The Relevance for Non-Revenue Water and Meter Replacement Policies

The Challenge

Household level water leakages are often not registered, and subsequently go undetected by conventional water meters due to the low flows associated with leaks in the domestic system. Currently the entire community pays for this unaccounted water as part of the water utility’s bulk supply charges. The purpose of this project was to obtain a greater understanding of the quantities of water that are not registered and identify a range of possible solutions that address the challenge.

The Project

The project consisted of four key phases.

1. The testing of meter flow rates to establish when non-registration occurs for several meter types (new and aged). Work on meter testing was undertaken by Wide Bay Water in the initial stages of this project and have been reported on in their reports:
   • Understanding Apparent Water Losses Through Non-Registration of Domestic Water Meters, (WBW 2009)
   • Understanding Apparent Water Losses Through Non-Registration of Domestic Water Meters, Stage 2 Refining the Predictions from Stage 1 (WBW 2011)
2. The testing of unmeasured flow reducers, conducted by the Centre for Infrastructure Engineering and Management (Griffith University) which serve to inform this report, viz.:
   • Experimental study on meter registration accuracy at low flow rates and benefits of UFR implementation
   • Determining average volumes of residential water consumption in flow rate interval categories
3. Analysis of flow profiles from South East Queensland and Melbourne to assess the number of water events that fall into low flor ranges (0-5 litres per hour) where non-registration is significant.
4. Determining the implications for meter replacement.

The Outcome

The findings from this study indicate that the non-registration is more significant than under-registration, and that both get progressively worsen with usage, which has specific relevance for the assumptions used in accounting for non-revenue water and for specifying the replacement schedule for a fleet of meters.

In summary the outcomes of the study indicate that:

1. Non-registration is more significant than under-registration in understanding non-revenue water, since it could account for a larger percentage of the non-revenue water passing through a meter (illustrated in Table A). The additional volume associated with the percentage of non-registration can help further explain the apparent losses (customer metering inaccuracies) when calculating the water balance and the overall water loss.
2. Consideration of both the non-registration and under-registration components of the unregistered volume should be made when preparing a meter replacement policy. The percentage of non-registered volume increased more significantly with higher total registered volumes as compared with under-registration (see Table A). The consideration of non-registration therefore has the effect of bringing forward the timing for meter replacements.
3. The unregistered volume increases with the usage of the meter. Therefore using an average for a fleet of meters is not reliable enough, and improved estimates can be achieved by assessing the losses per registration group. This is especially relevant when calculating the water loss due to meter error and/or specifying the replacement schedule.