Famous fires

Some of the best summaries of fires which have altered design and other advice are found at
Gordon Cooke’s summaries stop however at 2006. Doubtless Grenfell will alter design rules in time.

This page has my own take on a few in factories or where plastics and other materials with unusual properties have played a part.
The fire at Eastwood textile mill in 1956 at Keighley, Yorkshire is notable because it led directly to a revision of the Factories Act that the Factory Inspectors enforced, with a new Act in 1959, quickly incorporated into a larger revision as the 1961 Act, still being actively enforced when I joined HSE in 1989

Summerland, Isle of Man 1973
This is just one example in a long series, where UK fire legislation was modified following a disaster, the series starts with the first building codes following the Great Fire of London in 1666.

The tragedy reminds us of some continuing truths and because so much is easily forgotten, I have provided the text of the many recommendations that are still relevant here
Most fires in buildings start with the contents. If at an early stage the building itself is fuel for the fire, you are likely to be in big trouble. Bitumen coated panels, and at a later stage in the fire, plastic glazing were significant elements causing the fire to spread quickly at Summerland.

Small scale fire tests on solid materials often do not give you all the information you need to know, to assess the potential fire growth rate. Geometric factors, which influence surface spread of flame and the growth of fire by combinations of conduction, convection and radiation cannot be modelled by the cone calorimeter. Smoke yields from some plastics are much greater than from cellulosic products. For HSE I was involved in the 1990s looking at medium scale fire tests for materials with unusual fire properties found in bulk quantities on industrial premises. Examples were rolls of polyester wadding, polyurethane and polyethylene foamed plastics, rolls of carpet. The test rig is no more, and the guidance that came from it has been largely forgotten.

Fire spread through hidden voids within buildings are particularly dangerous,as fire can grow substantially before it is detected. In industrial premises, the equivalent problems comes from elevators and ductwork for pneumatic or mechanical conveyors, which may extend through compartment walls provided for limiting fire spread, or between floors.

The behaviour of large groups of people when the fire alarm goes off is far from simple. They may be reluctant to stop what they are doing, go in search of other family members, or simply not understand why they should respond. They are liable to try and leave the way they came in. Much has been done to understand human behaviour in the fire situation, and evacuation models give us a much better idea of the timescales needed to clear a building, but in places of assembly the role of staff remains important.

In public buildings in particular, delays in raising the alarm and locked fire doors are simply unforgivable. While a building is fully occupied, there is never an excuse.

Maysfield Leisure centre
The fire at the Maysfield leisure centre in Belfast in January 1984 was in many ways a one-off. In single story premises, 6 people died as a result of a fire that was started maliciously in a pile of foam rubber gym mats. The rapid rate of fire growth, and the very large amount of smoke generated by the fire were factors in the loss of life, but delays in raising the alarm were also a major issue.The fire came at a time when fire brigades were becoming worried by domestic fire deaths from foam furniture which had the same characteristics as the mats.

As a consequence the hazard was tackled by seeking to enforce storage of mats in fire resistant stores, provide automatic smoke detection, and make the mats more resistant to ignition. The ignition tests draw heavily on the tests developed to make foam furniture more resistant to ignition, first from cigarettes, and then from matches.

Fire involving gym mats does not seem to concern countries except the UK, so the fire tests are not contained in European standard EN 12503. However, they are given in a cross reference to BS 1892, the original version of which dates from 1990, and was a direct result of the Maysfield fire. The current BS sets out two standards of performance, based on tests which assess ignition resistance when exposed to a smouldering cigarette (source 0), a small butane gas flame for 70s (source 3), or a wooden crib which is equivalent to a few sheets of screwed up newspaper (source 5) . Two performance standards are given, resistance to sources 0 and 3, (low performance) and resistance to sources 0 and 5 ( high performance). BS 1892 also recommends that mats are stored in fire resisting storerooms. Depending on the fire risk assessment, the user may decide that low fire resistance mats are acceptable, or in higher risk circumstances, that they need mats which achieve the higher performance level.

Dudgeons Wharf
This is an explosion incident that followed a fire, rather than simply a fire. The disaster occurred at a demolition site at the Isle of Dogs on the Thames Estuary on 17th July 1969. The site comprised a tank farm that was being demolished. Not too surprisingly, the records of what the tanks had contained were patchy or entirely missing. The demolition contractors were unaware of the potential hazards. The work continued over a period of weeks, during which a number of fires broke out, and were put out, by the local fire brigade. The local acting principal factory inspector was consulted about the sites, and gave rather weak advice, when he should probably have passed the query on to specialists in the London HQ. Eight officer grade members of London FB attended at different times, and the acting principal factory inspector. It seems clear that none really appreciated the risks of an explosion. Lines of responsibility for those undertaking the actual demolition work were very muddled.

On the fateful day, flames emerged from tank 97 while the top was being burnt off. No one before or subsequently had tried either to visually inspect the amount of residual oil or solid material after it had been steamed. Neither did anyone at any stage try and measure the atmosphere inside the tank for the presence of flammable vapour. It appeared the fire was substantially out by the time the fire brigade responded to a call, but 5 members of the London fire brigade were stationed on the tank roof with their hose lines, ready to extinguish any residual fire inside the tank. With the station officer in attendance, the demolition contractors tried to shift some bolts on an access hatch at ground level and when they could not be turned, they applied a blow torch. The tank exploded, 5 firemen were blown high into the air, and died immediately. One demolition worker was also killed. The recriminations then started. The official report makes much of the composition of the residue inside the tank.

There had been many previous examples of tank explosions caused by hot work on tanks that had not been properly cleaned, and many since then. The chances of a fatality are always high, because the holder of the torch is so close to the explosion. What Dudgeons Wharf reminds us, is that measuring the atmosphere inside the tank, on its own, is not enough. The heat from the torch is liable to generate vapours, either from products with a low volatility at ambient conditions, or chemical decomposition of polymers, sludges and residues at the high temperatures used. Cold cutting is best if at all possible. Failing that a detailed visual inspection of the inside for residues is essential.
A copy of the report from the public inquiry is available on request.