Hurricane Harvey made landfall on the Texas Gulf Coast on August 25, 2017, as a Category 4 hurricane and caused widespread flooding. We explored spatial and temporal distributions of well testing and contamination rates; relationships between contamination and system characteristics and recovery behaviors; and efficacy of mitigation strategies. We estimated that over 500 000 well users (∼130 000 to 260 000 wells) may have been affected, but only around 15 000 well users (∼3800 to 7500 wells) had inundated systems based on inundation maps. Local health departments and our team sampled 8822 wells in 44 counties in the 10 months that followed. Total coliform occurrence was 1.5 times and Escherichia coli was 2.8 times higher after Hurricane Harvey compared to baseline levels. Microbial contamination was more likely (1.7–2.5 times higher) when wells were inundated and/or residents felt their water was unsafe. Although more wells in urban counties were affected, E. coli rates were higher in wells in rural counties. Disinfection did not always eliminate contamination, highlighting concerns about the implementation and efficacy of chlorination procedures. Despite this extensive well testing conducted after Hurricane Harvey, we estimate that only 4.1% of potentially affected wells were tested, underscoring the magnitude of recovery assistance needed to assist well users after flooding events.

No abstract

No abstract

Despite public concern, the risk of lead exposure from schools remains poorly understood. The Lead and Copper Rule Revisions (LCRR) include, for the first time, a five-sample lead testing requirement for all elementary schools. However, the United States Environmental Protection Agency does not define school-wide lead risk or provide clear guidance on how results should be interpreted. Using the Massachusetts Lead in School Drinking Water Database, we explored the application of the LCRR sampling approach and provide insight into the magnitude and distribution of lead in water in Massachusetts public schools. We observed that 12% of fixtures had first draw lead >15 ppb and 3% after a 30 s flushing. Approximately 90% of fixtures with lead >15 ppb were clustered in 34% of schools. We determined a school-wide 90th percentile of 10 ppb closely approximated this clustering of problem fixtures and were able to identify schools with problem fixtures using the five-sample results with a confidence >90%. Fixtures releasing lead >1 ppb occurred in >90% of schools and represented 58% of first draws and 33% of 30-s flushed samples. Overall, our study provides an approach to classify a school's lead risk, which could help water utilities and schools prioritizing testing and remediation efforts.

No abstract