Radionuclide concentration variations in the fuel and residues of oil shale-fired power plants: Estimations of the radiological characteristics over a 2-year period.

Several multi-day samplings were conducted over a 2-year period from an oil shale-fired power plant operating with pulverized fuel type of boilers that were equipped with either novel integrated desulphurization system and bag filters or with electrostatic precipitators. Oil shale, bottom ash and fly ash samples were collected and radionuclides from the U and Th series as well as K were determined. The work aimed at determining possible variations in the concentrations of naturally occurring radionuclides within the collected samples and detect the sources of these fluctuations. During the continuous multi-day samplings, various boiler parameters were recorded as well. With couple of exceptions, no statistically significant differences were detected (significance level 0.05) between the measured radionuclide mean values in various ash samples within the same sampling. When comparing the results between multiple years and samplings, no statistically significant variations were observed between U and Ra values. However, there were significant differences between the values in the fly ashes when comparing Pb, K, Ra and Th values between the various samplings. In all cases the radionuclide activity concentrations in the specific fly ash remained under 100 Bq kg, posing no radiological concerns when using this material as an additive in construction or building materials. Correlation analysis between the registered boiler parameters and measured radionuclide activity concentrations showed weak or no correlation. The obtained results suggest that the main sources of variations are due to the characteristics of the used fuel. The changes in the radionuclide activity concentrations between multiple years were in general rather modest. The radionuclide activity concentrations varied dominantly between 4% and 15% from the measured mean within the same sampling. The relative standard deviation was however within the same range as the relative measurement uncertainty, suggesting that the main component of fluctuations is derived from the measurement method and approach. The obtained results indicate that representativeness of the data over a longer time period is valid only when a fuel with a similar composition is used and when the combustion boilers operate with a uniform setup (same boiler type and purification system). The results and the accompanying statistical analysis clearly demonstrated that in order to obtain data with higher reliability, a repeated multi-day sampling should be organized and combined with the registered boiler technical and operational parameters.