THE VEGETATION-TO-AIR CONCENTRATION RATIO IN A SPECIFIC ACTIVITY ATMOSPHERIC TRITIUM MODEL

Specific activity models are frequently used to estimate the concentra tion of tritium oxide in vegetation. In such models, a single value re presents the ratio (R) of the specific activity of tritium oxide in ve getation to the specific activity of atmospheric tritium oxide. Federa l agencies such as the Nuclear Regulatory Commission and the Environme ntal Protection Agency have not established a consensus default for R. Literature on this topic suggests that a site-specific distribution o f R should be developed when feasible. In this study, a distribution o f R is established for the Savannah River Site. Environmental tritium concentrations in air and vegetation measured on and around the Savann ah River Site over a 9-y period form the basis for the analysis. For d ose assessments of chronic atmospheric tritium releases at the Savanna h River Site, R is best parameterized by a normal distribution with a mean of 0.54 and one standard deviation of 0,10. The Nuclear Regulator y Commission default for R is approximately equal to the Savannah Rive r Site site-specific estimate. Based on the results, the default value for R recognized by the Environmental Protection Agency overestimates tritium concentrations in vegetation and, therefore, doses from foods tuff consumption pathways at humid sites. For the Savannah River Site, the magnitude of the error is on the order of a factor of 2. This con sideration may be important if an estimated dose approaches an as-low- as-reasonably-achievable or regulatory threshold. Conversely, without the benefit of site-specific data, ingestion doses may be underestimat ed in regions with dry climates.