Flixborough Memorial

Flixborough: The Human Interaction and Decision Making

Nearly 48 years have passed since the Flixborough disaster; an unfortunate event that has provided much learning for the process industries and overall health and safety management. The Flixborough disaster claimed the lives of 28 people who were working at the time, injured another 36 onsite and 53 offsite as the explosion devastated the plant, the process industry and the village of Flixborough.

The disaster unfolded in June 1974 and was the result of a catalogue of errors, human and mechanical. One of the main functions of the plant was to produce a substance used in the production of nylon. This meant oxidising cyclohexane through a series of six reactors. On Saturday 1st June, 1974, a temporary pipe linking two reactors together jack-knifed, causing an escape of cyclohexane, which became a vapour cloud and ignited in a matter of minutes. Much has been written about the engineering failures that brought out such devastation. However, the events that took place 1st June, 1974 were not failures in isolation, but the result of management decisions and actions made much earlier. It was a combination of engineering errors, plant design and management decisions that once all aligned, proved catastrophic for Flixborough.

According to Reason (2008, p. 96) “accidents have their primary origins in the fallible decisions made by designers, builders and top-level management”.   Some of these decisions may be classed as a slip, lapse or mistake. It was the slips, lapses and mistakes that occurred throughout the management at Flixborough that led to the devastation in June 1974.

Management had not retained engineering competency on site. When the Flixborough management neglected to replace the Chartered Works Engineer after he resigned in early 1974, they lost valuable mechanical knowledge and expertise. Day to day maintenance had been given to an electrical engineer by the site management who had little mechanical experience. This left significant deficiencies in engineering related competency, which would have been invaluable when the management decided to remove reactor number 5 upon the discovery of a 1.8m crack in the vessel. This lack of knowledge meant that no real effort was made to understand the cause of such a crack and also to inspect the other reactors for similar issues. Examination after the explosion found the crack to be the result of nitrate stress corrosion due to using cooling water treated with nitrate to dilute minor cyclohexane leaks from the facility (Great Britain. Department of Employment, 1975, p36). This would prove to be another error of the management team, although it can be said that this type of corrosion was not fully appreciated even by the process industry at the time (Great Britain. Department of Employment, 1975, p36). Nonetheless, it can be concluded as a management short-coming. Had a full and thorough investigation followed, the discovery of the crack, this cause of corrosion could have been found.

A change to the integrity of the plant caused additional engineering failures when Flixborough management decided a by-pass pipe was to be manufactured in-house to fill the void left by the removal of reactor number 5. The by-pass pipe would link reactor 4 and 6 together so that production could continue. It was considered an everyday plumbing task and therefore no engineering drawings were made, no calculations and no hydraulic testing before it was put into use. Due to weight of the pipe, scaffold was needed to support it and was found wholly inadequate. Overall, the job was rushed in order for the plant to begin producing again. The failure to consider mechanical risks associated with the pipe was an error by the site management. Again, the involvement of a competent engineer would have ensured the modification was adequate and that mechanical risks had been addressed. The failure to have a robust management of change process at the plant permitted additional error and deviation. The modification made, inevitably destroyed the integrity of a well built and functioning plant.

Poor plant design and consideration for plant layout were also contributing factors. The control room and offices were designed and built next to hazardous areas of the plant. The control room was totally destroyed and where most people lost their lives. The offices too, were close to hazardous areas of the plant. Had the explosion occurred on a weekday and not a weekend the loss of life would have been significantly more. It can be said that the human error during the planning and design stage of the plant led to the excessive loss at Flixborough.

Since the disaster at Flixborough much has been learned to help improve health and safety in the workplace, not just the process industry, but in all industries. After the Flixborough accident, in 1974, the Health and Safety at Work Act (HASAWA) became the cornerstone of UK regulation. The nature of this self-regulating law may have put additional pressure onto the management at Flixborough to ensure the health, safety and welfare of its employees and others, as they would have been legally bound to control risk so far as is reasonably practicable; taking into account time, cost, effort and resource. Could following HASAWA 1974 prevented the Flixborough disaster? Perhaps. In 1988, when HASAWA 1974 had been in place for 14 years, a gas leak from pipe work being maintained on Piper Alpha caused an explosion which claimed the lives of 167 people. Despite HASAWA 1974, it would seem such a major incident could still occur. Human error and safety culture on the doomed platform played a significant part in knowledge-based decision making. Like Flixborough, changes to platform design after original installation brought hazardous areas of the platform close to personnel areas.

The design and construction of facilities were at the core of these major incidents and learning from that safer design and construction became very important. In 1994 legislation was passed to control construction and design in the Construction Design and Management Regulations. It has been since reviewed with particular emphasis on designers to design out risk. This would include keeping personnel areas, offices and control rooms away from major hazards and include such material as blast panels to protect personnel areas. After Piper, the off-shore industry also recognised that hazards should be designed out and therefore additional legislation were passed which included the Offshore Installations (Safety Case) Regulations and the offshore Installations and Wells (Design and Construction, etc) Regulations.

Changes to plant design and modifications to equipment can also lead to disaster. The management of change process is now an integral part of the COMAH (Control of Major Accident Hazards) Regulations in the UK. Insisting that part of the safety report be given to reviewing change management procedures and any changes that may have significant repercussions triggers a review of risk assessments and the necessary control measures for risk (Health and Safety Executive, 2009, p.3).

Another key learning point from Flixborough highlighted that more must be done to either reduce the number of hazardous substances stored on site or even substitute these substances for something less hazardous. A control championed by Trevor Kletz (Kletz, 2007, p.50) and a key part of the hierarchy of controls under COSHH Regulations.

The 1990’s saw little in the way of major disasters. By contrast, the new century saw a spate of major accidents despite a wealth of legislation. In 2003 an explosion in a chemical plant in Geleen saw three workers die. BP Texas City Refinery in 2004 experienced an explosion which led to 15 deaths (Okoh and Haugen, 2014, p. 12-14). In 2005, the Buncefield oil storage depot experienced a vapour cloud igniting, which led to a fire burning for five days. And in 2010, the Deep-Water Horizon claimed the lives of 11 people when an explosion caused the platform to sink. Business priorities, lack of maintenance, poor application of management of change and design deficiencies were citied as causes of these incidents, all underpinned by broader management or human failings.

Organisational structures have also changed, perhaps as a lesson learned from Flixborough. Management teams typically now include access to competent sources of advice such as health and safety management. The Management of Health and Safety at Work Regulations require employees to appoint competent people and arrange for appropriate training and instruction for employees.  Organisations have also realised that involving the workforce, listening and addressing hazards and concerns help improve workplace safety overall. “Good…safety management does not happen by chance and requires constant active engagement” (Health and Safety Executive, 2005, p. 28).  It is this understanding that may have seen a change in organisational structure.

It is clear that hazards are best eliminated or controlled at the design stage including the design of plants, layout and construction materials, but also the need for competency at key levels and especially around design and engineering. Competency from board level to the coal face can never be underestimated.

Change management procedures in regulations such as COMAH should not be exclusive to the major accident hazard industry. Such procedures are essential throughout all industry and must include changes to the physical work environmental and the organisation, namely people and processes. “Change, of any kind, is a powerful error producer” (Reason, 2008, p. 45).

And it is the human contribution that is proving to be an area for more learning. Despite the improvements to legislation, enforcement, design, construction and technology; one key element remains consistent throughout the investigations to major incidents and that is the human contribution. That human interaction with design, construction, processes, procedures or management decisions can lead to catastrophic events. We are not infallible and it is inevitable that mistakes will be made. People safety is becoming just as important as process safety and rightly so. Focussing on minimising opportunities for error will minimise the opportunity for accidents to occur.

Leave a Comment

Your email address will not be published. Required fields are marked *