-
-
Notifications
You must be signed in to change notification settings - Fork 64
Water Quality in the US
Flint, Michigan dominated news cycles in early 2016, when it was declared to be in a state of emergency due to the city's lead-contaminated drinking water. Early childhood exposure to lead is known to have irreversible negative side-effects. Water is a significant contributor to toxic lead exposure. People around the world were stunned at how this could happen in a nation as prosperous as the United States.
The water crisis in Flint was an especially egregious example of a problem that persists across many American communities: many Americans are drinking water that may be poisoning them, but it's almost impossible to know how bad lead poisoning is in your city. States have different requirements for pediatric screening of lead exposure, and only one state (Delaware) follows the Center for Disease Control's (CDC) recommendations for screening. The water in Flint was unmistakably poisonous due to its unnatural yellow hue and musky smell. But lead poisoning is usually invisible to us, and lead can exceed extremely dangerous thresholds without anybody's knowledge. There could be more Flints out there right now, and in most cases, we'd have no idea.
Lead is far from the only water contaminant worth worrying about; there are many more contaminants that are known to exist in some communities' drinking water and are known to have adverse side-effects. Not all of these contaminants are inorganic, either; although waterborne diseases such as cholera and typhoid are much rarer today, waterborne disease outbreaks caused by micro-organisms still occur.
In addition to known and regulated contaminants, there are various so-called "emerging contaminants" which are known to be present in some communities' drinking water, but are completely unregulated. These contaminants pose additional problems compared to known contaminants because they are less likely to be tested for in water samples, and their health risks are unknown.
For a more detailed overview of drinking water contamination, you can read this EPA report.
Safe drinking water is regulated federally by the Environmental Protection Agency (EPA), a federal agency devoted to the protection of the environment established in 1970. The EPA's regulatory authority over the safety of public drinking water is granted by Safe Drinking Water Act (SDWA) of 1974. (The cleanliness of public water more generally, including stormwater and navagable waters, is also regulated by the Clean Water Act of 1972.)
An amendment to the SDWA in 1996 established the Unregulated Contaminant Monitoring Rule (UCMR), which requires the EPA to create a list every 5 years of no more than 30 unregulated contaminants to be monitored by public water systems. The two most recent UCMRs are here:
We share news stories, including on regulatory developments, in our Slack channel.
You can see the latest news in EPA rulemaking and proposed rulemaking related to the SDWA at FederalRegister.gov.
SDWIS is our primary source of data for drinking water violations. It contains data on states, territories, and tribal nations (herein collectively referred to as states)
States collect and manage relevant data in either their own databases, or the EPA's own database, Safe Drinking Water Information System/State (SDWIS/State). These states transfer their violations and enforcement data to the EPA's centralized version of SDWIS known as SDWIS/Fed. (Note: throughout our project, "SDWIS" refers to SDWIS/Fed, unless stated otherwise. Within this document, however, we specify between SDWIS/Fed and SDWIS/State). SDWIS/Fed is the data that the EPA primarily utilizes to assess the state of public drinking water in the United States.
SDWIS contains the following tables:
- ENFORCEMENT_ACTION: Documents actions taken against a Public Water System (PWS), laboratory, or operator.
- GEOGRAPHIC_AREA: Information on political units established by geographic boundaries, such as state, town, or county served by a Water System.
- LCR_SAMPLE_RESULT: 90th percentile sample summary results data for lead or copper.
- LCR_SAMPLE: 90th percentile sample summaries data for lead or copper.
- SERVICE_AREA: A service area defines the sensitive populations that receive water from the water system.
- TREATMENT: Treatment objectives and process for treating water from a water system facility.
- WATER_SYSTEM: Inventory information on public water systems.
- WATER_SYSTEM_FACILITY: Inventory information on public water system facilities.
- VIOLATION: Documents a breach of a requirement. Violations are detected by assessment of sample results or reviews (including on site visits).
- VIOLATION_ENF_ACTION: Association between a violation and an enforcement action.
The EPA notes that its data is inaccurate. "EPA is aware of inaccuracies and underreporting of some data in the Safe Drinking Water Information System. We are working with the states to improve the quality of the data."
The Government Accountability Office (GAO) has a comprehensive report on issues in the SDWIS. Some issues they've noted:
- According to an EPA audit in 2007, from a sample of 19 states audited, 20% of health-based violations were reported inaccurately or were not reported at all. In 2009, 26% of health-based violations were reported inaccurately or were not reported from a sample of 14 states. (p. 13)
- 91% of these errors in 2009 were because the state (or EPA region acting as a primacy agency) failed to issue a violation notice to a non-compliant water system. The other 9% are instances where a violation notice was issued but this was not reflected in SDWIS. (p. 15)
- In 2009, 84% of monitoring violations were inaccurately reported or not reported in SDWIS/Fed. (p. 16)
- The EPA does not check the quality of the data it receives from state agencies. (p. 17)
- EPA's routine estimates of the completeness and accuracy of SDWIS/Fed are based on a comparison of it and SDWIS/State, not based on original source documents, so if both data sets are inaccurate but consistent with one another, a violation may be inaccurately reported or not reported. (pp. 17-18)