Estimation of ex-vessel steam explosion pressure loads

An ex-vessel steam explosion may occur when, during a severe reactor accident, the reactor vessel fails and the molten core pours into the water in the reactor cavity. A steam explosion is a fuel coolant interaction process where the heat transfer from the melt to water is so intense and rapid that the timescale for heat transfer is shorter than the timescale for pressure relief. This can lead to the formation of shock waves and production of missiles that may endanger surrounding structures. A strong enough steam explosion in a nuclear power plant could jeopardize the containment integrity and so lead to a direct release of radioactive material to the environment.In this article, different scenarios of ex-vessel steam explosions in a typical pressurized water reactor cavity are analyzed with the code MC3D, which is being developed for the simulation of fuel–coolant interactions. A parametric study was performed varying the location of the melt release (central, right and left side melt pour), the cavity water subcooling, the primary system overpressure at vessel failure and the triggering time for explosion calculations. The main purpose of the study was to establish the influence of the varied parameters on the fuel–coolant interaction behaviour, to determine the most challenging cases and to estimate the expected pressure loadings on the cavity walls. For the most explosive central, right side and left side melt pour scenarios a detailed analysis of the explosion simulation results was performed. The study shows that for some ex-vessel steam explosion scenarios higher pressure loads are predicted than obtained in the OECD programme SERENA phase 1.     

Fuel coolant interaction experiments in TROI using a UO2/ZrO2 mixture

This paper discusses the results of steam explosion experiments using reactor material carried out under “Test for Real cOrium Interaction with water (TROI)” program. About 4–9 kg of corium melt jet is delivered into a sub-cooled water pool at atmospheric pressure. Spontaneous steam explosions are observed in four tests among six tests. The dynamic pressure, dynamic load, and morphology of debris clearly indicate the cases with steam explosion. The initial conditions and results of the experiments are discussed.     

Questions on safety margins of components at loads beyond design — failure modes of the primary system of PWRs

The reliability and load-carrying capabilities of structures are an important part of any risk analysis in two aspects. One is the probability of failure as an initiating event, the other is the probability of or time to failure in response to load situations beyond design conditions. The methods to predict the probability of failure of the primary pressure boundary as an initiating event for a loss-of-coolant accident have already been published by Beliczey and Schulz in 1986.For the analysis of the structural behaviour of components of the primary system at loads beyond design different questions have to be answered, e.g. - most probable sequence of loading; - most probable sequence of failure; - failure loads or times connected to a “high confidence of low probability of failure”. The failure modes of the primary circuit system and the respective times to failure were investigated for core melt-down under high pressure (HPC) and low pressure (LPC) conditions.Particular interest was directed towards the behaviour of steam generator tubes, to surgeline and main coolant piping, to the upper head and flange connections of the reactor pressure vessel, and to the lower head.     

Heat transfer from a partially uncovered reactor core

This paper documents a model which has been developed for predicting the temperature distribution along a “flow channel” of a pressurized water reactor during simulated, uncovered core conditions. In the model, heat conduction along the fuel element, convection from the surface to the coolant, radiation exchange between the clad surface and steam, and surface exchange between adjacent fuel rods are considered. Variations of the thermophysical properties of the fuel road and of the coolant with temperature are accounted for, but oxidation of Zircaloy is not modeled. Extensive sensitivity studies on the effects of heat generation in the core, steam velocity, pressure level, uncovered core height, presence of hydrogen gas in the coolant, power skew, clad emissivity, and convective heat transfer correlations have been examined. The results show that the importance of radiation in comparison with convection increases with an increase in the fuel rod temperature, pressure, and clad emissivity.     

Fundamental Experiment on Vapor Explosion with Mass of Grains

With the analysis of hypothetical accident in a nuclear power reactor, a molten fuel and coolant interaction (MFCI) leads a vapor explosion under certain circumstances. The author has performed fundamental experiment on the vapor explosion with a mass of grains of certain particle sizes which simulate the molten fuel fragments, to verify the relation between the particle size and the magnitude of pressure pulses.

The standard temperatures of water and liquid nitrogen used as cold liquid set prior to the test are 25°C and 77K respectively, and that of grains (Sic) are 600°C for the former and 25°C for the latter experiments. For both experiments, the maximum pressure pulse has the greatest value at the grain size of 0.27mmφ. This value of diameter agrees approximately with the median of the size distribution of the fragments measured in some vapor explosion experiments with a hot molten metal.

In the results of using water as cold liquid, boiling pressure traces show the oscillations of higher frequency than 100 Hz with particle sizes ranging 0.2–0.5 mm. The initial temperatures of grains and water little effect on generating such oscillations as far as it is tested in this study.     

ATWS analysis for PWR using the coupled code system DYN3D/ATHLET

The ATWS transient “Loss of main feed water supply” in a generic four-loop PWR at the nominal power of 3750 MW was analyzed using the coupled code system DYN3D/ATHLET. A variation of the MOX-fuel-assembly portion in the core has an effect on the reactivity coefficients of the fuel temperature and the moderator density. These two parameters mainly influence the behaviour of the coolant pressure, which is safety-relevant. It has been demonstrated that the pressure maximum decreases with an increasing portion of MOX. For all core loadings considered, both primary-circuit mechanical integrity and sufficient core cooling are guaranteed.     

Structural shock tests of prototype FBR “MONJU” scale models

This paper presents some results of experiments which simulate the structural dynamic response of a LMFBR primary coolant boundary to a hypothetical core disruptive accident (HCDA) based on scale models and high explosives. It was noted that high explosives are no longer a good simulant of the HCDA. However, the main purpose of the program, which included this experiment, is not to experimentally predict the dynamic response of the reactor structure at the HCDA, but to validate computer codes, which describe the pressure wave propagation and damage process in the reactor structures, using data obtained from these model experiments. The experiments were undertaken using many 1/15 scale simple models of the reactor vessels and internal structures, as well as 1/15 and 1/7.5 scale complex models of the interim design of prototype LMFBR ‘MONJU’. Simple model experiments involved a series of shock tests using pentolite to investigate the configuration effects of the vessel restraining section, the dipped-plate effect and the core barrel effect, respectively.     

A method for quantifying the reliability of instrumentation for engineered safety features to improve accident sequence recognition in reactor control rooms

This work adapts fault trees from plant-specific probabilistic risk analyses (PRAs) to quantitatively evaluate the reliability of the instrumentation for engineered safety features (ESFs). The purpose is to help improve reactor operator recognition and identification of potential accident sequences. The PRA system fault trees provide a framework for assessing the plant indicators so that the plant conditions are made more readily apparent to plant personnel through the conversion of system fault trees to alarm trees. In the alarm tree, possible states of each instrumented alarm are identified as “true” or “false”. In addition, a “warning” status is also defined and integrated into the alarm analysis routine. The impact of this additional status condition on the Boolean laws used to evaluate the alarm trees is examined. An application is described for a BWR high pressure coolant injection system (HPCI) that would be utilized during many severe reactor accidents.     

Reactor structural response to molten fuel-coolant interactions

In order to realistically determine the structural response of a liquid metal fast breeder reactor to a molten fuel-coolant interaction (MFCI), an MFCI region was incorporated into the two-dimensional, hydrodynamic containment code, REXCO-H. In this way, it is possible to account for the two-dimensional hydrodynamic response, as well as for the effect of vessels and plates, upon the expansion process in the MFCI region.The MFCI model has been extended in order to increase the usefulness of the code under a variety of conditions. The sodium equation of state has been improved using basic thermodynamic relations and recent data to incorporate temperature dependent properties. Heat transfer models available to describe the MFCI include not only a quasi-steady-state model, but also a parametric model, including the fuel heat of fusion. Nonhomogenous MFCI regions can be treated by assigning different parameters to each zone within a region, including volume fractions of fuel, sodium, steel, and void, as well as initial fuel and coolant temperatures and fraction of molten fuel.Several cases have been studied in order to delineate the effect of various parameters on the peak pressures generated in the MFCI zones. These include effect of initial fuel and coolant temperatures, void fraction, amount of molten fuel and/or vessel wall compliance. The response of a typical reactor configuration is evaluated for a given set of initial conditions.     

Three pass core design proposal for a high performance light water reactor

The paper describes a novel core concept for a nuclear reactor cooled with supercritical water, in which the coolant is heated up from 280 °C at the reactor inlet to 500 °C at the outlet in four steps: a first heat-up step is provided by heat transfer from fuel assemblies to the moderator water in gaps and moderator boxes, a second step is foreseen in a central “evaporator” and two further steps in a first and a second superheater surrounding it. The coolant flow scheme includes upward and downward flow through the core with intermediate mixing in chambers above and below the core to eliminate hot streaks. A preliminary single channel analysis, concentrating on an average flow channel and on the hottest one only, indicates that such core design can match the limits of cladding materials available today. Even though the resultant pressure drop of the coolant will be higher than usual, it is expected that the assembly boxes can be designed with acceptable deformations.     

Crack initiation and arrest in a pressurized thermal shock test for a model pressure vessel made of VVER-440 reactor pressure vessel steel

A joint pressure vessel integrity research programme involving three partners is being carried out during 1990–1995. The partners are the Central Research Institute of Structural Materials “Prometey” from Russia, IVO International Ltd (IVO) from Finland, and the Technical Research Centre of Finland (VTT). The main objective of the research programme is to increase the reliability of the VVER-440 reactor pressure vessel safety analysis. This is achieved by providing material property data for the VVER-440 pressure vessel steel, and by producing experimental understanding of the crack behaviour in pressurized thermal shock loading for the validation of different fracture assessment methods. The programme is divided into four parts: pressure vessel tests, material characterization, computational fracture analyses, and evaluation of the analysis methods. The testing programme comprises tests on two model pressure vessels with artificial axial outer surface flaws. The first model vessel had circumferential weld seam at the mid-length of the vessel. A special embrittling heat treatment is applied to the vessels before tests to simulate the fracture toughness at the end-of-life condition of a real reactor pressure vessel. The sixth test on the first model led to crack initiation followed by arrest. After the testing phase, material characterization was performed. Comparison of calculated and experimental data generally led to a good correlation, although the work is being continued to resolve the discrepancies between the measured initiation and arrest properties of the material.     

Heat transfer during molten corium-concrete interactions

Severe accidents in light water-cooled nuclear power plants involved heat transfer from molten reactor core materials or “corium” penetrating the reactor pressure vessel and coming to rest upon the containment building concrete floor covered by water. This paper discusses the difficulties of getting good information about the properties of the components and the flow structure during molten corium-concrete-water interactions. Also, potential heat transfer mechanisms are described and available prototypical tests are utilized to show that the enhancement in heat transfer by rising gas bubbles is the most likely mechanism, particularly if heat transfer by iradiation across the gas bubbles is included.     

The prediction of 2D thermal detonations and resulting damage potential

The main purpose of this paper is to introduce a new concept for the processes responsible for the escalation and propagation of steam explosions. The concept recognizes that initially only a small quantity of coolant around each coarsely premixed melt mass “sees” the fragmenting debris coming off it, hence it is called the concept of “microinteractions”. We also derive the analytical basis for it, define the nature of the requisite constitutive laws and related experimental data, and demonstrate that this concept is essential for the prediction of steam explosion energetics in large-scale premixtures in 2D geometries. We also provide the first numerical illustrations of this concept, implemented in the computer code .m. Further, we provide the first numerical results of steam explosions in large water pools, i.e. ex-vessel explosions. These results reveal two important mechanisms for explosion “venting” and thus for reducing the dynamic loads on adjacent structures. We conclude that, taken together, the “microinteractions” and “venting” make realistic predictions of steam explosion loads feasible and within reach in the near future.     

Molten fuel-coolant interaction phenomena with application to carbide fuel safety

This paper reviews the major phases occurring during an energetic molten fuel/coolant interaction (MFCI), the categories of interaction and modes of contact between molten fuel and liquid coolant, the film boiling destabilization and collapse mechanisms, and the important fragmentation mechanisms of the melt. Two major models that describe the processes involved in an MFCI event are discussed: the spontaneous nucleation model and the pressure detonation model. Finally, the MFCI experiments involving carbide fuel and liquid sodium are reviewed and the potential for an energetic interaction between molten carbide fuel and liquid sodium is discussed. Recommendations are given for future work on MFCI phenomena relative to the carbide fuel/sodium system.     

Numerical analysis of pressure load in a PWR cavity in an ex-vessel steam explosion

Ex-vessel steam explosion may happen as a result of melting core falling into the reactor cavity after failure of the reactor vessel and interaction with the coolant in the cavity pool. It can cause the formation of shock waves and production of missiles that may endanger surrounding structures. Ex-vessel steam explosion ener- getics is affected strongly by three dimensional （3D） structure geometry and initial conditions. Ex-vessel steam explosions in a typical pressurized water reactor cavity are analyzed with the code MC3D, which is developed for simulating fuel-coolant interactions. The reactor cavity with a venting tunnel is modeled based on 3D cylin- drical coordinate. A study was performed with parameters of the location of molten drop release, break size, melting temperature, cavity water subcooling, triggering time and explosion position, so as to establish parame- ters＇ influence on the fuel-coolant interaction behavior, to determine the most challenging cases and to estimate the expected pressure loadings on the cavity walls. The most dangerous case shows the pressure loading is above the capacity of a typical reactor cavity wall.     

Small break loss of coolant accidents: Bottom and side break

A loss of coolant accident (LOCA) in a nuclear reactor can be caused, e.g., by a small break in the primary cooling system. The rate of fluid escaping through such a break will define the time until the core will be uncovered. Therefore the prediction of fluid loss and pressure transient is of major importance to plan for timely action in response to such an event. Stratification of the two phases might be present upstream of the break, thus, the location of the break relative to the vapor-liquid interface and the overall upstream fluid conditions are relevant for the calculation of fluid loss. Experimental results and analyses are presented here for small breaks at the bottom or at the side of a small pressure vessel.It was found that in such a case the onset of the so-called “vapor pull through” is important but swelling at sufficient depressurization rates of the liquid due to flashing is also of significance. It was also discovered that in the bottom break the flow rate is strongly dependent on the break entrance quality of the vapor-liquid mixture. The side break can be treated similarly to the bottom break if the interface level is above the break.The analyses developed on the basis of experimental observations showed reasonable agreement of predicted and measured pressure transients. It was possible to calculate the changing interface level and mixture void fraction history in a way compatible with the behavior observed during the experiments.Even though the experiments were performed at low pressures, this work should help to get a better understanding of physical phenomena occurring in a full scale small break loss of coolant accident.     

Core temperature transient in the early phase of core uncovering at TMI-2

This paper describes a best-estimate analysis of the initial core boil-down and heat-up transient at Three Mile Island Unit (2) on 28 March 1979. This transient began shortly after all reactor coolant pumps were secured (100 min after reactor trip) and was terminated by a period of sustained high pressure injection of emergency cooling water, starting at 202 min.

The analysis is primarily directed to understanding the progression of core damage, rather than providing a detailed characterization of the core end-state condition. The latter objective can be achieved only after vessel head removal and visual examination.

The thrust of the present effort has been to: (1) develop a core coolant mixture level (dry-out level) calculation which satisfies the boundary conditions implied by various instrument responses and system operational characteristics; (2) couple the level calculation with a core heat-up modelto simulate the accumulation of thermal damage in the exposed, upper regions of the core; (3) compare calculated gross damage to the core with measurements of hydrogen and fission product releases subsequent to the accident.

Results indicate that:

1. (i) Observed containment hydrogen levels were due to Zircaloy/stainless steel corrosion that occurred during the period of core uncovering between the de-activation of the loop A reactor coolant pump (100 min after trip) and sustained operation of the high pressure injection system 100 min later. Appreciable zircaloy oxidation probably commenced at 150 min after trip, and continued at a high rate until the sustained high pressure injection at 202 min caused a major core quench.

2. (ii) There was some potential for fuel liquefaction. Calculations imply that peak fuel temperatures did not exceed the UO₂ pellet melting temperature, but 30% of the fuel was exposed to temperatures where liquid U---Zr---O alloys could have formed.

3. (iii) A substantial fission product release was obtained from fuel over-heating; however, an apparent disparity between the expected fission product release by calculation and the high range of fission product estimates obtained from plant measurements suggests that a significant release fraction may have originated from powdered or rubbilized fuel during cooldown. Additional gas releases may have developed from hot spots which persisted after core quench.

4. (iv) Steam temperatures in the upper plenum, at the outlet nozzle elevation, were generally below 900°C (1650°F) although this value was probably exceeded for a few min during the partial fuel quench caused by activation of the loop 2B reactor coolant pump, at 174 min after trip. The metal-work in the upper plenum, above the upper tieplate did not experience appreciable heating.

Thermal damage to the fuel and consequential weakening and mechanical disruption of the core was essentially complete 230 min after turbine trip.     

Fragmentation of molten debris during a molten fuel-coolant interaction

The potential for an energetic molten fuel-coolant interaction (MFCI) during a hypothetical core meltdown accident is of concern in nuclear safety analysis. An important aspect of a MFCI is the fine fragmentation and intermixing of molten core debris with the core coolant. The fragmentation characteristics of the molten debris (a mixture of UO₂ and zircaloy cladding) particles produced during a recent high-energy in-pile experiment are analyzed. The experimental results suggest that two mechanisms contributed to the fragmentation of the molten debris in this experiment, in which an MFCI occured. Phenomenological modelling of these two mechanisms and the effects of the governing parameters are presented.     

Accident management following loss-of-coolant accidents during cooldown in a Westinghouse two-loop PWR

Operation of pressurised water reactors involves shutdown periods for refuelling and maintenance. In preparation for this, the reactor system is cooled down, depressurised and partially drained. Although reactor coolant pressure is lower than during full-power operation, there remains the possibility of a loss-of-coolant accident (LOCA), with a certain but low probability. While the decay heat to be removed is lower than that from a LOCA at full power, the reduced availability of safety systems implies a risk of failing to maintain core cooling, and hence of core damage. This is recognised though probabilistic safety analyses (PSA), which identify low but non-negligible contributions to core damage frequency from accidents during cooldown and shutdown. Analyses are made for a typical two-loop Westinghouse PWR of the consequences of a range of LOCAs during hot and intermediate shutdown, 4 and 5 h after reactor shutdown respectively. The accumulators are isolated, while power to some of the pumped safety injection systems (SIs) is racked out. The study assesses the effectiveness of the nominally assumed SIs in restoring coolant inventory and preventing core damage, and the margin against core damage where their actuation is delayed. The calculations use the engineering-level MELCOR1.8.5 code, supplemented by the SCDAPSIM and SCDAP/RELAP5 codes, which provide a more detailed treatment of coolant system thermal hydraulics and core behaviour. Both treatments show that the core is readily quenched, without damage, by the nominal SI which assumes operation of only one pump. Margins against additional scenario and model uncertainties are assessed by assuming a delay of 900 s (the time needed to actuate the remaining pumps) and a variety of assumptions regarding models and the number of pumps available in conjunction with both MELCOR and versions of SCDAP. Overall, the study provides confidence in the inherent robustness of the plant design with respect to LOCA during cooldown to cold shutdown, and in the validity of a two-tier calculational method. The results have been directly used in updating the plant shutdown PSA, by changing the success criteria for core cooling during cooldown of the plant and showing a reduction in overall risk.     

Gas-cooled fast reactors—Status of CEA preliminary design studies

The gas-cooled fast reactor (GFR) is one of the six reactor concepts selected in the frame of the Generation IV initiative. The most significant GFR option is the use of a helium high temperature primary coolant. The helium option is very attractive (chemical inertness, neutron transparency, etc.) but it leads to very specific thermal-hydraulic issues.As far as the reactor core design is concerned, a ceramic fuel concept with a good thermal conductivity has been chosen. The main requirement is to obtain an average exit core temperature of 850 °C (energy conversion efficiency) with a maximum fuel temperature of about 1200 °C and with a low core pressure drop (in order to ease the decay heat removal). The main core characteristics have been determined for two reactor powers: a medium one (600 MWth) and a large one (2400 MWth). For various reasons, this latter became the CEA reference choice. A consistent set of core parameters has been determined taking into account the different constraints including the thermal-hydraulics. The reference arrangement proposed is based on plate fuel elements.A significant issue for the GFR is the decay heat removal. An innovative approach has been chosen in case of loss of coolant accidents (LOCAs). A “guard containment” enclosing the primary system is used to maintain a medium gas pressure (10 bar) in order to remove the decay heat by low power forced convection systems in the short term and natural convection systems in the long term. This guard containment is not pressurized during normal operations and can be a metallic structure.As far as the energy conversion system is concerned, an indirect-combined cycle has been chosen. The significant advantages of this choice are: a moderate core inlet temperature (400 °C instead of 480 °C for the direct cycle) and an attractive compactness of the primary system (facilitating the guard containment design).Due to the novelty of these options, a significant effort of components pre-sizing and design calculations has been achieved. Following this effort, a CATHARE model of the reactor system has been made and the calculation of the reactor steady-state confirms the consistency of the overall system pre-sizing. This model has been used for a first transient calculation. Other types of transients have to be analyzed, however, it is thought that the proposed GFR design can reach the safety requirements of Generation IV systems.