Effect of source water blending on iron and lead release: Thermodynamic and statistical modeling

The effects of distribution water quality changes caused by blending different source waters on iron/lead corrosion and release to drinking water were systematically investigated based on pipe surface characterization, equilibrium modeling, and statistical analysis of pilot observations. A large-scale pilot pipe distribution system was constructed including seven water treatment facilities, eighteen pipelines, and corrosion loops.; XRD indicated that FeCO₃, β-FeOOH, α-FeOOH, γ-Fe ₂O₃, Fe₃O₄ were the primary corrosion products on cast iron, while PbO, PbO₂, PbCO₃ and Pb ₃(OH)₂(CO₃)₂ were the main corrosion products for lead. No obvious crystalline corrosion products were found on lined iron and PVC pipes. XPS verified the results of XRD except that no significant FeCO₃ was found by XPS since FeCO₃ is very unstable and could transform into Fe2O₃ easily once the coupons were taken out from aqueous environment. Fairly good linear correlations were observed between iron release and XPS peak area % of Fe₂O₃, and between lead release and XPS peak area % of Pb₃(OH)₂(CO ₃)₂. According to our experimental configuration, diffusion theory, and maximum lead concentrations predicted by thermodynamic model, lead release was estimated to be 0.2 μg/L for typical pilot conditions, which was consistent with the majority of pilot observations that were below the detection limit of AAS (≈5 μg/L). SEM pictures showed some evidences that sulfate and/or chloride affected the surface structure of corrosion scales and consequently influenced lead release, which may explain the difference between model predictions and pilot observations.; The pilot study showed that water quality (i.e., pH, alkalinity, chloride and sulfate) could influence lead release significantly. A statistical model was developed based on one-year pilot study data and agreed relatively well with actual data. A series of additional three-month controlled studies verified the trends observed in previous pilot study and statistical models. The results suggested that surface water has low potential for lead release because of its high sulfate concentration, but groundwater and RO desalted water have higher lead release potential because of low sulfate, high chloride or low pH. (Abstract shortened by UMI.)...