Page 165 - Water Loss Control

Modelling W ater Losses 143

a first estimate for a compound N3 value could be N3 = 0.25. In this case, the impact of
reduction in leakage will mostly be proportionally greater than the impact of reduced
demand. In some cases, reduction of demand might actually be desired as in the case of
water conservation projects.

10.6.4 Modeling Break Frequency Factor BFF 7
The need for a better understanding of the influence of maximum system pressure on breaks
has recently been addressed by the IWA Water Loss Task Force’s pressure management team.
An extended data set of 112 systems from 10 countries as reported by Thornton and Lambert
8
in IWA Water 21 is summarized in Table 10.10. The following can be noted:
• “ Before” pressure (meters) ranges from 23 to 199, median is 57 and average 71.
• Percent pressure reduction ranges from 10 to 75%, median 33%, average 37%.
• Percent reduction in breaks ranges from 23 to 94%, median 50%, average 53%.
• The data shows no significant difference between average % break reductions
on mains and service connections.
The data from Table 10.10 are also shown in Fig. 10.2 as a plot of % reduction in
pressure versus % reduction in new break frequency, for mains and services together.
A simple interpretation, likely to give generally conservative predictions, is to
assume that the % reduction in new breaks = BFF × % reduction in maximum pressure,
where BFF is a break frequency factor, this can be checked against the data in Fig. 10.2.
The average value of BFF for mains and services together from Fig. 10.2 is 52.5%/
38% = 1.4, so a line drawn through the data in Fig. 10.2 with a slope of 1.4 gives an average
prediction.
An Upper line, with a BFF of 2.8 (twice the average) encompasses all but two of the
data points which give larger reduction in new break frequencies
A ‘Lower’ line, with a BFF of 0.7 (half the average) encompasses all the data points
which give smaller reductions in new break frequencies
When applying this simplified prediction approach, it is important to ensure that in cases
where both the BFF and the % reduction in maximum pressure are both large, the prediction
does not reduce the break frequency below the values used in the UARL formula.

10.6.5 The Latest Conceptual Approach
The latest conceptual approach currently being used by the Pressure Management Team
of the IWA WLTF, in attempting to develop an improved practical understanding of
pressure/break frequency relationships, is shown in the following series of figures.
In Fig. 10.3 the X-axis represents system pressure and the Y-axis represents failure
rates. When a new system is created, mains and services are normally designed to with-
stand maximum pressures far greater than the range of daily and seasonal operating
pressures for a system supplied by gravity. The system operates with a substantial fac-
tor of safety, and failure rates are low. Even if there are pressure transients in the system
as shown in Fig. 10.4, the maximum pressures do not exceed the pressure at which
increased failure rates would occur.
As the years pass, adverse factors based on age (including corrosion) gradually reduce
the pressure at which the pipes will fail as shown in Fig. 10.5. Then, depending upon local
factors such as traffic loading, ground movement, and low temperatures (which will vary