Advanced non-destructive methods for criticality safety ... · Advanced non-destructive methods for...

Advanced non-destructive methods for criticality

safety and safeguards of used nuclear fuel

Riccardo Rossa

Thèse présentée en vue de l’obtention du grade de Docteur en Sciences de l’Ingénieur

Promoteur: Prof. Pierre-Etienne Labeau

Co-Promoteur: Prof. Nicolas Pauly

Mentor (SCK•CEN): Dr. Alessandro Borella

Co-Mentor (SCK•CEN): Ir. Klaas van der Meer

Academic year 2016 - 2017

Acknowledgements

First of all I would like to express my sincere gratitude to my mentor at SCK•CEN Alessandro Borella

that guided me during the Ph.D. research. He taught me a lot about NDA techniques and Monte

Carlo simulations and this thesis would not be in this form without his patience, efforts, and

dedication.

Remaining in the SPS expert group at SCK•CEN I would like to thank the other colleagues for the nice

time spent together. In particular a big thank you to Klaas van der Meer for everything he does for

the group both during working hours and with social events.

Special thanks go to my promotor and co-promotor at ULB Pierre-Etienne Labeau and Nicolas Pauly

for their advices during those years, and to the rest of the Ph.D. jury: Pierre Capel, Alain Dubus, Paolo

Peerani, and Peter Schillebeeckx for their review of the manuscript.

I would like to acknowledge the scientists at JRC-IRMM in Geel working at the GELINA facility for their

support during the experimental measurements and data analysis: Peter Schillebeeckx, Carlos

Paradela, Jan Heyse, Stefan Kopecky, Ruud Wynants, Gery Alaerts.

Thank you to all the friends that shared these years in the Boeretang Kingdom: I would need another

book to thank you all individually! You may be scattered now all over the world, but the nice times

together will stay forever in my mind.

Last but not least I would like to thank my family for the continuous support during these years.

Switching to my native language, un grazie speciale alla mia famiglia che mi ha continuato a

supportare in tutti questi anni.

Riccardo Rossa

Mol, September 2016

This work is sponsored by GDF SUEZ in the framework of the cooperation agreement CO-90-07-2124

between SCK•CEN and GDF SUEZ.

i

Table of Contents List of figures ............................................................................................................................................ v

List of tables ............................................................................................................................................ ix

List of acronyms .................................................................................................................................... xiii

Summary .............................................................................................................................................. xvii

Résumé ................................................................................................................................................ xxiii

1 Introduction .................................................................................................................................. 1

1.1 The framework of nuclear safeguards .................................................................................... 1

1.2 Renewed interest for spent fuel measurement methods ....................................................... 2

1.3 Structure and objectives of the research project .................................................................... 3

2 Safeguards challenges of spent nuclear fuel ................................................................................ 5

2.1 Properties of spent nuclear fuel .............................................................................................. 5

2.2 Safeguards requirements for spent fuel verifications ............................................................. 6

2.3 Current non-destructive assays for spent fuel measurements ............................................... 7

2.3.1 Overview of NDA techniques .......................................................................................... 7

2.3.2 Digital Cherenkov viewing device .................................................................................... 9

2.3.3 Spent fuel attribute tester ............................................................................................. 10

2.3.4 Fork detector ................................................................................................................. 11

2.4 Techniques investigated in this Ph.D..................................................................................... 13

2.4.1 Self-indication neutron resonance densitometry ......................................................... 13

2.4.2 Partial defect tester ....................................................................................................... 13

3 Approach used for the study of the non-destructive techniques ............................................... 15

3.1 Overview of literature study ................................................................................................. 15

3.1.1 Previous research on SINRD .......................................................................................... 15

3.1.2 Previous research on PDET ............................................................................................ 16

3.1.3 Contributions from this Ph.D. project ........................................................................... 17

3.2 Description of the Monte Carlo models ................................................................................ 18

3.2.1 Principles of the Monte Carlo methods ........................................................................ 18

3.2.2 Spent fuel assembly geometry ...................................................................................... 18

3.2.3 Storage configuration .................................................................................................... 19

3.3 Definition of the source term ................................................................................................ 20

3.3.1 Use of the spent fuel library in the research project .................................................... 20

ii

3.3.2 Structure of the spent fuel library ................................................................................. 21

3.3.3 Data processing to generate the fuel material composition ......................................... 22

3.3.4 Data processing to generate the source term characteristics ...................................... 23

3.4 Determination of the detectors response ............................................................................. 23

3.4.1 Neutron detectors ......................................................................................................... 23

3.4.2 Gamma-ray detectors .................................................................................................... 25

4 Monte Carlo assessment of the self-indication neutron resonance densitometry .................... 27

4.1 Structure of the study ........................................................................................................... 27

4.2 Influence of the moderator on the neutron flux ................................................................... 27

4.3 Definition of the SINRD signature ......................................................................................... 29

4.4 Setup optimization ................................................................................................................ 32

4.4.1 SINRD filters ................................................................................................................... 32

4.4.2 Comparison of detector types ....................................................................................... 35

4.5 Expected performances in realistic scenarios ....................................................................... 39

4.5.1 Influence of the spent fuel composition on the SINRD signature ................................. 39

4.5.2 Investigation of systematic effects on the SINRD technique ........................................ 45

4.6 Conclusions ............................................................................................................................ 50

5 Benchmark of the self-indication neutron resonance densitometry ......................................... 53

5.1 Objectives of the benchmark experiments ........................................................................... 53

5.2 Description of the GELINA Time-of-Flight facility .................................................................. 53

5.3 Overview of the experimental setup ..................................................................................... 54

5.3.1 Transmission measurements ......................................................................................... 54

5.3.2 Benchmark measurements ........................................................................................... 56

5.4 Results of the validation experiments ................................................................................... 57

5.4.1 Transmission measurements ......................................................................................... 57

5.4.2 Benchmark measurements ........................................................................................... 60

5.5 Conclusions ............................................................................................................................ 65

6 Monte Carlo assessment of the partial defect tester ................................................................. 67

6.1 Structure of the study ........................................................................................................... 67

6.2 Reference conditions for the PDET detector ......................................................................... 67

6.2.1 Contribution of single fuel pins ..................................................................................... 67

6.2.2 Comparison among several detector types .................................................................. 72

6.2.3 Influence of spent fuel irradiation history ..................................................................... 75

iii

6.3 Influence of the spent fuel assemblies in the storage rack ................................................... 78

6.3.1 Impact of the central fuel assembly burnup ................................................................. 78

6.3.2 Impact of the lateral fuel assemblies on the reference distributions ........................... 88

6.3.3 Impact of the corner fuel assemblies on the reference distributions ........................... 93

6.4 Conclusions ............................................................................................................................ 95

7 Analysis of the partial defect capabilities for SINRD and PDET .................................................. 97

7.1 Description of the diversion scenarios .................................................................................. 97

7.2 Response of SINRD to the diversion scenarios ...................................................................... 98

7.3 Response of PDET to the diversion scenarios ..................................................................... 102

7.4 Conclusions .......................................................................................................................... 107

8 Discussion and conclusion ........................................................................................................ 109

8.1 Self-Indication Neutron Resonance Densitometry .............................................................. 109

8.2 Partial Defect Tester ............................................................................................................ 111

8.3 Outlook ................................................................................................................................ 112

References ........................................................................................................................................... 115

Annex A. Further considerations on the influence of individual nuclides on the SINRD signature .... 125

A.1. Results with a 235U fission chamber ..................................................................................... 125

A.2. Results with a 3He proportional counter ............................................................................. 127

A.3. Results with a 10B proportional counter .............................................................................. 128

Annex B. Additional diversion scenarios for the PDET detector ......................................................... 131

B.1. Description of the diversion scenarios ................................................................................ 131

B.2. Results for the 235U fission chambers .................................................................................. 132

B.3. Results for the 238U fission chambers .................................................................................. 133

B.4. Results for the ionization chambers .................................................................................... 135

v

List of figures Figure 2-1: Material composition of LEU spent fuel with 3.5% initial enrichment, 33 GWd/tHM burnup,

and 3 years cooling time. (IAEA, 2012b) ................................................................................................. 5

Figure 2-2: Cumulative inventory of spent fuel generated worldwide. (IAEA, 2012b) ........................... 6

Figure 2-3: Improved Cherenkov viewing device (ICVD, left, (IAEA, 2011)). Digital Cherenkov viewing

device (DCVD, right, photo by Andy Gerwing, Channel Systems). .......................................................... 9

Figure 2-4: Color-enhanced image obtained with the DCVD detector of BWR 8x8 (left, adapted from

(Chen, 2003)) and a PWR 17x17 fuel assembly (right, (Chen, 2009)). .................................................. 10

Figure 2-5: Spent fuel attribute tester (SFAT) and water-tight collimator. (Chichester, 2009) ............ 11

Figure 2-6: Fork detector developed by LANL (left, (Antech, 2015)). Fork detector developed by

SCK•CEN (right, (Borella, 2011))............................................................................................................ 12

Figure 2-7: Total microscopic cross-section of 239Pu according to the ENDF-B/VII.0 nuclear data library

(Chadwick, 2006). The right plot is a zoom on the energy region close to 0.3 eV. ............................... 13

Figure 2-8: Prototype of the PDET detector developed by LLNL. (Ham, 2013) ..................................... 14

Figure 3-1: Monte Carlo model of the PWR 17x17 fuel assembly. ....................................................... 19

Figure 3-2: Monte Carlo model of the fuel storage configuration for SINRD (dry case, left) and for

PDET (right). .......................................................................................................................................... 20

Figure 3-3: Extract of the output file generated from the data processing to obtain the spent fuel

material composition. ........................................................................................................................... 22

Figure 3-4: Extract of the output file generated from the data processing to obtain the gamma-ray

source energy distribution. ................................................................................................................... 23

Figure 3-5: Microscopic cross-sections used for the estimation of the neutron detector responses.

Values are based on the ENDF/B-VII.0 nuclear data library. ................................................................. 24

Figure 3-6: Response functions used for the estimation of the gamma-ray detector responses. ....... 26

Figure 4-1: Energy distributions of the neutron flux calculated in the central guide tube for several

moderators. All simulations had 12 cm of moderator outside the fuel assembly, and for the wet

conditions the water was also included within the fuel assembly. ...................................................... 28

Figure 4-2: Energy distribution of the neutron flux calculated in the central guide tube for fresh and

borated water. In this case the water is only outside the fuel assembly and the place among the fuel

pins is filled with dry air. The dry case obtained with polyethylene is included for comparison. ........ 29

Figure 4-3: Energy distribution of the difference between the detector response through the Gd and

Cd SINRD filters...................................................................................................................................... 34

Figure 4-4: SINRD signature as a function of the 239Pu content for fuel with different compositions.

The results refer to a 239Pu fission chamber for the measurement of the neutron flux in the 0.3 eV

resonance region, and the values are normalized to the case of fuel containing only 238U and 16O. ... 39

Figure 4-5: Energy distribution of the detector response in the regions selected for the SINRD

signature. Results for fuel with only 238U and 16O, and fresh fuel with 3.5% initial enrichment. ......... 41


signature. Results for fuel with burnup of 15 GWd/tU. ......................................................................... 42


signature. Results for fuel with burnup of 60 GWd/tU. ......................................................................... 42

Figure 4-8: SINRD signature as a function of the 239Pu content for fuel with different irradiation

histories. The fuel composition contained the 50 main neutron absorbers. ........................................ 43

Figure 4-9: RTH ratio as a function of the 239Pu content for fuel with different irradiation histories. The

fuel composition contained the 50 main neutron absorbers. .............................................................. 44

vi

Figure 4-10: View of the central guide tube in the Monte Carlo model and comparison of different

positioning of the detector.................................................................................................................... 45

Figure 4-11: Comparison of detectors with different lengths (L1 and L2). The track of a neutron

crossing both detectors is also shown with the indication of the incoming angles ( and ). ............. 46

Figure 4-12: Zoom of the central region of the fuel assembly and comparison of the detector cover by

the SINRD filter. ..................................................................................................................................... 47

Figure 4-13: Energy distribution of the difference between the detector response calculated for a 239Pu fission chamber covered by Gd and Cd filters. The fuel in these simulations contained 238U and 16O. ......................................................................................................................................................... 49

Figure 5-1: Aerial view of the GELINA Time-of-Flight facility at the JRC-IRMM in Geel. ....................... 54

Figure 5-2: Schematic view of a transmission experiment. .................................................................. 55

Figure 5-3: Schematic representation of the self-indication experiments carried out at GELINA. ....... 56

Figure 5-4: Transmission through different Gd and Cd foils. The experimental transmission is

compared with the analytical transmission based on the calculations in Chapter 4. ........................... 58

Figure 5-5: Transmission through different Gd and Cd foils using different nuclear data libraries. The

values were calculated with the analytical approach described in Chapter 4. The plot on the right is

focused on energy range below 0.1 eV. ................................................................................................ 59

Figure 5-6: Experimental setup for the self-indication experiments. The 0.027 mm Cd sample is

placed in the neutron beam and is surrounded by 4 C6D6 scintillator detectors. ................................. 60

Figure 5-7: Spectra of the self-indication experiments with Cd samples in the beam. The spectrum

obtained with the detector only is reported for comparison together with the background

contribution. All spectra were normalized to the same beam intensity. ............................................. 61

Figure 5-8: Spectra of the self-indication experiments with a 0.03 mm Gd (left) and 1.0 mm Cd (right)

filter in the beam. The spectrum obtained with the detector only is reported for comparison together

with the background contribution. ....................................................................................................... 62

Figure 5-9: Experimental observables RSI,1 and RSI,2 as a function of the areal density of the Cd sample

placed in the beam. The results were normalized to the measurements without Cd sample. ............ 63

Figure 5-10: Spectra obtained for the 235U fission chamber with a 0.03 mm Gd (left) and 1.0 mm Cd

(right) filter in the beam. Moreover, several Cd samples were used with the Gd filter to simulate the

neutron absorption by fuel pins containing 239Pu. ................................................................................ 64

Figure 5-11: Spectra obtained for the 10B ionization chamber with a 0.03 mm Gd (left) and 1.0 mm Cd

(right) filters in the beam. Moreover, several Cd samples were used with the Gd filter to simulate the

neutron absorption by fuel pins containing 239Pu. ................................................................................ 64

Figure 5-12: Experimental observable RSI,2 as a function of the areal density of the Cd sample placed

in the beam. The data refer to measurements with a self-indication detector, a 235U fission chamber,

and a 10B ionization chamber. ............................................................................................................... 65

Figure 6-1: Importance function for a 235U fission chamber placed in different guide tubes. The

neutron flux was calculated in the guide tube depicted in grey. The color bar ranges between 0 and

1%. ......................................................................................................................................................... 68

Figure 6-2: Importance function for a 238U fission chamber placed in different guide tubes. The


1%. ......................................................................................................................................................... 69

Figure 6-3: Importance function for an ionization chamber placed in different guide tubes. The


1%. ......................................................................................................................................................... 69

vii

Figure 6-4: Areas identified for the calculation of the integral contribution to the detector response in

the different guide tubes. ..................................................................................................................... 70

Figure 6-5: Neutron detector response of 235U fission chambers and 238U fission chambers. The results

for each plot were normalized to the maximum value obtained in the correponding simulation, and

the uncertainty of the values was lower than 0.2% .............................................................................. 73

Figure 6-6: Normalized detector responses for different guide tubes. The 235U and 238U fission

chambers were compared and the statistical uncertainty of the simulations was also included. The

results for the other guide tubes were not included due to the symmetry of the fuel assembly. ....... 73

Figure 6-7: Gamma-ray detector response of an ionization chamber with nitrogen as filling gas at 1

atm. The results were normalized to the maximum value obtained in the guide tubes, and the

uncertainty of the values was lower than 0.8% .................................................................................... 74

Figure 6-8: Normalized detector responses for different guide tubes. Only the response of ionization

chamber with nitrogen as filling gas at 1 atm was reported because the responses of the other

detector types were within the statistical uncertainty. The results for the other guide tubes were not

included due to the symmetry of the fuel assembly. ............................................................................ 75

Figure 6-9: Normalized detector responses for fuel with different burnup (BU). The results refer to 235U fission chambers and the values for the other guide tubes were not included due to the

symmetry of the fuel assembly. ............................................................................................................ 76

Figure 6-10: Normalized detector responses for fuel with different burnup (BU). The results refer to 238U fission chambers and the values for the other guide tubes are not included due to the symmetry

of the fuel assembly. ............................................................................................................................. 77

Figure 6-11: Normalized detector responses for fuel with different cooling time (CT). The results refer

to ionization chambers and the values for the other guide tubes are not included due to the

symmetry of the fuel assembly. ............................................................................................................ 77

Figure 6-12: Position of the guide tubes in the fuel assembly. ............................................................. 78

Figure 6-13: Average relative difference as defined in Formula (6.1) between the detector responses

calculated in the nine central guide tubes. ........................................................................................... 79


calculated in the sixteen guide tubes at the periphery of the fuel assembly. ...................................... 81




calculated in the nine guide tubes at the center of the fuel assembly. ................................................ 84




calculated in the peripheral guide tubes. The NONU card was used to neglect the emission of

secondary neutrons from fission. .......................................................................................................... 88

Figure 6-19: Storage rack configurations for the study of the influence of lateral fuel assemblies. .... 89

Figure 6-20: Relative difference between the normalized detector response calculated for storage

racks with the lateral fuel assemblies with different burnup and the values obtained in the reference

case. The results refer to 235U fission chambers, and the title of each plot indicates the storage rack

configuration described in Figure 6-19. ................................................................................................ 91


racks with the lateral fuel assemblies with different burnup and the values obtained in the reference

viii

case. The results refer to ionization chambers, and the title of each plot indicates the storage rack


Figure 6-22: Storage rack configurations for the study of the influence of corner fuel assemblies. .... 93


racks with the corner fuel assemblies with different burnup and the values obtained in the reference

case. The results refer to 235U fission chambers, and the title of each plot indicates the storage rack


Figure 7-1: Visualization of the diversion scenarios developed for the comparison of the NDA

techniques. The fuel pins are depicted in white, the dummy pins in grey, and the guide tubes in

yellow. ................................................................................................................................................... 97

Figure 7-2: Normalized detector responses in the reference case considering 239Pu fission chambers

and SINRD filters (left) and 238U fission chambers (right)...................................................................... 99

Figure 7-3: Average detector responses and standard deviation for the reference case and the

diversion scenarios for the nine central guide tubes. The range of the normalized detector response

calculated among the guide tubes is also reported. The values refer to 239Pu fission chambers covered

by the SINRD filters (left), and to bare 238U fission chambers (right). ................................................. 100


diversion scenarios for the peripheral guide tubes. The range of the normalized detector responses

calculated among the guide tubes is also reported. The values refer to 239Pu fission chambers covered

by the SINRD filters (left), and to bare 238U fission chambers (right). ................................................. 102


diversion scenarios for the nine central guide tubes. The range of the normalized detector responses

calculated among the guide tubes is also reported. The values refer to 235U fission chambers (left),

and to 238U fission chambers (right). ................................................................................................... 104


diversion scenarios for the nine central guide tubes. The range of the normalized detector responses

calculated among the guide tubes is also reported. The values refer to ionization chambers. ......... 104


diversion scenarios for the peripheral guide tubes. The range of the normalized detector response

calculated among the guide tubes is also reported. The values refer to 235U fission chambers (left),

and to 238U fission chambers (right). ................................................................................................... 106


diversion scenarios for the peripheral guide tubes. The range of the normalized detector response

calculated among the guide tubes is also reported. The values refer to ionization chambers. ......... 106

ix

List of tables Table 2-1: Overview of NDA techniques for spent fuel measurements. The acronyms of the

techniques are included in the list of acronyms at the beginning of the thesis. .................................... 8

Table 4-1: Energy-integrated neutron fluxes chosen for the definition of the SINRD signature. ......... 30

Table 4-2: Ratios calculated for the selection of the SINRD signature. The statistical uncertainty of the

values is within 1%. ............................................................................................................................... 31

Table 4-3: Ratios calculated for the selection of the SINRD signature. The statistical uncertainty of the

values is within 1%. ............................................................................................................................... 31

Table 4-4: First criterion for the optimization of the SINRD filters calculated for different thickness of

Gd and Cd and in case of a 239Pu fission chamber. The values are the ratios between the detector

response integrated over the 0.2 - 0.4 eV energy region and the absolute integral value. ................. 33

Table 4-5: Second criterion used for the optimization of the SINRD filters. The values are the ratio

between the integral of the negative contributions over the complete energy range and the absolute

integral value. ........................................................................................................................................ 35

Table 4-6: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 239Pu

fission chamber. .................................................................................................................................... 36

Table 4-7: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 235U

fission chamber. .................................................................................................................................... 36

Table 4-8: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 3He

proportional counter. ............................................................................................................................ 36

Table 4-9: Expected neutron counts for different combinations of SINRD filters. Results for a 239Pu

fission chamber. The uncertainty reported in the table takes into account the finite measurement

time. ...................................................................................................................................................... 38

Table 4-10: Expected neutron counts for different combinations of SINRD filters. Results for a 235U

fission chamber. The uncertainty reported in the table takes into account the finite measurement

time. ...................................................................................................................................................... 38

Table 4-11: Expected neutron counts for different combinations of SINRD filters. Results for a 3He

proportional counter. The uncertainty reported in the table takes into account the finite

measurement time. ............................................................................................................................... 38

Table 4-12: Expected neutron counts for different combinations of SINRD filters. Results for a 10B

proportional counter. The uncertainty reported in the table takes into account the finite

measurement time. ............................................................................................................................... 38

Table 4-13: Expected neutron counts for a bare 238U fission chamber. The uncertainty reported in the

table takes into account the finite measurement time. ....................................................................... 39

Table 4-14: Contribution of individual nuclides on the SINRD signature. The values have a statistical

uncertainty lower than 0.1%. ................................................................................................................ 40

Table 4-15: SINRD signature (RSI) and RTH ratio for several fuel compositions. The statistical

uncertainty of the values in the table was also reported. .................................................................... 45

Table 4-16: SINRD signature calculated for detectors with different active lengths. ........................... 46

Table 4-17: SINRD signature calculated for different detector cover by the SINRD filters. .................. 48

Table 4-18: SINRD signature calculated for changes in the nominal thickness value of the SINRD

filters. The statistical uncertainty is reported in the table and is comparable to the variation of the

SINRD signature for the cases considered in the study. ....................................................................... 49

Table 5-1: Characteristics of the Gd and Cd samples used for the transmission measurements. All

samples were in the form of a metal foil or disc. .................................................................................. 55

x

Table 5-2: Characteristics of the Gd and Cd samples used for the validation measurements. All

samples were in the form of a metal disc of 80 mm diameter. ............................................................ 57

Table 5-3: Relative difference between the detector responses to the transmitted flux calculated with

different data libraries and different SINRD filters. For each SINRD filter the results compared to the

value obtained with ENDF/B-VII.0. ........................................................................................................ 60

Table 6-1: Number of fuel pins included in each area identified in Figure 6-4 for the different guide

tubes. The fuel assembly contains a total of 264 fuel pins. .................................................................. 70

Table 6-2: Percentage contribution to the importance function from different sections of the fuel

assembly. Values for the 235U fission chamber...................................................................................... 71


assembly. Values for the 238U fission chamber...................................................................................... 71


assembly. Values for the ionization chamber. ...................................................................................... 72

Table 6-5: Average normalized detector responses and standard deviations for the nine central guide

tubes. The maximum difference calculated among the guide tubes is also reported. The statistical

uncertainty was around 0.1% for the neutron detectors and around 0.4% for the ionization

chambers. .............................................................................................................................................. 79

Table 6-6: Average normalized detector responses and standard deviation for the peripheral guide



chambers. .............................................................................................................................................. 80



uncertainty was around 0.3% for fuel with burnup of 10 GWd/tU and around 0.1% for all other cases.

............................................................................................................................................................... 82

Table 6-8: Average normalized detector responses and standard deviation for the nine central guide


uncertainty was around 0.2% for the neutron detectors and within 0.8% for the ionization chambers.

............................................................................................................................................................... 83



uncertainty was around 0.2% for the neutron detectors and within 0.8% for the ionization chambers.

............................................................................................................................................................... 85



uncertainty was around 0.3% for fuel with burnup of 10 GWd/tU and around 0.1% for all other cases.

............................................................................................................................................................... 87



uncertainty was within 0.5% for the 239Pu fission chambers and around 0.1% for the 238U fission

chambers. ............................................................................................................................................ 100



uncertainty was within 0.5% for the 239Pu fission chambers and around 0.1% for the 238U fission

chambers. ............................................................................................................................................ 101

xi




chambers. ............................................................................................................................................ 103

Table 7-4: Average normalized detector responses and standard deviations for the peripheral guide



chambers. ............................................................................................................................................ 105

xiii

List of acronyms

AP Additional Protocol

ASEA Allmänna Svenska Elektriska Aktiebolaget (general Swedish electric company)

BU BUrnup

BWR Boiling Water Reactor

C/S Containment and Surveillance

CLAB Centralt mellanlager för använt kärnbränsle (central interim storage facility for spent

nuclear fuel)

CT Cooling Time

DCVD Digital Cherenkov Viewing Device

DDA Differential Die-Away

DDSI Differential Die-away Self-Interrogation

DG-ENER European Commission's Directorate General for ENERgy

ENDF Evaluated Nuclear Data File

EU European Union

EURATOM EURopean ATOMic energy community

eV electronVolt

GELINA GEel LINear Accelerator

GWd GigaWatt-day

HM Heavy Metal

ICVD Improved Cherenkov Viewing Device

IAEA International Atomic Energy Agency

IE Initial Enrichment

xiv

IRMM Institute for Reference Materials and Measurements

ITU Institute for TransUranium elements

JANIS JAva-based Nuclear Information Software

JEFF Joint Evaluated Fission and Fusion File

JENDL Japanese Evaluated Nuclear Data Library

JRC Joint Research Centre

LANL Los Alamos National Laboratory

LEU Low Enriched Uranium

LLNL Lawrence Livermore National Laboratory

LWR Light Water Reactor

MATLAB MATrix LABoratory

MCNPX Monte Carlo N-Particle eXtended

MOX Mixed OXide

NDA Non-Destructive Assay

NGSI Next Generation Safeguards Initiative

NGSI-SF Next Generation Safeguards Initiative – Spent Fuel

NMA Nuclear Material Accountancy

NPT Non-Proliferation Treaty

NNWS Non-Nuclear Weapon State

NRD Neutron Resonance Densitometry

NWS Nuclear Weapon State

ORIGEN-ARP Oak Ridge Isotope GENerator – Automatic Rapid Processing

ORNL Oak Ridge National Laboratory

xv

Pa Pascal

PDET Partial DEfect Tester

ppm part per million

PWR Pressurized Water Reactor

SCALE Standardized Computer Analyses for Licensing Evaluations

SCK•CEN StudieCentrum voor Kernenergie – Centre d'Étude de l'énergie Nucléaire (Belgian

nuclear research centre)

SFAT Spent Fuel Attribute Tester

SKB Svensk Kärnbränslehantering Aktiebolag (Swedish nuclear fuel and waste

management company)

SINRD Self-Indication Neutron Resonance Densitometry

ToF Time-of-Flight

UGET Universal Gamma-ray Emission Tomography

xvii

Summary Introduction

This Ph.D. project describes the development of non-destructive assay (NDA) methods for the

measurement of spent nuclear fuel and was conducted as a collaboration between the Université

Libre de Bruxelles (ULB) and the Belgian nuclear research centre SCK•CEN.

Spent nuclear fuel refers to fuel assemblies that are discharged from nuclear reactors after

irradiation and are transferred to an interim storage. This material contains radioactive elements that

are responsible for neutron and gamma emissions. Since the radioactive decay leads also to decay

heat, the spent fuel is normally stored under water to ensure appropriate cooling and provide

radiation shielding.

After irradiation spent fuel still contains about 2% of fissile materials (i.e. 235U and 239Pu), therefore

nuclear safeguards are applied to ensure that the material is used only for peaceful applications. The

plutonium contained in spent fuel is a major concern for the safeguards community because it

represents almost 80% of all material placed under safeguards today. Moreover, the total spent fuel

inventory increases with time due to the discharge of fuel assemblies from operating reactors.

Several NDA techniques are used for the safeguards verifications of spent fuel and additional

techniques are under development to provide more accurate measurements. Both passive and active

techniques are considered. Passive techniques rely on the spontaneous emission of radiation from

the spent fuel itself, whereas external sources are used for active techniques. The radiation

measured by each NDA method varies between neutrons, gamma-rays, and Cherenkov light. The

techniques investigated in this Ph.D. project are the Self-Indication Neutron Resonance Densitometry

(SINRD) and Partial Defect Tester (PDET).

Development of the spent fuel library

The SINRD and PDET techniques were investigated in this Ph.D. mainly through Monte Carlo

simulations, so the development of a reliable model is of importance. A reference spent fuel library

was developed in the first step of this Ph.D. research to obtain realistic material compositions and

source terms for spent fuel with different irradiation histories. The neutron and gamma-ray

emissions were defined in terms of source intensity and energy distribution.

The development of the fuel library allowed to understand the influence of the fuel irradiation

history on the isotopic composition and consequently on the neutron and gamma-ray source

strength of the spent fuel. Moreover, the comparison of the ORIGEN-ARP and ALEPH-2 codes used

for the calculations was also performed. The final goal in the frame of the Ph.D. research was to

xviii

generate material composition and source terms compatible with the format of the MCNPX code

that was used to investigate the NDA techniques. In a broader scope, the data from the fuel library

are publicly available and can be used for the development of other NDA methods or for studies on

the final disposal of spent fuel.

Monte Carlo study on SINRD

The passive neutron emission from spent fuel is measured with SINRD, and the attenuation of the

neutron flux around the 0.3 eV energy region is used to directly quantify the 239Pu mass. The

microscopic cross-section of 239Pu shows a strong resonance around 0.3 eV. The cross-section

expresses the interaction probability between a certain nuclide and an incoming neutron, and it is

specific for each nuclide. Therefore, significant neutron absorption is expected in correspondence of

the 0.3 eV resonance due to the presence of 239Pu in spent fuel.

A thin foil of either Gd or Cd is placed around the neutron detector during the SINRD measurements.

These elements were chosen because they show a cutoff energy for neutron absorption slightly

below and above 0.3 eV, respectively. Due to this property these materials are called SINRD filters. By

taking the difference of the neutron counts measured with the two filters, the neutron flux in the

energy region close to the 239Pu resonance is estimated.

The Monte Carlo modelling of the SINRD technique was used to identify the optimal measurement

setup. The approach proposed in this Ph.D. foresees the introduction of small neutron detectors in

the central guide tube of the PWR 17x17 fuel assembly.

The measurement of fuel assembly immersed in fresh and borated water was modelled, and

compared to the case of fuel kept in air and surrounded by a thick slab of polyethylene. The results

from the dry configuration showed the clearest reduction of the neutron flux due to the absorption

of 239Pu. The dry configuration was chosen as reference condition for the study and can be

representative of a measurement station in an encapsulation plant for the final verification of a fuel

assembly before the insertion in the storage canister for geological disposal.

The SINRD signature was defined as the ratio between the neutron counts in the fast and in the

0.3 eV resonance energy region. A 238U fission chamber was chosen as reference detector for the

estimation of the fast neutron flux, whereas a 239Pu fission chamber covered by a foil of either Gd or

Cd was proposed for the resonance region. The optimization of the SINRD filter thickness was also

carried out, by identifying a combination of filter thicknesses that has mainly contributions from

neutrons with energy close to 0.3 eV and minimizing the contributions from other energy regions.

xix

The detector response of a 239Pu fission chamber was compared with the results obtained with a 235U

fission chamber and with proportional counters containing 3He or 10B. The 239Pu fission chamber

showed the highest sensitivity to the 239Pu content in the fuel and this is the advantage of using the

self-indication technique. Similar results were calculated for the other detector types, with the

proportional counters that obtained the highest total neutron counts thanks to an higher neutron

sensitivity. Taking these results into account, a combination of SINRD filters of 0.1 mm Gd and 1.0

mm Cd was suggested for the measurements with fission chambers to maximize the total neutron

counts, whereas SINRD filters of 0.2 mm Gd and 1.0 mm Cd were proposed for the proportional

counters to maximize the SINRD signature.

The expected performance of SINRD in realistic scenarios was evaluated by considering a detailed

fuel composition. The masses of 239Pu and 235U were the parameters that influenced the most the

technique, while a few other nuclides had an impact in case of fuel with high burnup. The SINRD

signature increased with the burnup due to the 239Pu content, and with the initial enrichment due to

the 235U mass. Moreover, the SINRD signature was largely independent from the cooling time of the

fuel assembly, since the fissile content in the fuel does not depend on this parameter. The approach

proposed in this study showed also no significant effect from the positioning of detector in the guide

tube, detector length, and small variations from the nominal filter thickness. The incomplete

detector cover by the filters caused a change in the SINRD signature values due to the increase of the

thermal neutron components passing through the bare section of the detector.

SINRD experimental benchmark

The results from the Monte Carlo study of SINRD were supported by an experimental benchmark

carried out at the GELINA Time-of-Flight (ToF) facility of the Joint Research Centre (JRC) of Geel

(Belgium). The Time-of-Flight technique was chosen for the experimental validation of SINRD

because with this technique the energy distribution of a neutron beam can be measured. Time-of-

Flight measurements are traditionally used for neutron resonance spectroscopy and measure the

time that a neutron needs to travel a given distance. The measured time and the flight distance are

then related to the kinetic energy of the neutron.

Transmission measurements were performed to verify the quality of nuclear data used in the

simulations for the optimization of the SINRD filters. The comparison showed some differences

between the experimental transmission and the values calculated with the analytical approach.

However, the quality of nuclear data is sufficient to define the optimal thickness of the Gd and Cd

filters. The results of the experiments indicated that the combination of a Gd filter about

0.1 mm-thick and a 1.0 mm Cd filter is suggested for the measurement of spent fuel containing 239Pu.

xx

In addition, self-indication measurements were carried out to confirm the basic principle of SINRD.

Measurements on natural Cd samples with different thicknesses were used to mimic the presence of

239Pu in spent fuel. To obtain the self-indication technique a thin Cd sample was surrounded by 4 C6D6

liquid scintillators detecting the prompt -rays emitted after (n,) reactions. Measurements were

performed with Gd and Cd SINRD filters in the beam, and the thickness of these filters was optimized

for the detection of neutrons with energy close to the Cd resonance at 0.178 eV. The results obtained

with the self-indication geometry were compared with measurements with Frisch gridded ionization

chambers with thin deposits of 235U or 10B. The self-indication detector showed an enhanced

efficiency at the energy of the resonance of interest, i.e. the 0.178 eV resonance.

The results obtained for the self-indication experiments using the SINRD filters were very similar to

the values calculated with the ideal measurement using ideal filters and background subtraction. The

comparison of the results from the self-indication measurements with the values obtained for the

other detectors confirmed that the highest sensitivity is obtained using a neutron detector with an

enhanced efficiency for a resonance of the material of interest. Therefore, a 239Pu fission chamber is

recommended for the characterization of spent fuel by SINRD.

Monte Carlo study on PDET

The partial defect tester (PDET) consists of a set of neutron and gamma detectors to measure the

spontaneous emission from spent fuel. Several small detectors are simultaneously inserted from the

top in the guide tubes of a PWR fuel assembly. These locations are designed for the insertion of the

control rods when the assembly is loaded in the reactor core, and they are generally empty once the

assembly is stored in the spent fuel pool. The measurement is performed without moving the fuel

assembly from the storage location. This Ph.D. work considered 235U fission chambers and 238U fission

chambers for the detection of thermal and fast neutrons, respectively. In addition, ionization

chambers are used for the measurement of the gamma flux. The PDET detector was conceived for

the partial defect verification of spent fuel, as the removal of fuel pins alters the spatial distribution

of the neutron and gamma fluxes across the fuel assembly cross-section and allows the detection of

the diversion.

The model of the storage rack consisted of nine fuel assemblies in a 3x3 configuration with the PDET

inserted in the central fuel assembly. The importance function from each fuel pin of the fuel

assembly being measured was calculated with Monte Carlo simulations for the different neutron and

gamma-ray detectors. All fuel pins contributed in a significant way to the responses of both fission

chamber types, whereas the contributions to the ionization chambers were strongly localized in the

vicinity of the guide tube containing the detector.

xxi

The detector responses of the 235U fission chambers were compared with the results from 239Pu

fission chambers and proportional counters containing 3He and 10B, but no significant difference was

observed. Ionization chambers containing N and Xe as filling gas at different pressures were

compared as well, but the difference in the responses were within the statistical uncertainty.

The initial enrichment and cooling time of the spent fuel assembly did not affect significantly the

normalized detector responses across the fuel assembly cross-section, whereas the fuel burnup had

an effect within ±3% on the normalized neutron detector responses.

The influence of the irradiation history of the fuel assemblies in the storage rack on the normalized

detector responses calculated in the guide tubes of the central fuel assembly was evaluated. The

burnup of the central fuel assembly had the largest impact in case of storage racks with high burnup

fuel assemblies in the lateral and corner positions. The neutron detectors in the guide tubes at the

periphery of the assembly were mostly affected by the change in burnup of the central fuel

assembly, and differences within ±10% were calculated on the normalized detector responses

compared to the reference case.

Several storage rack configurations were also developed to estimate the influence of the lateral and

corner fuel assemblies, and for all cases the maximum differences on the normalized detector

responses compared to the reference case were obtained in the guide tubes close to the assemblies

with high burnup. Variations between -50% and +30% were calculated for different guide tubes in

the central fuel assembly. Because of the self-shielding effect of the fuel pins, the gamma-ray

detector responses were in general less influenced by the fuel assemblies with different burnup

compared to the responses of the fission chambers.

Comparison of the partial defect capabilities for SINRD and PDET

The SINRD and PDET detectors were compared with Monte Carlo simulations in terms of their ability

to detect fuel pins replaced by dummy pins. The current IAEA goal for partial defect testing is to

verify that at least 50% of the fuel pins are present in a fuel assembly. Therefore, a series of 12

diversion scenarios were created for this analysis, considering replacements from 50% to 15% of the

fuel pins of the measured fuel assembly. The diverted fuel pins were replaced by dummies made of

stainless steel with the same dimensions of the original spent fuel pins. A symmetric pattern was

chosen for all cases since it was considered the most challenging situation to detect. Uniform

diversion was taken into account, as well as cases with replacement of pins in the outer section and

from the inner section of the assembly.

xxii

For both techniques the insertion of multiple detectors in the different guide tube positions was

modelled and the detector responses were normalized to the maximum value obtained among the

guide tubes. Based on the Monte Carlo studies the detector responses of bare 238U fission chambers

as well as 239Pu fission chambers covered by SINRD filters were considered for SINRD, whereas the

responses of 235U fission chambers, 238U fission chambers, and ionization chambers were calculated

for PDET.

Significant differences in the normalized detector responses were observed for both SINRD and PDET

in the diversion scenarios with 50% of fuel pins replaced, and the most challenging scenario was

obtained with the replacement with a chess-board pattern.

In the case of SINRD the detector responses of the 238U fission chambers were more affected by the

fuel pins diversion than the 239Pu fission chambers covered by the SINRD filters. For both detector

types the average detector response showed a difference between -30% and +15% in the case of

50% dummy pins.

The results for the PDET detector showed that the detector responses of the gamma-ray detectors

have the highest sensitivity to the diversion scenarios. As for SINRD, the variation in the detector

responses among the guide tubes provides indications for the detection of diversion scenarios with

50% of dummy pins. Relative differences up to -30% from the reference case were calculated for the

ionization chambers in the guide tubes at the periphery of the assembly. The variations for the

neutron detectors were within ±20% for the peripheral guide tubes and within ±10% for the central

guide tubes.

The diversion of 50% of fuel pins was detected with both techniques in all scenarios considered in the

Ph.D. work; however, detailed considerations about the measurement uncertainty have to be

included in the analysis to determine the limit of detection for the fuel pins diversion.

xxiii

Résumé Introduction

Ce projet de doctorat décrit le développement de méthodes de contrôle non-destructif (CND) de

mesure du combustible nucléaire irradié et a été mené en collaboration entre l'Université Libre de

Bruxelles (ULB) et le centre de recherche nucléaire belge SCK•CEN.

Le combustible nucléaire irradié est l’ensemble des assemblages de combustible qui sont déchargés

des réacteurs nucléaires après irradiation et qui sont transférés dans un stockage intermédiaire. Ce

combustible contient des éléments radioactifs qui sont responsables des émissions de neutrons et de

rayons gamma. Puisque la désintégration radioactive produit également une quantité de chaleur

significative, le combustible irradié est normalement stocké sous eau pour assurer un

refroidissement approprié et pour fournir une protection contre les rayonnements.

Après l'irradiation, le combustible irradié contient encore environ 2% de matières fissiles (235U et

239Pu). Par conséquent, les mécanismes de protection nucléaire sont d’application afin de garantir

l’utilisation du matériau à des applications uniquement pacifiques. Le plutonium contenu dans le

combustible irradié est une préoccupation majeure pour la communauté de protection contre la

prolifération, car il représente près de 80% de toutes les matières placées sous protection

aujourd'hui. De plus, la décharge du combustible irradié des réacteurs en exploitation conduit à

l'augmentation cumulative de l'inventaire au fil du temps.

Plusieurs techniques de CND sont utilisées pour vérifier le combustible irradié et des techniques

supplémentaires sont en cours de développement pour fournir des mesures plus précises. Des

techniques passives et actives sont considérées. Les techniques passives reposent sur l'émission

spontanée de rayonnements provenant du combustible irradié, alors que des sources externes sont

utilisées pour des techniques actives. Le rayonnement mesurée est différent pour chaque méthode

CND, il est possible d’utiliser les neutrons, les rayons gamma ou le rayonnement Cherenkov. Les

techniques étudiées dans cette thèse de doctorat sont le Self-Indication Neutron Resonance

Densitometry (SINRD) et le Partial Defect Tester (PDET).

Développement de la bibliothèque du combustible irradié

Les techniques de SINRD et PDET ont été étudiées dans cette thèse principalement par le biais de

simulations Monte Carlo, de sorte que le développement d'un modèle fiable est d'une importance

primordiale. Une bibliothèque de référence pour le combustible irradié a été développée dans la

première étape de cette thèse pour obtenir des compositions de matériaux réalistes et termes

xxiv

sources pour le combustible irradié avec différentes histoires d'irradiation. Les émissions de neutrons

et de rayons gamma ont été définis en termes d'intensité de la source et de distribution d'énergie.

Le développement de la bibliothèque de combustible a permis de comprendre l'influence de

l'irradiation du combustible sur la composition isotopique et par conséquent sur l’intensité de la

source neutronique et de rayonnement gamma provenant du combustible irradié. En outre, une

comparaison entre ORIGEN-ARP and ALEPH-2.2 utilisés pour le calcul a également été réalisée.

L'objectif final dans le cadre du doctorat était de générer composition du matériel et termes de

source compatibles avec le code MCNPX qui a été utilisé pour étudier les techniques CND. Dans un

cadre plus large, les données de la bibliothèque de combustible sont accessibles au public et peuvent

être utilisées pour le développement d'autres méthodes CND ou pour des études sur le stockage

définitif du combustible irradié.

Étude Monte Carlo de la méthode SINRD

L'émission passive de neutrons du combustible irradié est mesurée avec SINRD, et l'atténuation du

flux de neutrons autour de la région d'énergie de 0.3 eV est utilisée pour quantifier directement la

masse de 239Pu. La section efficace microscopique de 239Pu montre une forte résonance autour de

0.3 eV. La section efficace exprime la probabilité d'interaction entre un certain nucléide et un

neutron entrant, et il est spécifique pour chaque radionucléide. Par conséquent, une absorption

importante de neutrons est prévisible en correspondance de la résonance à 0.3 eV dû à la présence

de 239Pu dans le combustible irradié.

Une mince feuille soit de Gd ou de Cd est placée autour du détecteur de neutrons pendant les

mesures de SINRD. Ces éléments ont été choisis parce qu'ils montrent une énergie de coupure pour

l’absorption de neutrons légèrement au-dessous et au-dessus de 0.3 eV. En raison de cette propriété,

ces matériaux sont appelés filtres SINRD. En prenant la différence entre les comptages de neutrons

mesurés avec les deux filtres, le flux de neutrons dans la région de l'énergie proche de la résonance

239Pu est estimé.

La modélisation Monte Carlo de la technique SINRD a été utilisée pour identifier la configuration

optimale de mesure. L'approche proposée dans cette thèse prévoit l'introduction de petits

détecteurs de neutrons dans le tube-guide central de l'assemblage combustible 17x17 d’un PWR.

La mesure de l'assemblage combustible immergé dans de l'eau pure et de l'eau borée a été

modélisée, et comparée au cas du combustible gardé dans l'air et entouré par une couche épaisse de

polyéthylène. Les résultats de la configuration sèche ont montré la diminution la plus claire de flux de

neutrons due à l'absorption de 239Pu. La configuration sèche a été choisie comme condition de

xxv

référence pour l'étude et peut être représentative d'une station de mesure dans une usine

d'encapsulation pour la vérification finale d'un assemblage combustible avant l'insertion dans le

container de stockage pour le stockage géologique.

La signature SINRD a été définie comme étant le rapport entre les comptages de neutrons dans la

zone rapide et dans la région d'énergie de résonance à 0.3 eV. Une chambre à fission 238U a été

choisie comme détecteur de référence pour l'estimation du flux de neutrons rapides, alors qu’une

chambre à fission 239Pu recouverte par une feuille soit de Gd ou de Cd a été proposée pour la région

de résonance. L'optimisation de l'épaisseur du filtre SINRD a également été réalisée, en identifiant

une combinaison de filtres conduisant principalement à des contributions de neutrons avec une

énergie proche de 0.3 eV et à minimiser les contributions provenant d'autres zones d'énergie.

La réponse du détecteur d'une chambre à fission de 239Pu a été comparée aux résultats obtenus avec

une chambre à fission de 235U et des compteurs proportionnels contenant du 3He ou du 10B. La

chambre à fission de 239Pu a montré la plus grande sensibilité à la teneur en 239Pu dans le

combustible, ce qui est l'avantage d'utiliser la technique de "self-indication". Des résultats similaires

ont été calculés pour les autres types de détecteurs, avec les compteurs proportionnels qui ont

obtenu le plus haut total de comptages de neutrons grâce à leur sensibilité élevée aux neutrons.

Compte tenu de ces résultats, une combinaison de filtres de SINRD de 0.1 mm de Gd et 1.0 mm de Cd

a été suggérée pour les mesures avec des chambres à fission pour maximiser le total des comptages

de neutrons, alors que les filtres SINRD de 0.2 mm de Gd et 1.0 mm de Cd ont été proposés pour les

compteurs proportionnels pour maximiser la signature SINRD.

Les performances attendues de la méthode SINRD dans des scénarios réalistes ont été évaluées en

considérant une composition de combustible détaillée. Les masses de 239Pu et 235U se sont avérées

les paramètres qui influencent le plus la technique, tandis que quelques autres nucléides ont eu un

impact en cas de combustible à haut burnup. La signature SINRD a augmenté avec le taux de burnup

en raison de la teneur en 239Pu, et avec l'enrichissement initial en raison de la masse en 235U. En

outre, la signature SINRD s’est révélée largement indépendante du temps de refroidissement de

l'assemblage combustible, étant donné que la teneur en matière fissile dans le combustible ne

dépend pas de ce paramètre. L'approche proposée dans cette étude n'a également montré aucun

effet significatif de la position du détecteur dans le tube-guide, de la longueur du détecteur, et de

faibles variations de l'épaisseur nominale du filtre. Le recouvrement incomplet du détecteur par les

filtres a provoqué un changement dans les valeurs de signature SINRD en raison de l'augmentation

des contributions en neutrons thermiques traversant la partie dénudée du détecteur.

xxvi

Référence expérimentale SINRD

Les résultats de l'étude Monte Carlo de la méthode SINRD ont été soutenus par une analyse

expérimentale réalisée au GELINA Time-of-Flight (ToF), une infrastructure du Joint Research Centre

(JRC) de Geel (Belgique). La technique Time-of-Flight a été choisie pour la validation expérimentale

de la méthode SINRD parce qu'avec cette technique, la répartition d'énergie d'un faisceau de

neutrons peut être mesurée. Des mesures Time-of-Flight sont traditionnellement utilisées pour la

spectroscopie par résonance neutronique et mesurer le temps dont un neutron a besoin pour

parcourir une distance connue. Le temps mesuré et la distance de vol sont ensuite liés à l'énergie

cinétique du neutron.

Des mesures de transmission ont été effectuées pour vérifier la qualité des données nucléaires

utilisées dans les simulations pour l'optimisation des filtres SINRD. La comparaison a montré

quelques différences entre la transmission expérimentale et les valeurs calculées par l'approche

analytique. Cependant, la qualité des données nucléaires est suffisante pour définir l'épaisseur

optimale des filtres Gd et Cd. Les résultats des expériences ont indiqué que la combinaison d'un filtre

de Gd environ 0.1 mm d'épaisseur avec un filtre Cd de 1.0 mm est appropriée pour la mesure du

combustible irradié contenant du 239Pu.

En outre, les mesures de self-indication ont été effectuées pour confirmer le principe de base de la

SINRD. Les mesures sur des échantillons de Cd naturel avec des épaisseurs différentes ont été

utilisées pour simuler la présence de 239Pu dans le combustible irradié. Pour construire un détecteur

avec une sensibilité élevée dans la région d'énergie proche de la résonance à 0.178 eV, un

échantillon de Cd mince a été entouré par 4 scintillateurs liquides C6D6 détectant les rayons

prompts émis par réactions (n, ). Pour améliorer la sensibilité aux alentours de 0.178 eV, les

mesures ont été effectuées avec des filtres SINRD en Gd et Cd interposés dans le faisceau. Les

résultats obtenus avec la géométrie de self-indication ont été comparées aux mesures effectuées

avec les chambres d'ionisation à grille de Frisch avec de minces dépôts de 235U ou 10B. Le détecteur

de self-indication présente une efficacité accrue à l'énergie de la résonance d'intérêt, à savoir la

résonance à 0.178 eV.

Les résultats obtenus pour les expériences de self-indication à l'aide des filtres SINRD sont très

semblables aux valeurs calculées avec la mesure idéale à l'aide de filtres idéaux et la soustraction de

fond. La comparaison des résultats des mesures de self-indication avec les valeurs obtenues pour les

autres détecteurs a confirmé que la sensibilité la plus élevée est obtenue en utilisant un détecteur de

neutrons avec une efficacité accrue pour une résonance du matériau d'intérêt. Par conséquent, une

xxvii

chambre à fission 239Pu est recommandée pour la caractérisation du combustible irradié par la

méthode SINRD.

Étude Monte Carlo sur PDET

Le Partial Defect Tester (PDET) consiste en un ensemble de détecteurs de neutrons et gamma pour

mesurer l'émission spontanée du combustible irradié. Plusieurs petits détecteurs sont insérés

simultanément par le haut dans les tubes-guides d'un assemblage combustible de PWR. Ces

emplacements sont conçus pour l'insertion des barres de contrôle, lorsque l'ensemble est chargé

dans le cœur du réacteur, et ils sont généralement vides une fois que l'assemblage est stocké dans la

piscine du combustible irradié. La mesure est effectuée sans déplacer l'assemblage de combustible à

partir de son emplacement de stockage. L’étude effectuée dans ce doctorat a considéré des

chambres à fission en 235U et 238U pour la détection des neutrons thermiques et rapides,

respectivement. En outre, des chambres d'ionisation sont utilisées pour la mesure du flux de rayons

gamma. Le détecteur PDET a été conçu pour la vérification de défaut partiel du combustible irradié,

comme le retrait des crayons combustibles modifie la distribution spatiale des flux à travers la

section transversale de l'assemblage combustible et permettre la détection de la déviation.

Le modèle du rack de stockage est composé de neuf assemblages combustibles dans une

configuration 3x3 avec le PDET inséré dans l'assemblage combustible central. La fonction

d'importance de chaque crayon combustible de l'assemblage combustible mesuré a été calculée par

simulations Monte Carlo pour les différents détecteurs de neutrons et de rayons gamma. Tous les

crayons combustibles ont contribué d'une manière significative aux réponses des deux types de

chambre de fission, tandis que les contributions aux chambres d'ionisation sont fortement localisées

dans le voisinage du tube-guide contenant le détecteur.

Les réponses des détecteurs des chambres à fission de 235U ont été comparées avec les résultats des

chambres de fission de 239Pu et des compteurs proportionnels contenant du 3He et 10B, mais aucune

différence significative n'a été observée. Les chambres d'ionisation contenant du N et Xe comme gaz

de remplissage à des pressions différentes ont été comparées de la même manière, mais la

différence dans les réponses se situe dans l'incertitude statistique.

L'enrichissement initial et le temps de refroidissement de l'assemblage combustible irradié n'ont pas

affecté de manière significative les réponses normalisées des détecteurs à travers la section

transversale de l'assemblage combustible, alors que le burnup du combustible a eu un effet de ± 3%

sur les réponses normalisées des détecteurs de neutrons.

xxviii

L'influence de l'irradiation des assemblages combustibles dans le rack de stockage sur les réponses

normalisées des détecteurs calculées dans les tubes-guides de l'assemblage combustible central a

été évaluée. Le burnup de l'assemblage combustible central a le plus grand impact dans le cas des

supports de stockage avec des assemblages combustibles à haut burnup dans les positions latérales

et de coin. Les détecteurs de neutrons dans les tubes-guides à la périphérie de l'assemblage ont été

principalement affectés par le changement de burnup de l'assemblage combustible central, et des

différences dans un intervalle de ± 10% ont été calculées sur les réponses normalisées des détecteurs

par rapport au cas de référence.

Plusieurs configurations de rack de stockage ont également été développées pour estimer l'influence

des assemblages de combustible latérales et de coin, et pour tous les cas, les différences maximales

sur les réponses normalisées des détecteurs par rapport au cas de référence ont été obtenues dans

les tubes-guides à proximité des assemblages avec des burnups élevés. Des variations entre -50% et

+ 30% ont été calculées pour les différents tubes-guides de l'assemblage combustible central. En

raison de l'effet d'auto-protection des barres de combustible, les réponses des détecteurs de rayons

gamma sont en général moins influencées par les assemblages de combustible avec un burnup

différent par rapport aux réponses des chambres à fission.

Comparaison des capacités de défauts partiels pour SINRD et PDET

Les détecteurs SINRD et PDET ont été comparés aux simulations Monte Carlo quant à leur capacité à

détecter les crayons combustibles remplacés par des crayons factices. L'objectif actuel de l'AIEA pour

les tests de défaut partiel est de vérifier qu’au moins 50% des crayons combustibles sont présents

dans un assemblage combustible. Par conséquent, une série de 12 scénarios de modification ont été

créés pour cette analyse, en tenant compte de remplacements de 50% à 15% des crayons

combustibles de l'assemblage combustible mesuré. Les crayons combustibles détournés ont été

remplacés par des substituts en acier inoxydable avec les mêmes dimensions que les crayons

combustibles irradiés d'origine. Un motif symétrique a été choisi pour tous les cas, car il a été

considéré comme la situation la plus difficile à détecter. Une modification uniforme a été prise en

compte, ainsi que les cas de remplacement des crayons dans la section extérieure et de la section

intérieure de l'assemblage.

Pour les deux techniques, l'insertion de plusieurs détecteurs dans les différentes positions des tubes-

guides a été modélisée et les réponses des détecteurs ont été normalisées à la valeur maximale

obtenue parmi les tubes-guides. Sur la base des études Monte Carlo, les réponses des détecteurs de

chambres à fission 238U non recouvertes, ainsi que de chambres à fission 239Pu recouvertes des filtres

xxix

SINRD ont été considérées pour SINRD, alors que les réponses des chambres à fission 235U, chambres

à fission238U, et les chambres d'ionisation ont été calculées pour PDET.

Des différences significatives dans les réponses normalisées des détecteurs ont été observées pour

SINRD et PDET dans les scénarios de dérivation avec 50% des crayons combustibles remplacés, et le

scénario le plus délicat a été obtenu avec le remplacement d'un motif en échiquier.

Dans le cas de SINRD les réponses des détecteurs des chambres à fission 238U ont été plus affectées

par le détournement des crayons combustibles que les chambres à fission 239Pu couverts par les

filtres SINRD. Pour les deux types de détecteurs des différences relatives comprises entre -30% et

+15% ont été calculées à partir du cas de référence compte tenu de ces deux types de détecteurs.

Les résultats obtenus pour le détecteur de PDET a montré que les réponses des détecteurs de rayons

gamma ont la plus grande sensibilité aux scénarios de modification. En ce qui concerne SINRD, la

variation dans les réponses moyennes des détecteurs et les écarts maximaux entre les tubes-guides

fournissent des indications pour la détection des scénarios de dérivation avec 50% de crayons

factices. Des différences relatives jusqu'à -30% par rapport au scénario de référence ont été calculées

pour les chambres d'ionisation dans les tubes-guides à la périphérie de l'assemblage. Les variations

pour les détecteurs de neutrons sont à moins de ±20% pour les tubes-guides périphériques et à ±10%

pour les tubes-guides central.

Ces résultats donnent une estimation initiale des capacités des deux techniques pour la détection

partielle de défauts. Cependant, des considérations détaillées sur l'incertitude de mesure doivent

être incluses dans l'analyse afin de déterminer la limite de détection de remplacement de crayons

combustibles.

1

1 Introduction

1.1 The framework of nuclear safeguards

After the discovery of nuclear fission in 1939, nuclear energy was used both for civil and military

purposes. On the one hand the first man-made self-sustaining nuclear chain reaction was achieved in

December 1942 with the Chicago Pile 1, and the first nuclear power plants were connected to the

power grid during the 1950s. On the other hand the first nuclear weapon was tested in July 1945 at

the Trinity Site in the USA, followed by the bombings of Hiroshima and Nagasaki in August 1945

(Rhodes, 1986).

The risk of the proliferation of nuclear weapons after the Second World War led to the foundation of

the International Atomic Energy Agency (IAEA) in 1957. According to the IAEA statute (IAEA, 1989),

the Agency shall support the contribution of nuclear energy to world-wide peace, health, and

prosperity. Moreover, it shall ensure that nuclear technology and knowledge is not used to pursue

military purposes.

The Article III.5 of the IAEA statute authorizes the Agency "to establish and administer safeguards

designed to ensure that special fissionable and other materials, services, equipment, facilities, and

information ... are not used in such a way to further any military purpose" (IAEA, 1989). The legal

basis was granted by the treaty on the non-proliferation of nuclear weapons (NPT) (IAEA, 1970). Five

countries signatories of the NPT – United States, Russian Federation, United Kingdom, France, China

– are defined nuclear weapon states (NWS) since they detonated nuclear weapons before the entry

into force of the treaty, whereas all other countries are considered non-nuclear weapon states.

By signing the NPT all parties agree to enhance collaboration on the peaceful uses of nuclear energy,

and accept safeguards measures to prevent the dissemination of nuclear weapons. Moreover, the

NWS agree to reduce the existing stockpiles and to reach nuclear disarmament.

The technical objective of safeguards was defined in the INFCIRC/153 (IAEA, 1972) as "the timely

detection of diversion of significant quantities of nuclear material from peaceful nuclear activities to

the manufacture of nuclear weapons or of other nuclear explosive devices or for purposes unknown,

and deterrence of such diversion by the risk of early detection."

The shortcomings of the NPT verification regime were exposed during 1990s with the discovery of

clandestine military programs for the development of nuclear weapons in Iraq and North-Korea.

These countries were part of the NPT but were able to achieve advanced nuclear weapons programs.

2

As a response to these findings the IAEA drafted the Additional Protocol (AP) (IAEA, 1997), which

supports the NPT and focuses on the detection of clandestine military programs. The AP grants more

access rights to the inspectors and the state declaration covers a wider range of facilities and

activities compared to the NPT verification regime based on the INFCIRC/153.

The two agreements are complementary to each other; the NPT ensures the absence of diversion of

declared nuclear material and verifies the declared use of facilities, whereas the AP ensures the

absence of undeclared nuclear material and activities.

Focusing on the European scale, the EURATOM treaty (European Union, 2010) states that the

European Commission shall ensure that "ores, source materials and special fissile materials are not

diverted from their intended uses as declared by the users". To satisfy this requirement, the

EURATOM safeguards system was established as a set of controls and verifications covering the

nuclear fuel cycle from mining to final storage. The provisions for supply and safeguards obligations

must be satisfied also for agreements with non-member states and international organizations

(European Union, 2010). EURATOM safeguards, as part of the European Commission’s Directorate-

General for Energy (DG-ENER), serve also as focal point in the EU member states for the

implementation of the safeguards agreement and additional protocol with the IAEA (European

Commission, 2014).

1.2 Renewed interest for spent fuel measurement methods

Among the safeguards verification activities, the measurement of spent fuel was identified by

EURATOM safeguards as one of the major challenges. Thus significant R&D activities are carried out

within the European Commission and through collaborations with EU and non-EU countries and the

IAEA (Schwalbach, 2014). The main projects deal with the validation of a software to compare the

expected neutron and gamma count rates and the measurements with a Fork detector (Vaccaro,

2014). Moreover, the development of the partial defect tester and the gamma-emission tomography

are carried out (Goncalves, 2014; Tobin, 2014).

Similarly to EURATOM safeguards, the IAEA department of safeguards long term R&D plan for 2012-

2023 identified as topic with high priority the development of NDA methods for the spent fuel

verification before the transfer to difficult-to-access storage (IAEA, 2013).

The spent fuel assemblies can be considered in a difficult-to-access area once they are placed in the

storage cask for the final disposal (Fattah, 1990). Since the first-of-a-kind geological repository is

scheduled to receive spent fuel from 2022 (Park, 2014), there is a strong interest at international

level to develop more precise NDA techniques for this purpose (Tobin, 2013).

3

The Swedish and Finnish support programs to the IAEA focused on the development of gamma-ray

emission tomography and Monte Carlo simulations and experimental tests gave encouraging results

(Honkamaa, 2014; Jansson, 2015). With this technique a series of medium-resolution gamma-ray

detectors rotates around a spent fuel assembly and by image reconstruction techniques a

tomographic image of the fuel pins is obtained. The instrument is dedicated to the partial defect

verification and burnup estimation of spent fuel.

The Next Generation Safeguards Initiative – Spent Fuel (NGSI-SF) is a joint project between several

universities and US national laboratories (Tobin, 2011). The NGSI project studied 14 NDA techniques

and after an external evaluation focused on four combinations of single techniques for further study

(Charlton, 2012). Both active and passive techniques were considered, covering both neutron and

gamma detectors. The integration with current NDA methods is foreseen in the design of the

prototypes.

1.3 Structure and objectives of the research project

This Ph.D. project aimed at investigating the Self-Indication Neutron Resonance Densitometry

(SINRD) and the Partial Defect Tester (PDET). These are two NDA techniques proposed for the

measurement of spent nuclear fuel. The research was based mainly on Monte Carlo simulations, and

a reference spent fuel library was developed as preparatory step to obtain a reliable composition and

source term of the spent fuel assembly.

An overview of the safeguards challenges for spent fuel measurements is given in Chapter 2. This

section includes also a description of the current NDA techniques used for spent fuel verifications, as

well as the two techniques studied in this Ph.D. topic.

The approach chosen in the study is outlined in Chapter 3, in addition to the overview of previous

work on SINRD and PDET. The Monte Carlo models are described in terms of fuel assembly geometry

and storage configurations. Chapter 3 includes also the explanation on the contribution of the

reference spent fuel library to this research project, and the approach developed for the neutron and

gamma-ray detectors responses concludes the section.

The results from the Monte Carlo study of the SINRD technique are summarized in Chapter 4. The

study was carried out by comparing the performance of SINRD in both dry and wet measurement

conditions. Then the SINRD signature was defined to obtain a parameter directly related to the 239Pu

content in the spent fuel assembly, and the thickness for the SINRD filters was determined. The

responses of several neutron detector types were compared to optimize the detection system.

4

Finally the performance in realistic scenarios were modeled by considering several spent fuel

compositions and a sensitivity study on the SINRD technique.

The experimental benchmark of SINRD is described in Chapter 5. The measurements were carried out

at the GELINA Time-of-Flight facility at JRC-IRMM in Geel (Belgium). The goals of the benchmark were

to verify the quality of the nuclear data used in the Monte Carlo simulations and to perform self-

indication experiments on a set of target samples. Moreover, the responses of several detector types

were compared.

The results from Monte Carlo study of the PDET detector are given in Chapter 6. Reference

conditions were defined first to calculate the contribution of single fuel pins on the detector

response. The responses of several detector types were compared through simulations. Then the

influence of the different fuel assemblies inserted in the storage rack was evaluated by considering

assemblies with different compositions and source terms.

The performance of SINRD and PDET for the partial defect detection is compared in Chapter 7.

Several diversion scenarios were developed with the Monte Carlo model by removing fuel pins from

complete fuel assemblies.

The final conclusions from the study of SINRD and PDET are included in Chapter 8.

5

2 Safeguards challenges of spent nuclear fuel

2.1 Properties of spent nuclear fuel

The term of spent nuclear fuel refers to fuel assemblies discharged from nuclear reactors after

irradiation (IAEA, 2012a). The irradiation of fresh fuel in the reactor core leads to the production of

several fission products and minor actinides, some of which are strong neutron absorbers. In

addition, the total fissile material content decreases due to the fuel burnup, and therefore the fuel

assemblies are unloaded from the core after a few irradiation cycles and transferred to an interim

storage.

Spent fuel contains many radioactive elements that are responsible for very strong neutron and

gamma emissions. Since the radioactive decay is responsible also for significant decay heat, the spent

fuel is normally stored under water to ensure appropriate cooling and provide radiation shielding.

The average composition of spent Low Enriched Uranium (LEU) fuel from Light Water Reactors (LWR)

with 3.5% initial enrichment, 33 GWd/tHM discharge burnup, and 3 years cooling time is shown in

Figure 2-1. About 95% of spent fuel consists of 238U, whereas slightly more than 3% are fission

products and minor actinides. The fissile materials, namely 235U and Pu, are almost in the same

quantity and combined account for 2% of the total mass (IAEA, 2012b).

Figure 2-1: Material composition of LEU spent fuel with 3.5% initial enrichment, 33 GWd/tHM burnup, and 3 years cooling time. (IAEA, 2012b)

6

Although it accounts only for a fraction of the total spent fuel, the residual fissile material is a main

concern for the safeguards community. The plutonium contained in spent fuel and in fuel assemblies

in reactor cores represents almost 80% of the material placed under safeguards today (IAEA, 2014).

According to (IAEA, 2012b) more than 11000 tHM of spent fuel are discharged every year worldwide

and the total amount of spent fuel will reach approximately 450000 tHM by 2020. Figure 2-2 shows

the trend of the cumulative amount of spent fuel, taking into account storage and reprocessing.

Figure 2-2: Cumulative inventory of spent fuel generated worldwide. (IAEA, 2012b)

2.2 Safeguards requirements for spent fuel verifications

The spent fuel verifications are carried out through nuclear material accountancy (NMA) and

containment and surveillance (C/S) measures, according to the principles outlined in INFCIRC/153

(IAEA, 1972).

The NMA is performed by counting the number of spent fuel assemblies and by NDA measurements

to detect the diversion of nuclear material. These activities are normally carried out during the

annual physical inventory verification (IAEA, 1998). The NDA techniques are used for the gross and

partial defect testing. The gross defect testing aims at differentiating spent fuel assemblies from

other materials such as fresh fuel or structural materials. The IAEA safeguards criteria (IAEA, 2009)

defines that the partial defect testing "should ensure that at least half of the fuel pins are present in

each fuel assembly".

The C/S measures are complementary to NMA and they are used to maintain the so-called

Continuity-of-Knowledge. C/S systems are seals and cameras installed to survey areas containing

nuclear material, with both operator and safeguards inspectorate keeping record of the transfer of

7

material in the surveillance zone. The records are then compared with the declared inventory change

to confirm the operator declaration. (IAEA, 1998)

2.3 Current non-destructive assays for spent fuel measurements

2.3.1 Overview of NDA techniques

Several NDA techniques were developed in the past for spent fuel measurements, and other

methods are investigated to improve the current capabilities.

A list of such NDA techniques is included in Table 2-1, and those techniques span from in-field use to

research projects. Both passive and active techniques are considered. Passive techniques rely on the

spontaneous emission of radiation from the spent fuel itself, whereas external radiation sources are

used for active techniques. The radiation used by each NDA method is also specified, and varies

between neutron, gamma-ray, and Cherenkov light. No radiation is mentioned for calorimetry, since

the technique measures the increase of temperature of the medium surrounding the fuel assembly

due to the residual decay heat. Finally, the proposed goals of each technique are mentioned with

known issues. The known issues and limitations of the techniques will be discussed in detail in the

next chapters.

The first part of the table includes the NDA techniques used currently in-field, and more information

on each method is provided in the next sections. The SINRD and PDET techniques investigated in this

Ph.D. project are placed in the second part of the table and are described in details in the other

chapters of this thesis. The last section of the table includes a list of other NDA methods under

research, and detailed information on each technique can be found in (Honkamaa, 2014; Jansson,

2015; Tobin, 2013; Schillebeeckx, 2014).

Focusing on the two techniques investigated in this Ph.D., SINRD is the only passive NDA with the

goal of 239Pu quantification, and previous study on PDET suggested its use also for the burnup

estimation. Both SINRD and PDET are tested in this thesis for the partial defect detection.

8

Table 2-1: Overview of NDA techniques for spent fuel measurements. The acronyms of the techniques are included in the list of acronyms at the beginning of the thesis.

Technique Status Type Radiation Proposed goals Known issues

DCVD Field use Passive Cherenkov Gross defect; Detector alignment; measurement of long-cooled fuel partial defect

SFAT Field use Passive Gamma-ray Gross defect Self-shielding effect of fuel pins

Fork Field use Passive Neutron; gamma-ray

Gross defect Need for calibration curve

SINRD Research Passive Neutron 239Pu quantification; Detector availability;

partial defect measurements in air

PDET Prototype Passive Neutron; gamma-ray

Partial defect; Intrusive;

burnup estimation detector calibration

Calorimetry Prototype Passive --- Residual heat Long measurement time

UGET Prototype Passive Gamma-ray Partial defect Indirect measurement

DDA Prototype Active Neutron Multiplication; Need for external neutron source

verify operator data

DDSI Prototype Passive Neutron Multiplication Indirect measurement

NRD Demonstr. Active Neutron; gamma-ray

Isotopic quantification

Need for external neutron source

9

2.3.2 Digital Cherenkov viewing device

The Cherenkov radiation is emitted when a charged particle travels in a medium with a speed higher

than the speed of light in that medium (Phillips, 1991).

Cherenkov radiation in spent fuel is due to high-energy gamma-rays emitted from the radioactive

decay of fission and activation products. The interactions between these gamma rays and fuel

cladding or storage water produce electrons and positrons that have sufficient energy to emit

Cherenkov radiation (Phillips, 1991).

The Improved Cherenkov Viewing Device (ICVD) and the Digital Cherenkov Viewing Device (DCVD)

are among the main NDA instruments used by IAEA for gross and partial defect testing (Chen, 2009;

IAEA, 2011; Branger, 2014). These instruments detect the Cherenkov radiation induced when spent

fuel is stored under water in a spent fuel pool, and are shown in Figure 2-3.

Figure 2-3: Improved Cherenkov viewing device (ICVD, left, (IAEA, 2011)). Digital Cherenkov viewing device (DCVD, right, photo by Andy Gerwing, Channel Systems).

Both detectors intensify the ultraviolet light associated with the Cherenkov radiation and filter most

of the visible light coming from the measurement environment. The ICVD is a hand-held device

where one inspector can see the filtered image, whereas the DCVD is equipped with an external

monitor allowing the complete inspection team to view the image.

The measurements with both ICVD and DCVD are performed on the crane over the spent fuel pool

and therefore fuel movement or insertion of components in the storage pool is not required.

Figure 2-4 shows the color-enhanced images obtained from DCVD measurements of a BWR 8x8 and

of a PWR 17x17 fuel assembly. Since the Cherenkov radiation is emitted in water, the bright spots

visible on the images are the regions next to the fuel pins, whereas the dark areas are either the fuel

pins or the handle in the case of the BWR assembly.

10

Figure 2-4: Color-enhanced image obtained with the DCVD detector of BWR 8x8 (left, adapted from (Chen, 2003)) and a PWR 17x17 fuel assembly (right, (Chen, 2009)).

The measurements of Cherenkov radiation with ICVD and DCVD are relatively fast and do not require

the movement of fuel assemblies. On the other hand, the correct alignment between the detector

and the fuel assembly plays an important role in the data analysis. The detection of missing fuel pins

is challenging when the pins are substituted at the periphery of the assembly, or under the handle of

the BWR fuel assembly (Parcey, 2011).

Current research is carried out on the measurement of long-cooled fuel when this is stored close to

relatively short-cooled assemblies (Branger, 2014).

2.3.3 Spent fuel attribute tester

The Spent Fuel Attribute Tester (SFAT) measures the gamma-ray energy spectra from spent fuel and

it is used for the gross defect verification (IAEA, 2011).

The SFAT is a small medium resolution gamma-ray detector (e.g. NaI, CdZnTe) enclosed in a lead

collimator. A water-tight steel pipe is attached to the detector to provide additional collimation and

reduce the influence of neighboring assemblies in the storage rack. The detector system is immersed

in the spent fuel pool and the measurement is performed from the top of the fuel assembly. The

detector is connected through a water-tight cable to a multi-channel analyzer and data acquisition

system on the bridge crane above the storage pool (IAEA, 2011). An image of the SFAT with the

water-tight equipment is shown in Figure 2-5. A high-resolution HPGe detector has also been used

for R&D activities (Honkamaa, 2003).

11

Figure 2-5: Spent fuel attribute tester (SFAT) and water-tight collimator. (Chichester, 2009)

The presence of spent fuel is verified with SFAT by measuring the main gamma emission from 137Cs at

662 keV. Moreover, the measurement of the gamma-ray line from 144Pr at 2182 keV is possible for

fuel assemblies with cooling time shorter than 4 years, as well as emissions from 134Cs, 154Eu, and 60Co

(Janssen-Maenhout, 2008).

To confirm the presence of spent fuel, the intensity of the detected gamma-rays is compared with

measurements with the detector placed in the gap between neighboring fuel assemblies.

As for the Cherenkov viewing devices, the measurements with SFAT do not require the movement of

spent fuel. The use of SFAT is of interest when the measurement of Cherenkov radiation is difficult,

for instance when measuring fuel with low burnup and long cooling time, or when the water is not

clear enough (IAEA, 2011).

However, the SFAT gives information only from the top part of the fuel assembly, since the gamma-

rays emitted from the lowest section of the fuel are shielded by the fuel assembly itself. This

limitation rules out the SFAT for the verification of assemblies with partial length rods (Arit, 1995).

2.3.4 Fork detector

The Fork detector provides a measurement of the neutron and gamma emissions from a spent fuel

assembly. Most of the spontaneous neutrons are emitted by 242Cm and 244Cm, whereas 137Cs is the

main gamma emitter for fuel with a few years of cooling time. Several detector designs were

developed and are currently used in several locations by IAEA and DG-ENER inspectors (Rinard,

1988).

The version developed by Los Alamos National Laboratory (LANL) has a characteristic U-shape

polyethylene block, which hosts in each arm two 235U fission chambers and one ion chamber for the

12

neutrons and gamma-rays detection, respectively (Rinard, 1984). One of the fission chambers is

wrapped by a cadmium foil in order to be sensitive mainly to epithermal neutrons, whereas the other

fission chamber is bare and is sensitive mainly to thermal neutrons.

Another version of the Fork detector was developed by SCK•CEN (Carchon, 1987; Carchon, 1994),

and contains only one fission chamber and one ionization chamber in each detector arm. In this

design, the polyethylene block is surrounded by a cadmium sheet to absorb thermal neutrons and a

stainless steel housing (Borella, 2010). Both versions are shown in Figure 2-6.

Figure 2-6: Fork detector developed by LANL (left, (Antech, 2015)). Fork detector developed by SCK•CEN (right, (Borella, 2011)).

The Fork detector is introduced in the spent fuel pool and the measured fuel assembly is lifted

partially from the storage rack position. The arms of the detector surround the fuel assembly and the

neutron and gamma counts are collected simultaneously. In general, only one measurement close to

the central axial region of the assembly is required for the verification (IAEA, 2011).

The measurement of neutron and gamma emissions is used for the gross defect testing, whereas the

neutron signals coming from the bare and Cd-covered fission chambers give information on the

boron content in the spent fuel pool (Phillips, 1991). The ratio between the neutron and gamma

signals gives information about the fuel assembly burnup and cooling time; however, calibration

curves must be established to achieve reliable burnup estimation (Rinard, 1986; Rinard, 1988;

Borella, 2011).

The operator declaration is also needed for the partial defect detection, even for some diversion

scenarios with 50% of fuel pins removed (Tiitta, 2002; van der Meer, 2005).

13

2.4 Techniques investigated in this Ph.D.

2.4.1 Self-indication neutron resonance densitometry

The self-indication neutron resonance densitometry (SINRD) measures the passive neutron emission

from spent fuel (Menlove, 1969). The SINRD technique aims at measuring the attenuation of the

neutron flux around the 0.3 eV energy region as a way to directly quantify the 239Pu mass in the spent

fuel.

The microscopic cross-section of 239Pu is shown in Figure 2-7 and a clear peak, called resonance,

around 0.3 eV is observed. The cross-section expresses the interaction probability between a certain

nuclide and an incoming neutron, and it is specific for each nuclide. Therefore, significant neutron

absorption is expected in correspondence of the 0.3 eV resonance due to the presence of 239Pu in

spent fuel.

Figure 2-7: Total microscopic cross-section of

239Pu according to the ENDF-B/VII.0 nuclear data library (Chadwick, 2006). The

right plot is a zoom on the energy region close to 0.3 eV.

A thin foil of either Gd or Cd is placed around the neutron detector during the SINRD measurements.

These elements were chosen because they show a cutoff energy for neutron absorption slightly

below and above 0.3 eV. Due to this property these materials are called SINRD filters. By taking the

difference of the neutron counts measured with the two filters, the neutron flux in the energy region

close to the 239Pu resonance is estimated.

2.4.2 Partial defect tester

The partial defect tester (PDET) consists in a set of neutron and gamma detectors to measure the

spontaneous emission from spent fuel (Sitaraman, 2007). With the PDET detector several small

detectors are simultaneously inserted from the top in the guide tubes of a PWR fuel assembly. These

10-9

10-7

10-5

10-3

10-1

101

100

101

102

103

104

105

To

tal cro

ss-s

ectio

n (

b)

Neutron energy (MeV)

239

Pu

10-7

10-6

101

102

103

104

105

T

ota

l cro

ss-s

ectio

n (

b)


239Pu

14

locations are designed for the insertion of the control rods when the assembly is loaded in the

reactor core, and they are generally empty once the assembly is stored in the spent fuel pool.

A prototype of the detector was developed by Lawrence Livermore National Laboratory (LLNL) for

the measurement of PWR 17x17 fuel assemblies, and Figure 2-8 shows a picture of the prototype.

The neutrons are measured with 235U fission chambers, whereas ionization chambers are used for the

measurement of the gamma flux.

The geometrical location of the guide tubes lead to a reference profile of the neutron and gamma

fluxes across the fuel assembly cross-section. The removal of fuel pins from a fuel assembly alters the

spatial distribution of the fluxes and should allow the detection of the missing fuel pins. Therefore,

the PDET detector was conceived for the partial defect verification of spent fuel.

Figure 2-8: Prototype of the PDET detector developed by LLNL. (Ham, 2013)

15

3 Approach used for the study of the non-destructive techniques

3.1 Overview of literature study

3.1.1 Previous research on SINRD

The SINRD technique has been studied in the 1960’s among other methods using the unique

resonance structure of the microscopic cross-sections of different nuclides (Putz, 1966; Menlove,

1969).

The first experiments on SINRD were conducted in air and used the epithermal neutron flux

produced in a nuclear reactor (Menlove, 1969). The neutron beam was collimated and transmitted

through a neutron filter and a sample of fissile material. A thin foil of Gd or Cd was used as neutron

filter in the measurements.

Several fission chambers were used as neutron detectors, using the same isotope as measured

sample and as active material in the detector according to the self-indication technique (Putz, 1966;

Harvey, 1970; Massimi, 2011). The results from the experiments showed that the mass of 239Pu

metallic samples thinner than 4 mm can be estimated within 1-3%.

The application of SINRD to spent fuel verification was recently proposed by Los Alamos National

Laboratory for the direct quantification of 239Pu (LaFleur, 2011; LaFleur, 2012; Hu, 2012; LaFleur,

2013; LaFleur, 2015). The Monte Carlo simulations considered both PWR 17x17 and BWR 9x9 fuel

geometries, with a single fuel assembly immersed under water. The measurement approach was to

place a set of fission chambers on one side of the assembly, and a comparison between 235U and

239Pu fission chambers was made through simulations.

A prototype was built using four 235U fission chambers with different configurations to measure

neutrons within specified energy windows. A bare fission chamber was chosen for the estimation of

the thermal neutron flux, whereas the fast neutron flux was measured with a fission chamber

surrounded by polyethylene and a thick boron carbide case. In addition, the neutron flux close to the

0.3 eV resonance of 239Pu was estimated by taking the difference between the count rate of two

fission chambers covered with either Gd or Cd filters. An additional Hf filter was added to reduce the

contribution from the neutron absorption of 240Pu, as the Hf filter absorbed the majority of neutrons

with energy close to the 240Pu resonance at 1.056 eV.

The simulations and measurements showed promising results for burnup estimation and for partial

defect detection (LaFleur, 2011). However, the outer three fuel rows accounted for the majority of

16

the detector signals, and the detector had a very high sensitivity to the precise positioning of the

detector close to the fuel assembly (Hu, 2012).

3.1.2 Previous research on PDET

The PDET detector was developed by Lawrence Livermore National Laboratory (LLNL) for partial

defect verification and for the burnup estimation of spent fuel.

Monte Carlo models of the PWR 14x14, 16x16, and 17x17 fuel assemblies were developed for the

study, and both a single detector and clusters of multiple detectors were considered (Sitaraman,

2010). The thermal neutron and gamma-ray fluxes in the guide tubes of the fuel assembly were

determined for a complete fuel assembly and for several diversion scenarios. The gamma-to-neutron

ratio was then calculated for all guide tubes and the values were normalized to the maximum ratio

among them. The results showed that the PDET can detect fuel pin diversions in the 10% range of the

total (Sitaraman, 2007; Sitaraman, 2009; Ham, 2009a). The experimental validation using a single

detector confirmed the promising results (Ham, 2009b; Ham, 2010).

The influence of neighboring assemblies was analyzed in (Ham, 2009a) by considering a 3x3 storage

rack filled with fuel assemblies with a burnup gradient. The nine fuel assemblies were then randomly

rotated and the neutron and gamma-ray detector responses were compared to the case with an

isolated fuel assembly. It was found that the fuel assembly being measured is the neutron signal was

heavily influenced by the neighboring assemblies, whereas the gamma signals were within 10% of

the isolated case (Ham, 2010). The presence of boron in the storage pool was also investigated. It

was found that the boron in the pool water leads to a flattening of the neutron profile over the fuel

assembly cross-section due to the increased neutron absorptions.

The burnup estimation was investigated for PWR 14x14 fuel assemblies estimating the neutron and

gamma-ray fluxes in the different guide tubes. Both the average value calculated for the sixteen

guide tubes and the average considering the four central guide tubes were considered (Ham, 2011).

For fuel assemblies with a few years of cooling time a linear relationship was found between the fuel

burnup and the average values of the gamma-ray flux.

A prototype for the measurement of PWR 17x17 fuel assemblies has been developed by LLNL with a

customized data acquisition system (Ham, 2012; Ham, 2015). The prototype contains 12 fission

chambers and 12 ionization chambers included in a stainless steel frame identical to the spider used

for the insertion of the control rods. By using this instrument, the complete spatial distributions of

the neutron and gamma fluxes can be obtained with two insertions in the fuel assembly. No

movement of the assembly from the storage rack is needed to perform the measurement.

17

The prototype was tested in January 2014 at the Swedish interim spent fuel storage facility CLAB in

Oskarshamn with the collaboration of SKB and DG-ENER (Ham, 2015). The experiments confirmed

the technical feasibility of the insertion of the PDET in the spent fuel assembly, although few

detectors experienced non-normal signals.

To complement the research carried out for PWR fuel assemblies, the use of PDET for the verification

of BWR fuel assemblies was studied by JRC-ITU (Rossa, 2013). The focus of the study was on the gross

defect detection considering the 8x8 ASEA BWR fuel geometry. This fuel assembly has only one water

hole that is used for better neutron moderation and uniform neutron profile across the assembly.

The use of this single measurement position as well as the insertion of multiple detectors among the

fuel pins was proposed. This approach showed that the replacement of complete fuel assemblies can

be detected using the normalized neutron and gamma-ray fluxes, and the normalized gamma-to-

neutron ratio. The operator declaration is needed for the verification.

3.1.3 Contributions from this Ph.D. project

The investigation of SINRD and PDET was the main topic of this Ph.D. project, and alternative

approaches were considered compared to the previous research on both techniques.

A different positioning of the detectors was proposed for the SINRD technique, by inserting the

detectors in the central guide tube of a fuel assembly to reduce the sensitivity to the detector

positioning. In addition, the SINRD technique was studied in the case of wet and dry storage to assess

the influence of the storage medium. The thickness of the SINRD filters was optimized to focus on

the energy region close to the 0.3 eV resonance and to minimize the contributions from other energy

regions. Several detector types were compared by considering both fission chambers and

proportional counters. The performance of SINRD in realistic scenarios was evaluated by including

detailed fuel composition in the model and by performing a sensitivity study on the detector

positioning and characteristics of the SINRD filters. A set of benchmark experiments were performed

to confirm the choice of the SINRD filters thickness and to compare the response of several detector

types.

The Monte Carlo study of the PDET calculated the contributions of individual fuel pins on the

detector signals, and compared the response of fission chambers and proportional counters

proposed during the study of SINRD. The effect of the neighboring fuel assemblies was investigated

in detail by modeling several spent fuel storage configurations with fuel assemblies with different

burnup.

18

The two NDA techniques were compared for their capabilities of partial defect detection. Several

diversion scenarios were developed by replacing spent fuel pins with dummies made of stainless

steel.

3.2 Description of the Monte Carlo models

3.2.1 Principles of the Monte Carlo methods

Monte Carlo methods refer to a broad group of statistical techniques that are intensively used for the

study of particle transport and for the development of radiation detectors (Spanier, 1969), (Lux,

1991), (Ancius, 2015). With these simulations the path of several particles in complex systems is

tracked as a sequence of free flights and collisions. The nature of the collisions is selected by random

numbers with probabilities based on nuclear data (e.g. reaction cross-section). (Weber, 2007)

The central limit theorem ensures that macroscopic quantities in the system (e.g. neutron flux) can

be inferred from the average behavior of the simulated particles (Spanier, 1969). Due to the

statistical nature of the method, the Monte Carlo simulations require the simulation of many

particles in order to reduce the statistical uncertainty of the results. (Pelowitz, 2011)

The Monte Carlo N-Particle eXtended (MCNPX) code is a radiation transport code developed by Los

Alamos National Laboratory (LANL) that was used in this Ph.D. study. The code started as an

extension of the general Monte Carlo N-Particle (MCNP) code in order to simulate the behavior of

additional particle types (e.g. proton) over a broad energy range. (Pelowitz, 2011) The input file

structure is almost identical between the two Monte Carlo codes.

3.2.2 Spent fuel assembly geometry

The SINRD and PDET techniques were investigated in this Ph.D. mainly through Monte Carlo

simulations, so the development of a reliable model is of importance. The geometrical specifications

included in (Gauld, 2009a) were used for the modeling of a PWR 17x17 fuel assembly.

The model is shown in Figure 3-1. This fuel assembly geometry contains 264 fuel pins and 25 guide

tubes that are used for the insertion of the control rods during the reactor operation. The fuel pins

had equal length, uniform composition, and the fuel pellets were enclosed in a zirconium cladding

(Williams, 2006). In general the guide tubes were modeled as hollow zirconium cylinders, but a thin

Gd or Cd foil was included for some simulations on SINRD to model the SINRD filters. The fuel spacer

grids, top and bottom nozzles were not included in the model, and the particles were discarded when

exiting the top and bottom surfaces.

19

The Monte Carlo simulations for SINRD calculated the neutron flux tally (so-called F4 type) in the

central guide tube of the fuel assembly, whereas for PDET the neutron and gamma-ray flux tallies

were computed for all guide tubes. In the case of the PDET detector the guide tubes were numbered

sequentially from the top left to the bottom right corner as shown in Figure 3-1. The tallied guide

tubes were left empty in the central axial region to calculate the radiation flux in the position

envisaged for the detector.

Figure 3-1: Monte Carlo model of the PWR 17x17 fuel assembly.

The fuel material composition were defined according to the reference spent fuel library (Rossa,

2013b) using the approach described in Section 3.3. The neutron source was modeled as the energy

distribution of the 244Cm spontaneous fission, which is a major neutron emitter for spent fuel with

cooling time between 1 and 100 years (Rossa, 2013b). This source distribution is included in the

MCNPX code as a Watt fission spectrum shown in Formula (3.1). The parameters a=0.906 MeV and

b=3.848 MeV−1 were used in the simulations (Pelowitz, 2011).

𝑓(𝐸) = 𝐶 exp (−𝐸

𝑎) sinh(√𝑏𝐸) (3.1)

The energy distribution of the gamma-ray source was specific of the fuel irradiation history and was

obtained from the reference spent fuel library, as well as the neutron and gamma-ray source

intensities (Rossa, 2013b).

3.2.3 Storage configuration

The model of the storage configuration in the case of SINRD consisted in a single fuel assembly and a

moderator. Both fresh water and water with 2200 ppm of boron were compared to assess the

20

influence of boron. In addition, the case of fuel stored in air and surrounded by a thick slab of

polyethylene was studied. The model for this configuration is shown on the left-hand side of Figure

3-2. The spent fuel with short cooling time is not traditionally stored in air because of the significant

decay heat emitted, but this configuration can be representative of an encapsulation plant where

spent fuel with long cooling time is verified before the final disposal (Park, 2014).

The model developed for the PDET detector considered nine fuel assemblies stored in a 3x3 storage

rack and it is shown on the right-hand side of Figure 3-2. The material composition and source

intensity of each fuel assembly can be defined in the simulation. The storage basket was immersed in

fresh water, and the walls separating the fuel assemblies are made of borated steel with 1.6% of

boron. This model was based on the Swedish interim storage facility CLAB (Eliasson, 2012). The

neutron and gamma-ray flux tallies were calculated only for the central fuel assembly in the storage

rack.

Figure 3-2: Monte Carlo model of the fuel storage configuration for SINRD (dry case, left) and for PDET (right).

3.3 Definition of the source term

3.3.1 Use of the spent fuel library in the research project

A reference spent fuel library was developed using the ORIGEN-ARP and ALEPH-2 simulation codes

(Gauld, 2009a; Stankovskiy, 2012). The library contained the material composition and source terms

for spent fuel with different irradiation histories. The neutron and gamma-ray emissions were

defined in terms of source intensity and energy distribution.

The first objective of the fuel library was to understand the influence of the fuel irradiation history on

the isotopic composition and consequently on the neutron and gamma-ray source strength of the

21

spent fuel. Moreover, a comparison of the two codes used for the calculations was also performed.

The final goal in the framework of the Ph.D. research was to generate automatically several input

cards compatible with the MCNPX code that was used to investigate the NDA techniques. In a

broader scope, the data from the fuel library are publicly available and can be used for the

development of other NDA methods or for studies on the final disposal of spent fuel.

3.3.2 Structure of the spent fuel library

Two simulation codes were used for the development of the spent fuel library. ORIGEN-ARP

(Gauld, 2009a) is a burnup code part of the SCALE package (ORNL, 2009) developed by Oak Ridge

National Laboratory (ORNL), whereas ALEPH-2 is a Monte Carlo burnup code developed by the

Belgian nuclear research centre SCK•CEN (Stankovskiy, 2012). Specific information on the codes can

be found in (Gauld, 2009a; Gauld, 2009b; Stankovskiy, 2010).

The PWR 17x17 fuel geometry is one of the most common design in nuclear power plants and

therefore was chosen to develop the library. The same irradiation conditions were set for both

codes, and the reference irradiation cycle had an average power of 40 MW/tU, duration of 360 days,

and 30 days of cooling time between two cycles. The irradiation history was first varied according to

three variables:

Initial enrichment (IE): 7 values from 2% to 5%, with increments of 0.5%;

Discharge burnup (BU): 14 values from 5 up to 70 GWd/tU, with increments of 5 GWd/tU;

Cooling time (CT): 30 values ranging from immediate discharge up to 3 million years.

The analysis of the neutron emissions obtained in these simulations is found in (Rossa, 2013b).

The second stage of the fuel library kept the initial enrichment fixed at 4.5%, whereas the burnup

and cooling time were chosen as in the previous section. This stage focused on the parameters of the

irradiation history, varying the average power, duration of the irradiation cycle, and the cooling time

between two irradiation cycles. The simulations were performed with the ORIGEN-ARP code and the

results are included in (Borella, 2014).

The fuel library was extended also to MOX fuel using the ORIGEN-ARP code (Borella, 2015). Four

different plutonium percentages were considered on the total fuel mass, ranging from 4 to 10%. The

Pu isotopic vector was obtained from (Thorne, 2002) and was kept constant for all simulations. The

cooling time values were chosen as in the previous sections, whereas the final burnup was limited to

60 GWd/tU due to the range of applicability of the ORIGEN-ARP code. The reference values for

22

average power, duration of the irradiation cycle, and the cooling time between two irradiation cycles

were chosen.

3.3.3 Data processing to generate the fuel material composition

The output files from ORIGEN-ARP and ALEPH-2 contain detailed information about the spent fuel as

a function of the irradiation time. This includes the isotopic composition, neutron and gamma

emission terms, source activity, and decay heat. Given the significant extension of the file, several

scripts were developed to retrieve only the information relevant for the spent fuel library.

As first step in the data processing the complete fuel isotopic composition as a function of irradiation

time was included in a text file. A MATLAB (MATLAB, 2015) script was written to process this input

file and to obtain the fuel material composition in a format compatible with MCNPX.

A separate script was prepared to use the results from ORIGEN-ARP or from ALEPH-2, since the codes

give the material composition with different normalization units (as g/tU and g/cm3, respectively).

The script calculates the reaction rate for each nuclide present in spent fuel as the product between

the mass concentration and a weight factor. The weight factor was obtained with dedicated Monte

Carlo simulations computing the total microscopic cross-section averaged over an isolethargic

neutron flux from 10−9 and 100 MeV (Borella, 2015). Finally the nuclides reaction rates are sorted in

descending order and the material card is written in the output file in g/tHM. The top of the output file

contains the parameters of the fuel irradiation history, the input and output filenames, the

percentage of the total reaction rate, and the percentage of the total mass. An extract of one output

file is shown in Figure 3-3.

Figure 3-3: Extract of the output file generated from the data processing to obtain the spent fuel material composition.

23

3.3.4 Data processing to generate the source term characteristics

Another MATLAB script was written to generate the gamma-ray source energy distribution in a file

format compatible with MCNPX. The energy distribution of the gamma-ray source is included in the

ORIGEN-ARP output file in a 74-group structure defined by the user from 10 keV to 10 MeV. The

distribution is first extracted in a text file from the ORIGEN-ARP output file. The script then generates

the source term distribution according to the group structure defined in ORIGEN-ARP and saves the

results in an output text file according to the MCNPX file format. At the top of the output file the

parameters of the irradiation history are summarized, together with the input and output filename,

and the total gamma-ray source intensity. An extract of the output file generated from the script is

shown in Figure 3-4.

A similar script was written to obtain the neutron source term definition, by summing the

contributions from (α, n) reactions and spontaneous fissions. In this case, a 238-group energy

distribution was used in the ORIGEN-ARP simulations. A set of simulations were performed varying

the parameters of the neutron energy distribution defined in Formula (3.1) and no influence was

observed in the results, therefore the neutron source energy distribution obtained from the MATLAB

script was not used in the Monte Carlo simulations.

Figure 3-4: Extract of the output file generated from the data processing to obtain the gamma-ray source energy distribution.

3.4 Determination of the detectors response

3.4.1 Neutron detectors

The responses of several neutron detectors were simulated and compared in this study. However,

the detectors were not included in the Monte Carlo model of the spent fuel but their response was

calculated analytically by using Formula (3.2). The detector response can also be obtained with

MCNPX by developing a complete model of the measurement equipment (Borella, 2013), but the

24

approach described in this section allows to reach a lower statistical uncertainty in the simulations.

Moreover, a detailed model of the detector is not needed for the early investigation of the SINRD

technique.

𝑁 = ∫ 𝜑𝑁(𝐸𝑛) 𝜎𝐷𝐸𝑇(𝐸𝑛) 𝑑𝐸𝑛𝐸𝑛

(3.2)

In general the response (N) of a neutron detector can be calculated as the product between the

incoming neutron flux (N) and the microscopic cross-section (DET) of the active material in the

detector itself. Both quantities are a function of the incoming neutron energy En. The neutron flux

was obtained from the MCNPX simulations, whereas the cross-section values were taken from the

ENDF/B-VII.0 nuclear data library (Chadwick, 2006) and averaged over 600 logarithmically-

interpolated energy bins between 10−9 and 20 MeV. The total neutron count is then calculated as the

product between the neutron detector response and the neutron source term calculated with the

reference spent fuel library.

Figure 3-5: Microscopic cross-sections used for the estimation of the neutron detector responses. Values are based on the ENDF/B-VII.0 nuclear data library.

The reference thermal neutron detector for SINRD was a 239Pu fission chamber, whereas a 235U fission

chamber was the reference detector for PDET. The fission cross-sections for these two isotopes were

used to simulate the response of the corresponding detectors. Moreover, proportional counters

containing either 3He or 10B were also taken into account by using the (n,p) and (n, α) reactions cross-

sections, respectively. The fission cross-section of 238U was used to estimate the fast neutron flux for

25

both SINRD and PDET. Figure 3-5 shows the microscopic cross-sections used for the estimation of the

neutron detector responses.

3.4.2 Gamma-ray detectors

The response of the gamma-ray detector was estimated for the study of the PDET detector, and

ionization chambers with different filling gases were compared. These detectors are normally

operated in the so-called current mode, and the measurement output is an electric current that is

related to the ionization of the filling gas in the detector (Knoll, 2010).

In this Ph.D. study the detector response was calculated with Formula (3.3) where the gamma-ray

detector response (P) was calculated as the product between the gamma-ray flux (P) and the

response function of the specific detector (fDET).

𝑃 = ∫ 𝜑𝑃(𝐸𝛾) 𝑓𝐷𝐸𝑇(𝐸𝛾) 𝑑𝐸𝛾𝐸𝛾

(3.3)

Both quantities were obtained from MCNPX simulations and are function of the incoming gamma-ray

energy E. The gamma-ray flux was calculated with the model of the spent fuel described in Section

3.2.2., whereas the response function was calculated to estimate the energy deposition in the

detector filling gas. The energy deposition is further related to the ionization in the filling gas and for

this reason Formula (3.3) was chosen to estimate the detector response of the ionization chamber.

The response function fDET was obtained by modelling the detector alone, as an aluminum cylinder

filled with either nitrogen or xenon. Two pressure values, 0.1 and 1 MPa, were compared for both

materials. These characteristics represent commercial detectors that can be used for spent fuel

measurements (Photonis, 2016). The transport of both photons and electrons was simulated to

obtain the response function, which was calculated as a function of the gamma-ray source energy as

the integral value of the energy deposition tally (F6 type) in the gas-filled cavity. The energy range of

the source term was divided into 23 bins from 50 keV to 5 MeV, and separate simulations were

performed defining the source with a uniform histogram distribution over a single energy bin. The

response functions calculated for the different detector types are compared in Figure 3-6.

26

Figure 3-6: Response functions used for the estimation of the gamma-ray detector responses.

27

4 Monte Carlo assessment of the self-indication neutron resonance

densitometry

4.1 Structure of the study

The results from the Monte Carlo study of the SINRD technique are described in this chapter. A

comparison of different moderators placed around the fuel assembly was first performed to identify

the setup with the strongest reduction of the neutron flux around 0.3 eV due to the 239Pu neutron

absorption. The energy-integrated neutron flux was then calculated in several energy regions and the

ratio between the values obtained in two energy regions was taken. The SINRD signature was

defined as the ratio with the highest sensitivity to the 239Pu content in the fuel.

The thickness of the SINRD filters was optimized with a set of criteria to maximize the detection of

neutrons with energy close to 0.3 eV and to minimize the contributions from other energy regions. In

addition, several detector types were compared and the thickness of the SINRD filters was adapted

taking into account the neutron counts and SINRD signature values obtained with each detector type.

Finally the influence of the spent fuel composition on the SINRD signature and the systematic effects

due to the detector positioning and non-ideal conditions of the SINRD filters were evaluated.

4.2 Influence of the moderator on the neutron flux

The Monte Carlo study of the SINRD technique first compared the detector response with different

moderators surrounding the fuel assembly. Both wet and dry conditions were considered in the

simulations. Fresh and borated water with a boron concentration of 2200 ppm were the moderators

chosen for the wet conditions, since borated water is generally used in spent fuel pools to provide

margin for criticality safety (Borella, 2011), and fresh water was taken as a comparison to evaluate

the influence of boron. The dry conditions were obtained with the fuel modelled in air and by placing

a thick slab of polyethylene around the assembly. A gap of two centimeters was placed between the

assembly and the polyethylene. Spent fuel with short cooling time is traditionally not stored in dry

conditions because air does not ensure enough heat removal capability. However, the dry conditions

are representative of an encapsulation plant where spent fuel assemblies with long cooling time are

inserted in the storage canister before entering the final repository (Nyström, 2007).

A set of simulations was carried out with increasing thickness of the moderator layer around the

spent fuel assembly. The fuel included in the simulations contained 238U, 16O, and 1% of 239Pu. It was

found that for all moderators a thickness larger than 12 cm does not change significantly the integral

value of the neutron flux (Rossa, 2015a). Given this result a moderator thickness of 12 cm was used

for the rest of the Monte Carlo simulations.

28

The energy distribution of the neutron flux calculated for the different moderators is shown in

Figure 4-1, and the values are expressed per unit of source particle. The results calculated for the dry

condition show the clearest reduction of the neutron flux close to the 239Pu resonance at 0.3 eV,

whereas the reduction is not as clear when the fuel is immersed in water.

The reason for this difference was analyzed with additional simulations by replacing the water within

the fuel assembly with dry air. The energy distribution obtained in these conditions is shown in

Figure 4-2. The results when the water is removed from the internal section of the fuel assembly

resemble the distribution obtained with polyethylene around the fuel. Once the water is included

among the fuel pins, the neutron scattering occurring within the fuel assembly causes a flattening of

the neutron flux energy distribution in the central guide tube.

Figure 4-1: Energy distributions of the neutron flux calculated in the central guide tube for several moderators. All simulations had 12 cm of moderator outside the fuel assembly, and for the wet conditions the water was also included within the fuel assembly.

10-9

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

101

10-9

10-8

10-7

10-6

10-5

Gu

ide

tu

be

flu

x (

cm

-2)


Fresh water

Borated water

Polyethylene

29

Figure 4-2: Energy distribution of the neutron flux calculated in the central guide tube for fresh and borated water. In this case the water is only outside the fuel assembly and the place among the fuel pins is filled with dry air. The dry case obtained with polyethylene is included for comparison.

Therefore, the reduction of the neutron flux due to the 239Pu absorption in the 0.3 eV resonance

region is less visible when the water is included within the fuel assembly. By removing the water

among the fuel pins, the neutron moderation occurs mainly outside the fuel assembly and the energy

distribution of the neutron flux in the central guide tube resembles the results obtained in the dry

conditions.

Considering that the SINRD technique is based on the attenuation of the neutron flux in the 0.3 eV

energy region, the optimal condition is met when there is no moderation within the fuel assembly.

Therefore the dry case obtained with the fuel stored in air and surrounded by a thick slab of

polyethylene was chosen for the rest of the work.

4.3 Definition of the SINRD signature

The neutron flux calculated in the guide tubes of the fuel assembly is proportional to the neutron

emission of the assembly and net multiplication of the considered geometry (Lebrun, 2001). For a

given geometry, by taking the ratio between the integrated neutron flux in two energy regions, the

neutron source term and multiplication factor cancel out and therefore a quantity that is sensitive

only to the energy dependence of the neutron flux is obtained. The goal of this section is to define

the SINRD signature as the ratio between the neutron flux in two energy regions that is related to the

239Pu content in the fuel assembly.

10-9

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

101

10-9

10-8

10-7

10-6

10-5

Gu

ide

tu

be

flu

x (

cm

-2)


Fresh water

Borated water

Polyethylene

30

Therefore, a set of energy-integrated neutron fluxes were chosen between 10-9 and 20 MeV for the

definition of the SINRD signature and are reported in Table 4-1. The table includes the identification

name, the corresponding energy interval, and a brief description of each integrated flux.

Table 4-1: Energy-integrated neutron fluxes chosen for the definition of the SINRD signature.

Name Emin Emax Description

TOT 1 meV 20 MeV Total neutron flux

TH 1 meV 0.2 eV Thermal neutron flux

RES 0.2 eV 0.4 eV Neutron flux close to 0.3 eV

EPI 0.4 eV 20 MeV Epithermal neutron flux

FA 0.1 MeV 20 MeV Fast neutron flux

The integral values of the neutron flux in the energy windows were calculated for several fuel

compositions. The fuel contained 238U, 16O, and a 239Pu mass corresponding to spent fuel with

different burnup. The quantities of 239Pu were taken from the reference spent fuel library (Rossa,

2013b).

A set of ratios was then calculated by taking the combinations between the energy-integrated fluxes

defined in Table 4-1 and the results are reported in Tables 4-2 and 4-3. The values of the ratios are

normalized to the case of fuel without 239Pu, and the statistical uncertainty is within 1%. The highest

sensitivity to the 239Pu content was obtained with the ratios using as numerator either the epithermal

or the fast neutron flux and as denominator the flux close to the resonance region. The epithermal

neutron flux is expected to be influenced by the many neutron absorbers contained in spent fuel,

therefore the SINRD signature (RSI) was defined in Formula (4.1) using the fast neutron flux (FA) and

the flux close to the 0.3 eV resonance (RES).

𝑅𝑆𝐼 =𝐹𝐴

𝑅𝐸𝑆 (4.1)

The selection of the energy regions performed for the results in Tables 4-2 and 4-3 is the ideal

calculation of the SINRD signature, as neutrons with energy outside the regions defined for the ratio

are not considered. However, in practice the neutron flux in these energy regions is estimated by

using appropriate neutron detectors and foils of neutron absorbing materials.

31

Table 4-2: Ratios calculated for the selection of the SINRD signature. The statistical uncertainty of the values is within 1%.

Fuel BU (GWd/tU)

239Pu content (kg/tU)

EPI / TOT EPI / TH EPI / RES EPI / FA

0 0.00 1.00 1.00 1.00 1.00

5 2.18 1.09 1.68 2.17 1.00

10 3.51 1.12 2.14 3.01 1.00

15 4.37 1.14 2.45 3.66 1.00

20 4.97 1.15 2.68 4.14 1.00

25 5.34 1.15 2.84 4.45 1.00

30 5.57 1.16 2.93 4.66 1.00

35 5.70 1.16 2.98 4.76 1.00

40 5.77 1.16 3.01 4.83 1.00

45 5.81 1.16 3.02 4.87 1.00

50 5.81 1.16 3.02 4.87 1.00

55 5.80 1.16 3.02 4.86 1.00

60 5.80 1.16 3.02 4.86 1.00

Table 4-3: Ratios calculated for the selection of the SINRD signature. The statistical uncertainty of the values is within 1%.

Fuel BU (GWd/tU)

239Pu content (kg/tU)

FA / TOT FA / TH FA / RES TOT / TH TOT / RES

0 0.00 1.00 1.00 1.00 1.00 1.00

5 2.18 1.09 1.68 2.17 1.54 1.99

10 3.51 1.12 2.13 3.01 1.90 2.68

15 4.37 1.14 2.45 3.66 2.15 3.22

20 4.97 1.15 2.68 4.13 2.34 3.60

25 5.34 1.15 2.84 4.45 2.46 3.86

30 5.57 1.16 2.93 4.66 2.53 4.03

35 5.70 1.16 2.98 4.76 2.57 4.11

40 5.77 1.16 3.01 4.83 2.60 4.16

45 5.81 1.16 3.02 4.86 2.61 4.19

50 5.81 1.16 3.02 4.86 2.61 4.19

55 5.80 1.16 3.02 4.86 2.60 4.19

60 5.80 1.16 3.02 4.86 2.60 4.19

Formula (4.2) shows the ratio of neutron counts used in this work for the calculation of the SINRD

signature. A bare 238U fission chamber was chosen for the measurement of the fast neutron flux (CF)

because of the energy threshold of the 238U fission cross section In addition, the neutron flux in the

resonance region was taken with two measurements with a 239Pu fission chamber wrapped with a

thin foil of either Gd or Cd. These foils are also called SINRD filters and are chosen because they

present a cutoff energy for the neutron absorption slightly below or above the resonance region.

Therefore, by taking the difference between the neutron counts in the two measurements (CGd, CCd)

32

the neutron flux around 0.3 eV is estimated. The 239Pu fission chamber was the reference detector

type to increase the sensitivity in that energy region due to the self-indication principle

(Fröhner, 1966). In addition, several other detector types were compared to the 239Pu fission

chamber in Section 4.4.2.

𝑅𝑆𝐼 =𝐶𝐹

𝐶𝐺𝑑 − 𝐶𝐶𝑑 (4.2)

Previous work from (LaFleur, 2011) proposed the use of a 235U fission chamber enclosed in

polyethylene and surrounded by boron carbide for the fast neutron flux measurement. In addition,

the neutron flux in the resonance region was taken with two measurements with a 235U fission

chamber wrapped with a thin foil of either Gd or Cd.

4.4 Setup optimization

4.4.1 SINRD filters

The SINRD filters are foils of Gd or Cd placed around the fission chamber to absorb neutrons below a

certain energy defined as cutoff. However, the cutoff energy depends on the filter thickness

(Rossa, 2015b) and several thickness values are compared in this section.

A direct method for the optimization of the SINRD filters thickness is to perform Monte Carlo

simulations including the SINRD filter in the model. A drawback for this approach is the long

computational time, since the simulation has to be repeated for each filter thickness.

The approach used for this study was to calculate the neutron flux in the guide tube of the fuel

assembly without the SINRD filter, and to calculate the transmitted flux through the filter with

Formula (4.3).

𝜑𝑡𝑟(𝐸𝑛) = 𝜑0(𝐸𝑛) exp [− ∑ 𝑛𝑘 𝜎𝑡𝑜𝑡,𝑘𝐷

𝑘

(𝐸𝑛)] (4.3)

The transmitted flux (tr) through a thin sample can be expressed with Formula (4.3), where En is the

neutron energy, 0 is the incoming neutron flux determined with the Monte Carlo simulation, 𝜎𝑡𝑜𝑡,𝑘𝐷

is the Doppler broadened total cross section and nk is the areal density of nuclide k. This relationship

is valid for a parallel neutron beam which is perpendicular to a homogeneous sample, under the

assumptions that neutrons are either absorbed or traverse the sample without interactions

(Schillebeeckx, 2014).

33

The Doppler broadened cross section values were averaged in MCNPX by dividing the energy range

between 1 meV and 20 MeV in 600 evenly-spaced logarithmically bins. The ENDF/B-VII.0 was the

nuclear data library used for the calculations (Chadwick, 2006).

A set of criteria was defined and calculated for the optimization of the SINRD filters for a 239Pu fission

chamber. The first criterion was to maximize the ratio between the difference of the detector

response through the Gd and Cd filters integrated over the 0.2-0.4 eV energy range and the absolute

value of the detector response integrated over the whole energy range. The values are reported in

Table 4-4 for fuel containing only 238U and 16O and the results refer to several combinations of the

SINRD filters. The share of the detector response in the resonance region compared to the whole

energy range increases with the increase of the Gd filter thickness and, for most cases, the decrease

of the Cd filter thickness.

Table 4-4: First criterion for the optimization of the SINRD filters calculated for different thickness of Gd and Cd and in case of a

239Pu fission chamber. The values are the ratios between the detector response integrated over the 0.2 - 0.4 eV energy

region and the absolute integral value.

Share RES No 239Pu

Gd 0.025 mm

Gd 0.050 mm

Gd 0.075 mm

Gd 0.100 mm

Cd 0.25 mm 0.403 ± 0.003 0.553 ± 0.005 0.641 ± 0.006 0.703 ± 0.007

Cd 0.50 mm 0.399 ± 0.002 0.566 ± 0.004 0.659 ± 0.005 0.722 ± 0.005

Cd 0.75 mm 0.406 ± 0.002 0.571 ± 0.004 0.664 ± 0.004 0.725 ± 0.005

Cd 1.00 mm 0.409 ± 0.002 0.573 ± 0.003 0.664 ± 0.004 0.724 ± 0.005

Cd 1.25 mm 0.410 ± 0.002 0.572 ± 0.003 0.663 ± 0.004 0.721 ± 0.005

Cd 1.50 mm 0.410 ± 0.002 0.571 ± 0.003 0.661 ± 0.004 0.719 ± 0.005

Cd 1.75 mm 0.410 ± 0.002 0.570 ± 0.003 0.659 ± 0.004 0.716 ± 0.005

Cd 2.00 mm 0.409 ± 0.002 0.569 ± 0.003 0.657 ± 0.004 0.713 ± 0.005

Share RES No 239Pu

Gd 0.150 mm

Gd 0.200 mm

Gd 0.250 mm

Gd 0.300 mm

Cd 0.25 mm 0.784 ± 0.009 0.836 ± 0.010 0.872 ± 0.012 0.898 ± 0.014

Cd 0.50 mm 0.801 ± 0.007 0.848 ± 0.007 0.879 ± 0.008 0.901 ± 0.009

Cd 0.75 mm 0.801 ± 0.006 0.845 ± 0.007 0.873 ± 0.007 0.892 ± 0.008

Cd 1.00 mm 0.797 ± 0.006 0.839 ± 0.006 0.866 ± 0.007 0.882 ± 0.007

Cd 1.25 mm 0.793 ± 0.006 0.834 ± 0.006 0.858 ± 0.007 0.874 ± 0.007

Cd 1.50 mm 0.789 ± 0.006 0.828 ± 0.006 0.852 ± 0.006 0.866 ± 0.007

Cd 1.75 mm 0.785 ± 0.006 0.823 ± 0.006 0.846 ± 0.006 0.859 ± 0.007

Cd 2.00 mm 0.781 ± 0.005 0.819 ± 0.006 0.841 ± 0.006 0.853 ± 0.007

Figure 4-3 shows the energy distribution of the difference between the detector response through a

thick Gd and a thin Cd SINRD filter (NGd, NCd). The plot shows a peak close to the 239Pu resonance at

34

0.3 eV but also significant negative contributions to the difference around the thermal neutron

energy region.

Figure 4-3: Energy distribution of the difference between the detector response through the Gd and Cd SINRD filters.

A second criterion was defined for the optimization of the SINRD filters to consider the impact of the

negative contributions observed in Figure 4-3. The negative components act as suppression of the

net detector response in the resonance region; therefore, the criterion is to minimize the share of

the negative contributions over the absolute integral value of the difference between the detector

response through the SINRD filters. The values calculated for the second criterion for different SINRD

filters are reported in Table 4-5. The negative contributions on the absolute value decrease with the

decrease of the Gd filter thickness and with the increase of the Cd filter thickness. The share of the

negative contributions shown in Table 4-5 are lower than 1% by combining a Cd filter thicker than

0.75 mm with any Gd filter.

The values included in Tables 4-4 and 4-5 show that the two optimization criteria for the SINRD filters

are mainly influenced by the thickness of the Gd filter, whereas they are almost independent from

the Cd filter thickness. Therefore a 1.0 mm Cd SINRD filter was selected for the rest of the

calculations to reduce the contributions from thermal neutrons and this is in line with the thickness

of Cd foils used as anti-overlap filters in previous measurements at GELINA (Brienne-Raepsaet, 1999).

The choice of the Gd SINRD filter thickness is based on the detector response and the comparison

among different detector types is carried out in the next section.

10-9

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

101

-5.0x10-5

0.0

5.0x10-5

1.0x10-4

1.5x10-4

NG

d -

NC

d


Gd: 0.30 mm, Cd: 0.25 mm

35

Table 4-5: Second criterion used for the optimization of the SINRD filters. The values are the ratio between the integral of the negative contributions over the complete energy range and the absolute integral value.

Share NEG No 239Pu

Gd 0.025 mm

Gd 0.050 mm

Gd 0.075 mm

Gd 0.100 mm

Cd 0.25 mm 0.037 ± 0.003 0.174 ± 0.004 0.255 ± 0.005 0.319 ± 0.006

Cd 0.50 mm <0.001 ± 0.002 0.003 ± 0.003 0.009 ± 0.004 0.013 ± 0.004

Cd 0.75 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.004 0.001 ± 0.004

Cd 1.00 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.004 <0.001 ± 0.004

Cd 1.25 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.004 <0.001 ± 0.004

Cd 1.50 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.004 <0.001 ± 0.004

Cd 1.75 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.003 <0.001 ± 0.004

Cd 2.00 mm <0.001 ± 0.002 <0.001 ± 0.003 <0.001 ± 0.003 <0.001 ± 0.004

Share NEG No 239Pu

Gd 0.150 mm

Gd 0.200 mm

Gd 0.250 mm

Gd 0.300 mm

Cd 0.25 mm 0.433 ± 0.007 0.549 ± 0.009 0.680 ± 0.011 0.835 ± 0.013

Cd 0.50 mm 0.019 ± 0.005 0.024 ± 0.006 0.030 ± 0.006 0.036 ± 0.006

Cd 0.75 mm 0.002 ± 0.005 0.003 ± 0.005 0.005 ± 0.005 0.007 ± 0.006

Cd 1.00 mm 0.001 ± 0.004 0.002 ± 0.005 0.003 ± 0.005 0.005 ± 0.005

Cd 1.25 mm 0.001 ± 0.004 0.002 ± 0.005 0.003 ± 0.005 0.004 ± 0.005

Cd 1.50 mm <0.001 ± 0.004 0.001 ± 0.005 0.002 ± 0.005 0.004 ± 0.005

Cd 1.75 mm <0.001 ± 0.004 0.001 ± 0.005 0.002 ± 0.005 0.003 ± 0.005

Cd 2.00 mm <0.001 ± 0.004 0.001 ± 0.005 0.002 ± 0.005 0.003 ± 0.005

4.4.2 Comparison of detector types

The optimization criteria used in the previous section were calculated for a 239Pu fission chamber as

this is the reference detector for the estimation of the neutron flux close to the 0.3 eV resonance

region. This detector is compared in this section in terms of SINRD signature and neutron counts with

a 235U fission chamber and with proportional counters containing 3He or 10B. The proportional

counters are not traditionally used for spent fuel measurements because of the sensitivity to

radiation. Therefore the results contained in this section are meant to offer an optimal performance

for these detectors, while more detailed considerations on the effective applicability to field

measurements are not tackled.

The SINRD signature was calculated by taking the ratio between the detector response of a bare 238U

fission chamber for the fast neutron flux and the response of the different detector types covered by

the Gd or Cd SINRD filter for the resonance region. The results refer to fuel containing a 239Pu mass

corresponding to fuel with 3.5% initial enrichment, 10 years of cooling time, and burnup up to 60

GWd/tU. The material composition was taken from (Rossa, 2013b), and the values were normalized

to the results obtained with fuel containing only 238U and 16O. Several Gd SINRD filters were

compared by calculating the SINRD signature in combination with a 1.0 mm Cd SINRD filter.

36

Table 4-6: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 239

Pu fission chamber.

Fuel burnup (GWd/tU)

239Pu mass (kg/tU)

Gd 0.05 mm Cd 1.0 mm




0 0 1.00 ± 0.01 1.00 ± 0.01 1.00 ± 0.01 1.00 ± 0.01 5 2.18 1.96 ± 0.01 2.09 ± 0.01 2.17 ± 0.02 2.21 ± 0.02

10 3.51 2.58 ± 0.01 2.83 ± 0.02 2.97 ± 0.02 3.05 ± 0.03 15 4.37 3.04 ± 0.02 3.37 ± 0.03 3.56 ± 0.03 3.67 ± 0.04 20 4.97 3.35 ± 0.02 3.75 ± 0.03 3.98 ± 0.04 4.12 ± 0.04 40 5.77 3.80 ± 0.02 4.30 ± 0.03 4.58 ± 0.04 4.76 ± 0.05 60 5.80 3.82 ± 0.02 4.32 ± 0.03 4.61 ± 0.04 4.79 ± 0.05

Table 4-7: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 235

U fission chamber.


239Pu mass (kg/tU)





0 0 1.00 ± <0.01 1.00 ± 0.01 1.00 ± 0.01 1.00 ± 0.01 5 2.18 1.65 ± 0.01 1.71 ± 0.01 1.75 ± 0.02 1.78 ± 0.02

10 3.51 2.05 ± 0.01 2.14 ± 0.02 2.19 ± 0.02 2.22 ± 0.03 15 4.37 2.33 ± 0.01 2.43 ± 0.02 2.49 ± 0.03 2.52 ± 0.03 20 4.97 2.52 ± 0.01 2.62 ± 0.02 2.69 ± 0.03 2.72 ± 0.04 40 5.77 2.79 ± 0.01 2.90 ± 0.02 2.97 ± 0.03 3.00 ± 0.04 60 5.80 2.80 ± 0.01 2.91 ± 0.02 2.98 ± 0.03 3.01 ± 0.04

Table 4-8: SINRD signature as a function of fuel burnup for several Gd SINRD filters. Results for a 3He proportional counter.


239Pu mass (kg/tU)





0 0 1.00 ± <0.01 1.00 ± 0.01 1.00 ± 0.01 1.00 ± 0.01 5 2.18 1.63 ± 0.01 1.67 ± 0.01 1.70 ± 0.01 1.72 ± 0.02

10 3.51 2.02 ± 0.01 2.08 ± 0.01 2.12 ± 0.02 2.14 ± 0.02 15 4.37 2.29 ± 0.01 2.35 ± 0.02 2.39 ± 0.02 2.42 ± 0.03 20 4.97 2.47 ± 0.01 2.54 ± 0.02 2.58 ± 0.02 2.60 ± 0.03 40 5.77 2.73 ± 0.01 2.80 ± 0.02 2.84 ± 0.03 2.86 ± 0.04 60 5.80 2.74 ± 0.01 2.81 ± 0.02 2.85 ± 0.03 2.87 ± 0.04

The results are shown in Tables 4-6 – 4.8 and for all detector types the SINRD signature increases

with the 239Pu content. The highest sensitivity was obtained with the 239Pu fission chamber and this

proves the advantage of using the self-indication technique. Almost identical values were calculated

for the SINRD signature in the case of 3He and 10B proportional counters, and the values for the latter

are not reported in this chapter. For all detector types the use of a thick Gd SINRD filter determined

an increase on the sensitivity of the SINRD signature to the 239Pu content.

The comparison of the detector types included also an estimation of the total counts obtained for

each case. The total count (C) of a neutron detector was estimated with Formula (4.4), where Ntot is

the total neutron emission from the spent fuel assembly, t is the measurement time, ndet is the mass

37

of active material in the detector, det is the microscopic cross section of the active material in the

detector, in is the neutron flux obtained with the MCNPX simulation, nfil is the number of atoms per

unit area of the neutron filter, and σfil is the filter total neutron cross section.

𝐶 = 𝑁𝑡𝑜𝑡 𝑡 𝑛𝑑𝑒𝑡 ∫ 𝜎𝑑𝑒𝑡 𝜑𝑖𝑛 exp(−𝑛𝑓𝑖𝑙 𝜎𝑓𝑖𝑙) 𝑑𝐸𝑛 (4.4)

The neutron emission of spent fuel was taken from (Rossa, 2013b) and the measurement time was

fixed to 1 hour for all calculations. The 239Pu and 235U fission chamber contained 263.9 mg of active

material, corresponding to a detector with 7 mm diameter and 2 m length, which is an extension of

the Photonis CFUE43 235U-loaded fission chamber (Photonis, 2016). The 3He proportional counter

design was based on a tube with 6.8 mm diameter, 76.2 mm length, and 40 atm of filling gas

pressure (GE-Reuter-Stokes, 2016). The 10B proportional counter design had a diameter of 7.5 mm, a

length of 2 m, and a 10B mass of 49 mg as in the Proportional Technologies Boron-Coated Straw

detector (Proportional, 2016).

The calculated total neutron counts are reported in Tables 4-9 – 4.12 for the different detector types.

The values are the differences between the neutron counts obtained with the individual SINRD filter.

Several burnup values were considered and also combinations of SINRD filters to evaluate the

influence of the Gd filter thickness. The neutron counts increase significantly with the fuel burnup

because of the increase of the fuel assembly source term (Ntot), and decrease by increasing the

thickness of the Gd SINRD filter. The results obtained for the proportional counters are higher

compared to the values of the fission chambers due to the detector sensitivity and large (n,p) and

(n,) neutron cross sections of respectively 3He and 10B. Considering these results, a 0.1 mm Gd

SINRD filter was proposed for the measurements with the fission chambers to maximize the total

neutron counts, whereas a 0.2 mm Gd SINRD filter was chosen for the proportional counters to

maximize the SINRD signature. Taking into account the optimization criteria defined in the previous

section, the use of the thinner Gd SINRD filter leads to a decrease of the share of the resonance

region on the total neutron flux and a decrease of the negative contributions. The combinations of

the SINRD filters identified in this section are significantly different from (LaFleur, 2011), that used a

0.025 mm Gd filter and a 3.0 mm Cd filter.

The fast neutron count was also estimated by considering a bare 238U fission chamber with the same

characteristics of the fission chambers considered earlier in this section. The results are shown in

Table 4-9 and the total counts are significantly lower compared to the other detector types due to

the low fission cross section of 238U.

38

Table 4-9: Expected neutron counts for different combinations of SINRD filters. Results for a 239

Pu fission chamber. The uncertainty reported in the table takes into account the finite measurement time.






5 1.62 x 103 ± 3.1% 1.06 x 103 ± 4.2% 8.31 x 102 ± 5.0% 6.91 x 102 ± 5.8%

10 5.41 x 103 ± 1.8% 3.47 x 103 ± 2.5% 2.68 x 103 ± 3.0% 2.21 x 103 ± 3.6%

15 1.94 x 104 ± 1.0% 1.23 x 104 ± 1.4% 9.44 x 103 ± 1.7% 7.76 x 103 ± 2.0%

20 6.51 x 104 ± 0.5% 4.09 x 104 ± 0.8% 3.12 x 104 ± 1.0% 2.56 x 104 ± 1.1%

40 1.30 x 106 ± 0.1% 8.09 x 105 ± 0.2% 6.13 x 105 ± 0.2% 5.01 x 105 ± 0.3%

60 6.33 x 106 ± 0.1% 3.93 x 106 ± 0.1% 2.98 x 106 ± 0.1% 2.44 x 106 ± 0.1%


U fission chamber. The uncertainty reported in the table takes into account the finite measurement time.






5 4.79 x 102 ± 7.3% 2.34 x 102 ± 13.4% 1.57 x 102 ± 19.2% 1.19 x 102 ± 24.8%

10 1.70 x 103 ± 4.2% 8.26 x 102 ± 7.8% 5.53 x 102 ± 11.2% 4.19 x 102 ± 14.5%

15 6.33 x 103 ± 2.2% 3.07 x 103 ± 4.2% 2.06 x 103 ± 6.1% 1.56 x 103 ± 8.0%

20 2.16 x 104 ± 1.2% 1.05 x 104 ± 2.4% 7.03 x 103 ± 3.4% 5.34 x 103 ±4.5%

40 4.43 x 105 ± 0.3% 2.15 x 105 ± 0.5% 1.44 x 105 ± 0.8% 1.10 x 105 ± 1.0%

60 2.16 x 106 ± 0.1% 1.05 x 106 ± 0.2% 7.04 x 105 ± 0.4% 5.35 x 105 ± 0.5%

Table 4-11: Expected neutron counts for different combinations of SINRD filters. Results for a 3He proportional counter. The

uncertainty reported in the table takes into account the finite measurement time.






5 2.09 x 104 ± 1.0% 1.02 x 104 ± 1.7% 6.78 x 103 ± 2.4% 5.08 x 103 ± 3.1%

10 7.44 x 104 ± 0.5% 3.64 x 104 ± 1.0% 2.41 x 104 ± 1.4% 1.81 x 104 ± 1.8%

15 2.77 x 105 ± 0.3% 1.36 x 105 ± 0.5% 9.02 x 104 ± 0.8% 6.77 x 104 ± 1.0%

20 9.49 x 105 ± 0.2% 4.65 x 105 ± 0.3% 3.09 x 105 ± 0.4% 2.32 x 105 ± 0.5%

40 1.95 x 107 ± <0.1% 9.54 x 106 ± 0.1% 6.36 x 106 ± 0.1% 4.78 x 106 ± 0.1%

60 9.49 x 107 ± <0.1% 4.65 x 107 ± <0.1% 3.10 x 107 ± <0.1% 2.33 x 107 ± 0.1%


B proportional counter. The uncertainty reported in the table takes into account the finite measurement time.






5 1.61 x 104 ± 1.1% 7.86 x 103 ± 2.0% 5.22 x 103 ± 2.8% 3.91 x 103 ± 3.6%

10 5.72 x 104 ± 0.6% 2.80 x 104 ± 1.1% 1.86 x 104 ± 1.6% 1.39 x 104 ± 2.1%

15 2.13 x 105 ± 0.3% 1.04 x 105 ± 0.6% 6.94 x 104 ± 0.9% 5.21 x 104 ± 1.1%

20 7.30 x 105 ± 0.2% 3.57 x 105 ± 0.3% 2.38 x 105 ± 0.5% 1.78 x 105 ± 0.6%

40 1.50 x 107 ± < 0.1% 7.34 x 106 ± 0.1% 4.89 x 106 ± 0.1% 3.67 x 106 ± 0.1%

60 7.29 x 107 ± < 0.1% 3.58 x 107 ± < 0.1% 2.38 x 107 ± < 0.1% 1.79 x 107 ± 0.1%

39

Table 4-13: Expected neutron counts for a bare 238

U fission chamber. The uncertainty reported in the table takes into account the finite measurement time.


Bare detector

5 1.54 x 101 ± 25.5%

10 6.81 x 101 ± 12.1%

15 2.88 x 102 ± 5.9%

20 1.06 x 103 ± 3.1%

40 2.41 x 104 ± 0.6%

60 1.18 x 105 ± 0.3%

4.5 Expected performances in realistic scenarios

4.5.1 Influence of the spent fuel composition on the SINRD signature

The SINRD signature was calculated in the previous section for fuel containing only 239Pu apart from

238U and 16O, while in this section the influence of other nuclides is estimated. The results of the

study were also published in (Rossa, 2015c).

The fuel composition was taken from the reference spent fuel library (Rossa, 2013b) by considering

spent fuel with 3.5% initial enrichment, 10 years of cooling time and burnup up to 60 GWd/tU. A set

of Monte Carlo simulations was carried out by adding in a stepwise approach individual nuclides to

the material composition.

Figure 4-4: SINRD signature as a function of the

239Pu content for fuel with different compositions. The results refer to a

239Pu fission chamber for the measurement of the neutron flux in the 0.3 eV resonance region, and the values are

normalized to the case of fuel containing only 238

U and 16

O.

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11

Detector: 239

Pu fission chamber

Filters: Gd 0.1 mm

Cd 1.0 mm

40/60 GWd/tU

20 GWd/tU

15 GWd/tU

10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

239Pu +

16O [1]

[1] + 235

U [2]

[2] + 241

Pu [3]

[3] + 240

Pu [4]

[4] + 241

Am [5]

50 nuclides

40

The SINRD signature was calculated by taking the ratio of the detector response in the fast and in the

resonance energy region, considering a bare 238U fission chamber and a 239Pu fission chamber

covered by SINRD filters of 0.1 mm Gd or 1.0 mm Cd. The results obtained with the other detector

types included in Section 4.4.2 are reported in Annex A. The values were normalized to the case

obtained for fuel containing only 238U and 16O and the results are shown in Figure 4-4.

The plot shows the SINRD signature for fuel containing only 239Pu, and results obtained when 235U,

241Pu, 240Pu, and 241Am were added sequentially in the simulation. These nuclides were chosen

because they have the highest macroscopic cross section in the region close to 0.3 eV and therefore

are expected to influence the most the SINRD signature. The results obtained with fuel containing

the 50 main neutron absorbers were added for comparison. The main nuclides that contribute to the

SINRD signature are 239Pu and 235U, where the first gives the main influence at burnup higher than 20

GWd/tU and the second is the main contributor at lower burnup. The other nuclides included in the

plot have a significant impact for fuel with burnup higher than 40 GWd/tU.

The percentage contribution to the SINRD signature of the individual nuclides is shown in Table 4-14

for two burnup values. In the case of low burnup fuel almost 90% of the SINRD signature value is due

to 239Pu and 235U, and the latter has the major contribution. Considering fuel with burnup of 60

GWd/tU 239Pu is responsible for 50% of the SINRD signature value, but many other nuclides give

significant contributions.

Table 4-14: Contribution of individual nuclides on the SINRD signature. The values have a statistical uncertainty lower than 0.1%.

Fuel burnup 10 GWd/tU 60 GWd/tU

239Pu 39.8 % 50.3 % 235U 48.7 % 5.9 %

241Pu 1.9 % 7.5 % 240Pu 3.2 % 10.6 %

241Am 0.8 % 6.7 %

Figure 4-5 shows the energy distribution of the detector responses in the two energy regions defined

for the SINRD signature in the case of fresh fuel with 3.5% initial enrichment. The energy distribution

for fuel containing only 238U and 16O is given for comparison. The results reveal that the detector

response around the 0.3 eV resonance region is reduced when 235U is added in the fuel composition,

whereas the opposite effect is observed in the fast energy region. These effects are due to the

fissions occurring in the fuel assembly. The neutron-induced fission is mainly initiated by the

absorption of thermal neutrons, leading to a decrease in the thermal neutron flux, whereas mainly

41

fast neutrons are emitted from the fission reaction, causing the increase in the detector response in

the fast energy region.

Figure 4-5: Energy distribution of the detector response in the regions selected for the SINRD signature. Results for fuel with only

238U and

16O, and fresh fuel with 3.5% initial enrichment.

The analysis of the energy distributions of the detector response shown in Figure 4-5 was also

conducted for spent fuel with burnup of 15 and 60 GWd/tU and the results are reported in Figures 4-

6 and 4-7. The detector response in the 0.3 eV resonance region is reduced when 239Pu is added in

the fuel for both burnup values. At low burnup a significant contribution is seen also from 235U,

whereas 241Pu, 240Pu, and 241Am show effects at high burnup. The energy distribution in the fast

energy region is mainly affected by 239Pu and 235U. The integral value of the detector response

increases when the fissile materials are added in the fuel composition, with main contribution from

239Pu for both burnup values and significant effect from 235U at low burnup. The reduction of the

detector response in the resonance region coupled with an increase in the fast energy region when

the individual nuclides are added in the fuel composition explains the increase of the SINRD signature

value with the 239Pu content shown in Figure 4-4.

0.1 1

0

1x10-4

2x10-4

3x10-4

N x

de

tecto

r re

sp

on

se

(co

un

ts)

Neutron energy (eV)

238U +

16O

Fresh fuel (3.5% IE)

0.1 1 10

0

5x10-7

1x10-6

2x10-6

2x10-6

Detector: 238

U FC

No filter

Detector: 239

Pu FC

Filters: Gd 0.1 mm

Cd 1.0 mm

N x

de

tecto

r re

sp

on

se

(co

un

ts)


42

Figure 4-6: Energy distribution of the detector response in the regions selected for the SINRD signature. Results for fuel with burnup of 15 GWd/tU.

Figure 4-7: Energy distribution of the detector response in the regions selected for the SINRD signature. Results for fuel with burnup of 60 GWd/tU.

0.1 1

0

1x10-4

2x10-4

3x10-4

N x

de

tecto

r re

sp

on

se

(co

un

ts)

Neutron energy (eV)

238U +

16O [0]

[0] + 239

Pu [1]

[1] + 235

U [2]

[4] + 241

Am

50 nuclides

0.1 1 10

0

5x10-7

1x10-6

2x10-6

2x10-6

BU: 15 GWd/tU

Det: 238

U FC

No filter

BU: 15 GWd/tU

Det: 239

Pu FC

Filters: Gd 0.1 mm

Cd 1.0 mm

N x

de

tecto

r re

sp

on

se

(co

un

ts)


0.1 1

0

1x10-4

2x10-4

3x10-4

N x

de

tecto

r re

sp

on

se

(co

un

ts)

Neutron energy (eV)

238U +

16O [0]

[0] + 239

Pu [1]

[1] + 235

U [2]

[4] + 241

Am

Full

0.1 1 10

0

5x10-7

1x10-6

2x10-6

2x10-6

BU: 60 GWd/tU

Det: 238

U FC

No filter

BU: 60 GWd/tU

Det: 239

Pu FC

Filters: Gd 0.1 mm

Cd 1.0 mm

N x

de

tecto

r re

sp

on

se

(co

un

ts)


43

Figure 4-8: SINRD signature as a function of the

239Pu content for fuel with different irradiation histories. The fuel

composition contained the 50 main neutron absorbers.

The impact of the irradiation history on the SINRD signature was evaluated by comparing the results

for fuel with several initial enrichment, burnup, and cooling time. The 50 main neutron absorbers

were included in the fuel composition and the results are shown in Figure 4-8. The cooling time

ranged between 1 day and 50 years but the results are included only for the cases with 3.5% initial

enrichment; the rest of the values refer to a cooling time of 10 years. The values were normalized to

the case of fuel containing only 238U and 16O.

The SINRD signature increases with the fuel burnup and initial enrichment, due to the increase of

239Pu and 235U content, respectively. The fissile material does not change considerably with the

cooling time, and this explains the independence of the SINRD signature from this parameter.

Figure 4-8 shows that fuel with different initial enrichment and burnup have a similar SINRD

signature value; therefore, by using the SINRD signature only, the operator declaration on the fuel

irradiation history has to be used for the quantification of 239Pu. To provide the 239Pu mass without

information from the operator an additional parameter was calculated. The RTH ratio was defined in

Formula (4.5) as the ratio between the neutron counts in the fast (CF) and in the thermal energy

region (CT). The detector responses of a bare 238U fission chamber and of a bare 239Pu fission chamber

were calculated for CF and CT, respectively.

𝑅𝑇𝐻 =𝐶𝐹

𝐶𝑇 (4.5)

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11Detector:

239Pu fission chamber

Filters: Gd 0.1 mm

Cd 1.0 mm

40/60 GWd/tU

15 GWd/tU

20 GWd/tU

10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

IE: 3.5%

IE: 4.0%

IE: 4.5%

IE: 5.0%

44

The ratio RTH was calculated for the set of spent fuel compositions considered for the SINRD signature

and the values are shown in Figure 4-9. The results in the figure were normalized to the value

obtained for fuel containing only 238U and 16O. As for the SINRD signature, the RTH ratio increased

with the initial enrichment and it was largely independent on the cooling time. Only for cooling times

of 1 day the values were about 5% higher than for fuel with longer cooling times because of the

contribution from short-lived nuclides. The RTH values were almost constant for burnup values

between 5 and 20 GWd/tU and decreased by about 20% with burnup of 60 GWd/tU.

Figure 4-9: RTH ratio as a function of the

239Pu content for fuel with different irradiation histories. The fuel composition

contained the 50 main neutron absorbers.

Table 4-15 includes pairs of fuel assemblies with different initial enrichment (IE) and burnup (BU) that

had SINRD signatures within a 6% difference. The RTH ratio for the corresponding simulations is also

reported in the table, and it is significantly different for the two simulations in each pair. Therefore,

the combination of RSI and RTH can be used to determine the 239Pu mass without the support of the

operator declaration. The drawback for this approach is the need for the measurement with a bare

239Pu fission chamber in addition to the measurements required for the calculation of the SINRD

signature.

0 1 2 3 4 5 6 70

3

6

9

12

15

60 GWd/tU

Detectors: 239

Pu fission chamber (CT)

238

U fission chamber (CF)

40 GWd/tU

15 GWd/tU

20 GWd/tU

10 GWd/tU5 GWd/t

U

Fresh fuel

RT

H

239Pu content (kg/t

U)

IE: 3.5%

IE: 4.0%

IE: 4.5%

IE: 5.0%

45

Table 4-15: SINRD signature (RSI) and RTH ratio for several fuel compositions. The statistical uncertainty of the values in the table was also reported.

IE BU (GWd/tU)

RSI RTH (%)

4.5 5 5.41 ± 0.01 10.49 ± 0.02

3.5 10 5.64 ± 0.01 8.59 ± 0.01

4.5 10 6.58 ± 0.01 10.96 ± 0.02

3.5 20 6.85 ± 0.01 8.51 ± 0.01

3.5 40 7.51 ± 0.01 7.69 ± 0.01

4.5 15 7.43 ± 0.01 11.03 ± 0.02

4 40 8.07 ± 0.02 8.63 ± 0.01

5 15 7.99 ± 0.01 12.30 ± 0.02

4.5 40 8.82 ± 0.02 9.72 ± 0.01

5 20 8.64 ± 0.02 12.22 ± 0.02

4.5.2 Investigation of systematic effects on the SINRD technique

4.5.2.1 Effects due to the detector positioning and length

Previous work (LaFleur, 2011) showed that the accurate positioning of the SINRD detector is

important when the detector is placed on one side of the fuel assembly. Therefore, the impact of the

detector position within the guide tube on the calculated neutron flux was evaluated. The SINRD

filters were included in the model for the simulations performed in this section. The neutron flux tally

was calculated in a cavity in the guide tube that was completely covered by the SINRD filter.

Figure 4-10: View of the central guide tube in the Monte Carlo model and comparison of different positioning of the detector.

The study presented in (Rossa, 2015d) considered the reference case with the detector and SINRD

filter placed in the central position of the guide tube, as well as the two variations depicted in Figure

4-10. In the reference case the detector and the filter are centered in the guide tube, whereas with

the design variations the detector and filter are placed against the guide tube in the direction of a

46

fuel pin (Variation A) or facing the pitch between two fuel pins (Variation B). The results showed

almost no influence on the neutron flux calculated with the simulations.

The effect due to the active length of the detector was also investigated by comparing the results

obtained with different simulations. Starting from the reference value of 200 cm the detector active

length was reduced to a minimum of 1 cm. Previous work showed that the decrease of the detector

length lead to an increase in the transmitted flux in the energy region lower than 0.3 eV (Rossa,

2015d). The SINRD signature was calculated for the detectors with different lengths in the case of

fuel with 3.5% initial enrichment, burnup of 60 GWd/tU, and a cooling time of 10 years. The values

are shown in Table 4-16 and are normalized for each detector length to the corresponding case

obtained with fuel containing only 238U and 16O.

The SINRD signature has a slight increase with the decrease of the detector length, although in most

cases the variations are within the statistical uncertainty. The increase is due to the increase of the

difference between the transmitted neutron flux through the Gd and Cd filters.

Table 4-16: SINRD signature calculated for detectors with different active lengths.

Detector active length (cm)


235U fission chamber

3He proportional counter

10B proportional counter

200 7.52 ± 0.03 5.17 ± 0.03 4.86 ± 0.04 4.86 ± 0.04

100 7.54 ± 0.03 5.16 ± 0.03 4.84 ± 0.04 4.85 ± 0.04

20 7.66 ± 0.08 5.24 ± 0.07 4.90 ± 0.08 4.90 ± 0.08

10 7.66 ± 0.11 5.23 ± 0.10 4.87 ± 0.12 4.87 ± 0.12

5 7.71 ± 0.09 5.33 ± 0.08 4.97 ± 0.10 4.98 ± 0.10

1 7.78 ± 0.16 5.33 ± 0.13 4.98 ± 0.16 4.98 ± 0.16

Figure 4-11: Comparison of detectors with different lengths (L1 and L2). The track of a neutron crossing both detectors is

also shown with the indication of the incoming angles ( and ).

The increase of the neutron flux in the case of short detector is explained with Figure 4-11. Two

detectors with different lengths (L1 and L2) are depicted in the figure. Both detectors are fully

47

covered by a SINRD filter but this is not shown to simplify the drawing. The track of a neutron

entering the two detectors is also shown in the picture. The neutron crosses the SINRD filter of

detector with length L1 with incoming angle , and the filter of the detector with length L2 with angle

.

In general terms the straight path (l) of a neutron within the SINRD filter is given by Formula (4.6),

where d is the filter thickness and is the incoming angle.

𝑙 =𝑑

sin 𝜃 (4.6)

Comparing the two detectors shown in Figure 4-11 the straight path is longer for the longer detector,

since > . By increasing the straight path through the filter the transmitted neutron flux decreases

because the neutrons have to travel a longer distance through the filter. Therefore, the increase of

the straight path of the neutrons in the filters of the longer detector explains the reduction of the

neutron flux.

4.5.2.2 Effects due to the SINRD filters

The characteristics of the SINRD filters are of importance for the calculation of the SINRD signature,

therefore non-ideal filters were analyzed in this section. The effect of incomplete cover of the

detector by the SINRD filter was studied, as well as cases where the filter partially overlaps. The

scenarios are represented in Figure 4-12 and angles up to 30° were considered for both conditions.

Previous work (Rossa, 2015d) indicated that the overlap of the filters does not influence the energy

distribution of the detector response. On the contrary, the incomplete detector cover leads to the

increase of the thermal neutron contribution.

Figure 4-12: Zoom of the central region of the fuel assembly and comparison of the detector cover by the SINRD filter.

The impact of the imperfect detector cover on the SINRD signature was evaluated for the different

detector types. The SINRD signature is reported in Table 4-17 and refers to spent fuel with 3.5%

48

initial enrichment, burnup of 60 GWd/tU, and 10 years of cooling time. The values are normalized to

the case of fuel containing only 238U and 16O.

The SINRD signature calculated in the reference case with an ideal detector cover with the SINRD

filters is also reported in Table 4-17, and the title of each column describes the cover by the SINRD

filter. As in Figure 4-12, the filter overlap is indicated with +30° and the incomplete cover with -30°.

The values calculated for the 10B proportional counter were identical to the results for the 3He

proportional counter and therefore are not included in this chapter.

The highest SINRD signature was obtained for all detector types combining a measurement with the

incomplete detector cover by the Gd filter and with the overlap by the Cd filter. The lowest values

were obtained for a measurement with an overlapping Gd filter and an incomplete detector cover by

the Cd filter for the 239Pu fission chamber and in the reference case for the other detector types.

Negative values for the difference between the detector responses calculated with an overlap of the

Gd filter and with an incomplete detector cover by the Cd filter were obtained for the 235U fission

chamber and for the proportional counters. The corresponding values of the normalized SINRD

signature were not included in Table 4-17.

Table 4-17: SINRD signature calculated for different detector cover by the SINRD filters.

Detector Reference Gd: -30° Cd: -30°

Gd: -30° Cd: +30°

Gd: +30° Cd: -30°

Gd: +30° Cd: +30°


7.52 ± 0.03 7.75 ± 0.06 7.84 ± 0.04 7.14 ± 0.14 7.66 ± 0.06


5.17 ± 0.03 6.17 ± 0.08 7.36 ± 0.04 N. A. 5.20 ± 0.05


4.86 ± 0.04 6.12 ± 0.10 7.39 ± 0.04 N. A. 4.90 ± 0.06

The energy distribution of the difference between the detector response of a 239Pu fission chamber

covered by the Gd and by the Cd filters is shown in Figure 4-13. The values refer to fuel containing

only 238U and 16O. For all cases the detector response has a maximum around the 0.3 eV energy

region, but contributions to the total detector response were calculated in the thermal energy region

in the case of partial overlap of one of the two SINRD filters. Due to the incomplete cover the thermal

neutrons are not completely absorbed by the SINRD filter and are detected by the fission chamber.

To minimize this effect, an overlap of both SINRD filters is suggested to ensure conditions similar to

the reference case with an ideal cover of the detector.

49

Figure 4-13: Energy distribution of the difference between the detector response calculated for a


covered by Gd and Cd filters. The fuel in these simulations contained 238

U and 16

O.

The variations of the SINRD filter thickness were also considered in the study by simulating an

increase and decrease of 10% from the nominal value. Following the approach applied above, the

SINRD signature was calculated for the scenarios considered for fuel with 3.5% initial enrichment,

burnup of 60 GWd/tU, and 10 years of cooling time. Table 4-18 summarizes the results obtained in

the simulations, and contains the reference case and cases where the SINRD filters were increased or

decreased by 10%. The title of each column describes the filter combination. The SINRD signatures

calculated for the 10B proportional counter were identical to the values obtained for the 3He

proportional counter and therefore are not included in this chapter.

Table 4-18: SINRD signature calculated for changes in the nominal thickness value of the SINRD filters. The statistical uncertainty is reported in the table and is comparable to the variation of the SINRD signature for the cases considered in the study.

Detector Reference Gd: -10% Cd: -10%

Gd: -10% Cd: +10%

Gd: +10% Cd: -10%

Gd: +10% Cd: +10%


7.52 ± 0.03 7.52 ± 0.05 7.40 ± 0.04 7.63 ± 0.05 7.50 ± 0.05


5.17 ± 0.03 5.23 ± 0.04 5.11 ± 0.04 5.26 ± 0.05 5.11 ± 0.05


4.86 ± 0.04 4.98 ± 0.05 4.76 ± 0.05 5.06 ± 0.06 4.78 ± 0.05

For all detector types the variation of the SINRD signature due to the change from the nominal filter

thickness is comparable to the statistical uncertainty from the simulations. The maximum value of

the SINRD signature was achieved by increasing the thickness of the Gd filter and by decreasing the

0.01 0.1 1-1.0x10

-4

0.0

1.0x10-4

2.0x10-4

3.0x10-4

Detector: 239

Pu fission chamber

Filters: Gd 0.1 mm

Cd 1.0 mm

N x

de

tecto

r re

sp

on

se

(co

un

ts)

Neutron energy (eV)

Reference

Gd: -30°; Cd: -30°

Gd: -30°; Cd: +30°

Gd: +30°; Cd: -30°

Gd: +30°; Cd: +30°

50

thickness of the Cd filter. However, given the small impact on the SINRD signature, slight variations

from the nominal filter thickness can be tolerated for the foils used as SINRD filters.

4.6 Conclusions

The SINRD technique was investigated by using Monte Carlo simulations to define the optimal

measurement setup. The approach foresees the introduction of small neutron detectors in the

central guide tube of the PWR 17x17 fuel assembly.

The measurement of a fuel assembly immersed in fresh and borated water was modelled, and

compared to the case of fuel kept in air and surrounded by a 12-cm thick slab of polyethylene. The

results from the dry configuration showed the clearest reduction of the neutron flux due to the

absorption of 239Pu, therefore it was chosen as reference condition.

The SINRD signature was defined as the ratio between the neutron counts in the fast and in the

resonance energy region. A 238U fission chamber was chosen as reference detector for the estimation

of the fast neutron flux, whereas a 239Pu fission chamber covered by a foil of either Gd or Cd was

proposed for the resonance region.

The optimization of the SINRD filter thickness was also carried out, by identifying a combination that

has mainly contributions from neutrons with energy close to 0.3 eV and minimizing the contributions

from other energy regions. The selection of the SINRD filters was tailored for each detector type

considered in the study.

The detector response of a 239Pu fission chamber was compared with the results obtained with a 235U

fission chamber and with proportional counters containing 3He or 10B. The 239Pu fission chamber

showed the highest sensitivity to the 239Pu content in the fuel and this is the advantage of using the

self-indication technique. Similar results were calculated for the other detector types, with the

proportional counters that obtained the highest total neutron counts thanks to their size and high

neutron sensitivity. Taking these results into account, a combination of SINRD filters of 0.1 mm Gd

and 1.0 mm Cd was suggested for the measurements with fission chambers to maximize the total

neutron counts, whereas SINRD filters of 0.2 mm Gd and 1.0 mm Cd were proposed for the

proportional counters to maximize the SINRD signature.

The expected performance of SINRD in realistic scenarios was evaluated by considering a detailed

fuel composition and by assessing systematic effects on the SINRD signature. The masses of 239Pu and

235U were the parameters that influenced the most the technique, while a few other nuclides had an

impact in case of fuel with high burnup. The SINRD signature increased with the burnup due to the

51

239Pu content, and with the initial enrichment due to the 235U mass. Moreover, the SINRD signature

was largely independent from the cooling time of the fuel assembly, since the fissile content in the

fuel is almost independent on this parameter. From the results shown in this chapter it seems that by

combining the measurement of the SINRD signature with the RTH ratio the 239Pu mass in a fuel

assembly can be determined without the operator declaration.

The approach proposed in this study showed also no significant effect from the positioning of

detector in the guide tube, detector length, and small variations from the nominal filter thickness.

The incomplete detector cover by the filters caused a change in the SINRD signature values due to

the increase of the thermal neutron components passing through the bare section of the detector.

Therefore, a partial overlap of the SINRD filter is suggested to ensure conditions similar to an ideal

cover of the detector.

53

5 Benchmark of the self-indication neutron resonance densitometry

5.1 Objectives of the benchmark experiments

The study of the SINRD technique in the previous chapter was based on Monte Carlo simulations and

to validate the model an experimental benchmark was carried out at the GELINA Time-of-Flight (ToF)

facility of the Joint Research Centre (JRC) of Geel (Belgium).

The Time-of-Flight technique was chosen for the experimental validation of SINRD because with this

technique the energy distribution of a neutron beam can be measured. Time-of-Flight measurements

are traditionally used for neutron resonance spectroscopy and measure the time that a neutron

needs to travel a well-known distance (Harvey, 1970; Schillebeeckx, 2014). The measured time and

the flight distance are then related to the kinetic energy of the neutron.

Transmission measurements were first performed to verify the quality of nuclear data used in the

Monte Carlo simulations for the optimization of the SINRD filters thickness. In addition, self-

indication measurements were conducted to evaluate the advantage of using such technique, and

the response of different detector types was compared.

5.2 Description of the GELINA Time-of-Flight facility

The experiments for the benchmark of SINRD were conducted at the GELINA Time-of-Flight facility of

the JRC-IRMM. GELINA is a linear accelerator that is mainly used for high-resolution cross-section

measurements and for the development of measurement techniques for nuclear safeguards and

security (Mondelaers, 2006; Schillebeeckx, 2014). The aerial view of the facility is shown in Figure 5-

1.

The first section of the facility is the linear electron accelerator that delivers a pulsed white electron

beam. The injected electron pulses have a duration of 10 ns and a peak current of 12 A, whereas the

electrons leave the accelerator with energy varying linearly from 70 MeV to 140 MeV.

The electrons then travel along a 360° compression magnet, designed such that the electron pulse

will leave the target within a 1 ns time bin. Due to this compression the peak electron current rises to

about 120 A at the end of this section.

The high energy electrons leaving the compression section interact with the rotating neutron-

producing target, consisting of a thick disk of U-Mo alloy cooled with liquid mercury and sealed with

stainless steel. The electrons are rapidly decelerated in the target and high-energy photons are

produced via Bremmstrahlung reactions. Then the photons interact with the U atoms in the target to

54

produce neutrons via (,n) or (,f) reactions. Finally, to increase the contributions from thermal

neutrons, a light water moderator is placed around the target.

The neutrons produced in the previous section are then collimated in 12 flightpaths installed in a

star-like configuration around the neutron producing target. The flightpaths have a nominal length

between 10 and 400 m and allow multiple Time-of-Flight measurements at the same time.

During the measurements for the benchmark of SINRD, the accelerator was operated at a frequency

of 50 Hz to measure neutrons with energy lower than 0.3 eV, whereas the measurement stations at

10 m were chosen to obtain the maximum neutron flux.

Figure 5-1: Aerial view of the GELINA Time-of-Flight facility at the JRC-IRMM in Geel.

5.3 Overview of the experimental setup

5.3.1 Transmission measurements

Transmission measurements can be used for the determination of total neutron cross-section data

(Fröhner, 1966; Massimi, 2011; Schillebeeckx, 2014), and the schematic view of the measurement

setup is shown in Figure 5-2. The experiments were carried out to measure the transmitted neutron

flux through several samples of Gd and Cd as a function of the time-of-flight. In this way the quality

of the nuclear data for these nuclides and the optimal thickness of the SINRD filters obtained in

Chapter 4 were verified.

During the SINRD benchmark experiments an automatic sample changer was positioned at 7.7 m

from the neutron producing target, allowing an automated alternation of sample-in (tr) and sample-

out (0) measurements. A second sample changer was placed close to the sample position and is not

included in Figure 5-2, was used for anti-overlap and background filters. Anti-overlap filters are used

55

to absorb neutron produced during previous electron pulses of the accelerator, whereas background

filters are used to estimate the neutron background during the experiments according to the black

resonance technique (Schillebeeckx, 2012).

Figure 5-2: Schematic view of a transmission experiment.

Neutrons traversing the sample and the filters are finally detected by a 6.35 mm x 76 mm x 76 mm

Li-glass scintillator, placed at 11 m from the neutron producing target. The detector, which is

enriched to 95% in 6Li, is directly connected to a photomultiplier tube. A good transmission geometry

is realized by proper collimation of the neutron beam. A set of Li, B, Cu, Ni and Pb collimators with

decreasing diameter are placed between the neutron producing target and the sample; a similar

sequence of collimators with increasing diameter are placed between the sample and the detector.

Experiments in good transmission geometry are required to ensure that all detected neutrons

traverse the sample and scattered neutrons do not reach the detector (Schillebeeckx, 2012).

Table 5-1: Characteristics of the Gd and Cd samples used for the transmission measurements. All samples were in the form of a metal foil or disc.

Sample-ID Element Nominal thickness

(mm)

Weight (g)

Area (cm2)

Areal density (at/b)

TP-NP 07-32 Gd 0.030 1.0982 0.0001 50.3859 0.0007 (8.3468 0.0008) x 10-5

SN3S-2015-01-04 Gd 0.100 2.093 0.001 25.0455 0.0022 (3.2003 0.0016) x 10-4

Gd-disc Gd 0.200 8.242 0.001 50.2084 0.0003 (6.2865 0.0008) x 10-4

NS06001A Cd 0.500 8.6474 0.0001 19.6309 0.0005 (2.3598 0.0001) x 10-3

NS06001C Cd 1.000 17.1120 0.0001 19.6433 0.0017 (4.6668 0.0004) x 10-3

NS06001D Cd 1.000 17.1446 0.0001 19.6397 0.0030 (4.6765 0.0007) x 10-3

Transmission measurements using Gd and Cd metal foils or discs of different thicknesses were

carried out. The characteristics of the samples are summarized in Table 5-1. The beam size at the

sample position had a 10 mm diameter. To reduce the influence of the -ray flash in the detector a

permanent Pb filter was installed in the beam line.

The transmission is defined as the ratio between a sample-in (tr) and sample-out (0) measurements

and was calculated according to Formula (5.1). The experimental value of the transmission was

56

computed as the ratio between the ToF spectra obtained with the sample in the beam (Ctr) and the

unperturbed case (C0). Both contributions were corrected for the corresponding background

component, Btr and B0 respectively.

𝑇(𝐸) =𝜑𝑡𝑟(𝐸)

𝜑0(𝐸)=

𝐶𝑡𝑟(𝐸) − 𝐵𝑡𝑟(𝐸)

𝐶0(𝐸) − 𝐵0(𝐸) (5.1)

The ToF spectra were corrected for losses due to the dead time in the detector and electronics chain,

and all spectra were normalized to the same neutron intensity and ToF-bin width. The background,

which is a sum of ToF independent and dependent components, was determined by applying the

black resonance technique (Schillebeeckx, 2012). The resonance dips resulting from Co and Na filters

placed permanently in the beam were used to account for the impact of the presence of the sample

on the background. The average background (Btr, B0) below 1 eV was about 4% of the sample-in and

sample-out measurements (Ctr, C0).

5.3.2 Benchmark measurements

Validation experiments were carried out to demonstrate the basic principle of the SINRD techniques

and to compare the response of different detector types. The setup of the experiments is reported in

Figure 5-3.

Figure 5-3: Schematic representation of the self-indication experiments carried out at GELINA.

The incoming neutron flux (0) interacts with a sample under investigation and with the so-called

SINRD filter to obtain the transmitted flux (T) and the filtered flux (f), respectively. The sensitivity

to the sample is enhanced by combining measurements with different SINRD filters and a detector

sensitive to a specific resonance of a material of interest in the sample. Self-indication measurements

without the SINRD filter in the neutron beam were also carried out.

Due to safety regulations and limitations of available material, measurements on samples with

different amounts of 239Pu were not carried out, but natural Cd-samples with different thicknesses

were used to mimic the presence of 239Pu. The neutron cross section of 113Cd has a strong resonance

at 0.178 eV, which is relatively close in energy to the 0.296 eV resonance of 239Pu. The areal density

57

of the Cd sample that is placed in the beam can be derived from the attenuation of the neutron

beam due to the 0.178 eV resonance.

To construct a detector with a high sensitivity to the 0.178 eV energy region, a thin Cd sample

(0.027 mm thick) was surrounded by 4 C6D6 liquid scintillators detecting the prompt -rays emitted

after (n,) reaction. This detection system was placed at a 13.2 m distance from the neutron

producing target. Measurements with a Gd or Cd SINRD filter in the beam were performed to

measure neutrons with energy higher than the cut-off energy of each filter. By calculating the

difference between the two measurements, the neutron flux close to 0.178 eV was estimated. The

thickness of these filters was optimized for the detection of neutrons with energy close to 0.178 eV,

therefore slight changes are anticipated from the values proposed for 239Pu measurements. The

SINRD filters were placed far from the detectors to avoid background contribution of neutron

capture reactions in the filters. Measurements were performed using cadmium samples with

different thicknesses, and the samples were placed at about 6.5 m from the neutron producing

target. The characteristics of these samples are summarized in Table 5-2.

Table 5-2: Characteristics of the Gd and Cd samples used for the validation measurements. All samples were in the form of a metal disc of 80 mm diameter.

Sample-ID Element Nominal thickness

(mm)

Weight (g)

Area (cm2)

Areal density (x 10-4 at/b)

TP-NP 07-14 Cd 0.030 1.2814 0.0001 50.4334 0.0011 1.3611 0.0001

TP-NP 07-13 Cd 0.050 2.0993 0.0001 50.4699 0.0018 2.2283 0.0001

TP-NP 07-12 Cd 0.075 3.1837 0.0001 50.3927 0.0012 3.3845 0.0001

The results obtained with the self-indication geometry were compared with measurements with

Frisch-gridded ionization chambers with thin deposits of 235U or 10B. These detectors do not show an

enhanced efficiency at the energy of the resonance of interest, i.e. the 0.178 eV resonance. The

detectors were positioned at a 7.65 m distance from the neutron production target.

5.4 Results of the validation experiments

5.4.1 Transmission measurements

The results from the experimental transmission are shown in Figure 5-4 for the different Gd and Cd

filters. The plot includes also the values obtained with the analytical approach used in Chapter 4 for

the optimization of the SINRD filter thickness.

58

The analytical calculations do not include the response of the ToF-spectrometer; however, in the low

energy region the influence of this response is small compared to the broadening of the resonance

profiles due to the Doppler effect and the total width of the resonance.

Figure 5-4: Transmission through different Gd and Cd foils. The experimental transmission is compared with the analytical transmission based on the calculations in Chapter 4.

The spectra in Figure 5-4 illustrate that all Gd foils have a cut-off energy below the Cd resonance at

0.178 eV, whereas the cut-off energy for Cd is slightly above 0.3 eV. Evidently the cut-off energy

increases with increasing filter thickness. In order to reduce the contribution from thermal neutrons,

visible in Figure 5-4 in the case of 0.5 mm-thick Cd filter, a Cd filter with a minimum thickness of 1.0

mm is suggested. The results in Figure 5-4 show an overall good agreement between the

experimental and analytical transmissions.

From the results shown in Figure 5-4 a thickness of about 0.10 mm for the Gd filter is suggested to

obtain a cut-off energy slightly below the resonance of 239Pu at 0.296 eV. As mentioned above, a

thickness of 1.0 mm for the Cd filter is recommended to have a cut-off above the same resonance

energy and to reduce the contribution from neutrons below 0.1 eV. Therefore, the combinations of

SINRD filters identified in Chapter 4 to derive the amount of 239Pu are confirmed with these

experimental results.

Due to the lower energy of the Cd resonance compared to the one of 239Pu, the filter thickness of the

Gd filter has to be adapted for the benchmark measurement at GELINA. Following the criteria

described above, the optimum filter thicknesses are 0.03 mm for the Gd filter and 1.0 mm for the Cd

filter in case of self-indication measurements with Cd as material of interest.

0.01 0.1 110

-3

10-2

10-1

100

101

Experiment

Analytical

T

ran

sm

issio

n

Neutron energy (eV)





59

A series of calculations were performed using different nuclear data libraries and the analytical

transmissions are shown in Figure 5-5. The ENDF/B-VII.0 (Chadwick, 2006) and ENDF/B-VII.1

(Chadwick, 2011) data libraries were compared with JEFF-3.2 (Koning, 2010) and JENDL-4.0 (Shibata,

2011). The cross sections were obtained from the JANIS software (Soppera, 2014) by using a

logarithmic interpolation from 10-9 to 20 MeV and taking 100 points per decade. The transmission

was calculated with the analytical approach described in Chapter 4.

Figure 5-5: Transmission through different Gd and Cd foils using different nuclear data libraries. The values were calculated with the analytical approach described in Chapter 4. The plot on the right is focused on energy range below 0.1 eV.

Small differences were observed between the transmissions calculated for the data libraries,

especially for Cd filters at energy lower than 0.1 eV as shown in the right plot of Figure 5-5.

To quantify the impact of the data library, the detector response for a 239Pu fission chamber covered

by a SINRD filter was calculated with different data libraries. First, the neutron flux was calculated in

the central guide tube of a fuel assembly containing only 238U and 16O. Then the transmitted flux

through the SINRD filter covering the detector was calculated with the analytical approach described

in Section 4.4.1. Finally the detector response was obtained according to the approach described in

Section 3.4.1 as the energy-integrated product between the transmitted neutron flux in the guide

tube and the reaction cross-section of the active material in the detector. The values of the 239Pu

fission cross-section were taken from the JANIS software to keep the same energy groups considered

for the SINRD filters in Figure 5-5.

Table 5-3 shows the relative difference between the detector responses to the transmitted neutron

flux calculated for each SINRD filter with different data libraries and the corresponding value

obtained with the ENDF/B-VII.0. Similar results were obtained for the data libraries for the same

SINRD filter, and for all cases the differences were within 5% from the results with ENDF/B-VII.0. The

relative differences calculated for the Gd filters increase with the filter thickness and are in general

larger than those calculated for the Cd filters.

0.01 0.1 110

-3

10-2

10-1

100

101

ENDF/B-VII.0

ENDF/B-VII.1

JEFF-3.2

JENDL-4.0

Tra

nsm

issio

n

Neutron energy (eV)





0.01 0.110

-3

10-2

Tra

nsm

issio

n

Neutron energy (eV)

60

Overall, the relative differences observed in Table 5-3 might be important for nuclear safety

calculations, but they do not have an impact on the calculation of the optimum filter thickness. Based

on the results shown in the table, the Gd and Cd total cross sections included in the data libraries

considered in this section can be used for the study of the SINRD technique.

Table 5-3: Relative difference between the detector responses to the transmitted flux calculated with different data libraries and different SINRD filters. For each SINRD filter the results compared to the value obtained with ENDF/B-VII.0.

Gd 0.03 mm

Gd 0.10 mm

Gd 0.13 mm

Gd 0.20 mm

Cd 0.5 mm

Cd 1.0 mm

Cd 1.5 mm

Cd 2.0 mm

ENDF/B-VII.1 0.9% 2.7% 3.3% 4.4% 1.2% 0.4% 0.2% 0.1%

JEFF-3.2 0.6% 1.9% 2.2% 2.9% 0.9% 0.7% 0.7% 0.6%

JENDL-4.0 0.9% 2.2% 2.6% 3.4% 1.0% 0.6% 0.6% 0.6%

5.4.2 Benchmark measurements

Self-indication experiments using a neutron detector consisting of a 0.027 mm thin Cd sample

combined with 4 C6D6 detectors were carried out with and without Gd and Cd filters in the beam.

Measurements were performed with 0.03 mm, 0.05 mm, 0.075 mm and 0.105 mm Cd samples

placed in the neutron beam. The experimental setup is shown in Figure 5-6.

Figure 5-6: Experimental setup for the self-indication experiments. The 0.027 mm Cd sample is placed in the neutron beam and is surrounded by 4 C6D6 scintillator detectors.

The obtained responses without and with SINRD filters are shown in Figures 5-7 and 5-8,

respectively. All spectra were normalized to the same neutron intensity using the total counts of a

BF3 proportional counter which is installed in the concrete ceiling of the GELINA target hall. The

spectrum obtained without a Cd sample is also shown for comparison. Figure 5-7 shows that the

61

detection system chosen for the self-indication measurements enhances the detection efficiency in

the energy region around 0.178 eV. The spectra included in Figure 5-8 with the Gd filter also reveal

the reduction in detector response due to the presence of the Cd samples. In addition, the spectra

resulting from the measurements with the 1.0 mm thick Cd filter are practically insensitive to the

presence of the Cd samples. Since almost all neutrons with energy below 1 eV have been absorbed

by the filters, the spectra in this energy region reflect the background contribution. The background

can be expressed as the sum of time-independent and time-dependent contributions, as discussed in

detail in (Schillebeeckx, 2012). The total background (B) below 5 eV was approximated as a function

of the neutron energy (E) with Formula (5.2).

𝐵(𝐸) = 𝑎 + 𝑏 (𝐸

1 𝑒𝑉)

𝑐

(5.2)

The free parameters in the analytical expression were determined by a fit of the experimental data in

the energy regions between 0.0022 eV and 0.004 eV, 0.008 eV and 0.009 eV, 1 eV and 1.3 eV, and

between 3 eV and 4 eV. The average background spectra calculated with such an adjustment is also

reported in Figures 5-7 and 5-8.

Figure 5-7: Spectra of the self-indication experiments with Cd samples in the beam. The spectrum obtained with the detector only is reported for comparison together with the background contribution. All spectra were normalized to the same beam intensity.

0.01 0.1 1 100.00

0.05

0.10

0.15

0.20

0.25

Co

un

ts (

1/n

s)

Neutron energy (eV)

Detector only

Cd 0.03 mm

Cd 0.05 mm

Cd 0.075 mm

Cd 0.105 mm

Aver. background

62

Figure 5-8: Spectra of the self-indication experiments with a 0.03 mm Gd (left) and 1.0 mm Cd (right) filter in the beam. The spectrum obtained with the detector only is reported for comparison together with the background contribution.

The experimental observable RSI,1 was defined from the results in Figure 5-7 to quantify the areal

density of the Cd sample that is placed in the neutron beam. The RSI,1 was calculated according to

Formula (5.3) by integrating the ToF spectrum C(E) corrected for its background contribution B(E)

between 0.08 eV and 0.4 eV. The SINRD signature defined in Formula (4.1) in Chapter 4 foresees also

the measurement of fast neutrons to avoid the dependence on the neutron source term of the

measured fuel assembly. However, since all spectra of the benchmark measurements are normalized

to the total neutron beam intensity, the measurement of fast neutrons is not required and it is not

included in the Formula (5.3). The RSI,1 observable can be considered an ideal SINRD signature

because only neutrons in the energy region close to the Cd resonance at 0.178 eV are considered.

𝑅𝑆𝐼,1 =1

∫ 𝐶(𝐸) − 𝐵(𝐸) 𝑑𝐸0.4 𝑒𝑉

0.08 𝑒𝑉

(5.3)

A second observable was defined for the measurements with the SINRD filters, by taking the

difference between the total counts resulting from the measurements with the Gd and Cd filters in

the beam. The RSI,2 signature is defined in Formula (5.4), and CGd and CCd are the total counts of the

spectra taken with the Gd and Cd filters, respectively. The neutron background calculated with

Formula (5.2) was subtracted for both spectra.

𝑅𝑆𝐼,2 =1

𝐶𝐺𝑑 − 𝐶𝐶𝑑 (5.4)

0.01 0.1 1 100.00

0.05

0.10

0.15

0.20

0.25

Co

un

ts (

1/n

s)

Neutron energy (eV)

Detector only

Gd 0.03 mm [1]

[1] + Cd 0.03 mm

[1] + Cd 0.05 mm

[1] + Cd 0.075 mm

[1] + Cd 0.105 mm

Average background

0.01 0.1 1 100.00

0.05

0.10

0.15

0.20

0.25

Co

un

ts (

1/n

s)

Neutron energy (eV)

Detector only

Cd 1.0 mm [1]

[1] + Cd 0.03 mm

[1] + Cd 0.05 mm

[1] + Cd 0.075 mm

[1] + Cd 0.105 mm

Average background

63

Figure 5-9: Experimental observables RSI,1 and RSI,2 as a function of the areal density of the Cd sample placed in the beam. The results were normalized to the measurements without Cd sample.

The two observables defined in this section are compared in Figure 5-9 as a function of the areal

density of the Cd sample, and the results are normalized to the value obtained without a Cd sample

in the beam. This figure reveals that the sensitivity of RSI,2 to the Cd areal density is very close to the

one of RSI,1. The latter results from an almost ideal measurement, using an optimised filter and a

dedicated procedure to account for the background contribution. The agreement between the two

observables in Figure 5-9 demonstrates the effectiveness of the SINRD filters.

The results obtained with the self-indication detector were compared with the ones obtained using a

235U fission chamber and a 10B ionisation chamber as neutron detectors. Figure 5-10 and 5-11 show

the spectra normalised to the total neutron intensity taken with the 235U and 10B chambers,

respectively. The spectra measured with the Gd and Cd filters are given separately, and the results

obtained with the bare 235U fission chamber are also shown. Measurements with a bare 10B ionization

chamber were not performed. To optimise the available beam time, for both detectors the

measurements with the Cd filter were limited to a measurement with only the 1.0 mm thick filter in

the beam without any additional Cd sample. However, considering the results in Figure 5-8, negligible

effects were expected from measurements with additional Cd samples.

The spectra in Figures 5-10 and 5-11 illustrate that the 235U and 10B detectors do not show an

enhanced efficiency close to the 0.178 eV resonance. The peak below 0.1 eV is due to the energy

distribution of the neutron spectrum available at GELINA. The impact of the 235U resonances above

0.1 eV is clearly visible in Figure 5-10. In particular, the 0.271 eV resonance of 235U slightly enhances

the efficiency for neutrons close to the 0.178 eV resonance of 113Cd.

0 1 2 3 4 50

1

2

3

4

5

6

RS

I

Cd sample areal density (10-4 at/b)

RSI,1

RSI,2

64

Figure 5-10: Spectra obtained for the

235U fission chamber with a 0.03 mm Gd (left) and 1.0 mm Cd (right) filter in the beam.

Moreover, several Cd samples were used with the Gd filter to simulate the neutron absorption by fuel pins containing 239

Pu.

Figure 5-11: Spectra obtained for the

10B ionization chamber with a 0.03 mm Gd (left) and 1.0 mm Cd (right) filters in the

beam. Moreover, several Cd samples were used with the Gd filter to simulate the neutron absorption by fuel pins containing

239Pu.

The observable defined in Formula (5.4) was calculated also for the 235U and 10B detectors as a

function of the areal density of the Cd sample inserted in the neutron beam. It was supposed that the

neutron background was constant during the measurements of the spectra in Figures 5-10 and 5-11,

therefore it cancels out in the calculation of the observable with Formula (5.4). The results are

compared in Figure 5-12 with the observable calculated from the measurements with the self-

indication detector.

The highest sensitivity to the Cd areal density is obtained with the self-indication detector and this

confirms the advantage of using the self-indication technique. The results calculated for the 235U

fission chamber are more sensitive to the Cd areal density compared to the values obtained for the

10B ionization chamber. This is due to the resonance of the 235U total cross-section at 0.271 eV, which

is close to the resonance energy of Cd.

0.01 0.1 10.0

0.1

0.2

0.3

0.4

0.5

Co

un

ts (

1/n

s)

Neutron energy (eV)

Detector only

Gd 0.03 mm [1]

[1] + Cd 0.03 mm

[1] + Cd 0.05 mm

[1] + Cd 0.075 mm

[1] + Cd 0.105 mm

0.01 0.1 10.0

0.1

0.2

0.3

0.4

0.5

De

tecto

r re

sp

on

se

(1

/ns)

Neutron energy (eV)

Detector only

Cd 1.0 mm

0.01 0.1 10.00

0.04

0.08

0.12

0.16

0.20

Co

un

ts (

1/n

s)

Neutron energy (eV)

Gd 0.03 mm [1]

[1] + Cd 0.03 mm

[1] + Cd 0.05 mm

[1] + Cd 0.075 mm

[1] + Cd 0.105 mm

0.01 0.1 10.00

0.04

0.08

0.12

0.16

0.20

D

ete

cto

r re

sp

on

se

(1

/ns)

Neutron energy (eV)

Cd 1.0 mm

65

The results in Figure 5-12 demonstrate that the optimal results are achieved by using a detector with

enhanced detection efficiency close to a resonance of the material of interest. Hence, to determine

the amount of 239Pu in a spent fuel assembly a 239Pu based fission chamber is recommended. This

result is in line with the conclusion of Chapter 4, where the SINRD signature calculated with the 239Pu

fission chamber showed the highest sensitivity to the 239Pu content in the fuel compared to the other

detectors.

Figure 5-12: Experimental observable RSI,2 as a function of the areal density of the Cd sample placed in the beam. The data refer to measurements with a self-indication detector, a

235U fission chamber, and a

10B ionization chamber.

5.5 Conclusions

The results of the SINRD benchmark experiments performed at the GELINA facility of JRC-IRMM in

Geel were summarized in this chapter.

Transmission measurements were performed to verify the quality of nuclear data used in the

previous Monte Carlo simulation for the optimization of the SINRD filters. The comparison showed

some differences between the experimental transmission and the values calculated with the

analytical approach. However, the quality of nuclear data is sufficient to define the optimal thickness

of the Gd and Cd filters. The results of the experiments indicated that the combination of a Gd filter

of about 0.1 mm and a 1.0 mm Cd filter is suggested for the measurement of spent fuel containing

239Pu.

In addition, self-indication experiments were carried out and the results obtained with the SINRD

filters are very similar to the values calculated with the ideal measurement using ideal filters and a

dedicated procedure for the background subtraction. The results from the self-indication

0 1 2 3 4 50

1

2

3

4

5

6

RS

I,2

Cd sample areal density (10-4 at/b)

Self-indication detector235

U fission chamber10

B ionization chamber

66

measurements were compared with the values obtained for a 235U fission chamber and a 10B

ionization chamber, and the results confirmed that the highest sensitivity is obtained using a neutron

detector with an enhanced efficiency for a resonance of the material of interest. Therefore, a

neutron detector based on a 239Pu convertor is recommended for the characterisation of spent fuel

by SINRD.

67

6 Monte Carlo assessment of the partial defect tester

6.1 Structure of the study

The Monte Carlo simulations performed to study the partial detect tester (PDET) are described in this

chapter. As described in Section 3.2.3 the model contained a storage rack with nine fuel assemblies

in a 3x3 configuration and the neutron and gamma-ray fluxes were estimated in the guide tubes of

the central fuel assembly.

The first section of this study considered the reference conditions for the PDET detector, with the

storage rack containing fuel assemblies with the same isotopic composition and source strength. The

contribution of the individual fuel pins on the detector responses was calculated, and the normalized

detector responses of several neutron and gamma-ray detector types were compared. In addition,

the influence of the spent fuel irradiation history on the normalized detector responses was

estimated.

The impact on the normalized detector responses of the different fuel assemblies in the storage rack

was then evaluated in the second section. The fuel composition and source strength of the central

fuel assembly was first varied to represent fuel with different burnup values. Both low-burnup as well

as high-burnup assemblies were considered in the rest of the storage rack. Finally, the influence of

lateral and corner fuel assemblies was evaluated with another set of simulations by considering a

reference storage rack with low-burnup fuel assemblies and placing high-burnup fuel assemblies in

different storage rack positions.

6.2 Reference conditions for the PDET detector

6.2.1 Contribution of single fuel pins

The PDET detector foresees the insertion of tiny neutron and gamma-ray detectors in the guide tubes

of a PWR fuel assembly, therefore the contribution of each fuel pin to the detector response was

calculated for different detector types. This value is also called the importance function of the

individual fuel pins. The SCX card (Pelowitz, 2011) was used for the MCNPX simulations to calculate

the contribution of each fuel pin in the central fuel assembly to the total flux tally value computed for

all fuel pins in the assembly. The detector types considered in this section were a 235U fission

chamber, a 238U fission chamber, and an ionization chamber with nitrogen as filling gas at 1 atm.

These detector types were chosen because they are mostly sensitive to thermal neutrons, fast

neutrons, and gamma-rays, respectively. The detector response was calculated with the approach

described in Section 3.4. The fuel assemblies included in the model had uniform material

composition corresponding to fuel with 3.5% initial enrichment, 30 GWd/tU of burnup, and 5 years of

68

cooling time. The values of initial enrichment and burnup represent average values in the reference

spent fuel library (Rossa, 2013b), and the cooling time allows for the decay of short lived nuclides.

Figure 6-1 shows the contribution of the individual fuel pins to the detector response of a 235U fission

chamber inserted in several guide tubes. For all results of the PDET detector the guide tubes were

numbered sequentially from the top left to the bottom right corner as shown in Figure 3-1. Given the

symmetry of the assembly, the values of only six guide tubes are reported in the figures. The neutron

flux was calculated for the guide tube colored in gray in the figures, whereas the other guide tubes

are depicted with a cross. In all cases shown in Figure 6-1 the fuel pins close to the detector position

have the largest values of the importance function, and the values decrease with the distance from

the guide tube.

A similar trend was obtained also for the importance function calculated for a 238U fission chamber,

and the results are shown in Figure 6-2. The values in this case are more localized than in Figure 6-1,

but all fuel pins give still a significant contribution to the total.

Figure 6-1: Importance function for a

235U fission chamber placed in different guide tubes. The neutron flux was calculated

in the guide tube depicted in grey. The color bar ranges between 0 and 1%.

69

Figure 6-2: Importance function for a

238U fission chamber placed in different guide tubes. The neutron flux was calculated


Figure 6-3: Importance function for an ionization chamber placed in different guide tubes. The neutron flux was calculated


The same approach was also used to calculate the importance function in the case of an ionization

chamber and the results are shown in Figure 6-3. The importance function is strongly localized

around the detector position, and several fuel pins give negligible contributions when the detector is

placed in the guide tubes at the periphery of the assembly. The significant shielding effect observed

for the gamma-rays is in line with physics principles and with previous work (Sitaraman, 2009).

70

To further analyze the importance function of the fuel pins on the detector responses calculated in

the different guide tubes, the fuel assembly cross-section was divided into three areas that are

shown in Figure 6-4. The areas are concentric from the considered guide tube containing the

detector. The fuel pins in the assembly are grouped depending on their distance from the guide tube,

and for each area the sum of the importance function of the fuel pins was calculated. This approach

takes into account the fact that the importance function of the single fuel pins is higher for the fuel

pins closer to the detector position, but also that many pins are in the area far from the detector

position. The number of fuel pins included in each area is reported in Table 6-1.

Figure 6-4: Areas identified for the calculation of the integral contribution to the detector response in the different guide

tubes.

Table 6-1: Number of fuel pins included in each area identified in Figure 6-4 for the different guide tubes. The fuel assembly

contains a total of 264 fuel pins.

Area A Area B Area C

Guide tube 1 36 60 168





Guide tube 13 40 104 120

The sum of the importance function corresponding to each area identified in Figure 6-4 is reported in

Tables 6-2 – 6-4 for different guide tubes in case of a 235U fission chamber, a 238U fission chamber,

and an ionization chamber, respectively. The statistical uncertainty of the values in Tables 6-2 – 6.4 is

lower than 0.3%.

71

Focusing on the results of the neutron detectors, the fuel pins in the area close to the detector

position (Area A) contribute for less than 50% of the total importance function calculated with all fuel

pins, and similar values were obtained for all guide tubes. In general the values are higher in the case

of a 238U fission chamber compared to a 235U fission chamber, and this confirms that the response

function is more localized in case of a 238U fission chamber. For both detector types the values

calculated for guide tube 4 show an increase compared to the other guide tubes because the number

of fuel pins in Area A for this position is larger compared to the other locations as shown in Table 6-1.

The fuel pins in the outer region of the fuel assembly (Area C) contribute for more than 25% to the

total integral value for both neutron detectors, and for the 235U fission chamber give the main

contribution to the guide tubes at the periphery of the assembly (i.e. guide tubes 1, 2, and 4).

The results calculated for the ionization chamber confirm the high localization of the importance

function, since for all guide tubes the fuel pins close to the detector position (Area A) account for

more than 80% of the total value. The self-shielding effect for gamma-rays is more pronounced than

for neutrons and therefore the pins far from the detector (Area C) give a cumulative contribution

around 2% of the total value.

Table 6-2: Percentage contribution to the importance function from different sections of the fuel assembly. Values for the 235

U fission chamber.


Guide tube 1 24.7 31.2 44.1

Guide tube 2 24.0 32.8 43.2

Guide tube 4 30.2 24.8 45.0

Guide tube 7 24.9 40.0 35.1

Guide tube 8 24.4 41.9 33.7

Guide tube 13 23.9 43.7 32.4

Table 6-3: Percentage contribution to the importance function from different sections of the fuel assembly. Values for the 238

U fission chamber.


Guide tube 1 38.9 27.0 34.1

Guide tube 2 37.7 28.4 33.9

Guide tube 4 44.5 20.9 34.6

Guide tube 7 37.0 35.2 27.8

Guide tube 8 36.1 36.8 27.1

Guide tube 13 35.0 38.5 26.5

72

Table 6-4: Percentage contribution to the importance function from different sections of the fuel assembly. Values for the

ionization chamber.


Guide tube 1 87.1 11.1 1.8

Guide tube 2 87.3 10.6 2.1

Guide tube 4 89.6 8.6 1.8

Guide tube 7 84.2 14.1 1.7

Guide tube 8 83.7 14.1 2.2

Guide tube 13 82.8 14.6 2.6

6.2.2 Comparison among several detector types

Taking into account the information on the current PDET prototype (Ham, 2015), the reference

neutron detector for PDET was a 235U fission chamber whereas the gamma-ray detector was an

ionization chamber with nitrogen as filling gas at 1 atm. The responses of several detectors were

compared in this section to evaluate the impact of the detector type on the results.

Following the SINRD study, a 239Pu fission chamber and a 3He proportional counter were considered

as alternative thermal neutron detectors, and a 238U fission chamber as fast neutron detector. In

addition, the detector responses of ionization chambers with nitrogen or xenon as filling gas were

calculated in case of filling gas pressures of 1 atm and 10 atm.

The neutron detector responses are shown in Figure 6-5 for the 235U fission chambers and for the

238U fission chambers. The results for the 239Pu fission chambers and for the proportional counters

were within the statistical uncertainty of the values for the 235U fission chambers, therefore they

were not included in the plot. The values in Figure 6-5 refer to fuel with 3.5% initial enrichment,

burnup of 30 GWd/tU, and cooling time of 5 years. The detector responses in the guide tubes were

normalized to the maximum value obtained in the corresponding simulation, and the statistical

uncertainty was lower than 0.2%. For both detector types the largest detector response was

obtained in the central guide tube, and a decreasing trend was observed towards the periphery of

the assembly.

73

Figure 6-5: Neutron detector response of

235U fission chambers and

238U fission chambers. The results for each plot were

normalized to the maximum value obtained in the correponding simulation, and the uncertainty of the values was lower

than 0.2%

Figure 6-6: Normalized detector responses for different guide tubes. The

235U and

238U fission chambers were compared

and the statistical uncertainty of the simulations was also included. The results for the other guide tubes were not included

due to the symmetry of the fuel assembly.

The normalized detector responses for the 235U and 238U fission chambers are also shown in Figure 6-

6. Only 6 guide tubes were reported in the plot due to the symmetry of the fuel assembly. For each

guide tube the detector responses of the 238U fission chambers were slightly higher than the results

of the 235U fission chambers. This effect is visible for guide tube 2 that has a normalized detector

response 3% higher in the case of the 238U fission chamber.

The gamma-ray detector types were compared and the normalized detector responses for the

ionization chambers with nitrogen as filling gas at 1 atm are shown in Figure 6-7. The statistical

1 2 4 7 8 130.75

0.80

0.85

0.90

0.95

1.00

Peripheral guide tubes Central guide tubes

De

tecto

r re

sp

on

se

Guide tube

235

U fission chamber

238

U fission chamber

74

uncertainty of the values in the plot is lower than 0.8%. As for the neutron detectors, the largest

detector response was obtained in the central guide tube, and a decreasing trend of the normalized

detector responses was observed moving towards the external section of the fuel assembly. The

gamma-ray detector responses were more uniform over the fuel assembly cross section than in the

case of neutron detectors. Due to the self-shielding effect of the fuel pins, the detector response

calculated in a guide tube is mostly due to the fuel pins close to the detector position and therefore it

is marginally influenced by the geometrical position of the guide tube within the fuel assembly.

Figure 6-7: Gamma-ray detector response of an ionization chamber with nitrogen as filling gas at 1 atm. The results were

normalized to the maximum value obtained in the guide tubes, and the uncertainty of the values was lower than 0.8%

The normalized detector responses for the ionization chambers with nitrogen as filling gas at 1 atm

are shown in Figure 6-8, and for each guide tube position the results for the other gamma-ray

detector types were within the statistical uncertainty of the simulations.

For the rest of the study the detector responses of 235U fission chambers and 238U fission chambers

were calculated to consider detectors sensitive to thermal and fast neutrons, respectively. In

addition, ionization chambers with nitrogen as filling gas at 1 atm were chosen as gamma-ray

detectors because of their higher gamma-dose rate range compared to the other detector types

(Photonis, 2016).

75

Figure 6-8: Normalized detector responses for different guide tubes. Only the response of ionization chamber with nitrogen

as filling gas at 1 atm was reported because the responses of the other detector types were within the statistical

uncertainty. The results for the other guide tubes were not included due to the symmetry of the fuel assembly.

6.2.3 Influence of spent fuel irradiation history

The results from the PDET detector are expressed in terms of the normalized neutron and gamma-

ray detector response across the fuel assembly cross section to avoid the dependence on the

corresponding source term. In this section the normalized detector responses obtained for fuel

assemblies with different irradiation histories are compared by considering fuel with different initial

enrichment (IE), burnup (BU), and cooling time (CT).

For this analysis the relative difference (DK,R) between the normalized detector responses in

simulation K (NK) and in the reference case (NR) was defined in Formula (6.1). The detector responses

were normalized as in the previous section by taking the maximum of the detector response

calculated in the guide tubes in the corresponding simulation. The fuel composition in the reference

case refers to fuel with 3.5% initial enrichment, burnup of 30 GWd/tU, and cooling time of 5 years.

Fuel with 2.5% and 4.5% initial enrichment, and cooling time with 30 days and 50 years were also

considered in this section. The fuel composition was adapted in each simulation by varying one

parameter (i.e. IE, BU, CT) from the reference case.

𝐷𝐾,𝑅 =𝑁𝐾 − 𝑁𝑅

𝑁𝑅 (6.1)

The DK,R ratio calculated for the 235U fission chambers showed low sensitivity to the fuel initial

enrichment and cooling time since the values of the relative difference were within ±1%. The

increase of the burnup leads to an increase of the difference with the reference case, with the DK,R

1 2 4 7 8 130.75

0.80

0.85

0.90

0.95

1.00


De

tecto

r re

sp

on

se

Guide tube

N ionization chamber (1 atm)

76

ratio within ±2% for the different guide tubes. The normalized detector responses for fuel with

different burnup are shown in Figure 6-9, and the differences are visible mostly for the peripheral

guide tubes.

Figure 6-9: Normalized detector responses for fuel with different burnup (BU). The results refer to

235U fission chambers

and the values for the other guide tubes were not included due to the symmetry of the fuel assembly.

The DK,R ratio calculated for 238U fission chambers showed that the sensitivity of this detector type to

the fuel composition was slightly larger than in the case of the 235U fission chambers. The values were

mostly within ±1% by considering fuel with different initial enrichment and cooling time, and within

±3% in the case of fuel with different burnup. Figure 6-10 shows the normalized detector responses

for fuel with different burnup values and, as for the 235U fission chambers, the peripheral guide tubes

are the most affected by the different fuel composition.

The ratio in Formula (6.1) was calculated also for the gamma-ray detector response, but most of the

values of the DK,R ratio were within the statistical uncertainty. Only for fuel with cooling time of 30

days the DK,R ratio for the guide tubes at the periphery of the assembly was about -2%. The

normalized detector responses for fuel with different cooling times are included in Figure 6-11. The

peripheral guide tubes show the largest difference with the variation of cooling time.

Given the results in this section, fuel assemblies with different burnup were considered in the rest of

the study, whereas only fuel with 3.5% initial enrichment and 5 years of cooling time was taken into

account.

1 2 4 7 8 130.75

0.80

0.85

0.90

0.95

1.00


D

ete

cto

r re

sp

on

se

Guide tube

Reference

BU: 10 GWd/tU

BU: 60 GWd/tU

77

Figure 6-10: Normalized detector responses for fuel with different burnup (BU). The results refer to

238U fission chambers

and the values for the other guide tubes are not included due to the symmetry of the fuel assembly.

Figure 6-11: Normalized detector responses for fuel with different cooling time (CT). The results refer to ionization

chambers and the values for the other guide tubes are not included due to the symmetry of the fuel assembly.

1 2 4 7 8 130.75

0.80

0.85

0.90

0.95

1.00


De

tecto

r re

sp

on

se

Guide tube

Reference

BU: 10 GWd/tU

BU: 60 GWd/tU

1 2 4 7 8 130.75

0.80

0.85

0.90

0.95

1.00


De

tecto

r re

sp

on

se

Guide tube

Reference

CT: 30 days

CT: 50 years

78

6.3 Influence of the spent fuel assemblies in the storage rack

6.3.1 Impact of the central fuel assembly burnup

6.3.1.1 Fuel assembly surrounded by low burnup fuel

The results obtained in the previous sections refer to a storage rack with identical fuel assemblies,

whereas this section evaluates the impact of fuel assemblies with different burnup on the normalized

neutron and gamma-ray detector responses. As shown in Figures 6-9 and 6-10, the fuel burnup is the

parameter of the irradiation history that influences the most the detector responses in the guide

tubes of the fuel assembly being measured.

For the analysis in this section the guide tubes in the fuel assembly were divided into two groups as

shown in Figure 6-12. The average value of the normalized detector response was calculated for the

16 guide tubes at the periphery of the assembly and for the 9 guide tubes in the central section.

Figure 6-12: Position of the guide tubes in the fuel assembly.

The reference case in this section was a uniform storage rack containing fuel assemblies with burnup

of 10 GWd/tU. Several simulations were performed increasing the burnup of the central fuel

assembly in the storage rack up to 60 GWd/tU. Both the fuel assembly material composition and

source emission term were adjusted in each case.

The average values of the detector responses in the nine central guide tubes are shown in Table 6-5

for different detector types. The standard deviation among the guide tubes, as well as the maximum

difference amongst them, is reported. The maximum difference is expressed as percentage of the

average detector response. The statistical uncertainty was similar for central and peripheral guide

79

tubes and was around 0.1% for the neutron detectors and around 0.4% for the ionization chambers.

For all detector types the variation of the average values as a function of burnup was within the

standard deviation of the reference case. Similar average values were obtained for the neutron

detectors, and the maximum difference between the guide tubes was about 10%. The average values

for the ionization chambers were larger than for the neutron detectors and the maximum difference

among the guide tubes was about 2%.

Table 6-5: Average normalized detector responses and standard deviations for the nine central guide tubes. The maximum

difference calculated among the guide tubes is also reported. The statistical uncertainty was around 0.1% for the neutron

detectors and around 0.4% for the ionization chambers.

Central

fuel ass.

burnup

235U fission chambers 238U fission chambers Ionization chambers

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.945 ± 0.028 9% 0.946 ± 0.028 9% 0.992 ± 0.005 1%

20 GWd/tU 0.928 ± 0.036 11% 0.936 ± 0.033 10% 0.991 ± 0.006 1%

30 GWd/tU 0.928 ± 0.036 11% 0.938 ± 0.033 10% 0.989 ± 0.006 2%

40 GWd/tU 0.930 ± 0.036 11% 0.940 ± 0.032 10% 0.988 ± 0.007 2%

50 GWd/tU 0.930 ± 0.035 11% 0.941 ± 0.031 9% 0.988 ± 0.007 2%

60 GWd/tU 0.931 ± 0.035 11% 0.943 ± 0.030 9% 0.988 ± 0.007 2%

The average relative difference DK,R as defined in Formula (6.1) was calculated for the nine central

guide tubes, and the results are shown in Figure 6-13 for all detector types. The values were within -

3% from the reference case and no trend was observed with the burnup of the central fuel assembly.

The statistical uncertainty of the results was within 0.3%.

Figure 6-13: Average relative difference as defined in Formula (6.1) between the detector responses calculated in the nine

central guide tubes.

0 10 20 30 40 50 60

-20

-10

0

10

20

Central guide tubes (10)

DK

,R

(%

)




Ionization chamber

80

The average values of the detector responses in the peripheral guide tubes are shown in Table 6-6

with the standard deviation and the maximum difference obtained among the guide tubes. For both

neutron detectors the average values decrease by increasing the central fuel burnup from 10 to 20

GWd/tU, but the values for burnup larger than 20 GWd/tU were within the standard deviation. The

maximum difference among the guide tubes was between 5% and 7% for all simulations.

The variations with the fuel burnup of the average values for the ionization chambers were smaller

than for the neutron detectors and were almost within the standard deviation among the guide

tubes. As for the nine central guide tubes, also for the peripheral guide tubes the average values for

the detector responses of the ionization chambers were larger than the values for the neutron

detectors. The maximum difference among the guide tubes was around 2%.

Table 6-6: Average normalized detector responses and standard deviation for the peripheral guide tubes. The maximum


detectors and around 0.4% for the ionization chambers.

Central

fuel ass.

burnup


Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.790 ± 0.013 5% 0.801 ± 0.018 6% 0.938 ± 0.005 2%

20 GWd/tU 0.743 ± 0.018 6% 0.766 ± 0.020 7% 0.926 ± 0.005 1%

30 GWd/tU 0.738 ± 0.018 7% 0.768 ± 0.020 7% 0.921 ± 0.005 1%

40 GWd/tU 0.741 ± 0.018 6% 0.774 ± 0.019 7% 0.917 ± 0.005 2%

50 GWd/tU 0.742 ± 0.018 6% 0.780 ± 0.019 6% 0.916 ± 0.005 1%

60 GWd/tU 0.744 ± 0.018 6% 0.785 ± 0.018 6% 0.916 ± 0.005 1%

Figure 6-14 shows the average relative difference with the reference case for the normalized

detector responses of the peripheral guide tubes. All detector types show negative values for the

relative difference. The results for the ionization chambers were within -3%, whereas the relative

differences were around -6% for the 235U fission chambers, and between -3% and -5% for the 238U

fission chambers. The statistical uncertainty of the results was within 0.2%.

The percentage difference for the ionization chambers has a decreasing trend with the burnup of the

central fuel assembly. By increasing the burnup of the central fuel assembly, most of the source

particles are generated in the central fuel assembly itself and the normalized detector responses in

the guide tubes at the periphery are lower than in the reference case.

The values calculated for the 235U fission chambers do not vary for fuel with burnup larger than

20 GWd/tU, whereas the absolute value of the average difference for the 238U fission chambers has a

81

maximum for fuel with burnup of 20 GWd/tU and then decreases with the fuel burnup. Given the

difference in the results between the gamma-ray and neutron detectors, another set of simulations

was performed to explain the trend shown for the neutron detectors.

Figure 6-14: Average relative difference as defined in Formula (6.1) between the detector responses calculated in the

sixteen guide tubes at the periphery of the fuel assembly.

The same storage rack configurations were used for the second set of simulations, but the NONU

card was used to neglect the production of secondary neutrons from fissions in the fuel assemblies

(Pelowitz, 2011). Therefore, the impact of neutrons emitted from neutron-induced fissions was

evaluated by comparing the DK,R ratio shown in Figure 6-14.

Table 6-7 shows for the simulations with the NONU card the average normalized detector responses

calculated in the peripheral guide tubes and the corresponding standard deviations. The maximum

difference among the guide tubes is also included.

The average value calculated for fuel with burnup of 10 GWd/tU with the 235U fission chambers was

larger than the corresponding case in Table 6-6, and the maximum difference among the guide tubes

was smaller. The average values shown in Tables 6-6 and 6-7 for fuel with higher burnup were within

the standard deviation, and the maximum differences among the guide tubes were comparable.

The average normalized detector response for the 238U fission chambers in Table 6-7 was always

larger than the corresponding values in Table 6-6. In addition the standard deviation among the

guide tubes and the maximum difference were smaller than in Table 6-6. Overall, the detector

responses for both neutron detectors were more uniform in the peripheral guide tubes by neglecting

the secondary neutrons.

0 10 20 30 40 50 60

-20

-10

0

10

20

Peripheral guide tubes (10)

DK

,R

(%

)




Ionization chamber

82


difference calculated among the guide tubes is also reported. The statistical uncertainty was around 0.3% for fuel with

burnup of 10 GWd/tU and around 0.1% for all other cases.

Central

fuel ass.

burnup

235U fission chambers 238U fission chambers

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.842 ± 0.008 3% 0.904 ± 0.009 3%

20 GWd/tU 0.753 ± 0.017 6% 0.840 ± 0.012 4%

30 GWd/tU 0.748 ± 0.017 6% 0.836 ± 0.013 4%

40 GWd/tU 0.747 ± 0.017 6% 0.835 ± 0.013 4%

50 GWd/tU 0.746 ± 0.017 6% 0.835 ± 0.013 4%

60 GWd/tU 0.746 ± 0.017 6% 0.835 ± 0.013 4%

The average relative difference (DK,R) for the normalized detector responses at the peripheral guide

tubes was calculated for the simulations with the NONU card, and the results are shown in Figure 6-

15. The detector responses obtained in the previous simulations for fuel with burnup of 10 GWd/tU

were used as reference case for the calculation of the relative difference. Constant values were

obtained for both neutron detector types in case of fuel with burnup larger than 20 GWd/tU. The

average difference calculated for the 235U fission chambers was around -5% whereas for the 238U

fission chambers was around +5%. The statistical uncertainty of the results was within 0.1%.

The results in Figure 6-15 do not show a trend for fuel with burnup larger than 20 GWd/tU, therefore

the small trend shown in Figure 6-14 for the 238U fission chambers is due to the secondary neutrons

emitted from fissions in the fuel assemblies, and their effect increases with the fuel burnup.



0 10 20 30 40 50 60

-20

-10

0

10

20

Peripheral guide tubes (10) - NONU

DK

,R

(%

)




83

6.3.1.2 Fuel assembly surrounded by high burnup fuel

The impact of the central fuel assembly burnup was also evaluated taking as reference case a

uniform storage rack composed by fuel with burnup of 60 GWd/tU. The burnup of the central fuel

assembly was then lowered to 10 GWd/tU considering the same intermediate burnup values used in

the previous section. Therefore, in this set of simulations the central fuel assembly has the lowest

source term compared to the other assemblies in the storage rack.

The average normalized detector response for the nine central guide tubes and the corresponding

standard deviations were included in Table 6-8 with the maximum difference among the guide tubes.

For both neutron detector types the average values calculated for fuel with different burnup were

within the standard deviations. Similar results were obtained for the 235U and 238U fission chambers.

The maximum difference among the guide tubes ranged between 3% and 8% for both detector types

and was the smallest for fuel with burnup of 30 GWd/tU.

The average values for the ionization chambers in case of fuel with burnup higher than 20 GWd/tU

were within the standard deviations and were larger than for the neutron detectors. In case of fuel

with burnup of 10 GWd/tU the average value of the detector response was 10% lower than in the

other cases. The gamma-ray source term for fuel with burnup of 10 GWd/tU was about 10% of the

other assemblies in the storage rack; therefore, the normalized detector responses are mainly due to

the lateral and corner fuel assemblies. Due to the self-shielding effect of the fuel pins, the guide

tubes at the center of the assembly have lower normalized detector responses compared to the

reference case.

Table 6-8: Average normalized detector responses and standard deviation for the nine central guide tubes. The maximum


detectors and within 0.8% for the ionization chambers.

Central

fuel ass.

burnup


Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.964 ± 0.018 6% 0.961 ± 0.020 6% 0.894 ± 0.012 4%

20 GWd/tU 0.974 ± 0.013 4% 0.971 ± 0.015 5% 0.989 ± 0.004 1%

30 GWd/tU 0.979 ± 0.010 3% 0.976 ± 0.013 4% 0.990 ± 0.004 1%

40 GWd/tU 0.972 ± 0.013 4% 0.972 ± 0.014 4% 0.990 ± 0.005 2%

50 GWd/tU 0.963 ± 0.019 6% 0.966 ± 0.018 5% 0.990 ± 0.005 1%

60 GWd/tU 0.953 ± 0.024 8% 0.959 ± 0.022 7% 0.991 ± 0.006 1%

84

Figure 6-16 shows the average relative difference (DK,R) between the normalized detector responses

compared to the reference case for the nine central guide tubes. The statistical uncertainty of the

values in the plot was within 0.3%.

The relative differences for the neutron detectors were within +3% and the results for the 235U fission

chambers were larger than those obtained for the 238U fission chambers. For both detector types the

maximum values were obtained for a central fuel assembly with burnup of 30 GWd/tU. The results

for the ionization chambers do not show a trend with fuel burnup larger than 20 GWd/tU, whereas

for fuel with burnup of 10 GWd/tU a difference of -10% was obtained due to the small source term of


Figure 6-16: Average relative difference as defined in Formula (6.1) between the detector responses calculated in the nine

guide tubes at the center of the fuel assembly.

The average normalized detector responses for the peripheral guide tubes, the corresponding

standard deviation, and maximum difference among guide tubes are shown in Table 6-9. For both

fission chambers the average detector responses were the highest for fuel with burnup of 30

GWd/tU, and similar values were obtained for the two detector types for each fuel burnup. The

maximum difference of the normalized detector responses in the peripheral guide tubes was within

+5% in all cases. The average normalized detector responses for the ionization chambers decreased

with the fuel burnup, and it was higher than the results for the neutron detectors. The maximum

difference among the guide tubes was within +4%.

0 10 20 30 40 50 60

-20

-10

0

10

20

Central guide tubes (60)

DK

,R

(%

)




Ionization chamber

85



detectors and within 0.8% for the ionization chambers.

Central

fuel ass.

burnup


Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.844 ± 0.008 3% 0.857 ± 0.014 5% 0.985 ± 0.013 4%

20 GWd/tU 0.874 ± 0.006 2% 0.892 ± 0.011 4% 0.990 ± 0.006 2%

30 GWd/tU 0.887 ± 0.004 1% 0.911 ± 0.010 3% 0.961 ± 0.004 1%

40 GWd/tU 0.871 ± 0.005 2% 0.897 ± 0.012 4% 0.946 ± 0.004 2%

50 GWd/tU 0.840 ± 0.008 3% 0.870 ± 0.013 4% 0.939 ± 0.004 1%

60 GWd/tU 0.809 ± 0.011 4% 0.843 ± 0.015 5% 0.935 ± 0.005 2%

Figure 6-17 shows the average relative difference (DK,R) compared to the reference case for the

sixteen guide tubes at the periphery of the assembly, and the statistical uncertainty of the results

was within 0.2%. By decreasing the burnup of the central fuel assembly from 60 GWd/tU, the gamma-

ray detector responses were higher than in the reference case and reached a maximum difference of

+6%. This was due to the strong source term of the lateral and corner fuel assemblies compared to


The responses of the 235U fission chambers were more influenced by the change in the burnup of the

central fuel assembly compared to the values for the 238U fission chambers. The relative difference of

the detector responses for the peripheral guide tubes reached a maximum for fuel with burnup of

30 GWd/tU, as shown also for nine central guide tubes in Figure 6-16. The maximum relative

difference with the reference case for the peripheral guide tubes was around +10% for the 235U

fission chambers and +8% for the 238U fission chambers.

86



The simulations with the central fuel assembly surrounded by high-burnup assemblies were

performed also by using the NONU card to neglect the emission of secondary neutrons from fissions.

The average normalized detector responses in the peripheral guide tubes are shown in Table 6-10,

with the standard deviation and the maximum difference among the guide tubes.

The average values for both detector types were larger for fuel with burnup lower than 50 GWd/tU

and the variation shown in the different simulations was within the standard deviation. In all cases

the standard deviation of the normalized detector response among the guide tubes decreased with

the increase of the fuel burnup. The maximum difference calculated among the guide tubes was

around 10% for the 235U fission chambers and around 6% for the 238U fission chambers.

Both the average normalized detector responses and maximum difference among the guide tubes

calculated with the NONU card were higher than in the results of Table 32, and the variation was

larger for fuel with low burnup.

0 10 20 30 40 50 60

-20

-10

0

10

20

Peripheral guide tubes (60)

DK

,R

(%

)




Ionization chamber

87


difference calculated among the guide tubes is also reported. The statistical uncertainty was around 0.3% for fuel with

burnup of 10 GWd/tU and around 0.1% for all other cases.

Central

fuel ass.

burnup

235U fission chambers 238U fission chambers

Average and

std. deviation

Maximum

difference

Average and

std. deviation

Maximum

difference

10 GWd/tU 0.955 ± 0.030 9% 0.981 ± 0.020 6%

20 GWd/tU 0.957 ± 0.030 10% 0.984 ± 0.019 5%

30 GWd/tU 0.964 ± 0.024 8% 0.986 ± 0.013 4%

40 GWd/tU 0.971 ± 0.014 5% 0.988 ± 0.005 2%

50 GWd/tU 0.905 ± 0.005 2% 0.960 ± 0.006 2%

60 GWd/tU 0.837 ± 0.007 3% 0.904 ± 0.009 3%

The average relative difference (DK,R) for the normalized detector responses at the peripheral guide

tubes was calculated for the simulations with the NONU card, and the results are shown in Figure 6-

18. The detector responses obtained in the previous simulations for fuel with burnup of 60 GWd/tU

were used as reference case for the calculation of the relative difference. Constant values of the DK,R

ratio were obtained for both neutron detector types in case of fuel with burnup smaller than 40

GWd/tU. The average relative difference was between +15% and +20% for both detector types. The

statistical uncertainty of the results was within 0.3%.

The results in Figure 6-18 do not show a trend for fuel with burnup smaller than 40 GWd/tU,

therefore the trend shown in Figure 6-17 for both neutron detectors is due to the secondary

neutrons emitted from fissions in the fuel assemblies, and their effect decreases with the increase of

the fuel burnup.

88


peripheral guide tubes. The NONU card was used to neglect the emission of secondary neutrons from fission.

6.3.2 Impact of the lateral fuel assemblies on the reference distributions

The storage rack considered in the simulations contained nine fuel assemblies in a 3x3 pattern. In

addition to the central fuel assembly, the storage rack contains four so-called lateral fuel assemblies

having one side facing the central fuel assembly, and four corner fuel assemblies placed at the

corners of the storage rack.

The storage rack configurations developed for the evaluation of the impact of the lateral fuel

assemblies on the normalized detector responses calculated in the central fuel assembly are shown

in Figure 6-19. The reference case consisted in a uniform storage rack containing fuel with burnup of

10 GWd/tU, and five scenarios were modelled by placing fuel assemblies with burnup of 60 GWd/tU in

the lateral positions. The material composition and source emission probability were adapted in each

case to reflect the scenarios depicted in Figure 6-19.

Fuel assemblies with burnup of 60 GWd/tU were chosen to evaluate the impact of the lateral fuel

assemblies on the detector responses in the guide tubes because of the very large neutron or

gamma-ray emissions compared to the assemblies with 10 GWd/tU.

0 10 20 30 40 50 60

-20

-10

0

10

20

Peripheral guide tubes - NONU

DK

,R

(%

)




89

Figure 6-19: Storage rack configurations for the study of the influence of lateral fuel assemblies.

The normalized neutron detector responses were calculated for the guide tubes in the central fuel

assembly and for cases 1, 2, and 4 the maximum detector response was obtained in the guide tubes

close to the fuel assemblies with high burnup. The maximum values were obtained in the central

guide tube for cases 3 and 5.

The maximum normalized detector response for the ionization chambers was obtained for all cases

in the peripheral guide tubes close to the fuel assemblies with high burnup. The shielding effect of

the fuel pins is more significant for gamma-rays than for neutrons, therefore the nine central guide

tubes were less affected by the lateral and corner fuel assemblies.

The relative difference DK,R compared to the reference case was calculated for the configurations in

Figure 6-19 and the results are shown in Figure 6-20. The values refer to 235U fission chambers, but

similar results were obtained also for 238U fission chambers. The statistical uncertainty of the values

was around 0.2%. In all cases the maximum difference compared to the reference case was

calculated for the guide tubes close to the assemblies with burnup of 60 GWd/tU. Differences up to

+25% were obtained for the results shown in Figure 6-20. The normalized detector responses for

those guide tubes are larger than in the reference case due to the strong neutron emission from the

assemblies with high burnup. As a consequence the normalized detector responses for the guide

tubes far from the high-burnup fuel assemblies are lower than in the reference case, and negative

relative differences were obtained. This is visible for cases 1 and 2, due to the asymmetry of the

storage rack configuration, and the reduction of the detector responses were up to -48% from the

reference case.

Fuel burnup

10 GWd/tU

60 GWd/tU

Case 1 Case 2

Case 3 Case 5Case 4

90

The influence of the lateral assemblies was also evaluated for the gamma-ray detectors considering

the storage rack configurations described in Figure 6-19. The values of the relative difference DK,R are

shown in Figure 6-21 for the different guide tubes, and the statistical uncertainty of the results is

within 0.9%. In all cases considered in the study, the maximum difference of the normalized detector

response was calculated for the guide tubes close to the assemblies with burnup of 60 GWd/tU. The

figure shows that the increase of the normalized detector response is localized only in the guide

tubes next to the fuel assembly with high burnup. Values within ±10% from the reference case were

obtained in all configurations.

Comparing the results of the neutron and gamma-ray detectors, the former is more influenced than

the latter by the difference in the burnup of the fuel assemblies in the storage rack. This is due to the

self-shielding effect that is more evident for gamma-rays than for neutrons emitted in the storage

rack.

91

Figure 6-20: Relative difference between the normalized detector response calculated for storage racks with the lateral fuel

assemblies with different burnup and the values obtained in the reference case. The results refer to 235

U fission chambers,

and the title of each plot indicates the storage rack configuration described in Figure 6-19.

92

Figure 6-21: Relative difference between the normalized detector response calculated for storage racks with the lateral fuel

assemblies with different burnup and the values obtained in the reference case. The results refer to ionization chambers,


93

6.3.3 Impact of the corner fuel assemblies on the reference distributions

A set of simulations were performed to study the impact of the corner fuel assemblies on the

normalized detector responses in the guide tubes of the central fuel assembly. As in the previous

section the reference case for the storage rack had fuel assemblies with burnup of 10 GWd/tU and

fuel assemblies with burnup of 60 GWd/tU were then introduced in the corner positions as shown in

Figure 6-22.

Figure 6-22: Storage rack configurations for the study of the influence of corner fuel assemblies.

The maximum normalized neutron detector responses were obtained for the guide tubes close to the

corner fuel assemblies with high burnup for cases 6, 7, and 9, whereas the maximum was obtained in

the central guide tube for cases 8 and 10. The normalized detector responses obtained for the

ionization chambers were almost identical to the reference case.

Figure 6-23 shows the relative difference DK,R of the normalized detector responses calculated for

235U fission chambers in the cases depicted in Figure 6-22 compared to the reference case. The

statistical uncertainty of the values was around 0.3%, and similar results were obtained also for the

238U fission chambers. As in the previous section, the maximum difference compared to the reference

case was calculated for the guide tubes close to the fuel assemblies with high burnup, and this is due

to the strong neutron emission from these assemblies. The increase of the normalized detector

response was as high as +30% compared to the reference case, and it is evident for case 6. The

detector responses in the guide tubes far from the high-burnup fuel assemblies were also lower

compared to the reference case, and the relative differences were within -50%.

Fuel burnup

10 GWd/tU

60 GWd/tU

Case 6 Case 7

Case 8 Case 9 Case 10

94

Figure 6-23: Relative difference between the normalized detector response calculated for storage racks with the corner fuel

assemblies with different burnup and the values obtained in the reference case. The results refer to 235

U fission chambers,


The impact of the corner fuel assemblies was also evaluated for the normalized gamma-ray detector

responses, but limited effects were observed. The relative difference between the normalized

detector responses calculated in the scenarios and in the reference case was for most of the values

within the statistical uncertainty, therefore the results were not included in this chapter. The small

95

impact of the corner fuel assemblies on the normalized detector responses in the guide tubes of the

central fuel assembly is due to the shielding effect of the fuel pins in the storage rack.

6.4 Conclusions

The results from the Monte Carlo study on the PDET detector were presented in this chapter. The

importance function from each fuel pin of the fuel assembly being measured was calculated in the

case of 235U and 238U fission chambers, and ionization chambers. All fuel pins contributed in a

significant way to the responses of both neutron detectors, whereas the contributions to the

ionization chambers were strongly localized in the vicinity of the guide tube containing the detector.

The detector responses from several neutron and gamma ray detector types were calculated but, as

no appreciable differences were observed by changing the detector type, 235U and 238U fission

chambers were selected in order to have detectors sensitive to thermal and fast neutrons,

respectively. In addition, ionization chambers with nitrogen as filling gas at 1 atm were chosen as

gamma-ray detectors.

The initial enrichment and cooling time of the spent fuel assembly did not affect significantly the

normalized detector responses across the fuel assembly cross-section, whereas the fuel burnup had

an effect within ±3% on the normalized neutron detector responses.

The influence of the fuel assemblies in the storage rack on the normalized detector responses

calculated in the guide tubes of the central fuel assembly was evaluated. The burnup of the central

fuel assembly had the largest impact in case of storage rack with high burnup fuel assemblies in the

lateral and corner positions. The neutron detectors in the guide tubes at the periphery of the

assembly were mostly affected by the change in burnup of the central fuel assembly, and differences

within ±10% were calculated on the normalized detector responses compared to the reference case.

Several storage rack configurations were also developed to estimate the influence of the lateral and

corner fuel assemblies, and for all cases the maximum differences on the normalized detector

responses compared to the reference case were obtained in the guide tubes close to the assemblies

with high burnup. Variations between -50% and +30% were calculated for the different guide tubes

in the central fuel assembly. Because of the self-shielding effect of the fuel pins, the gamma-ray

detector responses were in general less influenced by the fuel assemblies with different burnup

compared to the responses of the fission chambers.

97

7 Analysis of the partial defect capabilities for SINRD and PDET

7.1 Description of the diversion scenarios

The two NDA techniques investigated in this work were compared in terms of their capability to

detect partial defects. The current IAEA goal for partial defect testing is to verify that at least 50% of

the fuel pins are present in a fuel assembly (IAEA, 2009). Therefore, a series of 12 diversion scenarios

were created for this analysis, considering replacements from 50% to 15% of the fuel pins of the

measured fuel assembly. The diverted fuel pins were replaced by dummies made of stainless steel

with the same dimensions of the original spent fuel pins. This material was identified as the worst-

case scenario in (Rossa, 2013a).

Figure 7-1: Visualization of the diversion scenarios developed for the comparison of the NDA techniques. The fuel pins are depicted in white, the dummy pins in grey, and the guide tubes in yellow.

The diversion scenarios are shown in Figure 7-1, where the fuel pins are depicted in white, the

dummy pins in grey, and the guide tubes in yellow. A symmetric pattern was chosen for all cases

since it was considered the most challenging situation to detect (Sitaraman, 2009). Uniform diversion

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1

1 5 1 1 1 1 5 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 5 1 5 1 5 1 5 1 5 1 1 1 5 1 5 5 5 1 5 1 1 1 1 5 1 5 5 5 1 5 1 1 1 1 5 1 5 5 5 1 5 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 5 1 5 1 5 1 5 1 5 1 1 5 5 5 5 5 1 1 5 5 5 5 5 1 5 5 5 5 5

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 5 1 5 1 5 1 5 1 5 1 1 1 5 1 5 5 5 1 5 1 1 1 1 5 1 5 5 5 1 5 1 1 1 1 5 1 5 5 5 1 5 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 5 1 1 1 1 5 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1 1 1 5 1 1

1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1 1 1 1 5 1 5 1 5 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 5 5 5 1 1 1 5 5 5 1 1 1 1 5 5 5 1 1 1 5 5 5 1

5 1 1 1 1 5 1 1 5 1 1 5 1 1 1 5 5 1 1 5 1 1 5 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

5 5 5 5 5 1 5 5 5 5 5 1 5 5 5 5 5 1 5 5 5 5 5 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

5 5 5 5 5 1 5 5 5 5 5 1 1 5 5 5 5 5 1 1 5 5 5 5 5 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

5 5 5 5 5 1 5 5 5 5 5 1 5 5 5 5 5 1 5 5 5 5 5 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

5 1 1 1 1 5 1 1 5 1 1 5 1 1 1 5 5 1 1 5 1 1 5 1

1 5 5 5 1 1 1 5 5 5 1 1 1 1 5 5 5 1 1 1 5 5 5 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1

1 0 0 0 0 5 0 0 5 0 0 5 0 0 0 0 1 0 1 0 0 0 5 0 0 5 0 0 5 0 0 0 1 0 1 5 5 5 1 1 5 5 5 1

0 0 0 5 1 0 0 0 0 0 0 0 1 5 0 0 0 1 0 0 5 0 0 0 0 0 0 0 0 0 5 0 0 1 5 5 5 5

1 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1

0 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 0 1 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 1 5 5 5 5 5 5 5 5 5 5

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1

1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1

1 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 1 0 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 0 5 5 5 5 5 1 5 5 5 5 5 1

1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1

0 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 0 1 0 5 0 0 5 0 0 5 0 0 5 0 0 5 0 1 5 5 5 5 5 5 5 5 5 5

1 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1

0 0 0 5 1 0 0 0 0 0 0 0 1 5 0 0 0 1 0 0 5 0 0 0 0 0 0 0 0 0 5 0 0 1 5 5 5 5

1 0 0 0 0 5 0 0 5 0 0 5 0 0 0 0 1 0 1 0 0 0 5 0 0 5 0 0 5 0 0 0 1 0 1 5 5 5 1 1 5 5 5 1

1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1

1 1 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Diversion 9: 20% Diversion 10: 20% Diversion 11: 15% Diversion 12: 15%



98

was taken into account (Diversion 1), as well as cases with replacement of pins in the outer section of

the assembly (e.g. Diversion 2) and from the inner section of the assembly (Diversion 5). A set of

asymmetric diversion scenarios was also developed for the PDET detector and the results are

included in Annex B.

The simulations for the SINRD technique were carried out by modelling a single spent fuel assembly

in air and surrounded by a polyethylene slab, according to the optimal configuration defined in

Chapter 4. The fuel pins had a material composition and source emission probability corresponding

to spent fuel with 3.5% initial enrichment, burnup of 30 GWd/tHM, and cooling time of 5 years. The

neutron fluxes were calculated in all guide tubes of the assembly to mimic the insertion of multiple

detectors at the same time. The neutron fluxes were normalized to the maximum value obtained

among the guide tubes. The SINRD filters were not included in the model, but their influence was

calculated with the analytical approach described in Section 4.4.1.

The simulations for the PDET detector considered the storage rack with 9 fuel assemblies stored in

fresh water, and the diversion occurred only in the fuel assembly in the central position while the

rest of the fuel assemblies in the storage rack were not modified from the reference case. The

material composition and source terms of the spent fuel pins for the simulations on PDET were the

same as those defined for SINRD.

7.2 Response of SINRD to the diversion scenarios

The normalized detector responses around the 239Pu resonance and in the fast energy regions were

calculated to evaluate the capability of SINRD to detect missing fuel pins. Based on the results in

Chapter 4 a 239Pu fission chamber and SINRD filters of 0.1 mm Gd and 1.0 mm Cd were used to

compute the detector response around 0.3 eV, whereas a bare 238U fission chamber was chosen as

fast neutron detector.

The reference case for SINRD consisted in a single fuel assembly without pins replaced and

surrounded by a 12-cm thick slab of polyethylene. The normalized detector responses for the

reference case are shown in Figure 7-2. The values have a statistical uncertainty within 0.6% for the

239Pu fission chambers and 0.1% for the 238U fission chambers.

The normalized detector responses calculated for the 239Pu fission chambers show that the maximum

was obtained for the guide tubes at the periphery of the assembly, whereas for the 238U fission

chambers the maximum was calculated for the central guide tubes. Despite the different

measurement setup, the results for the 238U fission chambers were similar to the values obtained for

the PDET detector in the previous chapter. On the contrary the results for the 239Pu fission chambers

99

were different due to the measurement configuration modelled for the SINRD detector. The slab of

polyethylene placed around the fuel assembly for neutron moderation is responsible for the high

detector response calculated for the guide tubes at the periphery of the assembly, and the self-

shielding effect of the fuel pins causes a lower detector response in the central section of the fuel

assembly.

Figure 7-2: Normalized detector responses in the reference case considering 239

Pu fission chambers and SINRD filters (left) and

238U fission chambers (right).

The average normalized detector responses in the nine central guide tubes are shown in Table 7-1 for

the two detector types in the reference case and diversion scenarios. The standard deviation of the

values in the different guide tubes was also included with the maximum difference calculated among

the guide tubes. The maximum difference was expressed as percentage of the average detector

response.

The results in Table 7-1 show that the average values for the detector responses of the 239Pu fission

chambers were lower than for the 238U fission chambers both for the reference case and the

diversion scenarios. The maximum difference calculated among the guide tubes for the 239Pu fission

chambers in the diversion scenarios showed similar values compared to the reference case, but

values more than three times higher were obtained for the 238U fission chambers. Therefore, the

increase in the maximum difference among the nine central guide tubes provides an indication of the

fuel pins diversion.

The values included in Table 7-1 are also reported in Figure 7-3 for both detector types. In all cases

the average values of the normalized detector responses in the nine central guide tubes for the

diversion scenarios were within ±15% from the reference case. No significant variation was observed

for the range of the normalized detector responses calculated among the guide tubes for the 239Pu

fission chambers in the diversion scenarios, but the increase in the maximum difference was visible

100

for the 238U fission chambers for the diversion scenarios with 50% of replaced fuel pins. The smallest

range among the guide tubes was obtained for diversion scenario 5. In this case the dummy pins

were placed in the central section of the fuel assembly, whereas in all other cases were placed in the

outer region. In general, the average difference of the detector responses compared to the reference

case decreases by decreasing the percentage of dummy pins in the fuel assembly.

Table 7-1: Average normalized detector responses and standard deviation for the nine central guide tubes. The maximum difference calculated among the guide tubes is also reported. The statistical uncertainty was within 0.5% for the

239Pu

fission chambers and around 0.1% for the 238

U fission chambers.

239Pu fission chambers 238U fission chambers

Average and std. deviation

Maximum difference


Maximum difference

Reference 0.580 ± 0.053 27% 0.968 ± 0.017 5%

Diversion 1 0.588 ± 0.047 26% 0.936 ± 0.030 10%

Diversion 2 0.495 ± 0.057 34% 0.904 ± 0.051 16%

Diversion 3 0.501 ± 0.060 35% 0.896 ± 0.055 17%

Diversion 4 0.523 ± 0.067 37% 0.888 ± 0.060 19%

Diversion 5 0.644 ± 0.048 23% 0.989 ± 0.004 1%

Diversion 6 0.537 ± 0.064 33% 0.931 ± 0.039 12%

Diversion 7 0.560 ± 0.058 29% 0.949 ± 0.028 8%

Diversion 8 0.549 ± 0.063 32% 0.939 ± 0.035 10%

Diversion 9 0.561 ± 0.060 30% 0.948 ± 0.029 8%

Diversion 10 0.556 ± 0.055 28% 0.954 ± 0.024 7%

Diversion 11 0.566 ± 0.055 28% 0.960 ± 0.022 6%

Diversion 12 0.559 ± 0.057 29% 0.959 ± 0.023 7%

Figure 7-3: Average detector responses and standard deviation for the reference case and the diversion scenarios for the

nine central guide tubes. The range of the normalized detector response calculated among the guide tubes is also reported.

The values refer to 239

Pu fission chambers covered by the SINRD filters (left), and to bare 238

U fission chambers (right).

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

101

Table 7-2 shows the values calculated for the two detector types in the peripheral guide tubes. The

average values of the normalized detector responses of the 239Pu fission chambers in the reference

case and in the diversion scenarios were within the standard deviation, whereas the values for the

238U fission chambers showed significant differences between the reference case and the diversion

scenarios. The average normalized detector responses were lower than in the reference case for all

diversion scenarios apart from diversion 5, and reductions over -25% were obtained for diversions 2,

3, and 4.

The maximum difference calculated among the guide tubes for the 239Pu fission chambers in the

diversion scenarios with 50% of dummy pins was almost double than the value in the reference case,

and it was about five times higher for the 238U fission chambers. As for the central guide tubes, the

increase in the maximum difference among the peripheral guide tubes provides an indication of the

fuel pins diversion. For both central and peripheral guide tubes the 238U fission chambers were more

affected by the diversion.

Table 7-2: Average normalized detector responses and standard deviation for the peripheral guide tubes. The maximum difference calculated among the guide tubes is also reported. The statistical uncertainty was within 0.5% for the

239Pu

fission chambers and around 0.1% for the 238

U fission chambers.

Diversion scenario

239Pu fission chambers 238U fission chambers


Maximum difference


Maximum difference

Reference 0.975 ± 0.025 7% 0.863 ± 0.010 3%

Diversion 1 0.950 ± 0.048 12% 0.813 ± 0.040 11%

Diversion 2 0.952 ± 0.048 12% 0.628 ± 0.037 14%

Diversion 3 0.950 ± 0.048 13% 0.629 ± 0.037 16%

Diversion 4 0.950 ± 0.050 12% 0.651 ± 0.040 15%

Diversion 5 0.978 ± 0.022 6% 0.950 ± 0.020 6%

Diversion 6 0.976 ± 0.034 9% 0.736 ± 0.027 10%

Diversion 7 0.974 ± 0.032 8% 0.796 ± 0.023 8%

Diversion 8 0.972 ± 0.025 7% 0.776 ± 0.015 5%

Diversion 9 0.970 ± 0.030 9% 0.807 ± 0.022 8%

Diversion 10 0.977 ± 0.021 6% 0.800 ± 0.015 5%

Diversion 11 0.978 ± 0.025 7% 0.823 ± 0.017 5%

Diversion 12 0.976 ± 0.028 7% 0.815 ± 0.018 6%

The average detector responses and the standard deviations were also shown in Figure 7-4 for the

239Pu fission chambers and the 238U fission chambers in the peripheral guide tubes. The range of the

normalized detector response among the guide tubes was also included. No significant variations

102

were observed for the results with the 239Pu fission chambers, whereas the average values for the

238U fission chambers show a reduction of -30% for the diversion scenarios with 50% of dummy pins.

An increase of +10% was observed for diversion 5 and it is due to the position of the replaced pins in

the assembly.

For both detector types the increase of the maximum difference among the guide tubes was visible

for the diversion scenarios with 50% of dummy pins, and the maximum difference decreases by

considering smaller diversions.


peripheral guide tubes. The range of the normalized detector responses calculated among the guide tubes is also reported.


Pu fission chambers covered by the SINRD filters (left), and to bare 238


7.3 Response of PDET to the diversion scenarios

The normalized detector responses in the case of 235U fission chambers, 238U fission chambers, and

ionization chambers were calculated to evaluate the capability of PDET to detect fuel pin diversion.

The normalized detector responses in the reference case were included in Chapter 6 and are not

repeated in this section.

The average normalized detector responses in the central guide tubes are shown in Table 7-3 for the

different detector types in the reference case and diversion scenarios. The standard deviation of the

values in the guide tubes and the maximum difference are also included.

For all detector types the variation of the average detector responses in the reference case and in

the diversion scenarios was comparable with the standard deviation calculated for the different

guide tubes. Similar average values were obtained in the reference case for both neutron detectors,

but differences around 10% were calculated between the two detector types for the diversion

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

103

scenarios 2 and 3. The average values for the ionization chambers were lower in all diversion

scenarios compared to the reference case, and reductions over -10% were obtained for diversion

scenarios 4 and 5.

The maximum difference calculated among the guide tubes for the 235U fission chambers was in

general smaller in the diversion scenarios than in the reference case, and the values for diversion 2

was 25% of the reference case. The maximum differences for the 238U fission chambers and the

ionization chambers were higher in the diversion scenarios compared to the reference case. Values

over twice and 12 times the reference case were obtained respectively for the two detector types in

diversion scenario 4.

Table 7-3: Average normalized detector responses and standard deviation for the nine central guide tubes. The maximum difference calculated among the guide tubes is also reported. The statistical uncertainty was around 0.2% for the neutron detectors and around 0.4% for the ionization chambers.



Maximum difference


Maximum difference


Maximum difference

Reference 0.947 ± 0.026 8% 0.952 ± 0.025 8% 0.991 ± 0.005 1%

Diversion 1 0.970 ± 0.028 6% 0.947 ± 0.041 10% 0.935 ± 0.063 14%

Diversion 2 0.986 ± 0.006 2% 0.909 ± 0.050 15% 0.918 ± 0.044 14%

Diversion 3 0.992 ± 0.007 2% 0.901 ± 0.056 17% 0.902 ± 0.054 17%

Diversion 4 0.958 ± 0.013 3% 0.892 ± 0.062 19% 0.881 ± 0.064 21%

Diversion 5 0.974 ± 0.017 5% 0.971 ± 0.006 2% 0.870 ± 0.019 5%

Diversion 6 0.971 ± 0.012 4% 0.927 ± 0.041 12% 0.947 ± 0.038 10%

Diversion 7 0.953 ± 0.023 7% 0.945 ± 0.030 9% 0.976 ± 0.014 4%

Diversion 8 0.968 ± 0.013 4% 0.934 ± 0.037 11% 0.955 ± 0.033 9%

Diversion 9 0.962 ± 0.018 6% 0.940 ± 0.034 10% 0.967 ± 0.022 6%

Diversion 10 0.948 ± 0.025 8% 0.946 ± 0.028 9% 0.981 ± 0.011 3%

Diversion 11 0.949 ± 0.026 8% 0.950 ± 0.027 8% 0.988 ± 0.008 2%

Diversion 12 0.949 ± 0.025 8% 0.949 ± 0.027 8% 0.985 ± 0.011 3%

The values reported in Table 7-3 are also shown in Figure 7-5 for the two neutron detectors and in

Figure 7-6 for the ionization chambers. The average normalized detector responses for the 235U

fission chambers in the diversion scenarios were within the standard deviation of the result obtained

in the reference case. Values slightly higher than the reference case were calculated for diversion

scenarios 2, 3, and 4. For the 238U fission chambers and ionization chambers the average values were

in general smaller for the diversion scenarios compared to the reference case, but the differences

were in general within the standard deviation.

104

The range of the normalized detector responses for the 238U fission chambers and the ionization

chambers were in general higher in the diversion scenarios compared to the reference case. Ranges

between 2 and 10 times higher than the reference case were obtained for the two detector types in

the diversion scenarios with 50% of dummy pins. On the contrary, the range of the detector

responses calculated for scenario 5 was significantly smaller than in the reference case.


nine central guide tubes. The range of the normalized detector responses calculated among the guide tubes is also

reported. The values refer to 235

U fission chambers (left), and to 238



nine central guide tubes. The range of the normalized detector responses calculated among the guide tubes is also

reported. The values refer to ionization chambers.

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

105

Table 7-4 shows the average normalized detector responses in the peripheral guide tubes for the

different detector types in the reference case and diversion scenarios. The standard deviation of the

average values and the maximum difference among the guide tubes are also included.

The average normalized detector responses for the 235U fission chambers were higher for the

diversion scenarios than for the reference case, and the differences up to +20% were obtained for

the diversion scenarios with 50% of dummy pins. For the 238U fission chambers and ionization

chambers the average values were in general smaller for the diversion scenarios compared to the

reference case and reductions up to -30% were obtained. For the 238U fission chambers the average

value was almost 20% higher in diversion 5 than in the reference case.

The maximum differences calculated among the guide tubes for the 235U fission chambers were

about 50% larger for the diversion scenarios with 50% of dummy pins than in the reference case. The

maximum differences for the 238U fission chambers and the ionization chambers were in general

higher in the diversion scenarios compared to the reference case. Values between 2 and 10 times

higher than the reference case were obtained for the two detector types in diversion scenarios with

50% of dummy pins.

Table 7-4: Average normalized detector responses and standard deviations for the peripheral guide tubes. The maximum difference calculated among the guide tubes is also reported. The statistical uncertainty was around 0.2% for the neutron detectors and around 0.4% for the ionization chambers.



Maximum difference


Maximum difference


Maximum difference

Reference 0.799 ± 0.012 4% 0.822 ± 0.016 5% 0.936 ± 0.005 2%

Diversion 1 0.880 ± 0.015 4% 0.896 ± 0.041 10% 0.874 ± 0.066 17%

Diversion 2 0.957 ± 0.019 5% 0.657 ± 0.039 14% 0.633 ± 0.050 18%

Diversion 3 0.959 ± 0.019 5% 0.660 ± 0.040 16% 0.632 ± 0.051 20%

Diversion 4 0.941 ± 0.021 5% 0.685 ± 0.043 15% 0.654 ± 0.053 18%

Diversion 5 0.839 ± 0.018 6% 0.971 ± 0.019 6% 0.935 ± 0.045 14%

Diversion 6 0.887 ± 0.009 3% 0.734 ± 0.031 12% 0.758 ± 0.028 10%

Diversion 7 0.827 ± 0.008 3% 0.796 ± 0.023 8% 0.858 ± 0.021 7%

Diversion 8 0.852 ± 0.011 4% 0.774 ± 0.018 6% 0.826 ± 0.008 2%

Diversion 9 0.828 ± 0.011 4% 0.793 ± 0.026 9% 0.872 ± 0.018 6%

Diversion 10 0.823 ± 0.031 4% 0.788 ± 0.018 6% 0.860 ± 0.004 2%

Diversion 11 0.810 ± 0.031 4% 0.804 ± 0.018 6% 0.895 ± 0.006 2%

Diversion 12 0.821 ± 0.024 3% 0.795 ± 0.020 7% 0.874 ± 0.016 5%

106

Figure 7-7 shows the values calculated in Table 7-4 for the two neutron detectors and Figure 7-8

contains the results for the ionization chambers. The average normalized detector responses for the

235U fission chambers were higher for the diversion scenarios than for the reference case, and

differences up to +20% were obtained for diversion scenarios 2, 3, and 4. For the 238U fission

chambers and ionization chambers the average values were in general smaller for the diversion

scenarios compared to the reference case and reductions up to -30% were obtained.


peripheral guide tubes. The range of the normalized detector response calculated among the guide tubes is also reported.


U fission chambers (left), and to 238



peripheral guide tubes. The range of the normalized detector response calculated among the guide tubes is also reported.

The values refer to ionization chambers.

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Range

Ref 1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

De

tecto

r re

sp

on

se

Diversion scenario

Average value

107

7.4 Conclusions

The SINRD and PDET techniques were compared in this section in terms of their ability to detect fuel

pins replaced by dummy pins. For both techniques significant differences in the normalized detector

responses were observed for the diversion scenarios with 50% of fuel pins replaced, and the most

challenging scenario was obtained with the replacement with a chess-board pattern.

In the case of SINRD the detector responses of the 238U fission chambers were more affected by the

fuel pins diversion than the 239Pu fission chambers covered by the SINRD filters. The average

normalized detector responses in the central and peripheral guide tubes, as well as the maximum

difference among the guide tubes, were different for the diversions with 50% of dummy pins

compared to the reference case. Average relative differences between -30% and +15% were

calculated from the reference case considering both detector types.

The results for the PDET detector showed that the detector responses of the gamma-ray detectors

have the highest sensitivity to the diversion scenarios. As for SINRD, the variation in the average

detector responses and the maximum variations among the guide tubes provide indications for the

detection of diversion scenarios with 50% of dummy pins. Differences up to -30% from the reference

case were calculated for the ionization chambers in the guide tubes at the periphery of the assembly.

The variations for the neutron detectors were within ±20% for the peripheral guide tubes and within

±10% for the central guide tubes. The magnitude of the variation of the neutron detector responses

due to the fuel pins diversion is similar to the value obtained for a storage rack containing fuel

assemblies with different burnup, whereas the variation of the gamma-ray detector responses is

much larger in the case of diversion. Therefore, from the scenarios considered in this work it seems

that by combining the information from neutron and gamma-ray detectors the diversion of 50% of

the total fuel pins cannot be masked by the burnup declaration of the fuel assemblies in the storage

rack.

The results presented in this chapter give an initial estimation of the capabilities of both techniques

for the partial defect detection. However, detailed considerations about the measurement

uncertainty have to be included in the analysis to determine the limit of detection for the fuel pins

diversion.

109

8 Discussion and conclusion

This thesis described the development of two NDA techniques for the safeguards verification of

spent nuclear fuel. The Self-Indication Neutron Resonance Densitometry (SINRD) and the Partial

Defect Tester (PDET) were investigated considering the PWR 17x17 fuel geometry as reference. Both

techniques foresee the insertion of small neutron or gamma-ray detectors in the guide tubes of the

fuel assembly.

Preliminary work was focused on the development of a reference spent fuel library to obtain the

isotopic composition of spent fuel with different irradiation histories. The irradiation history was

defined in terms of initial enrichment, burnup, and cooling time. In addition, the neutron and

gamma-ray emissions were calculated both as magnitude and energy spectra. The simulations code

ORIGEN-ARP and ALEPH-2.2 were used to develop the spent fuel library, and the output files were

processed to be compatible with the format of the MCNPX Monte Carlo input file.

8.1 Self-Indication Neutron Resonance Densitometry

The SINRD technique was first investigated by means of Monte Carlo simulations to identify the

optimal conditions for the application of the method. The measurement of the fuel assembly kept in

air and surrounded by a thick layer of polyethylene showed a clearer reduction of the neutron flux

around 0.3 eV compared to the results obtained with the assembly stored under water. This is due to

the lack of moderation within the fuel assembly in the dry conditions. Previous work on SINRD

(LaFleur, 2011) considered the spent fuel assemblies stored under water, therefore the research

conducted in this Ph.D. represents an alternative approach for the technique. The SINRD

measurement station can be envisaged in an encapsulation plant for the final verification of a fuel

assembly before the insertion in the storage canister for geological disposal.

The 239Pu content in spent fuel was determined with the SINRD signature, defined as the ratio

between the neutron counts in the fast energy region and in the energy region close to 0.3 eV. A 238U

fission chamber was considered for the measurement of fast neutrons, whereas two measurements

with a 239Pu fission chamber covered with SINRD filters were chosen for the measurement of the

neutron flux around 0.3 eV. The difference between the neutron counts obtained in the two

measurements was taken as estimation of the neutron flux around the 239Pu resonance. The 239Pu

fission chamber showed larger sensitivity to the 239Pu content compared to the 235U fission chamber

and proportional counters with 3He and 10B. The thicknesses of the SINRD filters were adapted to the

different detector types considering both the values of the SINRD signature and total neutron counts.

A thickness of 1.0 mm of the Cd filter was suggested for all detector types to absorb thermal

neutrons. In addition, a Gd filter of 0.1 mm thickness was chosen for the fission chambers to

110

maximize the total neutron counts, whereas a 0.2 mm Gd filter was suggested for the proportional

counters to maximize the SINRD signature. The use of proportional counters for spent fuel

measurement has to be fully investigated in view of their sensitivity to gamma-rays. The optimal

thicknesses identified in this study are significantly different from the values used in previous study

(LaFleur, 2011).

The SINRD signature calculated with Monte Carlo simulations showed a clear trend with the fuel

burnup and with the initial enrichment of spent fuel, due to the increase in the 239Pu and 235U

contents, respectively. The cooling time of the assembly did not influence the SINRD signature, since

no significant variations were calculated for the 239Pu and 235U masses. However, the initial

enrichment value must be provided by the operator to estimate the 239Pu content with the SINRD

signature. As alternative an additional measurement with a bare 239Pu fission chamber can be used

for the calculation of the RTH ratio between the fast and thermal neutron flux. The use of the SINRD

signature and the RTH ratio can lead to the identification of fuel with a certain initial enrichment and

burnup. With both approaches a calibration curve between the 239Pu content and the SINRD

signature has to be established, e.g. by measuring spent fuel assemblies with known irradiation

histories.

The SINRD sensitivity study estimated the impact of the detector and the SINRD filters on the results

obtained with the SINRD technique. The positioning of the detector did not influence the calculated

SINRD signature, and this represents an advantage of placing the detectors in the guide tubes instead

of the side of the fuel assembly. The SINRD signature was also not influenced by the detector length;

however, the amount of active material in the detector is directly related to the detector size and

must be maximized to reduce the measurement time. Slight variations in the nominal SINRD filter

thickness did not affect the SINRD signature, but an impact was found with the incomplete detector

cover by the SINRD filter due to the detection of thermal neutrons. Therefore, the complete detector

cover or a partial overlap of the SINRD filter around the detector was recommended to limit this

effect.

To support the results obtained with the Monte Carlo simulations an experimental campaign was

conducted at the GELINA Time-of-Flight facility of JRC-IRMM in Geel. Transmission measurements

showed that the quality of nuclear data used in the Monte Carlo study is sufficient for the

investigation of the SINRD technique. Moreover the experiments confirmed the results from the

optimization of the SINRD filters thicknesses, and a good agreement was reached with analytical

calculations. The SINRD benchmark measurements consisted in self-indication experiments

performed to verify the principle of the SINRD technique. The self-indication detector obtained the

111

highest sensitivity to the target material compared to the measurements with a 235U fission chamber

and a 10B ionization chamber. These results support the choice of a 239Pu fission chamber for the 239Pu

quantification in spent fuel.

The capability of SINRD for partial defect testing was investigated by calculating with Monte Carlo

simulations the detector responses of 238U and 239Pu fission chambers in the different guide tubes of

the PWR fuel assembly. The effect of the SINRD filters was evaluated for the 239Pu fission chambers

with the analytical approach used for the filter optimization. Relative differences between -30% and

+15% were calculated for the average detector responses in the central and peripheral guide tubes

compared to the reference case. In addition, the standard deviation taking into account the variation

of the detector responses in the central and peripheral guide tubes was larger in the diversion

scenarios with 50% of replaced pins compared to the reference case. By decreasing the number of

dummy pins the results approach the ones obtained in the reference case, increasing the difficulty to

spot the diversion. Considerations about the measurement uncertainty and measurement time are

needed to determine the limit for the partial defect detection.

8.2 Partial Defect Tester

The PDET technique was also investigated with Monte Carlo simulations, and the model developed

for this study considered a storage rack containing 9 fuel assemblies stored under water. Therefore,

the PDET can be used for the safeguards verifications without fuel movement from the storage

position. In all cases the detectors were supposed to be inserted in the guide tubes of the fuel

assembly in the central position in the storage rack. The application of PDET to fuel geometries

without guide tubes was not investigated in this research.

The reference conditions were defined with fuel assemblies with identical fuel composition and

source emission probabilities to assess the contributions of the individual fuel pins in a fuel assembly.

The detector responses of 235U fission chambers and 238U fission chambers were taken into account

as neutron detectors sensitive to thermal and fast neutrons, respectively. In addition the responses

of ionization chambers were calculated as gamma-ray detectors. The results for the ionization

chambers proved to be more localized than the response of the neutron detectors, due to the self-

shielding effect of the fuel pins. Several neutron and gamma-ray detector types were considered, but

no significant effect was obtained from the change in the detector type.

The effect of the spent fuel composition was evaluated and the study found that only the burnup is

responsible for a change within ±3% of the normalized neutron detector response. The variations

112

due to initial enrichment and cooling time were smaller, and for the ionization chambers were

almost in all cases within the statistical uncertainty.

The impact of the fuel assemblies in the storage rack was evaluated in the study by varying the fuel

composition and source strength of the different fuel assemblies. The burnup of the central fuel

assembly had the largest influence in case of fuel assemblies with high burnup in the lateral and

corner positions. Variations within ±10% were calculated for the guide tubes at the periphery of the

assembly. Fuel assemblies with different burnup were also placed in the lateral and corner positions

and variations between -50% and +30% were computed for the different guide tubes. In all cases the

neutron detectors were more influenced than the gamma-ray detectors due to the self-shielding

effect of the fuel pins.

Following the approach used for SINRD, the capability of PDET to detect partial defects was

evaluated. The normalized detector responses for the 235U fission chambers, 238U fission chambers,

and ionization chambers were calculated. The gamma-ray detectors showed the highest sensitivity to

the fuel diversion, and relative differences up to -30% were calculated for the detectors in the

peripheral guide tubes compared to the reference case. The relative difference in the normalized

detector responses compared to the reference case as well as the variation in the detector responses

among the guide tubes provide indications for the detection of 50% diversions from a complete fuel

assembly. As in the case of SINRD, detailed study considering the measurement uncertainty and

measurement time is needed to define the limit for the detection of partial defects.

8.3 Outlook

This Ph.D. thesis contained the results obtained with the study of the SINRD and PDET technique for

the measurement of spent fuel. Based on the results shown in this work, both techniques show

promising potential for their use during the safeguards verifications and will be further investigated.

Future work on SINRD will focus on the development and testing of a prototype for the detector. The

availability of 239Pu fission chambers will be further explored as the calculation with these detectors

showed the most promising results for the quantification of 239Pu. The use of 235U fission chambers is

envisaged as alternative. The measurement setup and thickness of the SINRD filters will follow the

results described in this thesis. The technical feasibility of spent fuel measurements in air will be also

further examined.

Additional data analysis will be conducted for the PDET detector, with a focus on the combination of

the neutron and gamma signals obtained from the different detector types. Future work will also

consider additional diversion scenarios to assess the limit of the detector for partial defect testing.

113

The design of a prototype for the PDET detector will be considered and will take into account the

results from this Ph.D. thesis. Ionization chambers will be used as gamma-ray detectors, whereas 238U

fission chambers as fast neutron detectors and 235U fission chambers as thermal neutron detectors.

For the detection of partial defects multiple neutron and gamma-ray detectors will be used in the

PDET, and the number of detectors will be based also on cost considerations.

The testing of the prototypes will be first performed in well-known radiation fields to accurately

characterize the responses of the individual detectors. Moreover, the fine tuning of the prototypes

can be carried out at this stage. Measurements in strong neutron and gamma-ray fields are then

foreseen to test the equipment under harsh radiation environment. Finally the measurement of

actual spent fuel will be planned to test the prototypes in working conditions and to evaluate the

limitations of each technique.

115

References Ancius, D., De Baere P., Schwalbach P., 2015. "Monte Carlo modelling of EURATOM neutron detectors

for operational purposes". Proceedings of the 37th ESARDA annual meeting.

Antech, 2015. "http://www.antech-inc.com/products/b2102/". Last accessed 23/05/2016.

Arit R., Fotin V., Harms N., Tarvainen M., 1995. "SFAT in physical inventory verification of spent LWR

fuel". Proceedings of the 17th ESARDA annual meeting.

Borella A., Carchon R., van der Meer K., 2010. "Experimental and Monte Carlo methods for burnup

assessment of spent fuel elements". Proceedings of the International Conference on Management of

Spent Fuel from Nuclear Power Reactors.

Borella A., Carchon R., DeLimelette C., Symens D., van der Meer K., 2011. "Spent fuel measurements

with the Fork detector at the nuclear power plant of Doel". Proceedings of the 33rd ESARDA annual

meeting.

Borella A., Rossa R., van der Meer K., 2013. "Modeling of a highly enriched 235U fission chamber for

spent fuel assay". Annals of nuclear energy 62 (2013) 224-230.

Borella A., Gad M., Rossa R., van der Meer K., 2014. "Sensitivity studies on the neutron emission of

spent nuclear fuel by means of the ORIGEN-ARP code". Proceedings of the 55th INMM annual

meeting.

Borella A., Cometto M., Rossa R., van der Meer K., Remetti R., 2015. "Extension of the SCK•CEN

reference spent fuel inventory library". Proceedings of the 37th ESARDA annual meeting.

Branger E., Grape S., Jacobsson Svärd E., Wernersson E. L. G., 2014. "Improved DCVD assessments of

irradiated nuclear fuel using image analysis techniques". Proceedings of the 2014 IAEA safeguards

symposium.

Brienne-Raepsaet C., 1999. "Nouvelle détermination expérimentale des paramètres de résonances

neutroniques de 99Tc". Ph.D. dissertation at Univesite d'Aix-Marseille I – Note CEA-N-2841.

Carchon R., et al., 1987. "ION-1 Fork measurements on pressurized water reactor spent fuel

assemblies". Proceedings of the 1986 International symposium on nuclear material safeguards -

IAEA-SM-293/73.

Carchon R., De Boeck W., Bergmans G., Bragard R., Walrave C., 1994. "Fork detector measurements

on LWR spent fuel". ESARDA Bulletin n. 23.

116

Chadwick M. B., et al., 2006. "ENDF/B-VII.0 next generation evaluated nuclear data library for science

and technology". Nuclear data sheets, volume 107, issue 12.

Chadwick M. B., et al., 2011. "ENDF/B-VII.1 nuclear data for science and technology: cross sections,

covariances, fission product yields and decay data". Nuclear data sheets, volume 112, issue 12.

Charlton W. S., Humphrey M. A., 2012. "External review of the next generation safeguards initiative’s

spent fuel nondestructive assay project". Journal of Nuclear Materials Management JNMM 40(3) 18-

23.

Chen J. D., et al., 2003. "Cherenkov characterization of BWR assemblies using a prototype DCVD with

a back-illuminated CCD". Report prepared for the Canadian safeguards support program and the

Swedish support program (CSSP Report 2004-01-01, SKI Report ISRN SKI-R-03/45).

Chen J. D., et al., 2009. "Partial defect detection in LWR spent fuel using a digital Cerenkov viewing

device". Proceedings of the 50th INMM annual meeting.

Chichester D., 2009. "Introduction to nuclear safeguards: nondestructive analysis". INL/CON-09-

16307.

Eliasson A., 2012. "Private communications".

European Commission, 2014. "Report on the implementation of Euratom safeguards in 2014".

https://ec.europa.eu/energy/sites/ener/files/documents/20151211%20Annual_Report%202014.pdf.

Last accessed 23/05/2016.

European Union, 2010. "The Euratom treaty consolidated version". ISBN 978-92-824-2554-1.

doi:10.2860/52327.

Fattah A., Khlebnikov N., 1990. “International safeguards aspects of spent-fuel disposal in permanent

geological repositories”. IAEA Bulletin 1/1990.

Fröhner F. H., Haddad E., Lopez W. M., Friesenhahn S. J., 1966. "Accuracy of Neutron Resonance

Parameters from Combined Area and Self-Indication Ratio Measurements". Proceedings of the

Conference on Neutron Cross Section Technology, P. Hemmig (ed.), CONF-660 303, Washington D.C.

(1966) Book 1, pp. 55-66.

Gauld I. C., Bowman S. M., Horwedel J. E., 2009-a. “Origen-ARP: automatic rapid processing for spent

fuel depletion, decay, and source term analysis”. ORNL/TM-2005/39.

117

Gauld I. C., Hermann O. W., Westfall R. M., 2009-b. “ORIGEN-S: SCALE system module to calculate

fuel depletion, actinide transmutation, fission product buildup and decay, and associated radiation

source terms, SCALE6 Manual”. ORNL/TM-2005/39, Version 6, Vol. II, Sect. F7.

General Electric - Reuter-Stokes, 2016. "Helium-3 Position Sensitive Detector".

https://www.gemeasurement.com/radiation-measurement/neutron-scattering/helium-3-position-

sensitive-detector. Last accessed 23/05/2016.

Goncalves J., et al., 2014. “The European Commission Cooperative Support Programme: Activities and

Cooperation”. Proceedings of the 2014 IAEA safeguards symposium.

Ham Y. S., Sitaraman S., 2009-a. “Development of a safeguards verification method and instrument to

detect pin diversion from pressurized water reactor (PWR) spent fuel assemblies phase I study”.

Lawrence Livermore National Laboratory technical report LLNL-TR-409660.

Ham Y. S., Sitaraman S., Shin S., Eom S., Kim H., 2009-b. “Experimental validation of the methodology

for partial defect verification of the pressurized water reactor spent fuel assemblies”. Proceedings of

the 50th INMM annual meeting.

Ham Y. S., Sitaraman S., Swan R., Lorenzana H., 2010. “Partial defect testing of pressurized water

reactor spent fuel assemblies”. Proceedings of the 51st INMM annual meeting.

Ham Y. S., Sitaraman S., 2011. “Shipper/receiver difference verification of spent fuel by use of PDET”.

Proceedings of the 2011 ANIMMA conference.

Ham Y. S., Sitaraman S., Kerr P., Swan R., Wong A., 2012. ”A new versatile safeguards tool for

verification of PWR spent fuel”. Proceedings of the 53rd INMM annual meeting.

Ham Y. S., Sitaraman S., Kerr P., Wong A, Swan R., 2013. “PDET: tool for partial defect verification of

spent fuel”. Presentation to NGSI-spent fuel group.

Ham Y. S., et al., 2015. “Partial defect verification of spent fuel assemblies by PDET: principle and field

testing in interim spent fuel storage facility (CLAB) in Sweden”. Proceedings of the 2015 ANIMMA

conference.

Harvey J. A., ed., 1970. “Experimental neutron resonance spectroscopy”. Academic Press, New York

and London.

Honkamaa T., Hämäläinenet M., 2003. “How STUK verifies spent nuclear fuel - and why?”

Proceedings of the 25th ESARDA annual meeting.

118

Honkamaa T., et al., 2014. “A prototype for passive gamma emission tomography”. Proceedings of

the 2014 IAEA safeguards symposium.

Hsue S. T., Crane T. W., Talbert W. L. Jr., Lee J. C., 1978. "Nondestructive assay methods for irradiated

nuclear fuels". Los Alamos National Laboratory LA-6923.

Hu J., et al., 2012. "The performance of self-interrogation neutron resonance densitometry in

measuring spent fuel". Journal of Nuclear Materials Management, Volume XL, No. 3.

International Atomic Energy Agency (IAEA), 1970. “Treaty on the non-proliferation of nuclear

weapons”. INFCIRC/140.

International Atomic Energy Agency (IAEA), 1972. “The structure and content of agreements between

the Agency and States required in connection with the treaty on the nonproliferation of nuclear

weapons”. INFCIRC/153 (corrected).

International Atomic Energy Agency (IAEA), 1989. “IAEA Statute” - as amended up to 23 February

1989.

International Atomic Energy Agency (IAEA), 1997. “Model protocol additional to the agreement(s)

between state(s) and the International Atomic Energy Agency for the application of safeguards”.

INFCIRC/540.

International Atomic Energy Agency (IAEA), 1998. “Design measures to facilitate implementation of

safeguards at future water cooled nuclear power plants”. TRS-392.

International Atomic Energy Agency (IAEA), 2009. “Safeguards criteria: special criteria for

difficult-to-access fuel items”. SG-SC-Annex-04, Version 2.

International Atomic Energy Agency (IAEA), 2011. “Safeguards techniques and equipment: 2011

edition”. International nuclear verification series no. 1 (rev. 2).

International Atomic Energy Agency (IAEA), 2012-a. “Storage of spent nuclear fuel”. IAEA safety

standards series No. SSG-15.

International Atomic Energy Agency (IAEA), 2012-b. “Spent fuel reprocessing options”.

IAEA-TECDOC-1587.

International Atomic Energy Agency (IAEA), 2013. “IAEA department of safeguards long term R&D

plan, 2012-2023”. STR-375.

119

International Atomic Energy Agency (IAEA), 2014. “IAEA annual report for 2013”. IAEAGC(58)/3.

Janssen-Maenhout G., editor, 2008. “Nuclear safeguards and non-proliferation”. ESARDA working

group on training and knowledge management.

Jansson P., Andresson P., White T. A., Mozin V., 2015. “Monte Carlo simulations of a universal

gamma-ray emission tomography device”. Proceedings of the 37th ESARDA annual meeting.

Knoll G. F., 2010. "Radiation Detection and Measurement – Fourth edition". John Wiley & Sons, Inc.

Koning A. J., et al., 2010. "Status of the JEFF Nuclear Data Library". Proceedings of the International

Conference on Nuclear Data for Science and Technology, 2010, pp.1057-1062.

LaFleur A. M., 2011. “Development of self-interrogation neutron resonance densitometry (SINRD) to

measure the fissile content in nuclear fuel”. Ph.D. dissertation at Texas A&M University.

LaFleur A. M., Charlton W. S., Menlove H. O., Swinhoe M. T., 2012. “Comparison of fresh fuel

experimental measurements to MCNPX calculations using self-interrogation neutron resonance

densitometry”. Nuclear Instruments and Methods Section A, 680 (2012) 168-178.

LaFleur A. M., Charlton W. S., Menlove H. O., Swinhoe M. T., Lebrun A. R., 2013. “Development of

self-interrogation neutron resonance densitometry to improve detection of partial defects in PWR

spent fuel assemblies”. Nuclear Technology, 181, (2013) 354-370.

LaFleur A. M., Menlove H. O., 2015. “Analysis of experimental measurements of PWR fresh and spent

fuel assemblies using self-interrogation neutron resonance densitometry”. Nuclear Instruments and

Methods Section A, 781 (2015) 86-95.

Lux I., Koblinger L., 1991. "Monte Carlo particle transport methods: neutron and photon calculations".

Library of Congress ISBN 0-8493-6074-9. CRC Press, Inc.

Massimi C., et al., 2011. “Neutron resonance parameters of 197Au from transmission, capture and

self-indication measurements”. Journal of the Korean Physical Society 59, 1689 - 1692.

MathWorks, Inc., 2015. ”MATLAB ® Primer”.

Menlove H. O., Tesche C. D., Thorpe M. M., Walton R. B., 1969. “A resonance self-indication

technique for isotopic assay of fissile material”. Nuclear applications, volume 6, 401-408.

Mondelaers W., Schillebeeckx P., 2006. "GELINA, a neutron time-of-flight facility for high-resolution

neutron data measurements". Notiziario Neutroni e Luce di Sincrotrone, II(2), 19-25 (2006).

120

Nyström A., 2007. "Encapsulation plant at Forsmark". SKB Rapport R-07-36.

Parcey D. A., et al., 2011. “Detection of partial defects in LWR spent fuel using a DCVD”. Proceedings

of the 33rd ESARDA annual meeting.

Park W. S., et al., 2014. “Safeguards by design at the encapsulation plant in Finland”. Proceedings of

the 2014 IAEA safeguards symposium.

Pelowitz D., editor, 2011. “MCNPX user’s manual version 2.7.0”. Los Alamos National Laboratory

LA-CP-11-00438.

Phillips J. R., 1991. “Irradiated fuel measurements”. Passive nondestructive assay of nuclear

materials, LA-UR-90-732.

Photonis-Gamma detector,2016. ”https://www.photonis.com/uploads/datasheet/ngd/CRGE10.pdf”.

Last accessed 23/05/2016.

Proportional Technologies, Inc., 2016. "Sealed Boron-Coated Straws for Neutron Imaging".

http://www.proportionaltech.com/new_site/images/stories/Sealed_7straw_Imaging_Detector_7_30

_12.pdf. Last accessed 23/05/2016.

Putz F., Al-Khafaji T, Weinzierl P., 1966. “Determination of plutonium-239 content in nuclear fuel by

neutron transmission technique”. Proceedings of the symposium on nuclear materials management,

p. 809-816.

Oak Ridge National Laboratory (ORNL), 2009. “SCALE: a modular code system for performing

standardized computer analyses for licensing evaluation”. ORNL/TM-2005/39.

Rhodes R., 1986. “The making of the atomic bomb”. Simon & Schuster publishing.

Rinard P. M., 1984. "Neutron measurements in borated water for PWR fuel inspections". Los Alamos

National Laboratory report LA-10068-MS.

Rinard P. M., 1986. "BWR spent-fuel measurements with the ION-l/Fork detector and a calorimeter".

Los Alamos National Laboratory report LA-10758-MS.

Rinard P. M., Bosler G. E., 1988. "Safeguarding LWR spent fuel with the FORK detector". Los Alamos

National Laboratory report LA-11096-MS.

121

Rossa R., Peerani P., Ham Y. S., Sitaraman S., 2013-a. "Application of the PDET detector to BWR fuel

assemblies: gross defect testing using the spatial distribution of neutron and photon flux".


Rossa R., Borella A., van der Meer K., 2013-b. "Development of a reference spent fuel library of 17x17

PWR fuel assemblies". ESARDA Bulletin n. 50.

Rossa R., Borella A., van der Meer K., Labeau P. E., Pauly N., 2014. "Optimization of the filters

thickness for the SINRD technique applied to spent fuel verification". Proceedings of the 55th INMM

annual meeting.

Rossa R., Borella A., van der Meer K., 2015-a. "Investigation of the self-interrogation neutron

resonance densitometry applied to spent fuel using Monte Carlo simulations". Annals of Nuclear

Energy 75 (2015) 176-183.

Rossa R., Borella A., Labeau P. E., Pauly N., van der Meer K., 2015-b. "Neutron absorbers and detector

types for spent fuel verification using the self-interrogation neutron resonance densitometry". Nuclear

Instruments and Methods in Physics Research A 791 (2015) 93–100.

Rossa R., Borella A., Labeau P. E., Pauly N., van der Meer K., 2015-c. "Influence of fuel composition on

the spent fuel verification by Self-Indication Neutron Resonance Densitometry". ESARDA Bulletin n.

52.

Rossa R., Borella A., Labeau P. E., Pauly N., van der Meer K., 2015-d. "Sensitivity study of the Self-

Interrogation Neutron Resonance Densitometry for the safeguards verification of spent fuel".

Proceedings of the 56th INMM annual meeting.

Schillebeeckx P., et al., 2012. "Determination of Resonance Parameters and their Covariances from

Neutron Induced Reaction Cross Section Data". Nuclear Data Sheets 113, 3054-3100.

Schillebeeckx P., Becker B., Harada H., Kopecky S., 2014. "Neutron resonance spectroscopy for the

characterization of materials and objects". JRC Science and Policy Reports - Report EUR 26848 EN.

Schwalbach P., et al., 2014. "Euratom safeguards: improving safeguards by cooperation in R&D and

implementation". Proceedings of the 2014 IAEA safeguards symposium.

Shibata K., et al., 2011. "JENDL-4.0: A New Library for Nuclear Science and Engineering". Journal of

Nuclear Science and Technology 48(1), 1-30.

122

Sitaraman S., Ham Y. S., 2007. "Characterization of a safeguards verification methodology to detect

pin diversion from pressurized water reactor (PWR) spent fuel assemblies using Monte Carlo

techniques". Proceedings of the 48th INMM annual meeting.

Sitaraman S., Ham Y. S., 2009. "Symmetric pin diversion detection using a partial defect detector

(PDET)". Proceedings of the 50th INMM annual meeting.

Sitaraman S., Ham Y. S., 2010. "Signatures for different PWR assemblies using partial defect tester

(PDET)". Proceedings of the 51st INMM annual meeting.

Soppera N., Bossant M., Dupont E., 2014. "JANIS 4: an improved version of the NEA Java-based

nuclear data information system". Nuclear Data Sheets 120, 294–296.

Spanier J., Gelbard E. M. , 1969. "Monte Carlo principles and neutron transport problems". Addison

Wesley, Reading, MA.

Stankovskiy A., van den Eynde G., Vidmar T., 2010. "Development and validation of ALEPH Monte

Carlo burn-up code". Nuclear Measurements, Evaluations and Applications - NEMEA-6.

Stankovskiy A., van den Eynde G., 2012. "Advanced method for calculations of core burn-up,

activation of structural materials, and spallation products accumulation in accelerator-driven

systems". Science and Technology of Nuclear Installations, 2012:1(2012), p. 1-12.- ISSN 1687-6075.

Thorne P. R., O'Connor G. J., Bowden R. L., 2002. "Problem specification for the OECD/NEANSC

burnup credit benchmark phase IV-B: mixed oxide (MOX) fuels". BNFL, Risley, Warrington, Cheshire,

U.K.

Tiitta A., et al., 2002. "Investigation on the possibility to use Fork detector for partial defect

verification of spent LWR fuel assemblies". Final report on Task JNT A 1071 (BEL, FIN, SWE) of the

Members States’ support programme to IAEA safeguards.

Tobin S. J., Menlove H. O., Swinhoe M. T., Schear M. A., 2011. "Next generation safeguards initiative

research to determine the Pu mass in spent fuel assemblies: purpose, approach, constraints,

implementation, and calibration". Nuclear Instruments and Methods in Physics Research A 652

(2011) 73-75.

Tobin S. J., Jansson P., 2013. "Nondestructive assay options for spent fuel encapsulation". SKB

technical report TR-13-30.

123

Tobin S. J., et al., 2014. "Experimental and analytical plans for the non-destructive assay system of the

Swedish encapsulation and repository facilities". Proceedings of the 2014 IAEA safeguards

symposium.

Vaccaro S., Gauld I. C., Hu J., De Baere P., Schwalbach P., 2014. "Uncertainty quantification of fork

detector measurements from spent fuel loading campaigns". Proceedings of the 2014 IAEA

safeguards symposium.

Van der Meer K., Coeck M., 2005. "An assessment of the FORK detector as a partial defect tester".


Weber A.L., 2007. "Monte Carlo simulation applied to non-destructive assay techniques". ESARDA

Bulletin n. 37.

Williams III R. G., Gesh C. J., Pagh R. T., 2006. "Compendium of material composition data for

radiation transport modeling". PNNL-15870.

125

Annex A. Further considerations on the influence of individual

nuclides on the SINRD signature

A.1. Results with a 235U fission chamber

The influence of individual nuclides was estimated in this section for a 235U fission chamber following

the approach used for a 239Pu fission chamber in Section 4.5.1, and Figure A.1 shows the results. The

impact of SINRD filters of 0.1 mm Gd and 1.0 mm Cd was considered with the analytical approach

described in Chapter 4.

The SINRD signature calculated for fuel containing the 50 main neutron absorbers increases with the

239Pu content in the fuel. However, the values calculated for a 235U fission chamber are significantly

lower compared to the results obtained with a 239Pu fission chamber. For both detectors 239Pu and

235U are the nuclides with major influences on the SINRD signature, but several nuclides show

significant effects for fuel with high burnup.

Figure A.1: SINRD signature as a function of the

239Pu content for fuel containing different nuclides. Results with a

235U

fission chamber.

The percentage contribution of the individual nuclides on the SINRD signature is shown in Table A.1

for two burnup values. In case of fuel with burnup of 10 GWd/tU 239Pu and 235U account for almost

90% of the SINRD signature value and the other nuclides have limited contributions. The impact of

235U on the SINRD signature is larger than in the results obtained with the 239Pu fission chamber in

Chapter 4, and this is due to the self-indication effect on 235U. In fact, by using a 235U fission chamber

the sensitivity to the 235U content is increased and this results can be taken into account for future

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11

Detector: 235

U fission chamber

Filters: Gd 0.1 mm

Cd 1.0 mm

40/60 GWd/tU

20 GWd/tU15 GWd/t

U10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

239Pu +

16O [1]

[1] + 235

U [2]

[2] + 241

Pu [3]

[3] + 240

Pu [4]

[4] + 241

Am [5]

50 nuclides

126

applications. The results calculated for fuel with burnup of 60 GWd/tU show that 239Pu has the major

impact to the SINRD signature, and several other nuclides give significant contributions.

Table A.1: Contribution of individual nuclides on the SINRD signature. The values have a statistical uncertainty lower than 0.1%.

Fuel burnup

10 GWd/tU 60 GWd/tU 239Pu 32.4 % 45.2 % 235U 53.9 % 6.3 %

241Pu 1.5 % 7.6 % 240Pu 4.7 % 13.1 %

241Am 0.4 % 5.3 %

The influence of the fuel irradiation history on the SINRD signature was analyzed by including the 50

main neutron absorbers for fuel with different initial enrichment, burnup, and cooling time. The

results are shown in Figure A.2 and resemble the trend obtained for a 239Pu fission chamber. The

SINRD signature increases with the fuel burnup due to the increase of the 239Pu content and with the

initial enrichment due to the 235U mass. No significant influence was observed with the cooling time

of the assembly.


239Pu content for fuel with different irradiation history. Results with a

235U

fission chamber.

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11Detector:


Filters: Gd 0.1 mm

Cd 1.0 mm

40/60 GWd/tU15 GWd/t

U

20 GWd/tU

10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

IE: 3.5%

IE: 4.0%

IE: 4.5%

IE: 5.0%

127

A.2. Results with a 3He proportional counter

The approach described in Section 4.5.1 was also applied to a 3He proportional counter to evaluate

the influence of individual nuclides on the SINRD signature. The impact of SINRD filters of 0.2 mm Gd

and 1.0 mm Cd was considered for both 3He and 10B proportional counters according to the analytical

approach described in Chapter 4.

The SINRD signature calculated for spent fuel with different material composition is shown in Figure

A.3, and the results are very similar to the values obtained for a 235U fission chamber. As in the

previous section, the percentage contributions on the SINRD signature was calculated for several

nuclides and the results are shown in Table A.2 for different burnup values. The values calculated for

fuel with burnup of 10 GWd/tU confirm that 239Pu and 235U are responsible for almost 90% of the

SINRD signature value. Both nuclides have a similar percentage contribution due to the lack of the

self-indication effect. The results for the high burnup fuel show that 239Pu gives the major

contribution to the SINRD signature but, as for the fission chambers considered in the previous

sections, other nuclides have a significant impact.

The influence of the irradiation history on the SINRD signature is reported in Figure A.4, and similar

results to Figure A.2 were obtained. Therefore, the comments made in the previous section apply

also for the 3He proportional counter.



3He

proportional counter.

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11

Detector: 3He proportional counter

Filters: Gd 0.2 mm

Cd 1.0 mm

40/60 GWd/tU

20 GWd/tU15 GWd/t

U10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

239Pu +

16O [1]

[1] + 235

U [2]

[2] + 241

Pu [3]

[3] + 240

Pu [4]

[4] + 241

Am [5]

50 nuclides

128


Fuel burnup

10 GWd/tU 60 GWd/tU 239Pu 37.4 % 44.3 % 235U 46.6 % 4.8 %

241Pu 1.4 % 5.3 % 240Pu 7.5 % 19.1 %

241Am 0.7 % 7.7 %



3He


A.3. Results with a 10B proportional counter

The effect of individual nuclides on the SINRD signature was evaluated also for a 10B proportional

counter and the results are shown in Figure A.5. Following the approach applied for the other

detector types, the percentage contribution of the individual nuclides on the SINRD signature was

computed and the results are shown in Table A.3. Moreover, the influence of the fuel irradiation

history was calculated and the results are plotted in Figure A.6. The values are identical to the results

obtained for the 3He proportional counter; therefore, the comments made in the previous section

are valid also for this detector type.

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11Detector:


Filters: Gd 0.2 mm

Cd 1.0 mm

40/60 GWd/tU

15 GWd/tU

20 GWd/tU

10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

IE: 3.5%

IE: 4.0%

IE: 4.5%

IE: 5.0%

129



10B



Fuel burnup 10 GWd/tU 60 GWd/tU

239Pu 37.4 % 44.3 % 235U 46.6 % 4.8 %

241Pu 1.4 % 5.3 % 240Pu 7.5 % 19.1 %

241Am 0.7 % 7.7 %



10B


0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11

Detector: 10

B proportional counter

Filters: Gd 0.2 mm

Cd 1.0 mm

40/60 GWd/tU

20 GWd/tU15 GWd/t

U10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

239Pu +

16O [1]

[1] + 235

U [2]

[2] + 241

Pu [3]

[3] + 240

Pu [4]

[4] + 241

Am [5]

50 nuclides

0 1 2 3 4 5 6 70

1

2

3

4

5

6

7

8

9

10

11Detector:

10B proportional counter

Filters: Gd 0.2 mm

Cd 1.0 mm

40/60 GWd/tU

15 GWd/tU

20 GWd/tU

10 GWd/tU

5 GWd/tU

Fresh fuel

RS

I

239Pu content (kg/t

U)

IE: 3.5%

IE: 4.0%

IE: 4.5%

IE: 5.0%

131

Annex B. Additional diversion scenarios for the PDET detector

B.1. Description of the diversion scenarios

The diversion scenarios developed in Chapter 7 considered the replacement of fuel pins with

symmetric patterns, whereas in this annex the asymmetric diversion is taken into account. Figure B.1

shows the scenarios considered in this section. As for the previous scenarios the dummy pins had the

same dimensions of the spent fuel pins and were made of stainless steel. Spent fuel with 3.5% initial

enrichment, burnup of 30 GWd/tU, and cooling time of 5 years was modelled as material composition

and source term.

The asymmetric diversion was considered for the PDET detector since it can be masked by placing

fuel assemblies with different burnup in the storage rack. The same cases were not simulated for the

SINRD technique because only one fuel assembly was included in the model for this technique, and

the symmetric diversion was considered as the limiting case for the detection.

Figure B.1: Visualization of the asymmetric diversion scenarios developed for the PDET detector. The fuel pins are depicted in white, the dummy pins in grey, and the guide tubes in yellow.



132

B.2. Results for the 235U fission chambers

The normalized detector response was calculated for the different guide tubes and Figures B.2 and

B.3 show the results for the 235U fission chambers. The plots show the relative difference DK,R

calculated according to Formula 6.1 in Chapter 6.

The diversion scenarios with 50% of replaced fuel pins had relative differences up to 20% from the

reference case. In most cases the replacement of pins leads to a decrease of the normalized detector

response in the guide tubes close to the dummy pins, and a consequent increase of the normalized

detector response in the other guide tubes. Therefore, the maximum positive differences in general

were calculated for the guide tubes far from the replaced pins. For diversion scenarios 23, 25, and 28

the positive differences compared to the reference case were obtained for the guide tubes close to

the dummy pins. The relative difference compared to the reference case increases with the amount

of replaced fuel pins.

Figure B.2: Relative difference for the normalized detector responses in the diversion scenarios compared to the reference case. The dummy pins are depicted in grey and the fuel pins are shown in white. The results refer to

235U fission chambers.

133



The diversion scenario 21 resembles the results obtained for a storage rack with high-burnup in the

lateral position close to the guide tubes 23, 24, and 25, whereas the DK,R ratios for scenario 26 are

similar to the case with a corner fuel assembly with high burnup close to the guide tube 21. The

diversion scenarios 23 and 25 had results that cannot be obtained by placing fuel assemblies with

different burnup in the storage rack, whereas the other diversion scenarios with 25% of dummy pins

showed small differences compared to the reference case. Given the results in this section, the

normalized detector responses with other detector types were calculated to assess the capability to

detect these diversion scenarios.

B.3. Results for the 238U fission chambers

The relative difference DK,R was calculated also for the 238U fission chambers and Figures B.4 and B.5

show the results. Negative values were obtained for the guide tubes close to the dummy pins in all

diversion scenarios for this detector type.

The relative difference for the different guide tubes ranged between +20 and -70% compared to the

reference case, and similar values were obtained for scenarios with 25% or 50% of dummy pins. The

variation of the normalized detector responses compared to the reference case can be used to

detect the asymmetric diversion in all scenarios considered in this study.

134

The comparison of the results obtained with the 235U and 238U fission chambers gives an additional

indication of the pins replacement. Similar results were obtained in Chapter 6 by comparing the

detector responses of the two detector types for a storage rack with different fuel assembly,

whereas the results shown in Figures B.2 – B.5 show a significant difference between the two

detector types.

The difference in the results obtained for the two neutron detector types can be explained with the

importance functions calculated in Chapter 6. The detector response of the 238U fission chambers

were due mainly to the fuel pins close to the guide tube containing the detector, whereas the whole

fuel assembly contributed in a significant way to the detector response of the 235U fission chambers.

The fuel diversion influences all guide tubes in the case of 235U fission chamber; the combination of

the remaining pins with the neighboring fuel assemblies can lead to an increase of the normalized

detector response in the guide tubes close to the dummy pins in some diversion scenarios.



135



B.4. Results for the ionization chambers

The DK,R ratio was calculated in the case of ionization chambers for all asymmetric diversion scenarios

and the results are shown in Figures B.6 and B.7.

As for the 238U fission chambers negative differences for the normalized detector responses

compared to the reference case were obtained for the guide tubes close to the dummy pins.

Differences between -50% and -90% were calculated, whereas the results for the guide tubes close to

the spent fuel pins were similar to the reference case. Considering the strong shielding effect of the

fuel pins, the guide tubes far from the dummy pins are almost not affected by the diversion. Similar

values were obtained for diversion scenarios with 25% or 50% of replaced pins.

The significant variations of the detector responses compared to the reference case can be used for

the detection of the fuel pins diversion, and support the indications given by the neutron detector

types considered in this study.

136

Figure B.6: Relative difference for the normalized detector responses in the diversion scenarios compared to the reference case. The dummy pins are depicted in grey and the fuel pins are shown in white. The results refer to ionization chambers.

Figure B.7: Relative difference for the normalized detector responses in the diversion scenarios compared to the reference case. The dummy pins are depicted in grey and the fuel pins are shown in white. The results refer to ionization chambers.

Advanced non-destructive methods for criticality safety ... · Advanced non-destructive methods for...

Documents

Transcript of Advanced non-destructive methods for criticality safety ... · Advanced non-destructive methods for...