Development of radiation dose prediction program based on measurement data

Area of Science:

• Nuclear Engineering
• Radiation Protection
• Computational Science

Background:

• Accurate dose assessment is critical for nuclear power plant (NPP) worker safety.
• Predicting worker radiation doses during accidents is challenging due to missing source term data.

Purpose of the Study:

• To develop a dose prediction program utilizing readily available dose rate measurements.
• To enhance worker safety protocols in NPPs during emergency scenarios.

Main Methods:

• Developed a dose prediction model using Kriging methodology for spatial data analysis.
• Implemented the model and user interface in C# and Windows Forms.
• Validated predictions against MCNP code simulations under various measurement densities.

Main Results:

• The developed program predicts worker doses with mean absolute percent errors (MAPE) ranging from 14.70% to 18.95% depending on measurement density.
• The Kriging model effectively analyzes spatial correlations in measured data.
• The program generates 3D dose maps and comparative dose charts for clear visualization.

Conclusions:

• The developed dose prediction tool offers a viable method for estimating worker radiation doses in NPP accidents.
• The program's ability to use measurement data enhances its applicability when source term information is unavailable.
• Findings support the establishment of optimized work plans and improved worker safety during nuclear emergencies.