Page 161 - Water Loss Control

Modelling W ater Losses 139

1. ICF based on system-wide ILI: The ILI is a performance indicator calculated in
relation with the top-down water balance. It is a dimensionless indicator
describing the ratio between the unavoidable annual real loss volume and the
current annual real loss volume calculated by the water balance. A quick first
estimate of ICF can be taken from the ILI of the entire system. The system-wide
ICF can be assumed to have a similar value to the ILI.
2. ICF based on initial sensitivity analysis: Undertake a sensitivity analysis which
averages the two extreme possibilities of the ICF. A minimum ICF equals one,
where real losses volumes are composed of the unavoidable background losses
and recoverable losses. The maximum ICF happens when all leakage is due to
background leakage except for a ratio of 1 for the components of reported and
unreported leakage. For example, if the maximum ICF is 6 with the other two
components at 1 and the minimum ICF is 1 with the other components higher,
then the average ICF would be 3.5 and initial estimations could be made for the
other components of leakage and potential solutions. It is recommended however
that field testing is undertaken to validate this simple estimation process.
3. ICF based on N1 step test: If the system is predominantly a rigid or metallic
system, an N1 step test is a valuable tool to estimate the ICF value in DMAs.
Based on a representative sample of N1 step test results a system-wide ICF can
be calculated. If the system is not predominantly rigid or metallic the principles
behind the N1 step test and its calculations do not fully apply, and may result
in an overestimation of the background leakage component as the breaks
themselves may have a variable leakage path.
4. ICF based on removal of all detectable leaks: Once a DMA has been installed, even
on a temporary basis, and all recoverable leakage has been identified and
repaired, then the remaining background leakage level can be measured. In an
ideal situation, night time consumers are temporarily turned off so that there is
little doubt that the measure flows represent background leakage. Where this is
not possible then it is necessary to use a process similar to that described in the
previous example to build up a picture of night consumption (including toilet
leakage) and subtract that from the measured night flows. However, confidence
in such a result would not be as good as the first option of turning off consumers
for the period of the test. Results from representative DMAs can be used to
estimate a system-wide ICF.

10.6.1 Calculating Losses from Reported and Unreported Breaks
After collecting the annual numbers of reported breaks on mains and service connec-
tions (and other system components such as valves and hydrants if so desired), flow
rates and durations have to be established. Unless the utility has investigated average
leak flow rates and has detailed data available the figures from Table 10.8 can be used
as a starting point.
The break/leak duration can be split in three elements—time needed for
Awareness duration: The length of time taken from a leak first occurring—whether
it is reported or unreported—to the time when the utility first becomes aware that
a leak exists, although not necessarily aware of its exact location. For reported
leaks and breaks, this duration is usually very short, while for unreported leaks
and breaks, it is a function of the active leakage control policy interventions.