Addressing gaps in the US EPA Lead and Copper Rule: Developing guidance and improving citizen science tools to mitigate corrosion in public water systems and premise plumbing

Lead and copper in drinking water are known to pose aesthetic and health concerns for humans and pets. The United States Environmental Protection Agency (US EPA) Lead and Copper Rule (LCR) set 90th percentile action levels for lead (15 ppb) and copper (1.3 mg/L), above which utilities must implement systemwide corrosion control. However, gaps in the US EPA LCR leave at least 10% of residents using municipal water and all private well users vulnerable to elevated lead and copper in their drinking water. To help address these gaps in the LCR, this dissertation 1) Evaluates accuracy of at-home lead in water test kits to help residents identify lead problems, 2) Refines orthophosphate corrosion control guidance to help reduce cuprosolvency, 3) Identifies challenges to mitigating cuprosolvency by raising pH, and 4) Develops guidance that can help residents assess and address cuprosolvency problems.

Lead in drinking water can pose a variety of health concerns, particularly for young children. The revised LCR will still leave many residents unprotected from elevated lead in their drinking water and potentially wondering what to do about it. Many consumers concerned about lead may choose to purchase at-home lead in water test kits, but there is no certification authority to ensure their accuracy. Most off-the-shelf tests purchased in this work (12 of 16) were not able to detect dissolved or particulate lead at levels of concern in drinking water (i.e. near the lead action level of 15 ppb) due to high detection limits (5,000-20,000 ppb). Binary type tests, which indicate the presence or absence of lead based on a trigger threshold of 15 ppb, were often effective at detecting dissolved lead, but they failed to detect the presence of leaded particles that often cause high lead exposures in drinking water problems. Some of these problems detecting particles could be reduced using simple at-home acid dissolution with weak household acids such a vinegar or lemon juice. Our analysis points out the strengths and weaknesses of various types of at-home lead in water tests, which could be particularly important considering potential distrust in official results in the aftermath of the Flint Water Crisis.

Elevated cuprosolvency, or copper release into drinking water, can be an aesthetic concern due to fixture staining, blue water, and green hair and can pose health concerns for residents and pets. In addition to the general gaps in the LCR described above, compliance sampling in the LCR focuses on older homes at highest risk of elevated lead, rather than the newer homes at highest risk of elevated copper. Problems with elevated copper can sometimes go undetected as a result. Guidance was developed to help proactive utilities address cuprosolvency issues through the addition of orthophosphate corrosion inhibitors or pH adjustment as a function of a water's alkalinity. Linear regressions developed from pipe cuprosolvency tests (R2>0.98) determined a "minimum" orthophosphate dose or a "minimum" pH for a given alkalinity that was expected to almost always reduce copper below the 1.3 mg/L EPA action level in a reasonable length of time. The subjective nature of the terms "almost always" and "reasonable length of time" were quantitatively discussed based on laboratory and field data.

Orthophosphate addition was generally very effective at cuprosolvency control. Orthophosphate treatment in copper tube cuprosolvency tests produced cuprosolvency below the action level within the first week of treatment. As expected, orthophosphate treated waters sometimes resulted in higher long-term cuprosolvency than the same waters without orthophosphate corrosion control treatment. This is consistent with the formation of phosphate scales which have an intermediate solubility between the cupric hydroxide in new pipes and the malachite or tenorite scales expected in pipe aging without orthophosphate. A linear regression (R2>0.98) was used to determine the orthophosphate dose needed for a given alkalinity to yield copper below the 1.3 mg/L action level in the pipe segments with the highest, 2nd highest, 3rd highest copper concentrations (100th, 95th, or 90th percentile, n=20 replicates, five each from four manufacturers) after 4 or 22 weeks of pipe aging. This regression was generally in good agreement with a bin approach put forth in the 2015 Consensus Statement from the National Drinking Water Advisory Council, but in some cases the regression predicted that higher orthophosphate doses would be needed.

In contrast, due to the greater complexity of the reactions involved, a similar simplistic approach for pH adjustment is not widely applicable. A linear regression predicted that higher "minimum" pH values would be needed to control cuprosolvency compared to those suggested by the 2015 National Drinking Water Advisory Consensus Statement. Results indicate that factors such as the potential for calcite precipitation, pipe age, and significant variability in cuprosolvency from pipes of different manufacturers may warrant further research. Field LCR monitoring data indicated that 90th percentile copper concentrations continued to decline over a period of years or decades when orthophosphate is not used, and our laboratory results demonstrate a few cases where copper levels even increased with time. Consideration of confounding effects from other water quality parameters such as natural organic matter, silica, and sulfate would be necessary before the "minimum" pH criteria could be broadly applied.

Guidance was then developed to help address cuprosolvency issues on a single building or single home basis for residents with private wells or those with high copper in municipal systems meeting the LCR. A hierarchy of costs and considerations for various interventions are discussed including replumbing with alternative materials, using bottled water or point use pitcher, tap, or reverse osmosis filters to reduce copper consumption, and using whole house interventions like more conventional orthophosphate addition and pH adjustment, or unproven strategies like granular activated carbon filtration, reverse osmosis treatment, and ion exchange treatment. Laboratory and citizen science testing demonstrated that some inexpensive at-home tests for pH and copper, were accurate enough to serve as inputs for this guidance and could empower consumers to diagnose their problems and consider possible solutions. Citizen science field testing and companion laboratory studies of potential interventions indicate that short-term (<36 weeks) use of pH adjustment, granular activated carbon, anion exchange and reverse osmosis treated water were not effective at forming a protective scale for the resident's water tested. In this case-study, cuprosolvency problems were ultimately related to water chemistry and linked to variability in influent water pH.

Overall, this work highlighted weaknesses in the current US EPA Lead and Copper Rule. It attempted to close some of these gaps by assessing the accuracy of at-home citizen science tests for lead and copper detection and developing guidance to support voluntary interventions by utilities or consumers. Ideally, local authorities (utilities, health departments, cooperative extension programs) could adapt this guidance to account for local water quality considerations and support consumers in resolving cuprosolvency issues. This guidance may also serve as a citizen science approach that some consumers could use to make decisions on their own. Future work could extend and improve on these initial efforts.