour

Figure 6. Relative distriubtions of phosphorus. Figure 7. Relative contributions of

facecal coliforms.

As will be appreciated, the effect of the various overflow-derived impacts will depend upon the
response of the receiving environment to those pressures, the frequency of the overflow and the
degree of flushing that occurs as a result of storms (which often coincide with overflows). The key
point to stress, however, is that quantification of leakage impacts and, by extension, the degree of
remedial action required will depend on the circumstances and the degree of risk we judge acceptable.
That leakage causes environmental impacts on receiving waters, land (e.g. waterlogging and nutrient
enrichment), recreational amenity, flora and fauna, and air quality is, however, undeniable.

Sewerage systems are designed to cater for wet weather ingress. International best practice allows
for sewer overdesign to cater for such flows. Sewerage systems may be designed to cater for up to
10 times average dry weather flow. It stands to reason, that such overdesign increases the cost of
sewerage systems. Viewed over time, however, it might be suggested that this overdesign represents
underutilisation of an expensive asset (as peak capacity may be utilised only a small percentage of
the time). Such underutilisation is counter to the principles of modern business practice which seeks
to drive assets as hard as possible. Without going into significant detail, it should be noted that pipe
costs rise exponentially with size. CSIRO’s Urban Water Program has gathered extensive data on
pipe and laying costs in different environments. As a generalisation, this shows that 1 m of installed
pipe costs 30% more than a pipe of half the capacity (85% more for pipe only), as shown in Table 1

A reduction in peak pipe flow therefore has the capacity to reduce costs and increase asset
utilisation. Concepts that might lead to such a reduction would include improved pipe materials,
characterisation of the performance of pipes in different laying environments (enabling targeting of
materials and techniques at specific circumstances) and transport of sewage from a common
holding pit under pressure.

As well as environmental impacts and construction/operation costs caused by pipe leakage, an

additional category of costs – “those accruing to third parties” – exists, however, these are generally
not accounted for. These costs arise where there is a cost to a third party, in this case arising from
sewerage system leakage, that is borne by that party rather than by the operator or user of the
sewerage system. Such costs may include diminution of recreational opportunities, loss of income
from, say, oyster farming or fishing, and potentially increased health costs. These costs too must be
considered in the characterisation of sewerage system performance.
Table 1. Class 16 PVC pipe cost and installed cost
ND Approx. internal cross-sectional area Pipe cost Installed cost
(m2) ($) (nominal) ($) (nominal)
100 0.0078 3.77 75.00
140 0.015 7.00 90.00
200 0.031 13.75 113.00
275 0.059 25.00 145.00
375 0.110 45.00 195.00
500 0.196 76.00 265.00
 Pipes Wagga Wagga, Australia 1999

LEAKAGE DETECTION
Economic pressure, concern over public health risk, the need to conserve water and the increased
treatment costs associated with infiltration (White et al. 1997) motivate water system operators to
implement leakage control programs. Leakage control can forewarn asset managers of potential
problems, including the impending collapse of a pipeline, which usually damages adjacent utilities
such as gas and telephone, or damages nearby assets including roadways and buildings. Significant
efforts were made in the past to develop such programs, and as a result procedures for systematic
water loss control programs are now well established and widely used. There are two major steps in
any systematic leakage control program:
• Water audits.
• Leak detection surveys.
Water audits involve detailed accounting of water flow into and out of the distribution system or
parts of it. The American Water Works Association manual Water Audits and Leak Detection
provides detailed procedures for water audits (AWWA 1990). The audits help to identify parts of
the distribution system that have excessive leakage and hence they are an important part of any
effective leakage control program. However, they do not provide information about the exact
location of leaks requiring attention. In order to locate leaks in areas that have been identified by
water audits as suffering from leakage, leak detection surveys must be undertaken. Water system
operators now have several new techniques to help them find and control water losses. In today’s
world of ever-increasing costs of treatment and distribution and tight capital, efficiency in detecting
water leakage is critical to maintain the efficiency of our urban water systems (Keeling 1999).
Leak Detection in Water Distribution Systems
Monitoring of water flows
Over the past few years several new non-instrumental techniques have been developed to provide
better and more cost-effective leakage detection and control programs. A very useful tool in
determining the most effective leakage control program for a particular system is to monitor
minimum night flows. Since leaks flow 24 hours a day, the night flows minus any flows from
legitimate usage, such as commercial cleaning or 24-hour process loads, give a true volume of
water lost to leakage. If night flows minus any legitimate usage flows are close to zero, leakage is
also close to zero and an intensive leakage survey to pinpoint leaks is not warranted. On the other
hand, if night flows minus the legitimate usage flows are high, then a major leakage detection
survey may be warranted. When such leakage surveys are undertaken, the volume of lost water can
serve as a benchmark against which subsequent surveys can be gauged and can provide a sound
basis for cost/benefit studies. Once the survey has been completed and the located leaks are
repaired, monitoring of night flows can be repeated to determine the volume of water saved.
The monitoring of flows can be taken one step further for an even more effective survey. Instead of
monitoring the entire system, individual zones can be monitored. The resulting information can then
be used to plan the most effective leakage detection program where zones of high night flows (and
hence high leakage) are surveyed first. Decisions can be made before surveying each zone as to
whether or not there is sufficient leakage to justify the survey and repair costs. This technique can
again be taken one step further. Zones (ideally of 1000–3000 properties) that can be supplied by one
or two pipes are monitored individually. The technique involves successively closing valves within
the zone to isolate sections of pipe and then recording the corresponding reduction in the water
flow. A large reduction in the flow indicates the existence of a leak between the boundaries of the
last two isolated sections. As a result of this procedure, leak locations can be narrowed down to a
very small area and the leak(s) can then be pinpointed easily using traditional acoustic methods.
Flow monitoring techniques can also be used for ongoing leakage control. For instance, by
monitoring night water flows continuously, unusual changes in water volumes can be detected.
 Pipes Wagga Wagga, Australia 1999

Based on experience with the water system, the system operator can then determine if the volume
and flow increase is due to new leaks or not. If new leaks are the cause, they can be bracketed by
step testing and then quickly pinpointed by acoustic methods as described below.
Detection of leak sounds
The water distribution system can be systematically checked for leaks by using acoustic equipment,
which detects the sound or vibration, induced by water as it escapes from pipes under pressure. The
water passing through the orifice or leak causes a sound in the pipe, normally in the 500–800 Hz
range. The leaking water impacting on the soil in the leak area causes a different sound, normally in
the 20–300 Hz range. A third sound, also normally in the 20–300 Hz range, is caused by the leaking
water circulating in the water remaining in the cavity in the soil adjacent to the leak. The first sound
can be transmitted along the pipe wall for significant distances depending on the type of pipe, so it
is important that initial rough surveys be carried out. The other two sounds are generally limited to
the immediate area of the leak and therefore they are useful for pinpointing the leak. Leak sounds
transmitted through the soil or the pipe can be detected by means of commonly used listening
devices placed on the soil or road surface directly above the pipe or placed in direct contact with a
pipe or its accessories, e.g. fire hydrants.
Listening devices utilise sensitive mechanisms or materials, e.g. piezo-electric elements, for sensing
leak-induced sound and vibration. They could be either of the mechanical or electronic type.
Modern electronic devices may include signal amplifiers and noise filters, which could be very
helpful in adverse environments. The use of listening devices is usually straightforward but their
effectiveness depends on the experience of the user. Leak noise correlators, on the other hand, are
state-of-the-art portable computer-based devices, which can locate leaks automatically. They work
by measuring vibration or sound signals at two points that bracket the location of a suspected leak.
Vibration sensors (normally accelerometers) are attached to fire hydrants, valves or any other
contact points with water pipes. Alternatively, hydrophones (or underwater microphones) can be
used. These are inserted into fire hydrants through modified hydrant caps. Vibration or sound
signals are usually transmitted from the sensors to the correlator via wireless radio transmitters. In
order to pinpoint a suspected leak, noise correlators first determine the time lag between the
measured leak signals by calculating the cross-correlation function. The location of the leak relative
to one of the measurement points is then easily calculated by the correlator based on a simple
algebraic relationship between the time lag, distance between the measurement points, and sound
propagation velocity in the pipe. Normally, leak noise correlators are more efficient and yield more
accurate results compared to listening devices. Since the introduction of these devices in the early
1980s, they have significantly improved the “art” of pinpointing leaks.
The effectiveness of existing acoustic methods and equipment has been demonstrated successfully for
metallic pipes (Fantozzi et al. 1993; Fuchs & Riehle 1991; Liston & Liston 1992). Until recently,
however, the effectiveness of acoustic methods for locating leaks in plastic pipes was not well
understood or established. Existing acoustic equipment was developed mainly with metallic pipes in
mind, but the acoustical characteristics of leak signals in plastic and metallic pipes differ significantly
– plastic pipes are “quieter” and do not transmit sound or vibration as efficiently as metallic ones
(Hunaidi & Chu 1999). A recently completed study of leak detection methods for plastic pipes
(Hunaidi et al. 1999) demonstrated that leaks in plastic pipes can be located using acoustic equipment
and established the necessary procedures and guidelines. This study also identified several
modifications to field procedures and equipment that will improve their effectiveness.

Alternative methods
Leaks can also be detected using several alternative methods or equipment developed by other
industries. These non-acoustic techniques, such as the tracer gas methods, infrared photography (or
thermography) and ground-penetrating radar (Hunaidi et al. 1999), have been used in a limited way,
however, their effectiveness is not well established.
 Pipes Wagga Wagga, Australia 1999

The tracer gas method – This involves the introduction of non-toxic, water-insoluble and lighter-
than-air gases into the water distribution system. Commonly used gases are helium and hydrogen.
Tracer gases injected into pipes escape at leak locations and then, being lighter than air, they
permeate through the soil and pavement. Highly sensitive gas detectors are employed to locate
escaping gas directly above the pipe. The tracer gas method has been used extensively in the past by
the telecommunication industry for leak testing of pressurised telephone cables.
Thermography – The principle behind the use of thermography for leak detection is that water
leaking from an underground pipe changes the thermal characteristics of adjacent soil and in turn
makes it a more effective heat sink relative to the surrounding dry soil. Infra-red scanners are used
to detect thermal anomalies above pipes, which may be taken as an indication of water leaks.
Available scanners can be either hand-held or vehicle-mounted.
Ground-penetrating radar – In this method a short pulse of electromagnetic waves is transmitted
into the ground using a radar-transmitting antenna. When a radar pulse encounters an interface
between two materials, it is partially reflected to the surface where it is sensed by a radar-receiving
antenna. Partial reflection of electromagnetic waves at an interface between two materials, or any
anomaly, is due to a contrast in dielectric properties. The time lag between transmitted and reflected
electromagnetic waves is then used to determine the depth of the reflecting surface. By transmitting
radar pulses at regular surface positions, i.e. performing a scan, the size and depth of buried objects
could be determined. Ground-penetrating radar may be used to detect water leaks in two ways: (a)
identifying soil cavities created by the turbulent flow of leaking water; and (b) identifying pipe
segments which appear deeper than expected because of the increase of the dielectric constant of
adjacent soil saturated by leaking water (Hunaidi & Giamou 1998).
Sewer Leakage Detection
The current inspection technology available on the market is dominated by closed circuit television
(CCTV)-based systems. These are remotely controlled vehicles, which travel through the sewer
pipes carrying TV cameras. A range of other new techniques has, however, also been established
and is becoming commonly used worldwide. These include techniques which involve flow
measurement and statistical techniques.
Existing technologies
Statistical methods cover the full range from dry weather flow to modern stochastic computer models.
For sewer leakage work, such methods can only give a very broad-brush picture, and thus should only
be used for a general view of the system. The main disadvantages are cost and time, as most systems
use deterministic models that require many flow and sewer surveys to be undertaken before they are
useful. They are, however, the best starting point for identifying problem areas. The very latest
computer programs, based on stochastic methods, do however show great promise. Accuracy of
within 10% has been obtained from a very limited database (e.g. flows at key points of a catchment).
Flow monitors (or meters) are only useful as part of a statistical modelling exercise, as their
accuracy is insufficient to gauge minor variance between manholes. Data on water use, and hence
discharge to sewers, is at present mostly statistical and thus not accurate enough for micro studies.
A few strategically placed monitors, however, can quickly include or exclude parts of a system
from a study. When combined with network modelling, a very accurate overall picture of a
catchment can be obtained at a reasonable cost. Development of more accurate meters may improve
the effectiveness of this technique further.
Dye dilution is a method that relies on injecting a dye into the system at a known dilution and
measuring the dilution rate downstream. Obviously this can act as a form of monitoring where
infiltration is suspected, but by adding a flow monitor at the point of sampling, exfiltration can be
possibly gauged.
 Pipes Wagga Wagga, Australia 1999

CCTV is a very useful micro tool for infiltration. Its use is very limited in exfiltration as it relies
totally on a visual record. Whilst visually it is easy to identify defects which may leak, defective
joints are almost impossible to spot. It is also almost impossible to survey the parts of the sewer
under the flow, unless the effluent is very clear. CCTV has also been combined with a sonar unit on
partially surcharged sewers to give a complete picture of the sewer both above and under the
sewage flows.

Infra-red thermography as for water mains detection can be carried out using an aeroplane
overflying the area or by using specially equipped vans. Although good results have been obtained,
the technique is susceptible to environmental conditions (particularly rain) and is prohibitively
expensive for general use.

Dye and smoke testing is a technique much used in Australia and America where records of sewers
are reasonably accurate. It is of less use in European sewers as all connections must be known prior
to tests.

Air pressure tests are the standard test for new sewers. Until very recently, they were of little use on
older sewers as it was almost impossible to seal laterals etc. against air leakage. New robotic
techniques have enabled the insertion of stoppers in sewers. This has made air testing of older
sewers feasible and cost effective, the limitation being that only the main sewer is tested but none of
the laterals or manholes.

Water tests are the only testing systems which give 100% accurate results for both infiltration and
exfiltration. They are unfortunately very labour intensive and disruptive. It is therefore important to
target the lengths for testing by other methods.

Manual surveys simply involve opening manholes in a progressive manner at periods of low flows
(night) and noting any inexplicable clear flows (interim stoppers may be needed). Probably the most
cost-effective method of tracing infiltration, its use in exfiltration studies is very limited.

New techniques
Sonar can give a fairly accurate picture of the profile of the pipe wall and surrounding soil. The
results, however, are very much open to interpretation, and a very skilled operator is vital. It can,
however, show flow regimes under water, so it can be used to spot infiltration into surcharged pipes.

Ground probing radar until recently was of little use as each type of soil and depth of pipe required
a different frequency of radar, hence selection of the correct frequency was critical. A new system,
developed initially to trace plastic land mines, uses multifrequency radar and hence works in a
variety of ground conditions. At present, however, this system has had very few trials on sewers and
its effectiveness is difficult to gauge.

Current sonde is a very new system, developed in Germany. The sonde emits a current
perpendicular to itself (and the pipe) which increases in strength when the sonde passes a leak point.
The main disadvantages are that the computer read-out is difficult to interpret, and that there is no
difference on the screen between a leak and a connection. New developments of a German R&D
project on this method will possibly solve this problem (Eiswirth & Hötzl 1998).

Optical triangulation is a technique used for the optical 3D measurement of a sewer pipe. A new
system has been developed and integrated in Germany (KARO) as well as in Australia (PIRAT)
which permits automatic measurement of the tube shape (diameter and deviations from the circular
shape) during the robot motion. Since the 3D sensor used in these systems looks straightforward
into the direction of robot motion, the sensor scope is comparably large; however, the resolution
and corresponding accuracy remains low. Cracks less then 1 mm cannot be detected (Kuntze et al.
1995; Campbell et al. 1995).
 Pipes Wagga Wagga, Australia 1999

Microwave signals can be used to inspect the state behind sewerage pipes. Commercial microwave
sensors are on the one hand too big for inspecting small pipes and on the other too expensive for
many standard missions. In the KARO project, a smaller and cheaper microwave backscattering
sensor has been developed in Germany which is able to explore anomalies in a medium range
behind the outer pipe surface (Kuntze et al. 1995). During the last year, a wall scanning microwave
sensor has been developed which is carried by a vehicle and can be rotated around the pipe axis.
Thus, the complete wall surface can be scanned with optimal resolution (Eiswirth & Hötzl 1998).

Hydrochemical sensors used in preliminary tests in Germany with mobile probes within sewers
indicate the applicability of this method for the detection of ground water infiltration with high
accuracy. Since electrochemical sensors are sensitive to only a few chemical compounds and they
have a comparable low resolution, they must be tuned to the wastewater composition. The detection
of ground water infiltration has been shown to be possible; however, sewage exfiltration detection
remains nearly impossible. Detailed investigations concerning the improvement of resolution and
accuracy of these sensors have been done in laboratory experiments as well as in field experiments
(Eiswirth & Hötzl 1998).

Neutron and gamma ray probes have been used in geophysical boreholes for the detection of cracks
as well as in ground water observation wells for the investigation of soil density, soil moisture
content, etc. Tests with radioactive sensors within sewers indicate the applicability of this methods
to the detection of active sewer leaks (significant changes in soil moisture) and holes behind the
sewer pipes (different density). Since radioactive sensors are affected by different physical
underground parameters, their sensitivity and resolution was low until now. Detailed investigations
with radioactive sensors are planned in laboratory experiments, as well as in test beds under
different situations in Germany.

Air ultrasonic sensors have been adapted and integrated into the German KARO multisensor system
(Kuntze et al. 1995) for the detection and quantification of cracks and wall thickness. Air ultrasonic
sensors used in the KARO system have a large inspection area, however, the resolution and
corresponding accuracy is rather small. Fine fissures, their depth and extent, are difficult to detect.

Acoustic systems are based on detecting vibrations and other phenomena caused by the spreading of
mechanical waves, and are suitable for detecting cracks as well as for determining the state of
connections and the bedding. Vibrations in the low frequency range (100–10,000 Hz) are excited
through hammer taps (Klingmoller 1995). Until now the resolution and corresponding accuracy of
this method has not been suitable for detecting leaks in sewers.

Application of methods
The most suitable method(s) applied for identification of sewer leakage will depend on the size of
the sewer, location (under structure, in road, class of road, urban/rural etc.), flow, defect(s), nature
of sewage and percentage of sewer capacity already used. Therefore, most detection techniques
need to be combined with others to give the best results. Most water companies have implemented
drainage area studies (DASs) and drainage area plans (DAPs). A DAS is a combination of a desk
study, gathering all records of the system and collating, ground surveys, such as manhole and flow
surveys, and CCTV. In a DAP, the data from the DAS is analysed and the system graded. After
further surveys (CCTV, etc.) recommendations can be made based on the survey reports and
computer modelling. A common strategy, much used throughout the world, would be (Jones 1998):

1. Use existing records to identify critical position in the catchment (node points such as major
branch connections, etc.).
2. Monitor at these positions.
3. Model the system on stochastic computer models.
4. Separate inflow and infiltration (using rainfall data).
 Pipes Wagga Wagga, Australia 1999

5. Prioritise the catchments.

6. Identify the dominant defect class in the priority catchments (by random survey).
7. Conduct the appropriate physical inspection for the defect.
8. Determine optimal system investment by evaluating the cost of rehabilitation versus the cost of
replacement required for sufficient hydraulic capacity.
9. Carry out a cost/benefit analysis for each section or catchment.
10. Develop a sustainable capital investment program that addresses both replacement and
rehabilitation projects.
11. Secure appropriate financing to bring the system into acceptable operating conditions.

Exfiltration, whilst as easy to cure as infiltration, is far harder to identify. It is possible to first
identify potential problem areas by statistical methods, and then to monitor down to progressively
smaller catchments. This is despite reservations about the accuracy of monitoring. To identify
actual lengths that are leaking, the best method seems to be water tests, although the current sonde
system and other new techniques are showing promise. At present, however, exfiltration work is
generally ruled out on cost/benefit grounds, and seems likely to remain this way unless legislation,
or public pressure, forces changes on the industry.

LEAKAGE SOLUTIONS AND CONTROL

Pressure Control
Pressure management has been around for many years in its basic form as a means of hydraulic
system control and more recently has been very successfully introduced in many countries to
combat leakage. Pressure management, as a means of combating leakage, can be used in most
systems whether they are pumped or gravity fed, although the design of the scheme will change
dramatically due to different hydraulic patterns. Often the reduction of pressure from one level to
another can be a controversial subject, and one which sometimes utility managers prefer to ignore
as there is a potential for customer dissatisfaction.

Pressure management may not be the answer for all utility systems, but it may prove to be very
successful for others. Systems with a large percentage of uPVC pipes, systems with high service
densities and systems in dense urban areas, would probably be among the best candidates to benefit
from an advanced pressure management scheme.

Recent studies have proven that the relationship between leakage and pressure is not merely related
to the square root of the pressure, but rather an expanding power law. This is because PVC leaks,
and many other types of leaks and joints in particular, are subject to a change in area as pressure
changes, and consequently the typical laws relating pressure and leakage do not apply. This means
that potential benefit in pressure reduction, based on the volume of these leaks, has a much greater
impact as not only the velocity of leak flow changes, but also the leak area.

Some of the potential benefits of a pressure management scheme are:

• Leakage reduction.
• Water conservation.
• Efficient distribution of water.
• Reduced hydraulic impact.
• Reduced customer complaints.

Some utilities are not comfortable with pressure management as a means of loss control for the
following reasons:
• Fire flow concerns.
• Potential loss of revenue.
• Reservoirs not filling at night.
 Pipes Wagga Wagga, Australia 1999

Where fire flows are a concern, sectors can have multiple feeds, controlled by pressure-reducing
valves (PRVs) with flow-modulating capability. Therefore, if there is a fire, the system has
sufficient hydraulic capacity to maintain pressures and flows for firefighting, as required for
example in the USA in the NFPA regulations. The valves will automatically regulate pressure as
determined by the demand requirement plus the minimum safe operating limit at residual
conditions. Obviously when setting up potential pressure-controlled sectors, these limits along with
insurance regulations for the type of property, should be taken into account.

As far as the loss of revenue is concerned, systems with high leakage will almost always see a
positive benefit from pressure management, even when stacked against a potential loss of revenue,
due to reduction of pressure in the residence or industry. This is especially true for systems with
high water production or purchase costs, where the associated reduction in water losses can make
significant savings. In situations where a loss of revenue cannot be tolerated, pressure management
can be limited to night hours, when legitimate consumption is at its lowest and system pressures are
at their highest. It should be remembered also that many systems are enforcing water-conservation
programs where pressure-reduction programs can play an important role.

When considering pressure management for a sector, the per capita use must be considered. If it is
excessive, then pressure management will become a natural part of a conservation program. If it is
not, the components of consumption within the sector (residential, commercial, industrial), the
volumetric consumption and the consumption directly tied to pressure must be decided. Then, the
potential benefits of loss reduction over reduction in revenue can be analysed.

Many pressure management programs concentrate on the smaller mains, therefore allowing
reduction of losses in selected areas while allowing normal system pressure in the larger trunk or
transmission lines. Reservoirs are usually connected with the larger pipes, so there should not, in
many cases, be a problem. Most utilities find that non-visible leakage tends to be on the smaller
pipes and service connections, so the effectiveness of a potential pressure management program
should not be reduced significantly, by the exclusion of larger pipes in the control area.

Pressure control can be characterised into a number of different types as follows.

Sectorisation
This is one of the most basic forms of pressure management and is still very effective. Subsectors
are divided either naturally or by physical valving. The sectors are usually quite large and often
have multiple feeds, therefore they do not usually develop localised hydraulic problems because of
valve closures. Systems with gravity feeds are usually sectorised by ground level and systems with
pumped feeds are usually sectorised depending on the level of elevated tanks or storage.

Pump control
Many utilities use pump control as a method of controlling system pressure. Pumps will be
activated or de-activated depending upon system demand. This method is effective if the reduced
level of pumping (usually at night) can still maintain reservoir levels. With recent energy
conservation concerns, this methodology should be carefully reviewed as to the efficiency of energy
use if the pump is operating outside of the normal profile.

Throttled system valves

Many system operators recognise the need for reducing system pressure and partially close a gate or
butterfly valve to create a headloss and reduce pressure. This method is the least effective, as the
headloss created will change as system demand changes. At night when a distribution system needs
the least pressure, the pressure will be higher, and during the day when the distribution system
needs the most pressure to supply demand, the pressure will be lower.
 Pipes Wagga Wagga, Australia 1999

Fixed outlet hydraulic control

This is the traditional method of control and uses a basic hydraulically operated control valve. This
method is effective for areas with low headlosses, demands which do not vary greatly due to
seasonal changes, and areas with uniform supply characteristics.
Time-based modulation of a hydraulic valve (Ecowat)
This can be effected by using a controller with an internal timer. Control is effected in time-bands in
accordance with demand profiles. This methodology is very effective for areas with stable demand
profiles and headlosses and is usually used where cost is an issue, but advanced pressure
management is desired. Time-based modulation controllers can be supplied with or without data
loggers and/or remote links. Some manufacturers connect the controller to the pilot valve and alter
the set point of the pilot valve by introducing a force against the existing force of the pilot spring.
Other manufacturers use a timer and a solenoid valve to re-route control through preset pilots.
Flow-based dynamic modulation of a hydraulic valve (Autowat)
This is the best type of control for areas with changing conditions, headloss, fire flow requirements
and the need for advanced control. This type of control is affected by controlling outlet pressure in
relation to demand by connecting the controller to a metered signal output. Modulation of outlet
pressure is achieved by altering the force against the pilot spring. The controller is normally
supplied with a local data logger and optional remote communications. Control can be effected with
a preset profile which shows the changing relationship of demand and headloss in the sector.
Alternatively a direct communications link can be made between the controller and the critical
point. Obviously the second option involves communications and therefore higher costs, which are
not always necessary or cost effective. In general, installation costs are higher for this type of
control, however, additional savings and guaranteed fire flows due to more intelligent control
usually make this type of control more desirable.
Pressure Control – Water Savings
Pressure control can offer significant savings in water loss, as shown in Table 2, for sites in the
metropolitan area of Sao Paulo, Brazil, utilising demand-based dynamic control methods, time-
based methods and fixed traditional control. It should be noted that these water savings are not just
a function of the reduced pressure, but also of the size of the zone in kilometres, service density and
number of connections, and also the amount of pressure controlled and for how many hours per day.
Figure 8 shows how pressure control can significantly flatten the pressure profile across a system,
even though the actual pressure reductions are very small, being in the order of only 10 m.
Table 2. Water savings achieved utilising pressure control
Site location Pressure Pressure Pipe φ pipe/ Savings Control
before after length PRV
(m) (m) (km) (mm) (L/s)
Rua do Curtume x Av. Hermano Marchetti (Lapa) 45–63 20–50 5.2 150/150 12.10 Autowat
Rua Francisco da Cunha Menezes 28–72 20–40 11.6 300/150 11.50 Autowat
Rua Antônio Portugal (Guaraú) 86–130 60 7.9 200/150 9.40 Fixed
Rua José Arnoni (Guaraú) 77–105 20 4.25 200/75 6.20 Fixed
Rua Pilões (Sacomã) 42–87 20–47 30.5 400/200 49.00 Autowat
Avenida Paulo de Queiroz (Sapopemba) 54–67 25 6.7 400/100 4.89 Fixed
Rua S. Francisco Magiano (Itaquera) 20–42 20–33 41.3 300/200 10.80 Ecowat
Av Cantidio Sampaio X Rubens Raul (Jaraguá) 60–97 20–40 3.7 150/100 2.00 Autowat
Rua Alvarenga (Butantã) 43–63 32 10.5 200/150 4.98 Fixed
Rua Xavier de Moraes (Caieiras) 110 70 3 200/75 1.00 Fixed
Rua Tomás Xavier de Almeida (Perús) 63–72 26 15.2 300/200 13.20 Fixed
Av. Santo Amaro x R. Mal. Deodoro (Sto. Amaro) 41–69 20–44 10.4 200/100 3.80 Ecowat
Rua Ambrosina do Carmo (Caieiras) 21–53 17–30 30 200/200 33.30 Autowat
Rua Lacerda Franco (Vila Deodoro) 45–54 22–42 20.1 350/250 16.70 Autowat
Total 178.87
 Pipes Wagga Wagga, Australia 1999

35

30

Pressure (m) 25

20

15

10

0
00

45

30

15

00

45

30

15

00

45

30

15

00

45

5
00
:3

:1

:0

:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1
0:

0:

1:

2:

3:

3:

4:

5:

6:

6:

7:

8:

9:

9:

0:
10

11

12

12

13

14

15

15

16

17

18

18

19

20

21

21

22

23
Figure 8. Pressure before and after control.

Even though the amount of pressure reduction was small, the savings in water loss were very
significant, as can be seen in Figure 9. We should also note that the pressure after control shown by
the bottom area in Figure 8 is much more stable than the top area, which is uncontrolled and subject
to wide variation. This smoothing of the pressure cycle also reduces new bursts within the system.

REHABILITATION, REPAIR OR REPLACEMENT

Once a defect has been identified in a pipeline, a decision has to be made to undertake a specific
type of rehabilitation.

Maintenance involves the cleaning and repair of the pipeline, to optimise hydraulic capacity by the
removal of foreign or dislodged materials such as tree roots, encrustations and bricks, and it is fairly
easy to make a decision on the correct procedures to apply.

Repair involves fixing local damage to the pipeline and may involve such techniques as part lining
repair of pipelines using small areas of epoxy linings etc., grout injection to stabilise backfill soil or
to stop areas where water infiltration/exfiltration occur, or brick repointing.
1600

1400
Volume before control Mains length 63.7 Km
1200
26,730 m3/day
309,4 l/sec
1000
Flow (m3/h)

800

600 Reduction in leakage = 76,5 l/s

400 Volume after control

20,118 m3/day
200 232.9 l/sec

0
00

45

30

15

00

45

30

15

00

5
:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1

:0

:4

:3

:1
:

:
00

00

01

02

03

03

04

05

06

06

07

08

09

09

10

11

12

12

13

14

15

15

16

17

18

18

19

20

21

21

22

23

Figure 9. Volumes and flow before and after control.

 Pipes Wagga Wagga, Australia 1999

Renovation is where the original pipeline is maintained and the structural condition of the whole
pipeline is improved. This may involve lining the pipeline with a lining using pipe or sprayed
materials and may involve a reduction of hydraulic performance. Alternatives include soft lining,
folded lining, reduced diameter lining, spiral lining, manual lining, unmodified slip lining, spray-on
lining, grouting and brick repointing.

Replacement is where the pipeline is replaced by either laying a new pipeline alongside the existing
pipeline using techniques such as pipe bursting, back reaming, microtunneling, jetting, conventional
tunnelling, trenching or off-line replacement techniques such as trenchless steerable, guided
drilling, directional drilling, microtunneling, conventional tunnelling or trenching.

Rehabilitation Selection Criteria

A large range of procedures is available for pipeline rehabilitation, however, the selection of a
suitable rehabilitation technique can be a difficult decision. This choice requires decisions to be
made concerning:
• The required lifetime to be obtained.
• The structural integrity of the existing pipeline and the structural integrity needed.
• The hydraulic capacity needed.
• Whether disruption to service can occur.
• The ability to divert sewer or potable water flows.
• The extent and location of surface works.
• Soil and ground conditions.
• The number and method of reconnecting laterals.
• The chemical, corrosion and abrasion resistance needed.
• The degree of cleaning needed to undertake the recommended renovation technique.

Systems to allow rational selection of a pipeline rehabilitation technique based upon the above
points are not readily available. Decision support systems to evaluate replacement or preventive
maintenance priorities have been suggested or implemented by several groups (Kulkarni & Reid
1991; Madiec et al. 1996; Fenner & Sweeting 1999). These do not necessarily address the issue of
selection of the repair/replacement technology. Kleiner et al. (1998) discuss the choice of an
optimal rehabilitation strategy. Their proposal is based on dynamic programming and (sometimes)
implicit enumeration of schemes to select, for each pipe in an existing network, the rehabilitation
alternative and the time of its implementation so as to minimise the cost of the rehabilitation.
However, none of these systems look at a full lifecycle costing methodology taking into account the
expected life of the repair or maintenance, the costs associated with all aspects of the rehabilitation
and the costs borne by third parties (externalities).

Also important will be quantification of the benefit gained for the expenditure committed. As
indicated earlier, environmental impacts of, for example, sewer overflows are not consistent across
all receiving environments. The level of rehabilitation required to achieve the desired goal and its
cost should be taken into account in prioritising rehabilitation programs. This is not to suggest that
the environment should be compromised, but that a rational decision-making framework is required
to ensure that money available for environmental rehabilitation programs is directed to those
rehabilitation programs that will produce the greatest effect.

Recent major failures in Australian urban water infrastructure highlight the problems that can occur
if a consistent approach is not taken to the development of a rational asset management system to
deal with pipe leakage. Development of an asset management system incorporating the components
of leakage detection, pressure management and rehabilitation selection has not occurred in an
Australian water authority, although some aspects are currently under development. For this process
 Pipes Wagga Wagga, Australia 1999

to occur, a decision-support system to allow selection of rehabilitation techniques incorporating

lifecycle costing methodologies, to assess the viability of each selection decision, needs to be
developed. The AWWA has recently commissioned a project in this area and as these tools are
developed the implementation of a rational asset management system in this area becomes possible.

CONCLUSIONS

It is recognised that leakage of sewer and water reticulation pipes is a problem worldwide. Although
the problem in Australia is not as great as countries with older systems such as Germany, as our
infrastructure continues to age the problems will increase. This is compounded by recent changes to
the way water authorities operate, where capital expenditure needs to be justified as being cost
effective before any replacement or maintenance of pipes occurs.

Leakage in both water and sewerage pipelines can have significant ecological and design effects,
with overdesign of sewerage systems and the associated pipework being a major cost factor in
providing a service to customers. If overdesign does not occur, as is the case for some water
authorities in Australia, then sewer overflows are a real possibility, with their associated health risks
and bad publicity. Whilst most attention has been focussed on sewer overflows, sewer leakage has
also been highlighted as a potential source of environmental problems, although one that is
currently receiving little attention.

To solve leakage in both sewer and water reticulation systems, sophisticated leak detection
techniques are required. The best method of leakage detection and control is very dependent on the
type of system and customer base in the utility. Whilst new systems may appear, the most likely
developments will be in the combining of current systems. A combination of CCTV, current sonde,
microwave sensor, neutron and gamma ray probe, and hydrochemical sensors would be a powerful
tool, provided it was not prohibitively expensive. The most significant factor requiring water
authorities to undertake such exercises, however, will be the public perception of sewer leakage.
This is already happening in Germany where leaking sewers are becoming of environmental
concern to the public (Misstear et al. 1996). The focus of the media up to now has largely been on
water pipe leakage, but the future focus will be on sewers.

A number of procedures can be undertaken to reduce leakage in both sewers and water supply
systems. In sewers, the only existing technology is rehabilitation or replacement. For water
reticulation pipelines, either rehabilitation can be undertaken or pressure management techniques
instigated. Pressure management can be a very cost-effective means of reducing losses or
controlling demand, for many utilities, with a wide diversity of conditions. As with any potential
investment, in loss control or demand management it is important to carefully model or calculate
the effects and benefits of the control prior to implementation to ensure ongoing system
requirements while cutting out unnecessary losses and system inefficiencies.

The choice of which rehabilitation technique or whether rehabilitation needs to be undertaken at all is
currently difficult, as little information is available to aid selection of the appropriate technology.
Asset management systems need to be developed to allow technology selection, taking into account
lifecycle costing methodologies as well as externalities, such as customer disruption and health risks.

REFERENCES

AWWA (1987), Leaks in Water Distribution Systems – A Technical/Economic Overview

(translation of a 1980 report published by the Association Québécoise des Techniques de
L’eau, Monteral, Quebec), American Water Works Association, Denver, CO
 Pipes Wagga Wagga, Australia 1999

AWWA (1990), Water Audits and Leak Detection, Manual of Water Supply Practices No. M36,
American Water Works Association, Denver, CO.
Campbell, G., Rogers, K. & Gibert, J. (1995), PIRAT – a system for quantitative sewer assessment,
Proc. Int. Conf. No Dig '95, Dresden, Germany, 19–22 September 1995, pp. 455–462.
Cheong, L. C. (1991), Unaccounted for Water and the Economics of Leak Detection, 18th Int.
Water Supply Congress & Exhibition, Copenhagen, 15–31 May 1991. Also Water Supply,
9(3&4) (1991).
Eiswirth, M. & Hötzl, H. (1998), Sewer defect characterization with multi sensor techniques,
EntsorgungsPraxis, 4, 49–53 (in German).
Fantozzi, M., Di Chirico, G., Fontana, E. & Tonolini, F. (1993), Leak inspection on water pipelines
by acoustic emission with cross-correlation method, Proc. AWWA Annual Conf., pp. 609–621,
American Water Works Association.
Fenner, R.A. & Sweeting, L. (1999), A decision support model for the rehabilitation of “non-
critical” sewers, Water Sci. & Tech., 39(9) 193–200.
Fuchs, H.V. & Riehle, R. (1991), Ten years of experience with leak detection by acoustic signal
analysis, Applied Acoustics, 33, 1–19.
Fuhrmann, D. n.d., New Developments on the Field of Sewer Sanitation – A Main Funding Activity
of the Federal Minister for Research and Technology (BMFT), Nuclear Research Centre,
Karlsruhe, Water Technology Agency.
Jones, M. (1998), Sewer leakage – detection and cures, http://klingon.util.utexas.edu/jones/
leaksewer.html.
Hunaidi, O., Chu, W., Wang, A. & Guan, W. 1999, Leak detection methods for plastic water
distribution pipes, Proc. AWWA Research Foundation Technology Transfer Conf., Fort
Lauderdale, Florida, 18–19 February 1999, pp. 249–264, also to appear in J. AWWA
(available online at http://www.nrc.ca/irc/leak/ leakdetect.html).
Hunaidi, O. & Chu, W. (1999), Acoustical characteristics of leak signals in plastic water
distribution pipes, J. Applied Acoustics, 58, 235–254 (available online at
http://www.nrc.ca/irc/leak/ leakdetect.html).
Hunaidi, O. & Giamou, P. (1998), Ground-penetrating radar for detection of leaks in buried water
distribution pipes, Proc. 7th Int. Conf. on Ground-penetrating Radar – GPR’98, Lawrence,
Kansas, 27–30 May 1998, 2, 783–786 (available online at http://www.nrc.ca/irc/leak/
leakdetect.html).
Keeling, D.A. (1999), New techniques in water leak detection, http://www.heathltd.com/
techpaper.html.
Kleiner, Y., Adams, B.J. & Rogers, J.S. 1998, Selection and scheduling of rehabilitation
alternatives for water distribution systems, Water Resources Research, 34(8), 2053–2061.
Klingmoller, O. (1995), Development of sound reflection analysis in the form of remote-controlled
non-destructive knocking testing on non-negotiable sewers, Proc. Int. Conf. No Dig '95,
Dresden, Germany, 19–22 September 1995, pp. 222–234.
Kulkarni, R.B. & Reid, F.A. (1991), Replacement/maintenance priorities for gas distribution
systems, Proc. Annual Conf. of the Urban & Regional Information Systems Assoc., 11–15
August 1991, 1, 156–66.
Kuntze, H.B, Schmidt, D., Haffner, H. & Loh, M (1995), KARO – a flexible robot for smart sensor-
based sewer inspection, Proc. Int. Conf. No Dig '95, Dresden, Germany, 19–22 September
1995, pp. 367–374.
Liston, D.A. & Liston, J.D. (1992), Leak detection techniques, J. New England Water Works
Assoc., 106(2), 103.
Madiec, C., Botzung, P., Bremond, B., Eisenbeis, P., Skarda, B.C., Ray, C.F. & Matthews, P.
(1996), Implementation of a probability modal for renewal of drinking water networks, Water
Supply, 14(3–4), 347–350.
Misstear, B., White, M., Bishop, P. & Anderson, G. (1996), Reliability of Sewers in
Environmentally Sensitive Areas, CIRIA Publication PR 44, London.
 Pipes Wagga Wagga, Australia 1999

Sydney Water Corporation (1998), Licensing Sewerage Overflows, Environmental Impact

Statement: Volume 1 – Sydney-wide Overview.
White, M., Johnson, H., Anderson, G. & Misstear. B. (1997), Control of Infiltration to Sewers,
CIRIA Publication R175, London.
WSSA Facts 98 (1998), The Australian Urban Water Industry, Water Services Association of
Australia.

View publication stats