NEI 12 16, “Guidance for Performing Analyses of Fuel ...Slide subtitle 18pt Using Measured Flux Map Data To Determine Fuel Burnup Reactivity Uncertainty Kord S. Smith (MIT) Albert

1

NEI 12‐16, “Guidance for Performing Criticality Analyses of Fuel Storage at Light‐

Water Reactor Power Plants”

NRC/NEI Meeting on SFP Criticality GuidanceFebruary 19th 2014 • Rockville, MD

2

Introduction and Meeting Objectives

Kristopher CummingsSenior Project Manager, Used FuelNRC/NEI Meeting on SFP Criticality Guidance

February 19th, 2014 • Rockville, MD

3

AgendaTime Subject Presenter

8:30 am Welcome, Introductions and Meeting Purpose Kris Cummings, NEI

8:45 am Overview of EPRI Depletion Benchmarks Albert Machiels, EPRI

9:00 am Use of Measured Flux Map Data to Determine Depletion Uncertainty

Kord Smith, MIT

10:15 am Break

10:30 am Application of EPRI Depletion Method Dale Lancaster, NuclearConsultants.com

11:00 am Isotopic Benchmarking Bob Hall, Dominion

12:00 am Lunch

1:00 pm BWR Cold Critical Experiments Kristen Bennett, GEH

2:00 pm Burnup Uncertainty Andrew Blanco, Westinghouse

4

Agenda (cont.)Time Subject Presenter

2:30 pm Content of Technical Specification Kris Cummings, NEI

3:00 pm Break

3:15 pm Multiple Misloads Kris Cummings, NEI

4:15 pm Eccentric Positioning Uncertainty Kris Cummings, NEI

4:45 pm Summary of Actions Captured/Overview of Next step

Kris Cummings, NEI

5

Meeting Purpose

• Reach resolution on methods to be used in spent fuel pool criticality analysis.

• Focus on issues around depletion code benchmarking, Tech Spec content, uncertainties and multiple misloads.

• Identify areas needing additional description, justification or explanation in NEI 12‐16.

Albert J. MachielsSenior Technical Executive

NRC/NEI Meeting on SFP Criticality GuidanceFebruary 19, 2014

Rockville, MD

PWR Fuel Reactivity Depletion Benchmarks Using Flux Map Data

A Roadmap to Available Documents

2© 2014 Electric Power Research Institute, Inc. All rights reserved.

Contents

• Contributors

• Motivation

• Main Documents

• Interactions with OECD/NEA

• MIT’s BEAVRS

• Concluding Remarks


Contributors

• Prof. Kord Smith, MIT (Presenter)– Prof. Benoit Forget– Geoffrey Gunow– Bryan Herman– Nick Horelik

• Dr. Dale Lancaster, NuclearConsultants.com (Presenter)• Studsvik Scandpower Team

– Mehdi Asgari– Tamer Bahadir– Rodolofo Ferrer

• Made possible by Duke Energy (44 cycles of operational data from four reactors)


Problem StatementExisting criticality benchmarks have been portrayed as

insufficient by regulators in light of operational/licensing changes being sought by utilities

Depletion uncertainty approaches that could increase licensee flexibility in addressing spent fuel criticality

concerns

Desired Outcome


Depletion Uncertainty Quantification

Quantification through critical benchmarks and chemical

assays

Leverage operational data

to generate benchmarks

EPRI approach ORNL approach


Fuel Reactivity Depletion Benchmarks1. “Benchmarks for Quantifying Fuel Reactivity Depletion

Uncertainty,” EPRI Report 1022909 (August 2011)http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000000001022909

ICNC 2011 “Experimental Benchmarks For Quantifying Fuel Reactivity Depletion Uncertainty” by K. Smith et al. (Edinburgh, Scotland, Sept. 19 – 22, 2011)

K. Smith et al., “Depletion Reactivity Benchmarks – Part 1: Experimental Benchmarks For Quantifying PWR Fuel Reactivity Depletion Uncertainty,” Nuclear Technology, Vol. 185, pp. 39-56 (January 2014)

2. “Utilization of the EPRI Depletion Benchmarks for Burnup Credit Validation,” EPRI Report 1025203 (April 2012)http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000000001025203ICNC 2011 “Utilization Of The EPRI Fuel Reactivity Depletion Benchmarks In PWR Spent Fuel Pool Criticality Analysis” by D. Lancaster et al. (Edinburgh, Scotland, Sept. 19 – 22, 2011)

D. Lancaster et al., “Depletion Reactivity Benchmarks – Part 2: Utilization In PWR Spent Fuel Pool Criticality Analysis,” Nuclear Technology, Vol. 185, pp. 57-70 (January 2014)


International Handbook of Evaluated Reactor Physics Benchmark Experiments (IRPhEP)• June 2012 – January 2013: OECD/NEA Review

– Focus should be on Hot Full Power (HFP) conditions• Cold, Cooled Reactivity Depletion Benchmarks (Appendix B).

– Use of CASMO-SIMULATE as a reactivity meter• Would other tools give the same results?

– Derivation of uncertainties• Completeness?• Provide data set?

– OECD/NEA Report formatting• Improvements required.

• Published as a “draft” in the “International Handbook of Evaluated Reactor Physics Benchmark Experiments,” NEA Nuclear Science Committee, NEA/NSC/DOC (2006)1, May 2013 Edition. (ISBN 978-92-64-99208-5). Entry: DUKE-PWR-POWER-001


MIT’s BEAVRS Project Opportunity: U. S. Utility made

available two cycles of detailed, measured PWR operational data

N. Horelik, B. Herman, B. Forget, and K. Smith. “Benchmark for Evaluation and Validation of Reactor Simulations (BEAVRS), v1.0.1,” Proc. Int. Conf. Mathematics and Computational Methods Applied to Nuc. Sci. & Eng., Sun Valley, Idaho, (May 2013)

Benchmark for Evaluation and Validation of Reactor Simulations,Rev. 1.1.1, MIT Computational Reactor Physics Group (October 31, 2013). Available at the MIT website: http://crpg.mit.edu/pub/beavrs


MIT and OECD/NEA Projects Using BEAVRS

• Completed– “PWR Fuel reactivity Depletion Verification and Uncertainty

Using Flux Map Data,” EPRI Report 3002001948 (January 17, 2014)http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000003002001948

• Ongoing– MIT

• Focus on perturbation protocol and treatment for quantifying uncertainties

• Planned– Joint MIT/OECD-NEA project

• MIT: CASMO/ENDF/SIMULATE• OECD/NEA: LWRWIMS/JEFF/PANTHER


Other Publications

• Submitted to PHYSOR 2014– “SOLUTION OF THE BEAVRS BENCHMARK USING THE nTRACER

DIRECT WHOLE CORE TRANSPORT CODE,” by Min Ryu (Seoul National University) et al. “This new PWR benchmark addresses the shortcomings of the previous light water reactor (LWR) benchmarks by providing highly-detailed as-is design data and also measurement data for two cycles. Since most institutions cannot have access to this kind of utility proprietary information, this benchmark is very much valuable.”

• “ANALYSIS OF SELECT BEAVRS PWR BENCHMARK CYCLE 1 RESULTS USING MC21 AND OPENMC” by D. J. Kelly (Knolls Atomic Power Laboratory) et al.

–Presented and published at NCSD 2013• “EPRI DEPLETION BENCHMARK CALCULATIONS USING

PARAGON” by V.N. Kucukboyaci (Westinghouse)


Concluding Remarks

• Experimental benchmarks of burnup reactivity decrement, which are independent of analysis codes, have been created– Benchmarks can be used to determine bias and

uncertainty of applicant’s specific analysis tools• Ongoing work ongoing at MIT provides opportunities for

other analysts to independently verify the approach retained for creating the depletion reactivity benchmarks– Planned collaboration with OECD/NEA

• The team is looking forward to regulatory feedback as part of the review of NEI’s Guidance for Performing Criticality Analyses of Fuel Storage at Light-Water Reactor Power Plants


Other EPRI-sponsored Contributions Not Cited in the Main Presentation– “Utilization of the EPRI Depletion Benchmarks for Burnup

Credit Validation” by D. Lancaster and A. Machiels, PHYSOR 2012 (Knoxville, TN, April 15 – 20, 2012)

− “Benefits of the Delta k of Depletion Benchmarks,” NEI Used Fuel Management Conference, by D. Lancaster and A. Machiels (St. Petersburg, FL, May 9, 2012)

− “Depletion Reactivity Benchmarks Derived from Measured Pressurized Water Reactor Flux Maps” by A. Machiels, RIC 2013 (Washington DC, March 13, 2013)

− “Use of EPRI Depletion Benchmarks for Transport Criticality Burnup Credit ,” by D. Lancaster and A. Machiels, PATRAM 2013 (San Francisco, CA, August 18 – 23, 2013)

− “Benefits of Using EPRI Depletion Reactivity Benchmarks for Burnup Credit Validation ,” by D. Lancaster and A. Machiels, NCSD 2013 (Wilmington, NC, September 29 – October 3, 2013)

Slide title 36pt

Slide subtitle 18pt

Using Measured Flux Map Data To Determine Fuel Burnup Reactivity Uncertainty

Kord S. Smith (MIT) Albert Machiels (EPRI) Hatice Akkurt (EPRI) Presentation to NRC Feb. 19, 2014

Slide title

36 pt

Text 24 pt

Bullets level 2 20 pt

2

What is the Bottom Line?

•  EPRI experimental benchmarks of burnup reactivity decrement have been created which are independent of analysis codes.

•  Total uncertainty of cold experimental fuel reactivity decrements are < 650 pcm (< 0.65% delta-k).

•  EPRI benchmarks can be used to determine biases and uncertainty of applicant’s specific analysis tools.

Feb. 19, 2014 Using Measured Flux Map Data to Determine Fuel Burnup Reactivity Uncertainty

Slide title

36 pt

Text 24 pt


3

Outline

•  Overview: flux maps and assembly reactivities

•  Duke Energy’s PWR models/data/qualification with Studsvik CMS

•  EPRI methodology: implementation details

•  HFP reactivity decrement biases and uncertainties

•  HFP to cold uncertainties with TSUNAMI

•  Cold biases and uncertainties

•  Reactivity decrement benchmarks

•  Summary


Slide title

36 pt

Text 24 pt


4

Flux Maps: In-Core Fission Rate Measurements •  Miniature fission chambers are inserted

in the central instrument tubes of selected assemblies.

–  BOC calculations required by NRC to be within a prescribed tolerance of measurement - to assure core loading.

–  Required every 30 days by NRC to guarantee that the core is operating within design margins

–  Measurement used to calibrate certain safety systems (e.g., ex-core detectors)

•  High precision (<1% statistical error) measurement of the core-wide distribution of fission rates.

•  Historical flux maps are used in NRC submittals to determine 95/95 confidence intervals on predicted core power distributions and assembly burnups.


R P N M L K J H G F E D C B A0.499 0.4830.491 0.4830.008 0.000

0.369 0.998 0.9960.389 1.008 0.991-‐0.020 -‐0.006 0.004

1.195 1.081 0.998 0.3701.175 1.069 0.985 0.3820.020 0.012 0.012 -‐0.012

0.610 0.997 1.3150.627 0.994 1.323-‐0.017 0.003 -‐0.008

1.309 1.330 1.309 1.1771.298 1.331 1.304 1.1700.010 -‐0.001 0.005 0.007

0.484 1.082 1.3310.479 1.071 1.3320.004 0.011 -‐0.001

1.191 1.276 1.236 1.2701.179 1.287 1.224 1.2740.011 -‐0.011 0.011 -‐0.004

0.475 1.196 1.232 1.209 1.317 1.195 0.9960.478 1.183 1.224 1.206 1.319 1.178 0.995-‐0.003 0.013 0.008 0.003 -‐0.002 0.017 0.001

0.958 1.278 1.332 0.4970.951 1.272 1.317 0.5010.007 0.006 0.014 -‐0.004

1.2961.298-‐0.002

0.355 1.306 1.233 1.308 0.3540.372 1.323 1.212 1.305 0.370-‐0.018 -‐0.017 0.020 0.004 -‐0.015

1.294 1.192 1.1701.305 1.182 1.173-‐0.011 0.009 -‐0.003

0.876 1.170 1.196 0.3690.876 1.188 1.183 0.3830.000 -‐0.018 0.013 -‐0.0140.369 0.956 1.000 0.6110.395 0.938 0.996 0.642-‐0.026 0.018 0.003 -‐0.032

0.354 0.4750.366 0.476-‐0.012 -‐0.001

S3-‐MEAS

13

14

15Reaction RateSIMULATE-‐3MEASURED

12

1

2

3

4

5

6

7

8

9

10

11

Slide title

36 pt

Text 24 pt


5

Basis For The EPRI Uncertainty Methodology

1.  Critical LWR cores provide a continuous source of measured reactivities for both fresh and depleted fuel assemblies at full-power conditions.

2.  Predictions of in-core reaction rate spatial distributions are very sensitive to the accuracy of computed assembly reactivities.

3.  Evaluation of the (batch-wise) fuel assembly reactivity that minimizes the average (r.m.s.) difference between computed and measured reaction rates is used to deduce the error in computed assembly reactivities.

4.  By examining hundreds or thousands of in-core measurements, the burnup dependence of the error in computed assembly reactivities (and its uncertainty) can be deduced.


Slide title

36 pt

Text 24 pt


6

Simplest of Nodal Core Models: FLARE Model


6

1

( ),p

p p qpp qp

q

kS w S w Sλ∞

=

= +∑

2 2 2

6

1

2

(1 )( / 2 ) ( / )

1

/

pq p p

pp pqq

p p p

w g M h g M h

w w

M D=

= − +

= −

= Σ

∑

•  Source (e.g. Power, Detector Signal, etc.) are sensitive to reactivity of the assembly.

•  PWR Migration areas are almost a constant (e.g., not very uncertain).

•  Chao’s diffusive homogeneity/backward diffusion theory use the same principles. •  Modern nodal codes are also very sensitive to k-infinity of the fuel assembly.

Slide title

36 pt

Text 24 pt


7

Implementation of EPRI Methodology •  Assumptions:

–  Assembly reactivity is modified by evaluating nodal parameters (e.g., group cross sections) by multiplying nodal burnups by an arbitrary multiplier, MB

–  A single multiplier is used for all radial and axial zones of each sub-batch

–  Discrete values of burnup multiplier are searched to find the value that minimizes the difference between computed and measured reaction rates


reactivity error = k-infinity (MB=Mmin) minus k-infinity (MB=1.0)

0.7000

0.8000

0.9000

1.0000

1.1000

1.2000

1.3000

0.00 20.00 40.00 60.00 80.00

Assembly k-infinity

Burnup (GWd/T)

Slide title

36 pt

Text 24 pt


8

Sensitivity of Flux Maps To Sub-Batch Reactivity


R P N M L K J H G F E D C B A-‐0.005 -‐0.0230.008 0.0000.020 0.022

-‐0.046 -‐0.045 0.012-‐0.020 -‐0.006 0.0040.006 0.032 -‐0.003

0.032 -‐0.017 -‐0.042 -‐0.0390.020 0.012 0.012 -‐0.0120.007 0.039 0.064 0.013

-‐0.056 -‐0.051 0.061-‐0.017 0.003 -‐0.0080.020 0.055 -‐0.062

0.032 0.069 0.027 -‐0.0420.010 -‐0.001 0.005 0.007-‐0.010 -‐0.066 -‐0.015 0.055

-‐0.013 -‐0.018 0.0380.004 0.011 -‐0.0010.026 0.038 -‐0.040

0.039 0.031 0.064 -‐0.0110.011 -‐0.011 0.011 -‐0.004-‐0.014 -‐0.053 -‐0.039 0.002

-‐0.008 0.025 0.080 0.042 0.057 0.029 0.009-‐0.003 0.013 0.008 0.002 -‐0.002 0.017 0.0010.001 0.001 -‐0.059 -‐0.037 -‐0.056 0.005 -‐0.007

-‐0.009 0.048 0.084 -‐0.0170.007 0.006 0.014 -‐0.0040.022 -‐0.036 -‐0.051 0.009

-‐0.005-‐0.0020.001

-‐0.039 0.005 0.092 0.026 -‐0.036-‐0.018 -‐0.017 0.020 0.004 -‐0.0150.003 -‐0.037 -‐0.047 -‐0.016 0.005

-‐0.015 0.037 -‐0.048-‐0.012 0.009 -‐0.003-‐0.009 -‐0.016 0.040

-‐0.057 -‐0.068 0.025 -‐0.0410.000 -‐0.018 0.013 -‐0.0140.054 0.029 0.001 0.011-‐0.052 0.002 -‐0.036 -‐0.070-‐0.026 0.018 0.003 -‐0.0310.000 0.033 0.041 0.006

-‐0.033 -‐0.005

-‐0.012 -‐0.001 MB=0.9 4.2%0.008 0.003 MB=1.0 1.2%

MB=1.1 3.3%

6

1

2

3

4

5

13

14

r.m.s. diff15

7

8

9

10

11

12

Slide title

36 pt

Text 24 pt


9

Each Flux Map Produces One Data Point Per Sub-Batch


•  Each flux map results in one burnup multiplier that produces the minimum r.m.s. difference between calculated and measured reaction rates

•  The sub-batch multipliers are then converted to reactivity error at each flux map

reactivity error = k-infinity (MB=Mmin) minus k-infinity (MB=1.0)

Slide title

36 pt

Text 24 pt


10

Combing All Flux Map Data Produces Sub-Batch History


Slide title

36 pt

Text 24 pt


11

Analysis Procedure •  For each Flux Map:

- 10 iterative passes to find minimum r.m.s. difference of calculated

and measured reaction rates

- Loop over all sub-batches in the core - Loop over burnup multipliers, MB, in increments of 0.01 (from 0.85 to 1.15) - Save MB that minimizes r.m.s. difference

•  > 4000 SIMULATE-3 calculations per flux map


Slide title

36 pt

Text 24 pt


12

Duke Energy’s Reactor Data


Table 5.1 Reactor and Fuel Data

Unit

Cycles

Cycle Length (EFPD)

Enrichment Range

(%)

HZP Boron (ppm)

Maximum LBP

# Pins

Maximum IFBA

# Pins

Maximum WABA+IFBA

# Pins McGuire-1 10 to 21 363 - 514 3.40 - 4.95 1576 - 2000 24 128 24 + 128 McGuire-2 10 to 20 428 - 512 3.64 - 4.90 1690 - 2037 24 128 24 + 128 Catawba-1 9 to 19 407 - 522 3.45 - 4.75 1501 - 2104 24 128 16 + 128 Catawba-2 8 to 17 451 - 527 3.50 - 4.90 1819 - 2109 24 128 20 + 128

Table 5.2 Feed Fuel Characteristics

Cycle Enrichment (%)

# of sub batches Cycle Enrichment

(%) # of sub batches

McGuire-1 McGuire-2 10 3.40 4 10 3.85 / 3.95 7 11 3.40 / 3.55 9 11 3.90 / 4.15 7 12 3.67 9 12 3.78 6 13 3.92 8 13 4.09 / 4.39 11 14 4.14 / 4.50 7 14 3.77 / 4.33 7 15 4.40 / 4.75 6 15 4.16 / 4.56 12 16 3.92 / 4.35 7 16 4.37 / 4.67 8 17 4.45 / 4.74 9 17 4.35 / 4.75 8 18 4.01 / 4.64 5 18 4.05 / 4.70 6 19 4.00 / 4.68 7 19 3.90 / 4.80 8 20 4.00 / 4.85 9 20 3.65 / 4.90 9 21 3.60 / 4.95 10

Catawba-1 Catawba-2 9 3.86 8 8 3.98 6 10 3.65 / 3.92 6 9 4.32 / 4.42 9 11 4.02 6 10 4.54 5 12 4.50 6 11 3.90 / 4.20 6 13 3.81 / 4.31 6 12 4.35 / 4.66 8 14 4.19 / 4.46 9 13 4.00 / 4.75 8 15 4.18 / 4.53 10 14 4.45 /4.75 8 16 4.42 / 4.67 11 15 3.80 / 4.73 11 17 3.88 / 4.51 9 16 4.38 / 4.90 8 18 4.05 / 4.51 8 17 3.80 / 4.82 9 19 3.96 / 4.75 7

Slide title

36 pt

Text 24 pt


13

CASMO/SIMULATE Models Set Up from Design Data


0

200

400

600

800

1000

1200

1400

1600

0 100 200 300 400 500 600

Bor

on C

once

ntra

tion

(ppm

)

Cycle Burnup (EFPD)

Fig. 5.1 Comparison of SIMULATE and Measured Boron Letdowns for Catawba Unit-2, Cycle 13

Measured

Comparison of CASMO-4/SIMULATE-3 and Measured Reaction Rates

Reactor

# of Flux Maps

Mean Radial (2D) r.m.s. Difference

(%)

Mean Nodal (3D) r.m.s. Difference

(%) McGuire-1 164 1.22 2.56 McGuire-2 160 1.16 2.78 Catawba-1 174 1.23 2.51 Catawba-2 169 1.26 2.91

Average 1.22 2.69

Studsvik CMS INTERPIN-4 (1D Fuel Pin) CASMO-4 (2D Lattice) CASMO-5 (2D Lattice) CMSLINK (Data linking) SIMULATE-3 (3D Core/T-H)

Slide title

36 pt

Text 24 pt


14

Question: Minimize 2D or 3D r.m.s. Differences?


•  2D and 3D r.m.s. minimizations examined – no significant

difference in Burnup Multipliers observed •  Final implementation:

•  Search on 3D r.m.s. difference to find best burnup multiplier •  2D r.m.s. are improved - as result of 3D r.m.s. minimization

Slide title

36 pt

Text 24 pt


15

Measured Sub-Batch Reactivity Data Points


•  4 reactors, 44 cycles, 667 flux maps, 1 million measured signals •  2.7 million CASMO-4/SIMULATE-3 cases •  Produced >2500 sub-batch measured reactivities

reactivity decrement error = k-infinity (MB=Mmin) minus k-infinity (MB=1.0)

Slide title

36 pt

Text 24 pt


16

Sources of Spread in Data Points


•  Errors in computed sub-batch burnups (used as the plot ordinate) •  Reaction rate measurement errors or uncertainties •  Imperfect knowledge of reactor configuration (fuel bowing, crud) •  Differences in sub-batch spectra vs. CASMO lattice assumption (zero leakage) •  Differences in intra-assembly spatial flux distributions vs. lattice assumption •  Errors in SIMULATE-3 nodal and detector physics models •  Errors in SIMULATE-3 cross section data fitting models •  Thermal-hydraulic uncertainties

•  Imprecision in iterative analysis method

(particularly for low sensitivity sub-batches) •  Discrete interval for burnup multiplier search

•  Errors in CASMO nuclide number densities and

neutron cross sections

Slide title

36 pt

Text 24 pt


17

Eliminate Low Sensitivity Points and Batches < 12


Slide title

36 pt

Text 24 pt


18

•  Linear and quadratic regression examined •  Constrained to 0.0 decrement error at 0.0 GWd/T •  Variance assumed to be linear in burnup (heteroscedastic) •  Confidence intervals here assume total correlation within sub-

batch/cycle (each measurement is not independent)

Reactivity Bias Determined From Regression Fits


Slide title

36 pt

Text 24 pt


19

Linear and Quadratic Regressions Are Similar


Burnup (GWd/T) 10 - 20 20 - 30 30 - 40 40 - 50 50 - 60 # of Data Points 176 458 809 357 47

Slide title

36 pt

Text 24 pt


20

Small Differences Between Individual Reactor Units


Slide title

36 pt

Text 24 pt


21

No Significant Differences vs. Enrichment


Slide title

36 pt

Text 24 pt


22

•  Create a library with known defect - no lumped fission product absorption •  Perform all 2.7 million SIMULATE cases with defective library •  Perform regression fit to measured sub-batch decrement error •  Compare measured decrement with known decrement (e.g., compute error)

The reactivity decrement of the library is measured to within 145 pcm.

Do Regression Fits Accurately Determine Reactivity Bias?


0

20

40

60

80

100

120

140

160

180

200

0.00 10.00 20.00 30.00 40.00 50.00 60.00

Difference (pcm)

Burnup (GWd/T)

Difference in Reactivity Decrement Errors (Measured vs. Known Decrement Change)

Slide title

36 pt

Text 24 pt


23

CASMO HFP Reactivity Bias vs. Critical Borons •  Measured reactor boron concentration data at HFP

•  Flux map data is consistent with BOC to EOC boron bias •  Small unit dependencies are consistent with flux map data •  145 pcm error from defective library test is consistent with

both boron data and flux map regression data


Slide title

36 pt

Text 24 pt


24

CASMO HFP Reactivity Bias and Uncertainty •  Use quadratic regression fits to determine bias •  Conservatively assume:

–  145 pcm difference is taken as a one-sigma uncertainty –  2-sigma uncertainty applied independent of fuel burnup

Real uncertainty is much closer to confidence than prediction intervals


Table 7.1 Measured CASMO Reactivity Decrement Bias (change needed to match measurement) Burnup (GWd/T) 10.0 20.0 30.0 40.0 50.0 60.0 CASMO-4 Bias (pcm) 92 153 182 181 148 84 CASMO-5 Bias (pcm) 22 49 78 110 146 185 Uncertainty (pcm) 290 290 290 290 290 290

Slide title

36 pt

Text 24 pt


25

HFP to Cold Uncertainties: TSUNAMI Analysis •  Measured CASMO HFP uncertainties account for both:

–  Isotopic inventory (number density) uncertainties –  Cross section uncertainty (at hot conditions) of k-infinity

•  Isotopic inventory not changed in going to cold conditions, assume bias due to isotopics is the same at HFP and cold

•  Use CASMO number densities vs. lattice burnup •  Compute cross section uncertainty of reactivity with TSUNAMI


Burnup (GWd/T) 0.5 10. 20. 30. 40. 50. 60. k-inf Hot 1.02068 1.08597 1.08298 1.03600 0.97772 0.92176 0.87200

Uncertainty (pcm) 1034 1104 1211 1265 1295 1311 1323 k-inf Cold 1.16103 1.23979 1.25419 1.21057 1.14557 1.07995 1.01883

Uncertainty (pcm) 1015 1087 1209 1281 1326 1361 1388

Slide title

36 pt

Text 24 pt


26

HFP to Cold Uncertainties: TSUNAMI Analysis (2) •  HFP and cold k-infinity uncertainties are highly correlated •  Use TSUNAMI tools to account for correlations on HFP to cold

reactivity change


Burnup (GWd/T) 0 0.5 10. 20 30. 40. 50. 60. Uncertainty (pcm) 347 427 459 508 527 530 521 509

Correlation Coefficient, ck, vs. Burnup (GWd/T)

0 0.5 10 20 30 40 50 60

Bas

e

HFP 1 1 1 1 1 1 1 1 HFP No Xe 1 0.9843 0.9817 0.9831 0.9854 0.9878 0.9898 0.9914 HZP 0.9992 0.9992 0.9992 0.9995 0.9994 0.9994 0.9995 0.9995 Cold 1000 ppm 0.9807 0.9609 0.9608 0.9646 0.9678 0.9702 0.9723 0.9734 Cold 0 ppm 0.9694 0.9506 0.9514 0.9560 0.9592 0.9611 0.9634 0.9640 Decay 100 hr 0.9705 0.9512 0.9519 0.9558 0.9591 0.9613 0.9633 0.9643 Cold in Rack 0.9717 0.9543 0.9510 0.9543 0.9582 0.9605 0.9623 0.9641

Slide title

36 pt

Text 24 pt


27

•  Assume largest observed uncertainty applies to all lattices

•  Cold BOL uncertainty of SFP/cask analysis come from criticals,

so we need the additional uncertainty in HFP to Cold reactivity

HFP to Cold Rack Reactivity vs. Burnup (GWd/T)

Burnup (GWd/T)

0.0 10. 20. 30. 40. 50. 60.

Base Reactivity (pcm) -10145 -6291 -4658 -4002 -4097 -4601 -5344 Uncertainty 222 324 353 358 356 349 339

128 IFBA 24 WABA

Reactivity (pcm) -10405 -6684 -4858 -4094 -4130 -4626 -5347 Uncertainty 212 317 352 364 354 346 341

3.50 % Enrichment


4.25 % Enrichment


Small Fuel Radius


HFP to Cold Uncertainty


Table 8.6 HFP to Cold Uncertainty (2 sigma) vs. Burnup Burnup (GWd/T) 0.0 10.0 20.0 30.0 40.0 50.0 60.0 Uncertainty (pcm) 365 498 555 550 537 521 512

Table 8.7 HFP to Cold Additional Uncertainty (2 sigma) vs. Burnup Burnup (GWd/T) 10.0 20.0 30.0 40.0 50.0 60.0 Uncertainty (pcm) 367 441 435 419 397 385

HFP to Cold Reactivity vs. Burnup (GWd/T)

Burnup (GWd/T)

0.0 10. 20. 30. 40. 50. 60.

Base Reactivity (pcm) 9867 11425 12605 13919 14986 15891 16527 Uncertainty 347 459 508 527 530 521 509

128 IFBA 24 WABA

Reactivity (pcm) 9977 11367 12556 13916 15009 15888 16612 Uncertainty 337 450 508 533 531 521 512

3.50 % Enrichment


4.25 % Enrichment


Small Fuel Radius


Slide title

36 pt

Text 24 pt


28

Correlation Coefficients Between Lattice Types •  All lattices are extremely similar because:

–  U-238 is almost constant –  Structural materials are almost constant –  Enrichment range is very narrow in LWRs


Correlation Coefficient, ck, vs. Burnup (GWd/T)

0 0.5 10 20 30 40 50 60

HFP

Base 1 1 1 1 1 1 1 1 128 I 24 W 0.9989 0.9990 0.9998 0.9999 0.9999 1 1 1 3.5 % Enrichment 0.9927 0.9931 0.9831 0.9870 0.9898 0.9919 0.9939 0.9958 4.25 % Enrichment 0.9985 0.9984 0.9963 0.9969 0.9976 0.9980 0.9984 0.9988 Small Fuel Rad. 0.9999 0.9998 0.9999 0.9999 0.9999 0.9999 0.9999 0.9999

Col

d 0

ppm

D

ecay

100

hr Base 1 1 1 1 1 1 1 1

128 I 24 W 0.9988 0.9986 0.9997 0.9999 1 1 1 1 3.5 % Enrichment 0.9941 0.9939 0.9818 0.9836 0.9860 0.9885 0.9912 0.9937 4.25 % Enrichment 0.9988 0.9987 0.9962 0.9963 0.9968 0.9972 0.9977 0.9983 Small Fuel Rad. 1 0.9999 0.9999 0.9999 0.9999 0.9999 0.9999 0.9999

Col

d R

ack

Geo

met

ry Base 1 1 1 1 1 1 1 1

128 I 24 W 0.9988 0.9992 0.9997 0.9999 1 1 1 1 3.5 % Enrichment 0.9940 0.9941 0.9817 0.9835 0.986 0.9884 0.9910 0.9936 4.25 % Enrichment 0.9988 0.9988 0.9961 0.9963 0.9968 0.9972 0.9977 0.9982 Small Fuel Rad. 1 0.9999 0.9999 0.9999 0.9999 1 0.9999 0.9999

Slide title

36 pt

Text 24 pt


29

Fuel Temperature Uncertainty


Table 8.1 Fuel Temperature Effect on Hot and Cold Lattice Reactivity

Hot Depletion (HFP) Branch to Cold (Bor=0, Xen=0, 293K)

Burnup (GWd/T)

k-inf (946K)

k-inf (897K)

Difference (pcm)

946K - 897K k-inf

(from 946K) k-inf

(from 897K) Difference

(pcm) 946K - 897K

0.0 1.07825 1.07944 -‐119 1.14864 1.14864 +0 10.0 1.13287 1.13415 -‐128 1.19546 1.19542 +4 20.0 1.13405 1.13536 -‐131 1.20511 1.20492 +19 30.0 1.08500 1.08601 -‐101 1.15734 1.15679 +55 40.0 1.02438 1.02499 -‐61 1.09117 1.09019 +98 50.0 0.96681 0.96699 -‐18 1.02544 1.02399 +145 60.0 0.91598 0.91577 +21 0.96580 0.96391 +189

•  HFP and cold reactivities have uncertainty because HFP fuel temperatures are not known exactly

•  Assume 100% of INTERPIN-4 changes in temperature, from 0.0 to 60 GWd/T, can be taken as an additional uncertainty

•  Assume maximum instantaneous and maximum history changes

are a 2σ uncertainty and apply to all burnups •  Computed at nominal 5.58 kw/ft linear heat rate - assume linear

dependence for other power densities

Slide title

36 pt

Text 24 pt


30

Experimental Benchmarks For Reactivity Decrements •  Compute CASMO-5 reactivity decrement vs. burnup and cooling interval for each lattice •  Assume CASMO-5 decrement bias is independent of lattice and cooling time •  Add CASMO-5 HFP measured decrement bias from regression fits to determine measured decrements •  Statistically combine uncertainties:

1.  HFP regression fit uncertainty (290 pcm) 2.  Instantaneous fuel temperature uncertainty (131 pcm) 3.  Historical fuel temperature uncertainty (189 pcm) 4.  HFP to cold cross sections uncertainty from TSUNAMI (367 to 441 pcm)


Measured Cold Reactivity Decrement Bias and Uncertainty Burnup (GWd/T) 10.0 20.0 30.0 40.0 50.0 60.0

CASMO-4 Bias (pcm) 92 153 182 181 148 84 CASMO-5 Bias (pcm) 23 49 78 110 146 185 95% Uncertainty (pcm) 521 576 571 559 543 534

Table 13.1 Benchmark Lattice Cases

1 3.25% Enrichment2 5.00% Enrichment3 4.25% Enrichment4 off nominal pin diameter depletion5 20 LBP depletion6 104 IFBA depletion 7 104 IFBA plus 20 LBP depletion8 high boron depletion=1500 ppm9 branch to hot rack (150F coolant/fuel)=338.7K10 branch to high rack boron = 1500 ppm11 high power depletion*(pde, coolant/fuel temp)

Slide title

36 pt

Text 24 pt


31

11 Reactivity Decrement Benchmarks


Case 10 20 30 40 50 601 -‐0.1779 -‐0.2754 -‐0.3589 -‐0.4303 -‐0.4876 N/A2 -‐0.1606 -‐0.2457 -‐0.3218 -‐0.3932 -‐0.4600 -‐0.52063 -‐0.1683 -‐0.2589 -‐0.3393 -‐0.4131 -‐0.4791 -‐0.53534 -‐0.1682 -‐0.2592 -‐0.3404 -‐0.4152 -‐0.4825 -‐0.53975 -‐0.2462 -‐0.2757 -‐0.3400 -‐0.4092 -‐0.4722 -‐0.52616 -‐0.2165 -‐0.2643 -‐0.3371 -‐0.4101 -‐0.4765 -‐0.53347 -‐0.2911 -‐0.2834 -‐0.3383 -‐0.4062 -‐0.4694 -‐0.52398 -‐0.1684 -‐0.2569 -‐0.3343 -‐0.4045 -‐0.4668 -‐0.51979 -‐0.1698 -‐0.2602 -‐0.3398 -‐0.4126 -‐0.4775 -‐0.532610 -‐0.1289 -‐0.2082 -‐0.2803 -‐0.3467 -‐0.4056 -‐0.454911 -‐0.1750 -‐0.2631 -‐0.3392 -‐0.4078 -‐0.4685 -‐0.5200

Table 14.2 Measured Reactivity Decrement

Burnup (GWd/T)

Table 13.1 Benchmark Lattice Cases1 3.25% Enrichment2 5.00% Enrichment3 4.25% Enrichment4 off nominal pin diameter depletion5 20 LBP depletion6 104 IFBA depletion 7 104 IFBA plus 20 LBP depletion8 high boron depletion=1500 ppm9 branch to hot rack (150F coolant/fuel)=338.7K10 branch to high rack boron = 1500 ppm11 high power depletion*(pde, coolant/fuel temp)

Slide title

36 pt

Text 24 pt


32

Sensitivity of Benchmarks to Lattice Code/Library •  Generate benchmarks with different code/libraries and compare:

–  CASMO-5 ENDF/B-VII vs. CASMO-4 (ENDF/B-IV&VI,JEF-1&2)

<2% reactivity decrement dependence on cross section library/code

Benchmarks are insensitive to the lattice code/library


Case 10 20 30 40 50 601 -‐0.65 -‐0.03 0.58 1.11 1.57 N/A2 -‐1.22 -‐0.60 -‐0.19 0.27 0.80 1.333 -‐1.03 -‐0.41 0.04 0.55 1.12 1.604 -‐1.02 -‐0.43 -‐0.01 0.54 1.06 1.575 -‐0.35 -‐0.17 0.07 0.56 1.06 1.556 0.10 -‐0.01 0.15 0.51 0.96 1.397 0.12 0.20 0.27 0.58 1.00 1.408 -‐0.97 -‐0.30 0.15 0.68 1.22 1.679 -‐1.22 -‐0.62 -‐0.13 0.39 0.96 1.4410 -‐1.04 -‐0.37 0.15 0.76 1.39 1.9611 -‐0.45 0.07 0.51 0.99 1.51 1.97

Table 9.1 % Difference in Reactivity Decrement (C4-‐C5) No Cooling

Slide title

36 pt

Text 24 pt


33

Fuel Reactivity Burnup Decrement Summary •  Flux maps provides integral data that has been converted into measured fuel

reactivity decrements, which: –  Include both nuclide inventory and cross section uncertainties –  Have HFP uncertainties < 300 pcm –  Include ~2000 data points, spanning burnups from 10 to 55 GWd/T

•  TSUNAMI uncertainty analysis demonstrates:

–  LWR assemblies at HFP and cold conditions are extremely similar –  LWR fuels are all very similar (enrichment, bps, etc.) –  HFP-to-cold reactivity cross section uncertainties < 450 pcm

•  CASMO-4/CASMO-5 have HFP burnup decrement biases < 200 pcm. •  Flux map reactivity decrement data are consistent with BOC to EOC boron biases.

•  EPRI experimental benchmarks are independent of analysis codes, and they have total uncertainty < 600 pcm (0.6% delta-k)


Slide title

36 pt

Text 24 pt


34

Bottom Line

•  EPRI benchmarks can be used to determine biases and uncertainty of applicant’s SFP or cask criticality tools.

•  Cold reactivity decrement uncertainties far smaller than Kopp’s 5% criteria. (< 2.0% uncertainty is supportable at high burnups)


Slide title

36 pt

Text 24 pt


35 Feb. 19, 2014 Using Measured Flux Map Data to Determine Fuel Burnup Reactivity Uncertainty

Thank You!

1

Application of EPRI Depletion Method

Dale LancasterNuclearConsultants.com

NRC/NEI Meeting on SFP Criticality GuidanceFebruary 19th 2014 • Rockville, MD

2

Topics

• Theory• Bias and Uncertainty • Example Application• Comparison To Other Methods• Advantages of the EPRI Benchmarks for Depletion Reactivity Validation

• Conclusions

3

Theory

• The latest version of ANSI/ANS‐8.27 allows the following equation:kp + kp + kd + kb kc ‐ kc ‐ km

• kc is derived from fresh UO2 critical experiments and addresses the geometric and material concerns not related to burnup.

• kc is the uncertainty in the method derived from the fresh UO2 criticals• km is the administrative margin (typically 5%)

• kp is the system k that has the appropriate modeling of the axial and horizontal burnup variations.

• kp is the uncertainty in the system model (eg manufacturing uncert.)

• kd is the depletion uncertainty covering all non‐spatial changes from the initial UO2 criticals.

• kb is an allowance for uncertainty in kpdue to uncertainty in the assigned burnup value.

4

Depletion Reactivity Bias & Uncertainty

• The Depletion Reactivity Bias and Uncertainty, kd , must be based on measured data.

• The EPRI depletion benchmarks provide that measured data.

• Historically, this approach was taken but with an assumed uncertainty (no bias) of 5% of the reactivity of depletion (Kopp memo).

• Alternatively, rather than use kd , one can use uncertainties in isotopic content and isotopic worth (NUREG/CR‐7108 and 7109)

5

Bias and Uncertainty

• The difference between the calculated delta k of depletion and the EPRI benchmark is a bias.

• Although there are multiple benchmarks each represents a different condition.

• The variation in the calculated bias is not due to a random event but the difference in the condition.

• The uncertainty is the uncertainty in the benchmarks (not the variation in biases).

6


• The temperature model is a large uncertainty in the EPRI benchmarks so the uncertainty in the benchmarks are a function of kw/ft.

• The uncertainty should be a function of burnup but the uncertainties associated with going from hot to cold conditions overwhelmed the burnup dependence. (Due to the small size of the HFPuncertainty a conservative constant uncertainty was selected.)

7


• Two uncertainties are reported:- 0.00576 for 38.1 W/gm U, and - 0.00643 for 57.2 W/gm U

• 57.2 W/gm corresponds to a 1.5 peaking factor.

• Conservatively use 0.0064.

8

Example Application

• EPRI sponsored “Utilization of the EPRI Depletion Benchmarks for Burnup Credit Validation,” EPRI Report 1025203 (April 2012)http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000000001025203

• Used SCALE 6.1 and various libraries

• This example will use 238 Group ENDF/B‐VII library

9

Benchmark Lattice Cases

Benchmark Lattice Cases

1 3.25% Enrichment 2 5.00% Enrichment 3 4.25% Enrichment 4 off-nominal pin diameter depletion 5 20 WABA depletion 6 104 IFBA depletion 7 104 IFBA plus 20 WABA depletion 8 high boron depletion = 1500 ppm 9 branch to hot rack (150°F coolant/fuel) = 338.7K

10 branch to high rack boron = 1500 ppm 11 high power depletion (power, coolant/fuel temp)

High Power is 1.5 times average power of 38.1 W/gm

10

Benchmark Data

Benchmark Uncertainty (same for all cooling)

Measured Reactivity Decrements

Case1234567891011

0.005760.00643

0.005760.005760.005760.005760.00576

0.005760.00576

0.005760.00576

delta‐k

15 years Cooling

5 years Cooling

100 hours Cooling

11

Example of Bias Calculations

Burnup GWD/MTU

Calculated k

Monte Carlo

Calculated Decrement

EPRI Benchmark Decrement

Bias Calc–EPRI

Decrement 0 1.4712 0.0001

10 1.3484 0.0002 0.1228 0.1223 0.0005 20 1.2553 0.0002 0.2159 0.2157 0.0002 30 1.1721 0.0001 0.2991 0.2990 0.0001 40 1.0958 0.0001 0.3754 0.3758 -0.0004 50 1.0267 0.0001 0.4445 0.4445 0.0000 60 0.9688 0.0001 0.5024 0.5029 -0.0005

Case 3 (4.25 wt% U-235)Negative Bias is conservative and will be ignored

12

EPRI Benchmark Biases

13

Case 30 GWD/T

13.25% enrichment

depletion -0.0017

25.00% enrichment

depletion 0

34.25% enrichment

depletion -0.0005

4off-nominal pin

depletion -0.0018

5 20 WABA depletion 0.0001

6 104 IFBA depletion 0.0001

7104 IFBA, 20 WABA

depletion 0.0007

8high boron depletion

= 1500 ppm -0.0001

9branch to hot rack =

338.7K -0.0005

10branch to rack

boron = 1500 ppm -0.0018

11high power density

depletion -0.0005

Higher Enriched Case 2 -3 0.0005Bigger Pin Case 3-4 0.0013WABA + IFBA Case 7-3 0.0012Higher ppm depltete Case 8-3 0.0004Hotter Rack Case 9-3 0.0000Poisoned Rack Case 10-3 0.0000Higher Power Case 11-3 0.0000

Sum of Case 3 plus deltas 0.0029

Several of the effects can happen at the same time.

The bias needs to come from the combined effects

Need to cover all burnups and cooling times.

Final Bias Selected was 0.003

14

Advantages of the EPRI Benchmarks for Depletion Reactivity Validation

• Analysis of EPRI benchmarks match closely the analysis for the pool.- Chemical Assay analysis require special treatment to get pin isotopics – Pool analysis use assembly isotopics.

• The EPRI benchmarks are easy to perform.• All the isotopes and changes with the fuel are contained in the depletion reactivity.

15

Advantages of the EPRI Benchmarks for Depletion Reactivity Validation

• Since individual benchmarks have low biases so the analysis can help improve modeling - The uncertainty associated with each chemical assay is so high little can be learned from an individual assay.

• The documentation is easy to find, including example applications

16

Conclusions

• The EPRI Depletion Benchmarks use power distributions to determine measurement inferred depletion reactivities.

• The depletion reactivities include the change in isotopic concentrations as well as the macro cross sections.

• The EPRI benchmarks are easy to perform and closely match the analysis methods for the spent fuel pool.

17

Conclusions

• The depletion reactivity includes all isotopes and material changes.

• The uncertainties and bias from using the EPRI benchmarks are similar to the much more complex chemical assay analysis.

• The uncertainties and bias are smaller than the Kopp memo.

1

Isotopic and Worth Benchmarking

Bob HallConsulting Engineer, Dominion

NRC/NEI Meeting on SFP Criticality GuidanceFebruary 19th, 2014• Rockville, MD

2

NUREG/CR‐7109

• Addresses adequacy of critical experiment data for actinides and fission products used in burnup credit code validation

• Addresses minor actinides and fission products (burnup credit nuclides not well represented in benchmark criticals)

• Recommends a K‐eff bias of 1.5% of minor actinide and fission product worth to bound cross section uncertainty and bias

• 1.5% applies to all minor actinide and fission product• Applies to all codes that use ENDF‐B/V through ENDF‐B/VII

3

NUREG/CR‐7108• Addresses K‐eff bias and uncertainty associated with predicted burnup credit nuclide concentrations

• Compares radiochemical assay (RCA) measurements of depleted fuel samples to SCALE 6.1 calculated nuclide concentrations

• Provides two methods to determine bias and uncertainty using isotopic validation

• Monte Carlo method• Direct Difference method

• K‐eff bias and uncertainty for Direct Difference are “comparable to, but smaller than” for Monte Carlo

• Monte Carlo method is not accessible for independent analyses

4

NUREG/CR‐7108

• Key Conclusions:

• Using SCALE 6.1 and the ENDF/B‐VII nuclear data results in a small overprediction of K‐eff

• Uncertainties in the calculated U‐235 and Pu‐239 concentrations contribute ~90‐95% of the K‐eff bias uncertainty of all 28 nuclides evaluated

• Uncertainties in calculated FP concentrations are small (< 3% of the K‐eff bias uncertainty)

5

Monte Carlo Method Results

Positive bias means K‐eff is overpredicted

6

Monte Carlo and Direct Difference Results

Positive difference means K‐eff is under‐predicted

7

Observations• K‐eff uncertainty derived by the two methods is similar at discharge burnups

• The burnup bins used in the Monte Carlo approach results in poor resolution of uncertainty as a function of burnup • 3 Major actinide bins: 0‐15, 15‐40, > 40 GWd/T• 2 Minor actinide bins: 0‐40 and >40 GWd/T• Fission products: all burnups lumped together

8

Observations• Results are for SCALE 6.1 and ENDF/B‐VII

• Depletion codes using ENDF/B‐VII give very similar results

• Some variation versus model type• Poisoned rack, un‐poisoned rack, cask

• Uncertainty is dominated by U‐235 and Pu‐239• Results can be extended to more than 28 nuclides

• No allowance for nuclide concentration measurement bias and uncertainty • Overly conservative treatment assigns entire ∆K‐eff to code mis‐

prediction

9

• Isotopic Composition Measurement Uncertainty

• NUREG/CR‐6811 caution is not addressed in NUREG/CR‐7108:

“This source of error [RCA uncertainty] can represent a potentially large source of the overall uncertainty in isotopic validation. Radiochemical analysis of spent fuel is complex and difficult, and the uncertainties associated with the nuclide measurements can be significant.

When computer code predictions are benchmarked against experimental data, the differences between calculations and observation are often incorrectly attributed to the bias and uncertainty of the code, because it is usually not possible to differentiate between code errors and experimental errors.”

10

Chemical Assay Critique• NUREG/CR‐7108 uses 100 assay candidates• Sample N‐9C‐J for H. B. Robinson should be eliminated• Taken from under an Inconel grid

• Some TMI data is questionable • Has no effect if using bounding approach• Increases standard deviation if using statistical method

• Two samples from Calvert Cliffs BT03 were incorrectly analyzed (removed permanent BAs)

11

Chemical Assay Critique• Gosgen sample U3 cooling time is not clear

• Two different cooling times used in two ORNL reports.Believe the NUREG/CR‐7108 cooling times are incorrect

• Vandellos II samples not included in NUREG/CR‐7108 due to late availability. • Have been analyzed for a public domain LAR soon to be issued. Did not change the limiting curve so use of only the NUREG/CR‐7108 minus poor assays is sufficient.

12

Analysis Alternatives• Use direct difference results (exclude some assays)

• Demonstrate consistency with CR‐7108 rack model • Use SCALE 6.1 or demonstrate consistency with SCALE 6.1

burnup worth• Develop conservative ∆K uncertainty vs burnup function• Extend ∆K uncertainty vs burnup function to all nuclides used

• Fraction of burnup credit beyond 28 nuclides is not large• Can convert ∆K uncertainty to % uncertainty if burnup credit ∆K is determined (apply % uncertainty to total burnup credit taken)

13

14

15

Analysis Alternatives• Calculate direct difference (exclude 4 assays)

• Model direct difference cases with depletion model• Calculate direct difference ∆K using LAR rack model • Develop conservative ∆K uncertainty vs burnup function • Extend ∆K uncertainty vs burnup function to all nuclides used

• Fraction of burnup credit beyond 28 nuclides is not large• U‐235 and Pu‐239 differences dominate

16

17

Direct‐Difference Bias and Uncertainty

18

Comparison To Other Methods

• The EPRI benchmark uncertainty is 0.0064 and the bias is 0.003 (burnup independent)

• The isotopic uncertainty at 40 GWd/T is 0.008 (details in later presentation). The bias is 0.0

• The fission product and minor actinide bias is 0.0016.

• Notice the close agreement.• The Kopp Memo would be an uncertainty of 0.019.

19

Summary• Apply K‐eff bias of 1.5% of minor actinide and fission product worth

• Use of direct difference data is inherently conservative due to assignment of all differences to code uncertainty

• ENDF/B‐VII tends to conservatively underestimate burnupcredit

• Direct difference data can be used to derive values for burnup credit uncertainty

• Considering this recent isotopic benchmarking and the EPRI depletion methodology the 5% depletion uncertainty is acceptable.

1

Burnup Uncertainty

Andrew BlancoTechnical Lead – Spent Fuel Pool Criticality Safety

Westinghouse Electric CompanyNRC/NEI Public Meeting on NEI 12‐16

2/19/2014 • Rockville, MD

2

Overview

• Definition• Calculation of Burnup Uncertainty• Size of Burnup Uncertainty• Application of Burnup Uncertainty• Recommendations

3

Definition

• The Burnup Uncertainty, or Burnup Measurement Uncertainty, (BMU) is an uncertainty representing the maximum potential reactivity impact of deviations between an assembly’s “true” and estimated burnup based on reactor records.

• Incorporated in analyses because burnup measurement campaigns to confirm true assembly burnup are not part of used fuel management at site.

4

Calculation of Burnup Uncertainty

• There are a number of ways to calculate BMU, two common methods are:

BMU = 0.0X*(Burnup)*(δk/δBU); BMU = (k(1‐0.0X)BU – kBU)

- X = percent deviation between true and estimated burnup

• X is traditionally assumed to be 5% (0.05)

• Alternatively, utilities can reduce burnup of assemblies by 5% instead of incorporating the uncertainty

5

Size of Burnup Uncertainty

• To calculate BMU, knowledge of an appropriate value for “X” is needed- Historically 5% has been used

• Both EPRI and ORNL have performed studies to evaluate the accuracy of reactor records:- TR‐112054, “Determination of the Accuracy of Utility Spent‐Fuel Burnup Records”, EPRI, 1999.

- NUREG/CR‐6998, “Review of Information for Spent Nuclear Fuel Confirmation,” ORNL, 2009.

6

Size of Burnup Uncertainty (cont’d)

• The EPRI and ORNL reports agree that burnup estimations are within 5% of the true assembly burnup- When using properly calibrated core follow software which is updated with in‐core measurements the uncertainty is less than 2%

7

Application of Burnup Uncertainty

• Burnup Uncertainty has traditionally been a separately assessed uncertainty- RSSed with the other uncertainties in the sum of biases and uncertainties

8

Summary

• Burnup Uncertainty should be accounted for by including a separately assessed uncertainty in the calculation of the sum of biases and uncertainties- Use of 5% is conservative without being unduly penalizing

• Alternatively, discharge burnups can be directly modified to account for the burnup uncertainty.

1

Technical Specification Content

Kristopher CummingsSr. Project Manager, Used Fuel Programs


2

Topics of Discussion

• Plant Systems (Section 3.7)• Design Features (Section 4.0)• Future Tech Spec Changes (after NEI 12‐16 endorsement)

• Summary/Conclusions

3

Plant Systems (Section 3.7)

• LCO for Fuel Storage Pool Boron Concentration:- If LCO found to not be met, required actions are to suspend movement of fuel and initiate actions to restore boron concentration immediately

- Surveillance Requirement:• Frequency is 7 days or in accordance with the Surveillance Frequency Control Program (typically weekly)

4

Plant Systems (Section 3.7)• An LCO for fuel storage loading patterns

- If LCO found to not be met, required action is to immediately initiate action to move non‐complying fuel assembly to an acceptable storage location

- Surveillance Requirements:• Each fuel assembly, fuel insert, neutron absorber insert, hardware shall meet the storage loading patterns in the Design Features section (Section 4.0) prior to storing the fuel assembly in the rack.

5

Design Features (Section 4.0)

• Section 4.3 covers Fuel Storage:- Requirements for new and spent fuel storage, burnup/enrichment, neutron absorber insert locations, fuel inserts (i.e., RCCAs)

- Burnup/enrichment/cooling time requirements should be specified as an equation or in tabular form. Graphical representation (figures) are not necessary and should not be solely relied upon for loading curves

6

Future Tech Spec Changes• Currently the Design Features section of the Tech Spec for the spent fuel pool can be quite large, depending on the number of storage loading patterns

• This is in contradiction with the size/detail associated with the Tech Specs for the reactor.

• Propose developing a “Pool Operating Limits Report” (POLR) to place detailed loading patterns, with Tech Specs reserved for appropriate design criteria:- Max enrichment, keff limits, rack pitch.

7

Summary/Conclusions

• Current approach to Tech Specs (fuel loading patterns and soluble boron level) provide appropriate protection of public health and safety.

• Future efforts to simplify Tech Spec and develop a more flexible approach (i.e., POLR) will be pursued subsequent to the NRC endorsement of NEI 12‐16.

1

Multiple Misloads



2


• Administrative Controls• Event Tree Analysis/Human Factors• Post Movement Verification• Blocking Devices• Multiple Misload Scenarios• Summary/Conclusions

3

Administrative Controls

• Additional procedural controls can be instituted to ensure multiple misloads are prevented.- Validated software implementation to provide:

• Surveillance reports to show acceptability of storage configurations

• Graphical representation to augment manual verification• Pre‐verify planned fuel moves and configuration• Visual (color coded) maps showing acceptability

- Need detailed administrative procedures for implementation

4

Administrative Controls

• To prevent categorization and storage cell selection errors:- Training, qualification and experience of nuclear engineers and operators

- Documentation of fuel assembly characteristics- Operating experience with validated software- Independent verification of validated software output (Fuel Transfer Logs ‐ FTLs)

- Independent categorization of affected assemblies and prep of FTLs by qualified individuals

- Training of qualified individuals in new TS criteria.

5

Event Tree Analysis• EPRI has previously analyzed

the likelihood of a criticality event due to fuel assembly misload in a transportation cask

• This method could be used to assess the event tree and likelihood of a misload in a spent fuel pool.

• Requires a specific storage configuration to be assessed (difficult to assess generically).

Source: EPRI Report 1016635

6

Post Movement Verification

• There is a distinct difference between the appearance of fresh fuel and once‐burned fuel.

• This can be used to restrict the potential misloaded fuel assemblies.

• Post movement verification of low‐burnup fuel assemblies can also be used to restrict the type of misloaded fuel assemblies.

7

Post Movement Verification

• After fuel assemblies are placed in the pool a confirmation of the fuel assembly location is performed.

• Verification of the location of high reactivity fuel (i.e., fresh or once‐burned fuel assemblies) in the spent fuel pool to ensure proper placement.

8

Storage Cell Blocking Devices• Blocking devices are physical hardware installed to prevent

inadvertent placement of a fuel assembly in the storage cell.• Features needed in a blocking device:

- Semi‐permanent (cannot be moved by normal handling tool)- “Lock in‐place” feature to ensure no inadvertent movement of

blocking device.- Ability of blocking device to be moved needs to be balanced with

ability for flexibility in storage configurations- Needs to be simple to ensure cost‐effective implementation- Allows for continued buoyancy driven fluid flow in storage cell- Does not need to prevent a dropped fuel assembly from entering the

storage cell (but single misload needs to address this situation)• Use of blocking devices can be credited to prevent a misloaded

assembly in a location with a blocking device

9

Multiple Misload Scenarios

• The use of admin controls, event tree analysis, post move verification and blocking devices will be used to inform the appropriate misloading scenarios to be addressed via analysis.

• Must be determined on a case‐by‐case basis depending on complexity of storage configurations

• Example: A straightforward Region 1 – Region 2 pool could analyze misloaded once‐burned fuel assemblies in Region 2.

• Demonstrate that keff < 0.95 at Tech Spec soluble boron content to ensure regulatory compliance

10

Summary/Conclusions• Multiple misloads are generally not applicable to BWR racks with maximum reactivity criteria.

• Soluble boron (controlled by Tech Spec LCO) provides the primary safety feature to prevent a criticality event, even in the case of multiple misloads.

• Combination of robust administrative controls, event tree analysis, post move verification and blocking devices can be used to inform a reasonable assumption for multiple misload analysis

1

Eccentric Positioning Uncertainty



2


• Bias versus an Uncertainty• Why Eccentric Positioning is an Uncertainty• Comparison to Other Biases/Uncertainties• Precedence of Eccentricity as an Uncertainty• How to Determine the Effect of Eccentricity• Preliminary Results• Summary/Conclusions

3

Bias versus Uncertainty• Bias has traditionally been defined in the context of the criticality code validation/analysis methods (mean difference between experiment and calculation – Kopp)

• Bias has since been extended to other areas (temperature bias, fuel assembly design) where known effects that effect all assemblies are applied conservatively (versus making a statistical argument for application as an uncertainty).

4

Bias versus Uncertainty• Uncertainty has traditionally been defined in the context of manufacturing tolerances – maximize the parameter of interest within tolerance range (Kopp)

• Effects of uncertainties are statistically combined (square‐root sum of the squares).

• An uncertainty is applied when the parameter of interest can vary (independently) on a per‐unit basis (i.e., each fuel pellet, fuel rod, storage cell, etc.)

5

Why is Eccentric Positioning is an Uncertainty• The location of assemblies in the cell is random within the boundaries of the cell walls.

• How close the assembly can be placed to the storage cell wall is dictated by the fuel assembly envelope (from manufacturing drawings)

• This is directly analogous to the treatment of manufacturing tolerances, where there is a nominal value and a range (plus/minus) of possible values.

• Each assembly placement is independent of the placement of other assemblies.

6

Examples of Other Biases/Uncertainties

Biases Uncertainties

Criticality Code Bias Fuel Assembly Manufacturing Tolerances

Moderator Temperature Effect Uncertainty in Depletion Calculations

Fuel Assembly Type Uncertainty in Calculated Burnup

Storage Cell Bulging Storage Rack Manufacturing Tolerances

Effect of Residual Carbon (Carborundum)

7

Precedence of Eccentricity as an Uncertainty

• NRC has found acceptable the application of eccentric positioning as an uncertainty:- Prairie Island (Sept 2013) – ML11236A134- Palisades (Jan. 2012) – ML110380083- Turkey Point (Nov. 2011) – ML110560336

• Applications prior to 2011 were also approved by the NRC with eccentricity being applied as an uncertainty.

8

How to Determine the Effect of Eccentricity

Reflective/Periodic Boundary Conditions

• Assume 4 assemblies are eccentrically located and determine the reactivity difference from the reference condition (centered).

• The calculated reactivity is the maximum displacement from center and is statistically combined with other uncertainties.

• The limiting condition is often with centered assemblies but both conditions shall be analyzed.

• Larger arrangement (greater than 2x2) of assemblies skewed to a central point is unnecessary since the effect is small and we are already counting 4 random events as one.

9

Preliminary Results • Eccentric positioning is only positive when no fixed neutron absorber is present.

• Going to lattice sizes larger than 2x2 (i.e., 4x4, 8x8, 12x12) shows an asymptotically increasing effect. No additional increase in the 20x20 case.

• Effect is larger for Region 1 (flux‐trap) racks than Region 2

• Maximum effect was ~ 0.0080 k at high enrichment, no burnup (high reactivity).

10

Summary/Conclusions

• Fuel location is random within the storage cell dimensions; appropriate to be applied as an uncertainty.

• NRC has already approved eccentricity as an uncertainty of multiple applications in past and recent applications

• Need to reach agreement on how to model eccentricity to be treated as an uncertainty.