Masanori Naitoh

Development of severe accident analysis code SAMPSON in IMPACT project

IMPACT is the name of a program and of specific simulation software, which will perform full-scope and detailed calculations of various phenomena in a nuclear power plant for a wide range of event scenarios. The four years of the IMPACT project Phase 1 have been completed, and each analysis module of the prototype version of the severe accident analysis code SAMPSON has been developed and verified by comparison with separate-effect test data. Verification of the integrated code with combinations of up to 11 analysis modules has been conducted, with the Analysis Control Module, to demonstrate the code capability and integrity. A 10-inch cold leg failure Loss of Coolant Accident in the Surry Plant was the assumed initiating event. The system analysis was divided into two cases; one was an in-vessel retention analysis when gap cooling was effective, the other was an analysis of phenomena when the event was extended to ex-vessel due to the reactor pressure vessel failure when gap cooling was not sufficient. U...

Critical Power Analysis with Mechanistic Models for Nuclear Fuel Bundles, (I) Models and Verifications for Boiling Water Reactor Application

The critical power analysis code for BWR fuel bundles, “CAPE-BWR”, was developed. The objective of the development is to predict dryout phenomena of liquid film on fuel rod surfaces without tuning any parameters even for fuel bundle design improvements. The major features of the code are modular structure with mechanistic models and parallel computation. The calculation methods were divided into three steps: subchannel, liquid film flow and spacer effect analyses. The code was validated by the rod bundle test analyses. The overall comparison of calculated critical power with 166 measured data points showed -0.3% average difference with the standard deviation of 6.3%. The spatial domain decomposition method was applied for parallel computation of the spacer effect analysis. The parallelization efficiency was about 80%. The calculated dryout location agreed well with the measured one at the full-scale 8×8 bundle test. The code could trace the tendencies of the critical power depending on power distribution, spacer geometry and fluid conditions within a practical range of difference. From the calculation, difference of the critical power due to the spacer geometry was clarified to be caused by the difference of droplet deposition characteristics onto the liquid film.

Evaluation Methods for Corrosion Damage of Components in Cooling Systems of Nuclear Power Plants by Coupling Analysis of Corrosion and Flow Dynamics (III) : Evaluation of Wall Thinning Rate with the Coupled Model of Static Electrochemical Analysis and Dynamic Double Oxide Layer Analysis

Flow accelerated corrosion (FAC) is divided into two processes: a corrosion (chemical) process and a flow dynamics (physical) process. The former is the essential process to cause FAC and the latter is the accelerating process to enhance FAC occurrence. The chemical process in the surface boundary layer is analyzed to evaluate FAC rate. Contributions of flow dynamics on wall thinning rate due to FAC are expressed as a function of mass transfer coefficient but not that of flow velocity. FAC evaluation procedures were divided into 5 steps as follows. (1) Flow pattern and temperature in each elemental volume along the flow path were obtained with 1D computational flow dynamics (CFD) codes, (2) corrosive conditions, e.g., oxygen concentration and electrochemical corrosion potential (ECP) along the flow path were calculated with a hydrazine oxygen reaction code, (3) precise flow patterns and mass transfer coefficients at the structure surface were calculated with 3D CFD codes, (4) danger zones were evaluated by coupling major FAC parameters, and then, (5) wall thinning rates were calculated with the coupled model of static electrochemical analysis and dynamic double oxide layer analysis at the identified danger zone. Anodic and cathodic current densities and ECPs were calculated with the static electrochemistry model and ferrous ion release rate determined by the anodic current density was used as input for the dynamic double oxide layer model. Thickness of the oxide film and its characteristics determined by the dynamic double oxide layer model were used for the electrochemistry model to determine the resistances of cathodic current from the bulk to the surface and anodic current from the surface to the bulk. Two models were coupled to determine local corrosion rate and ECP for various corrosive conditions. The calculated results of the coupled models had good agreement with the measured ones.

Evaluation Methods for Corrosion Damage of Components in Cooling Systems of Nuclear Power Plants by Coupling Analysis of Corrosion and Flow Dynamics (II): Evaluation of Corrosive Conditions in PWR Secondary Cooling System

Flow accelerated corrosion (FAC) is divided into two processes: a corrosion (chemical)process and a flow dynamics (physical) process. The former is the essential process to cause FAC and the latter is the accelerating process to enhance FAC occurrence. The chemical process in the surface boundary layer can be analyzed to evaluate FAC rate. In this paper, corrosive conditions along the flow path of the PWR secondary cooling system were evaluated. To do this, flow velocity and temperature in each elemental volume along the flow path were obtained with 1D computational flow dynamics (CFD) codes, distribution of oxygen concentration along the flow path was calculated with a oxygen hydrazine reaction code, and then electrochemical corrosion potential (ECP) was evaluated by using the Evans diagram. In the proposed calculation procedures for corrosive conditions, the oxygen hydrazine reactions were divided into bulk and surface reactions and the oxidation reaction of hydrazine on the surface was considered to obtain ECP under hydrazine coexisting conditions. Calculations of precise flow patterns and mass transfer coefficients at the structure surface made with 3D CFD codes and calculations of wall thinning rates made with the coupled model of static electrochemical analysis and dynamic double oxide layer analysis agreed with the calculations of corrosive conditions to evaluate FAC rate.

Evaluation Methods for Corrosion Damage of Components in Cooling Systems of Nuclear Power Plants by Coupling Analysis of Corrosion and Flow Dynamics (I) : Major Targets and Development Strategies of the Evaluation Methods

Problems in major components and structural materials in nuclear power plants have often been caused by flow induced vibration and corrosion and their overlapping effects. In order to establish safe and reliable plant operation, future problems for structural materials should be predicted based on combined analyses of flow dynamics and corrosion and they should be mitigated before becoming serious issues for plant operation. Three approaches have been prepared for predicting future problems in structural materials: 1. Computer program packages for predicting future corrosion fatigue on structural materials, 2. Computer program packages for predicting future corrosion damage on structural materials, and 3. Computer program packages for predicting wall thinning caused by flow accelerated corrosion. General features of evaluation methods and their computer packages, technical innovations required for their development, and application plans for the developed approaches for plant operation are introduced in this paper.

BWR Loss of Coolant Integral Tests with Two Bundle Loop, (I) Thermal-Hydraulic Characteristics in Parallel Channels

Two simulation tests of a BWR loss-of-coolant accident (LOCA) by a postulated guillotine rupture of a recirculation suction line were conducted using the Two Bundle Loop (TBL), which was volumetrically scaled to a BWR/5 plant with 764 fuel bundles. The major objective of the tests was to clarify thermal-hydraulic difference in parallel bundles. In the tests, the failure of a diesel generator for two low pressure coolant injection (LPCI) pumps was assumed, and the initial bundle power combinations were 4.0 and 5.9 MW in the first test, and 5.0 and 4.9 MW in the second. In one of the two bundles, the rods heated up locally in the radial direction. In the other, the rods heated up rather uniformly and later than in the former bundle. Much water fell locally into the former bundle, while ascending steam flow from the lower plenum was larger in the latter. A difference in thermal-hydraulic responses was observed even in the case of nearly identical bundle powers, but the difference was less than in the case of...

Verification and Validation of Evaluation Procedures for Local Wall Thinning Due to Flow-Accelerated Corrosion and Liquid Droplet Impingement

A six-step evaluation procedures have been proposed to evaluate the local wall thinning due to flow-accelerated corrosion (FAC) and that due to liquid droplet impingement (LDI). Corrosive conditions were calculated with a N2H4-O2 reaction analysis code. Precise flow turbulence at major parts of the system was analyzed with the three-dimensional computational flow dynamics code to obtain mass transfer coefficients at structure surfaces. Then, wall thinning rates were calculated with the coupled model of electrochemical analysis and oxide layer growth analysis by applying the corrosive conditions and the mass transfer coefficients. To apply computer simulation codes for wall thinning due to FAC and LDI to evaluate residual life and the effectiveness of countermeasures, accuracy and applicability of the codes were confirmed based on verification and validation processes. From comparison of the calculated wall thinning rates due to FAC with hundreds of measured results for secondary piping of an actual pressurized water reactor plant, it was confirmed that the calculated wall thinning rates agreed with the measured ones within a factor of 2 and the accuracy of the evaluation model for residual pipe wall thickness after 1 year of operation had an error of <20%. Finally, just the FAC simulation code was applied to evaluate the effects of oxygen injection into the feedwater line. From comparison of the calculated wall thinning rates due to LDI with measured results for vent lines of an actual boiling water reactor plant, it was confirmed that the calculated local wall thinning rates agreed with the measured ones within about a factor of 2, though there were still some outside that region.

Experiments on Convection Heat Transfer Along a Vertical Flat Plate Between Pools with Different Temperatures

This paper reports that to evaluate the heat removal capability of an external water wall-type containment vessel, which is a passive system for containment cooling, thermal-hydraulic behavior in the suppression and outer pools has been examined experimentally. The following results are obtained: A thermal stratification boundary, which separates the pools into an upper high-temperature region and a lower low-temperature region, is observed just below the vent outlet. The natural-convection heat transfer coefficients (HTCs) for the downward and upward flows that appear inside and outside the primary containment vessel wall are measured. The condensation HTCs in the presence of non-condensable gas, which affect heat transfer between the wet well and the outer pool, are measured along the long wall. The capability for decay heat removal in the external water wall-type containment vessel for a 600-MW (electric) plant is evaluated based on these results and is found to be large enough.

Analysis on Pipe Rupture of Steam Condensing Line at Hamaoka-1, (I) Accumulation of Non-condensable Gas in a Pipe

A pipe rupture occurred in the steam condensing line of the residual heat removal system at the Hamaoka Nuclear Power Station Unit-1 on November 7, 2001. The detonation of the hydrogen accumulated in the pipe was considered to be the likeliest cause of the rupture. As the first step of the accident analysis, fluid behavior in the line was analyzed in order to give boundary conditions for hydrogen combustion analysis. From the analysis, it was concluded that: (1) Condensed water filled up the hollowed part at the downstream end of the line within 4 days after the start-up of the cycle operation. (2) The temperature of the accumulated water became almost the ambient temperature at 14 days after the start-up. (3) Hydrogen and oxygen, generated by radiolysis of reactor water, accumulated at the downstream end of the line for a length of 6.9 m from the accumulated water surface, and concentration and temperature were distributed along the accumulated region. (4) At the boundary between the accumulated non-condensable gas and the steam regions, temperature and concentration fluctuated due to operation of the HPCI valve. The fluctuating region was limited within 200–300 mm upstream and downstream across the boundary.

Analysis on Pipe Rupture of Steam Condensation Line at Hamaoka-1, (II): Hydrogen Combustion and Pipe Deformation

A pipe rupture occurred in the steam condensing line of the residual heat removal system at the Hamaoka Nuclear Power Station Unit-1 on November 7, 2001. As the accident analysis, the three-dimensional hydrogen combustion behavior was first solved with the one-step irreversible overall reaction model, the Magnussen eddy dissipation model and the SIMPLE method. The ignition point was given at the upstream boundary surface of the non-condensable gas region. The temperature and concentration distributions in the pipe, which were obtained from the analysis of the non-condensable gas accumulation, were given as the initial conditions. The analysis result showed that the detonation pressures in the straight pipes were about 120 MPa and the peak pressures at the elbows were 2.0–2.5 times higher than those in the straight pipes, due to reflection and overlapping of the pressure waves. Then, a three-dimensional dynamic response of pipe deformation was analyzed, with the time transients of the pressure distribution as boundary conditions inside the pipes. The result showed that the strain at the elbow above the surface of the accumulated water exceeded the critical strain and the pipe ruptured there. This was generally consistent with the result of actual pipe deformation observed after the accident.