Estimating future doses from long-lived radionuclides discharged to sea from the British Nuclear Fuels reprocessing plant at Sellafield

A number of long-lived radionuclides have been discharged, under authorization, into the waters of the Irish Sea as part of the low-level liquid wastes arising from the British Nuclear Fuels (BNFL) reprocessing plant at Sellafield. These discharges began in 1952 and have continued to the present day. Throughout this period intensive monitoring programmes of external and internal exposure, including surveys of the habits of local people and analysis of large numbers of environmental materials, have ensured compliance with authorized limits of discharge and consequently with the necessary level of protection of the general public. The rates of discharge have been considerably reduced in recent years, and this has resulted in a concomitant decrease in dose rates. Nevertheless, the long half-lives of some of the nuclides discharged are such that some level of dose will be received by the public for many years to come. This has necessitated detailed surveys of the distribution of these nuclides in the environment, studies of the processes which control this distribution, and the development of comprehensive mathematical models in order to estimate their future distribution under different environmental circumstances. The most important long-lived radionuclides, in terms of potential dose to man, are those of the transuranium series, in particular 239/240Pu, and 241Am, which also grows-in from 241Pu. These nuclides are primarily associated with sedimentary materials. This makes it difficult to determine their distribution in the environment, and requires a considerable range of studies to be made in order to estimate the rates at which they will continue to enter food chain pathways to man. Equally difficult to estimate are the dominating environmental processes which will affect these distributions over different time-scales. In the short term it is possible to extrapolate from past experience, but models of environmental processes are required to make longer term predictions. (author). 26 refs, 6 figs