[en] Nuclear energy is almost free of CO${}_2$ emissions, not only the nuclear power plant alone, but the entire life cycle from nuclear fuel production to nuclear waste disposal. Therefore, many countries are including nuclear energy in their strategy to decarbonize their electricity generation. In today's nuclear power plants, the main isotope that is split is ${}^{235}$U, which is only 0.72% contained in uranium. These nuclear power plants leave behind waste that is radioactive for a very long time. To eliminate these disadvantages, many places are working on a new generation of reactors. They are based on the conversion of the isotope ${}^{238}$U and alternatively thorium into new fissile material, a process called breeding. Furthermore, they can convert long-lived components in high-level waste into fission products that decay more quickly, which is called transmutation. The enormous potential expansion of the range of energy resources and the reduction of the burden on the disposal of long-lived, mainly alpha-toxic components are of great interest for the sustainable use of nuclear energy into the distant future. In addition to a further increase in nuclear safety, economic efficiency and proliferation resistance, this provided the main motivation for the development of these new types of nuclear reactors, which are intended to replace today's technology in the long term. This article presents the different technological approaches. It also discusses the extent to which this will result in new requirements for radiation protection. However, an exhaustive analysis is not possible due to the breadth of the subject matter. Fundamental differences to the plants currently in operation and questions for further investigations are identified.

[en] For the investigation of stratified two-phase flow, two horizontal channels with rectangular cross-section were built at Forschungszentrum Rossendorf. The channels allow the investigation of air/water co-current flows, especially the slug behaviour, at atmospheric pressure and room temperature. The test-sections are made of acrylic glass, so that optical techniques, like high-speed video observation or particle image velocimetry (PIV), can be applied for measurements. The rectangular cross-section was chosen to provide better observation possibilities. Moreover, dynamic pressure measurements were performed and synchronized with the high-speed camera system. CFD post test simulations of stratified flows were performed using the code ANSYS CFX. The Euler- Euler two fluid model with the free surface option was applied on grids of minimum 4.10⁵ control volumes. The turbulence was modelled separately for each phase using the k-ω based shear stress transport (SST) turbulence model. The results compare very well in terms of slug formation, velocity, and breaking. The qualitative agreement between calculation and experiment is encouraging and shows that CFD can be a useful tool in studying horizontal two-phase flow. (authors)

[en] The extension of the worldwide light water reactor fleet will cause the demand for uranium to grow. The static reach of identified resources might soon fall below the life time of new nuclear power plants which are usually designed for 60 years of operation, if the exploration of new uranium deposits will stop resulting in exploitable resources. The article discusses, if, as frequently claimed, the energy consumption in the uranium mines renders impossible to secure the nuclear fuel supply in the long term. (orig.)

[en] Highlights: • Code comparison on the performance of a Passive Containment Condenser. • Simulation of separate effect tests with pure steam and non-condensable gases. • Role of the secondary side and accuracy of pool boiling models are discussed. • GOTHIC and TRACE predict the experimental performance with slight underestimation. • Recirculatory flow pattern with injection of light non-condensable gas is inferred. - Abstract: Typical passive safety systems for ALWRs (Advanced Light Water Reactors) rely on the condensation of steam to remove the decay heat from the core or the containment. In the present paper the three-dimensional containment code GOTHIC and the one-dimensional system code TRACE are compared on the calculation of a variety of phenomena characterizing the response of a passive condenser submerged in a boiling pool. The investigation addresses the conditions of interest for the Passive Containment Cooling System (PCCS) proposed for the ESBWR (Economic Simplified Boiling Water Reactor). The analysis of selected separate effect tests carried out on a PCC (Passive Containment Condenser) unit in the PANDA large-scale thermal-hydraulic facility is presented to assess the code predictions. Both pure steam conditions (operating pressure of 3 bar, 6 bar and 9 bar) and the effect on the condensation heat transfer of non-condensable gases heavier than steam (air) and lighter than steam (helium) are considered. The role of the secondary side (pool side) heat transfer on the condenser performance is examined too. In general, this study shows that both the GOTHIC and TRACE codes are able to reasonably predict the heat transfer capability of the PCC as well as the influence of non-condensable gas on the system. A slight underestimation of the condenser performance is obtained with both codes. For those tests where the experimental and simulated efficiencies agree better the possibility of compensating errors among different parts of the heat transfer models is outlined. Besides, the pool side heat transfer coefficient is indicated to differ between GOTHIC and TRACE, revealing that the model implemented in TRACE might be more accurate. Finally, both codes deviate more from the experiments for the tests with helium, which is explained by the emerging of a complex circulatory flow pattern in the experiments that the models fail to predict

[en] Cyclic temperature fluctuations leading to the formation and propagation of cracks in industrial piping systems is an ongoing area of research and concern in the nuclear thermal hydraulics field. This paper focuses on turbulent penetration of a fast main pipe flow into the branch line of T-junctions, especially when the branch is not completely 'dead,' as in a leaking valve scenario with very high main pipe/branch line velocity ratios. A large eddy simulation of turbulent penetration in a T-junction using the WALE (Wall-Adapting Local Eddy-viscosity) subgrid scale model has been performed. A highly resolved prismatic-polyhedral grid (∼6 million cells) is utilized for the simulation of 13 seconds of flow time. A velocity ratio of 100 between the main pipe (D_m = 50 mm) and branch line (D_b = 26 mm) leads to turbulent penetration up to four branch diameters. Furthermore, a steady region of velocity shear leads to the formation of Kelvin Helmholtz Instabilities as the main flow boundary layer separates from the wall at the start of the branch line. The shearing of these waves is the likely cause of preferred scalar mixing frequencies in the branch. A comparison of the scalar mixing spectrum from the simulation with a broad peak in the mixing spectrum detected in experimental results, around 6 Hz, shows good qualitative agreement. Finally, the scalar average and RMS are found to be well reproduced by the LES simulations albeit with an underproduction of the mixing. (authors)

[en] Highlights: • Prediction of signals of electrical liquid film thickness sensors by potential field simulations. • Analysis of the sensor response to waves travelin gover them. • Quantitative information on measuring uncertainties in case of wavy films is given. The electrical liquid film thickness (LFT) sensor developed at ETH Zurich is applied to study wavy annular flows. The LFT sensor consists of a matrix of electrodes put flush to the surface of the wall at which the liquid film of interest is present. The conductance between transmitter and receiver electrodes is sampled and converted into a film thickness using a calibration function obtained for a flat liquid film without waves. Due to limited lateral resolution and non-linear sensor response, waves are characterized with measuring errors, which depend on the wave height, length and angle of attack. The paper presents the result of potential field simulations of waves passing over the sensor surface. The wave parameters obtained from the simulated sensor signals are compared to the input. The results are used to quantify the uncertainty of dynamic film thickness measurements. The obtained detailed information on the sensor response allows a better interpretation of experimental results.

[en] Wire-mesh sensors are widely used to characterize gas-liquid two-phase flows and single-phase mixing processes. The geometry of the electrode grids and the way of the signal readout generates a three-dimensional electrical field in the vicinity of the electrode wires. Resulting electrical currents at the receiver electrodes, representing the primary measuring information, are calculated by a three-dimensional potential field simulation within the sensitive volume formed by the electrode wires, whereas bubbles are taken into account as simplified, spherical or elliptical objects placed at different locations in the calculation domain. The response of the sensor to the passage of such synthetic bubbles is studied. A significant deviation from the linear dependency between the received current and the local instantaneous gas fraction is found. Overshoots of the current above the reference value obtained by calibration in plain liquid occur. Furthermore, the response of the sensor depends on the axial distance between the transmitter and the receiver electrode grids. Swarms of bubbles of small size passing through the grids of the wire-mesh sensor lead to an average decrease of the current which can be described by the average conductivity of an emulsion according to Maxwell.

[en] The reactor transient caused by a perturbation of boron concentration or coolant temperature at the inlet of a pressurized water reactor (PWR) depends on the mixing inside the reactor pressure vessel (RPV). Initial steep gradients are partially lessened by turbulent mixing with coolant from the unaffected loops and with the water inventory of the RPV. Nevertheless the assumption of an ideal mixing in the downcomer and the lower plenum of the reactor leads to unrealistically small reactivity inserts. The uncertainties between ideal mixing and total absence of mixing are too large to be acceptable for safety analyses. In reality, a partial mixing takes place. For realistic predictions it is necessary to study the mixing within the three-dimensional flow field in the complicated geometry of a PWR. For this purpose a 1:5 scaled model [the Rossendorf Coolant Mixing Model (ROCOM) facility] of the German PWR KONVOI was built. Compared to other experiments, the emphasis was put on extensive measuring instrumentation and a maximum of flexibility of the facility to cover as much as possible different test scenarios. The use of special electrode-mesh sensors together with a salt tracer technique provided distributions of the disturbance within downcomer and core entrance with a high resolution in space and time. Especially, the instrumentation of the downcomer gained valuable information about the mixing phenomena in detail. The obtained data were used to support code development and validation. Scenarios investigated are the following: (a) steady-state flow in multiple coolant loops with a temperature or boron concentration perturbation in one of the running loops, (b) transient flow situations with flow rates changing with time in one or more loops, such as pump startup scenarios with deborated slugs in one of the loops or onset of natural circulation after boiling-condenser-mode operation, and (c) gravity-driven flow caused by large density gradients, e.g., mixing of cold emergency core cooling (ECC) water entering the RPV through the ECC injection into the cold leg. The experimental results show an incomplete mixing with typical concentration and temperature distributions at the core inlet, which strongly depend on the boundary conditions. Computational fluid dynamics calculations were found to be in good agreement with the experiments

[en] Filtered Containment Venting System (FCVS) is a safety system equipped in nuclear power plants to release steam and non-condensable gases from the containment during a severe accident scenario, in order to avoid over-pressurizing the containment, while retaining most radioactive fission products inside the plant. This paper presents the RELAP5/Mod3.3-p3 and TRACE/V5.0-p2 simulations of the experiments performed in the VEFITA (Venting Filter Assessment) test facility at Paul Scherrer Institut (PSI). This experimental program is designed to investigate thermal-hydraulic phenomena in the FCVS such as internal recirculation, water inventory, and level swell. The water level swell during the gas injection is particularly interesting since it would affect the design and operation of the venting filtration system. Test section is a vessel with a 0.58 m diameter and 12 m total height. The lower wet scrubber section is filled with water to a certain level above the gas injection nozzle. Non-condensable gas at various mass flow rates are fed through the injection nozzle into the test section. RELAP5 and TRACE code models are developed for the test facility and several series of experiments are simulated. The system code simulation results, as well as a previous physical model based on homogeneous-heterogeneous regimes transition in bubble columns, are compared with the experimental data. (authors)

[en] Void fraction in fuel assemblies of a Boiling Water Reactor (BWR) is typically investigated using system codes through various closure relationships and assumed simplified interfacial structures. Interface tracking methods, on the other hand, solve directly for the interface and, hence, evaluate the interfacial terms accurately. This work investigates adiabatic annular flow downstream of a swirl type spacer in double subchannel geometry using the Volume of Fluid (VOF) method and compares with experiments, conducted at the ETHZ annular flow facility, to investigate liquid film behavior driven by turbulent air flow. Large eddies in the computations were resolved on the grid level while the subgrid scales were modeled using the Wall Adapting Local-Eddy-viscosity (WALE) model. We present time-averaged results, which compare reasonably well with the experiments. The simulations were performed using the commercial code STAR-CCM+. (authors)