Undisclosed tests showed low levels of MCHM in water
CHARLESTON, W.Va. -- The Tomblin administration on Friday released test results that showed there were low levels of the chemical MCHM in the regional water supply for longer than had previously been disclosed.
Dozens of samples showed levels of MCHM that were below 10 parts per billion -- the reporting limit the state used for results it posted online -- but above a lower, 2 part-per-billion reporting limit one lab contracted by the state was using.
About 20 samples taken between Feb. 1 and Feb. 8 showed levels of MCHM greater than 2 parts per billion. Samples taken since Feb. 8 have all come back with non-detect results at the 2-part-per-billion limit.
Still, the new results show examples of state officials telling the public they had not detected the chemical, where the same samples showed some level of the substance when tested using the lower detection limit.
The state Department of Military Affairs and Public Safety disclosed the results, and posted them online, in response to a Freedom of Information Act request from the Charleston Gazette.
Some of the results reflect testing of new samples at the lower detection limit, while some results are for re-testing of older samples at that lower limit, state officials said.
Thomas L. Kirk, the department's general counsel, emphasized in an e-mail message that the results "must be viewed within the context" of the 1 part per million health screening level set by the U.S. Centers for Disease Control and a separate 10-part-per-billion level Gov. Earl Ray Tomblin directed state officials to use. One part per million is equal to 1,000 parts per billion, so the new 2-ppb test results still show levels of MCHM far below the CDC's controversial health-screening level of 1 part per million.
State officials said they had been previously been posting online only the results of tests using the 10-part-per-billion, or ppb, method because they viewed those as more reliable than tests that could detect lower concentrations.
In his e-mail to the Gazette, Kirk said the lab in question "informed state officials this week that they now have the capability to test confidently at 2 ppb" and that the department's Division of Homeland Security and Emergency Management would now be regularly posting those results online.
The Gazette had been asking for the 2 ppb detection limit test results for two weeks, since state Homeland Security Director Jimmy Gianato mentioned the lower detection limit during testimony before a legislative committee.
Tomblin administration officials released the data early Friday afternoon, after issuing a news release Thursday night that announced a plan to retest the water in all schools in the counties impacted by the chemical leak using the 2-ppb detection limit.
"After testing thousands of lab samples, chemists can now confidently test at 2 ppb," Adjutant General James Hoyer of the National Guard said in a Department of Education press release. "This additional safety factor is one more example of how the governor has gone above and beyond the federal recommendations to bring added confidence to parents, students, and staff to ensure the highest level of public safety."
The news release said that school testing was expected to occur over the weekend, with results due back by mid-week. Any school that tests above 2 parts per billion would have its plumbing system "re-flushed" and then be retested, the release said.
"Student safety continues to be our top priority," Tomblin said in the release. "As a parent, I understand that families need the additional peace of mind that comes with this testing."
In addition, the release said, schools will continue to provide hand sanitizer and bottled water for drinking and cooking.
"I applaud the governor for taking these additional steps," said state schools Superintendent Jim Phares. "We are taking every precaution because we know that parents trust us to keep their children safe."
Shortly after the administration's press release, West Virginia American Water issued its own Thursday night statement saying that the company had since Feb. 14 been working with laboratories to measure levels of MCHM down to 2 parts per billion to "address the remaining odor issues."
"We share customer's concerns and anger over the impact the Freedom Industries spill has had on our community," said water company President Jeff McIntyre. "We know that odor has added to their concerns, regardless of levels, and we will continue to flush our distribution systems to help address this issue."
The water company said that water tested at its treatment plant and at "nearly all points tested to date," is below 2 parts per billion, with only four samples showing concentrations greater than that.
"This additional data allows us to pinpoint our flushing activities and address odor issues in an expedited fashion," McIntyre said. "These additional efforts are solely to address odor issues and are not related to any concern regarding health thresholds set by the CDC. Water testing results indicate that water in the distribution system has been under the health-protective threshold of 1 ppm since January 18."
For several weeks, test results posted by the state -- mostly from sampling at the water treatment plant -- have consistently shown "ND," which the state explains on its results sheet reflects "any lab result below 10 parts per billion."
During a meeting of the Legislative Joint Commission on Water Resources on Feb. 7, Gianato told lawmakers, "Every sample that has been taken is listed on that website."
Later in the meeting, Gianato said that some testing had been done at a 2-ppb detection limit.
Lawrence Messina, a spokesman for Gianato's parent agency, explained in an e-mail message later on Feb. 7 that the 2-ppb testing that Gianato mentioned was an MDL -- an abbreviation that typically stands for "method detection limit" -- for one of the labs the state was using for its MCHM testing.
"That facility has tested new samples and re-tested prior samples at that MDL," Messina wrote. However, Messina wrote at the time, the state planned to continue to post online only those results using the 10 ppb detection limit. Messina explained that the 10 ppb limit was what's known as a "PQL," or practical quantification limit.
State officials planned to continue to post results at the PQL of 10 ppb, "viewing those results with a much higher degree of confidence," Messina wrote at the time.
Messina said Hoyer mentioned the existence of the 2 ppb testing previously "in previous media interviews, but largely in passing."
"It is unclear if those mentions have been included in prior coverage," Messina wrote.
In his Friday email releasing the 2 ppb test results, Kirk said that, "The lab capable of testing at that level provided these records."
Earlier in the week, in a Tuesday e-mail message, Kirk said the office of DMAPS Secretary Joseph Thornton "maintains no records reflecting" the 2 ppb testing.
"All other testing results received by the Office of the Secretary of the Department of Military Affairs and Public Safety have been relayed to the Division of Homeland Security and Emergency Management, which have been continuously posted on the water State of Emergency resource page," Kirk wrote.
Kirk said he checked with other agencies within DMAPS and "identified data that reflects" the 2 ppb testing.
Experts say that testing for extremely small concentration of chemicals can be difficult, especially if -- as in the case of MCHM -- the substance isn't very common, with few published studies and not necessarily a great need for such testing. The terminology used by labs can also vary, and it can be difficult to explain what various types of detection limits mean.
Generally, though, the lower the levels being detected, the less confident scientists are in whether the concentration shown on lab results is the true concentration.
Often, chemists talk about one number, called the reporting limit, as the lowest concentration at which a substance can be detected in a sample and its concentration can be reported with a reasonable degree of accuracy and precision. Another term is the detection limit, which is the lowest concentration at which a substance can be detected. The detection limit is always less than the reporting limit. And for detection limits, it is often more difficult to say exactly what the concentration is than to confirm that some amount of a substance is present.
Marc Glass, an environmental consultant with the Morgantown-based firm Downstream Strategies, said that both reporting limits and detection limits are important in understanding the results of chemical contamination tests.
"Some environmental regulatory programs require that you report both," Glass said. "You report both the reporting limit and the detection limit. You should report both."
In an unrelated news conference on Friday, one of the consultants hired by the Tomblin administration to do more analysis of the chemical spill's potential impacts said it only makes sense that labs have become more confident in the results of lower detection limits as they process more MCHM samples.
Andrew Whelton, a University of South Alabama engineer, said there's been a "mad rush" at the lab facilities to improve the ability to test water for chemicals from the Elk River spill.
Initially, Whelton noted, some labs were providing testing with a reporting limit of only 50 parts per billion. Soon, he said, they had improved that to 10 parts per billion.
Whelton said that the lab his home-testing project will use has a reporting limit of 1 part per billion of MCHM and a detection limit of 0.5 parts per billion.
"We're approaching the limit that we're actually able to quantify," Whelton said.
@tag:Reach Ken Ward Jr. at email@example.com or 304-348-1702.