Background Notes

 Introduction

The data used to compile this release were received by the CSO from Irish Water in response to a request made under the Statistics Act, 1993.

Revisions

All figures are subject to revision. Water meters can store data for previous readings and they are typically read using drive-by technology. Not all meters may have been read for the full release period e.g. if the last data collection for an area was done during December then data for December will not be available until the meter is read again and the data are made available to the CSO. Revisions for any missing readings will be included in the next release e.g. meters with no December data.

Thresholds

Table 3 in the release examines the impact of using consumption thresholds to exclude data that are unusually high. These thresholds show the substantial impact that high usage has on the average consumption figures. Typically high consumption is due to a leak. The difference between the average and the median gives an indication of the extent of high consumption as the median is less influenced by the small number of meters with high consumption. In Table 3 the median is calculated based on all 2015 data for a meter whereas in Tables 1, 5B, 6B and 7 it is calculated from monthly data.

Coverage

The data used in this release only cover metered usage. Apartments, many group water schemes, and abstractions from wells are not included. The January 2015 data are based on 519,781 meters. By December this had increased to 771,348 meters. Data from meters which were installed during the year have been included provided a full month's data could be calculated from the meter readings.

Methodology

Irish Water provided monthly files of meter readings to the CSO. The files contained both meter readings and derived consumption data. The dates of the meter readings are generally from a date that is at the end of the month but readings can also occur on other days of the month and in some cases a meter may not have been read in a particular month.

In a small number of cases, a meter was read more than twice in a month. In these cases the CSO has used the first and last readings in a month to calculate monthly consumption.

Meter readings that were irregular in the series, e.g. readings lower than previous readings, have been excluded. Before excluding any readings, the CSO has examined in detail the series of readings for meters where there was a decrease of 100 cubic metres or more between two readings. The higher reading was excluded if it was the irregular reading in the series. Only usage figures for a full month have been included in this release e.g. a meter that was last read before the end of a data month is excluded for that month.

The CSO has converted the readings into monthly usage figures by allocating the difference in two readings proportionately between the dates that the meters were read. Months with partial data have been excluded e.g. if the last reading for a meter was during December, and there were no meter readings available yet from the following year, then that month was excluded. The first month's reading for every meter has been excluded as the CSO does not know the period covered by that reading - for the 2015 release many of these first months occurred in 2014.

Measurement units

The meter readings are in cubic metres. One cubic metre equals one thousand litres. The readings in cubic metres have been converted to average and median consumption in litres per meter per day. In Table 1, the data were converted into average and median consumption per meter per day in each month.

County data

There was a substantial meter installation programme during 2015. This did not happen concurrently in all counties. In a small number of cases, we have excluded all data for a county for a month because there were only a small number of meters installed e.g. county Cavan in January and February 2015. Similarly there were a relatively small number of meters installed in Dublin 17 and Dublin 24. The consumption levels in these two postal districts was considerably higher than elsewhere and considered to be not representative of all households in the two areas. Accordingly all data for 2015 for these two districts have not been used in the release.

Hide Background Notes