Protection of non-human biota against radiation in freshwater—Effect of time dependence in tiered exposure assessment

Abstract

The ERICA's European framework that have been proposed to assess the exposure of biota to radionuclides is based on tiered approach: the Tier 1 assessment is based on equilibrium assumptions (constant radioactive releases over time, equilibrium at the interfaces water-suspended matter and water-biota); the Tier 2 assessment allows to use time-dependent and site-specific data to perform a more precise evaluation. The objective of this paper was to compare these two tiers in the context of routine releases from nuclear power plants in freshwater ecosystem. The Loire River where 14 nuclear power plants (NPPs) operate was chosen as a case study and the transport of ⁶⁰Co in the river, its partition between water and suspended particles and its bioaccumulation was calculated under equilibrium and time-dependent assumptions respectively. Using this case study allowed to quantify the overpredictions linked to conservatisms assumed in Tier 1. Recommendations were identified on when a Tier 2 approach is actually needed.

Introduction

So far, protection of the environment against radiation was based on the assumption that non-human biota are protected when exposure levels are protective for humans (e.g. ICRP, 1990, IAEA, 1992). This assumption was however criticized because: (i) non-human biota exists in zones where humans do not; (ii) specific exposure pathways and/or effects can occur for some species, including rare or endangered species. Thus, there is now an international consensus whether a dedicated framework is needed to assess that biota is protected against ionising radiation (Holm, 2004, Janssens, 2004).

Approaches proposed so far or that are under construction (i.e. Jones et al., 2003, Larsson et al., 2004; the ERICA approach (European Commission, 2007a; http://www.erica-project.org)) structure the assessment on the classical steps of environmental risk assessment (ERA), i.e. (1) the problem formulation, (2) the exposure and (3) effect assessment, and (4) the risk characterisation. The scope of this paper is focused on the step (2), that aims at calculating the external and internal exposure levels to biota. The paper is limited to freshwater ecosystems.

The ERICA assessment framework (European Commission, 2007a) is built on a tiered approach, each tier requiring more or less information on the context to be investigated: (i) the Tier 1 assessment is designed to be simple and conservative, requiring a minimum of input data (e.g. maximum releases in the system); (ii) the Tier 2 assessment allows the user to be more interactive and to use ‘personal’ data to perform a more precise evaluation; (iii) Tier 3 is a probabilistic risk assessment in which uncertainties may be determined.

When using such a tiered approach, the time dependency of input data and parameters is a key element for formulating the assessment problem and for defining which tier is actually required for the investigated context. Two types of ‘time dependency’ in the problem formulation can be distinguished:

• Time dependency of input data. Tier 1 assessment generally considers that releases are constant and continuous over the year. Time dependency of the radioactive releases into the ecosystem is therefore neglected and this assumption is considered to be conservative. However, actual releases from NPPs are intermittent. For example, in France, NPPs generally discharge once a month during some hours and releases are constrained by the flow rate of the river (releases being forbidden when the dilution rate in the river is too low or when the consequences of flooding must be avoided). Such industrial and environmental constraints lead to temporal variation in the concentration of radionuclides in the river system (Ciffroy et al., 2005). When conducting a more detailed evaluation by Tiers 2 or 3, it may be justified to consider time-dependent releases into the freshwater system instead of ‘time-averaged’ or ‘time-maximized’ releases.

• Time dependency of transfer parameters. Current approaches are based on two main parameters: (i) the radionuclide distribution coefficient K_d, expressed as the concentration ratio between the particulate phase and the dissolved phase under equilibrium conditions; (ii) the radionuclide Concentration ratio CR, expressed as the concentration ratio between the bioaccumulated concentration and the concentration in dissolved water under equilibrium conditions. In Tier 1, it is generally considered that radionuclides released into the river are immediately equilibrated within all the compartments of the system, especially at the interface between water and particulate suspended matter, and at the interface between water and organisms. However, radionuclides are initially released in a dissolved form and their sorption to particles can be slow (Durrieu et al., 2006). As for suspended particulate matter, it is well known that the accumulation and elimination of radionuclides by organisms (e.g. fish) can be slow (Adam, 1997). Thus, the dynamics of radionuclide uptake/elimination by freshwater organisms can significantly modify the internal exposure, both in terms of concentrations and time scales. Assumptions used in Tier 1 become therefore questionable when conducting evaluation by Tiers 2 or 3. In such cases, it is justified to study the relevance of the “equilibrium” assumption for exposure assessment.

The objective of this paper is to study the effect of ‘time dependency’ when conducting risk assessment in freshwater systems submitted to routine releases from NPPs. For this purpose, two risk assessments were carried out in parallel: the first one followed the ERICA's Tier 1 method and then used equilibrium and conservative assumptions; the second one could be relevant for a Tier 2 approach, since it considered time-dependent and site-specific data. Such a comparison will be used to provide some guidance on how time dependency concepts can be incorporated in Tiers 2 and 3 of frameworks like ERICA.

It has to be noted that only external and internal exposures due to radionuclides present in the water column were considered in this paper and that exposure from contaminated sediments to benthic biota were not studied. In order to study the effect of time-dependent input data and transfer parameters and to compare time-dependent approaches with equilibrium ones, ⁶⁰Co releases into the Loire River, where 14 NPPs operate, was chosen as a case study.