Atmospheric transport of radioactive debris to Norway in case of a hypothetical accident related to the recovery of the Russian submarine K-27

Abstract

The Russian nuclear submarine K-27 suffered a loss of coolant accident in 1968 and with nuclear fuel in both reactors it was scuttled in 1981 in the outer part of Stepovogo Bay located on the eastern coast of Novaya Zemlya. The inventory of spent nuclear fuel on board the submarine is of concern because it represents a potential source of radioactive contamination of the Kara Sea and a criticality accident with potential for long-range atmospheric transport of radioactive particles cannot be ruled out. To address these concerns and to provide a better basis for evaluating possible radiological impacts of potential releases in case a salvage operation is initiated, we assessed the atmospheric transport of radionuclides and deposition in Norway from a hypothetical criticality accident on board the K-27. To achieve this, a long term (33 years) meteorological database has been prepared and used for selection of the worst case meteorological scenarios for each of three selected locations of the potential accident. Next, the dispersion model SNAP was run with the source term for the worst-case accident scenario and selected meteorological scenarios. The results showed predictions to be very sensitive to the estimation of the source term for the worst-case accident and especially to the sizes and densities of released radioactive particles. The results indicated that a large area of Norway could be affected, but that the deposition in Northern Norway would be considerably higher than in other areas of the country. The simulations showed that deposition from the worst-case scenario of a hypothetical K-27 accident would be at least two orders of magnitude lower than the deposition observed in Norway following the Chernobyl accident.

Joint Nordic nuclear research to strengthen nuclear emergency preparedness after the Fukushima accident

“…For each of two source terms defined and each of eight numerical weather prediction models and four atmospheric dispersion models results were generated for each of a number of simulated release dates. It should be mentioned that both longscale Lagrangian and Gaussian (Bartnicki et al, 2014;Havskov Sørensen et al, 2016) and mesoscale puff (Brandt et al 1996) dispersion models were employed. These models were made and continuously developed in the Nordic countries, but are also incorporated in the European decision support systems.…”

Archives of Environmental Protection

“…Almost two hundred thousand trajectories were calculated during the period of interest. The methodology of calculating the trajectory was identical to that in (Bartnicki et al 2010) but extended to a three-dimensional problem, such as in (Pongkiatkul & Kim Oanh 2007, Draxler 2007.…”

Improved source term description in Eulerian models in ARGOS

“…In the first step, we compared selected analytical solutions with the numerical results from the Lagrangian, Eulerian and the hybrid methods. In the second step, the hybrid method was applied and tested in the real meteorological conditions, as a combination of the Lagrangian SNAP model (Bartnicki et al, 2014) and Eulerian EEMEP model (Simpson et al, 2012). The results of both, analytical tests and tests in the real meteorological conditions indicate some small improvement when the hybrid approach was applied.…”