33
METHODS & USE OF PROBABILISTIC SEISMIC HAZARD ANALYSIS IN US NUCLEAR POWER PLANTS Dr. Annie Kammerer, PE Pacific Earthquake Engineering Research Center, UC Berkeley Annie Kammerer ConsulLng Korean Atomic Energy Research InsLtute April 2015 Acknowledgements to : Dr. Kevin Coppersmith, Dr. Julian Bommer, and Dr. Jon Ake

METHODS(&(USE(OF(PROBABILISTIC(SEISMIC(HAZARD(ANALYSIS( …

  • Upload
    others

  • View
    5

  • Download
    1

Embed Size (px)

Citation preview

METHODS  &  USE  OF  PROBABILISTIC  SEISMIC  HAZARD  ANALYSIS  

IN  US  NUCLEAR  POWER  PLANTS  

Dr.  Annie  Kammerer,  PE  Pacific  Earthquake  Engineering  Research  Center,  UC  Berkeley  

Annie  Kammerer  ConsulLng  

Korean  Atomic  Energy  Research  InsLtute  April  2015  

 

Acknowledgements  to  :  Dr.  Kevin  Coppersmith,    Dr.  Julian  Bommer,  and  Dr.  Jon  Ake  

 

ProbabilisLc  Seismic  Hazard  Analysis  (PSHA)  

¨ Method  for  assessing  seismic  load  ¨ ObjecLve  is  to  determine  the  best  esLmate  (and  associated  uncertainty)  of  ground  moLon  levels  at  a  parLcular  locaLon  over  Lmes  periods  of  interest  

¨ PSHA  considers  all  possible  earthquakes  from  all  seismic  sources  that  may  impact  a  site,  and  accounts  for  the  likelihood  of  any  parLcular  earthquake  

2  

Design  Level  &  Total  Hazard  3  

SPRA  

Seismic  Load   Capacity          Seismic  MoLon  Parameter  

Freq

uency  of  Exceedance  

Pi  

Seismic  MoLon  Parameter  

Cond

iLon

al  Probability  of  Failure  

i  

Systems  Analysis  

Event  trees,  Fault  trees,  Containment  Analysis  

Assessing  Hazard  &  Risk  4  

Early  AEC  Seismic  Hazard  

¨  Early  seismic  hazard  assessments  under  the  Atomic  Energy  Commission  were  for  a  Design  Basis  Earthquake  (DBE),  which  was  determinisLc  with  standard  spectral  shapes  Led  to  a  peak  ground  acceleraLon  (PGA)  based  on  past  earthquakes  in  the  region  (and  a  lot  of  expert  judgment).  

5  

NRC  Seismic  Hazard  RegulaLons  

¨  In  1971  the  NRC  established  General  Design  Criteria  (GDC  2),  which  required  that  SSCs  important  to  safety  be  designed  to  withstand  the  effects  of  natural  phenomena    with  “appropriate  consideraLon  of  the  most  severe  of  the  natural  phenomena  that  have  been    historically  reported  for  the  site  and  surrounding  region  with  sufficient  margin  for  the  limited  accuracy  and  quan?ty  of  the  historical  data  and  the  period  of  Lme  in  which  the  data  have  been  accumulated”  (Appendix  A  to  10  CFR  Part  50).  Codifies  seismic  design  &  requires  consideraLon  of  uncertainty.  

6  

NRC  Seismic  Hazard  RegulaLon  

¨  In  1973,  Appendix  A  to  10  CFR  Part  100,  “Seismic  and  Geologic  Si?ng  Criteria  for  NPPs”  was  established  to  provide  detailed  criteria  to  evaluate  the  suitability  of  proposed  sites  and  the  suitability  of  the  plant  design  basis  established  in  consideraLon  of  seismic  and  geological  characterisLcs.  Codified  minimum  standards  for  site  and  hazard  evaluaLon  and  site  suitability.  

¨  In  1996,  the  NRC  issued  10  CFR  Part  50  Appendix  S  defines  SSE  as  “Safe-­‐shutdown  earthquake  ground  mo5on  is  the  vibratory  ground  moLon  for  which  certain  structures,  systems,  and  components  must  be  designed  to  remain  funcLonal”    Provides  performance-­‐based  requirement  for  ground  moLon  that  replaces  the  determinisLc  DBE  

7  

Current  Risk-­‐Informed  Design  Guidance  

¨  In  1997  NRC  was  issued  a  commission  direcLve  to  move  towards  “Risk  Informed”  policies.    

¨  1997  issuance  of  RG  1.165,  which  was  the  first  seismic  hazard  guide  based  on  probabilisLc  techniques.  

¨  2007  NRC  issued  RG1.208,  based  on  PSHA  and  developed  for  use  with  ASCE  43-­‐05  ¤  RG1.208:  A  Performance-­‐Based  Approach  to  Define  the  Site-­‐Specific  Ground  MoLon  

¤  ASCE  43-­‐05:  Seismic  Design  Criteria  for  SSCs  in  Nuclear  FaciliLes  

8  

Recent  Seismic  Hazard  Re-­‐evaluaLon  

¨  In  mid-­‐2000s  interest  was  renewed  in  new  NPPs  and  the  NRC  received  3  Early  Site  Permit  (ESP)  applicaLons  for  reactors  to  be  co-­‐located  at  exisLng  sites.  ¤ North  Anna,  Vogtle,  and  Calvert  Cliffs  

¨ Based  on  the  difference  in  the  new  PSHA  data  and  the  DBEs  for  the  exisLng  reactors,  NRC  Staff  iniLated  Generic  Issue  (GI)  199.  

9  

Recent  Re-­‐evaluaLon  

¨ GI-­‐199  was  a  program  to  reevaluate  seismic  hazard  and  risk  for  central  &  eastern  US  NPPs  

¨ Prior  to  Fukushima,  the  NRC  performed  and  published  an  iniLal  screening  required  to  show  that  GI-­‐199  met  program  criteria  (risk-­‐significant  changes  to  mulLple  NPPs)  

¨ Ajer  Fukushima  GI-­‐199  was  extended  to  all  NPPs.  GI-­‐199  the  became  part  of  the  Fukushima  NTTF  recommendaLon  2.1  Program.  

10  

Seismic  Sources  Magnitudes  &  LocaLons  

Earthquake  Recurrence  

Seismic  Source  CharacterizaLon  Model  

Ground  MoLon  Model  

Seismic  Hazard  Curves  

Local  Site  Response  

GMRS  (ground  moLon)  

Overview  of  PSHA    

Seismic  Hazard  Advances  12  

• Central  and  Eastern  US  Seismic  Source  CharacterizaLon  project  for  Nuclear  FaciliLes  (CEUS  SSC  Project)  

• NUREG  2115  (NRC,  2012)  

Source  CharacterizaLon  

• EPRI  Ground  MoLon  Model  Update  Project  (EPRI,  2013)    • Next  GeneraLon  AlenuaLon  RelaLonships  for  the  Central  and  Eastern  (NGA-­‐East)  (PEER  UC  Berkeley,  2015)  

Ground  moLon  predicLon  equaLons  

• PracLcal  ApplicaLon  of  the  SSHAC  Guidelines  • NUREG  2117  (NRC,  2012)  

PSHA  process  guidance    

Senior  Seismic  Hazard  Analysis  Commilee  (SSHAC)    Original  report  provides  and  procedure  for  conducLng  mulLple-­‐expert  assessments  of  input  to  PSHA    New  report  provides  addiLonal  details.    

NUREG/CR-6372 NUREG 2117 (1989) (2012)

SSHAC  Guidelines  

ObjecLve  of  a  SSHAC  Study  

¨ Objective is to develop a model that represents the center, body and range of technically defensible interpretations of the available data ¤ Center-best estimate ¤ Body-shape of the distribution ¤ Range-extreme values of the distribution

¨ Achieved through a process with well defined evaluation and integration phases

Key  elements  of  a  SSHAC  study  

¨ Views of the larger technical community are fundamental

¨ Competing scientific hypotheses can be considered and uncertainties captured

¨ PSHA is a snapshot in time of our knowledge and uncertainties

¨ Notion of the “views of the larger informed technical community” leads to defined expert roles and responsibilities

¨ Experts must be “evaluators” not “proponents”

SSHAC  Process  Outcomes  

¨  Stability  ¤  Enjoys  public  confidence  that  the  views  of  the  larger  informed  technical  community  have  been  considered  and  properly  represented  

¤  Public  represented  by  regulator  ¤ Not  subject  to  significant  change  with  each  new  scienLfic  finding  

¨  Longevity  ¤  Results  expected  to  be  valid  for  applicaLons  up  to  at  least  10  years  in  the  future  

¤  The  technical  underpinnings  will  remain  valid  in  the  future,  despite  the  development  of  new  data,  unLl  significant  changes  in  the  data  or  the  science  

}  Gather  data  and  informaLon  from  literature  }  TI  makes  assessments  including  uncertainty  }  TI  confers  with  members  of  technical  community  to  understand  alternaLve  viewpoints  

}  Workshops  are  held  to  discuss:  ◦  Significant  issues  and  available  data  ◦  AlternaLve  hypotheses  ◦  Feedback  

}  ParLcipatory  peer  review  of  process  and  technical    }  TI  team  responsible  for  technical  assessments  }  Expert  panel  responsible  for  making  technical  assessments  

}  TFI  facilitates  expert  interacLons  and  aggregates  expert  assessments  

Leve

l 1

Leve

l 2

Leve

l 3

Leve

l 4

SSHAC  Levels  

¨  Stability  comes  from  idenLfying  and  quanLfying  uncertainLes  and  integraLng  them  into  the  PSHA  (and  SPRA)  model  

¨  Views  of  the  larger  technical  community  are  uncertain  ¤  AlternaLve  models  for  the  locaLons  and  rates  of  future  earthquakes  

¤  Parameters  that  define  the  models  are  uncertain  ¨  Likelihood  that  the  full  range  of  technically  defensible  interpretaLons  is  captured  increases  with  SSHAC  Study  Level  (2  versus  3  and  4)  

¨  Increasing  formalism  and  involvement  of  experts  ¨  Increasing  stability  (regulatory  confidence)  that  all  viable  hypotheses  have  been  considered  

Formal  Processes  for  UncertanLes  

Uncertainty  

Aleatory  

Natural  variability  

Not  reducible  

Addressed  through  integraLon  over    parameter  distribuLons  

Epistemic  Modeling  or  knowledge  

uncertainty  

Reducible  with  more  informaLon  

Addressed  through  use  of  a  logic  tree  

UncertainLes  in  Hazard  &  Risk  19  

Uncertainty  

Aleatory          

IntegraLon  over  distribuLon  of  expected  parameter  values  

Epistemic  

logic  tree  of    technically  defensible  interpretaLons  

UncertainLes  in  Hazard  &  Risk  20  

Process and Technical R

eview

PPRP

Evaluation of M

odels to Form C

omposite D

istribution

TI Team

Hazard  sensi?vity  calcula?ons  Preliminary  database  

WORKSHOP  1:  Hazard  Sensi?ve  Issues  and  Data  Needs  

Resource  Experts  

Addi?onal  data  collec?on  &  analysis  

WORKSHOP  2:  Review  of  Database  and  Discussion  of  Alterna?ve  Models  

Resource  Experts  

Proponent  Experts  

Final  database   Preliminary  SSC  and  GMC  models  

WORKSHOP  3:  Presenta?on  of  Models  and  Hazard  Sensi?vity  Feedback  

Final  SSC  and  GMC  models,  then  final  hazard  calcula?ons,  Documenta?on  of  all  technical  bases  

Database C

ompilation

Technical Staff & Contractors

NUREG  2117  (Level  3)  

Seismic  Hazard  Advances  22  

• Central  and  Eastern  US  Seismic  Source  CharacterizaLon  project  for  Nuclear  FaciliLes  (CEUS  SSC  Project)  

• NUREG  2115  (NRC,  2012)  

Source  CharacterizaLon  

• EPRI  Ground  MoLon  Model  Update  Project  (EPRI,  2013)    • Next  GeneraLon  AlenuaLon  RelaLonships  for  the  Central  and  Eastern  (NGA-­‐East)  (PEER  UC  Berkeley,  2015)  

Ground  moLon  predicLon  equaLons  

• PracLcal  ApplicaLon  of  the  SSHAC  Guidelines  • NUREG  2117  (NRC,  2012)  

PSHA  process  guidance    

Central  and  Eastern  US  Sites  

Process  

Recent  Advances  in  PSHA  

Source  Characteriza?on   Ground  Mo?on  Characteriza?on  

PUBLISHED    January  2012   IN  PROGRESS  Expected  2015  

23  

• Advances  in  applying  SSHAC  process  on  a  site-­‐specific  basis  in  the  US.  New  assessments  submiled  the  NRC  March  2015.  

Western  Sites  

• First  applicaLon  of  the  SSHAC  Level  3  process  in  a  developing  naLon  

• Assessment  for  new  NPPs  • Sponsored  by  uLlity  with  ongoing  review  by  the  regulator  

South  Africa  

Site-­‐Specific  Seismic  Hazard  24  

Bommer  et  al  (2013),  “A  SSHAC  Level  3  ProbabilisLc  Seismic  Hazard  Analysis  for  a  New-­‐Build  Nuclear  Site  in  South  Africa”  Earthquake  Spectra  

ProbabilisLc  Flood  Hazard  Advances  25  

¤ Ongoing  US  efforts  towards    probabilisLc  flooding  methods                                                      to  support  more  uniform                                                                    treatment  of  hazard  in  risk-­‐informed  decision  making  

¤ January  2013  Joint  Workshop  on  ProbabilisLc  Flood  Hazard                                        Assessment  Workshop    

¤ Recent  NRC  &  FERC  staff  workshop  on  PFHA  and  applicaLon  of  the  SSHAC  guidelines  (proceedings  expected  to  be  published)  

hlp://www.nrc.gov/public-­‐involve/public-­‐meeLngs/meeLng-­‐archives/

research-­‐wkshps.html  

Regulatory  Documents  

¨  Code  of  Federal  RegulaLons  ¤  10  CFR  parts  50  &  52,  Federal  law  

¨  Regulatory  Guidance  ¤  States  staff  posiLons  on  meeLng  the  code  

¨  NUREG-­‐Series  Documents  ¤  Technical  supporLng  documents  ¤  NUREG  (staff)  &  NUREG/CR  (contractors)  

¨  Standard  Review  Plan  (NUREG  0-­‐800)  ¤  NRC  review  standards  for  staff  

¨  Plant  OperaLng  Licenses  ¤  Become  part  of  10CFR  (and  are  therefore  federal  law)  

¨  CerLfied  Designs  ¤  Become  part  of  10CFR  

26  

AEC    Atomic  Energy  Commission  ASCE    American  Society  of  Civil  Engineers  ASME    American  Society  of  Mechanical  Engineers  CDF      Core  Damage  Frequency  CFR    Code  of  Federal  RegulaLons  COL    Combined  OperaLng  License  DBE    Design  Basis  Earthquake  EDMG    Extreme  Damage  MiLgaLon  Guideline  FOSID    Frequency  of  the  Onset  of  Significant  InelasLc        

     DeformaLon  HCLPF    High  Confidence  of  a  Low  Probability  of  Failure  GDC    General  Design  Criteria  GL      Generic    Leler  (to  Licensees)  IEEE    InsLtute  of  Electrical  and  Electronics  Engineers  

Acronyms 27  

JLD      Japan  Lessons  Directorate  NPPs    Nuclear  Power  Plants  NRC    Nuclear  Regulatory  Commission  NTTF    Near  Term  Task  Force  (on  Fukushima)  PEER    Pacific  Earthquake  Engineering  Research  PGA    Peak  Ground  AcceleraLon  QME    QualificaLon  of  Mechanical  Equipment  RG      Regulatory  Guide  SAMG    Severe  Accident  MiLgaLon  Guideline  SEP      SystemaLc  EvaluaLon  Program  SSE    Safe  Shutdown  Earthquake  SSC    Structures,  Systems  and  Components  SSC    Seismic  Source  CharacterizaLon       Acronyms

28  

Thank  You  

QuesLons?  

29  

Uncertainty  

                   Aleatory                                                    Epistemic  

AcceleraLon  (g)  

Annu

al  Prob  of    Exceedance  

Aleatory  variability  gives  the  curve  its  

shape.  

Epistemic  uncertainty  leads  to  uncertainty  

bands  85%

Median

15%

30  

UncertainLes  in  Hazard  &  Risk  

Uncertainty  

Median  Curve  

AcceleraLon  (g)  

Annu

al  Prob  of    Exceedance  

85%

Median

15%

31  

UncertainLes  in  Hazard  &  Risk  

Uncertainty  

Mean  Curve  

AcceleraLon  (g)  

Annu

al  Prob  of    Exceedance  

85%

Median

15%

The  mean  curve  is  used  in  risk  assessment  and  design  to  beler  account  for  epistemic  uncertainty.  

Mean  Curve  

The  mean  curve  exceeds  the  median  due  to  the  log-­‐normal  distribuLon  of  most  parameters  

UncertainLes  in  Hazard  &  Risk  32  

Uncertainty  

Mean  Curve  

AcceleraLon  (g)  

Annu

al  Prob  of    Exceedance  

85%

Median

15%

Greater  epistemic  uncertainty  due  to  lack  of  data  leads  to  a  higher  mean  curve.  This  leads,  in  turn,  to  higher  assessments  of  risk.  There  is  a  benefit  to  accumulaLng  data.  

 

Mean  Curve  

UncertainLes  Hazard  &  Risk  33