Validation of a Machine Learning–Based Nuclear Forensics Methodology for the Discrimination of a Chemically Separated Plutonium Sample from Low-Enriched Uranium

Patrick J. O’Neal; Sean P. Martinson; Sunil S. Chirayath

doi:10.1080/00295639.2023.2271711

Validation of a Machine Learning–Based Nuclear Forensics Methodology for the Discrimination of a Chemically Separated Plutonium Sample from Low-Enriched Uranium

Patrick J. O’Neal, Sean P. Martinson, Sunil S. Chirayath

Research output: Contribution to journal › Article › peer-review

Abstract

When the foundation of a method is simulated data, it is paramount that the method is validated with data from physical samples when possible. This study presents the results of validating a recently developed nuclear forensics methodology with a new low-burnup plutonium sample, chemically separated from low-enriched uranium irradiated in thermal neutron flux. The nuclear forensics methodology uses machine learning models to discriminate the reactor type of origin, fuel burnup, and time since irradiation (TSI) of chemically separated plutonium samples. The machine learning models use intra-elemental isotope ratios of cesium, samarium, europium, and plutonium as features; the isotopic ratio data for training the models were generated through fuel burnup simulations of various nuclear reactor types. The methodology predicted the reactor type and fuel burnup of the plutonium sample successfully. Initial difficulties to predict the TSI were resolved with the inclusion of a new intra-elemental isotope ratio of cerium.

Original language	English
Journal	Nuclear Science and Engineering
DOIs	https://doi.org/10.1080/00295639.2023.2271711
State	Accepted/In press - 2023
Externally published	Yes

Funding

This work performed was funded by the Consortium for Monitoring, Technology, and Verification under the U.S. Department of Energy (DOE) National Nuclear Security Administration (NNSA) award number DE-NA0003920. The opinions expressed in this paper are the authors’ own and do not necessarily state or reflect the view of the NNSA, the DOE, or the U.S. government or any agency thereof. Neither the U.S. government nor any agency thereof, nor any of their employees, makes any warranty, express or limited, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S. government or any agency thereof.

Funders	Funder number
U.S. Department of Energy
National Nuclear Security Administration	DE-NA0003920

Keywords

Plutonium forensics
fuel burnup
machine learning
reactor-type
time since irradiation

Access to Document

10.1080/00295639.2023.2271711

Cite this

@article{150300b319d8428b939ac23ab41e7703,

title = "Validation of a Machine Learning–Based Nuclear Forensics Methodology for the Discrimination of a Chemically Separated Plutonium Sample from Low-Enriched Uranium",

abstract = "When the foundation of a method is simulated data, it is paramount that the method is validated with data from physical samples when possible. This study presents the results of validating a recently developed nuclear forensics methodology with a new low-burnup plutonium sample, chemically separated from low-enriched uranium irradiated in thermal neutron flux. The nuclear forensics methodology uses machine learning models to discriminate the reactor type of origin, fuel burnup, and time since irradiation (TSI) of chemically separated plutonium samples. The machine learning models use intra-elemental isotope ratios of cesium, samarium, europium, and plutonium as features; the isotopic ratio data for training the models were generated through fuel burnup simulations of various nuclear reactor types. The methodology predicted the reactor type and fuel burnup of the plutonium sample successfully. Initial difficulties to predict the TSI were resolved with the inclusion of a new intra-elemental isotope ratio of cerium.",

keywords = "Plutonium forensics, fuel burnup, machine learning, reactor-type, time since irradiation",

author = "O{\textquoteright}Neal, {Patrick J.} and Martinson, {Sean P.} and Chirayath, {Sunil S.}",

note = "Publisher Copyright: {\textcopyright} 2023 American Nuclear Society.",

year = "2023",

doi = "10.1080/00295639.2023.2271711",

language = "English",

journal = "Nuclear Science and Engineering",

issn = "0029-5639",

publisher = "Taylor and Francis Group",

}

TY - JOUR

T1 - Validation of a Machine Learning–Based Nuclear Forensics Methodology for the Discrimination of a Chemically Separated Plutonium Sample from Low-Enriched Uranium

AU - O’Neal, Patrick J.

AU - Martinson, Sean P.

AU - Chirayath, Sunil S.

PY - 2023

Y1 - 2023

N2 - When the foundation of a method is simulated data, it is paramount that the method is validated with data from physical samples when possible. This study presents the results of validating a recently developed nuclear forensics methodology with a new low-burnup plutonium sample, chemically separated from low-enriched uranium irradiated in thermal neutron flux. The nuclear forensics methodology uses machine learning models to discriminate the reactor type of origin, fuel burnup, and time since irradiation (TSI) of chemically separated plutonium samples. The machine learning models use intra-elemental isotope ratios of cesium, samarium, europium, and plutonium as features; the isotopic ratio data for training the models were generated through fuel burnup simulations of various nuclear reactor types. The methodology predicted the reactor type and fuel burnup of the plutonium sample successfully. Initial difficulties to predict the TSI were resolved with the inclusion of a new intra-elemental isotope ratio of cerium.

AB - When the foundation of a method is simulated data, it is paramount that the method is validated with data from physical samples when possible. This study presents the results of validating a recently developed nuclear forensics methodology with a new low-burnup plutonium sample, chemically separated from low-enriched uranium irradiated in thermal neutron flux. The nuclear forensics methodology uses machine learning models to discriminate the reactor type of origin, fuel burnup, and time since irradiation (TSI) of chemically separated plutonium samples. The machine learning models use intra-elemental isotope ratios of cesium, samarium, europium, and plutonium as features; the isotopic ratio data for training the models were generated through fuel burnup simulations of various nuclear reactor types. The methodology predicted the reactor type and fuel burnup of the plutonium sample successfully. Initial difficulties to predict the TSI were resolved with the inclusion of a new intra-elemental isotope ratio of cerium.

KW - Plutonium forensics

KW - fuel burnup

KW - machine learning

KW - reactor-type

KW - time since irradiation

UR - http://www.scopus.com/inward/record.url?scp=85176955280&partnerID=8YFLogxK

U2 - 10.1080/00295639.2023.2271711

DO - 10.1080/00295639.2023.2271711

M3 - Article

AN - SCOPUS:85176955280

SN - 0029-5639

JO - Nuclear Science and Engineering

JF - Nuclear Science and Engineering

ER -

Validation of a Machine Learning–Based Nuclear Forensics Methodology for the Discrimination of a Chemically Separated Plutonium Sample from Low-Enriched Uranium

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this