What the BRE knew but did not say: part 1, 'nothing.. to add'?

The 1999 ETR Committee Inquiry on the fire safety of cladding systems received written evidence from the FBU, the CWCT, the FSDG (Fire Safety Development Group), Eternit UK Ltd, the Loss Prevention Council, Mark Heywood Associates, Matthew Smyth of Smyth Plastics Ltd, International Fire Consultants Ltd, thirteen local authorities, and the DETR:

but none from the one body probably best placed to give it, the Fire Research Station, which had been part of the privatised Building Research Establishment since 1997. Why was that?

An email from Peter Field, the Deputy Director of the FRS, to Tony Edwards of the DETR, on 7 July 1999, 13 days before the Committee took oral evidence from witnesses, sheds light on the question. Field had just been approached 'again' by the Inquiry Secretary and asked whether the BRE intended to submit evidence:

He had replied that the BRE had nothing to add that the evidence that the DETR 'proposed to give':

which BRE must presumably have seen.

Oddly, Field asks Edwards to let him know of any 'developments' in order to avoid potential 'conflicts of interest':

The most common sense of 'conflict of interest', where someone's personal interests prevent the the same individual discharging their official responsibilities in an even-handed way, doesn't seem to fit here. But the term can also be used simply of a divergence of interests between two different parties:

causing them to pull in different directions. But why shouldn't the BRE pull in a different direction to the government on occasion? Isn't that the very essence of being an 'independent' research, testing and consultancy body, as it claims to be?

Edwards agreed that this was the 'right approach', that is, for the country's leading fire research organisation not to give the benefits of its expertise to the Parliamentary Committee:

He would give him a copy of the final version of the DETR's submission to the Committee, which he himself had prepared ('my... submission'):

'Brian' is certainly Brian Martin, who was on secondment to the DETR, embedded within it 2 to 3 days a week, working in support of Tony Edwards:

In reality, BRE had a great deal not only to add, but also to subtract, from the DETR's submission. Their view, as implied even in the oral evidence that Field later gave to the Committee, was that Class 0 was inadequate. The Department claimed the contrary, and went on to reject the Committee's recommendation to change the guidance. Why did the BRE fail to explain the basis of their concerns about Class 0, supported as they were both by theoretical considerations and by extensive evidence from large-scale testing, in a written submission?

In this article, I examine the DETR's written submission to the 1999 Inquiry, and describe briefly some of the experimental results from 1993-4 that should have led BRE to make their own written submission to correct the government claim that the existing approach to cladding fire safety regulation was adequate. I intend in my next article to examine that earlier research in more detail.

The DETR's written evidence

The Department understood (p. 26) the need for 'high standards against fire propagation and spread of flame' where the height above ground would make fire fighting difficult:

Remarkably, it then stated that in high buildings, the combustibility of materials used in the external wall construction 'needs' to be limited, and went on to explain the concept of 'limited combustibility' as it used in the Approved Document:

I think it is reasonable for the reader to infer that the Department's opinion was that materials used in the external wall construction needed to be of 'limited combustibility', as defined in the Approved Document. I think it is unreasonable to expect the reader to consider the possibility that two entirely different concepts of the limitation of combustibility were being juxtaposed in the same paragraph.

The then current 1992 AD B required that 'insulation material' on high buildings should be of 'limited combustibility':

while 'external surfaces' at height were to be Class 0:

The innermost surface of the external wall was of course also subject to the requirements for internal linings - generally Class 1 for rooms and Class 0 for circulation surfaces:

It strikes me as misleading to point out that plasterboard is typically of 'limited combustibility':

from which one naturally I think infers that it was required to be such, and very possibly gains the impression that all the materials in the external wall were required to be of limited combustibility. But plasterboard is not an insulation material and was therefore not subject to that requirement.

I think it was also potentially misleading to make particular mention of insulation materials in regard to the general need for external wall materials in high buildings to be limited in combustibility:

tending to give the impression, I think, that what was required of insulation materials was required of all the materials.

In reality, the previous requirement for external walls of high buildings to be of limited combustibility (apart from cladding and internal linings), to be found in AD B 1985:

and dating back to a similar statutory requirement in the first national building Regulations of 1965:

had been dropped in 1992, allowing timber-frame (and polymer composite) construction for high buildings in England (for the first time since 1667, if I am not mistaken:)

as pointed out by Brian Martin in his evidence to the Inquiry:

but not by Colin Todd, in his expert report to the Inquiry, in which he did draw attention to the limited combustibility requirement in the 1985 AD:

but not to its removal from the 1992 AD:

merely reciting the new text:

which a) replaced the limited combustibility requirement with a general caution about the use of combustible materials, whose scope b) was itself restricted to insulation materials and cladding framework only, leaving the use of combustible materials for structural frame, backing boards, membranes, and so on, without even a caution:

The limited combustibility requirement, which followed this text, applied only to insulation materials, as shown above. Northern Ireland was unique amongst the jurisdictions, from 1994 onwards, in making both insulation materials and cladding framework subject to the limited combustibility requirement:

The account given by DETR of this section of the guidance in its submission to the 1999 Inquiry seems to contain faulty reasoning:

The 'therefore' is misplaced. It cannot follow from the risks of combustible insulation and cladding framework materials that insulation materials only, and not framework, have a limited combustibility requirement. For there to be logic, there must have been a further reason to impose the requirement on insulation only, whether it be that the fire risk was seen as greater, or the cost of a restriction to limited combustibility materials was considered to be less.

Inaccuracy concerning classification of 'surfaces'

The DETR claimed that 'surfaces', whether of 'materials' or whole 'cladding systems', were classified by reference to BS 476-7 (Surface Spread of Flame) and BS 476-6 (Fire Propagation):

But in fact both BS 476-7 and BS 476-6 define tests of products, not of surfaces, as even their titles show:

a product being defined as a 'material, composite or assembly':

The 1992 edition of Approved Document B required the 'external surfaces' of walls to meet the provisions of Diagram 36:

which itself was titled 'Provisions for external surfaces of walls':

and referred to Class 0, and BS 476-6 indices, as 'wall surface classifications':

whereas the 1985 edition required the 'cladding' at height on high buildings to be Class 0:

which was a meaningful requirement, since there exist cladding 'products' which can be tested to BS 476-6 and 476-7. Even an 'assembly', defined as a 'fabrication of materials and/or composites that can contain air gaps':

could be tested, within the limits set by the dimensions of the test specimen, which for part 6 was to be 225mm square, with a maximum thickness of 50mm:

Industrial practice continued as before after 1992, as Jonathan Evans has also discussed, since the test laboratories were subject to the standards, and therefore were constrained to test 'products' rather than 'surfaces', even if 'surface' could be defined. It seems however that key members of BRE believed (at least according to their evidence to the Grenfell Inquiry) that the AD B2 12.6 (2006) Class 0 requirement for the external wall surface at height on high buildings was intended to ensure good spread of flame characteristics of the surface coating, for example of ACM panels, while the limited combustibility requirement of 12.7 was intended to restrict the contribution to a fire of the core of the same product. For example, Debbie Smith agreed with Counsel's summary of her view as being that the core of a panel had to be of limited combustibility, while the 'surface' only had to meet Class 0 or Class B:

But who has ever tested an ACM core to ascertain whether it is of limited combustibility? An A2 ACM panel does not necessarily have an A2 core. The core, rather than the panel with its aluminium facing, would have to be subjected to the SBI test, and I am personally not aware of any instances of this being done prior to the Grenfell fire. (I understand that one major ACM company did classify their ACM core to EN 13501-1 after they saw DCLG post-Grenfell 'interpretation' of 12.7.) Would the 'filler' include the adhesive that binds the core to the aluminium facings? If so, with the flame attacking the adhesive, would it achieve FIGRA ≤ 120 W/s and THR[600s] ≤ 7.5 MJ, as required?

Calibration with real fires and large scale tests

The DETR claimed that the provisions expressed in terms of fire classifications derived from small scale tests had been calibrated with experience from real fires and data from larger scale tests (p. 26):

This is seriously misleading. The BRE's post-Knowsley test series, reported by Connolly in 1994 for the Department of Environment (DoE), included 6 cladding systems that were fully compliant with AD B2 (1992). As I will detail in my next article, the cladding sheets were Class 0 on both sides, the insulation was mineral wool and of limited combustibility, there were cavity barriers at every floor level, and even the caution about combustible framing was heeded, the support rails being aluminium. All six failed to meet the fire safety criteria set by Connolly, and by a large margin in at least four cases. The small-scale tests had indeed been 'calibrated' with larger scale fire research, and they had been found to be inadequate for external cladding, as both DoE and BRE knew.

Review of AD B (1992)

A review of the existing guidance was underway, but the only change to the cladding guidance the Department was intending to make was to adjust the definition of high buildings from > 20m to > 18m:

In fact, the AD B (2000) amendments included a major weakening of the cladding guidance. Whereas insulation material 'used in the external wall construction' had previously been required to be of limited combustibility, as shown above, it was now only insulation material 'used in ventilated cavities in the external wall construction' that were to meet this requirement:

The door was thereby opened for the employment of combustible insulation on high rise in sandwich panels (as at The Edge, for example), spandrel panels, and rendered systems. To my knowledge, the Grenfell Inquiry has not yet investigated why, and at whose instigation, this relaxation was made. If the DETR was telling the truth in its written evidence to the ETR Committee, then the proposals must have been revised after the submission of that evidence in July 1999.

The next year, in his March 2000 reply to the ETR Committee Recommendations, the Minister Nick Raynsford claimed that there had been 'no suggestion' during the review that the cladding guidance was inadequate:

This was seriously misleading in my view. AD B had been reviewed by the Building Regulations Advisory Committee (BRAC) following a public consultation:

It seems to me that the consultation was an integral part of the review process, and thus of the 'review' in ordinary parlance. We now know (thanks to Ian Abley's researches) that the Fire Safety Development Group (FSDG), in its response to the consultation, had called for 'substantial revision' of the guidance for external wall surfaces. The Class 0 requirement for the outer cladding material was 'entirely inappropriate':

It may be noted that the DETR's 2000 relaxation of the guidance to allow render systems with combustible insulation, discussed above, flew in the face of the FSDG particular warning about such systems, which had continued to be installed on the basis of BBA certificates contrary to the AD B guidance, as I have shown elsewhere. The FSDG did not here draw specific attention to the danger of combustible Class 0 panels, such as those used on Knowsley Heights.

The DETR claimed that the 'risk of serious fire spread' would be

'minimal'

if the AD B guidance was followed:

Ray Connolly had followed the AD B guidance in six of the systems he tested during 1993-4. According to his criteria for an acceptably safe system, flame could extend from the fire on the ground floor up the outside of the first storey but should not 'rise to the second floor level':

In four of these six systems, flame reached not only the second floor, but the third floor, and then not only the third floor but right to the top of the building, even in two tests emerging out of the top of the building, as for example in test 7:

with Connolly commenting that fire would have spread to further storey levels if a taller test building had been constructed:

How, may I ask BRE and DETR (DLUHC), could the 'risk of serious fire spread via external cladding' from such a system be considered to be 'minimal'? Since government had failed to act on the results of research they had commissioned BRE to undertake, then BRE should have made use of their independence to warn the ETR Committee of the danger they had exposed.

Andrew Chapman