Band 73, Heft 1-2

The Accident Localisation System (ALS) of the Ignalina nuclear power plant is a “pressure suppression” type confinement, which protects the population, employees and environment from radiation hazards. According to the emergency operating procedures for the Ignalina nuclear power plant ∼110 m 3 of hydrogen is released to the ALS compartments during the maximum design basis accident. However, in case of beyond design basis accidents, when the oxidation of zirconium starts, the amount of generated hydrogen could be significantly higher. The volume concentration of hydrogen in the compartment might reach 4 % of a combustible mixture. To prevent a possible hydrogen accumulation in the ALS of the Ignalina nuclear power plant during an accident, the H 2 control system is installed. The analysis performed resulted in the identification of places of possible H 2 accumulation in the ALS compartments during transient processes and the assessment of the mixture combustibility in these places for a postulated accident scenario. The results also allow to evaluate the influence of the proposed severe accident management strategy on the ALS conditions.

During the whole life of reactor equipment it has to be shown that safety relevant limitations of general reactivity characteristics are not exceeded. Usually, safety margins of these characteristics are listed in a technical certificate (reactor passport). This paper presents the calculation results of the Ignalina nuclear power plant unit 2 reactor passport characteristics obtained by the best estimate 3D computer code QUABOX/CUBBOX, adjusted for the RBMK-1500 reactor core. The characteristics determined are supported by an uncertainty and sensitivity analysis which was performed using the software package SUSA. The uncertainty of the reactivity coefficients and associated effects were quantified in terms of statistical tolerance limits. The sensitivity analysis provided a ranking of the uncertain input parameters with respect to their contribution to the uncertainty of the reactivity characteristics.

Recovered uranium (RU) is a by-product of many light-water reactor (LWR) fuel recycling programs. A fissile content in the RU of 0.9 to 1.0 % makes it impossible for reuse in an LWR without re-enrichment, but CANDU reactors have a sufficiently high neutron economy to use RU as fuel. The Institute for Nuclear Research (INR) Pitesti has analyzed the feasibility of using RU fuel with 0.9 – 1.1 w% 235 U in the CANDU-6 reactors of the Cernavoda Nuclear Power Plant (Cernavoda NPP). Using RU fuel would produce a significant increase in the fuel discharge burnup, from 170 MWh/kgU currently achieves with natural-uranium (NU) fuel to about 355 MWh/kgU. This would lead to reduced fuel-cycle cost and a large reduction in spent-fuel volume per full-power-year of operation. The RU fuel bundle design with recovered uranium fuel, known as RU-43, is being developed by the INR Pitesti and is now at the stage of final design verification. Early work has been concentrated on RU-43 fuel bundle design optimization, safety and reactor physics assessment. The changes in fuel element and fuel bundle design contribute to the many advantages offered by the RU-43 bundle. Verification of the design of the RU-43 fuel bundle is performed in a way that shows that design criteria are met, and is mostly covered by proof tests such as flow and irradiation tests. The most relevant calculations performed on this fuel bundle design version are presented. Also, the stages of an experimental program aiming to verify the operating performance are briefly described in this paper.

Three dimensional CFD simulations of the moderator flow in the calandria vessel of a heavy water reactor are performed using the PHOENICS CFD code. The model includes the entire calandria vessel consisting of three shells, calandria tubes and inlet and outlet nozzle openings. The computational model prepared in PHOENICS consists of (a) standard k-∊ turbulence model, (b) PARSOL technique for handling curved objects in cartesian grids and (c) Boussinesq formulation for handling variable density flows. PHOENICS is validated by applying it to three different flow cases. The flow pattern in the calandria vessel under normal operating conditions is obtained through simulation. The effect of the presence of calandria tubes and heat generation on moderator flow pattern is studied. The simulation is also performed for various heat loads and moderator mass flow rates. The maximum temperature achieved by the moderator flow under various heat loads and moderator mass flow rates is obtained.

The German regulatory PSA guide describes the fundamental requirements with respect to probabilistic safety assessments as a central element of comprehensive safety reviews. The technical details regarding the performance of PSA are set out in two technical documents (PSA Methods and PSA Data), which have been developed by a working group of PSA experts (FAK PSA). Based on the experiences gained from the first series of periodic safety reviews in Germany, on international experience and on the fact that comprehensive safety reviews are now mandatory since April 2002, all guides hat to be revised. The first guide updated was the PSA guide together with the corresponding technical documents, issued October and November 2005. This article describes the key points of the revised requirements on PSA methods and data as well as their associated potential to improve emergency planning on nuclear sites.

The safety evaluation of the primary core cooling system requires an accurate hydraulic analysis of the system during normal and abnormal conditions. A model of the piping components of a common research reactor system is obtained and illustrated using the FE/PIPE Finite Element Package. Different accidents leading to a loss of the system function are investigated, such as pipe rupture and valve failure. The hydraulic-structure interaction and dynamic parameters responded to the accident are obtained during the transient time. The scenario time domain is selected to be very short (2.0 s) to get more conservative calculations and more safety margins in the final design. The results reflect that the natural system frequency should be in the margin greater than 34 Hz otherwise resonance with the generated forces during the accident will happen, that could lead to a system failure.

The uncertainty in calculations of proton induced reaction cross-sections using modern nuclear models and codes has been investigated. Cross-sections calculated with TALYS-0.72, ALICE/ASH, DISCA, and MCNPX codes have been compared with experimental data available in EXFOR for (p, x) reactions at proton incident energies up to 150 MeV for target nuclei from 24 Mg to 209 Bi. In addition, the comparison of evaluated data from ENDF/B-VII, JENDL-HE, and PADF calculations with measured cross-sections has been performed. Various statistical criteria were applied for the quantification of the difference between results of calculations, evaluated and experimental data.

In this work we present a LTS N formulation for the isotropic neutron transport problem in a slab assuming c = 1. The LTS N solution for c = 1 is modified by applying the Schur decomposition and Heaviside expansion techniques. Numerical results are reported.

The critical slab problem of the one-speed and one dimensional neutron transport equation for an isotropic homogeneous medium was studied by using the Chebyshev polynomial approximation, i. e. U N method. The results obtained by the U N method for various c values, using the two type boundary conditions which are known as Mark and Marshak boundary conditions, are presented in tables. The tables also include the results obtained by the P N method for comparison.

The Chebyshev polynomial approximation is used to solve the reflected critical slab problem for one-speed neutrons in a slab with strongly anisotropic scattering. The scattering kernel which is a combination of backward-forward-isotropic scattering and linearly anisotropic scattering is chosen in a uniform finite slab. The critical slab thicknesses are given for different degrees of reflection, backward-forward and linear anisotropy. Calculated numerical results for the critical thickness are in good agreement with the results available in literature.

The fundamental mode of a CANDLE burn-up wave, propagating into fertile material, in one-dimensional geometry was derived analytically with the result that the fluence and the neutron flux shapes turned out to be a Gudermannian and a Secans Hyperbolicus, respectively, of a Galilean variable. Furthermore, the fundamental evolution equation for the neutron flux was represented and discussed thoroughly. All other higher dimensional cases, however, have to be treated by numerical methods.