Volume 67, Issue 2-3

During the annual maintenance of the VVER-440 type reactors, the reactor pressure vessel, the cooling pond and the transfer pond form a connected flow domain. The reactor is cooled by the natural circulation, which develops in one or two main loops. The cooling pond has its own cooling loops. CFD calculations have been performed with the code CFX-4.3 to investigate whether it is possible to cool the reactor core in case the main loops are lost and other emergency systems are not available. The results point out that the cooling system of the cooling pond is not capable of cooling the reactor core with the present connection. Therefore, modifications of the cooling system are investigated which would make it suitable for removing the remanent heat from the core.

A process-oriented simulation model (POS-model) for common cause failures (CCF) is presented. This model consists of the following sequence of stochastic variables: time of CCF impact, number of components in the group affected by the impact, times of failure of the impacted components, time of detection of the CCF-process. As compared to the considerable number of modeling approaches that directly yield expressions for the probability of the number of failed components the POS-model seems more complicated but – thanks to a variety of interesting properties – might be useful as a complementary analytical tool. The results of a couple of applications to available field data are presented. In these exploratory examples, the adaptability of the POS-model is demonstrated and results are compared with various other modeling approaches. Finally, further planned development steps for the POS-model are addressed, primarily the optimization of parameter estimation.

The effect of the asymmetric distribution of activated water in PNA (pulsed neutron activation) measurements has been investigated experimentally by depositing a small amount of colour, simulating the activated water, in a transparent Plexiglas pipe. Based on the colour experiments, a semi-empirical model has been developed that describes the distribution of the activated water at different distances from the activation point. The model shows that the combination of inhomogeneous activation and a radial velocity profile makes the mean velocity of the activity lower than the mean velocity of the water. It can also be seen that the velocity of the activity increases as the distance from the activation point increases. The model has been compared with experimental values from PNA measurements and the measured mean velocity shows a similar dependence on the distance from the activation point.

To solve the large-group diffusion adjoint equation the averaged group constants have to be pre-calculated. However, at this point the forward and adjoint fluxes are unknown. Thus, an alternative method, independent of the solution of this equation, has to be provided. This method is divided into two steps. In the first, the forward and adjoint spectra should be obtained for a unit cell assuming infinite medium. The second step consists of the correction for a finite medium by approximating the spatial dependence by a single Fourier fundamental mode in terms of the geometrical buckling B 2 . This paper is dedicated to the solution of the second step. Therefore, forward and adjoint P 1 equations were both obtained and numerically solved in lethargy space in order to provide the neutron fluxes in a finite unit fuel cell. Moreover, the direct, adjoint and bilinear weighting for calculation of macrogroup constants is discussed.

This paper describes experimental and numerical investigations of sub-cooled boiling, condensation, and void flashing in the HRTL-5 test loop, which simulates the primary loop of a 5 MW nuclear heating reactor. A drift-flow model of two-phase with four governing equations was used, in which sub-cooled boiling, condensation, and void flashing have been taken into account. Based on the mathematical model, a program has been developed for analyzing the natural circulation system. As parameters, inlet sub-cooling, system pressure, and heat flux are varied. For comparison, some simplified models, which are designed to reveal the importance of sub-cooled boiling, condensation, flashing in the HRTL-5 test loop, are adopted in the program. The results show: (1) sub-cooled boiling, condensation, and void flashing may have great influence on the distribution of the void fraction and more intense at low system pressure; (2) the calculation of them is correlative and interactive other than independent; (3) for a system with short heated section, long riser, and low pressure, it is possible to reach “boiling out of the core”, where there is almost no void in the heated section, but much in the riser.

Analytical investigations have been carried out to study the flow instability behaviour in a boiling two-phase natural circulation loop under low quality conditions. For this purpose, the computer code TINFLO-S has been developed. The code solves the conservation equations of mass, momentum and energy and equation of state for homogeneous equilibrium two-phase flow using linear analytical technique. The results of the code have been validated with the experimental data of the loop for both the steady state and stability. The study reveals that the stability behaviour of low quality flow oscillations is different from that of the high quality flow oscillations. The instability reduces with increase in power and throttling at the inlet of the heater. The instability first increases and then reduces with increase in pressure at any subcooling. The effects of diameter of riser pipe, heater and the height of the riser on this instability are also investigated.

This paper presents the outline of the core thermal hydraulic design and analysis (Operational Safety Analysis) of the Budapest nuclear research reactor (WWR-M2 type), which is a tank-type, light water-cooled nuclear research reactor with 36% enriched uranium coaxial annuli fuel. The research reactor is currently upgraded to 10 MW th of thermal power, while the cooling capacity of the reactor was designed and constructed for 20 MW th . This reserve in the cooling capacity serves redundancy today but can be used for future upgrading too. The core thermal hydraulic design was, therefore, done for the normal operation conditions so that fuel elements have enough safety margins both against nucleate boiling anywhere in the reactor core. Thermal hydraulic performance was studied. It is shown that the 36% enriched UAl x -Al fuel elements in WWR-SM fuel coolant channel do not allow to force up the reactor power to 20 MW th . The study was carried out for an equilibrium core, with compact load (223 fuel assemblies) under normal operation conditions only (steady state condition).

The second Egyptian research reactor ET-RR-2, is a pool type reactor. A sudden loss of pool water would leave the core region uncovered. The reactor core is surrounded by chimney chambers with water isolated from the pool water. This accident would lead to significant external doses. A model is developed and used to calculate the dose rates for key access-areas and traffic plans from indirect line of sight of the core which have a maximum dose rate. The model developed uses the discrete ordinate method as implemented in the code DOT3.5.

A Fire Probabilistic Safety Assessment – called the Fire PSA – is being carried out by the French Institute of Radiological Protection and Nuclear Safety (IRSN) to be used in the framework of the safety assessment of operating 900 MWe PWRs. The aim of this study is to evaluate the core damage conditional probability which could result from a fire. A fire can induce unavailability of safety equipment, notably damaging electrical cables introducing a significant risk contributor. The purpose of this paper is to present the electrical cable fire tests carried out by IRSN to identify the failure modes and to determine the cable damage criteria. The impact of each kind of cable failure mode and the methodology used to estimate the conditional probability of a failure mode when cable damage occurred is also discussed.

Für die ingenieurgemäße Untersuchung brandschutztechnischer Fragestellungen, die im Rahmen von individuellen Brandschutzkonzepten vorgenommen werden, ist häufig die Kenntnis von Umgebungstemperaturen im Umfeld des Brandereignisses wesentlich. Dabei stehen Fragen nach dem Auslösezeitpunkt von Sprinklern, die Temperaturbeaufschlagung von Bauteilen und das Erreichen kritischer Temperaturen Für Menschen oder Baustoffe im Vordergrund. Vielfach wird diese Berechnung mit Zonenmodellen oder überschlägig mit Plume-Ansätzen vorgenommen. Bei dieser Vorgehensweise werden häufig lokale Effekte nicht beachtet oder in Fällen, bei denen sich während des Brandverlaufes eine heiße Rauchgasschicht bildet, zu geringe Plume-Temperaturen berechnet. Mit Hilfe von modifizierten Ansätzen, wie sie in diesem Bericht vorgestellt worden sind, besteht die Möglichkeit, realistischere Werte zu berechnen, da sie insbesondere eine Vermischung der im Plume aufsteigenden Verbrennungsgasen mit den heißen Rauchgasen einer ausgebildeten Rauchgasschicht mit endlicher Temperatur in die Betrachtung einbeziehen.

For purposes of emergency planning for Swiss nuclear power plants the following scenarios were selected from among a great variety of possible sequences: scenarios without core damage, scenarios with core damage and correct functioning of the containment and the filtered venting system, and scenarios with core damage and without correct functioning of the containment – medium up to long pre-phase. The selection of these scenarios serves as basis for establishing measures to protect the population. This contribution presents alerting procedures for scenarios without warning time.