True exposure to lead at the tap: Insights from proportional sampling, regulated sampling and water use monitoring

This work aimed to evaluate the ability of four regulatory sampling protocols to accurately determine weekly water lead levels (WLLs) of exposure at the kitchen tap in twenty-nine households with or without a lead service line (LSL). Proportional sampling was used as the gold standard to which the other protocols, 5-min flush, 30-min stagnation, 6-h stagnation and random daytime were compared. Random daytime samples provided mean WLLs closest to true exposure in the households monitored overall compared to other sampling protocols. Strikingly, mean WLLs after 5 min of flushing underestimated lead exposure by 47%. Supporting these observations, water usage patterns revealed that full flushing only occurs in 3.4% of usage events within the service line and in 0.26% at the tap. The time between usage events in the service line was approximately 30 min but the 30-min protocol tended to slightly underestimate WLLs. These differences were explained by flushing prior to the 30-min stagnation sampling, which limited the contribution of the LSL to WLLs. Furthermore, the average stagnation at the kitchen time was 106 min and usage events rarely exceeded the water volume within premise plumbing (1.1 L). Mean WLLs after 6 h of stagnation without flushing overestimated exposure by 29% but provided a conservative indicator of WLLs of exposure. •High WLLs were more related to water use at the kitchen tap than to use of the LSL.•Inter-use stagnation at the kitchen tap is much greater than in the LSL.•RDT sampling provided mean WLLs closest to real exposure.•30MS and 5MF underestimated exposure while 6HS overestimated it.•Premise plumbing was confirmed as a significant source of lead exposure.