ML13255A376

From kanterella
Jump to navigation Jump to search
Soarca Peach Bottom Uncertainty Analysis (UA) Acrc Briefing - Sept 2013
ML13255A376
Person / Time
Site: Peach Bottom  Constellation icon.png
Issue date: 09/16/2013
From: Tina Ghosh
NRC/RES/DSA
To:
Ghosh T
Shared Package
ML13255A375 List:
References
Download: ML13255A376 (55)
v • d • e


Text

SOARCA Peach Bottom Uncertainty Analysis (UA)

ACRS Briefing Tina Ghosh, PhD RES/DSA/AAB September 16, 2013

Agenda

  • ACRS comments t on MACCS2 weatherth uncertainty integration and convergence of results and staff responses results,
  • MELCOR parameters of interest
  • MACCS2 parameters of interest 2

MELCOR - MACCS2 -

Weather Uncertainty Integration ACRS Comment:

  • For the combined MELCOR-MACCS2 results, the report currently presents only results averaged over the weather trials.
  • The report should also present results that include and di l the display th full f ll weather th aleatory l t uncertainty t i t 3

Conditional mean, individual latent cancer fatality y ((LCF)) risk (p (per event) for combined results (865) with LNT model 0 10 0-10 0 20 0-20 0 30 0-30 0 40 0-40 0 50 0-50 miles miles miles miles miles 5th 3.1x10 3 1 10-55 4 4.9x10 9 10-55 3 3.4x10 4 10-55 2 2.2x10 2 10-55 1 1.9x10 9 10-55 percentile Median 1.3x10-4 1.9x10-4 1.3x10-4 8.7x10-5 7.1x10-5 Mean 1.7x10-4 2.8x10-4 2.0x10-4 1.3x10-4 1.0x10-4 95th 4.2x10-4 7.7x10-4 5.3x10-4 3.4x10-4 2.7x10-4 percentile SOARCA UA 9.0x10-5 8.3x10-5 5.8x10-5 3.7x10-5 3.0x10-5 Base Case 4

Conditional Individual LCF Risk (per Event) CCDFs for Combined Aleatory and d Epistemic E i t i Uncertainty U t i t andd Epistemic Uncertainty with Aleatory Means 1 0-10 miles Aleatory Mean 0.9 0-20 miles Aleatory Mean 0-50 miles Aleatory Mean 0.8 0-10 miles Epistemic & Aleatory 0 20 miles 0-20 il E Epistemic i t i & Al Aleatory t

0.7 0-50 miles Epistemic & Aleatory 0.6 05 0.5 CCD DF 0.4 0.3 0.2 0.1 0

1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 Individual LCF Risk 5

MACCS2 and Weather Uncertainties for Prompt Fatality Risk ACRS Comment:

  • Select the MELCOR realization that produced the largest conditional prompt fatality consequences in the current SOARCA uncertainty results.
  • For that realization, sample from the 350 MACCS2 input parameters and for each epistemic sample generate 984 parameters, weather cases to derive an uncertainty distribution for the conditional prompt fatality consequences at each di t distance.
  • Demonstrate convergence of the combined MACCS2-weather uncertainty analysis results.

6

MACCS2 and Weather Uncertainties for Prompt Fatality Risk (cont.)

Approach:

  • MELCOR Replicate 2, Realization 291 identified as the source term that produced the largest conditional prompt fatality risk consequence
  • For that source term, three Monte Carlo runs of sample size 1000 were completed (Runs 3 3, 4 4, 5) using three different LHS random seeds for the 350 MACCS2 input parameters
  • The same 984 weather trials were used 7

Conditional, mean, individual prompt-fatality p p y risk (p (per event))

statistics for the MACCS2 Uncertainty Analysis for specified circular areas (Run 1) 0-1.3 0-2.5 0-3.5 0-7 0-10 miles miles miles miles miles Mean 4.5x10-7 8.9x10-8 3.5x10-8 8.3x10-9 4.8x10-9 Median 00 0.0 00 0.0 00 0.0 00 0.0 00 0.0 75th percent p 0.0 0.0 0.0 0.0 0.0

-ile 95th 1.9x10 percent 1 9x10-6 3 3.5x10 5x10-8 00 0.0 00 0.0 00 0.0

-ile 8

Run 3-5 conditional, mean, individual p prompt-fatality p y risk (p (per event) statistics for specified circular areas 0-1.3 0-2.5 0-3.5 0-10 0-7 miles miles miles miles miles Run 3 3.3E-06 1.0E-06 3.4E-07 4.7E-08 9.5E-09 M

Mean R Run 4 3 3E 06 3.3E-06 9 4E 07 9.4E-07 3 0E 07 3.0E-07 4 2E 08 4.2E-08 8 9E 09 8.9E-09 Run 5 3.2E-06 9.8E-07 3.0E-07 4.7E-08 1.3E-08 Run 3 4.9E-07 1.2E-07 0.0 0.0 0.0 Median Run 4 3 3E 06 3.3E-06 9 4E 07 9.4E-07 00 0.0 00 0.0 00 0.0 Run 5 3.2E-06 9.8E-07 0.0 0.0 0.0 75th Run 3 4.0E-06 1.0E-06 2.0E-07 3.8E-09 0.0 percent Run 4 3 7E 06 3.7E-06 8 8E 07 8.8E-07 2 2E 07 2.2E-07 1 1E 08 1.1E-08 00 0.0

-ile Run 5 3.9E-06 9.6E-07 1.9E-07 8.2E-09 0.0 95th Run 3 1.4E-05 4.1E-06 1.5E-06 2.1E-07 1.2E-08 percent Run 4 1 6E 05 1.6E-05 4 7E 06 4.7E-06 1 8E 06 1.8E-06 2 3E 07 2.3E-07 00 0.0

-ile Run 5 1.4E-05 4.4E-06 1.6E-06 2.0E-07 0.0 9

Runs 3-5 and Run 1 Conditional, Mean Individual Prompt Fatality Risk (per Event) Epistemic Uncertainty CCDF, at 1.3 Miles 1

0.1 0-1.3 miles Run 1 CCD DF 0-1.3 miles Run 3 0.01 0-1.3 miles Run 4 0-1.3 miles Run 5 0.001 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 Individual Prompt Fatality Risk per Event 10

Runs 3-5 and Run 1 Conditional, Mean Individual Prompt Fatality Risk (per Event) Epistemic Uncertainty CCDF, at 3.5 Miles 1

0.1 0-3.5 miles Run 1 0-3 0 3.5 5 miles Run 3 CCD DF 0-3.5 miles Run 4 0.01 0-3.5 miles Run 5 0.001 0 001 1.0E-12 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 Individual Prompt Fatality Risk per Event 11

MACCS2 and Weather Uncertainties for LCF Risk 1 ACRS Comment:

  • Select the MELCOR realization that produced the largest conditional LCF fatality consequences in the current SOARCA uncertainty results.
  • For that realization, sample from the 350 MACCS2 input parameters and for each epistemic sample generate 984 parameters, weather cases to derive an uncertainty distribution for the conditional LCF fatality consequences at each distance.
  • Demonstrate convergence of the combined MACCS2-weather uncertainty analysis results.

12

MACCS2 and Weather Uncertainties for LCF Risk 1 (cont.)

Approach:

  • MELCOR Replicate 3, Realization 46 identified as the source term that produced the largest conditional LCF risk consequence
  • For that source term, three Monte Carlo runs of sample size 1000 were completed (Runs 6 6, 7 7, 8) using three different LHS random seeds for the 350 MACCS2 input parameters
  • The same 984 weather trials were used 13

Run 6-8 Combined Aleatory and Epistemic Uncertainty Conditional Individual LCF Risk (per Event) CCDF 1

0.9 0-10 miles Run 6 0.8 0-10 miles Run 7 07 0.7 0-10 miles Run 8 0.6 0-50 miles Run 6 0.5 0-50 miles Run 7 CC CDF 0.4 0-50 miles Run 8 0.3 02 0.2 0.1 0

1.E-05 1.E-04 1.E-03 1.E-02 Individual Latent Cancer Fatality Risk per Event 14

Runs 6-8 and Run 1 Epistemic Uncertainty with Aleatory Mean, Conditional Individual LCF Risk (per Event) CCDFs 1

0-10 0 10 miles Run 1 0.9 0-10 miles Run 6 0.8 0-10 miles Run 7 07 0.7 0-10 0 10 miles Run 8 0-50 miles Run 1 0.6 0-50 miles Run 6 0.5 0-50 miles Run 7 CCD DF 0.4 0-50 miles Run 8 0.3 0.2 0.1 0

1.0E-05 1.0E-04 1.0E-03 Individual Latent Cancer Fatality Risk per Event 15

MACCS2 and Weather Uncertainties for LCF Risk 2 ACRS Comment:

  • Select a MELCOR realization that produced a small, but non-zero, contribution to the conditional LCF fatality consequences in the current SOARCA uncertainty results.

realization sample from the 350 MACCS2 input

  • For that realization, parameters, and for each epistemic sample generate 984 weather cases to derive an uncertainty distribution for the conditional diti l LCF ffatality t lit consequences att each h di distance.

t

  • Demonstrate convergence of the combined MACCS2-weather uncertainty analysis results.

16

MACCS2 and Weather Uncertainties for LCF Risk 2 (cont.)

Approach:

  • Three representative p source terms were chosen
  • First an initial MACCS2 run (Run 2) used all 865 source terms while all MACCS2 parameters were set to their SOARCA point estimate values values.

- To assess the influence of the source term when MACCS2 parameters are fixed 17

Run 2 Conditional Mean, Individual LCF Risk (per Event) for 865 Source Terms and Fixed CCDF 1

0.9 0-10 miles 0.8 0-20 miles 0.7 0-30 miles 0.6 0-40 miles 0.5 0-50 miles 04 0.4 CCD DF 0.3 0.2 0.1 0

1 0E 05 1.0E-05 1 0E 04 1.0E-04 1 0E 03 1.0E-03 1 0E 02 1.0E-02 Individual LCF Risk per Event 18

MACCS2 and Weather Uncertainties for LCF Risk 2 (cont.)

  • A set of 11 results have then been used as metrics to select three representative source terms:

- Latent Cancer Fatality (LCF) risk at 5 different locations (10 (10, 20 20, 30, 40 and 50 miles)

- Fraction of inventory released for 5 radionuclides (Cs, I, Ba, Ce, Te)

- Release time

  • Goal is to choose three source terms whose metrics ranks come closest to 1/6, 1/2, and 5/6 among the population 19

Results: Cobweb Graph for Selected Source Terms 1.0 0.8 0.6 CDF F 0.4 02 0.2 0.0 50 20 10 30 40 Cs Iodine Ba Ce Te (hr) miles miles miles miles miles Metric 20

MACCS2 and Weather Uncertainties for LCF Risk 2 (cont.)

Approach (cont (cont.):):

  • With respect to conditional LCF risk:

- MELCOR Replicate 3, Realization 187 identified as the representative low source term

- MELCOR Replicate 1, Realization 75 identified as the representative medium source term

- MELCOR Replicate 1, Realization 290 identified as the representative high source term

  • For each of these source terms,, three Monte Carlo runs of sample size 1000 were completed (Runs 9-11,12-14, 15-17 respectively) using three different LHS random seeds for the 350 MACCS2 input parameters
  • The same 984 weather trials were used.

21

Runs 9-11 (Low Source Term)

Conditional, Mean, Individual LCF Risk (per event) Statistics Run # 0-10 0 10 0-20 0 20 0-30 0 30 0-40 0 40 0-50 0 50 Statistic miles miles miles miles miles Run 9 1.1E-04 1.2E-04 8.3E-05 5.4E-05 4.4E-05 Mean Run 10 1 1E-04 1.1E-04 1 2E-04 1.2E-04 8 3E-05 8.3E-05 5 4E-05 5.4E-05 4 4E-05 4.4E-05 Run 11 1.1E-04 1.2E-04 8.3E-05 5.4E-05 4.4E-05 Run 9 8.8E-05 1.0E-04 7.2E-05 4.7E-05 3.9E-05 Median Run 10 8 6E-05 8.6E-05 1 0E-04 1.0E-04 7 4E-05 7.4E-05 4 8E-05 4.8E-05 3 9E-05 3.9E-05 Run 11 8.8E-05 1.0E-04 7.2E-05 4.7E-05 3.9E-05 5th Run 9 2.3E-05 3.8E-05 2.7E-05 1.7E-05 1.4E-05 percentile Run 10 2 2E 05 2.2E-05 3 8E 05 3.8E-05 2 6E 05 2.6E-05 1 7E 05 1.7E-05 1 4E 05 1.4E-05 Run 11 2.3E-05 4.0E-05 2.7E-05 1.8E-05 1.4E-05 95th Run 9 2.5E-04 2.4E-04 1.7E-04 1.1E-04 8.9E-05 percentile Run 10 2 6E 04 2.6E-04 2 4E 04 2.4E-04 1 7E 04 1.7E-04 1 2E 04 1.2E-04 9 5E 05 9.5E-05 Run 11 2.7E-04 2.4E-04 1.7E-04 1.1E-04 9.4E-05 22

Runs 9-11 and Run 1 Epistemic Uncertainty Conditional, Mean, Individual LCF Risk (per Event) CCDFs 1 0-10 miles Run 1 09 0.9 0 10 miles Run 9 0-10 0-10 miles Run 10 0.8 0-10 miles Run 11 07 0.7 0 50 miles 0-50 il R Run 1 0.6 0-50 miles Run 9 0.5 0-50 miles Run 10 CCD DF 0-50 miles Run 11 0.4 0.3 0.2 0.1 0

1.0E-06 1.0E-05 1.0E-04 1.0E-03 Individual LCF Risk per Event 23

Runs 12-14 (medium) and Run 1 Epistemic Uncertainty Conditional, Mean, Individual LCF Risk Ri k (per

( E Event) t) CCDFs CCDF 1 0-10 miles Run 1 09 0.9 0-10 miles Run 12 0.8 0-10 miles Run 13 0-10 miles Run 14 0.7 0-50 miles Run 1 0.6 0-50 miles Run 12 0.5 0-50 miles Run 13 CCD DF 0.4 0-50 miles Run 14 0.3 02 0.2 0.1 0

1.0E-06 1.0E-05 1.0E-04 1.0E-03 Individual LCF Risk per Event 24

Runs 15-17 (high) and Run 1 Epistemic Uncertainty Conditional, Mean, Individual LCF Risk (per Event) CCDFs 1

0-10 miles Run 1 0.9 0-10 0 10 miles Run 15 0-10 miles Run 16 0.8 0-10 miles Run 17 0.7 0-50 miles Run 1 0 50 miles Run 15 0-50 0.6 0-50 miles Run 16 0-50 miles Run 17 0.5 CCD DF 0.4 0.3 02 0.2 0.1 0

1.0E-06 1.0E-05 1.0E-04 1.0E-03 Individual LCF Risk per Event 25

Average difference between the three separate p LHS runs over all Aleatory Weather Distributions (1st to 99th percentile)

Conditional Conditional Source Term LCF Risk LCF Risk 0-10 miles 0-50 miles Highest Prompt Fatality 0.8% 0.8%

Risk - Runs 3-5 35 Highest LCF Risk -

0.8% 0.9%

Runs 6-8 L

Low - Runs R 9 9-11 11 0.9%

0 9% 0.8%

0 8%

Medium - Runs 12-14 0.8% 0.9%

High - Runs 15-17 15 17 1 0%

1.0% 0 6%

0.6%

Overall Average 0.9% 0.8%

26

MACCS2 Stability Analysis Using Bootstrap Approach Approach:

  • MACCS2 code modified to allow simple random sampling
  • The high g source term ((i.e., Replicate p 1 Realization 290))

and the SOARCA UA MACCS2 Analysis (Run 1) were selected to compare between Simple Random Sampling (SRS or MC) and Latin Hypercube Sampling (LHS) in order to validate the use of LHS

  • Bootstrapping performed (similar to approach with MELCOR results) to estimate confidence bounds

Conclusion:

Results of the uncertainty analysis are well converged and LHS use is valid 27

Run 1 (CAP17) Conditional, Mean, Individual LCF Risk (per Event) CCDF with LHS and MC Sampling 1

09 0.9 0.8 0.7 0.6 0.5 0-10 miles CAP17-LHS CCD DF 0.4 0-10 miles CAP17-MC 0.3 0-50 miles CAP17-LHS 0.2 0-50 miles CAP17-MC 0.1 0

1.0E-06 1.0E-05 1.0E-04 1.0E-03 Individual LCF Risk per Event 28

Run 1 (CAP17) Conditional, Mean, Individual Prompt Fatality Risk (per Event) CCDF with LHS and MC Sampling 1

01 0.1 CCD DF 0.01 0-1.3 miles CAP17-LHS 0-1.3 miles CAP17-MC 0 2 miles CAP17 0-2 CAP17-LHS LHS 0-2 miles CAP17-MC 0-3.5 miles CAP17-LHS 0.001 0 001 0-3.5 miles CAP17-MC 1.0E-12 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 Individual Prompt Fatality Risk per Event 29

10-mile Conditional, Mean, Individual LCF Risk (per Event) CDF for Run 15 (CAP37) and 95%

Confidence Interval Upper and Lower Bounds for Runs 16 & 17 (CAP38 & 39) with SRS 30

50-mile Conditional, Mean, Individual LCF Risk (per Event) CDF for Run 15 (CAP37) and 95%

Confidence Interval Upper and Lower Bounds for Runs 16 & 17 (CAP38 & 39) with SRS 31

MELCOR Parameters of Interest

SRVLAM - SRV stochastic failure to reclose 33

CHEMFORM - Iodine and cesium fraction Parameter Distribution CHEMFORM: Five alternative combinations of RN Discrete distribution classes 2, 4, 16, and 17 (CsOH, I2, CsI, and Cs2MoO4) Combination #1 = 0.125 Combination #2 = 0.125 0 125 Note the fraction cesium below represents the Combination #3 = 0.125 distribution of 'residual' cesium which is the mass of Combination #4 = 0.125 cesium remaining after first reacting with the amount of iodine assumed to form CsI.

CsI Combination #5 = 0.500 0 500 Five Alternatives Species (MELCOR RN Class)

CsOH (2) I2 (4) CsI (16) Cs2MO4 (17) fraction iodine -- 0.03 0 03 0.97 0 97 --

Combination #1 fraction cesium 1 -- -- 0 fraction iodine -- 0.002 0.998 Combination #2 fraction cesium 0.5 -- -- 0.5 fraction iodine -- 0 00298 0.00298 0 99702 0.99702 --

Combination #3 fraction cesium 0 -- -- 1 fraction iodine -- 0.0757 0.9243 --

Combination #4 fraction cesium 0.5 -- -- 0.5 fraction iodine -- 0 0277 0.0277 0 9723 0.9723 --

Combination #5 fraction cesium 0 -- -- 1 Fraction iodine -- 0.0 1.0 --

SOARCA estimate Fraction cesium 0.0 -- -- 1.0 34

FL904A - Drywell liner failure flow area 35

BATTDUR - Battery Duration 36

SRVOAFRAC - SRV open area fraction 37

SLCRFRAC - Main steam line creep rupture area fraction 38

Radial debris relocation time constants - RDSTC (solid) 39

Radial debris relocation time constants - RDMTC (liquid) 40

RRIDRFRAC, RODRFRAC - Railroad door open fraction 41

H2IGNC - Hydrogen ignition criteria 42

RHONOM - Particle density 43

FFC - Fuel failure criterion 44

FFC - Fuel failure criterion (continued) 45

SC1141(2) - Molten clad drainage rate 46

Other MELCOR Items of Interest

  • Surrogate parameters
  • Lower L h head d penetration t ti ffailures il
  • Drywell D ll liliner ffailure il model d l
  • Operator actions 47

MACCS2 Parameters of Interest

DOSNRM, DOSHOT - Normal and Hotspot Relocation Doses 49

TIMNRM, TIMHOT - Normal and Hotspot Relocation Times 50

ESPEED - Evacuation speed 51

GSHFAC - Groundshine Shielding Factor 52

Next Steps p

  • ANS PSA Conference presentation and papers and CSARP presentation

- September 2013

Questions and Comments Note that all results in these presentation slides are conditional (per event) on the potential occurrence of a long-term long term station blackout (LTSBO) scenario, and modeling the SOARCA unmitigated LTSBO.

Th LTSBO scenario The i frequency f is i estimated ti t d in SOARCA to be ~3x10-6 per reactor year.

Retrieved from "https://kanterella.com/w/index.php?title=ML13255A376&oldid=2455396"