King’s Cross Fire, 1987
On the evening of 18 November 1987, a fire at King’s Cross London Underground train station killed 30 people, including one of the first fire-officers on the scene. A further person died later in hospital.
Most of the fatalities occurred when a flashover engulfed the ticket hall at the top of the escalators (hot air ignited and produced a ball of fire). There was some delay before London Fire Brigade were called, and the flashover occurred within two minutes of their arrival in the ticket hall.
London Underground dates from 1863. At the time of this fire, 80% of the system was more than 70 years old. The network carried two and a half million passengers every weekday. King’s Cross itself was a complex intersection of five Underground and three Intercity lines, across five levels below ground. It was the busiest station on the network.
This disaster resonates with me personally. I joined the UK health and safety Regulator (the UK HSE) in 1995. My senior manager had provided evidence to the Inquiry and so I learnt about this event first-hand. Unfortunately, whilst working for the UK HSE, I investigated several major fires – some with fatalities – that had similarities with this event, particularly regarding emergency response.
The immediate cause of the fire was a discarded match falling onto wooden escalators with an accumulation of debris underneath and in the escalator tracks. Although smoking had been banned underground in 1985, it was still common practice for smokers to light-up on the escalators on their way out of the station.
Human factors issues
Leadership, priorities and culture
In the 30 years before this disaster, there had been over 400 escalator fires on the Underground system, many of these due to discarded matches. The investigation assessed 46 serious escalator fires, several of which led to serious damage and station evacuations. This clearly was not a one-off event.
Management considered fires to be inevitable. However, they were not referred to as fires, but were called “smoulderings”, partly so as not to alert or concern senior management. This history of events reinforced a culture where fires were not unusual – and Underground staff dealt with fires themselves, rather than call the London Fire Brigade.
Fire or smoke detection had never been installed. It was assumed that any fire would be detected by staff or passengers. It was also assumed that there would be sufficient time to evacuate. The Inquiry proposed a new, proactive approach to safety management.
“The management remained of the view that fires were inevitable on the oldest and most extensive underground system in the world. In my view they were fundamentally in error in their approach”.Investigation in the King’s Cross Underground Fire, Desmond Fennell OBE QC (“Fennell Report”), 1988, p.17
Incredibly, there was a sprinkler system on the escalators, at ground level. This system was not designed for occasional emergency use, but for regular nightly use, in order to wet the machinery, and dampen any of these the regular “smoulderings”. However, this sprinkler system caused high levels of corrosion and so became disused.
There had been repeated internal and external reports on both the fire hazards, and the lack of training in emergency procedures – including concerns raised by the London Fire Brigade.
It seems that management were complacent, and had been lulled into a false sense of security, as no previous escalator fire had caused a death.
Desmond Fennell, QC, from the Inquiry, concluded that “It is my view that a disaster was foreseeable”. In the last 30 years, I’ve heard this conclusion echoed in many other disasters.
Failure to learn
The possibility of a serious fire on the wooden escalators, as revealed by many earlier incidents, was a blind spot in the organisation. In particular, the danger from smoke inhalation had been ignored. Leadership had little interest in learning from previous serious fires – the failure to learn from these events was a significant factor in the 1987 disaster.
“Recommendations from internal inquiries into accidents either did not reach the right people or were not acted on or seen through”.“Fennell Report”, 1988, p.127
In August 1987, a few months before the disaster, an internal London Underground memo (below) recognised that a proactive approach to safety management was required. It is unfortunate that this was not the pervading view in the organisation, otherwise the disaster may have been averted. Clearly, the absence of fires causing death in the past is not a reliable guide to whether there might be such fires in the future. Over 30 years later, this internal memo still provides good advice today:
“A safe environment is not one in which there is an absence or a low number of serious injury incidents, but is the result of active participation by management and staff in identifying hazards and then doing something positive about them. In other words, the absence of accidents is a negative measure largely dependent on luck, while the identification then prompt elimination or control of hazards is a positive step and is essential to the discharge of our duties under current legislation”, Senior Personnel Manager (Operations), London Underground, August 1987.
There was no system in place to conduct or learn from safety audits. The Inquiry concluded that if the financial state of a company can be assessed by a financial audit, then the state of safety can be similarly established by a safety audit.
Due to the history of fires on the Underground, the London Fire Brigade wrote to the Operations Director in August 1985, requesting that the Brigade be called immediately to any fire on the underground railway network. This advice was not taken and the Underground continued with a two-stage procedure (i.e. staff attempt to fight fires themselves, and only calling the fire brigade if thought necessary). Notably, none of the fire extinguishers were used by London Underground staff in this disaster.
The Inquiry concluded that London Underground staff had not been adequately trained in emergency response or evacuation, and that there was little supervision.
Their response to the incident was uncoordinated and haphazard.
There was considerable delay before staff notified Station management. Staff were woefully ill-equipped to deal with the emergency, doing their best using common sense in the absence of training and supervision. The Inquiry concluded that better training of staff is the single most important recommendation.
“It seems to me that the staff were totally unprepared to meet the disaster which happened that night and had to do the best they could in the circumstances”.“Fennell Report”, 1988, p.64
As fires were clearly not unusual, you may have expected a robust approach to incident response. Unfortunately that wasn’t the case. There had been no thought into planning for the response to a major fire, there was no emergency response plan and no evacuation plan.
The Station Operations Room was no longer staffed, contributing to a lack of communications and control. In any case, much of the equipment was out of service, including many CCTV cameras.
Most senior managers didn’t realise the seriousness of the incident until it was almost over. The Inquiry stated that there was a general failure to appreciate the severity of the disaster and therefore a failure to act with the appropriate sense of urgency.
The lack of incident control allowed trains to continue to arrive and offload more passengers towards the fire. Communications with the general public were woefully inadequate throughout the incident – the Public Address system was not used and many passengers were evacuated into the line of fire.
London Fire Brigade were not met on arrival (indeed, there was no agreed rendezvous point) and they were unsure of the station layout. Senior Underground managers did not liaise early with the Brigade, or offer their detailed knowledge of the Station layout. The firefighters underground had no means of communicating with their colleagues on the surface. (Twenty years later, following several suicide bombs on London Underground trains, the same communications issues were raised).
There were many lost opportunities to exchange vital information between London Underground and the Fire Brigade; as well as between the fire, police and ambulance services.
It seems that staff and the response teams had been set up to fail.
Wider organisational issues
How did London Underground get to a position where fires were considered to be inevitable, but not managed? There are several wider organisational causes, relevant to both the incident and the emergency response. It’s interesting to read in the final investigation report how the Inquiry concentrated upon “the system in place which allowed the disaster to occur rather than seeking to make personal judgements upon the people involved“. This approach was quite possibly ahead of its time – I’ve seen many recent investigations that have taken a less mature approach.
Clearly, staff were totally unprepared to meet the disaster, but several aspects of the system enabled the disaster to occur:
- The “silo” structure of the organisation prevented effective internal communication. Surprisingly, no organisational charts existed – they had to be created for the investigation.
- London Underground was unwilling to take advice from outside, such as from the London Fire Brigade or the Health and Safety Executive. It was an inward-looking organisation. Outside appointments were rare.
- Not only did organisational changes create confusion and hamper communications, managing the many changes used up valuable resources. (I’ve written elsewhere on how to manage organisational change).
- No single person was charged with overall responsibility for safety.
- The Directors of both the Safety and Personnel Functions did not see passenger safety as being a part of their job. They were concerned only with the safety of staff.
Finally, the Inquiry found the organisation to be focused on financial matters at the expense of everything else.
The role of the Regulator was considered by the Inquiry: in this case the Railway Inspectorate. It was concluded that the Regulator misunderstood its responsibilities under health and safety law. Its relationship with London Underground was considered to be ‘too informal’ and did not create the ‘tension’ necessary to ensure safety. There was also little liaison between the Inspectorate and the London Fire Brigade.
The quote below has been referred to many times since, and neatly sums up the importance of these wider organisational issues:
Regardless of the industry, the human and organisational factors that “set people up to fail” are the same, and in the last 30 years the organisational causes of the King’s Cross Fire have been repeated in events across many different industries.
Disasters like the King’s Cross Fire are foreseeable; complacency and assumptions can kill; you get the system that you design and staff have to be prepared to respond, should the worst happen.
- Do you challenge the assumptions that you make?
- Do you question the way that you do things?
- What does your organisation accept as “inevitable”?
- How do you know that your emergency response will be effective when it’s needed?
Finally, I’d like to note that the Inquiry recognised the dedication of response teams, especially the London Fire Brigade:
“It is clear that a large number of members of the London Fire Brigade behaved with conspicuous courage and devotion to duty during the disaster in which they lost a very brave officer, Station Officer Townsley”Fennell Report, 1988, p.83
London Thames News, originally broadcast on 10 November 1988, following the launch of the Fennell report into the King’s Cross disaster:
London BBC News, broadcast 30 years after the King’s Cross Fire: