@book{94b754dc633e42ee970f12954aa4bea3,
title = "Level 2 PRA studies: Source term characteristics and hydrogen explosions",
abstract = "This document studies probabilistic modelling of severe nuclear power plant accidents. Factors that affect the height and temperature of radioactive releases are identified, and hydrogen explosions in boiling water reactor (BWR) nuclear power plant are studied. A containment event tree model is developed for a BWR plant based on a model developed in a previous study. Uncertainty analysis is implemented for release probabilities, and release heights and temperatures are included in the model. The height and temperature of radioactive releases have typically not been included in level 2 probabilistic risk analyses (PRA), even though they are needed as inputs for level 3 PRA. Release height is usually the height of the location where the reactor building leaks (which depends on the containment failure mode), or the height of the chimney if the release is controlled. In an uncontrolled accident case, both the containment failure location and the flow path of radionuclides in the reactor building need to be analysed to determine the release height. The temperature of release from containment is in most cases close to 100°C, but the temperature of radionuclides can potentially change during their migration in the reactor building. Higher temperatures can be caused by fires and explosions. There are computer codes that can be used to analyse radionuclide flows in reactor building and determine the release heights and temperatures. Hydrogen explosions can occur inside BWR containment if the containment is not inert. Therefore, the risk of hydrogen explosions comes mainly from reactor start-up, shut-down and refuelling. Hydrogen explosions outside containment have usually not been modelled in PRA, even though they can affect the releases significantly.",
keywords = "probabilistic risk analysis, severe accident, source term, hydrogen explosion, uncertainty analysis",
author = "Tero Tyrv{\"a}inen and Ilkka Karanta",
note = "Project code: 102420 ",
year = "2017",
language = "English",
series = "VTT Research Report",
publisher = "VTT Technical Research Centre of Finland",
address = "Finland",
}