Ongoing work and goals of the nuclear data team · TMC applied to burn-up calculations with SERPENT 13 / 31 PWR fuel element based on a Westinghouse 3-loop PWR design (array of 17x17),

1 / 31

Ongoing work and goals of the

nuclear data team

D. Rochman and A.J. Koning

Nuclear Research and Consultancy Group,

NRG, Petten, The Netherlands

Uppsala, Sweden, November 2011

Contents

2 / 31

① Method of work

② Applications: TMC, TENDL, TMC−1

③ Conclusions

A (very) brief and recent history of nuclear data in NRG

3 / 31

1997 2000 2003 2006 2009 2012

1st idea of

TALYS

TENDL

TMC

T6

Inverse

TMC

Manhattan-2

project

TALYS-1.0

released

TENDL-2011

released

TENDL-2010

released

TENDL-2009

released

TALYS-1.2

released

TENDL-2008

released

Backbone of our methodology: REPRODUCIBILITY

4 / 31

TALYSnuclear code

T6 softwarepackage

Librarycloning andcomplement ReproducibilityN

RG

Nuc

le

ardata team

Original nuclear datalibrary TENDL+ covariances

Uncertaintypropagation

TMC

OptimumSearch and find(Inverse TMC)

TMC: Total Monte Carlo

5 / 31

Uncertaintypropagation

TMC

• Started in 2008

• Many publications

• Applied to crit-saf and shielding benchmarks, reactor (keff, βeff, void,Doppler), burn-up inventory, radiotoxicity

• Still a controversial method

TMC: procedure for the random file production

6 / 31

AutoTalysTASMAN

n TANESinput files

n TALYSinput files

n TARESinput files

n TAFISinput files

TANES TALYS TARES TAFIS

n FissionNeutron Spect.

output files

n TALYSoutput files

n ResonanceParametersoutput files

n ν-baroutput files

TEFAL

1 ENDF file+

covariancesn×

ENDFrandom files

TMC: Total Monte Carlo

7 / 31

Control of nuclear data (TALYS)+ simple processing (NJOY)

+ system simulation (MCNP/ERANOS/CASMO...)

1000times

For each random ENDF file, the benchmark calculation is performed withMCNP. At the end of the n calculations, n different keff values are obtained. In theobtained probability distribution of keff, the standard deviation σtotal reflects twodifferent effects: σ2

total = σ2statistics +σ2

nuclear data.

hst1-1 (236U)

keff value

Num

ber

ofco

unts

/bin

s

1.0020.9990.9960.993

40

30

20

10

0

Examples with 63Cu(n,2n) and 65Cu(n,el)

8 / 31

n = 10063Cu(n,2n)

Incident Energy (MeV)

Cro

ssse

ctio

n(b

)

2018161412

1.0

0.8

0.6

0.4

0.2

0.0

n = 163Cu(n,2n)

Incident Energy (MeV)

Cro

ssse

ctio

n(b

)

2018161412

1.0

0.8

0.6

0.4

0.2

0.0

Skewness = 0.33

Sigma = 4 mb/Sr

Mean = 22 mb/Sr

cos(θ) = −0.41

En = 8 MeV65Cu(n,el)

Cross section (b/Sr)

Cou

nts

/bin

0.050.040.030.020.01

20

15

10

5

0

En = 8 MeV

65Cu(n,el) (n=100)

cos(θ)

dσ/d

θ(b

/Sr)

0.80.50.2-0.1-0.4-0.7-1.0

101

100

10−1

10−2

Nuclear data: examples on (n,2n) cross sections

9 / 31

Exp.ENDF/B-VII.0

This work

100 random 241Am(n,2n)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(m

bar

ns)

2015105

600

400

200

0Exp.

ENDF/B-VII.0This work

100 random 232Th(n,2n)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(b

arns)

2015105

3

2

1

0

Exp.ENDF/B-VII.0

This work

100 random 235U(n,2n)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(m

bar

ns)

2015105

1200

800

400

0

Exp.ENDF/B-VII.0

This work

100 random 239Pu(n,2n)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(m

bar

ns)

2015105

600

400

200

0

Nuclear data: examples in the resonance region

10 / 31

JEFF-3.1ENDF/B-VII.0

This workExp

50 random 241Pu(n,f)

Incident neutron energy (eV)

Cro

ssse

ctio

n(b

arns)

10010−110−2

104

103

102

101

JEFF-3.1ENDF/B-VII.0

This workExp

50 random 235U(n,f)

Incident neutron energy (eV)

Cro

ssse

ctio

n(b

arns)

4.03.02.01.00.40.1

102

101

TMC applied to a SFR void coefficient

11 / 31

Reduced Chi-square=0.215

SVR= 8.02510± 0.55092 $

Kalimer void coefficient (23Na)

Void coefficient value ($)

Num

ber

ofco

unts

/bin

s

9.38.77.97.36.7

20

15

10

5

0<Skewness>

Sample number

Updat

edsk

ewnes

s

8006004002000

0.0

-0.1

-0.2

-0.3

-0.4

< σ >

Updat

edunce

rtai

nty

($)

0.6

0.5

0.4

0.3

< void coefficient >

Sodium Fast Reactor (Kalimer-600)

Updat

edSV

R($

)

8.5

8.3

8.1

7.9

TMC applied to shielding benchmark (Mn Oktavian benchmark)

12 / 31

Emission Gamma Energy (MeV)

25

15

5

10532

Emission Neutron Energy (MeV)

Unce

rtai

nty

(%)

10.05.03.01.00.5

3020100

ExperimentJEFF-3.1This work

Oktavian 55Mn Neutron Leakage

Lea

kage

curr

ent/

leth

argy

/sou

rce

par

ticl

e 0.60

0.30

0.20

0.10

0.05

ExperimentJEFF-3.1This work

Oktavian 55Mn Gamma Leakage

0.100

0.050

0.010

0.005

0.001

TMC applied to burn-up calculations with SERPENT

13 / 31

PWR fuel element based on a Westinghouse 3-loop PWR design (array of 17x17),4 m in length, 21.5 cm in width, 400-500 kg of enriched uranium (4.8 % in 235U)

due to fission yieldsdue to transport data

238U+235U+239Pu

Burn up (GWd/tHM)

Unce

rtai

nty

onk

eff(%

)

6050403020100

2.0

1.5

1.0

0.5

0.0

due to fission yieldsdue to transport data

239Pu

Burn up (GWd/tHM)

Uncertain

tyon

keff

(%)

1.5

1.0

0.5

0.06050403020100

due to fission yieldsdue to transport data

235U

Burn up (GWd/tHM)

Uncertain

tyon

keff

(%)

1.5

1.0

0.5

0.0605040302010

Unce

rtai

nty

onk

eff(%

) 1.5

1.0

0.5

0.0

due to decay Datadue to fission yields

due to transport data

238U

Burn up (GWd/tHM)605040302010

TMC applied to burn-up calculations with DRAGON

14 / 31

Unce

rtai

nty

onk

eff(%

) 1.0

0.8

0.6

0.4

0.2

0.0

due to transport data

238U

Burn up (GWd/tHM)605040302010

due to nu-bardue to pointwise xsdue to resonance xs

due to transport data

238U

Burn up (GWd/tHM)

Unce

rtai

nty

onk

eff(%

)

6050403020100

1.0

0.8

0.6

0.4

0.2

0.0

SERPENT (Monte Carlo) DRAGON (deterministic)

TENDL: nuclear data production

15 / 31

Librarycloning andcomplement

Original nuclear datalibrary TENDL+ covariances

• Started in 2008

• TENDL libraries released every year

• First of its kind by the method of production

• First of its kind by separating evaluation work and library production

• ' 80 isotopes from TENDL to JEFF-3.2β, to FENDL

TENDL: Motivations

16 / 31

We need a consistent and complete nuclear data library to be integrated in reactorcalculations, including realistic covariance data.

(None of the existing libraries fulfill these requirements.)

� Use global, robust TALYS method for the bulk of nuclides,

� Use in-depth evaluation, adjustment... for important nuclides (e.g. 56Fe,239Pu),

� Reproducible library,

� ENDF, PENDF, ACE, EAF and text format,

� Available at www.talys.eu/tendl-2008 (2009, 2010 and 2011).

Produce TENDL-2008, -2009, -2010 ... with an increasing quality with regard to:differential data, model development, integral validation, completeness andcovariance data

TENDL: Standard and modern approaches

17 / 31

ManualCross section

evaluation

(semi) ManualFormatting

• Manual validation• Manual Feedback

JEFF, ENDF/B, JENDL...

ManualCross section

evaluation

AutomaticFormatting

• Automatic validation• Manual Feedback

TENDL

TENDL-2011: t1/2 > 1 sec, 200 MeV (n, p, d, t, a, γ reactions)

18 / 31

TENDL cloning of other library

19 / 31

When TALYS can not reproduce experimental cross sections, normalization isapplied

TALYS-1.2ENDF/B-VII.0

Exp.

238U(n,f)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(b

)

1815129630

1.5

1

0.5

0

TENDL cloning of other library

19 / 31

When TALYS can not reproduce experimental cross sections, normalization isapplied

TENDL-2010TALYS-1.2

ENDF/B-VII.0Exp.

238U(n,f)

Incident neutron energy (MeV)

Cro

ssse

ctio

n(b

)

1815129630

1.5

1

0.5

0

Monte Carlo Nuclear Data Adjustment: TMC−1

20 / 31

OptimumSearch and find(Inverse TMC)

• Started in 2010

• Two publications so far

• Controversial (if understood at all)

• We believe this is the future of nuclear data evaluation work

• It might be the only way to sensibly improve C/E

Inverse TMC on 239Pu

21 / 31

Total Monte Carlo + selection=⇒ 1

TMC

① Use TALYS to create a single 239Pu evaluation close or equal toENDF/B-VII.0 or JEFF-3.1.1

② Randomize all model parameters (resonances, nubar, fission neutronspectrum, TALYS parameters) to create 500 random 239Pu evaluations

③ Benchmarks the n ≥ 500 files with the same set of criticality benchmarks

④ Select the best random file

Example: 100 benchmarks, 500 random files =⇒ 500 TALYS + NJOY and 100×500 = 5 ·104 MCNP loops,1.4 years on a single processor, or 5 days on 100 processors (3 GHz)

Inverse TMC: simple example with 6 keff benchmarks

22 / 31

JEFF-3.1 & ENDF/B-VII.0

TENDL-2009JENDL-3.3 & CENDL-3

ENDF/B-VI.8

Jezebel (pmf1) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

TENDL-2009

JEFF-3.1ENDF/B-VII.0

Jezebel-240 (pmf2) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf12 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf5 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pst1-6 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

Thor (pmf8) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

αJEFF-3.1.1: 1.14e−4

JENDL-3.3: 1.71e−4

TENDL-2009: 3.66e−4

ENDF/B-VI.8: 1.72e−4

ENDF/B-VII.0: 1.69e−4

α = ∑ni=0

(Ci−Ei)2

Ci,

Inverse TMC: simple example with 6 keff benchmarks

22 / 31

JEFF-3.1 & ENDF/B-VII.0

TENDL-2009JENDL-3.3 & CENDL-3

ENDF/B-VI.8

Jezebel (pmf1) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

TENDL-2009

JEFF-3.1ENDF/B-VII.0

Jezebel-240 (pmf2) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf12 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf5 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pst1-6 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

Thor (pmf8) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

αJEFF-3.1.1: 1.14e−4

JENDL-3.3: 1.71e−4

TENDL-2009: 3.66e−4

ENDF/B-VI.8: 1.72e−4

ENDF/B-VII.0: 1.69e−4

random 0: 2.29e−4

random 1: 13.4e−4

Inverse TMC: 6 keff benchmarks with random 239Pu

23 / 31

JEFF-3.1 & ENDF/B-VII.0

TENDL-2009JENDL-3.3 & CENDL-3

ENDF/B-VI.8

Jezebel (pmf1) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

TENDL-2009

JEFF-3.1ENDF/B-VII.0

Jezebel-240 (pmf2) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf12 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pmf5 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

pst1-6 keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

Thor (pmf8) keff

keff value

Num

ber

ofco

unts

/bin

s

1.0201.0101.0000.9900.980

10

5

0

αJEFF-3.1.1: 1.14e−4

JENDL-3.3: 1.71e−4

TENDL-2009: 3.66e−4

ENDF/B-VI.8: 1.72e−4

ENDF/B-VII.0: 1.69e−4

Real case: 120 239Pu benchmarks

24 / 31

Table 1: List of plutonium benchmarks selected for the random search.

Name Cases Name Cases Name Cases Name Casespmf1 1 pmf2 1 pmf5 1 pmf6 1pmf8 1 pmf12 1 pmf13 1 pci1 1pmi2 1 pst1 6 pst2 6 pst3 8pst4 13 pst5 9 pst6 3 pst7 9pst8 29 pst12 22 pmm1 6

α =n

∑i=0

(Ci −Ei)2

Ci, (1)

Results independent of the type of factor α, χ2... or

F = 1−10√

1N ∑(log(Ei)−log(Ci))

2(2)

α values for random 239Pu evaluations

25 / 31

RandomJEFF-3.1

JENDL-3.3ENDF/B-VI.8

ENDF/B-VII.0

Random file number

αper

random

file

7006005004003002001000

1· 10−1

2· 10−2

5· 10−3

C/E values for the best 239Pu (run 307)

26 / 31

Run 307 (α = 0.0048)JEFF-3.1 (α = 0.0081)

Benchmark239Pu benchmarks

Benchmark type

C/E

pmm

1-6

pmm

1-1

pst1

2-19

pst1

2-14

pst1

2-9

pst1

2-4

pst8

-28

pst8

-23

pst8

-18

pst8

-13

pst8

-8

pst8

-2

pst7

-7

pst6

-3

pst5

-7

pst5

-2

pst4

-10

pst4

-5

pst3

-8

pst3

-3

pst2

-5

pst1

-5

pmi2

-1

pmf8

pmf1

1.04

1.02

1.00

Inverse TMC: best 239Pu for the ANDES project

27 / 31

NRG 95Mo, 56Fe, 239Pu, 28Si, 182W + JEFF-3.1JEFF-3.1

ENDFB/-VII.1Benchmarks

239Pu fast benchmarks

September 15, 2011 at 15:41

Benchmark type

C/E

pst1

-1

mcf

4

mcf

3-2

mcf

3-1

mcf

2

mcf

1

pmm

1-6

pmm

1-5

pmm

1-4

pmm

1-3

pmm

1-2

pmm

1-1

pmi2

pci1

pmf1

3

pmf1

2

pmf8

pmf6

pmf5

pmf2

pmf1

1.015

1.010

1.005

1.000

0.995

Inverse TMC: second example on natural copper and Oktavianbenchmark

28 / 31

ExperimentJEFF-3.1This work

Emission Neutron Energy (MeV)

Lea

kage

curr

ent/

leth

argy

/sou

rce

par

ticl

e

Lea

kage

curr

ent/

leth

argy

/sou

rce

par

ticl

e

10532

0.30

0.20

0.10

0.05

0.02

ExperimentJEFF-3.1This work

Emission Gamma Energy (MeV)

Lea

kage

curr

ent/

leth

argy

/sou

rce

par

ticl

e

0.05

0.02

0.01

0.01

10532

Software release: TALYS and T6

29 / 31

TALYSnuclear code

T6 softwarepackage

• TALYS: latest release in December 2012,

• T6: started 2011,

• T6: ”(almost) all our knowledge in a tar file”

• with T6, anyone can produce ENDF evaluations, ACE files, random files,TENDLs...

Software release: T6

30 / 31

t6.tar(a tar file)

TALYS TASMAN TEFAL TARES TANES TAFISTools

autotalys NJOY? PREPRO libs checkingtools acedit

t6 tested on many systems (Linux RedHat, Ubuntu, FreeBSD),with many compilers (g95, gfortran, ifort, lf95, g++, Intel C++)

from Li to Cf, from 0 to 200 MeV?: if authorized

Future work

31 / 31

➪ Continue with TMC (more reactor calculations, applied to current reactors),

➪ Improve the quality of the TENDL library,

1. in format (new needs or existing problems)2. in cross sections (better fit to differential and integral data)

➪ Apply TMC−1 to a large number of isotope (including crit-safety, reactor,dosimetry benchmarks)

➪ Release T6 (or part of it) ?

, And finally world domination (and world peace).