The disposal of Canada's nuclear fuel waste: the biosphere model, BIOTRAC, for postclosure assessment

The nuclear fuel waste management concept of Canada calls for disposal of the waste in a vault mined deep in plutonic rock of the Canadian Shield. The technical feasibility of this concept, and its impact on the environment and human health, will be documented in an Environmental Impact statement (EIS) by AECL. The present report is one of nine EIS primary references. The report describes the BIOTRAC model, which is used to trace nuclide movement from the geosphere through the biosphere and to calculate time-dependent environmental concentrations and radiological doses to humans and other biota for the postclosure phase. These concentrations and doses are crucial for evaluating the safety and environmental acceptability of the concept in terms of chemical and radiological toxicity. BIOTRAC was developed specifically to assess the impacts of a used-fuel disposal vault. It is a comprehensive, generic model with distributed or probabilistic parameter values to account for spatial and temporal variability and uncertainty. It is composed of four separate but closely linked submodels representing surface waters, soils, the atmosphere and the food chain. It also includes a unique model for predicting radiological doses to non-human biota, represented by generic target organisms. The mathematical formulation of each submodel is derived in detail and interpreted physically, and all the assumptions are fully evaluated and discussed. It is shown how the parameter values and distributions adopted for each submodel are derived from the available data. The interfaces between the submodels, and between BIOTRAC and the geosphere model, are presented in detail. Fluctuations in the physical state of the biosphere are accounted for through the parameter distributions. Major environmental changes, such as those caused by continental glaciation, are addressed quantitatively and through reasoned arguments, which indicate that radiological doses to humans will not increase suddenly or dramatically beyond 10 000 a. The reliability of BIOTRAC is discussed in terms of experimental validation, model and data evaluation, peer review, model intercomparisons, conservative assumptions, quality assurance procedures and natural analogs. A sensitivity analysis of the model is used to identify the nuclides, pathways and parameters that are important in determining radiological doses, and to show that the model is performing as one would expect on an intuitive basis. The changes required to apply the model to a specific site during siting of a disposal facility are identified and shown to be achievable. The information presented in this report indicates that BIOTRAC provides a suitable and satisfactory description of nuclide behaviour in the biosphere, and will not underestimate consequences when used to assess the concept for the disposal of Canada's nuclear fuel waste. (author)