An investigation into the representativeness of random daytime sampling for lead in drinking water, using computational modelling

Random daytime (RDT) sampling for lead in drinking water has been used in the UK since 1989 for regulatory compliance assessment and since 2004 in the Netherlands. In 2008, RDT sampling was recommended to the European Commission as the method to be used throughout the European Union but there are concerns about the representativeness of this method, being of relevance to the protection of human health in water supply systems. This issue has been investigated using an established computational modelling system, from which it was concluded that: (i) RDT sampling as practised in the UK is adequately representative of the range of circumstances that occur; (ii) for houses with daytime residency, RDT sampling is not sensitive to the time period of sampling, unless it is constrained to only a few hours; (iii) for houses without daytime residency, RDT sampling of houses elsewhere with daytime residency is adequately representative, for the total periods of water use; (iv) for houses without daytime residency, random sampling just before or after ‘normal office hours’ will not be representative for checking zonal compliance; (v) it is important that seasonal variation is accommodated; and (vi) adequate reproducibility can be achieved if at least 100 samples are taken annually and if results are aggregated for several years.