Loading...

Determination of Concentration Profile in a Transport Column by Gamma Spectroscopy combined with Neural Network Technique

Dara, Mojtaba | 2023

76 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 56320 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Samadfam, Mohammad
  7. Abstract:
  8. Safety assessment of the final disposal of radioactive waste is crucial to ensure the waste isolation from the biosphere for long periods of time. There are various scenarios for the radionuclide transport from the disposal site to the environment, among which the ground water scenario is considered as the main scenario of the radionuclides migration. In this scenario, the transport parameters of the radionuclides are very important in the safety assessment of the final disposal of radioactive wastes. These parameters are determined using either dynamic or static methods, which usually the dynamic method leads to more realistic results. In the dynamic method, one of the two approaches of extracting the spatial concentration profile inside the column or extracting the temporal concentration profile at a certain point of the column (breakthrough curve) is used. In this research, an attempt is made to provide a simple, cheap and fast method for extracting the spatial concentration profile based on gamma spectroscopy and optimization by genetic algorithm (one of the conventional neural network methods), with an accuracy comparable to common techniques. In this method, after injecting the radionuclide solution into the water stream entering the column (with a uniform flow rate), the detector is placed in several different positions along the axis of the column and counts the gamma rays emitted from the column. In the following, using the recorded counts and the genetic algorithm technique, the spatial concentration profile and then the transport parameters are estimated. The obtained concentration profile corresponds well with the results of gamma scanning method with an RMSE error of less than 0.046. In addition to the very good estimation of the concentration profile, the proposed method is much faster than the gamma scanning method (10 minutes versus 8 hours). Also, the uniqueness of the obtained concentration profile was confirmed with the help of MCNP simulation
  9. Keywords:
  10. Column Experiment ; Genetic Algorithm ; Neural Network ; Concentration Profile ; Segmented Gamma Scanner ; Radionuclides Transport Parameters ; Radioactive Waste

 Digital Object List

 Bookmark

No TOC