Debris interactions in reactor vessel lower plena during a severe accident II. Integral analysis

The integral physico-numerical model for the reactor vessel lower head response has been exercised for the TMI-2 accident and possible severe accident scenarios in PWR and BWR designs. The proposed inherent cooling mechanism of the reactor material creep and subsequent water ingression implemented in this predictive model provides a consistent representation of how the debris was finally cooled in the TMI-2 accident and how the reactor lower head integrity was maintained during the course of the incident. It should be recalled that in order for this strain to occur, the vessel lower head had to achieve temperatures in excess of 1000 °C. This is certainly in agreement with the temperatures determined by metallographic examinations during the TMI-2 Vessel Inspection Program. The integral model was also applied to typical PWR and BWR lower plena with and without structures under pressurized conditions spanning the first relocation of core material to the reactor vessel failure due to creep without recovery actions. The design application results are presented with particular attention being focused on water ingression into the debris bed through the gap formed between the debris and the vessel wall. As an illustration of the accident management application, the lower plenum with structures was recovered after an extensive amount of creep had damaged the vessel wall. The computed lower head temperatures were found to be significantly lower (by more than 300 K in this particular example) with recovery relative to the case without recovery. This clearly demonstrates the potential for in-vessel cooling of the reactor vessel without a need to externally submerge the lower head should such a severe accident occur as core melting and relocation.     

Results of examinations of pressure vessel samples and instrument nozzles from the TMI-2 lower head

15 prism-shaped steel samples were removed from the lower head of the damaged Three Mile Island Unit 2 (TMI-2) nuclear reactor pressure vessel to assess the effects of approximately 19 tonne of molten core debris that had relocated there during the 1979 loss-of-coolant accident. Metallographic examinations of the samples revealed that inside-surface temperatures of 800–1100°C were attained during the accident, in an elliptical ‘hot spot’ with dimensions of about 1 m × 0.8 m. Tensile, creep and Charpy V-notch specimens were cut from the samples to assess the mechanical properties of the lower head material at temperatures up to the peak accident temperature. These properties were used in a margin-to-failure analysis of the lower head. Examinations of instrument nozzles removed from the lower head region assisted in defining the relocation scenario of the molten core debris and showed that the lower head was largely protected from catastrophic failure by a solidified layer below the molten core debris that acted as a partial thermal insulator.     

Experimental results of integral effects tests with 1/10th scale zion subcompartment structures in the Surtsey test facility

Three integral effects tests (IET-1, IET-3, and IET-6) were conducted to investigate the effects of high-pressure melt ejection on direct containment heating. A 1:10 linear scale model of the Zion reactor pressure vessel (RPV), cavity, instrument tunnel, and subcompartment structures were constructed in the Surtsey test facility at Sandia National Laboratories. The RPV was modeled with a melt generator that consisted of a steel pressure barrier, a cast MgO crucible, and a thin steel inner liner. The melt generator/crucible had a hemispherical bottom head containing a graphite limitor plate with a 4 cm exit hole to simulate the ablated hole in the RPV bottom head that would be formed by tube ejection in a severe nuclear power plant accident. The reactor cavity model contained 3.48 kg water with a depth of 0.9 cm that corresponded to condensate levels in the Zion plant. 43 kg iron oxide/aluminum/ chromium thermite was used to simulate molten core debris. The molten thermite in the three tests was driven into the scaled reactor cavity by slightly superheated steam at 7.1, 6.1, and 6.3 MPa for IET-1, IET-3, and IET-6 respectively. The IET-1 atmosphere was pre-inerted with nitrogen, while the IET-3 atmosphere was nitrogen with approximately 9.0 mol% O₂. The IET-6 atmosphere was nitrogen with 9.79 mol% O₂ and 2.59 mol% pre-existing hydrogen. In IET-1, approximately 233 g mol hydrogen were produced but almost none burned because oxygen was not available. In IET-3, approximately 227 g mol hydrogen were produced and 190 g mol burned. In IET-6, approximately 319 g mol hydrogen were produced and 345 g mol burned. The peak pressure increases in the IET-1, IET-3 and IET-6 experiments were 0.098, 0.246, and 0.279 MPa respectively. In IET-3 and IET-6 hydrogen burned as it was pushed out of the subcompartments into the upper region of the Surtsey vessel. In IET-6, although a substantial amount of pre-existing hydrogen burned, it apparently did not burn on a time scale that made a significant contribution to the peak pressure increase in the vessel.     

Dryout Heat Flux for Core Debris Bed, (III)

Mixture of cylindrical steel pellets and Al₂O₃ balls, which simulated the intact fuel pellets and fragmented claddings, respectively, was inductively heated in a 50 mm I.D. pyrex glass cylinder filled with water, to investigate the coolability of TMI-2 type degraded core debris bed. The size of steel pellets was 11 mm dia. × 11 mm for BWR, 8 mm dia. × 12 mm for PWR and 5.5 mm dia. × 9 mm for FBR and Al₂O₃ balls were about 2 mm in diameter. The height of the debris bed was 25 cm or lower.

The dryout heat flux does not level off up to a bed height of 25 cm or over for the TMI-2 type bed while 8 cm or so in the bed of only steel balls. The dependence of dryout heat flux on the system pressure agrees with the Lipinski's 0-D model by adopting a proper equivalent diameter. When a simple number-weighted average is used as the equivalent diameter, the prediction gives a fairly good agreement with the experiment for FBR type bed but underestimations for the PWR and BWR type beds. It should be noted, that the small balls of less fraction, not the large pellets, substantially govern the dryout. When the coolant flow is allowed from the bottom, however, the dryout heat flux is enhanced up to the level for the complete vaporization of coolant, and small amount of mass flux or circulation head can greatly improve the coolability.     

增强IVR有效性的堆内注水策略研究

熔融物堆内滞留（IVR）是一项核电厂重要的严重事故管理措施,通过将熔融物滞留在压力容器内,以保证压力容器完整性,并防止某些可能危及安全壳完整性的堆外现象。对于高功率和熔池中金属量相对不足的反应堆,若下封头形成3层熔池结构,则其顶部薄金属层导致的聚焦效应可能对压力容器完整性带来更大的威胁。本文考虑通过破口倒灌及其他工程措施实现严重事故下熔池顶部水冷却,建立熔池传热模型,分析顶部注水的带热能力,建立事件树,分析顶部注水措施的成功概率及IVR的有效性。结果表明,通过压力容器内外同时水冷熔融物,能显著增强IVR措施的有效性。     

Ion guiding through capillaries in uncoated Al2O3 membranes

We made an experimental study on ion guiding through capillaries in uncoated Al₂O₃ membranes using a variety of ions such as O¹⁺, O³⁺, and O⁶⁺. The incident energy was varied within the range of 30-150 keV. The results were compared with others using coated PET and Al₂O₃ capillary membranes as well as with the so-called scaling law discovered by Stolterfoht and his co-workers. Good agreement of our results with the scaling law was found. However, our membranes showed extraordinarily strong guiding ability. The reason lies in that our membranes were uncoated. A slower charge drift speed along the insulating capillary wall and a much larger equilibrium charge Q_∞ seems to exist in our experiment.     

Sensitivity study of melt behavior of Fukushima Daiichi unit 1 type accident with MELCOR code and MPS method

Sensitivity calculation on melt behavior and lower head response at Fukushima Daiichi unit 1 reactor was performed with methods for estimation of leakages and consequences of releases (MELCOR) 2.1 and moving particle semi-implicit (MPS) method. Four sensitivity cases were calculated, considering safety relief valve (SRV) seizure, penetrations and debris porosity. The results indicated that the lower head failed due to creep rupture, not considering penetrations; otherwise it would have failed due to penetration tube rupture and ejection at an earlier time, resulting in part of debris dropping into the cavity of the drywell. The temperature of residual debris in pressure vessel kept low, and the vessel wall did not suffer creep failure up to 15 hours after reactor scram from which moment the water injection became available. Another aspect was that reactor pressure vessel (RPV) depressurization postponed the lower head creep failure time, and the low debris porosity brought forward the penetration rupture time. Either lower head creep failure or penetration rupture and ejection occurred in the central part of the pressure vessel. In MPS calculation, a slice of debris bed together with lower head, including an instrument guide tube, was chosen as the computational domain. Detailed temperature profiles in debris bed, penetration and vessel wall were obtained. The penetration rupture time calculated by MPS was earlier than the MELCOR result, while the vessel wall creep failure time was later.     

Scattering of light molecules from thin Al2O3 films

Molecular oxygen and hydrogen ions were scattered at grazing incidence from various thin Al₂O₃ films. The energy of incident particles was varied from 390 to 1000 eV. For scattered positive oxygen ions, negative ion fractions of up to 17% were recorded. For scattered positive hydrogen ions, the negative ion fractions reached up to 2%. These findings qualify thin films of Al₂O₃ as possible candidates for use as charge state conversion surfaces in neutral particle sensing instruments, which will work in space.     

Debris interactions in reactor vessel lower plena during a severe accident I. Predictive model

An integral predictive physico-numerical model has been developed to understand and interpret debris interactions in the reactor vessel plenum such as those which took place in the TMI-2 accident. The model represents the extent of debris jet disintegration by a jet-water entrainment model which can result in two types of debris configurations. One is particulated debris which eventually quenches in the water as a result of the entrainment process. The remainder of the debris penetrates to the bottom of the lower plenum and collects as a continuous layer. Each is treated as a separate region and has governing principles for its behavior. The potential for creating gap (contact) resistance and boiling heat removal is considered for heat transfer between the debris bed, the reactor vessel and steel structures and, most importantly, the vessel-to-crust gap water. The proposed in-vessel cooling mechanism due to material creep and water ingression into the expanding gap between the core debris and the vessel wall was found to explain the non-failure of the TMI-2 vessel in the course of the accident. The particulate debris bed is a mixture of metal and oxide, which is distributed as individual spherical particles of sizes determined at the time of entrainment. Energy is received from the continuum bed below by radiation and convection. The continuum debris bed is described by the crust behavior with the heat flux to the crust given by the natural convection correlations relating the Nusselt and Rayleigh numbers for the central region of debris. Using these governing principles, the rate laws for heat and mass transfer are formulated for each type of debris condition in the lower plenum. With the integration of the individual rates, the formation, growth and possible shrinkage of these regions are calculated. The potential reactor vessel breach is accounted for by considering the combined thermal and mechanical response of the vessel wall. The two-step failure model allows the vessel to fail at two different locations and at two different times.     

Global rupture of a nuclear reactor lower head with elevated temperature due to severe accidents

This paper is concerned with the global rupture of a reactor pressure vessel (RPV) with elevated temperature due to severe accidents in order to check if the RPV wall can retain the high-elevated pressure. The global rupture of an RPV is simulated by finite element limit analysis for the collapse load and mode to secure the safety criteria of a nuclear reactor under severe accident conditions. Finite element limit analysis is a systematic tool dealing with upper bounding and minimization technique to calculate the collapse load and mode. The finite element code (CALF, computer analysis of lower head failure) developed provides the temperature elevation in the lower head of a nuclear reactor under severe accident conditions as well as the collapse load and mode. The thermal analysis has to deal with heat transfer from the debris pool to the RPV wall and the top of the pool. The temperature distribution in such a system depends sensitively on the initial temperature of the debris pool and the thermal properties of a gap between the debris crust and the RPV wall. For accurate calculation, the thermal properties of a gap have to be determined in consideration of the gap size and conditions.     

Formation of diffusion barrier coating on superalloy 690 substrate and its stability in borosilicate melt at elevated temperature

Aluminized and thermally oxidized superalloy 690 substrates forming Al₂O₃ layer on (NiCr)Al + Cr₅Al₈ types aluminides and bare substrates were exposed in sodium borosilicate melt at 1248 K for 192 h. SEM–EDXS analysis along the cross-section of bare substrate with adhered glass revealed formation of a continuous, thick Cr₂O₃ layer at the substrate/glass interface due to its low solubility in borosilicate melt. XRD on aluminide coated and thermally oxidized specimen revealed existence of Al₂O₃ along with NiAl and Cr₅Al₈ type phases after the exposure in borosilicate melt. SEM–EDXS analysis along the cross-section of aluminide coated and thermally oxidized sample with adhered glass indicated good stability of coating in borosilicate melt without any phase formation at the coating/glass interface. However, some Al enrichment in glass phase adjacent to interface was noticed without any significant Ni or Cr enrichment.     

On the mechanism of aluminum ignition in steam explosions

An available theory [Epstein, M., Fauske, H.K., 1994. A crystallization theory of underwater aluminum ignition. Nucl. Eng. Des. 146, 147–164] of the ignition of aluminum melt drops under water, which is based on the assumption that the aluminum oxide (Al₂O₃) drop-surface skin first appears in a metastable molten state, is compared with existing experimental data on the ignition of aluminum drops behind shock waves in water [Theofanous, T.G., Chen, X., DiPiazza, P., Epstein, M., Fauske, H.K., 1994. Ignition of aluminum droplets behind shock waves in water, Phys. Fluids 6, 3513–3515]. The predicted and measured ignition temperature of about 1770 K coincides approximately with the spontaneous nucleation temperature of supercooled liquid Al₂O₃ (1760 K). This suggests that the crystallization of the oxide layer represents a strong ‘barrier’ to aluminum drop ignition under water. Apparently a similar interpretation is applicable to aluminum drop ignition in gaseous oxidizing atmospheres. We conclude from the theory that the low-temperature aluminum ignitions (in the range 1100–1600 K) that have been observed during steam explosions are a consequence of the short aluminum drop oxidation times in this environment relative to the characteristic time for Al₂O₃ crystallization. Several aspects of the aluminum drop/shock interaction experiments besides ignition are discussed in the paper. In particular, the experiments provide strong evidence that during the course of a vapor explosion metal fragmentation occurs via a thermal mechanism at low pressure and precedes the development of a high-pressure shock.     

Design and application of GDC on EAST Tokamak

On EAST Tokamak, DC glow discharge (GDC) is developed to clean the first wall of plasma. It can effectively control the recycling of H, C, O impurities and improve the wall conditions. There are four GDCs which distribute equally on the EAST Tokamak vacuum vessel wall. Each GDC is equipped with an anode, a stainless steel cover and four support legs. The anode is insulated from cover with Al₂O₃ ceramics. After a round of experiment, the value of insulation resistance of GDC decreases remarkably due to metallization. To protect the insulation parts and heighten the reliability, ceramic protection covers are used on the GDCs. The other measures which can heighten insulation grades are also taken. After upgrade, the insulation resistance of each GDC between anode and ground is raised highly. When the pressure reaches 4 Pa, H₂-GDC and He-GDC is strarted. Boronation and siliconization are also applied to the device wall conditioning. After GDC cleaning, the impurities and partial pressure of remainder gases in vacuum vessel (VV) is decreased greatly and vacuum degree of VV can reach high easily.     

Fabrication of Al<Subscript>2</Subscript>O<Subscript>3</Subscript>/FeAl Coating as Tritium Permeation Barrier on Tritium Operating Component on Quasi-CFETR Scale

Limited CFETR-scale experience of engineering preparation techniques of tritium permeation barrier (TPB) exists up to date. Aimed at processing some real components that are usually tubular components sealed in one end, in the tritium cycling systems of China Fusion Engineering Test Reactor (CFETR), an Al₂O₃/FeAl coatings as TPB was prepared on tubular components of 321 type stainless steel components with a length of 400 mm and an external diameter of 150 mm, by Al-electroplating followed by heat treating and selective oxidation. The ability to construct TPB coated components on quasi-CFETR scale was demonstrated, with fabricating a TPB of Al₂O₃/FeAl coating with a double-layered structure, consisted of an outer γ-Al₂O₃ layer with a thickness of 0.3 µm and an inner (Fe,Cr,Ni)Al/(Fe,Cr,Ni)₃Al layer of 40 µm in thickness. The tritium permeation reduction factors of the Al₂O₃/FeAl TPB on component were 229 and 96 at 500 and 600 °C respectively. Finally, signatures and gaps of TPB mass process on CFETR-scale were discussed.     

Al2O3-containing silver phosphate glasses as hosting matrices for radioactive iodine

Al₂O₃-containing silver phosphate glasses were synthesized to investigate the feasibility of phosphate glasses for the immobilization of radioactive iodine (¹²⁹I) present in spent nuclear fuel. Characterizations were performed by X-ray diffraction, Fourier transformed infrared spectroscopy, and scanning electron microscopy coupled with energy dispersive spectroscopy to examine structures, bonding properties, surface morphology, and elemental distribution of the synthesized glasses. The principal results showed that iodine became more strongly immobilized in the phosphate glasses with the addition of Al₂O₃, which was confirmed by the decrease of iodine leaching rates with approximately one order of magnitude. The present study would be helpful to decide whether Al₂O₃-containing silver phosphate glasses could be used as a candidate matrix to incorporate ¹²⁹I.     

Review of debris bed cooling in the TMI-2 accident

Still today after decades of severe accident research it is not well understood why the molten corium did not attack or even penetrate the lower head vessel wall in the TMI 2 accident. The findings can only be explained by additional assumptions which have been proposed by various authors. This paper is also looking for an explanation by examining the role of the debris cooling in the TMI lower head. The present knowledge of debris cooling is based on small-scale experiments with simulant debris. It is argued that the experiments have been stopped to early and therefore do not reveal the potential of debris bed coolability in case of a corium debris bed. It is also argued that in TMI the debris bed was in a state of turbulence and fluidization such that the coolabilty was much higher than in the small-scale experiments. Basically this paper hypothesizes that the unknown phenomenon in the TMI 2 accident is the strong interaction of the debris particles of a wide range of grain sizes and the strong turbulent motion of the cooling fluid and vapor mixture in conjunction with a “virtual gap” at the vessel wall.     

Study of the vaporization of Ba from UO2 nuclear fuel particles by high-temperature mass spectrometry

In connection with improving the retention of solid fission products in gas-cooled high-temperature reactor fuels, the vaporization of Ba from UO₂ model nuclear fuel particles with and without a pyrocarbon coating was studied by high-temperature mass spectrometry using a Knudsen cell. The UO₂ kernels of the particles were doped with BaO. In addition, some of them contained Al₂O₃. Whereas BaO mainly evaporated from the surface of the kernels as BaO, only Ba could be observed over the coated particles. Moreover, the BaO vapor pressure over kernels with and without the addition of Al₂O₃ was determined. From this it was determined that the BaO vapor pressure could be diminished by approximately two orders of magnitude by the admixture of Al₂O₃. Finally it was proved that the diminution of the BaO vapor pressure was caused by the formation of the compound BaAl₂O₄.     

Experimental measurement of frictional law under CCFL in porous debris

Counter-current flow limitation (CCFL) is the dominant phenomena for dryout in porous debris, which would be formed during a severe accident of a nuclear power plant. Since flow at CCFL in porous debris is far away from normal two-phase flow in a pipe, it is not clear whether the interfacial friction laws in a pipe can be applied to the CCFL in porous debris. In the present experimental research, the void fraction in porous debris is measured simultaneously with the differential pressure and flow rates for gas and liquid. The combination of these simultaneously measured data makes it possible to estimate the shear stresses and friction factors in a porous debris at CCFL. It results that the wall friction factor estimated is larger than the Ergun equation proposed for single-phase flow in a porous bed. Furthermore, the interfacial friction factor estimated is well correlated by the theoretical correlation derived here from a force balance for a hypothetical flow channel in porous debris at CCFL.     

Use of Al2O3 layer as a dielectric in MOS based radiation sensors fabricated on a Si substrate

The use of Al₂O₃ dielectric in MOS based radiation sensors has been investigated. Their response has been compared with conventional MOS capacitors with a SiO₂ dielectric. The study includes gamma radiation effects with dose up to 4 Gy. The effect of radiation has been determined from the valance band shift in C-V curves. The amount of charge induced by the radiation has been calculated and compared with the response of MOS capacitors with SiO₂ with the same and different thicknesses. Fading properties have been studied and compared. Results show that MOS capacitors with Al₂O₃ dielectric exhibit sensitivity greater than that obtained from MOS capacitors with SiO₂. This higher sensitivity is attributed to higher trapping efficiency in the Al₂O₃ layer.     

Interface strain study of thin Lu2O3/Si using HRBS

The interface of thin Lu₂O₃ on silicon has been studied using high-resolution RBS (HRBS) for samples annealed at different temperatures. Thin rare earth metal oxides are of interest as candidates for next generation transistor gate dielectrics, due to their high-k values allowing for equivalent oxide thickness (EOT) of less than 1 nm. Among them, Lu₂O₃ has been found to have the highest lattice energy and largest band gap, making it a good candidate for an alternative high-k gate dielectric. HRBS depth profiling results have shown the existence of a thin (∼2 nm) transitional silicate layer beneath the Lu₂O₃ films. The thicknesses of the Lu₂O₃ films were found to be ∼8 nm and the films were determined to be non-crystalline. Angular scans were performed across the [1 1 0] and [1 1 1] axis along planar channels, and clear shifts in the channeling minimum indicate the presence of Si lattice strain at the silicate/Si interface.