Leakage potential of LWR containment penetrations under severe accident conditions

This paper is an overview of a Sandia National Laboratories, Albuquerque (SNLA) study of the performance of mechanical penetrations in light-water reactor (LWR) containment buildings that are subjected to severe accident environments. The study is concerned with modes of failure as well as the magnitude of leakage. The following tests have been completed, are under way, or are planned: (a) seals and gaskets have been tested to register the effects of radiation aging, thermal aging, seal geometry, and squeeze on seal and gasket materials in severe accident environments; (b) the performance of a full-scale airlock will be evaluated at severe accident temperature and pressure levels; (c) personnel airlock and equipment hatch tests were made on a model of a steel containment building; and (d) tests of mechanical penetrations are planned as part of a test on a model of a reinforced concrete building. This program is part of an overall US Nuclear Regulatory Commission (USNRC) effort to evaluate the integrity of LWR containment buildings.     

Experiments to evaluate behavior of containment piping bellows under severe accident conditions

Bellows are an integral part of the containment pressure boundary in nuclear power plants. They are used at piping penetrations to allow relative movement between piping and the containment wall. In a severe accident they may be subjected to high pressure and temperature and a combination of axial and lateral deflections. A test program to determine the leak-tight capacity of containment penetration bellows is being conducted at Sandia National Laboratories, Albuquerque, NM. Several different bellows geometries representative of actual containment bellows are being subjected to extreme deflections along with pressure and temperature loads. The bellows geometries and loading conditions are described along with the testing apparatus and procedures. A total of 13 tests have been conducted. The tests showed that bellows are capable of withstanding relatively large deformations up to or near the point of full compression before developing leakage. The test data are presented and discussed.     

Failure mechanisms of LWR steel containment buildings subject to severe accident loadings

The potential failure mechanisms in LWR steel containment buildings subject to quasi-static pressurization and elevated temperature are identified. For each failure mechanism, the relevant structural response measures are discussed. For mechanisms involving leakage, the importance of seal performance is also discussed. Criteria that can be used to evaluate threshold environments are presented for several failure mechanisms. Results of tests on scale models and seal tests that support the criteria are referenced.     

SMART behavior under over-pressurizing accident conditions

SMART (System-integrated Modular Advanced ReacTor) is an integral reactor of 330 MW capacity with passive safety features under development in Korea. The design is developed by combining the firmly-established commercial reactor technologies with new and advanced technologies such as industry proven KOFA (Korea Optimized Fuel Assembly) based nuclear fuels, self-pressurizing pressurizer, helically coiled once-through steam generators, and new control concepts. The design of SMART focuses on enhancing the safety and reliability of the reactor by employing inherent safety features such as low core power density, elimination of large break loss of coolant accident, etc. In addition, in order to prevent the progression of emergency situations into accidents, the SMART is provided with a number of engineered safety features such as Passive Residual Heat Removal System, Passive Emergency Core Cooling System, Safeguard Vessel, and Passive Containment Over-Pressure Protection System. This paper presents an overview of the SMART design, characteristics of it’s safety systems, and results of over-pressure accident analyses. The results of the accident analyses show that the SMART provides the inherent over-pressure protection capability for design basis accidents without actuation of any protection devices such as safety valves, rupture disks, etc.     

Effects of accident management strategy on the severe accident environmental conditions

This paper presents a methodology utilizing an accident management strategy in order to determine accident environmental conditions to be used as inputs to equipment survivability assessments. In the case that there is a well-established accident management strategy for a specific nuclear power plant (NPP), an application of this tool can provide a technical rationale on equipment survivability assessment so that plant-specific and time-dependent accident environmental conditions could be practically and realistically defined in accordance with the equipment and instrumentation required for the accident management strategy or appropriate actions. For this work, three different tools are introduced; probabilistic safety assessment (PSA) outcomes, major accident management strategy actions, and accident environmental stages (AESs). In order to quantitatively investigate an applicability of accident management strategy on equipment survivability, the accident simulation for most likely scenario in Korean standard nuclear power plants (KSNPs) is performed with the MAAP4 code. The accident management guideline (AMG) actions such as the reactor coolant system (RCS) depressurization, water injection into the RCS, the containment pressure and temperature control, and hydrogen concentration control in containment are applied. The effects of these AMG actions on the accident environmental conditions are investigated by comparison to actions from previous normal accident simulation, especially focusing on equipment survivability assessment. As a result, the AMG-involved case shows the higher accident consequences along the accident environmental stages. This implies that plant-specific AMG actions need to be considered in order to determine accident environmental conditions in equipment survivability assessments.     

Iodine-steel reactions under severe accident conditions in light-water reactors

Owing to large surface areas, the reaction of volatile molecular iodine (I₂) with steel surfaces in the containment may play an important role in predicting the source term to the environment. Both wall retention of iodine and conversion of volatile into non-volatile iodine compounds at steel surfaces have to be considered. Two types of laboratory experiment were carried out at Siemens (KWU) in order to investigate the reaction of I₂ at steel surfaces representative for German power plants.

• 1. 

  (1) For steel coupons submerged in an I2 solution at T = 50, 90 or 140 °C the reaction rate of the I2−I− conversion was determined. No iodine loading was observed on the steel in the aqueous phase tests. I2 reacts with the steel components (Fe, Cr or Ni) to form metal iodides on the surface which are all immediately dissolved in water under dissociation into the metal and the iodide ions. From these experiments, the I2−I− conversion rate constants over the temperature range 50–140 °C as well as the activation energy were determined. The measured data are suitable to be included in severe accident iodine codes such as IMPAIR.

• 2. 

  (2) Steel tubes were exposed to a steam-I2 flow under dry air at T = 120 °C and steam-condensing conditions at T = 120 and 160 °C. In dry air, I2, was retained on the steel surface and a deposition rate constant was measured. Under steam-condensing conditions there is an effective conversion of volatile I2 to non-volatile I− which is subsequently washed off from the steel surface. The I2−T− conversion rate constants suitable for modelling this process were determined. No temperature dependence was found in the range 120–160 °C.

     

External cooling of a reactor vessel under severe accident conditions

The TMI-2 accident demonstrated that a significant quantity of molten core debris could drain into the lower plenum during a severe accident. For such conditions, the Individual Plant Examinations (IPEs) and severe accident management evaluations, consider the possibility that water could not be injected to the RCS. However, depending on the plant specific configuration and the accident sequence, water may be accumulated within the containment sufficient to submerge the lower head and part of the reactor vessel cylinder. This could provide external cooling of the RPV to prevent failure of the lower head and discharge of core debris into the containment.This paper evaluates the heat removal capabilities for external cooling of an insulated RPV in terms of (a) the water inflow through the insulation, (b) the two-phase heat removal in the gap between the insulation and the vessel and (c) the flow of steam through the insulation. These results show no significant limitation to heat removal from the bottom of the reactor vessel other than thermal conduction through the reactor vessel wall. Hence, external cooling is a possible means of preventing core debris from failing the reactor, which if successful, would eliminate the considerations of ex-vessel steam explosions, debris coolability, etc. and their uncertainties. Therefore, external cooling should be a major consideration in accident management evaluations and decision-making for current plants, as well as a possible design consideration for future plants.     

Rock-like oxide fuel behavior under reactivity initiated accident conditions

Pulse irradiation tests of two types of rock-like oxide (ROX) fuel, i.e. yttria stabilized zirconia (YSZ) and YSZ/Spinel composite, were conducted in the Nuclear Safety Research Reactor (NSRR) to investigate the fuel behavior under reactivity-initiated accident conditions. The ROX fuels failed with cladding burst at fuel volumetric enthalpies above 10 GJ m⁻³, which was comparable to that of UO₂ fuel. The failure of the ROX fuels, however, occurred with considerable fuel melting and was quite different to that of UO₂ fuel, which was caused by cladding melting and embrittlement due to heavy oxidation. Lower fuel melting temperature of the ROX fuels compared to that of UO₂ contributed to the different fuel failure modes. Certain amount of molten ROX fuel dispersed out at the failure. However, the mechanical energy generation due to the molten fuel/water interaction was negligible for the ROX fuels at peak fuel enthalpies below 12 GJ m⁻³.     

Optimization of material and production to develop fluoroelastomer inflatable seals for sodium cooled fast breeder reactor

The feasibility of producing thin-walled fluoroelastomer profiles under continuous, atmospheric-pressure vulcanization conditions in air has been demonstrated by successful manufacture of ∼2 m diameter test inflatable seals for the 500 MWe, Prototype Fast Breeder Reactor (PFBR) using a 50/50 blend formulation of Viton^® GBL-200S/600S based on advanced polymer architecture (APA). A commercial cold feed screw extruder with 90 mm diameter screw was used along with continuous cure by microwave (2.45 GHz) and hot air heating (190 °C) at a line speed of 1 m/min to produce the seals. The blend formulation promises significant improvement in the performance and safety of the seals. This article depicts the relevant characteristics of the original inflatable seal compound that was used as reference to achieve the objectives through synchronized optimization of material and production technologies. The production trials are outlined and the blend formulation used with minor factory modifications to produce the test seals is reported. Progressive refinements of the original, Viton^® A-401C based compound to the blend formulation is presented along with an assessment of potential performance gains. Possible uses of the reported formulation and production technique for other large diameter, high temperature seals of PFBR are indicated along with the envisaged activities en-route the production of perfected reactor inflatable seals.     

Hydrogen distribution tests under severe accident conditions at the large-scale HDR-facility

The E11 tests series covered a combination of important issues dominating the physical phenomena controlling the hydrogen distribution mechanisms, namely: large-scale, multi-compartment, geometry with large-sized dome volume, high gas release rates, multiple steam and gas injection phases at different axial positions and examinations of the efficiencies of mitigative system features including the impact of external sprays at the top of the dome.The test series consisted of a total of eight different experiments covering all aspects of the H₂-distribution and potential mitigation features.A total of 700 sensors were applied during these experiments.The paper outlines experimental and computational results of tests E11.2 and E11.4 which were chosen for two computational PHDR-Benchmark Exercises in the context of blind posttest predictions with broad international participation applying the majority of known computer codes. In addition test E11.2 was selected as an open post-test, OECD International Standard Problem No. 29 (H. Karwat, Distribution of hydrogen within the HDR-containment under severe accident conditions — Task specifications (July 1990)) which is presently in progress.     

Prediction of structural integrity of steam generator tubes under severe accident conditions

Available models for predicting failure of flawed and unflawed steam generator tubes under normal operating and design-basis accident conditions are reviewed. These rate-independent flow stress models are inadequate for predicting failure of steam generator tubes under severe accident conditions because the temperature of the tubes during such accidents can reach as high as 800°C where creep effects become important. Therefore, a creep rupture model for predicting failure was developed and validated by tests on unflawed and flawed specimens containing axial and circumferential flaws and loaded by constant as well as ramped temperature and pressure loadings. Finally, tests were conducted using pressure and temperature histories that are calculated to occur during postulated severe accidents. In all cases, the creep rupture model predicted the failure temperature and time more accurately than the flow stress models.     

严重事故条件下压力容器完整性评价的研究进展

堆芯熔融物堆内滞留(In-Vessel Retention,IVR)是以AP1000为代表的第三代轻水反应堆严重事故管理的重要策略之一,也是严重事故条件下保证压力容器完整性(Reactor Vessel Integrity,RVI)的典型方法之一.该文综述了国外在严重事故条件下压力容器完整性试验研究和理论分析的现状,总...     

A consistent approach to severe accident management

The approach adopted for severe accident management (SAM) at the Loviisa nuclear power plant (in Finland) is presented and discussed. The approach includes a number of significant hardware changes and procedures that allow lowering of the lower head thermal insulation and neutron shield assembly, opening of the ice condenser doors, and spraying (externally) of the steel shell of the containment. It is expected that with these changes we can assure in-vessel debris coolability and retention, gradual burning of the hydrogen with good access to the ice condenser, and long term stabilization of the containment pressure, even in the absence of the residual heat removal system. Methodological aspects of demonstrating these SAM objectives, and the status of work in support of related quantifications (of key phenomena), are included in sufficient detail to provide an integrated perspective of the approach taken. The detailed quantifications, separately on each task, will follow, as respective research and quantification programs come to completion.     

A simpler approach for predicting LWR vessel failure during severe accident conditions

Using closed-form solution techniques, models were developed for assessing the thermal and structural response of light water reactor (LWR) vessels and penetrations during severe accident conditions. Results from models are displayed as failure maps, generally developed in terms of non-dimensional groups, so that a broader range of reactor design parameters and severe accident conditions can be considered. In this paper, failure maps are used to compare LWR vessel response to three accident conditions. Results discussed within this paper illustrate the importance of vessel and tube geometrical parameters and material properties for predicting which vessel failure mode occurs first.     

Water boiling on the corium melt surface under VVER severe accident conditions

Experimental results are presented on the interaction of corium melt with water supplied on its surface. The tests were conducted in the ‘Rasplav-2’ experimental facility. Corium melt was generated by induction melting in the cold crucible. The following data were obtained: heat transfer at boiling water-melt surface interaction, gas and aerosol release, post-interaction solidified corium structure. The corium melt charge had the following composition, mass%: 60% UO_2+x–16% ZrO₂–15% Fe₂O₃–6% Cr₂O₃–3% Ni₂O₃. The melt surface temperature ranged within 1920–1970 K.     

Methodology for developing channel disassembly criteria under severe accident conditions for PHWRs

This paper presents a methodology to develop a model for disassembly of the coolant channels in Pressurized Heavy Water Reactors under severe accident conditions. This model gives criteria to decide when under severe accident condition coolant channels will rupture due to deterioration in material properties at high temperatures and increase in load due to creep sag of channels above it and hence get disassembled. Presently available severe accident codes use simplistic and optimistic criteria based on a predefined temperature to predict failure of fuel channels and an explicit criterion for disassembly of the channel is not covered. The coolant channel disassembly model developed in this paper is based on modeling the sag and pile up of channels. A uniform temperature along the length of the channel is assumed. The disassembly of the channel is assumed when the total strain at any location exceeds the failure strain for a given temperature. A 3D failure surface which is a plot of time to failure, temperature of the calandria tube and load on the calandria tubes (on account of no of channels piled up) is developed. This failure surface can be used as an input to severe accident codes to predict the progress of the core disassembly. A set of failure surfaces is recommended to be used if metal–water reaction on the outer surface is to be accounted for loss in ductility due to metal water reaction. The temperature transient of the calandria tube for a severe accident obtained from system thermal hydraulic codes can be mapped onto the failure surface. The time at which the mapped transient crosses the failure surface gives the time at which the calandria tube is disassembled. This disassembly model is an engineered model which is much more realistic as compared to the current temperature based conservative model for predicting severe accident progression.     

Finite element analyses of a BWR vessel and penetration under severe accident conditions

This study assesses the two-dimensional thermal response of a BWR vessel and drain line penetration to three types of debris bed; primarily metallic, primarily ceramic and metallic and ceramic layered, with sensitivity studies for the most severe case. Structural finite element analysis evaluates vessel elastic, plastic and creep response for two cases which bound the thermal challenge to the vessel.Thermal analysis results indicate that drain line failure does not occur for the case when metallic debris relocates to the lower head; structural analysis predicts that the vessel also remains intact for this case. In cases where ceramic debris relocates to the lower head, drain line temperatures peak near values where failure may occur within several minutes; whereas vessel failure is not predicted for 3.5 to 4.0 hours. Sensitivity studies indicate that large porosity debris or high heat removal rates from the vessel and drain line outer surfaces can preclude failure temperatures from occurring.     

Bottom nozzle failure mechanism of water reactor pressure vessel under severe accident conditions

Most of past studies devoted to the creep rupture of a nuclear reactor pressure vessel (RPV) lower head under severe accident conditions, have focused on global deformation and rupture modes. Limited efforts were made on local failure modes associated with penetration nozzles as a part of TMI-2 vessel investigation project (TMI-2 VIP) in 1990s. However, it was based on an excessively simplified shear deformation model. In the present study, the mode of nozzle failure has been investigated using data and nozzle materials from Sandia National Laboratory's lower head failure experiment (SNL-LHF). Crack-like separations were revealed at the nozzle weld metal to RPV interfaces indicating the importance of normal stress component rather than the shear stress in the creep rupture. Creep rupture tests were conducted for nozzle and weld metal materials, respectively, at various temperature and stress levels. Stress distribution in the nozzle region is calculated using elastic–viscoplastic finite element analysis (FEA) using the measured properties. Calculation results are compared with earlier results based on the pure shear model of TMI-2 VIP. It is concluded from both LHF-4 nozzle examination and FEA that normal stress at the nozzle/lower head interface is the dominant driving force for the local failure. From the FEA for the nozzle weld attached in RPV, it is shown that nozzle welds failure occur by displacement controlled fracture of nozzle hole not by load controlled fracture of internal pressure. Considering these characteristics of nozzle weld failure, new concept of nozzle failure time prediction is proposed.     

Development of a risk-informed accident diagnosis and prognosis system to support severe accident management

RISARD, risk-informed severe accident risk diagnosis system, is a computerized tool developed to improve a severe accident management (SAM) for a nuclear power plant and to effectively support the MCR and the TSC in executing the relevant SAM activities. In order to provide a diagnostic capability to a state of the plant and a prognostic capability for an anticipated accident progression, the system examines (a) a symptom-based diagnosis of a plant damage state (PDS) sequence in a risk-informing way and (b) a PDS sequence-based prognosis of key plant parameter behavior, through a prepared database (DB) containing plant-specific severe accident risk (SAR)-related information. For a given accident, the replicated use of these two processes makes it possible to obtain information about the functional states of the plant and containment safety systems expected at the time of a severe accident as well as future trend of the key plant parameters that are essentially required for taking the relevant SAM action, eventually leading to an answer about the best strategy for SAM. The foregoing concept for an accident diagnosis and prognosis can give the SAM practitioners more time to take action for mitigating the consequences of the potential accident scenarios since they are made in a simple, fast, and efficient way through a prepared SAR database and it is useful especially when the plant information available for SAM is incomplete and limited. The main purpose of this paper is to (a) introduce the concept of the RISARD system proposed to support SAM and its implementation through a prepared OPR1000 plant- and MAAP code-specific SAR database and (b) assess prediction capabilities of major events expected during the evolution of a severe accident through the system.