Review on condensation on the containment structures

In the last two decades condensation on the containment structures in presence of noncondensables has received substantial attention by nuclear scientists and engineers. The reason is that many Generation III and III+ reactors rely on passive systems operating under natural circulation. As a consequence, a vast number of publications have been made in the open literature. This paper reviews the specific physical phenomena that are involved in condensation and discusses how they have been considered in the different available models. In addition, it explores the data that have been used for validation and provides some insights on the effective suitability for this purpose. Finally, the paper summarizes the current codes' capabilities to deal with wall condensation in the presence of noncondensables according to the most recent available validation exercises.     

Aerosol growth by steam condensation in a PWR containment

The fraction of steam which condenses as an aerosol in the cooling through surfaces of a well-mixed steam-air mixture confined in a cavity is calculated. A previous theory which predicts this fraction is extended so as to allow supersaturations, S, to occur in boundary layers, and the fraction is shown to be rapidly varying towards its maximum value which occurs in the saturated limit, S = 1. Only when the cooling surfaces are cold is the fractional aerosol condensation predicted to be large, and the aerosol is expected to evaporate slowly for small bulk-wall temperature differences. Results applicable to the cooling of PWR containments by water sprays are obtained. They indicate that it is unlikely that any net aerosol growth is produced.     

A diffusion layer model for steam condensation within the AP600 containment

Steam condensation plays a key role in removing heat from the atmosphere of the Westinghouse AP600 containment in case of a postulated accident. A model of steam condensation on containment surfaces under anticipated accident conditions is presented and validated against an extensive and sound database. Based on the diffusion layer theory and on the use of the heat/mass transfer analogy, one can deal with large temperature gradients across the gaseous boundary layer under high mass flux circumstances. The thermal resistance of the condensate film, as well as its wavy structure, have also been considered in this model. As compared to Anderson et al. (1998) (Experimental analysis of heat transfer within the AP600 containment under postulated accident conditions. Nucl. Eng. Des. (submitted)) experimental database, an average error lower than 15%, within the experimental confidence range, has demonstrated its remarkable accuracy. In particular, the model has shown a good response to the influence of primary variables in steam condensation (i.e. subcooling, noncondensable concentration and pressure), providing a mechanistic explanation for effects such as the presence of light noncondensable gas (i.e. helium as a simulant for hydrogen) in the gaseous mixture. In addition, the model has been contrasted against correlations used in safety analysis (i.e. Uchida, Tagami, Kataoka, etc.) and occasionally to Dehbi’s database. This cross-comparison has pointed out several shortcomings in the use of these correlations and has extended the model validation to other databases.     

Iodine removal by condensation from containment atmosphere in post external event conditions

An integral solution is derived for the problem of passive removal rate of a typical fission product (iodine), from a gas-vapor mixture, by condensate liquid film adjacent to a containment wall. The analytical model consists of a coupled set of five conservation equations: momentum, energy and three matter conservation equations for each individual component of the gas mixture: air, steam and elemental iodine. The set is solved in conjunction with two balance equations for the mass and energy transport at the interface with the condensate layer. The model accounts for free convection due to temperature and concentration gradients, for mass and thermal diffusion and for variable properties in both the liquid and the gas-vapor regions. An economic solution procedure of this model is presented and employed for a wide range of parameters. The computational results of this study are used to derive an efficient correlation which provides a quick and simplified means of the calculation of the iodine mass removal coefficient, as a function of the bulk conditions. Some results are compared to other theoretical and experimental works showing good agreement within about 10%. The significance of the removal process in the “external event” scenario is analyzed and found to be much higher than in scenarios that start with a mechanical failure in the primary system.     

LOCA steam condensation loads in BWR Mark II pressure suppression containment system

Hydrodynamic loads induced in the BWR Mark II pressure suppression containment system during a loss-of-coolant accident (LOCA) were investigated using a large scale test facility. The maximum-bounding loading conditions on the pressure suppression pool-boundary structures were defined by conducting experiments for a wide range of parameters. The maximum-bounding loads occurred when steam, with air concentration less than 1% in weight, was injected at moderate rates ( 30 kg/m²·s) into a low-temperature (below 310 K) pool. Such conditions are most likely to be encountered during LOCAs with intermediate break sizes.     

Modeling of turbulent condensation heat transfer in the boiling water reactor primary containment

A condensation heat transfer model is developed for the purpose of predicting the atmosphere temperature response within the primary containment of a boiling water reactor during the initial forced convection heat transfer period following a postulated loss-of-coolant accident. The model utilizes simultaneous heat and mass transfer for the process of condensation in the presence of a non-condensible gas. The gas-vapor diffusion layer formed is in the mode of turbulent, forced convection. The predicted heat transfer is determined to be diffusion controlled with negligible resistance being contributed by the condensate film. The model is qualified through the analysis of the response of a containment test facility; the results compare favorably with experimental observations made by the General Electric Co. Predicted temperature responses for a typical containment are also shown and compared with those obtained through use of the Uchida heat transfer correlations.     

Reliability analysis of nuclear containment without metallic liners against jet aircraft crash

The present study presents a methodology for detailed reliability analysis of nuclear containment without metallic liners against aircraft crash. For this purpose, a nonlinear limit state function has been derived using violation of tolerable crack width as failure criterion. This criterion has been considered as failure criterion because radioactive radiations may come out if size of crack becomes more than the tolerable crack width. The derived limit state uses the response of containment that has been obtained from a detailed dynamic analysis of nuclear containment under an impact of a large size Boeing jet aircraft. Using this response in conjunction with limit state function, the reliabilities and probabilities of failures are obtained at a number of vulnerable locations employing an efficient first-order reliability method (FORM). These values of reliability and probability of failure at various vulnerable locations are then used for the estimation of conditional and annual reliabilities of nuclear containment as a function of its location from the airport. To study the influence of the various random variables on containment reliability the sensitivity analysis has been performed. Some parametric studies have also been included to obtain the results of field and academic interest.     

Application of fourier transform analysis to determine transient condensation heat transfer coefficients in reactor containment experiments

The method of Fourier transform analysis is used to determine the instantaneous values of condensation heat transfer coefficient at a point within the containment vessel of a simple blowdown rig. The shape of the measured heat transfer transient appears to be similar to that of the energy outflow from the blowdown pressure vessel, and a heat transfer coefficient which varies in this way is shown to give close fit to the shape of containment pressure transient when used in a lumped parameter calculation.     

Physical-mathematical model of volume condensation and transport of aerosols inside the containment shell of a vver reactor

Main Science Center of the Russian Federation — Physics and Power Engineering Institute. Translated from Atomnaya énergiya, Vol. 79, No. 5, pp. 376–381, November, 1995.     

Simulation of containment cooling with outside spray after a core meltdown

In the most severe hypothetical loss of coolant accident, the reactor core melts and falls into the containment sump, there evaporating much of the sump water and raising the pressure within the containment building. One possible method to remove the decay heat is to cool the steel containment shell with an outside spray system. To perform the structural analysis needed to confirm the integrity of the containment, the thermal profile in the containment wall must first be found. The purpose of this work is to develop a computer code to calculate this transient profile. Other aspects such as hydrogen build-up are not considered in this code.The method uses relationships for the natural convection-partial condensation phenomena occurring at the containment internal surfaces, iteratively coupled to a one-dimensional heat balance at the wall to solve for the wall temperature as a function of angular position. A differential calculation as a function of time treats the thermodynamic changes within the containment as quasi-steady state. The result is a quick-running code capable of analyzing the temperature transient for several hours following the LOCA with a few minutes of computing time.     

Non-iterative model for condensation heat transfer in presence of non-condensable gases inside passive containment cooling vertical tubes

Some contributions have been stated in order to improve the modeling of concurrent downflow condensation in presence of non-condensables inside vertical tubes. In particular, the influence of non-condensables over the liquid side heat transfer has been considered. The new proposed mechanistic models solve explicitly the real interface temperature by means of a cubic or a fourth order equation. As these models have a non-iterative nature, they can avoid the weakest point of the traditional mechanistic models, which is the slowdown computation if the model had to be implemented in a code. Moreover, as the main non-condensables effects can be accounted for in the heat and mass transfer processes, the new models will be more realistic. The models have been validated with the Vierow experimental data, obtaining a total average relative error, for the fourth order equation method model, of 21% for 268 points.     

Interaction between natural convection and condensation heat transfer in the passive containment cooling condensers of the ESBWR reactor

The passive containment cooling condensers (PCCC) are the crucial part of several new reactor designs. In this paper we have developed several models to compute the heat transfer coefficients for the following cases: (i) condensation in presence of noncondensable gases inside the PCCC tubes; (ii) laminar natural convection for vertical cylinders; (iii) turbulent natural convection for vertical cylinders. These models have been implemented in the TRAC-BF1 code, and we have studied the interaction between the conservation regime inside the condenser tubes and the natural convection outside.     

A numerical investigation of three-dimensional flows in large volumes in the context of passive containment cooling in BWRs

The paper describes Computational Fluid Dynamics (CFD) calculations undertaken in support of analyses of three-dimensional flows that take place in the drywell volumes of advanced boiling water reactors with passive decay–heat removal systems. Data for comparison are taken from the 1/40th-scale European Simplified Boiling Water Reactor (ESBWR) mock-up facility PANDA under conditions of symmetric steam injection and asymmetric outflow. Steady-state simulations for pure steam conditions illustrate how the separate flow streams mix to ensure balanced outflow conditions to the condenser units. A transient calculation has also been performed to examine how air, assumed to be released from solution in the PANDA boiler, would ultimately accumulate in the separate condenser units. Results provide a possible explanation for the rundown in performance of one of the condensers, behaviour which was repeatedly observed in some of the earlier PANDA tests. The work also provides more general insights on how trace amounts of non-condensable gases may accumulate in passive cooling equipment.     

Experimental study on the direct contact condensation of steam jet in the passive safety injection tank

The direct contact condensation and subsequent thermal mixing by the injected steam jet onto a quiescent coolant inside a tank were examined experimentally to simulate the phenomena in passive safety injection systems. Specifically, the influence of the steam injection velocity was studied. Even though the total flow rate of injected steam was unchanged, the pressure inside the tank increased quickly at the larger nozzle diameter. Additionally, at a larger nozzle diameter, the thickness of the thermal mixing zone decreased because the amount of direct contact condensation decreased. For the in-depth study on the role of the nozzle size for the thermal mixing, the particle image velocimetry method was used to understand the flow field of water inside the tank. The visualization results demonstrated the formation of a flow field in the coolant due to the expansion and contraction of the steam–air mixture boundary. Furthermore, the thermal mixing zone was found to be closely related to the penetration depth. Finally, a variety of penetration models were examined and compared against the experimental observation. The correlations based on the steam condensation approach under-predicted the penetration depth, whereas the approach that considers the momentum of non-condensable gas gave the reasonable prediction capability.     

Simulation of containment thermal-hydraulics in the Marviken Blowdown 16 experiment with ASTEC and CONTAIN codes

The boiling heavy water reactor Blowdown 16 experiment, which was performed in the Marviken experimental facility, was simulated with the ASTEC and CONTAIN codes. The main purpose of the work was the assessment of the codes for simulating thermal-hydraulic phenomena in a BWR containment at accident conditions. Simulated pressures, atmosphere temperatures and wetwell pool masses are compared to experimental measurements. The results show that both codes satisfactorily reproduced the overall containment thermal-hydraulic behaviour. The simulations also allow a more detailed understanding of the governing mechanisms during the performed experiment.     

Acoustic monitoring of containment tendons

Assured safety and operational reliability of post-tensioned concrete components of nuclear power plants are of great significance to the public, electric utilities and regulatory agencies. Prestressing tendons provide the principal reinforcement for 40% of the containment structures in the United States. This paper briefly examines current in-service inspection requirements for prestressed containments and also reviews the feasibility of developing a passive surveillance system for identification of ruptures in tendon wires and its application to a one-tenth scale ring model containment structure.     

Multi-dimensional modelling of multiphase flow physics: high-speed nozzle and jet flows — a case study

Multi-dimensional modelling of multiphase flows has become more prevalent as computer capabilities have significantly expanded. Such analyses are necessary if the flow physics demonstrates behavior that is fundamentally different from the estimates of one-dimensional analyses. Multiphase multi-dimensional behavior may involve physical mechanisms that interact with the flow field transverse to the main fluid direction and feedback into downstream processes. Consider the physics of high-speed internal nozzle flow, downstream external jet flow and the dynamics of jet breakup. This is a prime example of a coupled problem where multi-dimensional aspects may need to be considered. This paper examines multiphase physics as an illustration of the conditions under which multi-dimensional modelling would be required. Internal nozzle flow can involve cavitation phenomena, and as the geometry becomes more abrupt or asymmetric, multi-dimensional modelling is required. High-speed simulations using our internal flow model, CAVALRY, indicate that cavitation behavior can become oscillatory as the nozzle shape is altered. This exiting internal flow emerges as a multi-dimensional external jet flow, whose downstream breakup can be noticeably influenced by the inlet conditions as well as the jet breakup mechanisms. Jet breakup models first developed for the TEXASV model are utilized in the multi-dimensional KIVA code simulations for gas–liquid flows. The simulation results suggest that similar jet breakup mechanisms are operative for a multi-fluid system. Our comparisons to particular sets of data for high-speed nozzle flow and jet breakup in a gas suggest that the approach can be extended to multiphase systems using similar concepts; i.e. TEXAS-3d.     

Simulation on SIMS depth profiling of delta-doped layer including relaxation caused by defects

Using the dynamic Monte Carlo (MC) code, ACAT-DIFFUSE, the SIMS depth profiling of a multilayered thin film (Ta₂O₅ (18 nm)/SiO₂ (0.5 nm)) sample was investigated. The ACAT-DIFFUSE code is based on the binary collision approximation, taking into account the generation of interstitial atoms and vacancies, annihilation of vacancies, diffusion of interstitial atoms and primary ions and the relaxation of target materials according to the packing condition which include not only beam and target particles but also defects (interstitial atoms and vacancies). The observed 1–3 nm shift of the delta layer peak to the surface in SIMS depth profiles can be reproduced by the ACAT-DIFFUSE simulation. It is found that this peak shift is mainly due to the relaxation or expansion caused by defects produced behind the delta layer, not due to the collision mixing which results mainly in broadening the observed delta layer peak. Therefore, as ion energy decreases or the angle of incidence becomes large, the peak shift becomes small, because the total amount of defects produced behind the delta layer is small before the delta layer is sputtered off.     

Conceptual design and analysis of a semi-passive containment cooling system for a large concrete containment

An internal evaporator-only (IEO) concept has been developed as a semi-passive containment cooling system for a large dry concrete containment. The function of this system is to keep the containment integrity by maintaining the internal pressure not to exceed ultimate design pressure, i.e. 0.83 MPa (120 psia) in the absence of any other containment cooling following a severe accident, which postulates core damage and hydrogen combustion. The ability of the concept to protect the containment was evaluated for the design basis accident (DBA) large break loss of coolant accident (LB LOCA) and severe accident scenarios (LB LOCA without Emergency Core Cooling System (ECCS) and containment spray flow, 100% zirconium oxidation and complete hydrogen combustion). All were modeled using the GOTHIC computer code. It was concluded that a practical system requiring four IEO loops could be utilized to meet design criteria for severe accident scenarios.