Sodium boiling experiments in a 19-pin bundle under loss-of-flow conditions

An experimental study was conducted on transient sodium boiling in a 19-pin electrically heated LMFBR fuel subassembly mockup under loss-of-flow conditions. In each run the inlet flow was reduced or stopped at constant heater power. There was no strong effect of temperature ramp rate on incipient-boiling (IB) wall superheat. The observed coolant voiding was initially limited to the center subchannel because of steep temperature gradient in the bundle. The bulk pressure rise registered upon initial vaporization was markedly lower than the vapor pressure corresponding to the IB wall superheat. The pressure pulse generated at vapor bubble collapse correlated reasonably well with the re-entrant liquid velocity, but the measured value was very much smaller than the calculation by sodium hammer analysis.     

Steady state and transient sodium boiling experiments in a 37-pin bundle

As part of the fast breeder reactor safety analysis steady state and transient sodium boiling tests were performed out-of-pile in an electrically heated 37-pin bundle. The steady state boiling experiments served for investigations of the two-phase flow physics and to support the analysis of the transient experiments. The experimental work concentrated on the transient sodium boiling tests which simulated the unprotected loss of flow accident (ULOF) from the start of the flow run down via boiling inception to the onset of dryout. Special emphasis was laid upon the analysis of the transition from the spatial to the mainly one-dimensional growth of the boiling region during the flow transient. The experimental results from both types of tests serve as data basis for computer code validations. A reference test (L22) of the transient experiments was satisfactorily recalculated with a one-dimensional and with a three-dimensional computer programme.     

Flow rundown experiments in a seven pin bundle

A series of experiments has been carried out using an electrically heated seven pin bundle to simulate the conditions under which boiling of the sodium coolant could occur in the event of the loss of power to the circulator pumps of a fast reactor, coincident with a failure of the reactor to trip. Although it was not possible to represent the conditions of the reactor exactly, nor to continue the tests far into dryout, the results nevertheless give valuable qualitative information on the course of boiling development as well as useful quantitative information against which the predictions of computer codes can be checked. In particular, data have been obtained relating to the incidence of superheat, the location and time to dryout of the residual liquid films, the void fraction within the boiling region, and the types of flow regimes which may be expected within different parts of the boiling region at various stages of the transient.     

Theoretical interpretation of SCARABEE single pin in-pile boiling experiments

In the framework of LMFBR safety analysis a theoretical interpretation of some of the most representative of the single pin experiments of the in-pile SCARABEE project has been performed from both viewpoints of themohydraulic and fuel behavior using the computer codes CAPRI-2 and SATURN-1. The analysis aimed at investigating the pin behavior from the preirradiation history, through the observed sequence of events following a coolant mass flow reduction from boiling inception up to pin breakdown. A comparison of theoretical results with experimentally recorded data has allowed a deeper insight into the peculiar features of the experiments and enabled a valuable code verification.     

Sodium boiling experiments at low power under natural convection conditions

Sodium boiling experiments have been performed under natural convection conditions in an electrically heated 37-pin bundle. The pin power was in the range of the decay power of an LMFBR. In the tests which aimed at approaching the limits of cooling of a pin bundle three aspects have been investigated: (1) increase of the flow resistance in the cold leg of the natural convection loop, (2) pin powers exceeding the decay power, and (3) the maximum power which can be removed continuously from a grid spaced bundle with totally blocked inlet. The results related to items one and two reveal the large safety margin which exists between nominal operation conditions in the case of emergency cooling and cooling failure. Theoretical work concentrates on flow stability and on the predictions of a model based on counter-current vapour-liquid flow.     

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.     

A parametric study of multichannel molten cladding motion under unprotected loss-of-flow accident conditions in LMFBR's

One of the key assumptions of the present multichannel clad motion model was that the total pressure drop over the voided channel could be supplied as a boundary condition. The incoherency effect on cladding motion can be significant for a full-scale subassembly, and therefore parametric studies of the total pressure drop and oscillatory pressure effect due to sodium chugging were examined using the multichannel model.There is an axial blanket region in demonstration plant or commercial-power-plant designs instead of a reflector in FFTF design above the top of fuel. It was shown that due to the difference in the thermal conductivities between the blanket material and reflector, significant changes in the timings of various events of the cladding relocation might occur. It is also noted that depending on the effect of the sodium voiding on the reactivity, the fuel may become molten when the molten cladding is still around. The possibility of the occurrence of this situation is studied by increasing the power in the present model.     

Void-fraction distribution under high-pressure boil-off conditions in rod bundle geometry

Void fractions in a simulated pressurized water reactor (PWR) core rod bundle geometry were measured under boil-off conditions covering pressures from 3 to 12 MPa and mass fluxes from 5 to 100 kg m⁻² s⁻¹, with a particular interest in void fractions at higher pressures and relatively high mass fluxes. Test results showed that the Chexal-Lellouche model predicts best the present (volume-averaged) void-fraction data among correlations and models examined in this study. The volume-averaged void fractions obtained from differential pressure measurements are systematically smaller than the chordally averaged void fractions obtained from gamma densitometer measurements. Local void fractions were measured in the same bundle for non-heated steam-water two-phase flow of 3 MPa by using an optical void probe. It was found that the difference between the volume-averaged and chordally averaged void fractions mentioned above can be explained qualitatively by a local void-fraction distribution in the bundle measured in the latter tests.     

Separate effects study on sodium boiling and dryout under forced-and natural-convection conditions performed with a one-dimensional LOOP code

A comprehensive separate effects study has been performed with the one-dimensional code LOOP-1 on powers and times for sodium boiling initiation and dryout in a closed loop system. Two different kinds of transients were considered: loss-of-flow and loss-of-heat-sink. Loss-of-flow transients were studied under both forced- and natural-convection conditions. Loss-of-heat-sink transients were studied under natural-convection conditions. The results for loss-of-flow transients indicate that the boiling initiation time was reduced by a small amount, and the dryout time was reduced very significantly by increasing either the input power or the inlet temperature, or by decreasing the test section pressure for both forced- and natural-convection conditions. Under forced-convection conditions, a stabilizing effect occured by either increasing the test section valve setting or by decreasing the bypass ratio with a pump head adjusted to provide the same steady state and initial transient flows; thus, longer boiling times could be maintained before dryout occurred. For natural-convection loss-of-flow conditions, increasing the test section valve setting or decreasing the bypass ratio reduced the test section inlet flow, which resulted in boiling inception and dryout occurring more rapidly. A larger flow before the loss of flow transient starts yielded longer boiling initiation and dryout times. Under loss-of-heat-sink conditions, the higher the inlet temperature, the lower the boiling and the dryout powers. The margin between boiling and dryout powers increases with increasing inlet temperature. Results have been verified with experimental data. These results indicate that a margin between several seconds and several hours (depending on the type of transient) is available before core damage may occur in an actual reactor.     

Analysis of low-heat-flux sodium boiling test in a 37-pin bundle by the two-fluid model computer code SABENA

This paper summarizes the development of a new detailed multi-dimensional multi-field computer code SABENA and its application to an out-of-pile low-heat-flux sodium boiling test in a 37-pin bundle. The semi-implicit numerical method employed in the two-fluid six-equation two-phase flow model has proved in solving a wide spectrum of sodium boiling transients in a rod bundle under low pressure conditions. The code is capable of predicting the spatial incoherency of the boiling, dryout on fuel cladding surfaces and fuel pin heat transfer. Essential to the successful application of such a mechanistic model computer code are validational efforts aimed at the LMFBR accident phenomenology analyses. Through the simulation of the natural circulation boiling conditions, this study provides a consistent analytical interpretation of the experimental data. The important influences of such parameters as the inlet flow restriction and bundle geometry have been examined through interpretations of two-phase flow analysis including considerations of the flow instability induced dryout mechanism.     

Numerical simulation of bubble behaviors in subcooled flow boiling under swing motion

A numerical investigation of bubble behaviors in subcooled flow boiling of water under the effect of additional inertial forces has been performed considering energy and mass transfer during phase change based on the VOF (volume-of-fluid) method. The pressure ranges from 0.1 to 1.0 MPa, and heat flux from 200 to 500 kW/m². The mass flow rate and inlet subcooling are specified at 320 kg/m² s and 10 K, respectively. The liquid-vapor interface is captured using the piecewise linearity interpolation calculation (PLIC) geometry restructuring method. The simulations are carried out on upward water flow in a vertical, rectangular duct with single side heating surface. The pressure, velocity vector and temperature distribution around two isolated bubbles are studied firstly. The behaviors of bubble coalescence, sliding, detachment from the heated wall, and the bubble shape variation during lifetime are further examined. The bubble behaviors in the different pressure and heat flux are investigated. The simulated results of bubble growth rate and wall temperature are agreed well with the correlations in the literatures. The additional inertial forces caused by swing are negligible, but the fluctuation of mass flow rate caused by swing motion influences the forces acting on bubble significantly. Compared with the motionless condition, the pressure drop is increased and the fluctuation becomes acute as heat flux increases under the swing condition.     

A method for determining rewetting velocity under generalized boiling conditions

A method is presented for the solution of the quasi-steady state two-dimensional rewetting conduction problem. Using the method of separation of variables, a solution is presented for the case of an arbitrary heat transfer coefficient profile. Numerical results are also presented for the particular case in which a sputting region exists immediately behind the wet front. These results show that value of the heat transfer coefficient in the sputtering region determines the rewetting temperature if the sputtering region length exceeds one quarter of the class thickness. The numerical results in general show that the rewetting phenomenon is very localized with respect to the wet front. Results are compared favorably with experimental data in the low to moderate flow rate range.     

Sodium boiling experiments in a 7-pin bundle: Investigations into some safety aspects of the KNK II driver zone

A series of sodium boiling experiments, in which the thermohydraulic characteristics of KNK II driver subassemblies were simulated, has been carried out for the purpose of studying the effects of two types of cooling disturbances: rapid flow interruption and those flow reductions which develop rather gradually. Information about the spatial and temporal development of boiling, voidage and dryout was obtained. Furthermore the feasibility of individual subassembly temperature monitoring has been investigated as well as that of two integral boiling detection methods based on acoustic noise and reactivity measurements.     

Dynamics of liquid sodium pool spreading under sodium fire conditions

The spreading of burning liquid sodium has been investigated using a depth-averaged shallow water equation for isothermal and non-isothermal (burning) conditions. In the latter case, the spreading is one-way coupled with the flame through a separate energy equation for the pool, with appropriate source terms for radiative and conductive heat transfer from the flame, and a sink term (for the continuity equation) to account for loss due to burning. Pool fires on soil and concrete surface have been considered with appropriate friction and heat transfer terms in the momentum and energy equations, respectively. Using this model, numerical simulations have been carried out for a wide range of leak rates, and for a range of burning rates of liquid sodium. Results obtained from the non-isothermal model show that the non-isothermal effects of liquid sodium spreading can safely be neglected for the case or spreading of burning liquid sodium on a typical ground surface such as concrete or soil. Based on these conclusions, dimensionless correlations are proposed for the prediction of spreading parameters such as, equilibrium pool radius, pool formation time, and for mass inventory under pool fire conditions for liquid sodium. These parameters which are obtained from the spreading code can be specified, as input parameters for the existing sodium fire safety codes.     

Sodium boiling experiments in the THORS facility

Experiments have been performed with 19- and 61-pin test assemblies in the Thermal-Hydraulic Out-of-Reactor Safety (THORS) Facility at the Oak Ridge National Laboratory (ORNL) since 1971. The THORS Facility is a high-temperature sodium system operated for the US Liquid-Metal Fast Breeder Reactor (LMFBR) Safety Program. The facility is used primarily for testing simulated LMFBR fuel subassemblies (pin (bundles). High-performance, electrically heated fuel pin simulators (FPSs) duplicate the heat generating capabilities and the dimensional characteristics of the nuclear fuel pins. A number of test bundles have been built and operated to obtain base thermal-hydraulic data, inlet and heated zone blockage data, and transient boiling data. Five of these bundles have been operated under two-phase conditions. Sodium boiling for periods up to twelve minutes were sustained in one bundle. (The lengths of the periods were limited only by automatic data recording capability). Clad dryout occurred in several tests. Tests were run at widely varying conditions of flow and power density. Testing with nonuniform power distribution across the bundle was also a part of the program.A 19-pin bundle with 12 peripheral guard heaters and a 6-subchannel blockage around the center pin in the heated zone was tested. The test program for this bundle was designed to determine if local boiling in the wake of the blockage propagates radially or axially during quasi-steady-state conditions. Post-test inspection revealed that significant helical distortion of the FPSs occurred in the vicinity of the blockage plate. This distortion probably influenced the boiling behavior. In the more severe tests, boiling initiated at the outlet of the heated zone and propagated radially into the unblocked subchannels after it had progressed upstream to the blockage. The subchannel analysis codes, SABRE and COBRA, accurately predict the extent of the boiling region.Experimental and analytical studies of sodium boiling behavior in unblocked 19- and 61-pin bundles indicate that cooling can be maintained for a significant period of time beyond boiling inception in a flow-power transient. Quasi-steady-state boiling occurred under natural-convection conditions.Investigations of the temperature data indicate that the thermal-hydraulic behavior during boiling transients is determined by two-dimensional effects, and that one-dimensional models cannot accurately predict the important phenomena associated with sodium boiling in test bundles. The subchannel code SABRE-2P (with a simple two-phase multiplier boiling model) and the two-region equilibrium mixture code THORAX (developed at ORNL) accurately predict the two-dimensional behavior between boiling inception and dryout.Extrapolation of the data from the smaller bundle tests to full-size fuel assemblies shows that the time between boiling inception and dryout would be lower for a 217-pin bundle than for a 61-pin bundle for a comparable transient. However, the time delay would still be significant, especially in a heterogeneous reactor core.     

Development of the residual sodium quantification method for a fuel pin bundle of SFRs before and after dry cleaning

ABSTRACT

In a fuel handling system of sodium-cooled fast reactors (SFRs), it is necessary to remove the sodium remaining on spent fuel assemblies (FAs) before storing them in a spent fuel water pool (SFP). A next-generation SFR in Japan has adopted an advanced dry-cleaning system that consists of argon gas blowing to remove the metallic residual sodium on the FA, which increases economic competitiveness and reduces waste products thanks to a waterless process. In this R&D work, the performance of the dry cleaning process has been investigated.

This paper describes experimental and analytical studies focusing on the amount of residual sodium remaining on a fuel pin bundle before and after the argon gas blowing process. The experiments were conducted using a sodium test loop and a short (approximately 1 m) specimen consisting of a 7-pin bundle. The effects of the blowing gas velocity and the blowing time were quantitatively analyzed in the experiments. The blowing gas velocity was varied from 3.9 to 31.3 m/s, and 113 data-points of the residual sodium were collected during the experiment. On the basis of these experimental results, the residual sodium quantification method for the fuel pin bundle was constructed.     

CFD investigations of helical wire-wrap fuel pin bundle and its comparison with straight wire bundle

Computational Fluid Dynamics (CFD) investigations of a fast reactor fuel pin bundle wrapped with helical and straight spacer wires have been carried out and the advantages of using helical spacer wire have been assessed. The flow and temperature distributions in the fuel pin bundle are obtained by solving the statistically averaged 3-Dimensional conservation equations of mass, momentum and energy along with high Reynolds number k-ε turbulence model using a customized CFD code CFDEXPERT. It is seen that due to the helical wire-wrap spacer, the coolant sodium not only flows in axial direction in the fuel pin bundle but also in a transverse direction. This transverse flow enhances mixing of coolant among the sub channels and due to this, the friction factor and heat transfer coefficient of the coolant increase. Estimation of friction factor, Nusselt number, sodium temperature uniformity at the outlet of the bundle and clad hot spot factor which are measures of the extent of coolant mixing and non-homogeneity in heat transfer coefficient around fuel pin are paid critical attention. It is seen that the friction factor and Nusselt number are higher (by 25% and 15% respectively) for the helical wire wrap pin bundle compared to straight wire bundle. It is seen that for 217 fuel pin bundle the maximum clad temperature is 750 K for straight wire case and the same for helical wire is 720 K due to the presence of transverse flow. The maximum temperature occurs at the location of the gap between pin and wire. The ΔT between the bulk sodium in the central sub-channel and peripheral sub-channel is 30 K for straight wire and the same for helical wire is 18 K due to the presence of secondary transverse flow which makes the outlet temperature more uniform. The hotspot factor and the hot channel factors predicted by CFD simulation are 10% lower than that used in conventional safety analysis indicating the conservatism in the safety analysis.