Leaching of Heavy Metals Due to Changing Disinfectants in Drinking Water Distribution Systems

To comply with increasingly stringent controls of disinfection by-products, many utilities are converting from free chlorine disinfectant residual to chloramines in their drinking water distribution systems. At certain utilities conversion to chloramines has caused episodes of lead release resulting in lead concentrations far exceeding the action level of the USEPA Lead and Copper Rule (LCR) observed at household taps (HDR/EES, 2007). This investigation (Awwa-Research Foundation #3107) helps develop a more complete understanding of the effect of changing disinfectants on the levels and rates of lead and copper leaching from lead, copper, and bronze plumbing materials under a variety of water quality and distribution system conditions, so that future metal release events can be anticipated and avoided as more utilities convert to chloramination.To explore these issues, four pipe loops were constructed using representative plumbing materials at the HDR ARTC laboratory facility in Bellevue, Washington. The four loops contained lengths of unpassivated lead, copper, bronze, and passivated lead pipe. The passivated pipe was removed from a Seattle-area distribution system, and PbO2 similar to that observed in systems where chloramines-associated lead releases events have occurred was identified as the predominant constituent of the existing passivating scale on the pipe interior. Each loop was fitted with electrode ports for electrochemical (EC) corrosion analysis, and also contained an inline reservoir that housed penny-sized coupons of the corresponding loop material(s). These coupons were periodically removed for passivation scales analysis using scanning electron microscope (SEM). Observed formation and morphological transformations of solid corrosion product scales on coupon surfaces were correlated with EC measured corrosion rates and metal concentrations as disinfectants were alternated.The yearlong project was divided into four, ten-week test periods, each associated with a change in a water quality parameter significant to corrosion. Varied background water quality conditions included pH, alkalinity, and the presence of orthophosphate corrosion inhibitor.During each test period, a cycle of disinfectant changes was implemented. The cycle was comprised of two-week periods of no disinfectant, free chlorine at 1.5 mg/L-Cl2, and monochloramine at 3.5 mg/L-Cl2. Water was changed weekly. Grab samples were taken each week, and analyzed for total lead and total copper concentrations.The passivated lead pipe containing Pb(IV)-based PbO2 scales exhibited significant suppression of lead concentrations during free chlorine periods, and large increases in lead concentrations during monochloramine and no-disinfectant cycles. Measurements of oxidation-reduction potential (ORP) taken during the monochloramine and no-disinfectant periods were consistently within the pH/ORP region where PbO2 becomes thermodynamically unstable.Lead release from the new lead pipe showed minimal sensitivity to changes in disinfectant. SEM examination revealed that Pb(II)-carbonate scales formed rapidly on the new lead pipe surface, and trends in release were consistent with predictions based on Pb(II)-based solubility calculations.Copper release trends from the copper and bronze pipes indicate that copper release is sensitive to the presence of a disinfectant, but that changes between chlorine and chloramines had minimal impact on release levels. Copper and lead release from the bronze pipe were similar to those observed for the copper and new lead pipes, and were not notably impacted by changes in disinfectant.The results of this investigation have meaningful consequences for utilities considering conversion from free-chlorine disinfectant to chloramines. Any utility with a history of high ORP associated with free-chlorine residual should determine the composition of passivating scales on lead and lead bearing components in the distribution system prior to a conversion. If lead bearing materials in the distribution system have existing Pb(IV) scales that formed during historical high-ORP periods, a reduction in ORP associated with conversion to chloramines could result in increased lead release events. Determining the nature of existing passivating scales can be achieved by surface analysis. Interpretation of electrochemical-based prognostic methods is not always straight forward, and further investigation may be required before such tests can be reliably used to anticipate metal release events associated with changing disinfectants.