Report No DWI0773

May 1997

BACKGROUND

Lead in drinking water arises almost exclusively through contamination from lead pipes or other lead-containing materials. Lead in drinking water can be reduced by water treatment to reduce pick-up of lead, or practically eliminated by replacement of lead pipework and other lead-containing fittings. The concentration of lead in drinking water is strongly influenced by the standing time of water in lead pipework. This means that the method of sampling (e.g. flushed sample, random daytime sample) has a large effect on the measured lead concentration.

The European Commission has proposed a revision of the Drinking Water Directive; for lead it is proposed to reduce the limit from the present 50 µg/l to 25 µg/l for an interim period of 5 years with the ultimate limit of 10 µg/l to be achieved within 15 years of implementation of the Directive. The present standard applies to a flushed sample (the UK standard is numerically the same but is more stringent as it applies to any sample of the first water that issues from the tap). The proposed standards are to apply to "representative" water samples. Whilst the precise form of the proposed standards and monitoring requirements have yet to be determined, pending the outcome of an EC study, it is possible to make estimates of the likely compliance situation based on various assumed scenarios.

OBJECTIVES

The alms of this project are to estimate the likely extent of non-compliance with the interim and final proposed standards of 25 and 10 µg/l for various monitoring methods and interpretations of the standards; to estimate the extent to which further water treatment or pipe replacement would be needed to secure compliance and to calculate the costs of complying with the various scenarios; and to undertake a critical assessment of the various possible sampling regimes.

APPROACH AND METHODOLOGY

Data used for the study included information returns obtained from the Drinking Water Inspectorate, as well as water company responses to a questionnaire dealing with lead pipe occurrence, costs and the results of "special" lead surveys. For each proposed standard, three sample types were considered (random daytime, 30 minute stagnation and composite proportional) with three rules for pass/fail (maximum, percentile, mean). Comparative data on results for different sample types were used to derive relationships between measured random daytime concentrations and the concentrations that would be expected in the other types of sample. The zones which would pass or fail each scenario were then estimated based on the results of statutory monitoring, classification of lead risk for the zone, and the proportion of lead pipes. A simple model of the effects of treatment was used to determine whether treatment would be appropriate (if not already installed and for zones with a reasonable population of lead pipes) or if replacement would be necessary. This information was used together with unit cost data to estimate the costs of compliance for each scenario.

In a study of this type it is necessary to make various assumptions; e.g. in assessing the extent of compliance with various scenarios and calculating the costs of remedial measures. The key points to note are:

• statistically determined relationships between lead concentrations in different types of sample, together with compliance sampling data, were used to identify those zones which would pass or fail various scenarios;

• the suitability of water treatment to achieve compliance with each scenario was assessed using a simplified model of the effects of treatment on lead solubility, and various assumptions concerning the ability of treatment to achieve a 10 µg/l limit were included in the cost calculations;

• it was assumed that all lead pipes would be replaced in a zone where water treatment alone would not achieve compliance; and

• pipe replacement was not considered for zones which would pass a lead limit of 10 µg/l, even if these zones contain lead pipes.

The assumptions made are identified in the appropriate section of the report and should be taken into consideration for a full appreciation of the conclusions.

Comparison of results for different sampling techniques showed that concentrations in random daytime samples are in general slightly higher than in 30 minute stagnation samples; composite proportional samples give similar results to 30 minute stagnation; and lead concentrations in flushed samples are sensitive to the details of the procedure.

Copies of this report may be available as an Acrobat pdf download under the 'Find Completed Research' heading on the DWI website.