abstracted, others treated water.
System Input Volume may include or
exclude water exported, which can be
substantial in some systems, absent in
Another question to be asked is does
an individual NRW calculated as a
percentage of SIV include any water
exported? You don’t even know from
Figure 1 if a utility exports water or
not. If a utility’s NRW is 8 percent of
SIV including exports, but 13 percent
excluding exports, which fgure did
they choose to report? Did you realise
that the ‘best’ Utility in Figure 1 (NRW
of 3 percent of SIV) is a transmission
utility exporting almost all of its water;
that some systems in the data set have
intermittent supply, and in some others
customers have roof storage tanks
which signifcantly increase the meter
under-registration component of NRW?
Another common mix up is confusing
non-revenue water with leakage.
However, NRW consists of several
different components. For example,
unbilled authorised consumption and
apparent/commercial losses (customer
meter under-registration and theft) are
components of consumption which
reach customers, but are not paid for.
The remainder of NRW is Real Losses
(leakage) from utility systems, some
of which represents a potentially
recoverable water resource.
Due to different meter locations on
service connections, in some countries
calculated leakage includes leakage on
customers’ private underground pipes;
in other countries it doesn’t. You don’t
know which situation applies to each
utility in Figure 1.
If setting targets to reduce leakage,
you must frst separate the NRW
‘unpaid consumption’ components
(representing lost revenue) from the
NRW leakage components (representing
lost and potentially recoverable
water). However, you don’t have that
information in Figure 1, and unpaid
consumption components of NRW can
range from almost zero to almost half of
the percentage NRW.
The term ‘Percent of System Input
Volume is a Zero-sum calculation’ is
well known in fnancial and commercial
circles, but not in water loss control.
Here, using NRW or leakage as a
percent of SIV for setting targets and
tracking progress often has seriously
misleading consequences. If you doubt
this, set up and enter some fgures in a
simple Water Balance.
You can use real or imaginary
volumes of SIV and Consumption
for year one and year two, then
calculate NRW volume as (SIV –
Consumption). Each of these volumes
– SIV, Consumption and NRW – can
change by different positive or negative
percentages from year to year, as in the
In contrast, whatever volumes you
enter, the NRW and Consumption as
percentages of SIV for year one and
year two on the last two rows must
always total 100 percent of SIV, so their
percentages changes will always be
equal and opposite (except when there
is no change in water balance volumes
from year to year, when they will both
be 0 percent).
Often NRW volume will reduce
but NRW as a percentage of SIV will
increase, as in the above example.
If you increase consumption but
don’t change NRW volume, NRW as
percentage of SIV will decrease. Now
try increasing (or decreasing) both NRW
and consumption volumes by the same
percentage, and compare the results.
Why not play the percentages game
and try other options. Simple Zero Sum