Citation:

Rego, C. A. AND M. R. SCHOCK. Discovery of Unforeseen Lead Level Optimization Issues for High pH and Low DIC Conditions. Presented at AWWA Water Quality Technology Conference, Charlotte, NC, November 04 - 07, 2007.

Impact/Purpose:

Present at conference

Description:

A large northeast water utility serving over 500,000 retail and wholesale customers had historically been slightly below the 90th percentile Action Level for lead. The system had been operating at a pH of approximately 10.3, a DIC concentration of approximately 5 mg/L as C, and using free chlorine as a disinfectant. A further reduction of their lead levels and better optimization of their corrosion control was desired, to provide a higher level of public health protection. A desktop study was conducted to review historical water quality data, lead and copper sampling results, existing treatment, and system constraints to evaluate possible corrosion control treatment processes. Experiences of similar water systems were also reviewed. Theoretical modeling was used to simulate alternative treatment conditions for pH, alkalinity, and DIC. The results predicted that a 15% - 40% reduction in theoretical lead (II) solubility was possible by lowering the pH to a more-optimal range of 9.5 - 9.7 at the same DIC. After receiving primacy agency approval, the water utility slowly reduced the pH over a two and a half month period to a target pH of 9.7. After approximately one year at pH 9.7, six lead service line (LSL) specimens were collected from various areas of the distribution system and analyzed to investigate possible transitions in scale mineralogy and chemistry. These results could also be compared to the evolution of lead levels in tap monitoring, and lead profiles from select premises. Surprises in the mineral assemblage relative to classical Pb(II) solubility theory were immediately apparent. The scales were composed almost entirely of Pb(II) hydroxycarbonate and carbonate phases. However, qualitatively, there was a diverse range in relative ratios of plumbonacrite, hydrocerussite, and cerussite. The existence of considerable plumbonacrite strongly suggests that published estimates of the solubility constant for this mineral are in error, since it is not predicted to form at pH conditions as "low" as the historical treatment in this utility. More accurately determining the stability domain and decomposition pathway for plumbonacrite takes on great importance for high-pH utilities. Lowering the pH could inadvertently destabilize microcrystalline plumbonacrite, before it has the opportunity to recrystallize into more stable and more crystalline hydrocerussite, and cause elevated Pb release. Further, the truly optimal combination of pH and DIC for minimizing plumbosolvency in the current models may be erroneous, and needs to be revisited to properly revise treatment guidance.

Record Details: