Early Containment Failure Research Articles

Development of analysis program for direct containment heating

Direct containment heating (DCH) is one of the potential factors leading to early containment failure. DCH is closely related to safety analysis and containment performance evaluation of nuclear power plants. In this study, a DCH prediction program was developed to analyze the DCH loads of containment vessel. The phenomenological model of debris dispersal, metal oxidation reaction, debris-atmospheric heat transfer and hydrogen jet burn was established. Code assessment was performed by comparing with several separate effect tests and integral effect tests. The comparison between the predicted results and experimental data shows that the program can predict the key parameters such as peak pressure, temperature, and hydrogen production in containment well, and for most comparisons the relative errors can be maintained within 20%. Among them, the prediction uncertainty of hydrogen production is slightly larger. The analysis shows that the main sources of the error are the difference of time scale and the oxidation of cavity debris.

ToF-SIMS and XPS characterizations of model fission products (I, Cs) deposits after thermal treatment simulating late phase conditions of a nuclear power plant severe accident

During a Severe Accident (SA) occurring in a nuclear power plant, many Fission Products (FP) are released from the degraded fuel and are transported in the Reactor Coolant System (RCS). Depending on their volatility, FP can be either deposited on the surface of the Reactor Coolant System (RCS) or transported into the containment building from where they may be released into the environment in case of early containment failure. This was the case for the Fukushima Daiichi (FD) accident with important FP releases which were observed following the containment vessel failure which lasted up to several weeks after the accident. If early phase releases were well predicted by SA simulation codes, delayed releases were not correctly predicted [1]. Such discrepancies can be attributed to the revaporization of deposits inside the RCS which were not yet modeled in the SA simulation tools. Previous experiments over the past 20 years evidenced the possible formation and revaporization of CsI [2]. However, up to now, no clear CsI revaporization mechanism has been determined. In order to propose a mechanism, this study (scheduled within the OECD/NEA/ESTER project) aims at determining the speciation of each species remaining on the substrate after revaporization tests under different conditions using XPS and ToF-SIMS (X-ray Photoelectron Spectroscopy and Time of Flight Secondary Ion Mass Spectrometry. These two combined techniques allow the identification of the main species of the residue leading to the proposition of two revaporization reactions.

SOARCA uncertainty analysis of a short-term station blackout accident at the Sequoyah nuclear power plant

The U.S. Nuclear Regulatory Commission initiated the state-of-the-art reactor consequence analyses (SOARCA) project to develop realistic estimates of the offsite radiological health consequences for potential severe reactor accidents. The SOARCA analysis of an ice condenser containment plant was performed because its relatively low design pressure and its reliance on igniters make it potentially susceptible to early containment failure from hydrogen combustion during a severe accident. The focus was on station blackout accident scenarios where all alternating current power is lost. Accident progression calculations used the MELCOR computer code and offsite consequence analyses were performed with MACCS. The analysis included more than 500 MELCOR and MACCS simulations to account for uncertainty in important accident progression and offsite consequence input parameters.Consequences from severe nuclear power plant accidents modeled in SOARCA are smaller than previously calculated. The delayed releases calculated provide more time for emergency response actions. The results show that early containment failure is very unlikely, even without successful use of igniters. The modeled behavior of safety valves is very important to this conclusion, but there is sparse data and a lack of established expert consensus on the failure rates under severe accident conditions. Even for scenarios resulting in early containment failure, the calculated individual latent fatal cancer risks are very small. Early and latent-cancer fatality risks are one focus of this paper. Regression results showing the most influential parameters are also discussed.

Failure domain analysis and uncertainty quantification using surrogate models for steam explosion in a Nordic type BWR

Sever accident mitigation strategy adopted in Nordic Boiling Water Reactors (BWRs) employs a deep water pool below the reactor vessel to fragment and quench core melt and provide long term cooling of the debris. One of the risks associated with this strategy is early containment failure due to ex-vessel steam explosion. Assessment of the risk of steam explosion is subject to significant (i) epistemic uncertainties in modelling and (ii) aleatory uncertainties in scenarios of melt release. For quantification of the uncertainties and the risk a full model (FM) based on TEXAS-V code and a computationally efficient surrogate model (SM) have been previously developed. FM is used to provide a database of solutions that is used for development of a SM, while SM is used in extensive sensitivity and uncertainty analysis. In this work, we compare the risk of containment failure with non-reinforced and reinforced hatch door for metallic and oxidic melt release scenarios. We quantify the error of SM in the approximation of the FM and assess the effect of the approximation uncertainty on risk assessment. We analyze the results and suggest a simplified approach for decision making considering predicted failure probabilities, expected costs, and scenario frequencies.

Impact of probabilistic risk assessment and severe accident research in reducing reactor risk

The development of probabilistic risk assessment (PRA) as a safety analysis tool and the implementation of lessons learned from risk studies in the design, operation and regulation of nuclear power plants has resulted in a substantial reduction in reactor risk. The lack of a strong technical basis for realistically assessing severe accident behavior, including the release and transport of radionuclides to the environment, resulted in some conservatism in early risk studies that distorted the true nature of severe accident risk. This paper describes the evolution of PRA over the past four decades, the benefits that have been achieved in the reduction of reactor risk, and the changes in the perspective of the nature of severe accident risk associated with the development of a strong technical basis for assessing severe accident consequences. Based on these developments, we conclude that the probability of early containment failure leading to a large, early release of radioactive material to the environment was over stated in these early risk studies. Although it is not possible to preclude the possibility of offsite early fatalities in a severe accident, the probability is extremely small, perhaps below the level at which it should be a key consideration in regulatory oversight. Conversely, as highlighted by the Fukushima accident, the potential for the societal impacts of land contamination represents an important element of reactor accident risk that has received insufficient consideration in the past. These findings have implications regarding preferred strategies for emergency planning and appropriate metrics for risk-informed regulation.

Simulation benchmark based on THAI-experiment on dissolution of a steam stratification by natural convection

Locally enriched hydrogen as in stratification may contribute to early containment failure in the course of severe nuclear reactor accidents. During accident sequences steam might accumulate as well to stratifications which can directly influence the distribution and ignitability of hydrogen mixtures in containments. An international code benchmark including Computational Fluid Dynamics (CFD) and Lumped Parameter (LP) codes was conducted in the frame of the German THAI program. Basis for the benchmark was experiment TH24.3 which investigates the dissolution of a steam layer subject to natural convection in the steam-air atmosphere of the THAI vessel. The test provides validation data for the development of CFD and LP models to simulate the atmosphere in the containment of a nuclear reactor installation. In test TH24.3 saturated steam is injected into the upper third of the vessel forming a stratification layer which is then mixed by a superposed thermal convection. In this paper the simulation benchmark will be evaluated in addition to the general discussion about the experimental transient of test TH24.3. Concerning the steam stratification build-up and dilution of the stratification, the numerical programs showed very different results during the blind evaluation phase, but improved noticeable during open simulation phase.

Phenomenological uncertainty analysis of containment building pressure load caused by severe accident sequences

This paper illustrates an application of a severe accident analysis code, MAAP, to the uncertainty evaluation of early containment failure scenarios employed in the containment event tree (CET) model of a reference plant. An uncertainty analysis of containment pressure behavior during severe accidents has been performed for an optimum assessment of an early containment failure model. The present application is mainly focused on determining an estimate of the containment building pressure load caused by severe accident sequences of a nuclear power plant. Key modeling parameters and phenomenological models employed for the present uncertainty analysis are closely related to the in-vessel hydrogen generation, direct containment heating, and gas combustion. The basic approach of this methodology is to (1) develop severe accident scenarios for which containment pressure loads should be performed based on a level 2 PSA, (2) identify severe accident phenomena relevant to an early containment failure, (3) identify the MAAP input parameters, sensitivity coefficients, and modeling options that describe or influence the early containment failure phenomena, (4) prescribe the likelihood descriptions of the potential range of these parameters, and (5) evaluate the code predictions using a number of random combinations of parameter inputs sampled from the likelihood distributions.

‘MAYBE YOU DON'T ACTUALLY EXIST’: CONTAINING SHAME AND SELF‐HARM IN A SCHOOL COUNSELLING SERVICE

abstractIn this paper, I argue that the school counsellor occupies a liminal position in the school environment, on the boundary between the private and the public, and that this position intrinsically reflects the paradoxical nature of shame, at once hidden and viewed. I review arguments that locate the development of primitive shame in early containment failure, with shame being defended against by rage against others or the self. I argue that the school counsellor's position commonly oscillates through the three positions of the child–home–school triangle, but that this is felt particularly acutely when the dynamics projected through this triangle are those of shame and shaming. For adolescents, the paradoxical nature of shame also finds a counterpart in self‐harm, a simultaneously hidden and viewed act which I see as a mapping of shame on the body. In exploring the operation of shame in self‐harming students, I also argue that chance encounters between counsellor and student in the external school world render projections about intimacy, intrusion and shame particularly powerful.

ICONE19-43642 UNCERTAINTY ANALYSIS RELATED TO THE EARLY CONTAINMENT FAILURE PHENOMENON FOR ULCHIN UNITS 3&4 NUCLEAR POWER PLANTS

Up to now, even though various researches related to severe accident phenomena have been going on in Korea, research results conducted in the U.S. in the 1980s are still applied to the probabilistic safety assessment as reference materials. To invigorate the risk informed regulation and application that use probabilistic safety assessment (PSA) results and information, the quality of PSA must be improved. To improve the quality and reduce uncertainties of PSA, phenomenological uncertainty analysis of the level 2 PSA has been done using the MELCOR code. For uncertainty analyses, the MELCOR 1.8.6 model for Ulchin Units 3&4 nuclear power plants were developed according to their own design data. Especially, control volumes of the reactor coolant system were divided in detail to simulate the natural circulation flow phenomenon. And then, uncertainty variables that have an effect on the early containment failure phenomenon were selected through the sensitivity analysis and the expert judgment. The MELOR uncertainty engine was used to generate samples of uncertainty variables. A number of MELCOR calculations were performed for various uncertainty variable combinations using the batch MELCOR for convenience.

CORIUM COOLABILITY UNDER EX-VESSEL ACCIDENT CONDITIONS FOR LWRs

In the wake of the Three Mile Island accident, vigorous research efforts were initiated to acquire a basic knowledge of the progression and consequences of accidents that involve a substantial degree of core degradation and melting. The primary emphasis of this research was placed on containment integrity, with: i) hydrogen combustion-detonation, ii) steam explosion, iii) direct containment heating (DCH), and iv) melt attack on the BWR Mark-I containment shell identified as energetic processes that could lead to early containment failure (i.e., within the first 24 hours of the accident). Should the core melt fail the reactor vessel, then non-condensable gas production from Molten Core-Concrete Interaction (MCCI) was identified as a mechanism that could fail the containment by pressurization over the long term. One signification question that arose as part of this investigation was the effectiveness of water in terminating an MCCI by flooding the interacting masses from above, thereby quenching the molten core debris and rendering it permanently coolable. Successful quenching of the core melt would prevent basemat melt through, as well as continued containment pressurization by non-condensable gas production, and so the accident progression would be successfully terminated without release of radioactivity to the environment. Based on these potential merits, ex-vessel corium coolability has been the focus of extensive research over the last 20 years as a potential accident management strategy for current plants. In addition, outcomes from this research have impacted the accident management strategies for the Gen III+ LWR plant designs that are currently being deployed around the world. This paper provides: i) an historical overview of corium coolability research, ii) summarizes the current status of research in this area, and iii) highlights trends in severe accident management strategies that have evolved based on the findings from this work.

Modelling of PWR lower head failure under severe accident loading using improved shells of revolution theory

In the study of severe pressurized water reactor accidents, the scenarios that describe the relocation of significant quantities of liquid corium at the bottom of the lower head are usually investigated from the mechanical point of view. In these scenarios, the risk of a breach and the possibility of a large quantity of corium being released from the lower head exists. This may lead to an out of vessel steam explosion or to direct heating of the containment; both which have the potential to lead to early containment failure. Within the framework of the OECD Lower Head Failure (OLHF) programme, a simplified model based on the theory of shells of revolution under symmetrical loading was developed by IRSN. After successfully interpreting some other representative experiments on lower head failures, the model was recently integrated into the European integral severe accident computer ASTEC code. The model was also used to obtain the thermo-mechanical behaviour of a 900-MWe pressurized water reactor lower head, subjected to transient heat fluxes under severe accident conditions. The main objective of this paper is to present: (1) the full mathematical formulations used in the development of the model, including their matrices and integrals defined by analytical expressions; (2) the two creep laws implemented, one for the American steel SA533B1 and one for the French steel 16MND5; and (3) the various numerical interpretations of experiments using the simplified model. This paper can be considered as a theoretical manual to aid users of the simplified model during modelling of lower head failures under severe accident conditions. One of the applications presented in this paper concerns the determination of a diagram representing the vessel time to failure as a function of the pressure level and the heat flux intensity. This information has been used by IRSN in probabilistic safety assessment and severe accident management analyses.

Semi-analytical modeling of a PWR lower head failure under severe accident conditions using an axisymetrical shell theory

The USNRC/SNL OLHF program was carried out within the framework of an OECD project. This program consisted of four one-fifth scale experiments of a reactor pressure vessel (RPV) lower head failure (LHF) under well controlled internal pressure and large throughwall temperature differentials; the objectives were to characterize the mode, timing and size of a possible PWR lower head failure in the event of a core meltdown accident. These experiments should also lead to a better understanding of the mechanical behavior of the reactor vessel lower head, which is of importance both in severe accident assessment and the definition of accident mitigation strategies. A well-characterized failure of the lower head is of prime importance for the evaluation of the quantity of core material that can escape into the containment, since this defines the initial conditions for all ex-vessel events. A large quantity of escaping corium may lead to direct heating of the containment or ex-vessel steam explosion. These are important issues due to their potential to cause early containment failure. The experiments also provide data for model development and validation. For our part, as one of the program partners, a 2D semi-analytical model has been developed and used to simulate these experiments. The aim of this effort is to develop a simplified but well predicting code that can be then implemented in European integral severe accident computer codes (ASTEC, ICARE/CATHARE). This paper presents the detailed mathematical formulation of this simplified method which is used to interpret the experimental results. The axi-symmetric shell theory under internal pressure proposed by Timoshenko has been utilised. The solution to the equilibrium equations is presented, with particular attention to the Rabotnov analytical formula. The radius and the polar angle of the deformed structure have been written as analytical expressions in order to take the large displacements and large strains into account using our mathematical formulation. The Norton type creep law and the Kachanov damage law have been used. Several failure criteria were used in the calculations and their effect on the numerical results is discussed. This 2D semi-analytical model gives very satisfactory results when compared, with the experimental and numerical results that were presented recently in the Benchmark calculations based on the first test of the OLHF program. The performance of this model is also illustrated by its capacity to accurately simulate the deformation of the lower head, including the variation of wall thickness.

Simplified modeling of a PWR reactor pressure vessel lower head failure in the case of a severe accident

In order to characterize the timing, mode and size of a possible lower head failure (LHF) of the reactor pressure vessel (RPV) in the event of a core meltdown accident, several large-scale LHF experiments were performed under the USNRC/SNL LHF program. The experiments examined lower head failure at high pressures (10 MPa in most cases) and with small throughwall temperature differentials. Another USNRC/SNL LHF program, called the OLHF program, has been undertaken in the framework of an OECD project. This was an extension of the first program and dealt with low and moderate pressures (2–5 MPa) but with large through wall temperature differentials. These experiments should lead to a better understanding of the mechanical behavior of the reactor vessel lower head, which is of importance both in severe accident assessment and the definition of accident mitigation strategies. A well-characterized failure of the lower head is of prime importance for the evaluation of the quantity of core material that can escape into the containment, since this defines the initial conditions for all external-vessel events. The large quantity of escaping corium may lead to direct heating of the containment. This is an important severe accident issue because of its potential to cause early containment failure. The experiments also provide data for model development and validation. For our part, as one of the program partners, numerical modeling was performed to simulate these experiments. This paper presents a detailed description of three of our numerical models used for the simulation. The first model is a simplified semi-analytical approach based on the theory of a spherical shell subjected to internal pressure. The two other methods deal with 2D finite element (2D-FE) modeling: one combines the Norton–Bailey creep law with a damage model proposed by Lemaitre–Chaboche while the other uses only a creep failure criterion but takes into account thermo-metallurgical phase transformations. One of the objectives of the simulation is the development of a simplified and valid analytical code that can be implemented in integral severe accident computer codes. The numerical results are consistent with the experimental measurements. The effect on the numerical results of the multiphase transformation of the shell material and of the two failure criteria used, one involving necking (Considère's criterion) and the other involving creep damage (Lemaitre–Chaboche) is discussed.

Public health risks of substituting mixed‐oxide for uranium fuel in pressurized‐water reactors

The U.S. Department of Energy (DOE) has awarded a contract to the consortium Duke Cogema Stone and Webster (DCS) to dispose of up to 33 tonnes of excess weapons‐grade plutonium (WG‐Pu) by irradiating it in the form of mixed‐oxide (MOX) fuel in four U.S. commercial pressurized‐water reactors (PWRs). This paper estimates the increase in risk to the public from using WG‐MOX at these reactors and finds that it exceeds recently established Nuclear Kegulatory Commission (NRC) guidelines. Therefore, the NRC will have a technical basis for prohibiting the use of MOX at these reactors unless the risk that they will experience a severe accident can be significantly reduced. MOX fuel will displace a fraction of the low‐enriched uranium (LEU) fuel that these reactors currently use. Because MOX cores have greater quantities of plutonium and other actinides than LEU cores throughout the operating cycle, the source term for radiological releases caused by severe reactor accidents will be greater for MOX‐fueled PWRs. In this paper, the radiological consequences to the public from containment failure or bypass accidents at MOX‐fueled PWRs are calculated, and compared to those resulting from the same accidents at LEU‐fueled PWRs. This paper finds that compared to LEU cores, the number of latent cancer fatalities (LCFs) resulting from an accident with core melt and early containment failure would be higher by 39%, 81% or 131% for full WG‐MOX cores, depending on the fraction of actinides released (0.3%, 1.5% or 6%). Under the DCS plan, in which WG‐Pu will be purified using an aqueous process and only 40% of the core will be loaded with WG‐MOX, the number of LCFs would be 11%, 25% or 30% higher, respectively. The average LCF risk to individuals within ten miles of a severe accident approximately doubles for a full WG‐MOX core, and increases by 26% for a DCS core. These results are of particular concern for the nuclear plants in the DCS consortium, Catawba and McGuire. These plants have ice‐condenser containments, which Sandia National Laboratories estimates are at least two orders of magnitude more vulnerable to early failure than other types of PWR containments. The findings of this paper also apply to the proposed use of WG‐MOX in WER‐1000 reactors in Russia, which meet less stringent safety standards than U.S. reactors.

The impact of assessing lower head integrity on boiling water reactor probabilistic risk assessments

Abstract Over the last several decades, much effort has been directed at estimating the likelihood of a large early release of radioactivity during a nuclear accident. This effort has culminated in the Individual Plant Examinations (IPEs) for the over 100 US nuclear power plants and the NUREG 1150 study. The large early release of radioactivity requires core damage with loss of primary containment integrity during the accident. Given a successful reactor scram, early containment failure coupled with a large release of radioactivity will only occur if the reactor core vessel is breached by core debris. Most IPE/PRA studies performed to date have not considered the possibility of quenching core debris in the lower plenum. Consequently, lower head failure is presumed to closely follow the onset of core damage. Therefore, these assessments did not address the role that in-vessel debris retention plays in preserving primary containment integrity, nor do they propose a criterion for evaluating the integrity of the vessel lower head given that core damage has occurred. Yet preserving the vessel lower head integrity is a necessary condition for satisfying the plant design and licensing basis. Therefore, a more complete treatment of the risk associated with nuclear plant operation includes an evaluation of the ability to retain the core debris in-vessel. This paper presents a performance requirement for vessel integrity to be used in probabilistic risk assessments; evaluates the impact the core damage progression and lower plenum quenching models have on the likelihood of terminating the damage progression in-vessel; documents the significant reduction in BWR containment failure probability that can occur when appropriate core damage and lower head quenching models are used; reviews the implications of core debris quenching in the lower head on BWR PRA modeling; argues why crediting the capability to maintain vessel integrity is necessary from a safety point of view. These results and conclusions are derived from consideration of a BWR 4 plant with a 251 inch vessel. However, the concepts are generally applicable and results specific to other BWR designs can be developed using the methodology presented in this paper.

Rough Estimates of Severe Accident Containment Loads Accompanying Vessel Breach in Boiling Water Reactors

Several probabilistic risk assessments (PRAs) have identified containment loads accompanying reactor vessel failure as a major contributor to the probability of early containment failure during severe accidents. Two significant contributors to these loads are phenomena referred to as steam spike and direct containment heating. To date, direct application of experimental and analytical studies of these phenomena to boiling water reactors (BWRs) are constrained by two limitations: (a) they are based on applications of large, complex containment response analysis computer codes, for which values of many major input parameters are highly uncertain, or (b) they only address pressurized water reactor containment designs. Relatively simple, parametric models are developed which allow a PRA analyst to evaluate the range of conditions under which steam spike or direct containment heating may be important contributors to containment loads for postulated severe accidents in BWRs. The models have been applied to a representative BWR/4 Mark I containment design to illustrate calculated results.

A review of containment accidents

The consequences of severe reactor accidents depend greatly on containment safety features and containment performance in retaining radioactive material. If the containment function is maintained in a severe accident, the radiological consequences will be minor. If the containment function does fail, the timing of failure can be very important. The longer the containment remains intact relative to the time of core melting and radionuclide release from the reactor coolant system, the more time is available to remove radioactive material from the containment atmosphere by engineered safety features or natural deposition processes. Delay in containment failure or containment bypass also provides time for protective action, a very important consideration in the assessment of possible early health effects. Thus, in evaluating the performance of a containment, it is convenient to consider no failure, early failure, late failure, and containment bypass as separate categories characterising different degrees of severity. The paper reviews the containment challenges posed by the severe accidents, on the basis of an extensive list of references. First the phenomena that could lead to early containment failure are described: direct containment heating, steam explosions, hydrogen combustion, and isolation failures. Then the late containment failure modes are treated: gradual overpressurization, basemat meltthrough, and overheating. Finally, some words are said about containment bypass.

Sensitivity and Uncertainty Analysis of Accident Management Strategies Involving Multiple Decisions

A framework for assessing severe accident management strategies is presented using a new analytical tool, namely, influence diagrams. This framework includes multiple and sequential decisions, sensitivity analysis, and uncertainty propagation, and is applied to a proposed set of strategies for a pressurized water reactor station blackout sequence. The influence diagram associated with these strategies is constructed and evaluated. Each decision variable, represented by a node in the influence diagram, has an uncertainty distribution associated with it. Using the mean value of these distributions, a best estimate assessment is performed, and each strategy is ranked with respect to the conditional frequency of early containment failure (ECF). For the preferred alternative, the sensitivity of the results to values of the input variables is investigated. The sensitivity of the ranking itself is then considered. The distributions of the uncertain variables are also propagated through the influence diagram to rank the alternatives with respect to the uncertainty associated with the calculated conditional frequency of ECF. Finally, the sensitivity of the variance of the output distribution, given the preferred decision alternative, to the uncertainty of the input variables is investigated.

Recent Advances in Severe Accident Technology — Direct Containment Heating in Advanced Light Water Reactors

The issues affecting high-pressure melt ejection (HPME) and the consequential containment pressurization from direct containment heating (DCH), as they affect advanced light water reactors (ALWRs), specifically advanced pressurized water reactors (APWRs), were reviewed by the U.S. Department of Energy Advanced Reactor Severe Accident Program (ARSAP). Recommendations from ARSAP regarding the design of APWRs to minimize DCH are embodied within the Electric Power Research Institute ALWR Utility Requirements Document, which specifies (a) a large, strong containment; (b) an in-containment refueling water storage tank; (c) a reactor cavity configuration that minimizes energy transport to the containment atmosphere; and (d) a reactor coolant system depressurization system. Experimental and analytical efforts, which have focused on current-generation plants, and analyses for APWRs were reviewed. Although DCH is a subject of continuous research and considerable uncertainties remain, it is the judgment of the ARSAP that reactors complying with the recommended design requirements would have a low probability of early containment failure due to HPME and DCH.

Severe accident mitigation programme for the forsmark and ringhals plants

In February 1986 licensing requirements regarding severe accidents in nuclear power plants were given by the Swedish Government. This regulation constitutes conditions for operation of the plants beyond 1988. The requirements are based on the conditions previously given for the Barsebäck plant including construction of the filtered venting system, which was completed at Barsebäck in 1985. For the Forsmark and Ringhals plants a strategy is being implemented to meet the new requirements. A strong emphasis is put on both hardware and procedural measures to bring the reactor core back to stable cooling - even if it is severely damaged - and maintain the containment integrity during an accident. The hardware modifications include measures to prevent temperature or pressure induced early containment failure for the BWRs, reliable back-up water sources for containment spray and means for filtered venting of all plants to prevent late containment failure by overpressure. The ultimate aim is to minimize the environmental impact of a severe accident and meet a release limit set at 0.1% of the core fission product inventory excluding noble gases.