Docstoc

Comparative Review of FCI Comput

Document Sample
Comparative Review of FCI Comput Powered By Docstoc
					  Synthesis of Analytical
         Activities for
Direct Containment Heating

   ERMSAR-07, 2007, Karlsruhe

     R. Meignen, S. Mikasser - IRSN, France
         C. Spengler – GRS, Germany
           A. Bretault – EDF, France
Summary
1. Introduction
       DCH phenomena
       Scope of work
       Overview of FZK experimental activities
       Overview of analytical activities
          - Lumped parameter codes
          - JPA activity
          - Use of CMFD codes

2. Overview of DISCO-L1/FH02 results
3. Comparison of LP codes
       Major features of code results
       Major features of code models with respect to
          dynamical and thermal aspects

4. Conclusions and perspectives
Synthesis of analytical activities for DCH   ERMSAR-07   2
Introduction
    •      DCH phenomena
    •      Scope of work
    •      Overview of FZK experimental activities
    •      Overview of analytical activities
                     Lumped parameter codes
                     JPA activity
                     Use of CMFD codes




Synthesis of analytical activities for DCH   ERMSAR-07   3
    Direct Containment Heating
    phenomena
• The vessel failure under pressure leads to
  the dispersion of the melt. Thermal and
  chemical transfers lead to a pressurization
  of the containment.
• The major tasks for modeling are
  therefore :
    a. Prediction of the dispersion
              Fragmentation and coalescence processes
              Size of fragments:
                 - interfacial area for transfers
                 - entrainment by the gas
    b. Chemical transfers
              Oxidation of fuel
              Combustion of initially present and
               produced hydrogen




   Synthesis of analytical activities for DCH   ERMSAR-07   4
 Scope of work in SARNET context - 1
• Most of the models where built for US reactor type geometries
• However, model are very dependent on geometry




 Quite limited confidence
 Need at least an assessment on other reactors.

 Synthesis of analytical activities for DCH   ERMSAR-07   5
Scope of work in SARNET context - 2
• FZK engaged an experimental and
  analytical program for EPR geometry
  => DISCO.
• IRSN-FZK collaboration for French PWR
  Reactors (P’4)
• Objectives :
      • Better understanding of phenomena with the use of
        multi-dimensional multi-fluid codes
      • Assessment/improvement of ASTEC models

• In SARNET WP13-2 a collaborative
  work of comparison of available 0-D
  code
      • CONTAIN : GRS, MAAP : EDF, ASTEC : IRSN
      • 1st phase : dynamical and thermal aspects




Synthesis of analytical activities for DCH   ERMSAR-07      6
      Overview of FZK experimental activity
•   First DISCO program with EPR geometry
     • Moderate ejection pressure, up to ~ 15 bars.

•   ~ 50 tests with cold simulant (water, Wood’s
    metal)
     • Visualization, pressure measurements, dispersion.
     • Several failure mode => central mode gives
       conservative results




•   6 tests with hot simulant (Al2O3-Fe thermite)
      • 2 cases with direct path to containment.
      • Variable amount of initial H2 in cont.
      • Measurement of hydrogen characteristics:
            • Production through oxidation
            • Combustion


      Synthesis of analytical activities for DCH   ERMSAR-07   7
      Overview of FZK exp. Activity-2
• Second DISCO series in collaboration with IRSN                              5

  with French 1300 MWe P’4 reactor geometry                   6
    • Complex 3-D geometry, large pit, direct path to                             3

      containment (5), access door (7)                                4
    • Ejection pressure, up to ~ 20 bars.
                                                                              2

• ~ 15 tests with cold simulant :water, gallium alloy
  (d~6)
    • Only central failure
    • Some test with “2D” geometry for code qualification                 7

• 5 tests with hot simulant (Al2O3-Fe thermite)
    • Up to 6% initial H2 in containment
    • 1 test in neutral environment => LACOMERA L1, funded
      by EC
    • 1 test in “2D” geometry (no access door)




     Synthesis of analytical activities for DCH   ERMSAR-07       8
 Lumped parameter code used in
 SARNET
• ASTEC / RUPUICUV
   • Developed by IRSN and GRS
• MAAP
   • Developed by US utilities, used by EDF
• CONTAIN
   • Developed at SNL, used by GRS
• Codes were used with a limited knowledge of the
  models and their adequacy/accuracy
   • IRSN calculations made by DSR (FAR) whereas code developed
     by DPAM (Cadarache)
   • GRS uses CONTAIN with the perspective to get familiar with
     the physics and develop COCOSYS.
   • Limited open literature for MAAP.


 Synthesis of analytical activities for DCH   ERMSAR-07   9
 JPA task on lumped parameter code
 comparison

• Objective :
   • - compare ASTEC modules with experiment (LACOMERA L1)
     and other codes : MAAP (EDF) and CONTAIN (GRS).
   • - propose modifications for ASTEC modelling if necessary

• Status
   • Launched at mid-2005
   • Received result calculations from GRS and EDF during nov.
     and dec. 2005.
   • Report on calculation results delivered
   • New modelling for ASTEC under investigation at IRSN and GRS




 Synthesis of analytical activities for DCH   ERMSAR-07   10
   Use of CMFD codes
• Two codes used :
   • AFDM (FZK) and MC3D (IRSN)
   • Oxidation quite parametric
   • No combustion in MC3D, 0-D model in AFDM

• Global analysis and calculations
   • Necessarily quite rough meshes
      3-D tests possible with MC3D
   • Gives insights for scaling effects

• Local analysis of specific phenomenon
   – Possible use of 3-d fine meshes, particularly
     for dynamical aspects.
   – Gives some insight for details in designs or
     reactor with no experimental data



   Synthesis of analytical activities for DCH   ERMSAR-07   11
DISCO/LACOMERA test L1 (FH02)
Description and short analysis




Synthesis of analytical activities for DCH   ERMSAR-07   12
    DISCO-L1/FH02 : Main characteristics
                                                                                                5

• Geometry of P’4 french PWR.                                              6
•     Complex 3-D geometry                                                                          3

            - Direct path to containment (5)                                          4

            - Path to sub-compartments (6)                                                       2
            - Bottom access (7)
•     Scale 1:16

• Fuel : Alumina/Fe thermite                                                              7

• Neutral environment                               Breach hole diameter                      60 mm
•     No chemical interaction                       Thermite mass                             10.64 kg
                                                    RPV pressure at failure                   1.92 MPa
                                                    Gas composition in RPV                    100% N2
                                                    Gas composition in containment            97% N2 3% O2

                                                    Containment pressure                      0.2 MPa
                                                    Containment temperature                   303 K
                                                    Melt temperature                          2200 K
    Synthesis of analytical activities for DCH   ERMSAR-07                       13
   DISCO-L1/FH02 : short analysis



                                                                                       Dispersal results
                                                                         45
                                                                         40
                                                                         35




                                                            Fraction %
                                                                         30
                                                                         25
                                                                         20
                                                                         15
                                                                         10
                                                                          5

•  Thermal efficiency                                                    0
                                                                              undispersed         to             to       to pit bottom

  •Using a simple thermal model:
                                                                                (in cavity)   compartment   containment      access


      - Max possible pressure : 5.5 bar                                                               Peffective
      - Global efficiency ~25 %                                                      h
      - Considering only dispersed fuel energy :                                               Pthermal equilibrium
         Containment + sub-comp : h = 55 %
         Containment only : h ~ 100 %
   Synthesis of analytical activities for DCH   ERMSAR-07                                        14
DISCO-L1/FH02 : short analysis - 2
Particle diameter far smaller in containment than in
 sub-compartment. Explains :
        The low efficiency in sub-compartment (also due to smaller
         volume)
        The near thermal equilibrium in containment

                             450   Particle mass distribution
                             400
                                                   Containment
                             350                   Compartment
                             300                   Dome
                  Mass [g]




                             250
                             200
                             150
                             100
                             50
                              0
                                 01
                                 04


                                 08


                                 16




                                 45
                                 63


                                 25




                                 55
                                   6


                                   2


                                   4
                                   5




                                   9


                                   8
                                   5




                                                                                   1
                                                                                 10
                                                                             5
                                05


                                11


                                22
                                31




                                 0,


                                 1,
                                 2,




                                                                                 7,
                               0,
                               0,


                               0,


                               0,




                               0,
                               0,


                               1,




                               3,
                              0,


                              0,


                              0,
                              0,




                                                    particle diameter [mm]




Synthesis of analytical activities for DCH      ERMSAR-07                        15
     DISCO-L1/FH02 : short analysis - 3
Why is there a so big difference in particle diameter ?
               melt has to make a 90°turn to go to sub-compartments, but
              larger particles :
               Indicates :
                        - a recoalescence process before flow separation.
                        - particle size determined by flow at respective nozzles, not
               necessarily representative of particles in pit and of dispersal process
                                                                                      Cont.


           sub




     Synthesis of analytical activities for DCH   ERMSAR-07         16
Comparison of LP codes and results
       Major features of code
       Major code models with respect to dynamical and thermal aspects




Synthesis of analytical activities for DCH   ERMSAR-07   17
 1-d. Overview of the JPA task on
 lumped parameter code comparison-2

• Short first conclusions from calculations :
   •Unsatisfactory results due to arbitrary fittings for
    all code calculations.
     - Predicitive capabilities quite dubious, at least for
       ASTEC
     - Need for improved modelling.
   •Use of the different models very unclear
     - Unadapted geometrical models ?
     - Weak understanding of the conditions of use of the
       models and choice of parameters ? Relevance of
       models ?


 Synthesis of analytical activities for DCH   ERMSAR-07   18
  Comparison of 0-D calculation : ASTEC (IRSN)°
                                                           CPA

• Rough model quite                                              V = 11.429 m3                 CONTAINMENT
                                                                 Zmin = 1.04 m

  inadequate to complex
                                                                 Zmax = 3.94 m
                                                                 Diameter = 2.16 m


  geometries
                                     RUPUICUV

    Dispersion ratio                                             V = 2.451 m3
                                                                                     COMPARTMENT
                                                                 Zmin = 0 m
    = user input                       CAVITY                    Zmax = 1.04 m
                                                                 Dmax = 1.81 m
                                                                 Dmin = 0.6 m
                                                                 V = 0.148 m3
                                                                 Zmin = 0 m             PIT BOTTOM ACCESS
                                                                 Zmax = 0.2 m
                                                                 Dmax = 1.8 m
                                           CORIUM                Dmin = 0.6 m

  • Parametrical dispersion and particle size (heat transfer)
  • Geometry fitted to obtain adequated global dispersion

• Complex heat transfer modelling as CPA does not
  handle particle flow:
  • CORIUM module as interface to compute the heat source input
  • Heat transfer with a correlation.
     Synthesis of analytical activities for DCH     ERMSAR-07                             19
      Comparison of 0-D calculation : MAAP
• MAAP (Modular Accident Analysis
  Program)
•    Commercial tool similar to ASTEC (but
far older), developed for the US utilities,
used in particular by EDF.
• Limited open literature and
documentation.
• The DCH modelling has a rather low level
of complexity.
• Melt dispersal from cavity
  evaluated with:
• two steps : melt followed by gas
• a correlation evaluated at time 0;
• a kinetic based on the gas blow-down
time;
• Dynamical and thermal equilibrium
  between gas and debris.


• Geometry used in calculations not
  fully relevant
•   Flow sections inadequate
•   Intermediate volume necessary
      Synthesis of analytical activities for DCH   ERMSAR-07   20
    Comparison of 0-D calculation : MAAP

• Only 54 % of fuel ejected from vessel
•      The rest is unmelted => pb with definition of melt energy.
•       Computation of dispersion not relevant (although good results)

• But pressurisation well calculated with an
  homogeneous model




    Synthesis of analytical activities for DCH   ERMSAR-07   21
  Comparison of 0-D calculation : CONTAIN (GRS)°
• Developed for the USNRC. DCH modelling was an important
  focus of CONTAIN with quite detailed models.
   •  CONTAIN is one of the reference tools for DCH
• Melt and gas ejection from vessel incorporates a transient
  2-j period.
• The melt fragmentation/trapping processes is calculated
  using:
   • A setting of different classes for debris according to their size and their
     temperature. The shape of the spectrum is however parametric.
   • A process of fragmentation based on an entrainment rate correlation.
     (Alternatively, the user could also specify an entrained fraction
     correlation)
• Inter-cell flow with a dynamical equilibrium assumption
  (slip velocity could be used).
• Convective heat transfer with a standard correlation.
• Radiative heat transfer to gases and structures.

      Synthesis of analytical activities for DCH   ERMSAR-07   22
  Comparison of 0-D calculation : CONTAIN (GRS)°

• Numerous discussions during
  meetings and in report
  concerning the use of
  dispersal models
   • Correct use of correlations ?
   • Coefficient from 100 to 1000,
     whereas a value of 5 is
     theroretically recommended.
• Imposed spectrum of particle
  size
   • 10 groups
   • Not consistent with exp. difference
     between compartments and
     containment
                                                               Revised geometry in calculations
• Heat transfer reduction by                                   with CONTAIN (GRS)
  factor of 3

      Synthesis of analytical activities for DCH   ERMSAR-07                     23
 Melt dispersal computation
• ASTEC and MAAP use complex correlations
  difficult to read and understand.
  Dimensionless numbers with obscur meaning.
• ASTEC :                                                                                                    2
                                                                                                   g U Dcavity
   Fd  0.45  tanh2log 10 Y  4.4
             1                                                                               We 
                                                                                                       
                                                                                                                  2
                                             g   
                                                       0.327
                                                                Dcavity 
                                                                             1.431
                                                                                                            gU
                                                                                              Fr 
           Y  We                                               D 
                      1.1058        0.149
                               Fr                                                                    l gDcavity
                                             l                   b    
• MAAP :
                               
    Fd 0.4 1tanh 3.79 log10 t 1.2                     *
                                                                             t *  t e udebris / Rcavity  H cavity 
                                                   1 
                                                         
                                m0 2   u gc     1 
                                                                                                   H O
                          te               1                                u debris  u g          2
                               Qcavity 1   u g  
                                              
                                                                                                    l
 Synthesis of analytical activities for DCH             ERMSAR-07                                24
 Melt dispersal computation-2
• GRS calculations with CONTAIN used the
  Whalley-Hewitt correlation:
   •Correlation for entrainement in 2-j film flow in
    tubes
                  d                                                        
         K c t                     t  1 fg u g        1  360  
                                                    2
                                                          
                                          2                 
                                                                    Dcavity 
                                                                             
   •Kc originally = 5, values from 100 to 1000 used
    (~100 recommended in documentation)
   •Dependancy of viscosity quite dubious.
   •No influence on the melt density !




 Synthesis of analytical activities for DCH   ERMSAR-07               25
   Heat transfers
• A local model using a correlation as in ASTEC
  and CONTAIN seems unadequate in O-D model
  (choice of velocity ? Sensitivity ?)
  • Necessary fitting with large uncertainties for
   reactor applications
• Equilibrium assumption in MAAP
  •Relevant regarding exp. analysis
  •Needs further checking
• Important sensitivity in ASTEC
  •Contradiction with exp. Analysis
  •Investigations on-going.

   Synthesis of analytical activities for DCH   ERMSAR-07   26
  Conclusions of JPA Task.
• Use of existing correlations and models for dynamical
  aspects very uncertain:
    • Geometry in general not adapted
       - => Strong fitting hides the relevance.
       - => Need for geometry-specific models ?
               (already recognized with previous US studies, parameters in correlation
                 varied by 1 order of magnitude depending of the reactor)
    • Necessity for an increased comprehension:
        - Dispersal processes
        - Fragmentation processes.
    •  DCH not really closed as long as calculation adequacy not
      clarified and validated.
• Characteristic particle size should be different in each
  volume
• Simplified heat transfer modelling necessary and
  possible
    • Homogenous model might be sufficient.

  Synthesis of analytical activities for DCH   ERMSAR-07             27
 Perspectives
• 2nd phase of task with computation of tests
  FH01 and FH03 with oxidation and combustion
• Analysis of the building of a new model still
  on-going at IRSN.
• Use of CFMD codes showed to be promising for
  the dispersal problem.
   •As an alternative : a specific correlation can be
    build from calculations (at reactor scale) for each
    reactor geometry (on-going at IRSN with MC3D)




 Synthesis of analytical activities for DCH   ERMSAR-07   28

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:10/19/2010
language:English
pages:28