Comparative Analysis of Containment Response to Hydrogen Combustion under a Station Blackout for Different PWR Designs

Insights and comparisons of the level-2 results of recent probabilistic safety analyses

On the basis of passive containment cooling system concept, the one-dimensional codes are developed for the analysis of containment thermal-hydraulic characteristics under accident conditions, which can be applied to deal with large dry concrete containment in nuclear power plant. In order to build up the hypothesis flowing path, the containment space is divided into a rising channel and downward ring during the geometric modeling process. In this paper, the physical models, the identification of solving methods and the verification of the codes are introduced. It is assumed that the control volumes take the adiabatic condition in addition to heat exchanger and breaks, and there is no exchange of mass, momentum and energy between each volume in the rising channel and corresponding volume in downward ring. In the analysis, we also assume that the heat in the containment can only be transferred through natural circulation by passive containment cooling system. Furthermore, the break is supposed in the center of bottom of the containment. In this paper, the responses of the containment are predicted with the codes under large LOCA scenario. Under the same conditions, the characteristics of the natural circulation are also analyzed through the codes for the passive containment cooling system. The results can provide some references for the design of the passive containment cooling system.

Integrated phenomenological analysis of containment response to severe core damage accidents

Authors

Construction of the CAREM-25 full-size prototype, a very low power nuclear power station [25 MW(electric)], is scheduled to begin in Argentina in 2001. The CAREM-25 is designed based on principles of inherent safety, passive safety functions, and ease of operation. This paper analyzes the safety philosophy from the point of view of risk by performing a level-III probabilistic safety assessment (PSA) of this prototype. The specific PSA steps are discussed, including a specially developed method to obtain representative initiating events, system analysis by fault trees, event development in event trees, plant and containment response analysis, containment event tree development, consequence calculations, and risk representation. The PSA results are presented and discussed in terms of their own values as well as in comparison to other PSA results performed for larger nuclear power plants (NPPs). The advantages of the CAREM-25 from the risk point of view are studied in terms of the effective reduction of both the probability of severe accident sequences and the potential consequences of such sequences (radiological and emergency preparedness impact). The risk point of view also provides a perspective to analyze the impact of several design modifications in order to further reduce the residual risk of the NPP. These design modifications, several of which have already been included in the prototype, are discussed and evaluated.

The purpose of Steam condensation on cold plate experiment facility (SCOPE) and Water film test (WAFT) is to verify the steam condensation and water film evaporation correlation within the parameter variation range of CAP1400 passive containment cooling system. These correlations were used for containment response analysis. Uncertainty and sensitivity analyses were performed for SCOPE and WAFT tests in this paper. Sampling-based sensitivity analysis with uncertainty propagation is a new parameters sensitivity analysis method, and the importance of input parameters could be evaluated by calculating the correlation coefficients between input parameters and the output target parameter. This method was used to acquire the influence of the measured input parameters uncertainty on the output target parameter. The results show that air and steam flow rate, coolant flow rate, inlet and outlet water temperature are the main source of the uncertainty for SCOPE. Inlet film flow rate, inlet air flow velocity and plate surface temperatures are the main source of the uncertainty for WAFT. Sensitivity analysis results may provide support for experiment measurement system optimization to reduce the target parameter error range. Uncertainty analysis is one important aspect of test data analyses, which is meaningful to the assessment of test results. Conventionally the partial derivative with respect to the input parameters is used to transfer uncertainty from the input parameters to the output parameter. However, in this method the partial derivatives of the output parameter sub the input parameters must be calculated. For complex engineering problems, it is usually difficult to acquire theoretical correlations for the partial derivatives. WILKs formula is used to determine the parameter tolerance interval with certain probability content and confidence level. The tolerance interval is a good way to well describe the uncertainty of parameters. The nonparametric statistics with WILKS correlation were widely used in the best-estimate plus uncertainty (BEPU) accident analyses. However, little work has been conducted on the experiment results uncertainty analysis with that method. In this paper nonparametric statistics with WILKS correlation was used to acquire key parameters uncertainty. And the results show that key output parameters uncertainty for SCOPE and WAFT are within the reasonable range. Uncertainty Propagation Methods were implied for test results Sensitivity and Uncertainty Analysis in the paper, which may be conveniently applied to the other experiment data analyses and also valuable to the engineering project.

Lungmen Nuclear Power Plant in Taiwan is a twin-unit Advanced Boiling Water Reactor (ABWR) plant. In this study, a long-term GOTHIC model for the Lungmen ABWR primary containment response analysis is established. The wetwell space is vertically divided into several volumes to catch the pool temperature stratification effect. The long-term containment responses for a double-ended feedwater line break (FWLB) accident are calculated. The fuel decay heat is absorbed by the reactor coolant, and the coolant flows to the containment via the broken line. The suppression pool is gradually heated up by the high-temperature gas-water mixture following through horizontal vents. To reduce the pool temperature, the Residual Heat Removal (RHR) system will be required to operate in the suppression pool cooling mode. The RHR pumps have suction flow from suppression pool and discharge it to the RHR heat exchangers for cooling. The cooled water then returns to the pool. An elevated RHR return line is desired to avoid the cooled water being directly sucked again. The wetwell temperature stratification associated with the RHR return line elevation is investigated in this study. Effects of the RHR return line elevation on the pool temperature can be determined since the whole wetwell space is not lumped as a node only. The calculated peak pool temperature is 92.6°C based on the plant piping configuration. The peak temperature can be reduced to 88.9°C by returning the water via the wetwell spray spargers located in the top of the wetwell. However, it should be noted that using the wetwell spray also pressurizes the wetwell because the pool water temperature is higher than that of airspace during the late period of the event. Returning the pool water via the wetwell spray spargers is not suggested because it causes long-term wetwell pressurization.

A model of the Marviken Power Station has been built with the analysis code RELAP5. The model has been tested against the measurement data from the experiments run in the facility under the period August 1972 to April 1973 when a series of full-scale blowdown test have been performed. The aim of the work is the employment and the evaluation of the mono-dimensional code RELAP5 for the containment response analysis. Two different blowdown experiments have been selected and used as comparison for the validation test. The simulations have been performed using models that include the main hydraulic and thermal features described in the actual facility. The heat structures and the pressurizers, used to simulate the spray cooling system, have been introduced and treated in separated cases for a better understanding of the response of the model with regards to the presence of these components. In this way it has been possible to distinguish the effects caused by the heat structures from the results induced by the pressurizers. In this way three configurations have been used for each blowdown, one which describes the mere hydraulic system, a second one with the addition of detailed heat structures and a third one including the spray system. The temperature and the pressure initial values for the simulations are not in a steady state but they reproduce the conditions as described at the beginning of each experiments. The discharge rate and the specific enthalpy are accurately set according to the experimental data. The model response for the global fluid pressure and temperature behaviour is, in general, in good agreement with the experimental local data. The difference between the calculations and the experiments are mainly attributed to the inefficient heat transfer between the non-condensable phase (air) and the walls of the structure: the energy is therefore retained in the gas without the possibility to be released to the surroundings causing higher temperatures and pressures in the fluid itself in comparison with the experiments. Furthermore, a one-dimensional model is not able to predict a good mixing of the fluid in the wetwell as it happens in the reality thanks to the natural convection. The pressure, the temperature, the condensation rate as well as the energy in the drywell and in the wetwell are discussed in the paper and compared with the experimental results. In this paper, due to space limitations, only the results for the hydraulic and the heat structure configurations from one blowdown experiment will be discussed.

Compared with conservation evaluation model, best estimate plus uncertainty (BEPU) method can obtain more realistic results and gain larger license margins with respect to the safety criteria. In view of this, a BEPU method named 4S (SNERDI Statistical Solution for Safety) has been developed, according to the basic principles of evaluation model development and assessment of RG 1.203. The characteristics of 4S method are as follows: The output uncertainty is quantified by using random sampling and propagation of input uncertainties. Global sensitivity analysis is used to support PIRT establishment. Uncertainties of model parameters are calibrated and validated by using separate effects tests considering measuring uncertainties. DAKOTA code is used for uncertainty and sensitivity analysis. An automatic BEPU analysis platform has been developed by coupling DAKOTA and different reactor safety analysis codes, and code calculations can be performed in parallel. BEPU analysis of mass and energy release and containment pressure response of CAP1400 under a postulated double-ended cold leg break loss of coolant accident (DECL LOCA) has been carried out by coupling DAKOTA, a mass and energy release analysis code and a containment analysis code. In total, 21 uncertain input parameters are considered. To make the results more stable, the sample size is 124 and the third highest peak pressure is used as the pressure upper bound (with 95%/95% probability/confidence) based on Wilks’ formula. The calculated results show that the peak pressure upper bound is obviously lower than the present conservation method used in license application, with more than 10% analysis margin. Influences of input parameter uncertainties on the containment peak pressure have been analyzed, according to the partial rank correlation coefficients calculated by DAKOTA. The results show that the input parameters mainly affecting the peak pressure are the containment condensation heat transfer multiplier, initial containment temperature, break resistance, decay heat, initial containment pressure, Core Makeup Tank (CMT) resistance multiplier and initial containment humidity.

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.

Experimental investigation of steam-air condensation on containment vessel

Thermal-hydraulic analysis of NSSS and containment response during extended station blackout for Maanshan PWR plant

An analysis of thermohydraulic response during a station blackout (SBO) accident for the APR 1400 nuclear power plant is performed using MELCOR version 1.8.6. MELCOR 1.8.6 results for the SBO scenario are benchmarked with MELCOR 2.1. The simulation of the SBO accident with MELCOR 2.1 was done by the APR 1400 reactor designer company (KEPCO). This research consists of two parts; the first part is related to the results of MELCOR 1.8.6, and the thermo-hydraulic analysis of MELCOR1.8.6 has been done. Analysis of thermohydraulic response is focused on investigating thermohydraulic parameters, such as core pressure, fuel clad temperature, water mass flow rate in the core, time of fuel clad failure, time of lower head failure, and time of containment failure. In the second part, the results of MELCOR version 1.86 have been benchmarked with the results of MELCOR 2.1. The results of the analysis of containment pressure changes in version 1.8.6 showed that the effect of pressure increase in containment is mostly due to the increase in carbon dioxide mass, but in version 2.1, the increase in pressure is more due to water vapor.

Analyses of three long term PWR Station Black-Out (SBO) scenarios with and without mitigation are performed using the US NRC source term code MELCOR. Refilling of the secondary side of the steam generator (SG) with fire water pumps was previously identified as a potentially important accident management measure. To assess parametrically the effect of the restored availability to refill of SG secondary, SBO sequences are analyzed both without any mitigation and with mitigation at different times into the accident representing different stages of the accident progression. The scenarios studied were (i) base-line SBO without any thermally-induced Reactor Coolant System (RCS) breach, and then sequences with assumption of (ii) Surge Line failure and (iii) thermally-induced SG tube rupture (SGTR). A detailed model was used for the description of flows in the hot legs and in SGs to enable us to simulate the counter-current natural circulation which is inherent in these types of scenarios, with the cold leg plugged by water in the loop seal. These were long-term simulations, some of them up to 5 days of the transient, with the analyses of the containment response and fission product (FP) releases to environment. For the relevant cases, the molten core-concrete interactions (MCCI) are modeled in detail in the complex lower containment geometry with several distinct volumes (cavity and other rooms with concrete walls) subject to ablation by the molten corium; the challenges to containment barrier by the late overpressurization and by concrete ablation are evaluated. With the Passive Autocatalytic Recombiners, PARs, installed the hydrogen is found to be of less threat to containment integrity. For the thermally-induced SGTR the impact of the chosen mitigation strategy on potential bypass FP release to environment is also assessed for different times of the refill. The results indicate that mitigation by the SG secondary refill can be very effective for the base-line SBO sequences. For the thermally-induced SGTR it is effective only when the refill is achieved early after the SG tube failure. Mitigation by the pressurizer-loop SG refill was much less successful in the case of the Surge Line failure sequences.

A sensitivity analysis of thermal hydraulic response in containment during a 'station blackout' (the loss of all AC power) accident at Browns Ferry unit one plant was performed with the computer code MARCH 1.0. In the analysis, the plant station batteries were assumed to be available for 4h after the initiation of the accident. The thermal hydraulic response in the containment was calculated by varying several input data for MARCH 1.0 independently and the deviation among calculated results were investigated.The sensitivity analysis showed that (a) the containment would fail due to the overtemperature without any operator actions for plant recovery, which would be strongly dependent on the model of the debris-concrete interaction and the input parameters for specifying the containment failure modes in MARCH 1.0, (b) a core melting temperature and an amount of water left in a primary system at the end of the meltdown were identified as important parameters which influenced the time of the containment failure, and (c) experimental works regarding the parameters mentioned above could be recommended.