Methods for the radiological characterisation of the FiR 1 TRIGA research reactor decommissioning waste

Ackermann, Tarryn (2017-03)

Thesis (MSc)--Stellenbosch University, 2017

Thesis

ENGLISH ABSTRACT : Every nuclear reactor will eventually reach the end of its life time. With more than 60 new power reactors currently under construction, the need for understanding the safe decommissioning of a nuclear reactor is as important as ever. Among those reactors at the end of their life time is the VTT owned, FiR 1 TRIGA Mark II Research Reactor in Espoo, Finland. The FiR 1 decommissioning project is reviewed as a whole, as well as the requirements set by regulatory authorities relating to decommissioning waste and the environment. Special focus is placed on full site characterisation in terms of activity measurements. During this part of the procedure, an activity inventory of all activated components and structures is determined. The activity inventory has been determined computationally through the use of MCNP and ORIGEN-S, and this is explained on a general level. However, the results thereof require validation through physical measurements. These physical measurements are performed using γ spectroscopy for three different types of samples (concrete, Fluental and aluminium) using the In Situ Object Counting System (ISOCSTM). In addition, 13 barrels containing ion-exchange resin are characterised and compared with accepted release levels. Measurements are performed and interpreted using GenieTM 2000 software, to determine activity concentrations for validation of computationally determined activities. Algorithms used in the Genie 2000 software are outlined to better understand how the activity concentrations are determined. For the concrete sample, measured and predicted activities for 60Co and 152Eu agree with an uncertainty of 11% and 2%, respectively. While unsystematic differences are initially found for the cases of Fluental and aluminium, it is understood that the most probable cause is the assumed material compositions in the calculations. After ICP-MS leading to more accurate knowledge of the Fluental and aluminium sample compositions, there is much better agreement between computational and measured results. The differences are still large, but systematic. The calculated values are consistently one order of magnitude larger than the measured values, except for 59Fe in the aluminium sample, for which the values agree to the same order of magnitude. The 13 γ spectra for ion-exchange resin barrels all look similar, with some barrels more active than others, as would be expected for an inhomogeneous sample. In every barrel, 60Co and 137Cs is found, with the 137Cs suspected to come from a leaking fuel rod. Although much time and effort was spent on the analysis of a small number of samples relative to the number of samples in the full reactor decommissioning, it is a necessary exercise to learn lessons that may save time in the future. Most important is the study of the methods of measurement and calculation that will be applied later during full decommissioning

AFRIKAANSE OPSOMMING : Elke kernreaktor bereik eventueel die einde van sy leeftyd. Met meer as 60 nuwe kernkrag reaktore tans onder konstruksie, is die behoefte om die veilige ontmanteling van 'n kernreaktor te verstaan belangriker as ooit. Een van hierdie reaktore wat aan die einde van sy leeftyd gekom het, is die FiR 1 TRIGA Merk II Navorsingsreaktor van VTT in Espoo, Finland. Die FiR 1 ontmantelingsprojek word as geheel nagegaan, asook die vereistes wat deur die regulerende owerhede ingestel is wat betrekking het op ontmantelingsafval en die omgewing. Spesiale fokus word geplaas op heel-terrein karakterisering in terme van aktiwiteitsmetings. Tydens hierdie deel van die prosedure word 'n aktiwiteitsopgawe van alle geaktiveerde komponente en strukture bepaal. Die aktiwiteitsopgawe is bereken deur van MCNP en ORIGEN-S gebruik te maak. Dit word in die algemeen verduidelik. Die resultate hiervan vereis egter bevestiging deur fisiese metings. Hierdie fisiese metings word deur middel van γ spektroskopie uitgevoer vir drie verskillende soorte monsters (beton, Fluental en aluminium), deur gebruik te maak van die In Situ Voorwerp Tel Sisteem (ISVTSTM). Daarbenewens word 13 dromme, bevattende ioon-uitruil hars, gekarakteriseer en met aanvaarbare vrylatingsvlakke vergelyk. Metings word uitgevoer, en met behulp van die GenieTM 2000 sagteware ontleed, om aktiwiteits konsentrasies vas te stel vir die bekragtiging van die berekende aktiwiteite. Algoritmes wat in die Genie 2000 sagteware gebruik word, word uiteengesit om beter te verstaan hoe die aktiwiteits konsentrasies bepaal word. Vir die beton monster stem gemete en voorspelde aktiwiteite vir 60Co en 152Eu ooreen met 'n gepaartgaande onsekerheid van 11% en 2%, onderskeidelik. Terwyl nie-sistematiese verskille aanvanklik vir die Fluental en aluminium gevind is, word dit verstaan dat die waarskynlikste oorsaak die aangenome materiaal samesteling in die berekeninge is. Na induksie-gekoppelde plasma massaspektrometrie (IKP-MS), wat tot meer akkurate kennis van die samestellings van die Fluental en aluminium monsters lei, is daar baie beter ooreenkoms tussen gemete en berekende resultate. Die verskille is steeds groot, maar sistematies. Die berekende waardes is konsekwent een orde-grootte groter as die gemete waardes, behalwe vir 59Fe in die aluminium monster waarvoor die waardes tot dieselfde orde-grootte, ooreenstemend. Die 13 γ-spektra vir die ioon-uitruil hars dromme lyk almal dieselfde, met sommiges meer aktief as ander, soos wat verwag kan word vir 'n nie-homogene monster. 60Co en 137Cs word in elke drom gevind, met 137Cs vermoedelik afkomstig van 'n lekkende brandstofstaaf. Hoewel baie tyd en moeite aan die ontleding van 'n klein hoeveelheid monsters bestee is, in vergelyking met die hoeveelheid monsters in die volledige reaktor, is dit 'n noodsaaklike oefening om lesse aan te leer wat in die toekoms tyd kan spaar. Die belangrikste is die studie van die meet- en berekeningmetodes wat later tydens volledige ontmanteling toegepas sal word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/101305
This item appears in the following collections: