Optimisation of environmental remediation: how to select and use the reference levels

A number of past industrial activities and accidents have resulted in the radioactive contamination of large areas at many sites around the world, giving rise to a need for remediation. According to the International Commission on Radiological Protection (ICRP) and International Atomic Energy Agency (IAEA), such situations should be managed as existing exposure situations (ExESs). Control of exposure to the public in ExESs is based on the application of appropriate reference levels (RLs) for residual doses. The implementation of this potentially fruitful concept for the optimisation of remediation in various regions is hampered by a lack of practical experience and relevant guidance. This paper suggests a generic methodology for the selection of numeric values of relevant RLs both in terms of residual annual effective dose and derived RLs (DRLs) based on an appropriate dose assessment. The value for an RL should be selected in the range of the annual residual effective dose of 1-20 mSv, depending on the prevailing circumstances for the exposure under consideration. Within this range, RL values should be chosen by the following assessment steps: (a) assessment of the projected dose, i.e. the dose to a representative person without remedial actions by means of a realistic model as opposed to a conservative model; (b) modelling of the residual dose to a representative person following application of feasible remedial actions; and (c) selection of an RL value between the projected and residual doses, taking account of the prevailing social and economic conditions. This paper also contains some recommendations for practical implementation of the selected RLs for the optimisation of public protection. The suggested methodology used for the selection of RLs (in terms of dose) and the calculation of DRLs (in terms of activity concentration in food, ambient dose rate, etc) has been illustrated by a retrospective analysis of post-Chernobyl monitoring and modelling data from the Bryansk region, Russia, 2001. From this example, it follows that analysis of real data leads to the selection of an RL from a relatively narrow annual dose range (in this case, about 2-3 mSv), from which relevant DRLs can be calculated and directly used for optimisation of the remediation programme.