AIRNET Action Levels

• Current
• Historic

To identify and attend to potential environmental problems before they pose a threat to the environment or the public, data from AIRNET stations are compared to calculated action levels. There are two types of action levels: investigation levels and alert levels. Air concentrations exceeding the investigation level are significantly above the normal or expected value for that station. The alert level is exceeded when the investigation level is exceeded and the concentration is above the equivalent of a one-mrem dose (the Environmental Protection Agency limits the air pathway dose to a member of the public to 10 mrem per year). If the alert level is exceeded, a significant release may have occurred, and the appropriate personnel are notified. An investigation is launched to verify and determine the cause of the elevated air concentration if either type of action level is exceeded.

During 1999, we developed a new method for determining action levels based upon air concentrations (pdf) or estimated dose (pdf). This was needed primarily because fluctuations in alpha and beta activity levels and uranium concentrations, all arising from natural sources, impacted the suitability of existing action levels. We also determined that changes in atmospheric moisture and temperature have a significant effect on atmospheric tritium content around Area G, and that the use of action levels based on annual averages did not provide meaningful screening levels. Our action levels were, in some cases, too restrictive; data would exceed the action level and necessitate careful investigation when none was warranted. In other cases, the action levels were too large (to incorporate wide natural fluctuations) and might not indicate a problem where one existed. We also reviewed and improved our methods for identifying and removing outliers in the data (Pu and Am (pdf) or tritium (pdf)).

We began recalculating the action levels using the improved treatment of outliers and incorporating the effects of natural variations during 1999. For plutonium, americium, and tritium (excluding Area G stations), investigation levels were calculated as the five-year mean plus three standard deviations after removing outliers. For tritium at Area G stations, a five-year mean concentration at each station is calculated for each two-week sampling period. Then the investigation level for each individual two-week sampling period is calculated as a running mean of the two sampling periods before, the actual sampling period, and the two sampling periods after plus three standard deviations. This allows the investigation level to reflect the seasonal changes in tritium emissions from Area G.

New methods have also been developed for gross alpha, gross beta, and uranium investigation levels. We anticipate implementing these methods by the fall of 2000. For alpha and beta, the biweekly investigation level will be the average of all concentrations for that period plus three times the standard deviation of the mean for the preceding two to five years. For uranium, the investigation level will be set to determine if there is significant variation from the natural uranium-234 to uranium-238 air concentrations ratio.