Status of cross section evaluations, covariances and impact on AFC

1
ACE Workshop, NCSU, May 31-June 1, 2006
Status of cross section evaluations, covariances
and impact on AFC
Pavel Oblozinsky
Head of the National Nuclear Data Center
Brookhaven National Laboratory
2
Evaluated Nuclear Data File, ENDF

• Cross Section Evaluation Working Group, CSEWG
• Cooperative effort to the national laboratories,
  industry and universities in the United States
  and Canada responsible for the production of ENDF
  library.
• currently 60 scientists
• 20 laboratories (LANL, BNL, ANL, LLNL, ORNL,,
  Westinghouse, Bettis, )
• National Nuclear Data Center coordinates,
  provides support, archives the library

Historical perspective CSEWG founded 1966
ENDF/B-I released 1968 ENDF/B-II 1970
2y ENDF/B-III 1972 2y
ENDF/B-IV 1974 2y ENDF/B-V 1978 4y
Interest in covariances ENDF/B-VI 1990 12y
Last update in 2001 ENDF/B-VII 2006 16y 40
years of CSEWG ENDF/B-VIII ??? ???
3
Status of ENDF/B-VII

• Beta2 version released for testing in April 2006
• Currently extensive testing and validation
  against integral experiments
• Excellent performance reported so far, though
  not without problems
• Limited attention to specific AFC needs due to
  lack of funding
• Official release of ENDF/B-VII.0 expected later
  in 2006

www.nndc.bnl.gov April 25, 2006
4
Contents of ENDF/B-VII

• ENDF/B cross-section library serves many
  applications (reactor design, advanced fuel
  cycles, waste transmutation, nonproliferation
  national security, nuclear medicine, shielding,
  physics facility design, )
• ENDF/B-VII library contains 14 sublibraries (2 -
  new, 7 - many improvements and updates, 5
  unchanged)

5
ENDF/B-VII beta2 Neutron cross sections

• Contents
• Data for 393 materials (390 isotopes 3
  elements),
• The largest library (Europe 2005 381 materials,
  Japan 2002 337 materials )
• About 2/3 materials are of US origin, 1/3 taken
  over from other sources
• New features
• Major improvements in actinide cross sections
  (thermal fast)
• Fission products completely updated (219
  materials)
• Main US contributors of cross-section evaluations
• LANL (many actinides, 1-H, 16-O,)
• ORNL (resonances for important actinides, )
• BNL (70 fission products,)
• Completeness
• Includes resonances (resolved, unresolved) in
  modern representation
• Includes fast neutron cross sections up to 20 MeV
  or more
• Includes all reaction channels needed for
  transport calculation

6
Validation of ENDF/B-VII

• Overall performance looks very encouraging
• Validation process not yet completed, preliminary
  results available
• Some conclusions from the last CSEWG meeting,
  November 2005

7
Validation of ENDF/B-VII beta2
Data testing with ICSBEP criticality safety
benchmarks Fast metallic systems (LANL, May
2006)
Calc/Exp for k-eff criticality
Our new ENDF/B-VII database shows significant
improvements for calculated reactor criticality
(thermal fast) and transmutation
8
Validation of ENDF/B-VII beta2
Data testing with ICSBEP criticality safety
benchmarks Thermal U-O2 rods (LANL, May 2006)
Closed circles beta2 Closed squares beta1 Open
squares VI.8 Performance was considerably impro
ved.
Russian expts
Japanese expts
US expts
French expts
9
Validation of ENDF/B-VII beta2
Data testing with ICSBEP criticality safety
benchmarks Beryllium reflectors on fast
systems (LANL, May 2006)
Lines eye guides. Reflector thickness bias is
much reduced.
Closed circles beta2 Open squares VI.8
10
Neutron cross sections Covariances(uncertainties
correlation matrix)

• Covariances in ENDF/B-VII beta2 library
• Drastic reduction compared to ENDF/B-VI.8
• Covariances often from 1970-ties, mostly produced
  for ENDF/B-V
• CSEWG meeting, Nov 2005 decided to keep quality
  covariances only
• Review of covariances and recommendation by Don
  Smith (ANL, Jan 2006)
• About 90 of covariances were removed from beta1
• Only partial covariances for 13 materials were
  migrated to beta2
• New covariances for 9 materials
• 152,153,154,155,156,157,158,160-Gd
• Produced by BNL-ORNL-LANL, new evaluation
  including unresolved resonances
• Full set of 8 isotopes, complete covariances,
  show case for future effort
• 232-Th
• Produced by IAEA international project, new
  evaluation, complete covariances

11
Covariance visionPresented to DOE-SC, Office of
Nuclear Physics, Feb 2006

• Proceed in 3 steps, adopt flexible approach,
  establish strong dialog with users, produce
  usable results in each step
• 1st year Produce crude, yet reasonable
  covariances for all nuclei in ENDF/B-VII.0
  (Chadwicks idea, LANL), make results available
  via ENDF/A library, establish dialog with users
  (release in 2007).
• Next 2-3 years Improve all covariances so that
  they are of solid quality to justify their
  inclusion into ENDF/B-VII.1 (release in 2010).
• Next 4-5 years Produce quality results, include
  into ENDF/B-VII.2 (release 2015)
• Manpower and cost
• 2-4 FTE scientists, 1-2 post-docs
• cost initially 0.75 mil, increasing to 1.5
  mil in last years
• Important
• Leverage from CSEWG and international effort (NEA
  Paris, IAEA Vienna)
• Expertise in databases and services tailored to
  user needs

12
Covariance tools are being developed, example
BNL 2006
13
Impact of cross section uncertainties

• Cross section uncertainties have impact on key
  nuclear analysis parameters
• Criticality (multiplication factor)
• Doppler reactivity coefficient
• Coolant void reactivity coefficient
• Effective delayed neutron fraction
• Reactivity loss during irradiation
• Transmutation potential
• Peak power value
• Reactivity control
• Decay heat
• Radiation source at fuel discharge (isotopic
  depletion/burnup)
• Radiotoxicity

14
Impact of cross section uncertainties on reactor
design Example 1

• Uncertainties often estimated from differences of
  data in various libraries
• Example Accelerator driven system
• Uncertainties in cross sections for Np, Cm, Am
  isotopes led to significant differences in
  predicted criticality.
• Need to determine cross sections more precisely
• Need to quantify uncertainties (covariances)
• Similar sensitivities have been determined
  recently by Palmiotti (ANL).

Japanese analysis, Santa Fe 2004
15
Impact of cross section uncertainties

• Sensitivity/uncertainty (S/U) methods are used to
  propagate cross-section uncertainties to
  calculated quantities of interest in nuclear
  analysis
• In the U.S. ORNL and ANL developed S/U methods to
  analyze nuclear fuel cycle applications
  (reactors, transportation storage of spent
  nuclear fuel, shielding, )
• ORNL TSUNAMI software in SCALE based on adjoint
  perturbation theory
• Tool for Sensitivity and Uncertainty Analysis
  Methodology Implementation
• We will show one illustrative example only
• Subsequent talk by Brad Rearden (ORNL) will
  provide more details on the use of S/U methods
  and cross-section data uncertainties for nuclear
  analyses
• ANL ERANOS software based on generalized
  perturbation method, developed largely in CEA
  Cadarache, France

16
Impact of cross section uncertaintiesExample 2,
ORNL S/U analysis

• M. Dunn, ORNL
• ORNL produced new covariances for 233U (hopefully
  to be included in ENDF/B-VII)
• ORNL and LANL will complete 235U and 238U
  covariance evaluations in 2006
• ORNL demonstrated use of 233U covariance data
  with TSUNAMI and contribution of 233U data
  uncertainties to keff for selected ICSBEP
  criticality safety benchmarks (L. Leal, MC
  meeting in Avignon 2005)

Impact seems large. More discussion needed.
17
Impact of cross section uncertaintiesANL S/U
analysis

• M.Salvatores, adopted by international effort
  (NEA Paris, May 2006)
• Sensitivity analysis is performed, via GPT
  (Generalized Perturbation Theory), on performance
  parameters (core, fuel cycle) of representative
  models of the systems of interest
• Uncertainty (e.g. nuclear data covariance)
  propagation and assessment 
• Once the sensitivity coefficient matrix S and the
  covariance matrix D are available, the
  uncertainty on any integral parameter R can be
  evaluated

• Impact on design and target accuracy requirements
  can then be specified.
• ANL cross section covariance matrix is based on
  simple educated guess for uncertainties and the
  most simple estimate for correlation matrix.
• ERANOS code system used

18
Impact of cross section uncertaintiesANL S/U
analysis
Target accuracies assumed for integral
parameters, by Aliberti, Palmiotti, Salvatores

• Some conclusions
• Data uncertainties are significant for only a
  few parameters
• k_eff for all systems, in thermal systems at EOC
  due to high burnup
• burnup reactivity swing and related isotope
  density variations during core depletion
• some void coefficients in fast systems
• neutron source (thermal systems) at fuel
  unloading

19
Data needs for AFC and Gen-IV Cross sections
and covariances

• Actinides
• Pu-239 fission in 1MeV - 1keV lt 1eV
• Pu-240 capture at the first resonance
• Pu-241 fission in 1MeV - 1keV
• U-238 capture in 0.2MeV - 2keV
• U-238 capture in 400eV - 10eV
• U-238 inelastic
• Am-243 capture in fast and thermal energy range
• Am-241fission in fast range
• Structural/coolant materials
• Fe inelastic
• Na inelastic
• Pb inelastic
• Si inelastic

AFCI Gen-IV Physics Working Group S/U study
of fast and thermal systems established that the
following cross sections including covariances
are key for AFC and Gen-IV systems. More
studies needed to quantify these requirements.
20
Data needs for advanced reactorsImproved
actinide cross sections

• Cross sections needed for
• Simulation of nuclear criticality and
  transmutation rates (burnup)
• Simulation of radiation damage and heating

• Minor actinides
• Np, Am, Cm isotopes
• Experiments with extremely small radioactive
  targets at LANSCE
• Theory can be used to predict unknown actinide
  fission and capture
• Integral validation provides very accurate
  quality check

242mAm fission cross sections current status

• Major actinides
• 239Pu, 235,238U have high impact on AFCI because
  of their abundance
• Significant uncertainties (gt 10) in fast neutron
  region for capture

• Objectives for AFCI project
• Perform experiments
• Improve modeling
• Produce precise cross sections

21
Conclusions

• New U.S. evaluated cross section library
  (ENDF/B-VII)
• It will be released in 2006
• It is the largest and best neutron cross section
  library
• Specific AFC data needs should be addressed in
  near future
• Vast amount of covariances (uncertainties
  correlations)
• Improved cross sections for major and minor
  actinides
• Improved cross sections for structural and
  coolant materials
• Impact of cross section uncertainties on AFC
  systems
• Studied partially (lack of covariance data!),
  focus on reactor analysis
• More complete and more detailed studies required