Solution monitoring as a nuclear materials safeguards tool
The work presented in this thesis describes a solution monitoring system that has been developed to assist United Nations' inspectors performing nuclear materials safeguards, primarily pertaining to plutonium storage and nuclear fuel reprocessing facilities. Based on the concept of the 'ev...
Main Author: | |
---|---|
Published: |
University of Glasgow
1998
|
Subjects: | |
Online Access: | http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.266186 |
id |
ndltd-bl.uk-oai-ethos.bl.uk-266186 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-bl.uk-oai-ethos.bl.uk-2661862015-03-19T03:40:51ZSolution monitoring as a nuclear materials safeguards toolScothern, Stephen John1998The work presented in this thesis describes a solution monitoring system that has been developed to assist United Nations' inspectors performing nuclear materials safeguards, primarily pertaining to plutonium storage and nuclear fuel reprocessing facilities. Based on the concept of the 'event', which is essentially any process that occurs on the plant, it aims to construct a hypothesis of which events have actually occurred, and to decide if any of these have safeguards implications. The package developed is robust, portable, and easy to use. The system has been implemented in G2 with extensive call-outs to FORTRAN and C routines. Sensor data from the plant is first analysed, and salient features (sub-events) are extracted. A model based diagnostic algorithm is then used to determine all possible causes of these sub-events; based on topographical knowledge of the plant, this makes extensive use of a simulation model. A rule based system then examines permutations of these sub-events and diagnoses, to find all possible events which could explain the data. From the possibilities generated, the most likely events are chosen on the basis of user specified subjective probabilities and on supporting evidence; these probabilities reflect the view that some events are more likely to be acceptable to the operator than others. Bayesian evidential updating methods are used to achieve this. An automatic model generator is presented, which extends the portability and applicability of the diagnostic aid, and makes implementation a great deal easier. Amongst other things, this enables simulations to be constructed automatically using a library of unit process models. The nature and forms of the various user interfaces are discussed. In particular facilities are available for creating and maintaining databases of rules which are used to identify, classify and rank the events. The system has been tested using data from a number of plants, both hypothetical and real. The primary test facilities have pertained to plutonium nitrate solution storage areas. A hypothetical solvent-extraction and concentration facility has also been considered, to extend the range of applicability of the system. These studies have demonstrated that solution monitoring has the potential to be a valuable aid for inspectors responsible for nuclear materials safeguards.005T Technology (General)University of Glasgowhttp://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.266186http://theses.gla.ac.uk/4523/Electronic Thesis or Dissertation |
collection |
NDLTD |
sources |
NDLTD |
topic |
005 T Technology (General) |
spellingShingle |
005 T Technology (General) Scothern, Stephen John Solution monitoring as a nuclear materials safeguards tool |
description |
The work presented in this thesis describes a solution monitoring system that has been developed to assist United Nations' inspectors performing nuclear materials safeguards, primarily pertaining to plutonium storage and nuclear fuel reprocessing facilities. Based on the concept of the 'event', which is essentially any process that occurs on the plant, it aims to construct a hypothesis of which events have actually occurred, and to decide if any of these have safeguards implications. The package developed is robust, portable, and easy to use. The system has been implemented in G2 with extensive call-outs to FORTRAN and C routines. Sensor data from the plant is first analysed, and salient features (sub-events) are extracted. A model based diagnostic algorithm is then used to determine all possible causes of these sub-events; based on topographical knowledge of the plant, this makes extensive use of a simulation model. A rule based system then examines permutations of these sub-events and diagnoses, to find all possible events which could explain the data. From the possibilities generated, the most likely events are chosen on the basis of user specified subjective probabilities and on supporting evidence; these probabilities reflect the view that some events are more likely to be acceptable to the operator than others. Bayesian evidential updating methods are used to achieve this. An automatic model generator is presented, which extends the portability and applicability of the diagnostic aid, and makes implementation a great deal easier. Amongst other things, this enables simulations to be constructed automatically using a library of unit process models. The nature and forms of the various user interfaces are discussed. In particular facilities are available for creating and maintaining databases of rules which are used to identify, classify and rank the events. The system has been tested using data from a number of plants, both hypothetical and real. The primary test facilities have pertained to plutonium nitrate solution storage areas. A hypothetical solvent-extraction and concentration facility has also been considered, to extend the range of applicability of the system. These studies have demonstrated that solution monitoring has the potential to be a valuable aid for inspectors responsible for nuclear materials safeguards. |
author |
Scothern, Stephen John |
author_facet |
Scothern, Stephen John |
author_sort |
Scothern, Stephen John |
title |
Solution monitoring as a nuclear materials safeguards tool |
title_short |
Solution monitoring as a nuclear materials safeguards tool |
title_full |
Solution monitoring as a nuclear materials safeguards tool |
title_fullStr |
Solution monitoring as a nuclear materials safeguards tool |
title_full_unstemmed |
Solution monitoring as a nuclear materials safeguards tool |
title_sort |
solution monitoring as a nuclear materials safeguards tool |
publisher |
University of Glasgow |
publishDate |
1998 |
url |
http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.266186 |
work_keys_str_mv |
AT scothernstephenjohn solutionmonitoringasanuclearmaterialssafeguardstool |
_version_ |
1716733929469247488 |