Water Quality Index: Communicating System Performance

Water quality indexReporting water quality performance to senior management or customers can be problematic as it requires a myriad of numbers and difficult to pronounce parameters—impossible words such as clostridium perfringens, polydiallyldimethylam-monium or bromochloroacetonitrile are not part of the vocabulary of most people.1 Directors and customers of utilities are generally not water quality specialists and that need to be provided with easy to digest information for them to be able to assess how drinking water supply systems perform. To achieve this goal, a water quality index is currently in development.

The index is aimed at reducing complex data matrices to a single number, combining information from various sources. The index provides an overview of water quality performance, without mentioning technical details. The overall index consists of five parameters: treatment effectiveness, network protection, regulatory compliance and customer perception.

Given the broad nature of these parameters—from subjective assessments by customers to objective laboratory data—a certain level of subjectivity is unavoidable. The different aspects of the index will not contribute equally to the overall performance of water supply: How should we view customer complaints in relation to laboratory data?

Methodology

A crowd-sourcing tactic was employed in the form of a survey to seek the collective opinion of water quality experts.2

Respondents were asked about their involvement in water quality (such as level of education and amount of experience in the field). The main survey consisted of two question banks regarding the relative importance of each of the proposed index factors and network sub-factors. Data was analysed using the using statistical package R.3 Responses can be considered reliable as the average standard error is less than 5%. The complete survey results and detailed analysis can be viewed on Rpubs. The raw scores on the main questions are presented in the diagrams below. The levels on the Y-axis are the relative importance (0–100) given to each of the parameters by respondents.

Water Quality Index survey results (n=36). Click on diagram for high resolution image.Analysis

The individual results regarding the relative importance of the individual factors and sub-factors are self explanatory. The final index scores will be weighted in accordance with these survey results. Additionally some meta-analysis has been undertaken to obtain insight into the complexities of assessing water quality performance. Factor analysis with varimax rotation revealed that a one-factor solution is capable of explaining 49% of the variance. This is an indication that questions were answered consistently among respondents and that item scores can be interpreted as originating from one latent variable, i.e. water quality performance.

Ten respondents also provided additional comments regarding the water quality index. Some respondents mentioned that the questions were “simplistic”, “ambiguous“ and “inaccurate”. This problem is, however, inherent to the data reduction and simplicity objectives of the water quality index. The index’s ambiguity and inaccuracy are a reflection of the fact that information is sourced from paradigmatically different sources such as customer feedback and laboratory results.

Due to the reduction in data complexity, the index, its factors and sub-factors cannot be used for quantitative analysis. The index is in essence a qualitative expression of water quality performance only suitable for communication and not for analysis.

One respondent also commented on the relationship between physical and biological water quality parameters and customer’s perception of these:

Focus on water safety sometimes gets clouded by issues associated with customer aesthetic opinion.

This is an expression of the water quality paradox. Even if the quality of water is in accordance with regulations, customers might still not be satisfied. Providing safe water is a necessary condition, but not a sufficient condition to achieve customer satisfaction.

Conclusion

The survey has been successful and will aid in completing a water quality index that reflects the relative importance of the different aspects of water quality.

The comments made by water quality experts are a common expression of the difference in thought worlds between scientists and customer service professionals and aid in further developing a theoretical model for organisational culture in water utilities.

Notes


  1. These terms appear in the Australian Drinking Water Guidelines (2011)—Updated December 2013. 

  2. A total of 36 responses were received from Australia, New Zealand, the USA and Europe. The survey was closed on 31 January 2014. Questions can be viewed as a pdf file

  3. R Core Team. (2013). R: A Language and Environment for Statistical Computing. Vienna, Austria. Retrieved from www.R-project.org

Leave a Reply

Your email address will not be published. Required fields are marked *