Thursday, 13 July 2017

Summerland and Grenfell Tower: How to Learn A Lesson And Then Forget It.

The Summerland disaster involving a leisure centre on the Isle of Man, occurred on the night of the 2nd of August 1973, about a year before Grenfell Tower was opened. As built, Grenfell Tower had none of the key defects of the Summerland building. The public inquiry was opened in September 1973 and concluded in February 1974, which is the sort of prompt and efficient timescale that the Grenfell Tower residents seem to be looking for. The inquiry's basic conclusion was "misadventure" but it did blame the materials used in the novel building, which was also criticised for having "chimneys" in the form of atrium areas where there was open space from the ground floor to the transparent roof. 

Subsequently, information emerged that there had been prior knowledge that two of the materials used in the building were combustible, and that a waiver from building regulations had been sought for one of them "Galbestos." Other combustible materials, such as softwood instead of concrete, bituminous felt, logs and even polystyrene, just found their way into the actual building (but not the approved plan!) often as part of the decor, decided at the last minute. There are drawbacks to holding the inquiry too promptly and too efficiently! The truth sometimes takes time to percolate to the surface in these cases.

The Summerland disaster claimed between 50 and 53 lives: this should be taken as a warning that the ultimate death toll for the Grenfell Fire disaster may be expressed with a similar margin of error. The Wikipedia article mentions that there were 80 seriously injured: it does not mention that amongst the nearly 3,000 survivors, there were many less serious, but excruciatingly painful, injuries caused by drops of molten plastic falling on the crowds all trying to flee through the only unlocked avenue of escape. Some of these people might have been considered serious injury cases if the hospital had not been forced to confine itself to treating only the worst cases. Lesser casualties were treated by the island's citizens (at the request of desperate police officers), at the scene and in their own homes, and probably aren't recorded in any official way at all. Many were subsequently evacuated for treatment in Liverpool and Blackpool.

Given the difference between the emergency resources available on the IoM in 1973 and those available in London in 2017, the IoM shouldn't really be criticised  for the response, which did include relief fire crews being brought over from the mainland by the Steam Packet Company, which is something else not mentioned in the Wikipedia article. There was a delay in reporting the fire to the authorities, though, for which the Leisure Centre's management really should have censured.

Three types of material were blamed for the Summerland fire:

"Oroglas" a type of transparent acrylic sheeting; there were also opaque or "smoked" versions of similar material, used throughout the seventies and early eighties for everything from rulers, plates and coffee tables to large buildings, such as the "old" Bletchley Leisure Centre (replaced in 2009 by a new one.)

"Colour Galbestos" described in a report by the University of Birmingham as "plastic-coated steel" and by Wikipedia as steel coated with asbestos and bitumen. The tradename heavily implies an asbestos material, which people would associate with fireproof material, but in practice the outer layer of this stuff seems to have burned up and across in a self-supporting fashion -and enough heat penetrated through to ignite soundproofing material on the inside of the building. Steel does not burn easily, except in pure oxygen, but it does conduct heat.

"Soundproofing material" apparently polystyrene (according to the University of Birmingham). 

There was also a disco with a fairly improvised reflective ceiling. There were bean bags stuffed with polystyrene beads in the sunbathing facility. There was a rustic staircase built of logs and an artificial waterfall that was actually made of bituminous felt. Apart from the obvious fire risks, people in the (concrete) basement disco simply didn't believe a person who warned them of the fire, because nothing could be heard. It was only when the same person reappeared with singed clothes that the disco was evacuated. At Summerland, as at Grenfell Tower, alarms did not sound.

The building contained a safety feature that was never mentioned in the public inquiry or press reports at the time: the bottom ten feet of the transparent Oroglas walls were actually made of conventional tempered silica glass in hardwood frames, because the building operators and the architect were worried about the sort of vandalism that had happened on the mainland, with small fires being kindled against the walls of public buildings. This possibility was eerily similar to what actually happened, although the small external fire in question was accidental.

The actual fire: this started in a small plastic kiosk outside the main building, close to the part of the wall that was composed of "Colour Galbestos." Prefabricated kiosks made of moulded plastic (polypropylene?) sections were a common solution in the seventies, when a place to take money and issue tickets was needed in the open air, for part of the year. They were to be found all over the country, in public parks and in zoos and safari parks. If this one had been a few feet further away from the main building, the disaster would not have occurred (at least not in the form that it did.)

Three boys broke into the kiosk (the lock was a pretty token affair, as witnessed by the minimal compensation magistrates imposed on them for breaking it) and had an illicit smoke. A fire started, by what the authorities accepted was an accident. The only criminal charge raised against the boys was for breaking the lock, because that was the only deliberate harm that they did. The fire took hold of the materials inside the kiosk, and the material of which the kiosk was made, quite rapidly, and as it did so, the plastic structure slumped in the heat, and crucially, it slumped against the Galbestos wall, the outer layer of which caught fire, the fire spreading across it. The heat of the fire on the outside layer penetrated the steel core and ignited polystyrene soundproofing on the inside. 

This internal fire spread even more quickly that the fire on the outside layer of the Galbestos, and communicated with the "Oroglas" transparent wall, above the ten foot high firebreak of non-flammable silica glass. From this point on, things happened very quickly:

The fire raced up the Oroglas wall and into the Oroglas roof, where it rapidly spread across the whole roof area, raining drops of burning plastic down through the building's atrium areas onto the 3,000 people trying to escape. (The fire should have been detected and an evacuation commenced before this point, if the management and fire detection systems had been up to scratch.) The shower of burning droplets also started fires down throughout the building, except for the basement levels, which had concrete floors. (There should have been concrete floors at high levels, too, but these were replaced with softwood during construction, without those who'd passed the original plan being told of the change. One man thought his memory was playing tricks when he saw the wood!) The Oroglas panels spread the fire in two ways, therefore: from one panel to the next and by raining fire on everything below. The untested theory behind the structure was that in a fire, Oroglas panels would fall out of their frames from the heat before they ignited, this definitely did not happen.

Relevance to Grenfell Tower: First and foremost, neither building at the time of its fire, reflected the design and specifications which had been examined in detail and approved by planners and regulators. Summerland was heavily modified on the hoof as it was built, Grenfell Tower was built to specification, but modified and its fire-resistance completely compromised, several decades after it was built. 

Differences are that at the time Summerland was built, there wasn't a widespread awareness of the flammability of Oroglas and Galbestos, although something was known to some of the people involved in both cases: the manufacturer of Oroglas acrylic sheeting had admitted that it could prove highly flammable in some conditions, and a waiver to building regulations had been sought for Galbestos because it was known not to pass two of the tests specified in building regulations. (Allegedly, in the regulation-compliant plan which the Manx Parliament actually saw, the Galbestos wall was a concrete wall, and if it weren't for the waiver, which the Parliament didn't know about, it would have stayed a concrete wall.)

At the time Grenfell Tower was being built, the Ronan Point disaster had already turned the spotlight on building standards for high-rise buildings, and by the time it was completed there was great public awareness of the danger posed by careless choices of plastic panel materials, which had led to the Summerland disaster. The Daily Mail righteously declared that no structure such as Summerland could have been built in England, and at the time Grenfell Tower was built, this was true.

However, by the time Grenfell Tower was modified, with the addition of external cladding, it clearly was possible for a duplicate of Summerland to be created in England, because this is precisely what Grenfell Tower then became, in terms of flammability.  In the intervening years, the lessons about design integrity matching building integrity, which had been genuinely and sincerely learned in the aftermath of Ronan Point and Summerland, had been completely forgotten.

There is also a striking parallel between Galbestos, a thin sheet of steel covered in plastic or bitumen (depending on which source you believe) and "ACM", a sheet of aluminium covered in plastic, as in the skins of the cladding panels fitted to Grenfell Tower and, it now appears, a great many other buildings. Since aluminium has both a greater thermal conductivity than steel and a much lower melting point, it is screamingly obvious that if a small fire could be transmitted through steel-based Galbestos, a small fire can get through aluminium-based ACM without any difficulty whatsoever.

No comments: