Health effects of the Windscale Pile fire

On the 7th October 1957, at the Windscale nuclear complex in the northwest of the UK, a procedure was initiated to release stored Wigner energy from the graphite core of an air-cooled plutonium production reactor ('Pile 1'). In the early hours of the morning on 10 October 1957 temperatures in part of the core rose above the intended levels, and marked increases of radioactivity were observed both in the cooling air discharge and in on-site air samples. By about 4 pm in the afternoon it was apparent that both graphite and uranium fuel elements were involved in a fire in part of the core; strenuous efforts to control the fire culminated with the flooding of the core with water at 9 am on 11 October, and the fire being extinguished by that evening. About 18.6 PBq of radioactivity were released to the atmosphere, of which the most important components were 1.8 PBq of ¹³¹I and 42 TBq of ²¹⁰Po [1].

The event, generally known as the 'Windscale accident' or the 'Windscale pile fire', has been retrospectively rated at 5 on the INES scale [2].

The accident resulted in the first scientifically documented example of intervention, based on radiological protection criteria, to protect the public from an accidental release of radioactivity. The original paper has been recently re-published in this Journal to make it readily available [3, 4].

At that time there was no established guidance for the limitation of radiation dose to the public after an accident; the existing ICRP recommendations for internal radiation dealt only with continuous intake by workers. One can only imagine the pressures that the health physicists worked under to reach the conclusion that distribution of milk at concentrations in excess of 3700 Bq l⁻¹ (0.1 µCi l⁻¹) should be prevented, in order to limit the dose to children's thyroids to 200 mSv (20 rad). Distribution of milk from an area of 207 km² was banned on 12 October; following more extensive surveys the ban was extended to 518 km² on 14 October. As iodine concentrations in milk reduced below the thresholds the ban on milk distribution was progressively lifted, restrictions being completely removed by 23 November.

The accident has subsequently been the subject of many scientific studies, the majority of which have concerned the subsequent distribution of radioactivity in the environment, estimates of the release inventory, and modelling of the plume dispersion [1, 5–14].

Studies directly related to the health effects of the accident are much more limited. Theoretical estimates have been made of the expected number of excess cancers due to the releases, based on collective dose estimates and the then prevailing understanding of dose-response [15] and the mortality of the 470 workers most closely involved in the accident and subsequent on-site clean-up has been the subject of two epidemiological analyses [16, 17].

The study in this issue by McNally et al of thyroid cancer incidence in northwest England following the 1957 accident is the first observational study of possible health effects due to the accident in the general population [18]. It is therefore an important addition to the literature on the Windscale Pile Fire, particularly so against the background of the substantial excess of thyroid cancer amongst those exposed as children in the areas of Ukraine and Belarus most heavily affected by the Chernobyl accident in 1986.

McNally et al have been able to obtain data on all cases of thyroid cancer diagnosed between 1974 and 2012 in individuals resident in England at that time, and born between 1929 and 1973. The thyroid cancer incidence rates in Cumbria and Lancashire, the counties most highly affected by iodine deposition from the Windscale accident, are compared with those for the rest of England.

The study finds a significantly raised incidence rate ratio (relative to the rest of England) of 1.29, 95% CI 1.09–1.52, for those aged less than 20 years in 1958, and diagnosed during 1974–2012 while living in Cumbria. However the authors point out that the significantly increased rate ratio for those resident in Cumbria at the time of diagnosis extended to individuals born between 1959 and 1963, who would not have been exposed to iodine-131 from the accident. They conclude that 'although further investigations are required to properly understand the unusual patterns of thyroid cancer IRRs in Cumbria and Lancashire, the results of this preliminary study are not consistent with an effect of exposure to iodine-131 from the Windscale accident.'

As such, the study does provide assurance that an excess of thyroid cancer on the scale seen post-Chernobyl has not arisen from the Windscale accident; this is in line with expectation given the relative magnitude of the releases and assessed doses following the two accidents.

It is in the nature of epidemiological studies to leave some questions open, and this study is no exception. As the authors identify, incidence statistics were only available at the county level and in terms of place of residence at the time of diagnosis. The county level statistics will 'dilute' any effects of the accident as, for example, the incidence statistics for Cumbria will include both population and cases residing in the population centres to the north of the site which received very little iodine-131 deposition from the accident. Identifying cases in terms of residence at the time of diagnosis is an imperfect surrogate for residence during the critical period of exposure in late 1957 and early 1958 and is likely to further dilute any effects of the accident.

The authors suggest use of the 'Cumbrian birth cohort' [19] in a future study that would circumvent the above difficulties in the present geographical study. This would in principle allow individual estimates of thyroid dose to be related to thyroid cancer incidence. Notwithstanding the substantial complexity and difficulties of estimating individual doses, such a study would be a further valuable addition to the literature.

The continuing interest in studies of the Windscale accident, and the environmental impacts arising from the early years of operation of the UK nuclear industry in general, does raise an issue concerning continued availability of the important source data. Many high quality studies have been undertaken over the years and in most cases detailed results have been presented in reports issued by the research institutions concerned (such as the former Atomic Energy Research Establishment). At best, only a summary of these results will have reached the published scientific literature and hence the online journal archives that we have become used to in the present day. Some, but not all, of these reports will have reached the National Archives and can be found (albeit at the cost of some effort) by a determined researcher. As the older generation of researchers and health physicists fades away, the very existence of some of these studies may be forgotten. Is there a case on which an enterprising institution could seek funding to systematically identify, catalogue and preserve this material?