No tritium found in Santa Fe wells

-A A +A
By John Severance

Alex Puglisi is the environmental compliance officer with the City of Santa Fe. And he knows there are people that don’t always buy the Los Alamos National Laboratory¹s story.

Puglisi, though, is confident the lab is right on this one.

Back in March of last year, samples at three of the wells at Santa Fe Buckman Water Supply Wells came back positive for tritium, which is a radioactive isotope of hydrogen.

Ten months later, however, it was revealed that those findings were erroneous.

So how does something like this happen?

“We got explanations from the lab as to what caused the errors and we agreed there was an error in those values. If you look at the information provided, there is a whole list of things that happened. There was some sort of rounding up of results and there was a transcription error of data as well.

“We have had two conference calls with them and they have sent us information three different times. This was a reporting error. But we looked at the results of the tests and we were satisfied.” Puglisi said there is testing for tritium every quarter and the March 2011 sampling was the first time anything came back positive.

LANL environmental manager Danny Katzman said the tritium finding was out way out of line from other samplings. So they retested.

“When we see detection, we explore repeatability” Katzman said. “We recognize the analytic method that can create spurious values.”

In 2010, the lab’s contract with the University of Miami Tritium Laboratory expired and a new contract was awarded to the American Radiation Services. In a letter from LANL’s Michael Graham and NNSA’s George Rael to the City of Santa Fe water division, they said that results from the ARSL may not be directly comparable with those from the UMTL.

The reasons? There are differences in each laboratory’s minimum detectable activity and counting uncertainty.

Puglisi sent the lab a series of questions about the sampling of March 2011, according to a another letter dated Dec. 7 to the City of Santa Fe water division to from the LANL Environmental Programs.

The first question: What were these calculation errors?

The lab found that rounding functions programmed into its spreadsheet were incorrect and it also found input errors that were direct result of manual data entry.

Puglisi then asked how were they determined to have
occurred at this late point in time and why were they not discovered when the March 14 results were first reported.

The lab and NNSA responded that the lab identified two tritium defects from the March 14 sampling event as elevated compared with previous results immediately after pulling data from the lab’s database. The lab requested that an analytic review of the data packages to ensure the accuracy of the results but the review was not completed before the 120-day deadline for releasing data to the public or posting it to the RACER database. The letter went on to describe the counting errors and the transition from the different tritium testing labs.

The letter stated: “From these Buckman well samples, revised calculation activities were significantly different from initially reported. When recalculated using the correct spreadsheet, the results previously reported as detections are now considered non-detects.”

After the City of Santa Fe’s review period, the results were uploaded to the RACER database. The superseded results were retained in the RACER database but with new qualifiers that indicate the results were rejected.

“We have not seen any indication of tritium in the wells and we are confident these results are correct,” Puglisi said.

Katzman said a number of quality controls were implemented after the retesting to make certain those results were indeed correct.

In all, Katzman said, 18 totals were changed.

“This is a big deal,” Katzman said. “We take the business of changed data very seriously.”

 The RACER database, which was developed by the New Mexico Community Foundation, provides environmental data from LANL on a public website.