Six months after the government told people in Flint, Michigan, to use bottled water for drinking and cooking, new research shows that what's coming out of their taps remains unfit for human consumption.
A recent round of sampling shows lower levels of poisonous lead contaminating Flint's water, but the amount of lead is still above what the Environmental Protection Agency considers acceptable for a public water system.
"People have to continue using bottled water and filters until further notice," Marc Edwards, the Virginia Tech researcher who led the sampling effort, said Tuesday morning at a press conference at the university.
Lead is a handy pipe material that also happens to be a deadly neurotoxin. In Flint and other American cities where lead has been used for water pipes, the toxic metal can contaminate the water, exposing people who drink from the tap to a range of health problems -- including irreversible IQ loss in children.
For a public water system to be in compliance with the Safe Drinking Water Act, the EPA says it has to have lead levels below 15 parts per billion in 90 percent of samples taken. In March, Edwards' team found 22.8 parts per billion of lead in Flint's "90th percentile" sample. That's down from 28.5 parts per billion in August, but still above the EPA's limit. (Being below the limit doesn't necessarily mean the water is safe from lead.)
Last month Gov. Rick Snyder (R) touted the state's own testing results, which found that 91 percent of samples were beneath the federal limit, an improvement from the state's previous round of testing.
"We have always said that based on our testing, concerns still exist," Snyder spokesman Ari Adler said in a Tuesday email, adding that the governor is grateful for Edwards' work.
"It appears that many of their findings and recommendations track with what the state is seeing but we are continuing to test the water in Flint and are working to restore the system in terms of stability," Adler said.
The EPA also collected Flint water samples earlier this year, though it hasn't published "90th percentile" findings like Edwards and the state of Michigan have. Snyder's office said last week that federal, state and independent testing efforts all show improvement in Flint's water, though not enough for people to start drinking it.
Edwards and his graduate students first teamed up with Flint residents last summer to investigate Flint's water while the local, state and federal governments downplayed his concerns about its lead content. Since being proven right, Snyder has made Edwards one of his top water consultants, and the EPA funded his recent round of testing.
Flint's water soured after Snyder's government oversaw the city's switch to the Flint River as its source, then failed to treat the river water with anti-corrosion chemicals. Corrosion control treatment is something large public water systems are always supposed to do in order to prevent corrosive materials from leaching lead out of pipes and home plumbing materials.
Flint started adding extra corrosion controls to its water in December, two months after Snyder approved the city's return to the Detroit water system. On Tuesday, Edwards said his team's samples showed that the corrosion controls were working -- that they were successfully rebuilding the interior coating that services as a barrier between water and the pipe surface.
But Edwards noted that because Flint residents have been using much less tap water, it's taking longer than it normally would to flush out deposits of lead and rust that discolors the water. Edwards suggested Flint residents could help the system by running their taps, something Snyder's office also recommended last week.
"We have learned in the past few months it's going to take months or years to get these deposits out of the pipes," Edwards said. "If we want to get the system to recover, we have to get water moving through the pipes to speed up the rate at which these loose deposits are removed into the water."
Clarification: This story has been updated to reflect that the EPA has published some data related to its sampling in Flint, but has not released any conclusions.