Development of a numerical model for sector-average plume gamma dose and its validation with dose rate measurements at Kalpakkam NPP site, India.

A Gaussian Plume based simple numerical model, named DIFFUSE is developed to simulate the long-term sector-average plume gamma dose due to radioactive plume released during normal operation of nuclear facilities. DIFFUSE calculates site specific joint frequency distributions of wind speed, wind direction and atmospheric stability using micrometeorological observations. It performs the finite sector-average dose integration for any stack height and gamma energy using Simpson's 1/3rd method with sufficient computational efficiency within the site boundary up to 2 km. Plume dose contribution to the main plume sector from nearest and next nearest side plume sectors is also calculated. DIFFUSE is validated with a 3-month long, starting from February 2021 to April 2021, dose rate observation data during operational releases from 100 m stack of Madras Atomic Power Station, Kalpakkam, India. Meteorological data from onsite 50 m tower and continuous dose rate observation from two sets of Autonomous Gamma Dose Logger (AGDL) detectors, namely n-AGDLs and r-AGDLs, placed in two different configurations along the geometric arcs of wind sectors around the stack are used. Simulated doses are compared with look-up table based dose estimates by Hukkoo et al. (1988). Linear spatial averaging of cumulative AGDL doses on a sector arc is used as measured sector-average dose for model validation. Simulations performed for both n-AGDL and r-AGDL configurations show DIFFUSE estimated doses are ∼37% lower and Hukkoo estimated doses are at least ∼50% lower than the measured doses. Statistical analysis of DIFFUSE simulated doses shows a statistical correlation of R2∼98.3%, slope of the fit ∼1.36 for n-AGDL setup and R2∼75.3%, slope of the fit ∼1.57 for r-AGDL setup. Overall, DIFFUSE produces conservative doses compared to look-up table based doses as required by regulatory bodies.