With draft report, EPA takes major step to help communities assess risks from lead in drinking water

Tom Neltner, J.D.is Chemicals Policy Director

Communities around the country are testing their water for lead. But when they get the results, parents, public health officials, housing agencies and school officials have little guidance about what the number means and what actions to take or priorities to set. For lead in dust and soil in homes, child-care and schools, they have health-based numbers that serve as benchmarks for assessing risk. There is no such benchmark for drinking water. As a result, many are using the “Lead Action Level” of 15 parts per billion (ppb) as a surrogate. Yet, this level is based on the effectiveness of corrosion control; it has no relation to the associated health risks of lead exposure.

Yesterday, the Environmental Protection Agency (EPA) helped fill the void by releasing a draft report that provides three different approaches to setting a scientifically-robust “health-based benchmark” for lead in drinking water. The agency is seeking public comment on the draft and will convene a panel of scientific experts to consider each of the approaches.

The report is a critical step in implementing the recommendations of the agency’s National Drinking Water Advisory Council (NDWAC) which called for this type of health-based benchmark as part of an overhaul of the Lead and Copper Rule. The agency went a step further and provides alternatives to consider. We applaud EPA for its action and its rigorous, scientific analysis.

Accounting for the various models and assumptions, EPA developed a range of potential health-based benchmarks that range from 3 to 56 ppb of lead in water that people actually drink. However, you cannot readily compare these values to the typical water testing results reported by utilities or schools. Those tests are based on the first draw of water that has been sitting in the faucet and plumbing overnight and do not necessarily reflect what people drink over the course of a day. Later samples would likely be lower but could be higher if the building has a lead service line, especially if the line has been disturbed.

What approaches did EPA develop?

EPA developed values under three different approaches for three different age groups of children: 1) formula-fed infants from birth to six months; 2) toddlers from 1- to 2-years old; and 3) all children from birth to 7-years old. For each, the agency estimated different values that took into account if the child did or did not live in a home with lead-based paint. Because of their building’s age, homes with lead-based paint are more likely to have lead in dust and soil and higher levels of lead in the drinking water. An estimated 36 million homes have lead-based paint.

Each of the approaches EPA evaluated looks at the issue in a different way depending on how the issue is framed. The best approach depends on the objective.

  • Approach 1. Estimate the concentration of lead in drinking water that would result in a 1 or 5 percent increase in the probability of a child having an elevated blood lead level (EBLL). This is most similar to what NDWAC recommended in 2015 and how EPA established the regulations setting standards for lead in dust and soil in 2001. The current EBLL is 5 μg/dL; however, the Centers for Disease Control and Prevention is expected to lower it to 3.5 μg/dL later this month.
  • Approach 2. Estimate the concentration of lead in drinking water that would result in a 0.5 μg/dL or 1 μg/dL increase in children’s mean BLL. This is similar to the approach used by EPA when setting the National Ambient Air Quality Standard (NAAQS) for lead, although EPA used a much smaller increment, 0.1 μg/dL, because it corresponded to an IQ drop of 1 point.
  • Approach 3. Assumes sources of lead such as food, dust and soil remain unchanged and estimates the concentration of lead in water needed to protect the most highly exposed children from having an EBLL.

What are the numbers using NDWAC’s recommendation (Approach #1)?

For simplicity, assuming that CDC will reduce the definition of EBLL from 5.0 to 3.5 μg/dL, here are the draft health-based benchmark numbers:

  • For formula-fed infants living in homes without lead-based paint:
    • If drinking water used to make up the formula has more than 9.3 ppb of lead, there is a 1% increase in probability of the child having an EBLL.
    • The probability is 5% if there is 12.2 ppb of lead in the water.
  • If these infants lived in homes with lead-based paint and average levels of lead in dust and soil:
    • If drinking water used to make up the formula has more than 3.8 ppb of lead, there is a 1% increase in probability of the child having an EBLL.
    • The probably is 5% if there is 6.6 ppb of lead in the water.


EPA has taken a significant step forward in helping parents and public health professionals better assess the risk from lead in drinking water. The agency needs to update the numbers based on the feedback from external scientific review and public comments and move quickly to propose an overhaul of the Lead and Copper Rule by the end of 2017. EPA needs to recognize that until it settles on a number and corresponding sampling strategy, communities and utilities will be using these draft numbers to better understand the risk.

Note that I updated the upper range for the benchmark from 18 ppb to 56 ppb on Jan. 15.

This entry was posted in Drinking water, General interest, Health science, Lead, Regulation and tagged , , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.


  1. Morri Markowitz
    Posted January 13, 2017 at 2:18 pm | Permalink

    It is always useful to have hard numbers to work with when trying to avoid health risks from lead exposure. How were these estimates arrived at?

    • Tom Neltner
      Posted January 15, 2017 at 11:18 am | Permalink

      Morri, thanks for the feedback. EPA used two models to develop the estimates. It Integrated Exposure Uptake and Biokinetic Model (IEUBK) was the foundation. One of the approaches also used the agency’s Stochastic Human Exposure and Dose Simulation (SHEDS) model. For more details, best to dig into the report.