Shortly before this poverty-stricken city began drawing its drinking water from the Flint River in April 2014 in a cost-cutting move, officials huddled at the municipal water treatment plant, running through a checklist of final preparations.
Mike Glasgow, the plant's laboratory supervisor at the time, says he asked district engineer Mike Prysby of the Michigan Department of Environmental Quality how often staffers would need to check the water for proper levels of phosphate, a chemical they intended to add to prevent lead corrosion from the pipes. Prysby's response, according to Glasgow: "You don't need to monitor phosphate because you're not required to add it."
Recalling the meeting Tuesday in an interview with The Associated Press, Glasgow said he was taken aback by the state regulator's instruction; treating drinking water with anti-corrosive additives was routine practice. Glasgow said his gaze shifted to a consulting firm engineer in attendance, who also looked surprised.
"Then," Glasgow said, "we went on to the next question."
In hindsight, he said, it was a fateful moment. For nearly 18 months, Flint residents would drink water that had coursed through aging pipes and fixtures, scraping away lead from lines that ran from water mains to some homes and schools.
By the time Gov. Rick Snyder announced in October 2015 that Flint would return to the Detroit system, from which it had bought treated Lake Huron water for decades, scientists and doctors had reported dangerously high levels of lead in numerous water samples and a spike in the proportion of children with elevated lead in their blood. Even low amounts of lead are a health threat, especially for young children, as it is linked to lower IQs and behavioral problems.
Flint residents still are advised not to drink unfiltered tap water.
In a report last week, a task force appointed by Snyder to investigate the water crisis described the state as "fundamentally accountable," partly because of the DEQ's instruction to omit corrosion controls. It also assigned lesser blame to the state Department of Health and Human Services, local and federal agencies and emergency managers Snyder had appointed to oversee city operations.
The report did not fault Prysby alone among DEQ officials. The department's former director, Dan Wyant, and its chief spokesman resigned in December. Snyder later fired Liane Shekter Smith, former chief of the DEQ's Office of Drinking Water and Municipal Assistance. Another official has been suspended.
But during a hearing Tuesday in Flint before a legislative committee investigating the catastrophe, Glasgow said it was Prysby who told him that federal regulations on lead and copper pollution required testing the water for two consecutive six-month periods before deciding whether to apply corrosion controls.
The DEQ later would admit that was a misreading of the rules, which instead require systems serving more than 50,000 people to install and maintain corrosion control treatment.
Prysby has declined previous AP requests for an interview. He could not be reached for comment after the hearing.
"I did have some concerns and misgivings at first," Glasgow told the committee. "But unfortunately, now that I look back, I relied on engineers and the state regulators to kind of direct the decision. I looked at them as having more knowledge than myself."
Lee-Anne Walters, who helped draw official attention to the problem after high lead levels were discovered at her house, told the AP that learning of Prysby's instruction to the city made her "nauseous."
"That one meeting was the difference between this city being poisoned and not being poisoned," she said.
Walters, who says her children have been sickened by the lead-tainted water, was among numerous city residents who testified before the committee, some weeping and pounding the table in anger as they demanded Snyder's resignation and more funding to care for people who have suffered.
"I'm begging you — help us," a tearful Barbie Biggs said.
Walters described a frustrating struggle to convince government officials the problem was real. She eventually contacted Marc Edwards, a Virginia Tech engineering professor, whose team tested Flint's water and found elevated lead content.
Edwards later told the AP he was mystified by the failure to use corrosion controls, saying he had never heard of such a decision elsewhere in 25 years of studying lead contamination of drinking water.
"Corrosion control is the best investment a utility can make," he said.
State Sen. Joe Hune, a Republican from Hamburg, asked Glasgow during the hearing why he didn't disregard Prysby's instructions and add the phosphate. Glasgow responded that he had always respected the DEQ's judgment and added that it would have taken up to six months to acquire and install equipment for the treatments. (It wasn't until last December that the Flint plant finally got a corrosion control system.)
Glasgow, who last year became the city's utilities administrator, said pressure from superiors to move quickly also influenced him. Less than two weeks before the switchover to the Flint River, he had complained to another DEQ official that he needed more time for staff training and other preparations. "I will reiterate this to management above me, but they seem to have their own agenda," he wrote in an email.
But in the AP interview, he acknowledged faulting himself for not being more assertive.
"I kick myself every day," Glasgow said. "To know that if I could have screamed a little louder or questioned something a bit further, I could have maybe avoided all this, it's something I'll keep on my shoulders for the rest of my life."