Statistical Analysis of Seven Water Efficiency Studies using Control Groups and Weather Influences (2013)

You are here:
  • Main
  • Waterwise Publications
  • Statistical Analysis of Seven Water Efficiency Studies using Control Groups and Weather Influences (2013)
← All Topics

This project by Essex & Suffolk Water was concerned with analysing and reporting the extent to which external factors influence demand during periods when water efficiency studies are undertaken. Seven studies were included in the analysis, both domestic and commercial retrofit schemes. Measured water consumption data were used to calculate the mean pre and post intervention consumption per day at each property. The difference between the pre and post consumption was calculated for each property.

Executive Summary

Essex & Suffolk Water (ESW) has routinely carried out water efficiency projects for the last 15 years. As part of these schemes the Company has installed thousands of water efficiency devices and encouraged customers to embrace water saving habits through behavioural change campaigns. These initiatives are monitored through the collection of three separate meter reads; these are used to calculate overall study savings. Through this process ESW hasproduced a measurable decrease in their customers’ consumption.

However, ESW wanted to explore whether any further value can be extracted by re- analysing the results. This project was concerned with analysing and reporting the extent to which external factors influence demand during periods when water efficiency studies are undertaken. If external influences were found to be statistically significant a method of correctly adjusting for them was to be developed and reported upon in order that ESW staff can make use of the method on the analysis of future studies.

Objectives

The study objectives were as follows:

  1. a)  Calculate the overall savings achieved using the meter measurements provided for each study.
  2. b)  Characterise the pre and post measurement periods for each study.
  3. c)  Analyse and compare study group demographics with those contained within thecontrol group PCC properties.
  4. d)  Select most appropriate PCC pre and post consumptions for each study and test forstatistical significance.
  5. e)  Where appropriate use the calculations in d) above to adjust the results from a)above.
  6. f)  Where PCC adjustments have been made compare weather conditions for thoseperiods as evidence to support the reasons for the demand changes.

Approach

Seven studies were included in the analysis and these included both domestic and commercial retrofit schemes. Measured water consumption data was used to calculate the mean pre and post intervention consumption per day at each property. The difference between the pre and post consumptions was calculated for each property.