The history of liquid and atmospheric discharges from the Sellafield nuclear complex in Cumbria, UK over the period from 1951 (the start of operations at the site) to 1992 has been reviewed. The chronology of liquid effluent discharges is essentially unchanged from previously published figures, but the chronology of aerial effluent discharges has been very substantially amended, reflecting the results of experimental work to assess the efficiency of aerial effluent sampling systems on the site, the analysis of previously unreported data on plant based measurements of aerial effluents in the 1950s, and a number of other factors. The amended discharge chronologies have been validated by reference to the extensive environmental monitoring data which are available for the Sellafield site, comparing environmental concentrations predicted from the discharge chronology and environmental models with those actually measured. The revised discharge chronologies together with the model account for all the major features of spatial and temporal variations in environmental concentrations of radionuclides in the vicinity of the site, but overestimate the body content of plutonium in local residents compared to the limited number of available measurements. The revised discharge chronologies do not lead to any major changes in the assessed lifetime doses to local residents and do not alter the conclusion that doses from discharges appear to be far too small to account for the excess of childhood leukaemia in the local village of Seascale.

No abstract

No abstract

Radiological assessments have assumed that the mass of irradiated uranium oxide particles inadvertently released to the atmosphere from the Windscale Piles, two nuclear reactors at Windscale Works, Sellafield, England, during the 1950s was 20 kg. This paper re-examines the assumptions upon which this figure was based and concludes that the value is a realistically conservative estimate of the release, consistent with current radiological protection practice. The mass estimate is derived from a reanalysis of plant data produced at the time. The environmental data on which the initial estimates were based are reconfirmed, and additional support is provided by an interpretation of modelling studies of both the total deposition and milk concentrations resulting from that deposition. Milk-monitoring data from the time are shown to be consistent with the release assumptions used in the dispersion modelling exercise. Finally, the issue of statistical undersampling is addressed using the particle numbers and size distributions produced by the modelling exercise.