An Experimental Study of Thermal-hydraulic Phenomena in the Downcomer with a Direct Vessel Injection System of APR1400 during the LBLOCA Reflood Phase

The direct vessel injection (DVI) mode is adopted as a safety injection system in the place of a conventional cold leg injection (CLI) mode in the Advanced Power Reactor 1400MW (APR1400). It is expected that “sweep-out” and “direct ECC (Emergency Core Cooling) water bypass” are two most important bypass mechanisms of ECC water injected through the DVI lines during the LBLOCA reflood phase. Using the test facility of plane-channel type scaled down to 1/7 ratio of the prototype reactor (APR1400), we carry out the following tests for the investigation of the two mechanisms: water film spreading test, sweep-out test, and direct ECC water bypass test. From the water film spreading test, it was found that the curvature effect is negligible and the present modified linear scaling law is more appropriate than the linear scaling law. In the sweep-out test, the continuous onset is used to analyze the water height in the downcomer and the amount of ECC water bypass by sweep-out is compared with the previous correlations. The direct ECC water bypass test is performed in order to understand the ECC water film behavior in the downcomer.     

Air-water multidimensional impingement flow on vertical flat wall

The emergency core cooling (ECC) water is supplied from the direct vessel injection (DVI) system in the Advanced Power Reactor 1400 MWe (APR1400) during a postulated large-break loss-of-coolant accident (LBLOCA). The velocity of ECC water exceeds 10 m/s in the early high pressure phase of LBLOCA and then is decreased to 2-3 m/s in the late phase of reflood. During the injection the flow behavior exhibits a complex mode involving impingement, bypass, entrainment, sweepout and condensation in the reactor downcomer. There is currently no model to accurately simulate the local and complicated flow behavior in the APR1400 downcomer during a LBLOCA. This study is aimed at developing models for the water film flow and deformation, both of which are expected to sizably affect the other multidimensional flow characteristics in the downcomer. Experimental studies are conducted to benchmark the predictive model by furnishing the boundary conditions for the analysis resorting to the Accelerated Liquid Phase Hydrodynamics Apparatus (ALPHA) and the Kinetic Aerodynamic Physics Parallelepiped Apparatus (KAPPA). The Poisson equation and potential theory are applied to formulate the behavior of the water film and air flow. In both the experimental and numerical studies, the temperature-dependent thermodynamic properties and the reactor vessel curvature are neglected to render the problem at hand tractable. The model is found to reasonably describe the downward film flow behavior. The water film is developed in proportion to the initial injection velocity of the ECC water. The downward velocity of water film is increased with the heights of injection. Regarding the film deformation the calculated results tend to deviate from the experimental data as the injected air velocity is increased. The disagreement is attributed to limitations inherent in the two-dimensional treatment and point source approach.     

反应堆压力容器下降段水-蒸汽CCFL实验与模型研究

为探究反应堆压力容器下降段在喷放末期冷段安注过程中的水-蒸汽逆流特性,建立下降段逆向流动限制(CCFL)模型,开展了基于压力容器模化本体的下降段CCFL实验研究以及建模分析。通过实验研究获得了不同入口安注水流量、安注水过冷度、堆芯蒸汽流量等条件下的下降段环腔内的安注特性数据,并基于实验数据进行了CCFL建模分析。结果表明,开始发生CCFL的蒸汽无量纲流速与入口安注水无量纲流速呈现正相关,基于无量纲流速建立的模型斜率与入口安注水无量纲流速呈现高度指数关联。本文建立了适用于从不发生CCFL至不完全CCFL,再到完全CCFL的下降段水-蒸汽气液逆流全过程预测模型。     

Direct vessel inclined injection system for reduction of emergency core coolant direct bypass in advanced reactors

Multidimensional thermal hydraulics in the APR1400 (Advanced Power Reactor 1400 MWe) downcomer during a large-break loss-of-coolant accident (LBLOCA) plays a pivotal role in determining the capability of the safety injection system. APR1400 adopts the direct vessel injection (DVI) method for more effective core penetration of the emergency core cooling (ECC) water than the cold leg injection (CLI) method in the OPR1000 (Optimized Power Reactor 1000 MWe). The DVI method turned out to be prone to occasionally lack in efficacious delivery of ECC to the reactor core during the reflood phase of a LBLOCA, however. This study intends to demonstrate a direct vessel inclined injection (DVII) method, one of various ideas with which to maximize the ECC core penetration and to minimize the direct bypass through the break during the reflood phase of a LBLOCA. The 1/7 scaled down THETA (Transient Hydrodynamics Engineering Test Apparatus) tests show that a vertical inclined nozzle angle of the DVII system increases the downward momentum of the injected ECC water by reducing the degree of impingement on the reactor downcomer, whereby lessening the extent of the direct bypass through the break. The proposed method may be combined with other innovative measures with which to ensure an enough thermal margin in the core during the course of a LBLOCA in APR1400.     

Scaling for the ECC bypass phenomena during the LBLOCA reflood phase

As one of the advanced design features of the APR1400 (Advanced Power Reactor), a direct vessel injection (DVI) system is adopted instead of the conventional cold leg injection (CLI) system. It is known that the DVI system greatly enhances the reliability of the emergency core cooling (ECC) system. However, there is still a dispute on its performance in terms of water delivery to the reactor core during the reflood period of a large-break loss-of-coolant accident (LOCA). Thus, experimental validation is underway. In this paper, a new scaling method, using the time and velocity reduced “modified linear scaling law”, is suggested for the design of a scaled-down experimental facility to investigate the direct ECC bypass phenomena in the PWR downcomer.     

Effect of the yaw injection angle on the ECC bypass in comparison with the horizontal DVI

The comparison tests for the direct emergency core cooling (ECC) bypass fraction were experimentally performed with a typical direct vessel injection (DVI) nozzle and an ECC column nozzle having a yaw injection angle to the gravity axis. The ECC yaw injection nozzle is newly introduced to make an ECC water column in the downcomer region. The yaw injection angle of the ECC water relative to the gravity axis is varied from 0 to (±)90° stepped by 45°. The tests are performed in the air–water separate effect test facility (direct injection visualization and analysis (DIVA)), which is a 1/7.07 linearly scaled-down model of the APR1400 nuclear reactor. The test results show that (1) if the ECC water column is injected into the wake region which is induced by the hot leg blunt body in the downcomer annulus, the ECC bypass fraction is greatly reduced compared with the typical horizontal ECC injection which makes ECC film on the downcomer wall. At the same time, the ECC penetration toward the lower downcomer region becomes larger than those of a typical horizontal type of direct vessel injection on the downcomer wall vertically. (2) If the ECC water column is injected near the broken cold leg, the ECC water is directly bypassed. Thus, the ECC penetration fraction is greatly reduced compared with a typical film type of the horizontal ECC injection. (3) In order to minimize the ECC bypass fraction, the ECC water should be injected toward the wake region of the hot leg blunt bodies.     

CAP1400反应堆压力容器下降段内气-液逆向流试验研究

为研究真实工况下CAP1400反应堆压力容器下降段气-液逆向流现象,以CAP1400为原型,搭建压力容器下降段高度和直径比为1:1、60°切片的试验台架。试验工质为空气和水,试验研究了不同安注(DVI)供水量、不同气量的气-液两相流动和应急堆芯冷却剂(ECC)旁通现象。试验结果表明,DVI供水量相同时,随着供气量的增加,气-液逆向流现象明显,当质量流速达到4kg/s及以上时,安注水不能全部进入堆芯;Kutateladze经验关系式和UPTF经验关系式都与试验结果存在较大偏差,不适用于CAP1400压力容器下降段试验;基于试验数据,拟合了新的经验关系式,且通过比较有无DVI挡块的试验数据,验证了DVI挡块可以降低ECC旁通水量,增强安注能力。     

Sweepout model implementation in RELAP5/MOD3.3 to improve RCS coolant inventory calculation during a LBLOCA

According to the experiments of the Upper Plenum Test Facility (UPTF) and advanced power reactor 1400 MW_e (APR1400), the sweepout in the downcomer has been identified to play an important role in depleting the core coolant inventory during a Large-Break Loss-of-Coolant Accident (LBLOCA). In order to identify the sweepout mechanism and to estimate the amount of coolant discharged by sweepout, the separate-effect test was carried out in the plate type test apparatus, which was scaled down to 1/5 of the size of the APR1400 downcomer. In addition, the sweepout model was developed by correlating the experimental data on the critical void height and the discharge flow rate at the break to the values of analytically derived non-dimensional parameters. This model was implemented in RELAP5/MOD3.3 to improve its calculation of coolant inventory loss during a LBLOCA. To validate the modified RELAP5/MOD3.3 by implementing the sweepout model, the sweepout separate-effect test was simulated by both the original and the modified RELAP5/MOD3.3. The original one predicted the different discharge flow rates according to the node size of the donor volume, and these flow rates were larger than those of the experiment. On the other hand, the modified one calculated the discharge flow rate and the critical void height much more similar to those of the experiment than the original model did. In the future, the improved RELAP5/MOD3.3 adopted in an integrated analysis system will support a more realistic thermal hydraulic analysis.     

Experimental validation of the modified linear scaling methodology for scaling ECC bypass phenomena in DVI downcomer

In the direct vessel injection (DVI) system downcomer, the direct emergency core coolant (ECC) bypass is activated during the reflood phase of a large-break loss-of-coolant accident (LBLOCA) by the interaction between the downward-flowing liquid-film and the transverse gas flow. Direct ECC bypass is reportedly the major bypass mechanism of ECC, and various experiments have been performed to obtain detailed information about the ECC bypass in a DVI downcomer. These lead to a proposed new scaling methodology, named ‘modified linear scaling’, which is expected to preserve the phase distribution in the downcomer and the ECC bypass phenomena. In the present study, modified linear scaling was experimentally validated in air–water tests comprising Test 21-D of the upper plenum test facility (UPTF). The counterpart tests of UPTF Test 21-D were performed in 1/7.3 and 1/4.0 scale models of a UPTF downcomer, and the test results were compared with the experimental data of UPTF Test 21-D. Two important parameters of direct ECC bypass – the normalized liquid-spreading width on the downcomer wall and the direct ECC bypass fraction, which is the fraction of input water bypassed out the broken cold-leg – were considered in the validation. The comparison revealed that the scaling parameters of direct ECC bypass are well preserved in the prototype and reduced models, from which we conclude that the modified linear scaling methodology is appropriate for designing a reduced test facility and for a scaling analysis of direct ECC bypass in the reflood phase of an LBLOCA.     

Numerical analysis and visualization experiment on behavior of borated water during MSLB with RCP running mode in an advanced reactor

The core bypass phenomenon of borated water injected through direct vessel injection (DVI) nozzles in APR1400 (Advanced Power Reactor 1400MWe) during main steam line break (MSLB) accidents with a reactor coolant pump (RCP) running mode has been simulated using a two-channel and one-dimensional system analysis model code (MARS), and a three-dimensional computational fluid dynamics (CFD) code (FLUENT). A visualization experiment has also been performed using a scaled-down model of the APR1400. The MARS analysis has predicted a serious core bypass phenomenon of borated water, while the CFD analysis has shown results opposite to the MARS results. The CFD analysis has shown that the flow pattern in the downcomer is fully three-dimensional and that vortex flow structures are formed near the cold legs so that the borated water might pass without difficulty into the high flow region of the cold legs and flow well into the lower downcomer. The visualization experiment has shown that the borated water flows well to the lower plenum, as in the CFD analysis. Both the CFD analysis and visualization experiment have proved that a serious core bypass phenomenon of borated water might not happen in the APR1400. These results are quite different from those predicted by MARS.     

Two-phase natural circulation flow of air and water in a reactor cavity model under an external vessel cooling during a severe accident

As a part of a study on a two-phase natural circulation flow between the outer reactor vessel and the insulation material in the reactor cavity under an external reactor vessel cooling of the Advanced Power Reactor (APR) 1400, a Hydraulic Evaluation of Reactor cooling Mechanism by External Self-induced flow-HALF scale (HERMES-HALF) experiment has been performed by using the non-heating method of an air injection. This large-scale experiment uses a half-height and half-sector model of the APR1400. This experiment has been analyzed to verify and evaluate the experimental results by using the RELAP5/MOD3 computer code. The RELAP5/MOD3 results have shown that the water circulation mass flow rate is very similar to the experimental results of the HERMES-HALF, in general. Increases in the water inlet area and the water level in the reactor cavity lead to an increase in the water circulation mass flow rate. The effects of an air injection mass flow rate and the water outlet area on the water circulation mass flow rate are dependent on the water inlet area size. As the water outlet moves to a lower position, the water circulation mass flow rate increases slowly.     

Major findings from LBLOCA reflood tests using the ATLAS facility

A set of LBLOCA (large-break loss of coolant accident) reflood tests was performed in the first phase of the ATLAS (advanced thermal-hydraulic test loop for accident simulation) program. Their main objectives were to identify the major thermal-hydraulic characteristics during the reflood phase of a LBLOCA for APR1400 and to provide qualified data for APR1400 licensing. The ATLAS reflood test program could be divided into two phases (Phase-1 and Phase-2) according to the target period to be simulated. The Phase-1 tests were parametric effect tests for downcomer boiling in the late reflood phase of LBLOCA and the Phase-2 tests were integral effect tests for the entire reflood phase of LBLOCA. The experimental results from both Phase-1 and Phase-2 tests reproduced typical thermal-hydraulic trends expected to occur during the APR1400 LBLOCA scenario. A separate effect test was also performed under a low reflooding rate condition to provide data to validate the RELAP5 reflood models, and its experimental results showed a gradual reflooding in the core, a subsequent quenching of the core heater rods and the cooling of the reactor pressure vessel downcomer.     

Analytical evaluation of two-phase natural circulation flow characteristics under external reactor vessel cooling

This work proposes an analytical method of evaluating the effects of design and operating parameters on the low-pressure two-phase natural circulation flow through the annular shaped gap at the reactor vessel exterior surface heated by corium (molten core) relocated to the reactor vessel lower plenum after loss of coolant accidents. A natural circulation flow velocity equation derived from steady-state mass, momentum, and energy conservation equations for homogeneous two-phase flow is numerically solved for the core melting conditions of the APR1400 reactor. The solution is compared with existing experiments which measured natural circulation flow through the annular gap slice model. Two kinds of parameters are considered for this analytical method. One is the thermal–hydraulic conditions such as thermal power of corium, pressure and inlet subcooling. The others are those for the thermal insulation system design for the purpose of providing natural circulation flow path outside the reactor vessel: inlet flow area, annular gap clearance and system resistance. A computer program NCIRC is developed for the numerical solution of the implicit flow velocity equation.     

Advanced DVI for ECC direct bypass mitigation

An ECC direct bypass fraction during a late reflood phase of a LBLOCA is strongly dependent on the characteristics of the cross flow and the geometrical configuration of a DVI in the downcomer of a pressurized light water reactor. The important design parameters of a DVI are the elevation, the azimuthal angle, and the separator to prevent a steam-water interaction. An ECC sub-channel to separate or to isolate an ECC water from a high-speed cross flow is one of the important design features to mitigate the ECC bypass phenomena. A dual core barrel cylinder as an ECC flow separator is located between a reactor vessel and a core barrel outer wall in the downcomer annulus. A new narrow gap between the core barrel and the additional dual core barrel plays the role of a downward ECC flow channel or an ECC flow separator in a high-speed cross flow field of the downcomer annulus. The flow zone around a broken cold leg in the downcomer annulus has the role of a high ECC direct bypass due to a strong suction force while the wake zone of a hot leg has the role of an ECC penetration. Thus, the relative azimuthal angle of the DVI nozzle from the broken cold leg is an important design parameter. A large azimuthal angle from a cold leg to a hot leg needs to avoid a high suction flow zone when an ECC water is being injected. The other enhancing mechanism of an ECC penetration is a grooved core barrel which has small rectangular-shaped grooves vertically arranged on the core barrel wall of the reactor vessel downcomer annulus. These grooves have the role for a generation of a vortex induced by a high-speed cross flow. Since the stagnant flow in a lateral direction and rotational vortex provides the pulling force of an ECC drop or film to flow down into the lower downcomer annulus by gravity, the ECC direct bypass fraction is reduced when compared to the current design of a smoothed wall. An open channel of grooves generates a stagnant vortex, while a closed channel of grooves creates an isolated ECC downward flow channel from a high-speed lateral flow. In this study, new design concepts for a dual core barrel cylinder, grooved core barrel, and a reallocation of the DVI azimuthal angle are proposed and tested by using an air-water 1/5 scaled air-water test facility. The ECC direct bypass reduction performances of the new design concepts have been compared with that of the standard type of a DVI injection. The azimuthal angle of the DVI nozzle from a broken cold leg varies from −15° to +52° toward a hot leg. The test results show that the azimuthal injection angle is an effective parameter to reduce the ECC direct bypass fraction. The elevation of the DVI nozzle is also an important parameter to reduce the ECC direct bypass fraction. The most effective design for reducing the ECC direct bypass fraction is a dual core barrel. The reduction fraction when compared to the standard DVI is about −30% for the dual core barrel while it is −15% for the grooved core barrel.     

Investigation of Sweepout Mechanism and Critical Void Height in Annular Downcomer

The APR1400 (Advanced Power Reactor 1,400 MWe) has adopted the direct vessel injection (DVI) in lieu of the conventional cold leg injection for its emergency core cooling system (ECCS). In this reactor, sweepout from the water surface by gas (vapor or air) flow plays an important role in analyzing the mass and momentum transfer in the reactor downcomer of multidimensional geometry during a loss-of-coolant accident (LOCA) by decreasing the water level in the downcomer. The core water level will tend to decrease rapidly if a considerable amount of the entrained water stream and droplets bypasses through the break. The amount of entrained water is mostly determined by the interacting gas flow rate, the geometric condition, and the interfacial area between the gas and the water. The sweepout is observed to take place in three rather distinct steps: the beginning of undulation, the full wave and the wave peak (droplet separation). In view of these observations we investigated the relation between the gas flow rate and the amount of bypass as a function of time. The current experimental results shed light on the flow mechanism and the semi-empirical relations for the three-dimensional sweepout in a large-diameter annulus such as the reactor downcomer. A physico-numerical model is being developed to predict the multidimensional bypass flow rate resulting from the sweepout and entrainment in the downcomer.     

MELCOR 1.8.4 sensitivity analysis of the severe accident evolution during the APR 1400 LOCA

Unlike most other systems in which the emergency core cooling (ECC) water is injected into the cold-legs, the Advanced Power Reactor (APR) 1400 employs a concept of a direct vessel injection (DVI) to reduce the bypass effects of the ECC water via a break during a design basis LOCA. For this, the DVI piping is designed so that the ECC water taken from an in-containment refueling storage tank (IRWST) directly flows into the reactor pressure vessel (RPV) down-comer. The main objective of this paper is to provide the MELCOR 1.8.4 sensitivity analysis results on the evolution of the severe accidents that can be expected during the APR 1400 LOCA and the insights gained from the analysis. For this purpose, the present sensitivity analysis mainly focuses on: (1) the impact of the foregoing engineering features (i.e., DVI and IRWST) in mitigating a severe core degradation and (2) the APR 1400-specific impacts of different break locations and sizes, and an operation of the containment spray systems on the timings of the key thermal-hydraulic responses, the severe degradation of the core, and the evolution of the core materials. No significant accident management strategy that plays a great role in mitigating a further progression of severe accidents has been taken into account in the present analysis. As a result, the present analysis results can be taken as the technical basis for assessing the effectiveness of a potential severe accident management.     

Scaling methodology for a reduced-height reduced-pressure integral test facility to investigate direct vessel injection line break SBLOCA

A scaling methodology for a small-scale integral test facility was investigated in order to analyze thermal-hydraulic phenomena during a DVI (direct vessel injection) line SBLOCA (small break loss-of-coolant accident) in an APR1400 (advanced power reactor 1400 MWe) pressurized water reactor. The test facility SNUF (Seoul National University Facility) was utilized as a reduced-height and reduced-pressure integral test loop. To determine suitable test conditions for simulating the prototype in the SNUF experiment, the energy scaling methodology was propose to scale the coolant mass inventory and the thermal power for a reduced-pressure condition. The energy scaling methodology was validated with a system code (MARS) analysis for an ideally scaled-down SNUF model and that predicted a reasonable transient of pressure and coolant inventory when compared to the prototype model. For the actually constructed SNUF, the effect of scaling distortions in the test facility's thermal power and the loop geometry was analytically investigated. To overcome the limitation of the thermal power supply in the facility, the convective heat transfer between primary and secondary systems at the steam generator U-tubes was excluded and a modified power curve was applied for simulating the core decay heat. From the code analysis results for the actual SNUF model, the application of the modified power curve did not affect the major events occurring during the transient condition. The results revealed that the scaling distortion in the actual SNUF geometry also did not strongly disturb significant thermal-hydraulic phenomena such as the downcomer seal clearing. Thus, with an adoption of the energy scaling methodology, the thermal-hydraulic phenomena observed in the SNUF experiment can be properly utilized in a safety analysis for a DVI line break SBLOCA in the APR1400.     

Commissioning of the ATLAS thermal-hydraulic integral test facility

KAERI recently constructed a new thermal-hydraulic integral test facility for advanced pressurized water reactors (PWRs) – ATLAS. The ATLAS facility has the following characteristics: (a) 1/2-height&length, 1/288-volume, and full pressure simulation of APR1400, (b) maintaining a geometrical similarity with APR1400 including 2(hot legs) × 4(cold legs) reactor coolant loops, direct vessel injection (DVI) of emergency core cooling water, integrated annular downcomer, etc., (c) incorporation of specific design characteristics of OPR1000 such as cold leg injection and low-pressure safety injection pumps, (d) maximum 10% of the scaled nominal core power. The ATLAS will mainly be used to simulate various accident and transient scenarios for evolutionary PWRs, OPR1000 and APR1400: the simulation capability of broad scenarios including the reflood phase of a large-break loss-of-coolant accident (LOCA), small-break LOCA scenarios including DVI line breaks, a steam generator tube rupture, a main steam line break, a feed line break, a mid-loop operation, etc. The ATLAS is now in operation after an extensive series of commissioning tests in 2006.     

Downcomer and tie plate countercurrent flow in the Upper Plenum Test Facility (UPTF)

The limitation of vertical steam-water countercurrent flow, called flooding, is important for the operation of Emergency Core Cooling (ECC) Systems in Nuclear Reactors. The ECC water injection flooding behavior is scale dependent and the reactor size behavior cannot be extrapolated from small scale data.A new flooding correlation is presented, based on the classic flooding equation where the effects of gravity, interphase momentum exchange, and instability of the gas/liquid interface are considered. Development of a new correlation was necessary in order to correlate the reactor scale downcomer and upper tie plate countercurrent flow data gained in the Upper Plenum Test Facility (UPTF). The new correlation is an extension of the well-known Wallis-type and Kutateladze-type correlations. Each of the three correlations is valid for the experimental facilities on a certain scale. The range of applicability for each of the three correlations is defined for the case of downcomer and tie plate countercurrent flow.     

Experimental identification for flow distribution inside APR+ reactor vessel and direction of internal structure design improvement

A test program to quantify the reactor flow distribution has been performed using a test facility, named ACOP, having a 1/5 length scale referring to the APR+ reactor design. The flow characteristics of the prototype plant could be preserved by designing the test facility by adopting a linear reduced scaling principle. An Euler number is considered as a primary dimensionless parameter, which was preserved with a 1/41.0 Reynolds number scaling ratio based on the balanced flow conditions. The important measuring parameters are the core inlet flow, outlet pressure distribution, and sectional pressure drops along the major flow path inside the reactor vessel as well as static pressure and temperature at the vessel and boundary legs. The reactor flow distribution is identified by a series of three reactor flow balancing conditions: (1) balanced cold leg flow condition (2) 5% unbalanced cold leg flow condition, and (3) extreme unbalanced flow condition under the assumption of a single pump failure. This paper describes the design features for the test facility and the measuring method, and summarizes the reactor flow and pressure characteristics by ensemble averaging for each group of tests.