Sources and significance of variation in the dose estimates of 36Cl biosphere transfer models: a model...

A range of performance assessments have indicated that the long-lived activation product (36)Cl will be among the more significant contributors to dose following release to the biosphere from deep or near-surface repositories for radioactive wastes. Described here are results of a BIOPROTA model intercomparison study, investigating dose assessment uncertainties and variability on the basis of six (36)Cl models from three countries. The models share a compartmental approach with transfers between compartments handled on the basis of empirical transfer factors (IMARC, ERB2A, Aquabios), on the basis of defined specific activities (AquaCl36, SA_36Cl), or on a combination of these methods (MTA_Cl36). The dose estimates that these models produce for a consensus well-water irrigation scenario, as well as the effect of altering certain critical assumptions, are reported, and the causes of variation examined. For the scenario considered, the calculated doses are within a factor of 15 of each other. Major differences were attributable to the data used for stable Cl concentrations and (36)Cl transfer parameter values, both typically site-specific parameters. Additional critical assumptions were studied such as the impact of stable chloride in the diet on dose coefficients, the effect of irrigating pasture with contaminated water on (36)Cl concentrations in animal products, and the explicit consideration of foliar uptake.