3-D gamma dose rate reconstruction for a radioactive waste processing facility using sparse and arbitrarily-positioned measurements. (February 2022)

• Record Type: Journal Article
• Authors: Zhu, Shangzhen; Cao, Jianzhu; Fang, Sheng; Dong, Xinwen; Li, Wenqian; Liu, Xuegang; He, Qiange; Wang, Xinghai
• Is Part Of: Progress in nuclear energy. Volume 144(2022)
• Publication Date: 2022-02
• Subjects: Dose rate field -- Three-dimensional reconstruction -- Arbitrarily-positioned measurements -- Simulation and real scenarios data validation; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 333.7924
• DOI: 10.1016/j.pnucene.2021.104073
• Languages: English
• ISSNs: 0149-1970

Abstract

Abstract: The detailed 3-D dose rate field is important for optimizing radiation protection and planning the work in the radioactive area. However, in the presence of sparse and irregularly-positioned measurements, the current reconstruction method may not be available for obtaining a 3-D field, because they are focused on 2-D dose rates and they work with only measurement on regular grids. In this work, a novel method is proposed that can reconstruct a 3-D gamma dose rate field using sparse and arbitrarily-positioned measurements. This method is based on the modified Cahn-Hilliard equation, which recovers the missing information by promoting the continuity of the reconstructed field. To better recover the local details, the 3-D curvature function is incorporated into the modified Cahn-Hilliard equation. A numerical scheme is also introduced for solving the proposed method. This method is validated against two Monte Carlo simulations and the 1-D real measurement of a radioactive waste processing facility. The two simulations corresponds to the cases with and without shielding. The proposed method is applied to random and equispaced data. For the two simulation cases, the proposed methods can reconstruct the 3-D dose rate field using less than 4.2% data and achieve low error levels of 5.81%-10.03% . For the 1-D real measurements, the reconstructed profile is closer to the measurements than the Monte Carlo simulation, which the minimal error is below 2.1%. Highlights: Reconstruct detailed 3-D gamma dose rate fields using only less than 4.2% data. Works well both with randomly-sampled data and equispaced data. The reconstruction error is below 6.0% for 3-D radiation field without shielding. The reconstruction error is below 10.0 % for 3-D radiation field with shielding. Outperforms the Monte-Carlo simulation for the real measurements.