In order to provide reasonable assurance that clay barriers will ensure long-term waste isolation, it is essential to understand their behaviour under a variety of conditions. Due to the variability of argillaceous materials, to the complexity of the phenomena that might take place in a waste repository and to the longevity of the required isolation, an adequate understanding of the behaviour and the capability to model the evolution of the clay barriers are not easy tasks. The factors that need to be understood and modelled include stress evolution, long-term strain or creep, thermal effects on solid skeleton, on interstitial fluids and on mineralogy. The difficulty of the task is increased by the facts that many effects are coupled, that their rates must be extremely low, in order to be realistic, and that the time period to be modelled defies the possibility of direct experimental observation. Several critical issues are identified and discussed briefly, such as: constitutive law to describe the thermo-mechanical behaviour of the clay skeleton, modelling of the fluid phase in clays and its response to heating, thermal fracturing and healing. Strategies are suggested for a rational approach to the experimental investigation of some relevant processes. The study of suitable natural analogues, for example the thermo-metamorphic halo occurring at Orciatico in Tuscany, could provide valuable insights in the thermal effects of heating clay barriers. It is conceivable that models describing a variety of relevant phenomena, such as dehydration, fracturing and permeability changes could be tested through the study of the Orciatico analogue.

In the end performance assessments of clay barriers would benefit through improvements in modelling: this would involve progress in understanding the basic phenomena and their coupled nature, improved conceptual and mathematical models and increased reliability for their calibration/validation. The improved understanding of phenomena requires additional experimental activities on various levels: molecular, microscopic, macroscopic, medium scale and in situ.