Date of Award
Summer 8-15-2022
Degree Name
Doctor of Philosophy (PhD)
Degree Type
Dissertation
Abstract
Lead is a toxic heavy metal and legacy contaminant in drinking water that can be released from lead service lines and lead-containing premise plumbing. Lead concentrations in water are sensitive to changes in water chemistry. The extent of lead release is often driven by the solubility of the corrosion products formed on the interior of the pipes. Therefore, it is critical to advance our understanding of the extent of lead released from different lead-based plumbing materials at equilibrium and the ability to control lead when it exceeds the regulatory levels. Lead measured in the 1st liter of water collected from a tap can be significantly different from the lead measured in the water parcels beyond the 1st liter. The differences in the lead measured are usually driven by the differences in the types of pipe materials in contact with water during stagnation in the overall configuration of a service line that supplies a residence and the premise plumbing within the residence. The lead measured at the tap is regulated by the United States Environmental Protection Agency under the Lead and Copper Rule of 1991, which stipulates that the 90th percentile of lead concentrations should be below 15 µg/L. Recent revisions to the LCR introduced a 5th liter sample collection in addition to 1st liter sample for monitoring lead at residences with lead service lines (LSLs). However, there is limited information regarding the extent to which the 1st and 5th liter lead concentrations will differ. A systematic approach based on equilibrium solubility modelling can be used to accurately predict the lead in the 5th liter if the water stagnates within the LSL long enough to reach equilibrium. A national database was created with information on the 90th percentile 1st liter lead and key water chemistry variables from 434 very large public water systems in the United States in 2019. The solubility-based model was used to estimate dissolved lead concentrations in equilibrium with the lead corrosion products predicted to form in these systems. Solubility-based predictions exceeded the 90th percentile 1st liter lead for 98.9% of systems. While this was approach can be used for making conservative estimates for the 5th liter, it cannot be used to predict lead in the 1st liter. Process-based factors such as rates of dissolution, mass transfer and water usage patterns in addition to the water chemistry usually drive the 1st liter lead. In the absence of historical data required to implement a process-based model, an empirical approach built upon artificial neural network model was used for predicting 1st liter lead. Lead concentrations in drinking water can be lowered with the help of corrosion inhibitors. Sodium silicate is one such corrosion inhibitor and in typical deployments its addition results in an increase in the pH of water that can lower the solubility of lead corrosion products in addition to an increase in dissolved silicate. However, the role of sodium silicate independent of the pH-based lead control is not well understood. Pipe loop experiments were used to determine the effect of sodium silicate on lead release from LSLs harvested from a water system that has one of the Great Lakes as its source water. The LSLs were first conditioned with a synthetic water similar to the source water chemistry. The conditioned LSLs were then tested at two pH conditions: the original source water pH and an elevated pH. The corrosion scales were periodically assessed to investigate silicate accumulation in the scales. A similar approach was also used to investigate the effect of sodium silicate on lead release from new copper pipes containing lead solder. At the original source water pH and dissolved silicate concentrations of 0.33 mM, sodium silicate effectively lowered lead release from LSLs to lead concentrations below 10 g/L, but it did not have any effect on the lead release from copper pipes containing lead solder. To further probe the mechanism of corrosion inhibition exhibited by sodium silicate within LSLs, supplemental batch-scale experiments were carried out to explore the interactions between dissolved silicate and corrosion scales developed inside of LSLs. These batch scale experiments complemented the previous pipe loop experiments by allowing a focus on chemical interactions in the absence of any mass transfer limitations. Corrosion of service lines and premise plumbing can be sensitive to changes in the source water chemistry. Despite the resiliency that introduction of alternative water sources (e.g., direct potable reuse and desalination) into existing drinking water distribution systems offers, blending water supplies can impact the metal release and microbial communities within premise plumbing. Long-term pipe loop experiments with new copper pipes containing lead solder and with new brass rods with 3% lead were carried out to study the impact of gradual and abrupt introduction of advanced treated water (ATW) on drinking water quality with a primary emphasis on release of metals. The pipes were conditioned for six months with the baseline water (BLW) supply. Since the as-received ATW is highly purified with a low pH (<7) and low alkalinity (<10 mg/L as CaCO3), the ATW was stabilized with a calcite contactor before being blended with BLW. The effect of pH, alkalinity, monochloramine concentration and chloride/sulfate concentration on lead, copper and zinc release were monitored. To investigate the microbial diversity, samples were collected for DNA extraction and 16S rRNA amplicon sequencing during conditioning and blending. Introducing 100% ATW resulted in a dramatic increase in lead release and simultaneous decrease in copper release from the copper pipes with lead solder. The lead release was driven by galvanic corrosion, which was induced by the lowering of sulfate concentrations in the 100% ATW. Statistical analysis was carried out to determine the significance of the changes seen at the different stages of the experiment with different blends. The pipe loop experiments were supplemented with pipe scale analysis to detect changes in the mineral composition, scale morphology and elemental composition of scales after blending. The results from this study can be useful for water utilities that are considering potable reuse as they develop strategies to mitigate any adverse impacts of blending.
Language
English (en)
Chair
Daniel E. Giammar
Committee Members
Fangqiong Ling, Young-Shin Jun, Zhen (Jason) He, Jeffrey G. Catalano,