Zaffora B, Magistris M, Saporta G, Chevalier J-P
Radiation Protection Group, European Organization for Nuclear Research, 1211, Geneva 23, Switzerland.
Radiation Protection Group, European Organization for Nuclear Research, 1211, Geneva 23, Switzerland.
Appl Radiat Isot. 2017 Sep;127:142-149. doi: 10.1016/j.apradiso.2017.06.001. Epub 2017 Jun 3.
This paper describes the process adopted at the European Organization for Nuclear Research (CERN) to quantify uncertainties affecting the characterization of very-low-level radioactive waste. Radioactive waste is a by-product of the operation of high-energy particle accelerators. Radioactive waste must be characterized to ensure its safe disposal in final repositories. Characterizing radioactive waste means establishing the list of radionuclides together with their activities. The estimated activity levels are compared to the limits given by the national authority of the waste disposal. The quantification of the uncertainty affecting the concentration of the radionuclides is therefore essential to estimate the acceptability of the waste in the final repository but also to control the sorting, volume reduction and packaging phases of the characterization process. The characterization method consists of estimating the activity of produced radionuclides either by experimental methods or statistical approaches. The uncertainties are estimated using classical statistical methods and uncertainty propagation. A mixed multivariate random vector is built to generate random input parameters for the activity calculations. The random vector is a robust tool to account for the unknown radiological history of legacy waste. This analytical technique is also particularly useful to generate random chemical compositions of materials when the trace element concentrations are not available or cannot be measured. The methodology was validated using a waste population of legacy copper activated at CERN. The methodology introduced here represents a first approach for the uncertainty quantification (UQ) of the characterization process of waste produced at particle accelerators.
本文描述了欧洲核子研究组织（CERN）所采用的用于量化影响极低水平放射性废物特性描述的不确定性的过程。放射性废物是高能粒子加速器运行的副产品。必须对放射性废物进行特性描述，以确保其在最终处置库中的安全处置。对放射性废物进行特性描述意味着确定放射性核素清单及其活度。将估计的活度水平与废物处置国家当局规定的限值进行比较。因此，量化影响放射性核素浓度的不确定性对于评估废物在最终处置库中的可接受性至关重要，同时对于控制特性描述过程中的分类、体积减小和包装阶段也很重要。特性描述方法包括通过实验方法或统计方法估计产生的放射性核素的活度。使用经典统计方法和不确定性传播来估计不确定性。构建一个混合多元随机向量，以生成用于活度计算的随机输入参数。该随机向量是一种强大的工具，可用于考虑遗留废物未知的放射学历史。当痕量元素浓度不可用或无法测量时，这种分析技术对于生成材料的随机化学成分也特别有用。使用在CERN活化的遗留铜废物群体对该方法进行了验证。本文介绍的方法代表了对粒子加速器产生的废物特性描述过程进行不确定性量化（UQ）的第一种方法。
