Economic Leakage

Economic Leakage Level

Assessing Economic Level of Real Losses (ELL)

Assessments of ELL in England and Wales, since the mid-1990’s, have tended to concentrate almost exclusively on assessing the economic frequency of Active Leakage Control of unreported leakage, using a Minimum Total Cost approach, as shown in the graph below. For different levels of losses, the cost of lost water (the rising line) is added to the annual cost of active leakage control (which is influenced by the frequency of ALC interventions) to obtain a ‘dish-shaped’ Total Cost curve. The short-run economic level of leakage is then assumed to correspond to the lowest point on the Total Cost curve.

The short-run economic level of leakage is then assumed to correspond to the lowest point on the Total Cost curve.

The UK approach – which is basically just a calculation of Economic Intervention frequency – is based on continuous night flow measurements in numerous district metered areas (DMAs). However, it does not take account of how the management of excess pressure can impact all aspects of calculation of ELL through FAVAD N1 pressure:leak flow relationships, Pressure and Bursts relationships, and standards of service for repair times of reported leaks.

Simpler less data-hungry methods for calculating Economic Intervention Frequency, based on Rate of Rise of Unreported Leakage, and more applicable to international situations, were incorporated into calculation of Economic Leakage Levels outside the UK (2005M and 2005Q). The results obtained by the ‘Rate of Rise’ approach can also be shown in the form of minimum total cost curves. With the publication of the IWA Pressure and Bursts data for the 110 systems from 10 countries in 2006, Fantozzi and Lambert (2007R) considered that future ELL calculations ‘must surely include allowances for the influences of pressure management, if they are to be meaningful’.

ELL calculations are not widely used outside the UK, but specialised ILMSS software was developed and used for ELL calculations taking into account the effects of pressure management including burst reductions, for Sydney Water in Australia, and for several Utilities in Canada.  The graph below (from 2009E, for one of the Canadian Cities)  includes confidence limits, and shows how much difference there can be between Short Run ELL (SRELL) based only on economic intervention at current pressures (blue graph), and SRELL which includes the influence of pressure management.

SRELL calculations for a Canadian City, with and without pressure management.

This graph (from 2009E, for one of the Canadian Cities) includes confidence limits

 Source: Fanner and Lambert, 2009, using ELLCalcsV2a software

In the UK, a Report jointly commissioned by OFWAT, Defra and the Environment Agency on ‘Sustainable Levels of Leakage’ (SELL), published in 2012, concluded that ‘the degree to which pressure management can contribute to reduced leakage economically seems to have been underestimated. The SELL should be reassessed…… so that the SELL is consistent with what pressure management has achieved since.’

This conclusion relates to the UK situation that ELLs calculated for England/Wales Utilities in the late 1990’s were principally based only on assumptions of economic ALC intervention, and had not taken account of the substantial amount of pressure management carried out in the UK since 1995, after mandatory leakage targets were introduced in England and Wales. However, if OFWAT had accepted and implemented a House of Commons Environment Committee recommendation after the 1995/6 drought, to publish average pressures alongside Company leakage data (see KPI), this particular problem might have been identified at least 15 years previously.

This highlights one of the key problems with the overall concept of calculating Economic Leakage Levels. The validity of assessments of ELL can be as uncertain as calculations of ‘Unaccounted for Water’ or ‘% losses’ – the result depends to a substantial degree upon who did the calculation, how the calculation was done, what key factors were included and excluded, and what assumptions were made.