+ de critères

Risk prevention for radioactive pollutants

Yves Thiry (ANDRA) takes the floor

How observing forest ecosystems helps prevent the risks of radioactive pollution


Chronic low-level contamination and long time-scales are primordial criteria to consider when dealing with risks due to activities in the nuclear sector, including the storage of radioactive material.

The wide variety of contexts and the complex interactions among different environmental compartments (atmosphere, soil, water, flora and fauna) make it difficult to model the transfer of chemical elements into the biosphere; choices must be made concerning which concepts, parameters and hypotheses will best reflect the degree of accuracy desired. To avoid under-estimating the long-term risks involved with toxic radioactive pollutants in the environment, existing models (classically applied to the food chain and in environmental impact studies for radioactive waste storage sites) have adopted a simplified view of how contaminants are transferred into the environment: they presuppose a stylized, balanced environmental system and apply a precautionary strategy with a pessimistic bias on parameter values. This conservative approach is above all dictated by regulations and criteria for radiation protection. However, it is not well adapted when precise scientific information is the goal: for example, accurately describing the dynamics of how these pollutants accumulate in the soil or in vegetation, or determining speciation, whether they can be absorbed into the biosphere or how toxic they actually are. As a complement to the classic predictive models mentioned above, researchers need a more mechanistic description of contaminants' bio-geochemical cycle in our typical natural ecosystems to develop more rational dynamic models capable of handling uncertainty and leading to appropriate questioning.

Today forest ecosystems are often found near zones where radioactive waste has been stocked. In many different climatic zones, forests are the natural ecosystems that dominate areas where man does not occupy the territory. The long time-scales (100 to 10,000 years) required to evaluate the risks associated with nuclear waste storage are similar in length to the time-scales involved in forest ecosystem functioning. The long-lasting nature of forest ecosystems therefore makes them well-adapted to the study of the long-term behaviour of environmental contaminants. More precisely, this includes:

  •  Describing how contaminants are redistributed between the soil and the vegetation, and the time-scales involved under environmental conditions representative of future climates
  • Comparing the impact of potential contaminant releases with the impacts of the historic contaminant footprint (natural vs. artificial) in the environment
  • Specifying the risks of accumulation by modelling different vectors of contamination (atmospheric deposits vs. underground release)
  • Checking the validity of the simplified generic transfer models currently being used.

The main priority concerns stable or radioactive isotopes of the elements Cl, I, Se, Cs, C, B, As, Hg, Cs ..., the focus of the Andra programme, which are detectable in the natural background. Future models will aim for ecological realism, rather than complexity, and will be based on data such as recorded concentrations, stocks and fluxes (retention, transformation, volatilisation,...) collected from densely equipped sites or sites where access to a long time series of measurements is authorised.

In this context, the RENECOFOR Network was approached to furnish various samples (water, soil, biomass) representative of contrasted environmental conditions. For example, RENECOFOR sample collections were used to better quantify chlorine (inorganic vs. organic) transformation fluxes in the soil column, thus providing more explicit chlorine cycle models. Other studies are underway to identify the role of environmental factors influencing the distribution, speciation or residence time of other elements (chorine-36, iodine, selenium, caesium).

Pour en savoir plus