MultiPin Coupled Reflood Model

1
Multi-Pin Coupled Reflood Model

• Analysis and Assessment of the TRACE Code Reflood
  Model

2
OUTLINES

• Introduction
• Problem
• Physics of the Problem/ Background
• Objectives
• How to reach our goals?
• Thermal Hydraulics Goal Of The Work
• A Thermal Hydraulic Code TRACE code Overview
• TRACE Assessment the ACHILLES experiment
• Recent Work and Conclusion

3
INTRODUCTION

• Nuclear Reactor Safety Analysis
• Loss Of Coolant Accident (LOCA)
• Reference Design Basis Accident (DBA).
• Possible phenomena due to a Large-Break LOCA
• Clad deformation and burst
• Flow blockage of the fluid section
• Zircaloy cladding oxidation

4
INTRODUCTION
FIG.1 2-loop PWR system scheme cold leg LB LOCA
5
INTRODUCTION
FIG.2 Blockage caused by the ballooning
(ADMIRABILE 2003)
6
Problem

• PROBLEM
• Cladding of adjacent ballooning pins may touch,
  preventing access to the pins by the re-flooding
  coolant.
• The deformations of the pins change the coolant
  flow passages, and this changing of the cooling
  conditions modifies the subsequent mechanical
  response of the pins.

7
Physics of the Problem/ Background

• Physics of the Problem
• Two-Phase flow (droplet field modelling, flow
  diversion),
• Heat transfer,
• Mechanics (clad deformation),
• Fluid Structure Interaction
• LB-LOCA 4 main phases
• Blow Down
• Refill
• Reflood
• Long Term Cooling

8
Physics of the Problem/ Background

• Fuel Pin

FIG.3 Fuel pins are pressurised with helium to
maximise heat transfer and prevent creep buckling
9
Physics of the Problem/ Background
FIG.4 Clad ballooning and burst (Dr. JONES,
British Energy)
10
Objectives

• OBJECTIVES
• Prediction of Clad Ballooning Phenomenon during
  Reflood in PWRs.
• Fuel pins failure criteria to be respected.
• A peak clad temperature lt2200F (1204C),
• A maximum local clad oxidation of 17 of the clad
  thickness,
• A maximum hydrogen generation of no more than 1
  of the total amount that could be generated by
  clad oxidation,
• The maintenance of a coolable geometry,
• The maintenance of long term cooling.

11
How to reach our goals?

• To predict the clad ballooning, we will run
  simulations of a realistic LB-LOCA.
• By coupling 2 codes
• A thermal hydraulic code TRACE
• A structural mechanics code MABEL
• First, we will uncouple the problem and focus on
  the reflooding Assessment of the TRACE code,
  The ACHILLES Experiment.

12
Thermal Hydraulics Goal Of The Work

• Validation of the Reflood Model of TRACE v. 4.160
  through a detailed comparison between the codes
  predictions and experimental data
• STEPS
• TRACE Code Understanding
• Reflood Model Analysis
• Experimental Data Acquisition
• Modelling and Comparison
• Conclusions and Future Work Planning

13
TRACE Code Overview

• TRACE TRAC/RELAP Advanced Computational Engine
• Modernized thermal-hydraulics code designed to
  consolidate the capabilities of USNRC's three
  legacy safety codes - TRAC-P, TRAC-B and RELAP
• Productivity-enhancing graphical analysis
  environment (SNAP)
• Component-oriented reactor systems analysis code
  designed to analyze reactor transients and
  accidents up to the point of significant fuel
  damage.
• Finite-volume, two-fluid compressible flow code
  with one, two, and three dimensional flow
  geometry. Can model heat structures and control
  systems that interact with component models and
  the fluid solution.
• Typical TRACE reactor models range in size from a
  few hundred to a few thousand fluid volumes.
  TRACE can be run in parallel.
• TRACE can be run in a coupled mode with the PARCS
  three-dimensional reactor kinetics code.
• Exterior Communications Interface (ECI). TRACE
  could be coupled to detailed fuel models or CFD
  codes in the future using the ECI.

14
TRACE Code Overview
15
TRACE Code Overview
16
The REFLOOD Model

• TRACE contains specific flow and heat-transfer
  regime maps (Ishii Model) for reflood conditions.
  Currently reflood maps are turned on in VESSEL
  component core by a trip.
• Correlations/Models for reflood regimes are
  semi-empirical equations (not first principles).
  
• The Reflood model is applied to the definition
  of
• Interfacial Area
• Interfacial Heat and Mass Transfer
• Interfacial Drag
• Wall Drag
• Wall Heat Transfer
• The assessment is needed to verify the
  applicability of the whole model and of single
  regimes models to multi-pin subchannel
  geometries.

17
ACHILLES Reflood Experiment

• A series of reflooding experiments on a model PWR
  fuel assembly in the ACHILLES Rig at Winfrith
  Technology Centre (ACHILLES programme,CEGB,1989-19
  91) Sizewell B NPP Pre-Construction Safety
  Report.
• Aim study of heat transfer in the core of a PWR
  during the reflood phase of a loss of coolant
  accident.
• Unballooned and Ballooned series of experiments
  under the same reflooding conditions.

• 69 electrically-heated pins (Westinghouse)
• Shroud vessel heaters
• PWR core heated length (3.66m)
• Post-LBLOCA initial conditions
• Chopped-cosine power axial-profile
• Spacer-grids
• Fully instrumented

18
TRACE Model for ACHILLES Unballooned test

• Hydro Components
• 1 VESSEL - ACHILLES assembly
• 3 FILLS - Refllood water BCs
• 3 BREAKS - Pressure BCs
• 3 PIPES - Lower Plenum
• 3 PIPES - Upper Plenum

• Heat Structures
• 3 CYLINDRICAL HSs - Average pins
• 3 SLAB HSs - Shroud vessel heaters

• Control Procedures
• Reflood Trip - Reflood Model onset flag
• Edge Signal - Reflood Rate control

SNAP View of TRACE ACHILLES Model
19
TRACE Model for ACHILLES Unballooned test
20
Assessment
Cladding Temperature Profile

• TRACE predictions are compared to ACHILLES
  experimental data.
• Vertical grid lines show spacer-grids positions.
• the Black line shows experimental results
• the Red line shows TRACE prediction.

21
Assessment
22
Comments on the REFLOOD model

• TWO MAIN RESULTS
• TRACE predictions generally underestimate the
  cladding temperatures, especially far from the
  quenching front (Highly Dispersed Flow regimes).
  The maximum temperature difference between
  predictions and data is reached at 346 sec after
  the reflood onset ( 300 K).
• TRACE overpredicts the rewetting time. After 200
  seconds from the reflood onset, the collapsed
  level is 0.5 m higher than the experimental data.

• COMMENTS ON REFLOOD MODEL
• The Minimum Stable Film Boiling Temperature is
  overestimated, resulting in a premature rewetting
  of the cladding.
• In Highly Dispersed Flow Regimes (far above the
  quenching front) the droplets size is
  overestimated, resulting in an unrealistic heat
  transfer enhancement.
• The grids are partially simulated through an
  enhancements in the interfacial heat-transfer
  between the phases. There is no visible effect on
  the cladding temperature though.

23
Results

• TRACE Reflood Model is an intricate sequence of
  empirical correlations used to describe the
  inherently non-stationary thermal-hydraulic
  phenomenon of post-LBLOCA reflooding.
• An assessment of the TRACE Reflood Model against
  experimental results (ACHILLES test) has been
  performed through a detailed model of the
  facility using TRACE code and SNAP.
• The assessment showed a non-conservative
  behaviour of the current Reflood Model when
  applied to a multi-pin subchannel geometry. Peak
  cladding temperatures are generally
  underestimated and the quenching front motion is
  clearly over-predicted by the code.
• Some reasons for this unrealistic calculations
  have been identified and future work will involve
  an improvement of single flow regimes models
  using more recent and more reliable correlations.
• The ACHILLES Ballooned test has also been
  performed.

24
Recent Work and Conclusion

• Static deformation of the clad to estimate the
  temperature distribution in the core.
• We are coupling TRACE and MABEL to model a Vessel
  (3D component) to dynamically predict the clad
  ballooning.
• Comparison with experiments
• Fuel relocation modelling.
• High ductility clad materials

25

• THANK YOU