Modeling Soluble and Particulate Lead Release into Drinking Water from Full and Partially Replaced Lead Service Lines

Partial replacement of lead service lines (LSLs) often results in the excessive long-term release of lead particulates due to the disturbance of pipe scale and galvanic corrosion. In this study, a modeling approach to simulate the release and transport of particulate and dissolved lead from full and partially replaced LSLs is developed. A mass-transfer model is coupled with a stochastic residential water demand generator to investigate the effect of normal household usage flow patterns on lead exposure. The model is calibrated by comparing simulation results against experimental measurements from pilot-scale setups where lead release under different flow rates and water chemistry scenarios was reported. Applying the model within a Monte Carlo simulation framework, partial replacement of the LSL was predicted to result in releasing spikes with significantly high concentrations of particulate lead (1011.9 ± 290.3 μg/L) that were five times higher than those released from the simulated full LSL. Sensitivity analysis revealed that the intensity of flow demands significantly affects particulate lead release, while dissolved lead levels are more dependent on the lengths of the stagnation periods. Preflushing of the LSL prior to regulatory sampling was found to underestimate the maximum monthly exposure to dissolved lead by 19%, while sampling at low flow rates (<5.2 LPM) was found to consistently suppress the high spikes induced by particulate lead mobilization.