Band 69, Heft 1-2

The recent availability of powerful computer and computational techniques has enlarged the possibilities to perform best estimate simulations of complex scenarios in nuclear power plants. Nowadays, the coupled codes method, which consists in incorporating three-dimensional (3D) neutron modeling of the reactor core into system codes, is extensively used. It is particularly suited for simulating transients that involve core spatial asymmetric phenomena and strong feedback effects between core neutronics and reactor loop thermal-hydraulics. Within this framework the Peach Bottom BWR 2 Turbine Trip test was selected since it involves a rapid pressure induced positive reactivity addition into the core. It is also characterized by a self-limiting power course due to compensated inherent reactivity mechanisms. To perform a numerical simulation of the turbine trip a reference case was run for the coupled thermal-hydraulic system code RELAP5/mod3.3 and 3D neutron kinetic PARCS/2.3 code. The overall data comparison shows good agreements between the calculations and most of the significant global aspects observed experimentally. However, the test is very sensitive to the feedback modeling and requires a tightly accurate simulation of the thermal-hydraulic and the cross sections parameters. For this purpose, sensitivity studies have been carried out in order to identify the most influential parameters that govern the transient behavior. The considered cases showed that the self-limiting power amplitude as predicted by the coupled code calculation is mainly due to delayed feedback mechanisms whereas the experimental data shows that the power quenching before the Scram is governed by prompt feedback effects.

A theoretical and experimental investigation has been carried out to study natural circulation characteristics of an Indian PHWR under reduced inventory conditions. The theoretical model incorporates a quasi-steady state analysis of natural circulation at different system inventories. It predicts the system flow rate under single-phase and two-phase conditions and the inventory at which reflux condensation occurs. The model predictions were compared with test data obtained from FISBE (Facility for Integral System Behaviour Experiments), which simulates the thermal hydraulic behaviour of the Indian 220 MWe PHWR. The experimental results were found to be in close agreement with the predictions. It was also found that the natural circulation could be oscillatory under reduced inventory conditions.

During irradiation in the Halden Reactor data such as linear rating, coolant temperature, or reactor state are routinely stored every 15 minutes, in the case of special events even at shorter time intervals. Because of the long irradiation times of up to several years, a huge amount of data is produced which is considered to be too much for direct input into fuel performance codes. In order to reduce the amount of data, various data-condensation procedures have been developed, which are briefly discussed. An innovative data-condensation method based on multidimensional minimisation is presented. The condensation factor obtained so far for Halden irradiations is 25–40, which is sufficient for a modern fuel rod performance code like TRANSURANUS. This new condensation method avoids all loss of experimental data and doubtful averaging of temperatures associated with other more drastic condensation methods. Up to now, the so-called reactor state variable, which defines the reactor operation (e. g. constant, decreasing or increasing power), has not been used. The introduction of this variable should further improve this new condensation method.

For nuclear power plants fire events are of high safety significance since they have the potential to damage more than one redundant safety train simultaneously. Within the frame of the recently started second series of periodic safety reviews, a probabilistic fire risk assessment has to be performed in addition to the deterministic analysis. Therefore, realistic fire occurrence frequencies as well as appropriate validated fire models and simulation codes are needed. All these complex issues require internationally harmonized activities to gain consistent, reliable and traceable information as well as to reduce the costs. Presently, three international activities are ongoing: the OECD-FIRE project with the goal to establish a quality assured and consistent international fire event database, and two further activities. In the frame of one of these fire models and computer codes shall be evaluated and validated based on real scale experiments under as far as possible realistic boundary conditions in order to support their extended use for decision making. The other one aims on analyzing fire induced electric failures of safety related systems and components. The objectives of the international co-operations and the first results are presented.

A set of formulas for conservatively computing the fuel rod cladding strains caused by short and intense energy pulses is presented. Such pulses may be the consequence of a RIA. The formulas are based on the theory of stresses and deformations of tubes loaded by internal pressure and stressed beyond the elastic limit. The known solutions of this theory are applied to compute the cladding strains due to the intense interaction between fuel and cladding during a RIA. Simple and easy to evaluate expressions for the strains are obtained. They enable to quickly assess the impact of any energy pulse for any burn-up state. The actual burn-up state of the fuel rod at the moment in which the occurrence of a RIA shall be simulated is described by only three data. These data can be calculated in an enveloping sense in advance and then be used to study energy pulses of any intensity. Besides the cladding strains induced by the fuel-cladding interaction the significance of the gas pressure in the rod and of gas swelling for the cladding strains is also addressed.

Accelerator Driven Systems (ADS) are an old technological idea: Relativistic proton accelerators deliver their beams onto massive heavy element targets, thus producing abundant neutron fluences. Placing this target into sub-critical nuclear fission assemblies is yielding substantial fission reactions, thus additional fission energy (Rubbia called such a system “Energy Amplifier”). This technology has recently attracted considerable attention due to advances in the construction of powerful accelerators. It allows the safe and cheap production of nuclear energy simultaneously with the destruction (Transmutation) of long lived radioactive waste, in particular plutonium and other minor actinides (neptunium and americium). The principles and the present-state-of-the-art are described, including first experiments to transmute plutonium this way. This technology needs, however, many more years of further “research and development” before large scale ADS's can be constructed. It may be even necessary to investigate the question, if all basic physics phenomena of this technology are already sufficiently well understood.

The feasibility of an experimental methodology for in-situ quantification of oxygen in bulk oil using fast neutron activation analysis (FNAA) has been studied. The method was applied for determination of oxygen in 100 ml (∼ 90 g) rapeseed oil. The amount of oxygen in the rapeseed oil using the in-situ FNAA was estimated to 10.6 ± 2.6 weight %. Using cyclic fast neutron activation analysis (cFNAA), the amount of oxygen in the oil was determined in average 9.9 ± 0.4 weight %. Based on Monte Carlo calculations on water, the optimal radius and height of a cylindrical container where the activity is distributed through stirring of the water would be about 10 cm and 44 cm, respectively. These dimensions give a volume of about 14 liters, which is suitable for any type of oil. The accuracy in the in-situ FNAA can be increased by a more precise determination of oxygen in rapeseed oil in the beginning of a dynamic process using cFNAA or alternatively by a better background subtraction.

In investigations of the long-term development of the population dose in the highly contaminated regions of the Commonwealth of Independence States it was found that the external dose has not decreased as strongly as expected since 1992. Further investigations have shown that, contrary to expectations, no linear correlation can be observed between soil contamination and measured area dose rate. As a contribution towards clarifying these issues, the area dose rate and the soil contamination including the plant fraction were investigated in the Korma district, Belarus. It was found that it is necessary to cover and average over larger areas in order to determine from ground contamination the long-term development of the external dose commitment. This means that for this purpose the introduction of an “effective” surface contamination (sum of mineral and organic contamination components) is necessary. The phenomena observed are described in a model, which permits an analytical calculation of the contamination profile in soil taking migration and transfer effects into account. The differences observed between the measured soil contamination and the resulting external doses or the directly measured dose rate can be explained by the proposed model. Moreover, their long-term development can be calculated. The results show that a time decade after the accident the biological part of the “effective” soil contamination becomes dominant and cannot be neglected.