Operated by Los Alamos National Security, LLC for the U.S. … · 2017. 2. 28. · Pu series (0.126...

Operated by Los Alamos National Security, LLC for the U.S. Department of Energy's NNSA

Los Alamos National Laboratory

Subcritical Copper-Reflected α-phase Plutonium (SCRαP) Measurements and

Simulations

NCSP TPR March 2017

Operated by Los Alamos National Security, LLC for the U.S. Department of Energy's NNSA

Los Alamos National Laboratory *Institut de Radioprotection et de Sûreté Nucléaire (IRSN)

J. Hutchinson, R. Bahran, T. Cutler, W. Monange*, J. Arthur, M. Smith-Nelson, E. Dumonteil*


March 2017 | 3

Overview • Introduction • Experiment Design • Experiment Overview • Preliminary Results • Future work


March 2017 | 4

Introduction


March 2017 | 5

Design/Conduct/Analyze Subcritical Validation Experiments

o Nuclear Data and Transport Codes • Fill integral experiment database deficiencies + • Find differential nuclear data library deficiencies For different….

o Energy Ranges (Thermal, Intermediate, Fast) o Multiplication Ranges (Low, Medium, High) o Materials (Fissile, Moderator, Reflector) o Neutron Reactions

o Uncertainty Quantification


March 2017 | 6

Recent Advances in Subcritical Experiments • We have come a long way since the first subcritical

measurements at CP-1 in 1942. • Many organizations (LANL, LLNL, SNL, IAEA, IRSN, CEA,

universities, and others) have pursued subcritical experiments and/or simulations in recent years.

• The BeRP ball reflected by nickel benchmark evaluation was published in the 2014 edition of the ICSBEP handbook.

• This benchmark was the first: o Published benchmark evaluation of measurements performed at DAF.

o Benchmark evaluation using new MCNP capabilities for subcritical systems (the MCNP list-mode patch and MCNP6 list-mode capabilities).

o Benchmark using the Feynman Variance-to-Mean method.

o LANL-led subcritical experiment in the ICSBEP handbook.

• This benchmark was the culmination of several years of subcritical experiment research.

• BeRP-tungsten published in 2016 edition of ICSBEP handbook.


March 2017 | 7

Experiment Design (CED-1 and CED-2)


March 2017 | 8

Subcritical Copper-Reflected α-phase Plutonium (SCRαP) Integral Experiment

• SCRαP Preliminary Design (w/ MCNP®6) o BeRP (Beryllium-Reflected Plutonium).

• 4.5-kg WG α-phase stainless-steel clad plutonium sphere.

o High-purity nested copper shells • C101 Cu alloy (99.99 wt.% Cu).


March 2017 | 9


• SCRαP Preliminary Design (w/ MCNP®6) o High-density interleaved polyethylene

shells • Wide range of achievable subcritical

multiplication values will help: • Identify deficiencies and quantify

uncertainties in nuclear data • Validate computational methods related to

neutron multiplication inference.

There are two purposes for the configurations with polyethylene: • They allow for higher multiplication factor than with copper alone • They allow for a different neutron spectra (and resulting sensitivity)

for the same multiplication factor.


March 2017 | 10


• MC-15 (Multiplicity Counter 15) was used to estimate three benchmark parameters: o Detector singles count rate (R1) i.e. the

count rate in the detector system o Doubles count rate (R2) i.e. the rate in

the detector system in which two neutrons from the same fission chain are detected

o Leakage multiplication (ML) i.e. the number of neutrons escaping a system per starter neutron.

Photograph and MCNP® model of the MC-15 detector system. 15 He-3 tubes inside polyethylene.

Records list-mode data (a time list of every recorded neutron event to a resolution of 128 nsec).


March 2017 | 11


Photograph and MCNP® model of the MC-15 detector system. 15 He-3 tubes inside polyethylene.

Records list-mode data (a time list of every recorded neutron event to a resolution of 128 nsec).

For the SCRαP experiment, two MC-15 systems were present and collected data in the same time list.


March 2017 | 12

Sensitivity Results with MCNP®6

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.00.82

0.84

0.86

0.88

0.90

0.92

0.94

0.96

0.98

1.89

1.83

1.8

1.77

1.751.73

1.721.7

1.69 1.68 1.68 Pure Cu-Reflected BeRP Ball

kef

f

Subcritical Copper Shell Thickness [in]

avg. neutron energy causing fission [MeV]

• Cu-only base configurations shown.

• Decided to only measure configurations up to 4.0” thickness of total Cu: o Cu total neutron cross

section sensitivity (and average neutron energy causing fission) level out at the 5”-thick mark.

o Beyond 4.0” there are additional issues (weight, cost, criticality safety).


March 2017 | 13

Sensitivity Results with MCNP®6

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.00.82

0.84

0.86

0.88

0.90

0.92

0.94

0.96

0.98

1.89

1.83

1.8

1.77

1.75

1.731.72

1.71.69

1.68 1.68 Pure Cu-Reflected BeRP Ball

kef

f

Spherical Copper Shell Thickness [in]

avg. neutron energy causing fission [MeV]

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

k eff a

bsol

ute

sens

itivi

ty

63Cu (n,total)

65Cu (n,total)

• Cu-only base configurations shown.

• Decided to only measure configurations up to 4.0” thickness of total Cu: o Cu total neutron cross

section sensitivity (and average neutron energy causing fission) level out at the 5”-thick mark.

o Beyond 4.0” there are additional issues (weight, cost, criticality safety).


March 2017 | 14

Interleaving Poly Provides Additional Configurations

0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.891.83

1.8 1.771.75

1.731.72

1.7 1.691.68

239Pu (n,total) - Pure Cu-Reflected BeRP Ball avg. neutron energy causing fission [MeV] 63,65Cu (n,total) - Pure Cu-Reflected BeRP Ball

kef

f abs

olut

e se

nsiti

vity

keff


March 2017 | 15

Interleaving Poly Provides Additional Configurations

0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.891.83

1.8 1.771.75

1.731.72

1.7 1.691.68

1.741.66

1.55 1.53 1.52

239Pu (n,total) - Pure Cu-Reflected BeRP Ball 239Pu (n,total) - Pure Cu and Poly-Reflected BeRP Ball avg. neutron energy causing fission [MeV] 63,65Cu (n,total) - Pure Cu-Reflected BeRP Ball 63,65Cu (n,total) - Pure Cu and Poly-Reflected BeRP Ball

kef

f abs

olut

e se

nsiti

vity

keff

14% epithermal4.4% thermal

3.6% epithermal0% thermal


March 2017 | 16

Sensitivity Results with MCNP®6 • These results were compared to critical

configurations in the ICSBEP handbook. o Note that there are a limited number of

copper-reflected critical experiments (8 series with U fuel and 2 series with Pu fuel).

• Total Cu-63 (all energies): o For SCRaP experiment, the maximum

sensitivity (0.143) was greater than the two Pu series (0.126 max) but less than for some of the U series (0.200 max).

• Total Cu-63 (intermediate energy regime): o For SCRaP experiment, the maximum

sensitivity for the 16 configurations (0.018) is greater than the two Pu experimental series by nearly an order of magnitude (0.002), but similarly less than that for some of the U experimental series (0.051).

0.85 0.90 0.95 1.00-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14 Thermal Intermediate Fast

k eff a

bsol

ute

sens

itivi

ty

keff

Cu+Poly configurations

Cu only configurations

Cu-63 (n,total)

Configurations with HDPE have increased sensitivity in the intermediate energy range.


March 2017 | 17

0.85 0.90 0.95 1.00-0.010-0.0050.0000.0050.0100.0150.0200.0250.0300.0350.0400.0450.0500.0550.0600.0650.070

Thermal Intermediate Fast

k eff a

bsol

ute

sens

itivi

ty

keff

Sensitivity Results with MCNP®6 • These results were compared to critical

configurations in the ICSBEP handbook. o Note that there are a limited number of

copper-reflected critical experiments (8 series with U fuel and 2 series with Pu fuel).

• Cu-65 sensitivity results yielded same conclusions as Cu-63.

Cu-65 (n,total)

Configurations with HDPE have increased sensitivity in the intermediate energy range.


March 2017 | 18

Uncertainty estimates • Experimental uncertainties for 4 experimental

parameters were calculated. • Used criticality eigenvalue calculations for these

estimates as described in a previous work [J. HUTCHINSON, T. CUTLER “Use of Criticality Eigenvalue Simulations for Subcritical Benchmark Evaluations” Transactions of the ANS Winter Meeting, Las Vegas NV (2016)].

Parameter Experimental Uncertainty

Uncertainty

ML

Pu radius ± 2 mils 0.18 Pu isotopics ± 0.5% 0.19

Cu thickness ± 0.3 cm 0.03 Cu mass ± 0.5% 0.00006

R1

Pu radius ± 2 mils 1024 Pu isotopics ± 0.5% 1045

Cu thickness ± 0.3 cm 141 Cu mass ± 0.5% 0.34

R2

Pu radius ± 2 mils 37450 Pu isotopics ± 0.5% 41336

Cu thickness ± 0.3 cm 5252 Cu mass ± 0.5% 13.1

Estimate of experimental uncertainties for Configuration 15 (0.5 inch-thick HDPE surrounded by 3.5 inch-thick copper).

Cu mass was expected to be a minor uncertainty, which the table confirms.

Many lessons-learned from the previous Ni and W benchmarks were used to minimize experimental uncertainties.


March 2017 | 19

Experiment Overview (CED-3B)


March 2017 | 20

Experiment configurations • 17 total configurations: o 1 Bare o 8 Cu-only configurations o 7 Cu+HDPE configurations o 1 HPDE-only configuration

• In order to determine the detector efficiency, Cf-252 source replacement measurements were performed. o The source strength of the Cf-252

source at the time of the measurements was 7.59e5 fissions/sec +/- 1.0%.


March 2017 | 21

Experiment configurations

• 17 total configurations: o 1 Bare o 8 Cu-only configurations o 8 Cu+HDPE configurations

• In order to determine the detector efficiency, Cf-252 source replacement measurements were performed. o The source strength of the Cf-252

source at the time of the measurements was 7.59e5 fissions/sec +/- 1.0%.

Thickness (inches) Simulated keff

HDPE Cu HDPE+Cu LANL IRSN

0.0 0.0 0.0 0.774 0.777

0.0 0.5 0.5 0.837 0.829

0.0 1.0 1.0 0.871 0.862

0.0 1.5 1.5 0.894 0.884

0.0 2.0 2.0 0.911 0.900

2.0 1.0 3.0 0.917 0.907

0.0 2.5 2.5 0.924 0.914

2.0 2.0 4.0 0.929 0.921

2.0 2.0 4.0 0.935 0.919

0.0 3.0 3.0 0.935 0.923

0.0 3.5 3.5 0.944 0.933

0.0 4.0 4.0 0.951 0.939

1.5 2.0 3.5 0.951 0.939

1.0 2.5 3.5 0.957 0.943

0.5 3.0 3.5 0.958 0.942

0.5 3.5 4.0 0.965 0.948

4.0 0.0 4.0 - -


March 2017 | 22


Configurations 0-7


March 2017 | 23


Configurations 8-16


March 2017 | 24


Configurations 0-7


March 2017 | 25


Configurations 8-16


March 2017 | 26

Preliminary Results (CED-3B + CED-4A)


March 2017 | 27

Analysis method

• Neutron noise analysis o Rossi-alpha o Time interval analysis o Feynman variance to mean

• Hansen Dowdy • Hage-Cifarelli

o Others… • Analysis method used here

is documented in detail in the BeRP/Ni and BeRP/W ICSBEP evaluations.

Data are separated into gates (of time-width τ)

y-axis is the number of gates that contained exactly n events (Cn)

x-axis is the number of neutrons recorded in the gate (n)

∑= nCTime τ

τ


March 2017 | 28

Feynman histogram results

• Deviation from Poisson (solid lines) increases as system multiplication increases.

• Mean of histogram is proportional to the detector count rate.

• Width of histogram is proportional to the doubles count rate.

Configurations 0-5 (1024 micro-sec gate-width)

0 10 20 30 40 50 60 70 80 90 100102

103

104

105

106

107

Cn

n

bare


March 2017 | 29





0 25 50 75 100 125 150 175 200 225 250102

103

104

105

Cn

n



March 2017 | 30





0 25 50 75 100 125 150 175 200 225 250102

103

104

105

106

Cn

n



March 2017 | 31

Singles count rate (R1)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 161000015000200002500030000350004000045000500005500060000650007000075000800008500090000

R1 (c

ts/s

ec)

Configuration

SCRaP Singles Rate (R1)( )( )

!

)1()1(0

r

prnnnm n

n

r

∑∞

=

+−⋅⋅⋅−=

ττ

( ) ( )

( )∑∞

=

=

0nn

nn

C

Cp

τ

ττ

( ) ( )τ

ττ 1

1mR =

Reduced factorial moment:

normalized fraction of gates that recorded n events:

Singles count rate:

Gate-width (τ)


March 2017 | 32

Singles count rate (R1) Detector efficiency (from Cf-252 measurements):

This is the count rate from the Cf-252 measurements.

FS is the reported spontaneous fission emission rate of the Cf-252 source. is the average number of neutrons emitted per Cf-252 fission.

( ))1(

1

SSFR

ντ

ε =

)1(Sν

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1x104

2x104

3x104

4x104

5x104

6x104

7x104

8x104

9x104

SCRaP Singles Rate (R1)

R1 (c

ts/s

ec)

Configuration

0.000

0.005

0.010

0.015

0.020

0.025

0.030

SCRaP efficiency (ε)

effic

ienc

y

Plotted together to show that the reason that the count rate goes down significantly for the configurations with HDPE is due to the decrease in efficiency (which is caused by neutron absorption primarily in the hydrogen).


March 2017 | 33

Excess variance The excess variance (deviation of a Feynman histogram from a Poisson distribution) is proportional to Y2, given by:

( )[ ]

τ

τττ

212

2

)(21)( mm

Y−

=

0 250 500 750 1000 1250 1500 1750 2000

0.0

2.0x105

4.0x105

6.0x105

Y2

Time Width (µsec)

The amount of excess variance increases with Cu thickness (due to an increase in the system multiplication).

Configurations with Cu only


March 2017 | 34

Excess variance The excess variance (deviation of a Feynman histogram from a Poisson distribution) is proportional to Y2, given by:

( )[ ]

τ

τττ

212

2

)(21)( mm

Y−

=

The amount of excess variance increases with the system multiplication.

Configurations with Cu+HDPE

0 250 500 750 1000 1250 1500 1750 2000

0.0

2.0x105

4.0x105

6.0x105

Y2

Time Width (µsec)

HDPE can increase or decrease Y2 (due to a competition between multiplication and detector efficiency (due to absorption in H).


March 2017 | 35

Neutron lifetime/slowing-down time A fit can be performed on the Y2 curves to calculate the neutron lifetime/slowing-down time (1/λ):

λττλω

λτ−−−=

e11),(2

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16-50

0

50

100

150

200 SCRaP Lifetime (1/λ)

Neu

tron

lifet

ime/

slow

ing-

dow

n tim

e (m

icro

-sec

)

Configuration

0.000

0.005

0.010

0.015

0.020

0.025

0.030

SCRaP efficiency (ε)

effic

ienc

y

The MC-15 detector system has a slowing-down time of around 35 micro-seconds. For the configurations with Cu only, the result is approximately 35 micro-seconds as expected, but it is significantly larger for the configurations that include HDPE hemishells.

This can also be determined via Rossi-alpha analysis.


March 2017 | 36

Doubles count rate (R2) Doubles count rate:

),()()(

2

22 τλω

ττ

YR =

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1x104

2x104

3x104

4x104

5x104

6x104

7x104

8x104

9x104

SCRaP Singles Rate (R1)

R1 (c

ts/s

ec)

Configuration

103

104

105

106

SCRaP Doubles Rate (R2)

R2 (c

ts/s

ec)


March 2017 | 37

Leakage multiplication (ML)

( )( )

31224

1

)1(23

)1(

)2()1()2(2

)1(

)2()1(1

1

42

4

1

1

2

CCCC

RRC

C

C

withC

CCM

S

I

ISS

I

IS

L

−=

−=

−−=

−=

+−=

ετντ

νννν

ννν

)1(Sν

)1(Sν 1st factorial moment of Pu-240 Pν 2nd factorial moment of Pu-240 Pν 1st factorial moment of Pu-239 Pν 2nd factorial moment of Pu-239 Pν

)2(Sν

)1(Iν

)2(Iν

Assumes that there are no emissions from (α,n) neutrons.

Leakage multiplication is related to the multiplication factor (keff).


March 2017 | 38

Leakage multiplication (ML)

( )( )

31224

1

)1(23

)1(

)2()1()2(2

)1(

)2()1(1

1

42

4

1

1

2

CCCC

RRC

C

C

withC

CCM

S

I

ISS

I

IS

L

−=

−=

−−=

−=

+−=

ετντ

νννν

ννν

)1(Sν

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 160

2

4

6

8

10

12

14

16

ML

Configuration

SCRaP Leakage Multiplication (ML)


March 2017 | 39

Leakage multiplication (ML) )1(Sν

As previously discussed, preliminary simulations provided multiplication factor (keff) results. These results were used to approximate leakage multiplication.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150

2

4

6

8

10

12

14

16

18

20 LANL simplified model IRSN detailed model Measured

Leak

age

Mul

tiplic

atio

n (M

L)

Configuration

The LANL MCNP models were simplified models (perfect spherical reflectors with no materials present outside the reflectors) but the IRSN MORET models had additional details (MC-15 detectors, detailed reflector hemishells, etc.).


March 2017 | 40

Leakage multiplication (ML) )1(Sν

As previously discussed, preliminary simulations provided multiplication factor (keff) results. These results were used to approximate leakage multiplication.

The LANL MCNP models were simplified models (perfect spherical reflectors with no materials present outside the reflectors) but the IRSN MORET models had additional details (MC-15 detectors, detailed reflector hemishells, etc.).

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16-4

-2

0

2

Leak

age

Mul

tiplic

atio

n (C

-E)/E

(%)

Configuration

(C-E)/E (%) for IRSN MORET detailed models


March 2017 | 41

Future work (CED-4A + CED-4B)


March 2017 | 42

What’s next? • This experiment will be evaluated and

documented in an upcoming version of the ICSBEP handbook.

• Results will hopefully be used to improve cross-section libraries.

• Data set will also be used to validate subcritical analysis methods.


March 2017 | 43

Thank you for your attention.

This work was supported by the DOE Nuclear Criticality Safety Program, funded and managed by the National Nuclear Security Administration for the Department of Energy.

Operated by Los Alamos National Security, LLC for the U.S. … · 2017. 2. 28. · Pu series (0.126...

Documents

Transcript of Operated by Los Alamos National Security, LLC for the U.S. … · 2017. 2. 28. · Pu series (0.126...