Windscale Fire (UK, 10 October 1957)

Summary

On 10 October 1957 a fire broke out in the core of Windscale Pile No. 1, one of two air-cooled, graphite-moderated reactors on the Cumberland coast (now Sellafield) built to produce plutonium for Britain’s nuclear-weapons programme. During a routine operation to release stored energy from the graphite, a second heating was applied too soon; the core overheated, uranium cartridges ruptured and the graphite caught fire. It burned for around three days and released radioactive material — chiefly iodine-131 — across the UK and into Europe. The government banned milk from a roughly 200-square-mile area for about a month, then insisted there was little to worry about, and kept the full official report secret for three decades. It is the archive’s clearest case of production pressure and secrecy in a high-hazard system.

Systemic Features

Production pressure over safety. The Windscale piles were built quickly and pushed hard to produce weapons-grade plutonium — and later tritium, beyond their original specification — for a programme the government wanted delivered as fast as possible. Safety concerns were present from the start; operators had even shaved the fins of fuel cartridges to raise output. A high-hazard system was run to the edge of its margins in the service of a schedule (see normalisation of deviance).
Acting on an incomplete model of a tightly-coupled system. The behaviour of stored (Wigner) energy in the graphite was not fully understood; the annealing procedure was a workaround for a design phenomenon, and the fatal second heating was applied on the basis of that imperfect understanding. In a tightly-coupled physical system, the fire developed faster than the operators’ comprehension of it (see tight coupling).
Secrecy and reassurance over disclosure. William Penney’s inquiry reported within weeks and was sharply critical of technical and organisational deficiencies — but the Prime Minister, Harold Macmillan, permitted only a summary to be published and suppressed the full report for around thirty years, and the government publicly downplayed the accident. The institution’s account of itself was managed to protect the weapons programme’s standing, the same reassurance-over-disclosure pattern seen in BSE (see organisations as cognitive systems).
The precaution that was mocked, then vindicated. The chimney filters, insisted upon late and at cost by John Cockcroft against ridicule — “Cockcroft’s Folly” — captured much of the radioactive particulate and are credited with greatly reducing the release. A conservative safety margin, added over objections, was what limited the disaster: the inverse of the reflex to optimise a hazard away.

Cascading Systems Affected

Public health and the food chain (radioactive iodine in milk; a month-long milk ban over ~200 square miles)
The environment (contamination across the UK and parts of northern Europe)
Public trust in nuclear safety and in official statements
Nuclear regulation and the weapons programme

Impacts

The two Windscale piles were closed permanently. A later inquiry (Fleck) led to the creation of the National Radiological Protection Board in 1971.
Estimates of long-term harm are uncertain: a 1990 assessment suggested the release might eventually cause on the order of 100 fatal cancers, though the statistical power to detect effects is low.
The full Penney report was declassified only in the late 1980s, decades after the government’s reassurances.
The site, renamed Sellafield, continues to host nuclear activity; decommissioning of the fire-damaged pile is a project running to the far future.

Systemic Failure Archive and Research

Explorer

Windscale Fire (UK, 10 October 1957)

Summary

Systemic Features

Cascading Systems Affected

Impacts

Further Reading / Sources

Graph View

Table of Contents

Backlinks