Wed 1st April 12 – 1 pm Rm 312.222

One of the more difficult challenges involved in the disposal of radioactive waste in deep geological repositories is the very long time frames required by safety assessments. In the case of one Canadian project – Ontario Power Generation’s Deep Geological Repository (OPG DGR) project for the management of low and intermediate level radioactive reactor waste – the safety assessment time-frame is 1 million years. This type of time frame goes well beyond anything that can be considered in an experimental setting. Mathematical models that make long term estimates face challenges with considerable uncertainties over these time frames. There are also significant challenges with explaining the results of these models to the public and other stakeholders.

The safety case required for deep geological repository projects must use multiple lines of evidence to complement safety assessments that quantify the level of protection provided by a geological disposal facility. Geological time frames are particularly useful to support arguments associated with safety assessments, as geological time frames are often in excess of safety assessment time frames. For example, the OPG DGR project safety case considers a one million year period, and the sedimentary units that are proposed to host the repository have a well-studied, well-documented geological setting indicating that host sedimentary rocks were deposited and consolidated hundreds of millions of years ago, onto billion year old Canadian Shield rocks.

This talk will present some examples of research programs that focus on the safe long-term management of used nuclear fuel in a DGR, whose objectives are to build capabilities to assess the suitability of potential sites for a centralized deep geological repository, for all of Canada`s used nuclear fuel. Potential DGRs are assessed in terms of isolating and containing radioactive waste over geologically long time frames. Specific research topics focus on reducing uncertainties in models of long term repository performance and include (but aren`t limited to): modelling the results of large-scale shaft seal experiments, laboratory determinations of sealing material performance, modelling of the effects of past and future glacial events, natural tracers and natural analogues.