Report No FR0361

Apr 1993



The customer will benefit from using fluctuations in external parameters to predict an increased probability of coliforms in the distribution.

Identifying associations of the presence of coliforms with external parameters may provide information on the root causes and the origin such that steps may be taken to eliminate the problem.


To assess the significance of bacteriological failures in water supply and to evaluate available information on associated factors. In particular, assess associations between intermittent failures and external physical, chemical and biological factors which may promote release of coliforms from biofilms.


Bacteriological failures in samples of water collected from distribution systems can occur for a variety of reasons such as problems at the treatment works, infiltration and regrowth of organisms normally considered to be of some sanitary significance. Current monitoring procedures are unable to discriminate between these causes of failure and as a consequence remedial action has to be taken on the assumption that the failure indicates a risk to public health.

Better information on the source and type of organism associated with bacteriological failures would assist in deciding on the correct remedial action. In addition, identification of any physical, chemical or biological factors which may cause coliform release from biofilms would facilitate implementation of operational processes to minimise compliance failures.


Further associations of higher or lower values of certain external parameters of physical, chemical and biological nature have been identified. Of interest is the association of lower total chlorine concentrations (mean typically 0.023 - 0.057 mg l-1 lower) with the presence of coliforms in the water samples from some of the water companies. Data from other water companies exhibited no statistically significant association of total chlorine concentration with coliforms and for one water company a statistically significant association of higher total chlorine concentrations (mean 0.028 mg l-1 higher) with the presence of coliforms was observed. One day, two day and three day plate counts were typically higher (geometric means up to 25 times larger) for samples with coliforms compared to those without. These are statistically significant for the major proportion of data sets investigated. No statistically significant associations between the presence of coliforms and pH, concentrations of nitrite and nitrate ions have yet been demonstrated.

The Water Company A results are particularly interesting in that there appear to be differences between operational hydrants and statutory taps (both fixed and random) on the customers' premises. First the proportion of samples registering coliforms in each zone increases for the customer taps relative to the operational hydrants. Second, the presence of coliforms in the operational hydrant taps shows an association not only with lower total chlorine concentrations but also with higher plate counts (3 day, 22C). In the taps on customers' premises these statistical associations are not observed. This suggests an increase in coliforms somewhere within the distribution between operational hydrants and consumer taps that is not related to plate count or total chlorine concentrations.

Analyses of associations of coliform failures with [Al3+], [Fe3+], [NH4+] and conductivity have also been performed. No statistically significant associations were found except for Fe3+ in Water Company G where higher concentrations were associated with the presence of coliforms compared to those from samples without coliforms. Results for these external parameters will be fully discussed in the Final Report.


The following recommendations are made for further work:

  1. Assess associations between coliforms failures and TOC, source water type and water treatment process for individual zones.

  2. Confirm associations between external parameters and coliform failures identified for 1990 and/or 1991 data using data obtained from the same and/or other zones for 1992.

  3. Compare statistical parameters (geometric mean and logarithmic standard deviation) describing lognormal distributions of coliform concentrations with failure rate. Simulate the benefits of analysing larger volume water samples for statistical distributions with different geometric mean and logarithmic standard deviation.

  4. Assess feasibility of prospective studies using a pilot plant.

  5. Identify operational factors to minimise compliance failures.


This Progress Report details statistical analyses performed to date to identify associations of external parameters with water samples registering positive for total coliform bacteria. Statistical distributions of values for external parameters from samples with and without coliforms are compared graphically as probability plots.

Section 2 describes the graphical presentation and statistical treatment, discussing the use of probability plots in determining the most appropriate and best estimates for statistical parameters, namely the mean and standard deviation. Application of the z test is also described. Section 3 reports results of statistical analyses of associations between the presence of coliforms and external parameters with particular attention to concentrations of plate counts and total chlorine. Results for concentrations of nitrate (III) and (V), pH and temperature are also reported. Results of analyses of associations of coliform failures with [Al3+], [Fe3+], [NH4+] and conductivity have been performed and will be discussed fully in the Final Report. Data have not yet been received from one company. These will also be presented in the Final Report.

Copies of the Report are available from FWR, price 25.00 less 20% to FWR Members