Engineers need to study disasters. Disasters compress a lifetime's worth of hard lessons into a few hours or even moments. They show us the limits of our knowledge, and they make plain the consequences of shoddy decisions, easy arrogance, greed, laziness, and above all ignorance. In these failures are concealed every lesson that could ever be needed. They teach of technical issues, human failing, and organisational deficiency. Yet, when politicians say “lessons must be learned”, we shudder; for we know that signals another failure to come.
On the UK Column News on Thursday 15th June, we looked at the immediate cause of the rapid fire spread and huge (and still as I write unknown or undisclosed) loss of life. We identified the flammability of the outer cladding as the cause of the rapid fire spread and the smoke generated by the PIR insulation as the main contributory factor. These were well-known risks. None of this was new information:
The outer cladding was constructed of a thin sandwich of aluminium with a low-density polyethylene core. Polyethylene is flammable and melts at a relatively low temperature; the following video illustrates exactly the problem at Grenfell Tower
To move further, it is necessary to have a conceptual understanding of how disasters strike complex systems and mature industries. There is a simple model of this, called the Swiss Cheese Model; it was developed by Professor James Reason and is described in his book Human Error. This book bridges the gap between risks inherent in science and technology and understanding the human psychology of error. The author develops this into a general view of accident causation in complex systems. Put simply, the Swiss Cheese Model views the protective systems that prevent an accident as a series of barriers that stop unsafe events progressing. But these barriers have holes in them, like a Swiss cheese. Only on those occasions when all the holes line up can an unsafe event progress all the way to become a disaster.
There is a brief explanation with diagrams at the Duke University School of Medicine.
Professor Reason looks not just at unsafe acts but at the psychological precursors to those acts, the management failures, and the influence of the decision-makers. By this method, a full picture is revealed.
To illustrate where this approach might lead, we shall look again at just one of the materials we identified in our report of 15th June as being most directly involved: the outer cladding. What follows is an illustration of what needs to be done, nothing more. It is simply a starting point, an exhaustive examination of the issues requires a full inquest.
The cladding material was a relatively new construction innovation called Aluminium Composite Material or ACM. This is a thin rainscreen cladding. Several firms manufacture essentially identical products; in this case the supplier was the American giant corporation Alcoa and the product name was Reynobond.
First we must understand what this product is. It is made up of several layers, outer-paint finishes applied to 0.5mm thick aluminium with a polymer core. The core is available in several thicknesses and in several grades. The basic brochure gives a “technical overview” that includes the following table:
This would give the impression that all grades are similar in terms of fire safety; however, as we will see, that is not the position of Alcoa.
So we have one barrier against error that is completely ineffective: the manufacturer's sales literature – which is likely the first thing a specifier would read when considering the product. It does not identify the difference between the various grades of material in terms of fire safety, nor does it make clear the importance of the specifier's decision.
What are that differences between the various grades of the material? For that we go to another Alcoa document, also available on their website (but it took a while to find). This is a brochure entitled “Fire Safety in High-rise buildings; Our Fire Solutions".
This is the key information published by the manufacturer; did you understand what it is saying? If you did not, I am not surprised, because the key point is the small print below the building diagram. Let us look more closely:
So we learn that due to “accessibility of the fire brigade”, buildings more than 10m (3 storeys) high must be “conceived” with an incombustible material. Strange wording. What this diagram shows and the text fails to match is that:
1. Above 10 m / three storeys, the most commonly used product with a polyethylene core (PE) is actually not suitable and must not be used.
2. The fire-resistant (FR) core material option is good for buildings up to around 30m or eight to ten storeys, but no higher.
3. For high-rise buildings, the non-combustible mineral cored product is compulsory; any other option is dangerous.
On first reading, did you understand that? The wording is so poor that I would expect many readers to miss the significance. So here we have a second safeguard compromised: the manufacturer's fire safety document does not make the risks of using the wrong product clear; it implies it is to do with fire-fighter entry and conceals the essential issue- the flammability of PE.
And this is the only mention of PE in this brochure – it is not mentioned in the fire test data, because it is so flammable it would be pointless, and probably dangerous, to test. So it is simply omitted. What exactly is this document trying to achieve? Safety or sales? What psychology was it that produced this brochure, what management policies compromised the communication of a vital safety message?
Moving on from the cladding supplier, the next step in the chain is usually the architect, in this case a firm called Studio E LLP.
Here is their statement regarding the Grenfell Tragedy:
So we don’t know exactly how this material was specified; it may have been a full specification of the product by the architect (the more likely option) or a statement of requirements to be fulfilled, a so-called performance specification. If the specification was for the material that burned so quickly, that will raise many other questions: How was the practice managed? What was the resource allocation to the project? How experienced were the staff? What knowledge did they have of fire safety?
And this, in turn, will lead us to architect's education and the deficiencies therein; the lack of technical understanding; the inadequate understanding of fire safety, replaced by rote learning of building regulations, often without deeper analysis. Add to this the scope for inexperienced architects to specify materials on projects that are beyond their competence. The architectural profession will have serious questions to answer.
When looking at architectural drawing office practice, an obvious next step will be to ask, "Where would an architect look for guidance on product performance?" Well, the industry's best-respected source for such data is an organisation called the British Board of Agrement or BBA. They certify product performance, and their certificate for this product is available online.
Here is what the best-respected industry source says on the fire performance of the Grenfell cladding:
Did this mislead the specifier? It is most certainly another of the Swiss-cheese holes in alignment, for this is offering no warning at all. It gets worse – when the fire section is consulted:
This states that the standard LDPE product that caused the rapid spread of the fire at Grenfell Tower is little different from the fire-resistant grade. Both meet classification B of EN 13501-1, the relevant European standard. This means they are judged to be “Hardly Combustible” and to make only a “Limited Contribution” to the fire. I’ll just let that one sink in.
And the manufacturer's recommendation that the standard panels should be limited to buildings of 10m or less? That is nowhere to be found; it is not reproduced in the BBA certificate.
And this, in turn, leads us to the testing. How accurate are the the fire tests which inform organisations such as the BBA; how suitable are they for modern methods of construction; what do they really tell us?
Well, in the case of the British Standard test for spread of flame, not much. It is not really a fire test but a test on the surface paint to limit the rate of spread and the heat generated in a fire from the surface. Devised for fire safety on internal finishes within rooms, it neither assesses the amount of smoke given off by a product, nor identifies whether plastics drip from it when the material is heated. Crucially, it does not differentiate between combustible and incombustible materials. A "Class 0" rating puts many architects and specifiers to sleep, regulators too. But such a rating is related to a surface finish test that says little or nothing about cladding performance in a major fire.
The more modern European standard combines several separate tests to come up with a somewhat clearer picture, but, as we saw in the agreement certificate, this standard is still open to question as regards accuracy and relevance to external cladding.
Turning next to the regulators, we find that the traditional process of approving drawings was not carried out by the local authority; their web site states project is “completed, not approved”:
It would appear that rather than a building control officer based in the area, checking of the design was by one of more than 180 firms of approved inspectors operating across England. That arrangement will also need to be carefully examined.
And we have not yet considered the supply chain. A contractor does not stroll down to the local B&Q depot and pick up 100,000 square feet of ACM cladding. This was specially manufactured for Grenfell Tower. Each panel is shaped and dimensioned to fit. That means that a drawing office somewhere, probably one specialising in cladding contracts, drew every panel. Yet no-one noticed the grade was incorrect and that it did not meet the manufacturers recommendations.
Typically, an architect would lean heavily on the technical representative of a material supplier in a large contract such as this. What happened to that customary relationship? Did the representative give the correct advice or was he or she also ignorant of the situation?
These are just some of the issues that come out of an examination of the external cladding; and we have not even considered the PIR insulation yet, let alone wider issues. Wider issues such as firefighter response, training and policy; the decision not to employ sprinklers; problems with maintenance; the quality of the electrical infrastructure within the block, amongst a myriad of other issues.
But I hope this brief article has succeeded in illustrating the essential point: layer upon layer of protection against error are present, and it all failed simultaneously. No-one asked the necessary question, “Wait a minute, an LDPE core; isn’t that flammable?” Or if anyone did ask this question, no-one listened to them.
A full investigation will lead in many directions and may reveal systemic failure all across the construction industry. If we are to honour the memory of those who died, we must uncover the full story; all of it.