Validation of a physically based catchment model for application inpost-closure radiological safety assessments of deep geological repositoriesfor solid radioactive wastes

Abstract

The physically based river catchment modelling system SHETRAN incorporatescomponents representing water flow, sediment transport and radionuclidetransport both in solution and bound to sediments. The system has been appliedto simulate hypothetical future catchments in the context of post-closureradiological safety assessments of a potential site for a deep geologicaldisposal facility for intermediate and certain low-level radioactive wastes atSellafield, west Cumbria. In order to have confidence in the application ofSHETRAN for this purpose, various blind validation studies have beenundertaken. In earlier studies, the validation was undertaken againstuncertainty bounds in model output predictions set by the modelling team onthe basis of how well they expected the model to perform. However, validationcan also be carried out with bounds set on the basis of how well the model isrequired to perform in order to constitute a useful assessment tool. Herein,such an assessment-based validation exercise is reported. This exerciserelated to a field plot experiment conducted at Calder Hollow, west Cumbria,in which the migration of strontium and lanthanum in subsurface Quaternarydeposits was studied on a length scale of a few metres. Blind predictions oftracer migration were compared with experimental results using bounds set by asmall group of assessment experts independent of the modelling team. Overall,the SHETRAN system performed well, failing only two out of seven of the imposedtests. Furthermore, of the five tests that were not failed, three werepositively passed even when a pessimistic view was taken as to how measurementerrors should be taken into account. It is concluded that the SHETRAN system,which is still being developed further, is a powerful tool for application inpost-closure radiological safety assessments.