.png){kind=icon}
.png){kind=icon}
Return
Bayesian inference for total gamma clearance monitors: Accounting for both and information
.png){kind=icon}
PRE
AI
E
M
B
W
A
A
DOI：10.1016/j.nima.2025.171266.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This work introduces a novel Bayesian approach for interpreting total gamma clearance monitoring data. The approach integrates spatial and spectral information from the measurements with a surrogate efficiency model, all within a Bayesian inference framework, to estimate the spatial activity distribution across 15 subvolumes inside a 200-liter waste drum. When the 60Co/137Cs ratio is not known a priori, it can be jointly estimated alongside other variables of interest. Virtual and real mock-up experiments demonstrate that the method accurately identifies high-activity subvolumes and reconstructs the relative spatial activity distribution. Moreover, with reliable background count estimates, the total drum activity can be determined with a relative error below 10%. Tests on two real low-level waste drums, validated against HPGe-based gamma spectrometry, confirm a maximum relative error of 10%. Regarding inference of the 60Co/137Cs ratio, the approach correctly detects drums containing only 137Cs but underestimates the 60Co fraction when 60Co is present. Our proposed approach has also some practical limitations, mainly its reliance on prior knowledge of the drum's filling level being close to 100% and, to a lesser extent, its underlying assumption of a constant density across the drum. Future work will aim to solve that issue by incorporating filling degree into the surrogate efficiency model.
Keywords:
Radiological characterization
Total gamma
Bayesian inference
Clearance monitor
Markov chain Monte Carlo (MCMC)
Uncertainty quantification