«Practical Elimination» Principle in the Akkuyu Nuclear Power Plant Design

The paper considers the implementation of the “practical elimination” principle in the design of the Akkuyu NPP with VVER1200 reactors being under construction in Turkey. The “practical elimination” principle is defined as follows: for accident sequences or phenomena that contribute or lead to unacceptable radiological consequences for the public or the environment, it shall be shown with a high level of confidence that their occurrence is highly unlikely. “Practical elimination” is proved predominantly by results of a Level 2 probabilistic safety assessment. The implementation of the “practical elimination” principle was considered at the level of accident sequences leading to a large radioactive release. It has been shown that each accident sequence leading to an unacceptable release has the probability of occurrence below 4.45Е–8 per reactor per year, while their total probability not exceeding the value of 6.17Е–7 per reactor per year. For the phenomena inside the containment area during severe accidents, including hydrogen detonation, a large thermal explosion, direct containment heating, overpressure in the containment volume, and the containment damage at later stages due to the basement melting through, their “practical elimination” has been demonstrated. The paper also considers specific hard-to-assess scenarios of beyond design basis accidents, for which the applicability of the “practical elimination” principle is assessed as well: a major positive reactivity insertion, a rupture of the reactor pressure vessel and other largescale components, damage of fuel elements in the spent fuel pool, severe accidents with the containment bypass or containment failure, and severe accidents accompanied the means for mitigating with their consequences being unavailable. Criteria have been developed and used for the “practical elimination” assessment. A sensitivity analysis was undertaken as part of the Level 2 probabilistic safety assessment which has shown that estimated values slightly depend on the analytical assumptions, as well as on the random change in the parameters that the affect the progression of severe accidents.