[en] Modelling of severe accidents (SA) at nuclear power plants (NPP) includes assessing the radioactive material release into the environment (called the «source term»). The accuracy of the source term assessment using integral codes depends on the modeling error of individual phenomena such as the build-up of actinides and fission products (FP) in fuel prior to the accident, evolution of isotopic composition of fuel during the accident, release from fuel, transport, and deposition of actinides and FP in the reactor cooling system (RCS) and containment, their release into the environment.

sp0010^>The application of best estimate approach with uncertainty analyses, the deterministic support of probabilistic safety assessment (PSA) level 2, and calculations within emergency response require multiple calculations of SA scenarios, therefore the integral codes have to run pretty fast. For consideration of isotopic evolution both in the irradiated fuel and in chemical compounds released from the fuel a depletion model has to be regularly called during each calculation session. To fit the requirement of smaller runtime, the integral code has to be provided with a fast-running depletion model. The fast performance of such models is payed by the predictive accuracy; therefore, it is important to assess if the accuracy is acceptable for the mentioned issues of SA analyses. This paper presents the results of study of errors in the fast-running depletion model that is intended for use in integral code during multi-variant calculations of SA scenarios.

sp0015^>The validated parameters are concentrations of actinides and FP in light water reactor (LWR) spent fuel. Additionally, calculations of the decay heat power are provided. The basis for validation consists of the out-of-pile measurement results provided in Nuclear Energy Agency of Organization for Economic Co-operation and Development (OECD/NEA) database SFCOMPO-2.0, and the ORIGEN-2 code calculation results obtained by Japan Atomic Energy Agency (JAEA) for Units 1–3 of the Fukushima Daiichi NPP.

[en] Highlights: • Prototypical tests do not reveal detonation of steam explosions at in-vessel stage of SA. • Conversion ratios obtained in tests with steam explosion lie within 0,003–0,04. • Design features of VVER imply several factors which are limiting the likelihood of steam explosion. • In case of in-vessel steam explosions, the calculated peak pressure on VVER upper head is 6 MPa. - Abstract: This paper presents the results of calculations of the loadings that reactor main joint may experience after energetic fuel-coolant interaction in a hypothetical severe accident at VVER-1200 reactor. Estimates of the loadings are performed using a semi-empirical procedure with conservative assumptions. The initial data for estimates are deduced from the modeling of representative severe accident with an integral code SOCRAT/V1. The loadings can be used for further thermomechanical calculations of the integrity of elements that are fixing the upper head of VVER-1200 RPV.