Which is more effective for a water utility that has almost 400 DMAs and a service area slightly larger than one million customers—acoustic loggers or teams of acoustic-listening specialists?
Water has always been an integral
part of Bristol’s history. England’s sixth largest city, it is located on the
estuary of the River Severn, which has a tidal range of almost 40 feet, and
flows into the Bristol Channel between England and Wales. The city has been
major shipping port dating back to the 10th Century, when England first began
trading for Irish wool. In addition to the River Severn and the River Avon,
Bristol is also blessed with a large number of wells that produce relatively
inexpensive potable water.
In 1997 after two years of severe
drought, the British Office of Water Services introduced a new set of water
leakage standards for utilities in England and Wales. These changes amount to
what a 2002 American Water Works Association Worldwide Best Management Practices
report called, “the most advanced national system of water loss control in the
world today.”
In order to meet these mandates,
Bristol Water plc, a water-only utility with a 1,000-square-miles service area
and a daily output of more than 300 million liters, began a program of district
metering, which produced dramatic results. For example, between 1995 and 2002,
Bristol Water reduced its leakage levels from 66 million liters per day (17.4
million gallons) to 53 million liters per day (14 million gallons)—a 19.7%
reduction and a savings of 13 million liters per day.
In 2001, Bristol Water’s Water
Services Department, Network Technical Section, undertook a survey to determine
which was the most effective leak-prevention strategy: acoustic loggers or teams
of acoustic-“sounding” experts. The investigation was two-fold: 1. Determine
whether there were sound economic reasons to replace acoustic-listening teams
with loggers; 2. Test acoustic loggers under a variety of conditions.
 |
Photo: Adrian Pingstone, Wikipedia |
| The River Avon flows through the center of Bristol, England |
The question, posed by the study’s
authors Frank van der Kleij and Matthew Stephenson, eloquently states the nature
of Bristol Water’s investigation: “In an environment with a very high level of
district metering and a pool of qualified and fully trained leakage inspectors,
what is the practical experience of using acoustic logger in leakage
surveys?”
While the results of van der Kleij
and Stephenson’s work are interesting, perhaps the most telling aspect of this
study is that 97% of properties in Bristol Water’s service are within a District
Meter Area (DMA)—a level of meter coverage uncommon in the US. According to the
study, Bristol Water’s active use of district metering and qualified leakage
inspectors produced an “Infrastructure Leakage Index between 1.3 and 3.0 in 10
resource zones, with a weighted company average of 1.8.”
Bristol Waters tested three types
of loggers in their study: Type 1 (with radio transmitters), Type 2 (with the
choice between quick download of information or full analysis), and Type 3 (with
a correlating function). The initial investigation was broken down into two
types of surveys: detailed and general.
Detailed surveys were performed
when excessive losses occurred gradually over time, and multiple leaks were the
suspected culprits. This type of survey included listening to all fittings and
fixtures, including stop taps.
General surveys were conducted
when large or sudden loss of water occurred. This survey included acoustic
listening at all main valves/fittings and all hydrants in the area.
The temporary deployment of
loggers (left for several days, retrieved, downloaded, and analyzed) was
compared to leak inspection teams in terms of cost effectiveness. “All three
systems could not outperform leaking inspectors in total cost for survey, when
comparing with a standard-sounding survey,” the study says. “All systems found
the most significant leaks within the area, but all systems missed some leaks
that the leakage inspectors found.”
Leak inspectors cost 138 Euros to
undertake the survey, while the loggers ranged from 196 Euros for Type 1, to 184
Euros for Type 3. Worse still, the payback point on loggers covering of 50
kilometers of main came to 26 years on Type 1, and 41 years for Type 2. “This
payback period is too great to consider, and, as a result, Bristol Water have
justified maintaining their leakage inspector teams,” the report states.
According to van der Kleij, “The
price has come down [since the original study], but the payback period is still
probably not better than 15–18 years.”
The second stage of the study was
to determine if there was an optimum use for acoustic loggers. Both general and
detailed surveys were conducted in areas that tend to generate work for night
survey crews (busy streets in popular city centers and high-nighttime-traffic
areas).
The results didn’t fare well for
the loggers. Leak inspectors found 80% more leaks in the detailed surveys, than
were detected by the acoustic loggers (mostly smaller leaks in the
40–80-liters-per-hour range). In the general survey, the leakage teams found 20%
more leaks than the loggers. And, the general survey teams were able to do their
work three times faster than the acoustic-logger surveys.
However, the results were not all
black and white.
In the first survey, the cost
effectiveness for permanent deployment of loggers (Type 1 and Type 2 using radio
patroller units—Type 3 could not be interrogated without being collected first)
came out in favor of the loggers: 63 Euros for Type 1 and 104 Euros for Type 2,
versus 138 Euros for the leak inspectors.
In the second survey, loggers were
able to identify leaks on evenings when the ambient noise was low (Monday and
Tuesday) in areas that receive a high level of late-night background noise.
“Using normal-sounding techniques, a leakage inspector would need to return to
the same location on a number of nights in order to successfully locate any
leaks, which is clearly not a cost effective solution,” the survey says.
Van der Kleij and Stephenson also
acknowledge some potential experimental design flaws—such as the spacing of
acoustic loggers and a small data set—could skew the results, and that more work
is required to define the right niche for loggers in their system. He also
points out that the report on which all this information is based was written in
2002. Additionally, van der Kleij feels important to note that in order to
assess a level of “optimum” real water losses in the system for the purposes of
the Bristol Water’s Water Services Department’s study, the cost of water was at
measured at about nine pence per 1,000 litres of water.
Several non-DMA infrastructures in
the US have reported a payback time for loggers at just over two years. What
Bristol Waters’ extensive DMAs and inexpensive water supply help demonstrate is
that loggers are not a one-size-fits-all silver bullet solution for leak
management issues; they are an effective tool along with acoustic-listening
devices, correlators, and meters.