Menu
Nuclear Waste Management
I worked for a while in the Environmental Modelling Group of BNFL, testing, improving and applying a suite of risk assessment and groundwater transport models. The models are designed to study the long-term evolution (up to 100,000 years) of low-level waste dumps. Low-level waste consists mostly of contaminated material that may have been in contact with radioactive substances and then discarded. So, it might include waste paper, building debris, containers or old protective clothing. Typically, this material is disposed of by simply burying it a few metres underground. In the UK most low-level waste is sent for disposal at the Drigg site, which for historical reasons is rather near to Sellafield. It's important to understand and predict the long-term behaviour of such sites in order to be confident that the sites are reasonably safe. (It's also a legal requirement for BNFL to produce post-closure risk assessments of the site.) The most sophisticated, or at least the most comprehensive, of these models is the GRM/DRINK biogeochemical model. It considers the coupled interactions of many environmental processes including groundwater flow, microbiology, chemical reactions, corrosion, sorption, waste settling and (of course!) radioactive decay and ingrowth. Groundwater flow is calculated with a separate model and is used as input to GRM. The GRM output includes concentrations of radionuclides within groundwater leaving the site. The transport of such contaminated groundwater through the geosphere and biosphere can then be modelled using other codes, driven by the GRM output. Apparently, our use of such a scheme to construct source terms for a post-closure radiological safety assessment of Drigg was ``first direct use of a mechanistic reaction-transport model in [nuclear waste] risk assessment calculations''.