Dynamic Dose Modeling / Soil Segregation: A Method for Reducing Uncertainty and Increasing Efficiency during Radiological Deconunissioning

摘要

The regulatory release of sites and facilities (property) for restricted or unrestricted use has evolved beyond prescribed levels to model-derived dose and risk based limits. Dose models for deriving corresponding soil radionuclide concentration guidelines are necessarily simplified representations of complex processes. It is not practical to obtain data to fully or accurately characterize transport and exposure pathway processes. Similarly, it is not possible to predict future conditions with certainty absent durable land use restrictions. RESRAD incorporates the recommended framework to perform human exposure pathway analyses that support the development of remediation guidelines. The methodology for collecting input data for RESRAD and the ranges and typical values of input parameters are discussed in detail in the RESRAD Data Collection Handbook. The Handbook's RESRAD input data were chosen by its authors to be realistic yet reasonably conservative, thus generating remediation guideline values that should overestimate actual dose. This approach introduced a convenient derived guideline calculation standard (framework and uncertainties) easily adopted by decommissioning project managers. However, this convenience bears a risk that is the project manager's nemesis—excessive remediation. Uncontrolled, this pervasive risk escalates project costs far from its baseline and diminishes the end-state effectiveness. Because the calculation standard's uncertainties are carried forward in the derived remediation guideline values, Civil and Environmental Consultants, Inc. (CEC) health physicists developed a unique approach to dose modeling and remedial action design to effectively manage end-point uncertainty. This approach uses a dynamic feedback dose model and soil segregation technology to characterize impacted material with precision and accuracy not possible with static control approaches. Utilizing the remedial action goal "over excavation " and subsequent auto-segregation of excavated material for refill, the end-state (as-left conditions of the refilled excavation) RESRAD input parameters are re-entered to assess the final dose. The segregation process produces separate below and above criteria material stockpiles whose volumes are optimized for maximum refill and minimum waste. The below criteria material is returned to the excavation without further analysis, with the above criteria material packaged for offsite disposal. Using the activity concentration data recorded by the segregation system and the as-left configuration of the refilled excavation, an end state model of the site can be prepared with substantially reduced uncertainty. In conjunction with field controls on remediation work, this approach not only assures that the physical end state and its model are consistent, but also the variability in future dose is reduced by a significant margin. The major benefits of this approach are: 1) Dual 100% characterization and final status survey of impacted area, 2) Increased stakeholder confidence in dose projections brought about by uniquely thorough MARSSIM surveys, 3) Lowered project costs stemming from efficient analysis and abstraction of impacted material and reduced offsite waste disposal volume, and 4) Lowered project costs due to increased remediation/construction efficiency and decreased survey and radio-analytical expenses.