"General anomalies that occur in the management of apparent losses are identified and include determination of apparent loss indicators that fail to take into account both the random and bias errors inherent in metering. Estimating the volume of water losses that could be attributed to meter errors is also incorrectly assumed to be equivalent to the error limits (i.e. envelope) stipulated in standards. Performance indicators that are reliant on flow meter data must be expressed as a range and not as a single value which is common practice. It is shown that the measurement error range decreases as the number of meters increase when combining both their bias and random errors. A default value for reference annual apparent losses (RAAL) is quoted as 5% of the authorised consumption unless historic records can demonstrate otherwise. This 5% value relates to water meter errors that are only one component of apparent loss contributing to a volumetric imbalance and still requires inclusion of random errors in its calculation and therefore should be reported in the form of a range. Meter measurement errors include those of the meter fleet (e.g. revenue meters) as well as large (bulk supply) meters.
An Australian case history describes how an Apparent Loss Minimisation Model considers the influence that sequencing of the meter replacements has on determining the optimal meter replacement interval for a cohort of meters as well as how the model facilitates the defining of related metering components of apparent loss indicators. The model considers the justification for the replacement of water meters that compares the savings (e.g. benefits) achieved from an improvement in the measured volumes of water due to the installation of new meters (e.g. savings in apparent losses) with that of the costs of the associated meter replacement program. The availability of the Apparent Loss Minimisation Model in a user friendly software format together with a comprehensive manual tailor-made to a specific water utility’s organisational and operating conditions facilitates the analysis of various meter replacement scenarios for planning purposes and the preparation of budgets.
A logic diagram provides a useful illustration that assists with understanding how the model allows for the meter replacement sequence. Sensitivity analysis shows that the apparent loss volume is generally more sensitive to variations in input parameters that have a direct influence on the volumetric amounts registered by the old meters and the optimal replacement period has greater sensitivity to variances in inputs that influence the total cost of apparent losses.
Another case history describes the application of a Metering Lifecycle Model that assists with comparison of various types of metering technologies to determine which technology would provide the best “value for money”. The model also quantifies the apparent losses associated with errors occurring during the collection and transmission of data. Application of the Metering Lifecycle Model facilitates both the establishment of apparent losses for the initial part of the data pathway as well as the costs for potential failures in particular technologies. Failure of some meter reading technologies can potentially result in a ‘surrogate’ apparent loss of 13 L/connection/d if it is assumed that non-defaulting customers are required to pay for these costs through full cost tariffs."