The development of techniques for the surveillance of LMFBRs

The development of surveillance techniques of LMFBRs is determined by the interaction of three factors: the specification of requirements, improvements in technique and the physical analysis of the processes involved. The specification of requirements, which sets the structure for the discussion, is mainly concerned with public safety. Two main divisions are identified: those concerned with thermal events in the nuclear core and those concerned directly or indirectly with the mechanical integrity of components. The necessary developments are then discussed in terms of the signal analysis techniques to anticipate various modes of failures. The importance of an adequate understanding of the failure mode is emphasised in optimising the surveillance technique.

Core surveillance may be achieved by monitoring individual sub-assemblies or by monitoring bulk conditions. The important features of sub-assembly monitoring are discussed and the advantages of temperature analysis explained. The specification of the temperature-monitoring systems is identified and the conflicting requirements for the reactor sensor discussed, viz adequate band width as against a robust and reliable construction. A theoretical treatment using Monte Carlo techniques allows a full examination of the choice of method of temperature analysis. This shows that, although a filtered rms value has been the preferred choice for detecting either local blockage or sodium boiling, it may be possible to distinguish the temperature signals of blockages from those of power gradients by an amplitude probability density plot. The advantages of acoustic monitoring using the noise of boiling sodium to detect overheating, leading to core damage, are examined. An important consideration is the thermal-acoustic process of sodium boiling, and evidence is submitted from a range of out-of-pile experiments involving local sub-cooled boiling and bulk boiling in discussing the merits of pulse analysis and power spectral density techniques. An important factor in discriminating background from signal is the extent of cavitation in reactor components. Experiments are mentioned in which pulse techniques have been used to locate boiling sources by spatial correlation. The interpretation of reactor signals requires a detailed knowledge of the transmission of acoustic waves in reactor pools and structures and the effect of gas bubbles. Measurements in PFR and sodium loops have helped to lead to a more quantitative assessment of the sensitivity of the acoustic techniques.

Structural integrity depends on detecting failure modes, particularly those arising from crack propagation. Manufacturing defects or pre-existing cracks may be identified by ultrasonic inspection or by stress-wave emission. On-line monitoring for stress-loaded cracks by a stress-wave emission is seen as intrinsically difficult because of low signal strength and high attenuation but initial experiments have indicated possibilities for detecting stress-corrosion cracking. Mechanical failure from fatigue may be anticipated from a understanding of the vibrational modes of the sodium and its coupling with the structure. A one-eighth scale model of a LMFBR design has recently demonstrated the likely vibrational modes. A major handicap in supervising mechanical operation in sodium systems is the opacity of the sodium. Visualisation techniques of the major parts of the core structure are being developed. An important aspect is the study of the information processing required to present an image easy for the reactor operator to understand. Advances may be made using transform methods to improve object boundaries by modifying the spatial frequencies of the display or record.     

Results of acoustic and temperature noise measurements on sodium cooled electrically heated 7 rod bundle

Assembly cooling deficiency in a LMFBR is one of the most important safety problems for reactor design and operation.

Studies on early detection and diagnosis of local accident by means of noise analysis techniques have been initiated at CNEN. Acoustic and temperature noise measurements have been carried out on a 7 rod bundle during slow power transients up to boiling conditions. The test section, simulating the italian PEC reactor fuel element, was mounted on ENA-2 sodium loop located at the CSN Casaccia.

Acoustic noise spectral analysis up to 32 kHz shows the appearance, in presence of boiling, of power increase at certain frequencies. Power spectra and rms values are updated and recorded every 0.3 sec and show large variations going from single phase to boiling.

Temperature noise spectral analysis shows that the power, between 1 and 50 Hz, increases, in presence of boiling, by a factor bigger than 30. It has been tested the sensitivity of other indicators of the temperature fluctuations, like skewness and flatness, to reveal boiling.     

The detection of sodium vapor bubble collapse in a liquid metal fast breeder reactor

Sodium boiling detection utilizing the sound pressure emanated during the collapse of a sodium vapor bubble in a subcooled media is discussed in terms of the sound characteristic, the reactor ambient noise background, transmission loss considerations and performance criteria. Data obtained in several loss of flow experiments on Fast Test Reactor Fuel Elements indicate that the collapse of the sodium vapor bubble depends on the presence of a subcooled structure or sodium. The collapse pressure pulse was observed in all cases to be on the order of a kPa, indicating a soft type of cavitational collapse. Spectral examination of the pulses indicates the response function of the test structure and geometry is important. The sodium boiling observed in these experiments was observed to occur at a low (<50°C) liquid superheat with the rate of occurrence of sodium vapor bubble collapse in the 3 to 30 Hz range. Reactor ambient noise data were found to be due to machinery induced vibrations, flow induced vibrations, and flow noise. These data were further found to be weakly stationary enhancing the possibility of acoustic surveillance of an operating Liquid Metal Fast Breeder Reactor. Based on these noise characteristics and extrapolating the noise measurements from the Fast Flux Test Facility Pump (FFTP), one would expect a signal to noise ratio of up to 20 dB in the absence of transmission loss. The requirement of a low false alarm probability is shown to necessitate post detection analysis of the collapse event sequence and the cross correlation with the second derivative of the neutronic boiling detection signal. Sodium boiling detection using the sounds emitted during sodium vapor bubble collapse are shown to be feasible but a need for in-reactor demonstration is necessary.     

Borssele PWR noise: Measurements, analysis and interpretation

In the Borssele reactor — a 450 MWe PWR — reactor noise measurements have been performed during four fuel cycles. Measurements were made with a set of ex-core neutron detectors, on one occasion an in-core displacement transducer, and with primary coolant pressure sensors. Digital analysis was applied, where the most powerful tool was the computer programme FAST, which computes auto and cross power spectra for all combinations from a set of many simultaneously recorded signals.

Analyses of neutronic signals show a reactivity noise peak at 9.2 Hz, core barrel motion peaks at about 12 and 15 Hz, a damped oscillation at about 2 Hz. Results are given for begin and end of each fuel cycle. The r.m.s. value of the low frequency noise appears to depend linearly on the boron concentration over a wide range.

Also some results of primary coolant pressure noise are presented, with coherent peaks below 15 Hz and incoherent peaks above.

The second part of the paper describes an alternative way of analyzing and interpreting noise spectra, namely attempts to decompose the neutronic power spectra into physical components, using the information present in the CPSD's of all detector combinations. The components are characterized by their detector position dependency. Effects considered are: uncorrelated noise, global reactivity noise, core motion attenuation noise, and a possible coupling between reactivity and core motion. Results show excellent separation into reactivity and core motion components with possibilities to separate overlapping peaks. Weak peaks become more easily detectable. At low frequencies the decomposition of the spectra is not yet complete, however.     

Development of fuel failure detection system for a high temperaturegas cooled reactor. II

For pt.I see ibid., vol.34, p.567 (1987). A fuel failure detection (FFD) method based on selective detection of short-life and gaseous fission products, developed for a high-temperature gas-cooled reactor, is described. An improved precipitator was used as a detector for the fission products and the performance of the FFD system was tested using an irradiation rig at the Japan Material Testing Reactor. In the rig, three kinds of samples of coated-particle fuels were irradiated and each sample of the primary helium gas was fed to the FFD system. Failure rates of the three fuel samples called intact, normal, and slightly failed, were estimated at about 10^-6, 10^-5, and 10 ^-4, respectively. The FFD system showed a significantly increased response in counting rate for the sample gas with a failure rate of 10^-4. The FFD system did not respond to the sample gas with the smaller failure rate of 10^-5 even when the background level of long-life fission products in the primary coolant gas increased with fuel temperature and reactor power     

Incipient Boiling of Sodium in an Annular Channel Heated by a Single-Pin with Blockage

Abstract

In simulation of partial failure affecting a fast breeder core, experiments on sodium boiling were performed in a single vertical channel of annular cross section (15 mm I.D.), indirectly heated by high flux heater pin (6.5 mm O.D., 60 cm heating length) and provided with a blockage disk (11mm O.D., 1mm thick, obstructing 42% of the channel area). The experimental conditions were; Pressure of cover gas: 1.0 kg/cm² abs., Heat flux: 0–80 W/cm², Flow rate of sodium: 1–4/l min.

The results revealed that the degree of incipient boiling superheat is reduced to a fraction of the value obtained previously for direct joule heating under similar conditions, and that, consequently, the intensity of pressure pulses is correspondingly reduced. On the other hand, the incipient boiling superheat increases, together with the extent of its scattering, with rising intensity of the heat flux.

Local boiling was observed to precede the onset of bulk boiling, causing small pressure spikes to be detected by all three pressure sensors installed along the test channel. The pressures evaluated from the measured velocities of the liquid sodium column based on single-bubble model were compared with the measured pressure signals, and a fairly good agreement was obtained.     

Temperature fluctuations: An assessment of their use in the detection of fast reactor coolant blockages

In this paper, the temperature noise technique for the detection of local blockages in fast reactor subassemblies is discussed. The main factors involved in an assessment of the technique are outlined and the experimental and theoretical work that has been carried out at BNL on the various aspects of the problem is described. It is concluded that blockages appreciably smaller than those predicted to produce boiling should be detectable against a background noise level due to subassembly power tilts, on a time scale giving protection against rapidly developing incidents. Further work required to increase confidence in the application of the technique to the reactor is outlined, including measurements in fully representative geometries, data from sodium rigs and further information on reactor background noise levels.     

钠冷快堆蒸汽发生器背景噪声功率谱分析方法研究

为了了解钠冷快堆蒸汽发生器内部流体运行状态、内部成分等不易测量的变化，引入现代谱估计AR模型法和传统经典法谱估计的周期图法、Welch改进周期图法、BT法对钠冷快堆蒸汽发生器运行产生的背景噪声原始数据进行了分析和比较。结果表明，采用AR模型法对钠水反应背景噪声进行谱估计分析效果更加理想。      

Experimental analysis of the power curve sensitivity test series at ROSA-III

The rig of safety assessment (ROSA)-III facility is a volumetrically scaled (1/424) boiling water reactor (BWR/6) system with an electrically heated core designed for integral loss-of-coolant accident (LOCA) and emergency core cooling system (ECCS) tests. Seven recirculation pump suction line break LOCA experiments were conducted at the ROSA-III facility in order to examine the effect of the initial stored heat of a fuel rod on the peak cladding temperature (PCT). The break size was changed from 200% to 5% in the test series and a failure of a high pressure core spray (HPCS) diesel generator was assumed. Three power curves which represented conservative, realistic and zero initial stored heat, respectively, were used.In a large break LOCA such as 200% or 50% breaks, the initial stored heat in a fuel rod has a large effect on the cladding surface temperature because core uncovery occurs before all the initial stored heat is released, whereas in a small break LOCA such as a 5% break little effect is observed because core uncovery occurs after the initial stored heat is released. The maximum PCTs for the conservative initial stored heat case was 925 K, obtained in the 50% break experiment, and that for the realistic initial stored heat case was 835 K, obtained in the 5% break experiment.     

Spatial and model-order based reactor signal analysis methodology for BWR core stability evaluation

A new methodology for the boiling water reactor core stability evaluation from measured noise signals has been recently developed and adopted at the Paul Scherrer Institut (PSI). This methodology consists in a general reactor noise analysis where as much as possible information recorded during the tests is investigated prior to determining core representative stability parameters, i.e. the decay ratio (DR) and the resonance frequency, along with an associated estimate of the uncertainty range. A central part in this approach is that the evaluation of the core stability parameters is performed not only for a few but for ALL recorded neutron flux signals, allowing thereby the assessment of signal-related uncertainties. In addition, for each signal, three different model-order optimization methods are systematically employed to take into account the sensitivity upon the model-order.     

中国实验快堆全堆芯流量分配计算与试验

针对中国实验快堆(CEFR)堆芯和一回路的设计特点,开发水力特性计算程序DAEMON,完成不同工况下的全堆芯流量分配计算,给出流量分配不均匀性等参数。在反应堆调试阶段,进行全堆芯流量分配试验。结果表明,程序计算值与试验值符合较好。在此基础上,验证了CEFR堆芯的流体力学设计,并为反应堆调试和运行提供了基础数据。     

Sodium boiling experiments in a seven-pin bundle: Flow patterns and two-phase pressure drop

Two series of quasi-steady state sodium boiling experiments have been carried out in an electrically heated seven-pin bundle. The power levels (130–170 and 30–40 W/cm²) and other test conditions were selected to correspond to the core and radial breeder zones of a typical LMFBR. The test procedure involved the gradual reduction of mass flow rate through the bundle in a simulation of the consequences of a slowly growing blockage in the lower part of a reactor subassembly. By this means it was possible to study the development of quasi-steady state boiling up to the onset of permanent dryout. The results obtained provide information on flow regimes in the two-phase region, vapour qualities and flow rates at which cooling of the bundle can be effectively maintained, and the ultimate incidence of dryout. A relation for the two-phase pressure drop multiplier obtained from adiabatic pressure drop measurements in this geometry is given and compared with earlier results from single-channel geometry tests.     

Finite Element Analysis of Vibration of Thin Elastic Plates by Simplified Hybrid Displacement Method

Temperature noise, measured by thermocouples mounted at each core fuel subassembly, is considered to be the most useful signal for detecting and locating local cooling anomalies in an LMFBR core. However, the core outlet temperature noise contains background noise due to fluctuations in the operating parameters including reactor power. It is therefore necessary to reduce this background noise for highly sensitive anomaly detection by subtracting predictable components from the measured signal. In the present study, both a physical model and an autoregressive model were applied to noise data measured in the experimental fast reactor JOYO. The results indicate that the autoregressive model has a higher precision than the physical model in background noise prediction. Based on these results, an “autoregressive model modification method” is proposed, in which a temporary autoregressive model is generated by interpolation or extrapolation of reference models identified under a small number of different operating conditions. The generated autoregressive model has shown sufficient precision over a wide range of reactor power in applications to artificial noise data produced by an LMFBR noise simulator even when the coolant flow rate was changed to keep a constant power-to-flow ratio.     

Molecular Approach to Sodium Vapour Condensation,Rewetting and Vaporization in LMFBR Bundle under Hypothetical Accident Conditions

In the frame of safety analysis of Liquid Metal Fast Breeder Reactors (LMFBRs) under hypothetical Unprotected Loss-of-Flow (ULOF) conditions, two phase flow of sodium is simulated in a reactor core. Traditional approaches used in safety analysis codes to simulate sodium vapour condensation and vaporization rely upon application of macroscopic semi-empirical correlations for heat transfer and vapour condensation or evaporation rates. As an alternative to this macroscopic approach, we developed a microscopic methodology based upon the application of the basic laws of the kinetic theory for the determination of the evaporation and condensation rates of vapour in a reactor bundle. This microscopic approach is based upon a Monte Carlo simulation of the molecular trajectories, collision rates between vapour molecules and of the molecules with the surfaces of the claddings of the pins of a reactor bundle. The pins surfaces are treated in the Monte Carlo simulation as diffusely reflecting surfaces. Scattering of sodium particles is simulated with the “hard sphere” collision model. The “step splitting” technique is applied, which consists in separating the collisions dynamic calculation from collisionsless paths of the molecules. Vapour particles are assumed to condense on the surfaces of the pins when, after diffuse reflection, their velocity would be less than one third of the most probable velocity corresponding to the wall temperature. Rewetting of dried out regions of the cladding surfaces is simulated with a dynamic film model which computes the velocity distribution of the liquid across the film thickness and then the mean liquid film velocity. Evaporation of sodium molecules from the film yields a source of molecules which re-enter into the Monte Carlo calculation of the molecular dynamic approach. The coupling of the micro- and macroscopic models has been applied to the numerical simulation of an out-of pile sodium boiling experiment run at the Nuclear Research Center of Karlsruhe, Germany.     

钠沸腾噪声探测研究

文章叙述了钠沸腾噪声探测研究进展,建立了离线和在线均可进行的高频和低频信号采集和处理系统,引进、开展、改进和编制了信号处理、故障诊断、事故报警和自回归模型分析等软件包。应用这些硬软件对水和钠沸腾噪声进行了探测和分析。结果表明,沸腾噪声信号的自功率谱密度(APSD)的幅值明显大于沸腾时的值,用自回归模型判别因子分析,可实现钠沸腾在线实时诊断和监护。     

The reactivity risk in fast reactors and the related international experimental programmes CABRI and SCARABEE

CABRI and SCARABEE are two experimental reactors, located at Cadarache, France. During the last twenty years, they were operated by IPSN, together with other French and foreign research institutes, in order to conduct several experimental programmes to study the problems raised by the reactivity risk in fast reactors. Transient over-power tests were realized in CABRI, whereas in SCARABEE bundles with up to 37 pins were brought to melting and even boiling.

The results led to code developments and general engineering expertise, and helped to give a better understanding and prediction capability of different hypothetical accident scenarios, like core disruptions and subassembly blockages.

A new programme is underway to complement some important issues, mainly linked to other scenarios, like an accidental control rod withdrawal and the risk of recriticality of a molten core.     

State of development of the computer programme BACCHUS-3D/TP for the description of transient two-phase flow conditions in LMFBR fuel pin bundles

The three-dimensional transient two-phase flow version of the computer programme BACCHUS-3D/TP (Two Phase) relies upon the basis supplied by the single phase flow version of the code. The bundle geometry typical of LMFBRs is modeled by means of the porous body approach based on the concepts of volume porosities, surface permeabilities, distributed resistances and heat sources. Two phase flow is described by means of two physical models available in two distinct versions of the code. One of these two-phase models is a three-equations Slip Model (SM) which provides as a subcase the Homogeneous Equilibrium Model (HEM) if no slip between the phases is assumed. The second is a six equation model referred to as Separated Phases Model (SPM) in which two coupled systems of governing equations are solved for the vapour and liquid phases.

A fully implicit treatment of the conservation equations for the coolant flow is followed in the SM and a half-implicit approach in the SPM. The article outlines the present state of the code development and future activities aiming at unifying both variants in a comprehensive code version describing the transition between different two-phase flow regimes from bubbly flow to dispersed annular flow. An assessment of the present capabilities of the code has been made with the theoretical interpretation of out-of-pile sodium boiling experiments in a 7- and 37-pin bundle. Numerical results are discussed and compared with experimental data.     

Verifications of the design calculations of the immersed cooler system during the start-up tests of SNR-300

The sodium cooled fast breeder reactor SNR-300 in Kalkar is equipped with two redundant immersed cooler systems (ICS). They are used only after the occurrence of serious accidents such as loss of all three main heat sinks. As part of the start-up tests of the reactor, the cooling capacity of the ICS, including natural convection behaviour, was verified and compared to the design data and calculations of the thermohydraulic computer codes NOTUNG and NANO. Three tests were performed. The first test, natural circulation in the sodium loops and air stack of the ICS, and the second test, forced convection in the sodium loops and natural convection in the air stack, were carried out to demonstrate the natural convection potential, even though this is not a design criterion. As part of the commissioning programme the third test with forced convection in the ICS was performed to prove the designed cooling capacity. During all three tests in-vessel cooling was based on natural convection.