The D.C. Water and Sewer Authority violated federal law by claiming that the water in more than 400 District homes had safe levels of lead and by not replacing that number of lead service lines, federal regulators announced yesterday.
The U.S. Environmental Protection Agency ordered the water authority to notify the homeowners of the error, to offer new water tests and to advise them to flush their water lines or filter water before drinking it. Under an administrative order signed last week and posted on the EPA Web site this week, the agency also ordered WASA to replace an additional 500 lead service lines citywide to comply with federal law.
During a widespread testing program in 2003, the authority told an estimated 417 to 566 households that their tap water contained lead levels in the safe range. But WASA had flushed the lines for five minutes before conducting the tests, which is not allowed and could give artificially low readings. In January 2004, the public learned from news reports that the lead in the District's water had exceeded the federal safety standard for the previous year and a half.
WASA's chief engineer, John Dunn, said yesterday that the authority regretted the mistake but is aggressively working to replace far more than 2,000 lead service lines this year.
"People who got wrong data should be upset," Dunn said. "We're sorry about that. A mistake was made, that's all I can say. Nobody's trying to hide it."
Also this week, WASA released new test results that show lead levels in the District's water are falling slightly since chemical treatment was started this summer. Still, the overall lead results for the city remain nearly four times the official federal safety standard of 15 parts per billion, according to federal officials.
The EPA learned of WASA's 2003 testing error through a Covington & Burling law firm report that WASA commissioned of its handling of the lead problem. WASA's widespread testing that year was part of its effort to use a legal provision to avoid the expense of physically replacing lead service lines where it found low lead levels.
The authority was under federal mandate at that time to replace about 1,600 lead service lines citywide. The authority found 1,200 service lines with low levels after testing 6,000 homes. It physically replaced about 400.
Dunn said there's no evidence that anyone at the authority intentionally sought to flush service lines to get lower lead readings and reduce the replacement work.
"I haven't seen anything saying, 'We can cut the edge on this if we flush longer,' " Dunn said.
According to a WASA document provided to the EPA, the five-minute flushing method produced many more service lines with low lead levels. Of 847 homes tested after a five-minute flush, 66 percent had low readings. Of 592 homes tested with the proper method, 27.5 percent had low readings.
"They used a different protocol that was not allowed by the regulations," said Karen Johnson, an enforcement official in EPA's mid-Atlantic region office, which oversees the District.
The Washington Aqueduct, which treats the water, began adding the chemical orthophosphate in an attempt to reduce lead levels. WASA and the EPA have had slightly different presentations of water test results since the treatment began.
In the last six months of 2004, the tests show that 90 percent of homes where water was tested have lead levels of 59 parts per billion or lower. An analysis of water samples taken in the first six months of the year showed 90 percent of homes had lead levels of 63 parts per billion or lower.
WASA General Manager Jerry N. Johnson, in a prepared news release this week, pronounced the results "good news" and said the "promising drop in lead levels" is attributed to the addition of orthophosphate to water treatment. He did not mention the official 59 parts per billion reading, but he highlighted tests from October to December that showed "the average lead concentration plummeted to 10.6 parts per billion, which is below the EPA action level of 15 ppb."
EPA officials reported the official 59 parts per billion reading and were more cautious in a later news release about orthophosphate's role. They said that "additional rounds of monitoring are needed for conclusive evidence of its effectiveness."