The EPA Lead and Copper Rule set action limits for lead and copper concentrations in drinking water, but accelerated corrosion of lead in distribution systems due to a galvanic connection to copper. Prior research has demonstrated that the effects of galvanic corrosion can be controlled by water chemistry. This study not only investigated the main effects of alkalinity, natural organic matter (NOM), nitrate, disinfectant and inhibitor to galvanic corrosion, but also the interplay between these factors. A 2-level factorial (2v5-1) design was adopted which resulted in 16 testing conditions. Results of bench-scale experiments using static pipes with lead and copper segments demonstrated that alkalinity, disinfectant, inhibitor and alkalinity-inhibitor interaction had a significant impact on galvanic current. The significant factors affecting total lead release were alkalinity, NOM, disinfectant, alkalinity-inhibitor interaction, NOM-nitrate interaction, NOM-disinfectant interaction, NOM-inhibitor interaction, nitrate-disinfectant interaction and disinfectant-inhibitor interaction.

The EPA Lead and Copper Rule set action limits for lead and copper concentrations in drinking water, but accelerated corrosion of lead in distribution systems due to a galvanic connection to copper. Prior research has demonstrated that the effects of galvanic corrosion can be controlled by water chemistry. This study not only investigated the main effects of alkalinity, natural organic matter (NOM), nitrate, disinfectant and inhibitor to galvanic corrosion, but also the interplay between these factors. A 2-level factorial (2v5-1) design was adopted which resulted in 16 testing conditions. Results of bench-scale experiments using static pipes with lead and copper segments demonstrated that alkalinity, disinfectant, inhibitor and alkalinity-inhibitor interaction had a significant impact on galvanic current. The significant factors affecting total lead release were alkalinity, NOM, disinfectant, alkalinity-inhibitor interaction, NOM-nitrate interaction, NOM-disinfectant interaction, NOM-inhibitor interaction, nitrate-disinfectant interaction and disinfectant-inhibitor interaction.

Lead exposure poses as a risk factor for various adverse health effects including intellectual delays, reduced IQ, and behavioural problems in children, as well as cognitive decline in adults. Lead enters drinking water through corrosion of leaded materials such as lead pipes, solder, and brass devices. Three rounds of residential and non-residential lead monitoring were conducted to evaluate the corrosion control implemented by Halifax Water, and to identify sites with elevated lead concentrations. Follow-up testing was conducted at several sites to determine the sources of lead, and the factors that contributed to high lead release. Finally, a bench scale experiment was conducted to determine the impacts of plumbing flux on metal release. The lead action level for residential testing was exceeded only in the round that was conducted during the winter. Lead concentrations were also higher in the winter rounds than the fall round of non-residential sampling. The seasonal lead variation was likely caused by fluctuations in aluminum residuals in the water leaving the plant. Frequency of use, age, and outlet manufacturer were factors that were associated with elevated lead levels. Follow-up studies were conducted at several fountains to determine the source of elevated lead levels. These fountains typically contained several leaded components and received infrequent use. Fountains with leaded components that received high, regular usage had often provided samples with low lead levels. Drinking fountains that were banned and recalled in the US for potentially containing lead lined cooling tanks were found at eight locations throughout the study area. It was found that three of the eight likely contained the lined cooling tanks. High lead levels were present in samples collected from these fountains, even at sites with frequent usage. Low-use sites with the lead lined tank produced the highest lead levels in this study. Fountains suspected of containing lead lined tanks were removed and replaced, and the lead levels were significantly reduced at these sites. The impact of plumbing flux on metal concentrations was relatively short in duration, lasting only a week for most metals, with the exception of tin. Lead levels were found to stabilize under all flux conditions following roughly 40 L of flushing. Flux type was the main factor contributing to the elevated metals. The traditional petroleum flux was much more resistant to flushing than the water soluble flux, as it caused elevated tin levels for several weeks and a tacky flux deposition in the copper pipe remained even three months after the start of the experiment. The high amount of chloride from the flux was aggressive towards the copper corrosion, but it is unclear if this would have led to copper pitting corrosion.

Total lead release in drinking water in the presence of free chlorine and chloramine residuals was investigated in field, laboratory and fundamental investigations for finished waters produced from ground (GW), surface (SW), saline (RO) and blended (B) sources. Field investigations found more total lead was released in the presence of chloramines than in the presence of free chlorine for RO and blended finished waters; however, there were no statistical differences in total lead release to finished GW and SW. Laboratory measurements of finished waters oxidation-reduction potential (ORP) were equivalent by source and were not affected by the addition of more than 100 mg/L of sulfates or chlorides, but were significantly higher in the presence of free chlorine relative to chloramines. Development of Pourbaix diagrams revealed the PbO2 was the controlling solid phase at the higher ORP in the presence of free chlorine and Pb3(CO3)2(OH)2(s) (hydrocerussite) was the controlling solid phase in the presence of chloramines at the lower ORP, which mechanistically accounted for the observed release of total lead as PbO2 is much less soluble than hydrocerussite. The lack of differences in total lead release to finished GW and SW was attributed to differences in water quality and intermittent behavior of particulate release from controlling solid films.

Corrosion of distribution system piping and home plumbing materials is a major concern in the water community. Iron release adverse affects aesthetic water quality and the release of copper and lead is regulated by the Lead and Copper rule (LCR) and can adversely affect consumer health. Corrosion control is typically done by pH regulation and/or addition of corrosion inhibitors. Monitoring of corrosion control is typically done after the fact by monitoring metal release, functional group concentration of the selected chemical species or water quality. Hence, the associated laboratory analyses create a significant delay prior to the assessment of corrosion in drinking water systems. As corrosion in drinking water systems is fundamentally an electrochemical process, measurement of the electrical phenomena associated with corrosion can be use for real-time corrosion monitoring. This dissertation focuses on using parameters associated with electrochemical corrosion monitoring (EN) measurements in a field facility to predict and control the release of Iron, Copper and Lead in finished waters produced from ground, surface and saline sources with and without usage of corrosion inhibitors. EN data has not been used previously to correlate water quality and metal release; hence the use of EN data for corrosion control in drinking water systems has not been developed or demonstrated. Data was collected over a one year period from a large field facility using finished waters that are distributed to each of the fourteen pilot distribution systems (PDSs), corrosion loops and Nadles each. The PDSs have been built from aged pipes taken from existing distribution systems and contain links of PVC, lined cast Iron, unlined cast Iron and galvanized Steel pipe. The effluent for each PDS was split in two parts. One was delivered to the corrosion loops which are made from coiled copper pipe with lead-tin coupon inserted inside each loop and the other was delivered to the Nadles which housed the EN probes with electrodes for Fe or Cu or Pb-Sn. Finished water quality was monitored in and out of each PDS and total and dissolved Copper and Lead were monitored out of each corrosion loop. Photographs, scanning electron microscope (SEM) micrographs and energy disruptive x-ray spectroscopy (EDAX) conducted on all EN electrodes. EN electrodes showed dark brown to blackish voluminous scales for Fe, and EDAX revealed occurrence of two scales in distinct areas for all Fe electrodes; one comprised of porous, spongy looking structures and scales with more Fe content where the other had denser and more compact scales richer in Ca and P or Si. Cu electrodes had an orange to dark brown thin scale with blue green spots. Small pits were consistently observed mostly in the centre of such blue green spots which were identified as copper carbonates. The Pb electrodes visually showed a thin shiny transparent film with a surface very similar to the unexposed electrodes. Numerous pits were visually for pH controls and not seen for inhibitors; but SEM revealed that all electrodes had pits but the inhibitors reduced number and size of pits compared with pH controls. Thin hexagonal hydrocerussite plates were observed to occur in distinct growth areas and the presence of P or Si inhibitor seemed to increase the occurrence of hydrocerussite. Both Fe & Pb release were mostly in the particulate form while Cu release was mostly in the dissolved form. Total and dissolved Fe, Cu and Pb release models using EN parameters were developed by nonlinear regression. Fe release increased with localized corrosion (PF) and the EN model predicts that Fe release can be effectively controlled to the same degree by pH elevation or inhibitors. Cu release increased with general corrosion (LPRCR) and was also influenced by localized corrosion (ECNCR). However general corrosion was more significant for copper release which was mostly in the dissolved form. Pb release was depended on both general corrosion (LPRCR & HMCR) and localized corrosion (PF). The EN models predict that both Cu and Pb release is highest for pH control and all inhibitors reduced Cu and Pb release, which is consistent with the data. Inhibitors ranked by increasing effectiveness for reducing both Cu and Pb release are pH elevation, Si, ZOP, OP and BOP. EN monitoring is faster and less labor intensive than water quality monitoring and represents a significant advance for controlling metal release in drinking water distribution systems. The EN models were found to be comparable to water quality models developed from this study for metal release, and since EN is a real-time technique it offers a tremendous advantage over traditional water quality sampling techniques. Remote access of EN monitoring equipment is possible and the system requires little to no maintenance with the exception of a power supply or battery. The rapid turn around of corrosion rates from EN can be used to estimate metal release in drinking water proactively and mitigating measures can be implemented before the full adverse impacts are realized.