Chapter 20: Subsequent testing and marketing of Reynobond 55 PE

The Fred-Roderick Pohl presentation: September 2007

¶20.1 In September 2007, while Arconic was in the process of obtaining a BBA Certificate for Reynobond 55 PE, Gérard Sonntag, Arconic's Marketing Manager, and Didier Felder, its Area Sales Manager, attended a number of presentations organised by a metal and plastic distributor in Oslo. They included a presentation by Fred-Roderick Pohl, who had been invited by a manufacturing company, Otefal GmbH, for which he acted as a consultant.

¶20.2 In his report of the trip, Mr Sonntag recorded that Mr Pohl had explained the extreme danger presented by the use of ACM PE on residential buildings, comparing the combustibility of 5,000m2 of ACM PE to that of a truck containing 19,000 litres of oil, and drawing attention to the fact that even greater danger was posed by the toxic smoke emitted by burning polyethylene, which could kill in two or three minutes.¹⁵⁶¹⁵⁶ {META00001953/2}. It is quite possible that Mr Pohl was not entirely independent and was seeking to advance the benefits of pure aluminium by decrying the fire safety of ACM, but even making allowances for that, his presentation was graphic and, more importantly, was taken seriously by Mr Sonntag. Mr Sonntag had been particularly concerned by Mr Pohl's hypothetical question of what would happen if "one building made out [of] PE core is in [a] fire and will kill 60 to 70 persons, what is the responsibility of the ACM supplier?".¹⁵⁷¹⁵⁷ {META00001953/3}. Mr Sonntag's own recommendation in the light of that presentation was that Arconic ought to assess the financial consequences of selling only Reynobond with an FR core and in that case introduce a program of cost reduction to enable the FR version to be manufactured at the same cost as the PE version.

¶20.3 We have seen the slides from the presentation given by Mr Pohl,¹⁵⁸¹⁵⁸ {MET00081029}. which do indeed draw attention specifically to the combustibility of ACM PE¹⁵⁹¹⁵⁹ {MET00081029/4}. and the toxicity of the fumes emitted by burning polyethylene.¹⁶⁰¹⁶⁰ {MET00081029/14}. They also illustrate past fires in buildings covered in ACM cladding around the world, and refer specifically to the UK testing regime under BS 476.¹⁶¹¹⁶¹ {MET00081029/16}. Mr Sonntag also mentioned that Mr Wehrle had shown him a copy of a similar presentation made by Mr Pohl some two months previously. We have been unable to identify that document but have no reason to doubt what he said.

¶20.4 Despite those clear warnings and Mr Sonntag's reaction and recommendation, there is no evidence that anyone at Arconic took steps to examine the safety of Reynobond 55 PE or to ascertain the financial consequences of selling only panels with a fire-resistant core. However, shortly after Mr Sonntag made his report Arconic launched a project to reduce the cost of Reynobond FR to that of Reynobond PE in response to increasing demand for a fire-resistant product.¹⁶²¹⁶² Schmidt {MET00053187/14} page 14, paragraph 40. Despite various efforts over the ensuing years, however, the price of the fire-resistant version was never brought down to that of the unmodified version.¹⁶³¹⁶³ Schmidt {MET00053187/14-15} pages 14-15, paragraph 44.

¶20.5 In the light of Mr Sonntag's report, we think it is clear that by late 2007 Arconic had become aware that there was serious concern within the industry about the safety in fire of ACM panels with an unmodified polyethylene core (ACM PE) and had itself recognised the danger they posed. However, there is no evidence that Arconic took any steps to withdraw the sale of that product in the UK in the months and years that followed. On the contrary, not only did it continue to manufacture and sell ACM PE, it also sought to exploit what it perceived to be a weak regulatory regime in the UK while withholding from the market relevant information about the product's fire performance. That contrasts with the position in other countries, such as France, where, in 2016, Arconic insisted that its sales team specify only Reynobond 55 FR.

Internal and external communications: 2009 – 2010

¶20.6 On 17 July 2009, Mr Wehrle wrote to Claude Schmidt, Guy Scheidecker and Robert Quattrocchi (Arconic's plant manager) attaching some photographs of a building in Romania clad in ACM PE panels that had suffered a serious fire.¹⁶⁴¹⁶⁴ {MET00053158_P10/125}. Mr Wehrle wrote that the photographs demonstrated how dangerous ACM PE could be when it came to architecture.¹⁶⁵¹⁶⁵ {MET00053158_P10/122}.

¶20.7 On 16 March 2010, a sales executive, Isabel Moyses, wrote to Mr Scheidecker and Mr Wehrle, amongst others, telling them that a competitor had had to recommend fire-resistant panels for use in Spain because its standard ACM PE product did not have the classification of B-s3, d2 required under Spanish law.¹⁶⁶¹⁶⁶ {MET00064988/125}. She said, however, that Reynobond PE panels could still be used because they had achieved a classification of B-s2, d0. When he responded the following day, Mr Wehrle surmised that the competitor had based its view on tests of the product carried out in cassette form, which he acknowledged produced results much less favourable than those carried out on products in riveted form. He pointed out that in cassette form Reynobond PE did not meet European Class B, a matter, he said, that "we have to keep as "VERY CONFIDENTIAL!!!!".

¶20.8 Mr Scheidecker's response later the same day was much shorter: "[t]his shouldn't even have been mentioned".¹⁶⁷¹⁶⁷ {MET00064988/125}.

¶20.9 None of those emails were exhibited to Mr Wehrle's statement and he did not refer to them in any way. Since he did not give evidence, we have not had the benefit of any explanation of their contents he might have been able to give, so we are left to interpret them at face value. In our view they demonstrate clearly that by March 2010 Arconic was aware that the fire performance of ACM PE panels in cassette form was generally considerably worse than in riveted form and that test data relating to panels in riveted form did not apply to panels in cassette form. They also support the conclusion that the results of Test 5B had not been "rogue" results and that Arconic knew that any suggestion that Reynobond PE 55 in cassette form had achieved, or could achieve, Euroclass B was false. Finally, we think they show that Arconic was deliberately and dishonestly concealing from the market the true position in relation to Reynobond PE in cassette form. We were unable to invite Mr Wehrle to respond to those inferences, but Mr Schmidt was asked to do so and agreed that that had been the case.¹⁶⁸¹⁶⁸ Schmidt {Day93/51:8}-{Day93/55:21}.

¶20.10 In July 2010 a customer asked Ms Moyses for certification of the fire performance of Reynobond 55 PE in cassette form.¹⁶⁹¹⁶⁹ {MET00053158_P04/3}. She told him that Arconic had fire certification only for the riveted form, but that "soon" the cassette system would be better. Her message (written in English) is not entirely easy to understand, but when reporting to Mr Wehrle she said that she had told the customer that since the riveted form was the most unfavourable system, the certificate was acceptable for the cassette system. It seems, therefore, that her message had been intended to give the customer the impression that the cassette version performed better in fire than the riveted system. If so, that directly contradicted the results of Test 5B and the position as understood by Mr Wehrle. When the customer insisted on a certificate relating to the cassette form,¹⁷⁰¹⁷⁰ {MET00053158_P04/2}. Ms Moyses asked Mr Wehrle to "take care of it".¹⁷¹¹⁷¹ {MET00053158_P04/2}. Mr Wehrle told her that it was hard to commit anything to writing on the subject because, as he put it, Arconic was "not clean".

¶20.11 Mr Wehrle did not say in his witness statement what he had meant by the expression "not clean", but we think he must have been referring to the fact that Arconic had no certificate for Reynobond PE in cassette form and that its use of the certificate relating to riveted panels to support the use of cassette-fixed panels was not straightforward. Mr Wehrle's frankness with his own colleagues suggests that they knew as much as he did and that he had nothing to hide from them.

¶20.12 On 5 July 2010, Mr Wehrle wrote to the customer attaching what he described as "the document concerning the system we choose for the Reynobond PE fire certification".¹⁷²¹⁷² {MET00053158_P04/4}. It was a letter written by Mr Wehrle himself, in which he said:

"Reynobond PE, is classified B-s2, d0 in accordance with the European standard EN 13501. This test was done on the riveted system, which expose the core of the material to the flame contrary to the cassette system where the core is protected by the returns. Alcoa decided to check the behaviour of its composite panels in this worst case of system (exposed fasteners on flat panels) and to use it for all the other systems."¹⁷³¹⁷³ {MET00053158_P04/5}.

¶20.13 The assertion that Reynobond PE in cassette form would behave better in a fire than in the riveted form tested was not only false but was known by Mr Wehrle to be false. It demonstrates his willingness to resort to deliberate dishonesty in order to support the sale of the product for use in cassette form. Mr Schmidt, who was the only person who could be asked about it, accepted that Mr Wehrle had lied to the customer.¹⁷⁴¹⁷⁴ Schmidt {Day93/66:1-3}.

Further testing and review

CSTB Testing: June 2010

¶20.14 CSTB performed further tests of Reynobond 55 PE in riveted form in June 2010 and reported the results in February 2011. The grey riveted specimens tested achieved a classification of B-s1, d0, which was better than the results of Test 5A in 2005.¹⁷⁵¹⁷⁵ RA11-0032 classification report of 9 February 2011 {ARC00000383}; Test report of 9 February 2011 {ARC00000537}. However, it appears that the result was unlikely to be representative of the end use of Reynobond 55 PE, because, among other things, the sample tested included a backing board of calcium silicate to simulate a 50mm air gap behind the panel.¹⁷⁶¹⁷⁶ {ARC00000537/6}. (Test 5A had also had an air gap of 50mm.)

¶20.15 In subsequent correspondence with CSTB, Mr Wehrle agreed that the panel should be tested with a 20mm air gap between it and the insulation to reflect the way it was used in practice¹⁷⁷¹⁷⁷ {MET00053158/184}. and in his statement he acknowledged that variations in testing methods, such as the use of different insulation products and air gaps, were likely to influence the results.¹⁷⁸¹⁷⁸ Wehrle {MET00053190/19} page 19, paragraph 64. We do not know who was responsible for determining the test method or how the width of the air gap might have affected the outcome. However, when Reynobond 55 PE in rivet form was tested again in later years, it failed to achieve Class B.

CSTB testing of Reynobond 55 PE in cassette form: 2011

¶20.16 In late March 2011 Mr Wehrle decided to find out what was the best classification that Reynobond 55 PE in cassette form could obtain on the basis of Test 5B. On 29 March 2011 he wrote to the CSTB enclosing Test 5B and asked whether the cassette fixing variant could be classified as European Class D.

¶20.17 In May 2011 Mr Wehrle arranged for CSTB to test Reynobond 55 PE in cassette form and classify it under EN13501. On 29 June 2011 CSTB wrote to Arconic and said that, as had happened in 2004, the single burning item test had been stopped early because the heat emitted by the sample had been too great.¹⁷⁹¹⁷⁹ {MET00053158/172}. In response to his request for an F classification, he was told by CSTB that if the product were tested under the ISO 11925 single flame source test¹⁸⁰¹⁸⁰ See Chapter 5. it might be granted an E classification. Arconic duly instructed CSTB to conduct a further test of the product pursuant to ISO 11925 (on a small flat sample of the panel), which, when the specimen performed adequately, necessarily led to a classification of the product as Class E.¹⁸¹¹⁸¹ {ARC00000538}; {ARC00000386}.

¶20.18 Mr Wehrle said that the tests enabled him to understand that cassette-fixed panels were not safer than riveted panels because molten polyethylene from the core could collect in the internal ledges of the cassette and cause a 'flash-over'.¹⁸²¹⁸² Wehrle {MET00053190/18} page 18, paragraph 63. He claimed to have wanted a classification report for the product in cassette form because he had a potentially different understanding of the behaviour of the cassette variant.¹⁸³¹⁸³ Wehrle {MET00053190/18} page 18, paragraph 63. As we have said, however, we think that Mr Wehrle was aware well before 2011 that the product performed worse in cassette form than in riveted form. On any view he certainly knew that by June 2011 at the very latest.

The first BBA review: February – July 2011

¶20.19 The BBA certificate was due to be reviewed at three-yearly intervals and, having been issued in 2008, its first review was therefore due in 2011. The BBA opened a review project in August 2010 under a new contract with Arconic which incorporated the original terms and conditions, including one which required Arconic immediately to notify the BBA of any new or additional information concerning Reynobond or its suitability for use.¹⁸⁴¹⁸⁴ {BBA00008042/3} clause 7(g). It was signed by Mr Schmidt for Arconic in December 2010.¹⁸⁵¹⁸⁵ {BBA00008044/11}. It appears to have been signed by Claude Schmidt.

¶20.20 Between February and June 2011, the BBA asked Arconic for documents relevant to the review and, on 11 April 2011, Arconic provided it with a pack of documents¹⁸⁶¹⁸⁶ {BBA00008044/33}. precisely as requested.¹⁸⁷¹⁸⁷ {BBA00008044/79}. However, none of the documents that Arconic sent concerned the fire performance of Reynobond 55 PE in cassette form and no one at Arconic informed the BBA of the results of Test 5B. That was a striking omission in light of the contract terms, given that only two weeks earlier Mr Wehrle had sent the report of Test 5B to CSTB asking it to classify the product in cassette form as European Class D.

¶20.21 On 24 June 2011, the BBA confirmed that the certificate remained valid until 2014.¹⁸⁸¹⁸⁸ {MET00053158_P16/174}. The report was sent to Mr Wehrle on 1 July 2011,¹⁸⁹¹⁸⁹ {MET00053158_P16/179}. only two days after he had been informed that CSTB's most recent test of Reynobond 55 PE in cassette form had been stopped early and could not produce a classification.¹⁹⁰¹⁹⁰ {MET00053158/172}.

¶20.22 Nonetheless, there is no evidence that Mr Wehrle or anyone else at Arconic took any steps to inform the BBA of those results or gave any thought to whether the reliability and accuracy of the BBA certificate might be affected by the latest information about the fire performance of Reynobond 55 PE in cassette form. Mr Wehrle sought to explain that omission by saying that it simply did not occur to him that it might be necessary to provide the BBA with that recent test data.¹⁹¹¹⁹¹ Wehrle {MET00053190/17} page 17, paragraph59. That explanation is not plausible, however, given Mr Wehrle's keenness to re-test the product in cassette form only a few days earlier in the knowledge that it could achieve at best a European Class E. We also reject his assertion that the product would have achieved Class 0 if it had been tested again under BS 476.¹⁹²¹⁹² Wehrle {MET00053190/17} page 17, paragraph 59. In fact, the standard version of the panel had never been tested under BS 476 Parts 6 and 7 and had never achieved a national Class 0 classification. The chances of a product which could achieve no better than European Class E obtaining a Class 0 classification were slim.¹⁹³¹⁹³ As is explained in Part 2 Chapter 7, the UK Government's RADAR 2 report of 26 May 2000 {CLG00000951}, which was commissioned to find the appropriate transposition point between Euro classes and UK national classes, settled on Class B. Only one product achieved both Class E and Class 0 {CLG00000951/4}. That report was not in the public domain and not available to Claude Wehrle. On any view, the results of CSTB's latest tests were plainly relevant to the BBA's review of the certificate and Mr Wehrle should have provided them to it or alerted others at Arconic of the urgent need to do so. Arconic's failure to provide the BBA with information relating to the recent testing of the product in cassette form was entirely consistent with its deliberate failure to provide the BBA with information about Test 5B when it was preparing the original certificate. It was all part of a strategy on the part of Arconic to conceal from the BBA the fact that Reynobond 55 PE in cassette form had never achieved Class B but had in fact performed significantly worse.

¶20.23 Some months later, in November 2011, Mr Wehrle did tell a Spanish customer that Reynobond 55 PE cassettes were certified as Class E, which the customer described as being "close to spontaneous combustion".¹⁹⁴¹⁹⁴ {MET00053158_P04/54}. If Mr Wehrle was willing to disclose to his Spanish customer the true classification of Reynobond 55 PE in cassette form, the only reason he can have had for concealing that information from the BBA was a desire to withhold it from the UK market, a conclusion that is supported by the events described below.

Concentrating on the UK market: July 2011

¶20.24 On 5 July 2011, a week after the disastrous test by CSTB and five days after receiving confirmation that the BBA certificate remained valid, Mr Wehrle attended a meeting in Freiburg, Germany, with Peter Froehlich of Arconic and Frank Ritter of Alucobond, one of Arconic's competitors.¹⁹⁵¹⁹⁵ {MET00053158_P04/26}. In his report Mr Wehrle set out the results of the tests carried out on Reynobond 55 in riveted and cassette form (which in the case of Reynobond 55 PE in cassette form was Class E) and noted that Class B was the minimum requirement for a facade in Europe. He continued:

"For the moment, even if we know that PE material in cassette has a bad behaviour exposed to fire, we can still work with national regulations who are not as restrictive. … Some countries (Spain …) are already working with EN 13501 standards, and the PE in cassettes is no more usable there."¹⁹⁶¹⁹⁶ {MET00053158_P04/35}.

¶20.25 He also acknowledged that it would be very difficult for Reynobond 55 PE in cassette form to pass a test under BS 8414.¹⁹⁷¹⁹⁷ {MET00053158_P04/36}.

¶20.26 It is plain that by the summer of 2011 Arconic was well aware that Reynobond 55 PE when fabricated in cassette form performed much worse in a fire than in riveted form and was considerably more dangerous. Nonetheless, it was determined to exploit what it saw as weak regulatory regimes in certain countries (including the UK) to sell Reynobond 55 PE in cassette form, including for use on residential buildings.

¶20.27 In May 2012, Claude Wehrle attempted to discuss the European classification of Reynobond 55 PE with Claude Schmidt and Peter Froehlich with a view to deciding what information to give to the market.¹⁹⁸¹⁹⁸ {MET00053157/287}; {MET00053157/290}. We have seen no reference to any such discussion in the documents, although in his statement Mr Wehrle said that it had been agreed that reference to Class B should be removed from marketing material as it no longer accurately reflected the position.¹⁹⁹¹⁹⁹ Wehrle {MET00053190/28} page 28, paragraph 96. He said that the sales team in each jurisdiction was to be told that Reynobond 55 PE in cassette form was Euroclass E, but there is no evidence that anyone in the UK was told that at the time.

¶20.28 As far as we can see, the only step that Arconic took at that time in recognition of the true fire performance of Reynobond 55 PE in cassette form was to remove from its marketing literature the claim that Reynobond 55 PE²⁰⁰²⁰⁰ {ARC00000388/5}. was Class B.²⁰¹²⁰¹ {ARC00000378/3}. We think that that almost imperceptible change was designed to ensure that Arconic no longer made any positive claim for the fire performance of Reynobond 55 PE while not alerting the market to its dangers or to the fact that earlier claims that it was Class B could no longer be relied on.

¶20.29 Arconic claimed that CSTB openly published all test results on its website,²⁰²²⁰² Arconic Module 2 closing submissions {ARC00000770/1} page 1, paragraph 5; Arconic Module 2 closing submissions {ARC00000770/15} page 15, paragraph 65; Arconic Module 2 closing submissions {ARC00000770/24} page 24, paragraph 119. but the evidence of that was limited, particularly about the time of any publication. CSTB did publish summaries of results (but not the test or rating reports which were confidential to the customer) and would always ask the manufacturer in advance of publication.²⁰³²⁰³ Bonhomme {METCSTB00000105/12-13} pages 12-13. The earliest evidence we have been able to find is that on 22 July 2013, Claude Wehrle told a colleague that she could find the technical opinion relating to Reynobond on the CSTB website.²⁰⁴²⁰⁴ {MET00053158_P04/130}. Having regard to that description and the link that Mr Wehrle provided, that appears to have referred to an opinion relevant to the French national classification system. The first date on which a reference to the European classifications for Reynobond PE 55 appeared on the CSTB website was 5 December 2014.²⁰⁵²⁰⁵ {MET00053158_P02/62}. That is consistent with Mr Wehrle's own evidence that, in late December 2014, Arconic did not refuse CSTB's proposal to publish the results.²⁰⁶²⁰⁶ Wehrle {MET00053190/21} page 21, paragraph 70. Accordingly, the earliest that European classifications were likely to have been available on the CSTB's website was late 2014. However, that does not assist Arconic. It is entirely unclear how a UK entity would know that it should consult data published by a French testing house in order to obtain current results of reaction to fire tests. In any event, none of that explains why Arconic did not provide the information to the BBA, why it did not explicitly refer to the classifications in its marketing materials, or why it did not routinely disclose the classifications of Reynobond 55 PE to all UK purchasers.

The Tamweel Tower fire: November 2012

¶20.30 On 18 November 2012, a fire occurred at the Tamweel Tower in Dubai, which was clad in ACM PE rainscreen panels. Senior representatives of Arconic discussed the fire in an exchange of emails on 28 November 2012. An Arconic product had not been used in that case, but Mr Wehrle pointed out that all polyethylene composites react in the same way.²⁰⁷²⁰⁷ {MET00053157/28}. Mr Schmidt accepted that he had seen the emails at the time and that it was known within Arconic that all ACM PE products burned in the same way.²⁰⁸²⁰⁸ Schmidt {Day94/14:13-16}. Nonetheless, as Mr Schmidt accepted, Arconic took no action in response.²⁰⁹²⁰⁹ Schmidt {Day94/15:4-18}.

The Al Hafeet fire and the monitoring of projects

¶20.31 On 22 April 2013, there was a fire at the Al Hafeet tower in Dubai, which was clad with ACM PE rainscreen panels. In response to that fire, on 9 May 2013 Alucobond, one of Arconic's competitors, sought to allay its customers' concerns by confirming that it offered ACM with a fire-resistant core as standard.²¹⁰²¹⁰ {MET00053157/48}. The message was sent by Richard Geater, Alucobond's sales agent for UK and Ireland,²¹¹²¹¹ Richard Geater was employed by 3A Composites GmbH, the sales agent for Alucobond. and quoted a colleague's description of the extreme flammability of ACM panels with unmodified PE cores that were then widely used in Dubai in the interest of saving cost.

¶20.32 The message was sent to Deborah French, Arconic's UK sales representative, a few minutes later. On 10 May 2013 she passed it on to Mr Froehlich, Mr Wehrle, Mr Flacon, and Mr Schmidt and asked for their comments so that she could communicate with their customers.²¹²²¹² {MET00053157/47}. The fact that she chose to include Mr Schmidt, Arconic's president, shows how important she regarded it. Ms French said that she had discussed with Arconic's managers whether it should make the same offer to its customers but was told that it would not do so, at least in part for commercial reasons.²¹³²¹³ French {Day89/49:22}-{Day89/53:12}. We understand that to mean that Arconic would have had to charge more for the fire-resistant panels or would have made a smaller profit if it had sold them at the price it charged for the unmodified version. Mr Schmidt did not accept that version of events²¹⁴²¹⁴ Schmidt {Day94/21:19-22}. but he offered no other explanation and we accept Ms French's account. Arconic did not consider phasing out ACM PE in favour of the fire-resistant version then, or indeed later, because, according to Mr Schmidt, it needed time to think about it.²¹⁵²¹⁵ Schmidt {Day94/21:17-18}. Although Mr Wehrle continued to warn about the risks associated with unmodified ACM panels, Arconic's senior management does not appear ever to have seriously considered withdrawing the product from the UK market at any time before the Grenfell Tower fire.

¶20.33 Indeed, Arconic decided to adopt a different course. A few days later, on 13 May 2013, Ms French wrote to a number of fabricators, including CEP.²¹⁶²¹⁶ {CEP00049717}. In 2014 CEP fabricated and supplied the Reynobond 55 PE cassette panels to Harley for installation on Grenfell Tower. She referred to the recent fire at the Al Hafeet tower and offered them the following assurance:

"Regarding the supply of Reynobond in the UK, as you know we supply both PE and FR core and can control and understand what core is being used in all projects due to the controlled supply route we have. By only supplying Reynobond to a very small group of Approved Fabricators and working very closely with them on all projects we are able to follow what type of project is being designed/developed and then offer the right Reynobond specification including the core. At this stage we will continue to offer both PE & FR core and continue the close working relationship we have with our Approved Fabricators to make sure the right technical support, Reynobond Specification and Materials are being used and installed on Reynobond Projects."

¶20.34 In her witness statement, Deborah French said that the way in which the finished product was assembled and whether it complied with the relevant regulations were matters for the customer or designer.²¹⁷²¹⁷ French {MET00053162/12} page 12, paragraph 44. However, that was clearly not consistent with what she had said in her email of 13 May 2013, as she realistically accepted.²¹⁸²¹⁸ French {Day88/103:3-5}. Although she said that that email did not correctly reflect Arconic's position and had been badly drafted and that we should accept what she had said in her statement,²¹⁹²¹⁹ French {Day88/98:3-12}; {Day88/100:25}-{Day88/101:7}. we think that the truth is more complicated. Ms French's email, which may have reflected advice from Arconic personnel in Merxheim,²²⁰²²⁰ Froehlich {MET00053197/20} page 20, paragraph 57.1.2; French {Day88/88:23-25}; {Day88/90:14-19}. contained clear statements that Arconic knew the details of every project on which its products were used within its controlled supply routes and would ensure that the products supplied for use on those projects were suitable for their intended purpose. It was true that Arconic often did know for which projects its Reynobond products were destined and in many cases what form they would take, since those facts were sometimes recorded on Arconic's customer database (as the sale of Reynobond 55 PE to Harley for use on Grenfell Tower shows).²²¹²²¹ See Part 6, Chapter 55; Record of customer database for Grenfell Tower, indicating the form was to be cassette: {MET00019920/31}. Accordingly, Arconic was often in a good position to judge whether a particular product was suitable for a given project. Certainly, in the case of Grenfell Tower, Arconic did have a certain amount of relevant knowledge. However, it was not true that it intended to advise the fabricator, designer or customer whether the products were safe or suitable for use on any project. On the contrary, it expected the customer and designer to look after themselves. The purpose of Ms French's message was to persuade customers to continue to buy Reynobond PE in the expectation that it would tell them if it was unsuitable for the use to which they intended to put it, although in fact it had no intention of doing so.

Supply by Arconic for use at Grenfell Tower

¶20.35 In Part 6, Chapter 55 we have described how Reynobond 55 PE cassettes came to be supplied for the refurbishment of Grenfell Tower. The story confirms that Arconic often knew for which UK projects its products were bound, knew how they were likely to perform and whether they were safe for the intended use, but did not offer any relevant advice or warning to the designer or customer. So far as concerns Grenfell Tower, Arconic made no attempt to ensure that Harley knew about the performance of the product it supplied for the refurbishment.

¶20.36 Ms French was aware that Grenfell Tower was a high-rise building²²²²²² French {Day88/66:21-23}. because on 24 January 2013, she was sent several drawings relating to the project that showed that it involved a tall building.²²³²²³ {CEP00048962}, attaching {CEP00048966}. Moreover, during the course of the project she arranged for the supply of several samples of Reynobond 55 PE panels to Grenfell Tower, including a mock-up for the RBKC planning department.²²⁴²²⁴ {MET00019919}; {SEA00014518}; {SEA00014456}; {MET00053159/161}. It was also clear to her by August 2014 at the latest that ACM PE panels in cassette form were going to be used.²²⁵²²⁵ French {Day88/66:18-24}; {Day88/184:6-12}. She said that she did not think about the height of any building for which Arconic supplied its products,²²⁶²²⁶ French {Day88/66:18}-{Day88/69:3}. but in the light of her email of 13 May 2013, that seems unlikely. For her, however, it was enough that the product had a BBA certificate.²²⁷²²⁷ French {Day88/68:20}-{Day88/69:3}. Given that, by her own admission, that is all Ms French cared about, it was essential for Arconic, for whom she was selling Reynobond 55 PE, to ensure that the statements in the BBA certificate about fire performance and regulatory compliance were correct. That was all the more so if what she had said in her email of 13 May 2013 was to be taken at face value.

¶20.37 Ms French accepted that in the case of the Grenfell Tower refurbishment she had worked closely with the fabricator and the installer. It was therefore exactly the kind of project that she had described in her email.²²⁸²²⁸ French {Day88/183:11-23}. Despite that she said that she did not have sufficient knowledge to give advice about the risks of using the product in cassette form and did not make any recommendation about it.²²⁹²²⁹ French {Day88/184:13-18}. She did not seek advice about it from anyone else at Arconic because she assumed that others involved in the project would consider them.²³⁰²³⁰ French {Day89/4:2-8}.

¶20.38 Arconic itself must take responsibility for the use of Reynobond 55 PE on Grenfell Tower because it knew that the sale of the product had been obtained on the basis of the BBA certificate which it was well aware gave a misleading impression of the way in which the product in cassette form reacted to fire.

Further testing and classification: 2013 to 2014

¶20.39 By February 2013, Mr Wehrle was aware that there was potential for confusion in the market about the different fire performance classifications of Reynobond PE in cassette and riveted forms, since the former was Class E and the latter Class B.²³¹²³¹ Wehrle {MET00053190/19} page 19, paragraph 65. The FR product was also Class B. He therefore sought to obtain a lower classification for Reynobond 55 PE in riveted form in order to differentiate more clearly between the PE and FR products to assist marketing the FR product. He asked CSTB whether the PE product in both cassette and riveted form could be given a single classification.²³²²³² Wehrle {MET00053190/19} page 19, paragraph 65; {MET00053158/189}. When testing progressed in July 2013, Claude Wehrle told CSTB that Arconic wanted both forms of Reynobond 55 PE to be classed D.²³³²³³ {MET00053158_P02/10}. It is not clear whether he discussed that strategy with anyone in the senior management team at Arconic.

¶20.40 CSTB duly tested Reynobond 55 PE again in July 2013 in both riveted and cassette forms. On that occasion, the riveted specimen achieved only Class C-s2, d0. As on both previous occasions, the test of the product in cassette form had to be stopped early because of what CSTB described as "widespread ignition".²³⁴²³⁴ {MET00053158_P02/38-39}. Again, CSTB advised Arconic that the cassette form could be certified as achieving only Class E. The Class C result for the riveted form was not recorded in a classification report at the time.

¶20.41 Shortly afterwards, Arconic decided that all its Reynobond 55 PE products would be certified as Class E only and that revised classification for both cassette and riveted forms of the product was reflected in a report of CSTB dated 31 January 2014.²³⁵²³⁵ {MET00053158_P04/135}. That was almost exactly the time at which potential contractors were producing their final bids for the refurbishment at Grenfell Tower based on the NBS Specification which included Reynobond 55 PE as a possible rainscreen panel. All forms of Reynobond PE panel were now Class E and accordingly it was unlikely that any external wall on which they were used would comply with functional requirement B4(1) of the Building Regulations.

Claude Wehrle's message of 3 February 2014

¶20.42 On 3 February 2014, all of Arconic's sales staff (including Ms French) were told by Mr Wehrle that all Reynobond 55 PE products had been classified Class E, that the historic Class B classification for the riveted form could no longer be used and that all previous classification reports were immediately cancelled.²³⁶²³⁶ {MET00053158_P04/134}.

¶20.43 Ms French received the email and read it,²³⁷²³⁷ French {Day88/138:2-3}. but did not discuss it with anyone and did not pass the information on to any of her customers.²³⁸²³⁸ French {Day88/140:7-9}. She plainly should have done so, as she herself accepted,²³⁹²³⁹ French {Day88/155:14-16}. and she was unable to explain why she had failed to do so.²⁴⁰²⁴⁰ French {Day88/148:8-15}. She was aware that others in the sales team at Arconic had done so,²⁴¹²⁴¹ {MET00053159/282}. and she understood that the effect of Mr Wehrle's message was that Reynobond 55 PE was not as safe as had previously been thought.²⁴²²⁴² French {Day88/149:5-9}. Ms French was clearly willing and able to consider the fire performance of Arconic's products and to discuss it with her clients or to refer any questions to Mr Wehrle's technical sales support team. Curiously, she accepted that her failure to pass on the information had not been due to an oversight, although she insisted that it had not been deliberate.²⁴³²⁴³ French {Day88/155:18-19}. She said that she had not appreciated its importance because she understood that the relevant classification in the UK was Class 0 and she did not understand the relevance of European classifications to the UK market.²⁴⁴²⁴⁴ French {MET00053162/13} page 13, paragraph 48.2; French {Day88/141:9}-{Day88/142:24}. Her assertion was at odds, however, with the fact that the BBA certificate contained positive statements about the product's European classification which were plainly wrong in light of Mr Wehrle's email. Moreover, even if the details of the different testing regimes had eluded Ms French, she must at least have understood that any information about the fire performance of Reynobond 55 PE was likely to be of interest to customers in the UK and that they ought to be given it.

¶20.44 We think that Deborah French took the view that the change in the European classification of Reynobond PE did not matter as far as the UK market was concerned because Class 0 was all that was required in the UK for regulatory purposes. We also think that the existence of the BBA certificate was sufficient in her eyes and that she did not consider it her responsibility to worry about whether a change in the product's European classification might affect it.

¶20.45 The BBA certificate, however, remained unchanged. Arconic did not tell the BBA about the revised classification or invite it to reconsider the certificate in so far as it related to Reynobond 55 PE. Mr Schmidt sought to explain that as an oversight,²⁴⁵²⁴⁵ Schmidt {Day94/43:11-15}. but we think that unlikely. It must have been clear to some within Arconic, not least Mr Wehrle, that the change in classification was important information that undermined the existing certificate, which, as a result, would be likely to mislead anyone thinking of using Reynobond 55 PE in either form. Given the importance of the BBA certificate to sales in the UK and the close involvement of Mr Wehrle in its original production, we cannot accept that the failure to inform the BBA was the result of a simple oversight.

¶20.46 It remains an open question why Mr Wehrle went to the trouble of telling his sales force in different countries about the change of classification and instructed them not to represent Reynobond PE 55 as being Class B, and yet took no steps to contact the BBA to ask it to qualify or amend the certificate to reflect that fact. Whatever the explanation, however, the fact is that it was risky for Arconic to leave the BBA certificate in circulation in the UK market unamended, since designers would be very likely to continue to rely on it; and Mr Wehrle must have appreciated that risk.

¶20.47 Although she had been in close communication with Harley about the supply of Reynobond panels for the Grenfell Tower project since 2013, Ms French did not tell it about the revised European classification of Reynobond 55 PE or suggest that it might use the fire-resistant version instead. On the contrary, on 23 April 2014, some two months after having received Mr Wehrle's instruction on 3 February 2014, she sent Harley a copy of the BBA certificate without comment.²⁴⁶²⁴⁶ {CEP000000281}. That amounted to a clear invitation to accept everything it said at face value.

¶20.48 Ms French told us that she had not thought that the BBA certificate might be unreliable because she had not made a connection between it and Mr Wehrle's email.²⁴⁷²⁴⁷ French {Day88/162:2-21}. We do not accept that. In our view, she must, or certainly should, have recognised that a significant reduction in the European classification of Reynobond 55 PE might affect the statements made in the BBA certificate. By sending Harley a copy of the BBA certificate without any qualification she led it to think that Reynobond 55 PE was classified as European Class B and might therefore be regarded as achieving national Class 0, when that was not the case.²⁴⁸²⁴⁸ French {Day88/165:4-23}.

¶20.49 Deborah French was clearly at fault in failing to pass on the substance of Mr Wehrle's message to her customers in the UK and in using the BBA certificate without qualification to support the sale of Reynobond 55 PE. On receipt of that message, she should have told potential customers of the change of classification and should have drawn their attention to the availability of the fire-retardant version and the benefits of using it. Claude Wehrle, who was aware how dangerous the product could be in cassette form, was at greater fault in failing to ensure that the BBA was informed of the change in classification, as required by the contract, and in failing to instruct Arconic's sales representatives in the UK not to promote Reynobond 55 PE without drawing that change of classification to the attention of customers.

¶20.50 Arconic had initially been reluctant to recognise that Reynobond 55 PE reacted far worse to fire in cassette form than in riveted form, and was therefore much more dangerous, and had induced the BBA to issue a certificate that masked the difference in performance. As regulatory requirements in European countries became more restrictive, it sought to take advantage of jurisdictions in which regulatory requirements were less stringent to continue selling the product with an unmodified polyethylene core for use in both cassette and riveted form without distinction relying, in the case of the UK, on the support of the BBA certificate. Its commercial strategy strongly suggests that neither the failure of Ms French to pass on important information to her customers nor the failure of Mr Wehrle to give that information to the BBA was the result of an oversight. We are satisfied that they reflected a sustained and deliberate strategy by Arconic to continue selling Reynobond 55 PE in the UK based on a statement about its fire performance that it knew to be false. Deborah French's failure to pass on important information to her customers, wittingly or unwittingly, supported Arconic's strategy.

The Lacrosse Fire: November 2014

¶20.51 In the evening of 24 November 2014 there was a fire at the Lacrosse Building in Melbourne, Australia. A report of the fire circulated among several employees of Arconic, including Alain Flacon, Claude Wehrle and Gwenaelle Derrendinger, between April and June 2015.²⁴⁹²⁴⁹ {MET00053159/404}; {MET00053158_P10/183}. It is clear that Mr Wehrle read the report because, on 18 June 2015, he sent it to CSTB, describing it as "very interesting".²⁵⁰²⁵⁰ {MET00053158_P11/113}.

¶20.52 The report expressed a number of concerns about the use of ACM PE in high-rise buildings. It stated that the fire was directly attributable to the facade of the building and noted the "speed and intensity of the fire spread".²⁵¹²⁵¹ {MET00053158_P11/139}. It also included descriptions of other cladding fires around the world, including those at the Mermoz Tower in France in 2012, the three other fires in the UAE in that year, and the Torch Tower fire in Dubai on 21 February 2015,²⁵²²⁵² {MET00053158_P12/144}. all of which had been attributed to the use of ACM PE in the facade.

¶20.53 The report contained a further clear warning to Arconic of the dangers of using ACM PE in high-rise residential buildings, but although Arconic was aware that ACM with an unmodified polyethylene core was flammable, it took no action in response.²⁵³²⁵³ Schmidt {Day93/33:1-7}; {Day94/14:13-16}; {Day94/15:19}-{Day94/16:1}; {Day94/16:11-19}.

The BBA review: 2013 – 2015

¶20.54 Another review of the BBA certificate began in October 2013. By January 2015 the BBA had made at least twelve attempts to contact Arconic in writing to establish whether it had any new information that might be relevant.²⁵⁴²⁵⁴ {MET00053158_P17/182-186}. On 18 July 2014 Mr Wehrle was asked to provide, amongst other things, written confirmation that there was no new information that would invalidate the certificate.²⁵⁵²⁵⁵ {MET00053158_P17/182}. We deal with that in more detail in Chapter 21.

¶20.55 As we have explained, Arconic was aware in July 2014 that Reynobond 55 PE was not classified as Class B in any form and that the BBA certificate was therefore no longer correct. However, Arconic failed to respond to the BBA's requests for information and allowed the existing certificate to remain in circulation. Although the classification had been extended for five years and Reynobond 55 PE in rivet form had achieved Class B in 2010, Arconic knew by early 2014 that that test could not be relied on to sell Reynobond 55 PE for use in either form, hence Mr Wehrle's message of 3 February 2014. We return to that point in more detail in Chapter 21.

Further testing by CSTB: October to December 2014

¶20.56 Further tests were carried out on Reynobond 55 PE in late 2014. On 4 December 2014 CSTB issued a report classifying the panels in riveted form as Class C²⁵⁶²⁵⁶ {ARC00000397}; {MET00053158_P05/155}. and on the same date, at the request of Arconic, it reissued a report classifying the panels in cassette form as Class E.²⁵⁷²⁵⁷ {MET00053158_P02/119}; {ARC00000395}. However, Arconic did not report those test results to the BBA or ask for its certificate to be revised or withdrawn.

Testing of Reynobond 55 PE with a black core: 2015

¶20.57 In 2015, Arconic changed the colour of the core of Reynobond 55 PE from translucent to black. It then instructed CSTB to undertake a single new test of the product in riveted and cassette forms. The riveted form retained its C classification²⁵⁸²⁵⁸ {ARC00000402}. and the cassette form retained its E classification.²⁵⁹²⁵⁹ {ARC00000405}. The results of those tests were not provided to the BBA.

Arconic's involvement in the supply of Reynobond 55 PE for use at Grenfell Tower: 2015

¶20.58 Earlier in this chapter and elsewhere in the report²⁶⁰²⁶⁰ See Part 6, Chapter 55. we have covered the events by which Deborah French dealt with CEP and Harley in selling Reynobond 55 PE for use on Grenfell Tower. Ms French left Arconic at the end of 2014 and was replaced by Vince Meakins in May 2015.²⁶¹²⁶¹ French {MET00053162/3} page 3, paragraph 12; Meakins {MET00053164/2} page 2, paragraph 5. In the interim, Arconic's sales to the UK were managed by Peter Froehlich assisted by Gwenaelle Derrendinger, an administrative sales assistant.²⁶²²⁶² Froehlich {MET00053197/11} page 11, paragraph 39.

¶20.59 On 3 March 2015, Ms Derrendinger sent a formal quotation to CEP for the supply of Reynobond 55 PE for use on Grenfell Tower.²⁶³²⁶³ {MET00053161/27}. The order was confirmed on 18 March 2015.²⁶⁴²⁶⁴ {ARC00000010}; {MET00053161/28}; {MET00053161/40}; {ARC00000149}. Ms Derrendinger said that, although she could not recall whether she had discussed the type of core with CEP, unless she had been asked to quote for the FR version she would have quoted for the PE version, since it was a UK project.²⁶⁵²⁶⁵ Derrendinger {MET00053191/36} page 36, paragraph 104. See also Derrendinger {MET00053191/35} page 35, paragraph 100, in connection with orders placed by Booth Murie that also resulted in quotations for panels with PE cores. Even as late as March 2015, therefore, Arconic's marketing strategy in the UK was to sell ACM with a PE core unless an FR core was specifically requested.

¶20.60 At that time the panels in cassette form had a European Class E reaction to fire classification. There can be no doubt that Arconic understood that the panels it supplied for use at Grenfell Tower would be used in cassette form: Deborah French was herself aware of that fact and, in any event,²⁶⁶²⁶⁶ French {Day88/173:5-15}; {CEP000000443}. Arconic's client database had included that information from at least April 2014.²⁶⁷²⁶⁷ {MET00053159/86}; {MET00019920/31}. Nobody at Arconic made any attempt to tell CEP, Harley or anyone else working on the Grenfell Tower project that the panels in cassette form were Class E.

¶20.61 Mr Froehlich attended meetings with Harley and CEP in 2014²⁶⁸²⁶⁸ Froehlich {MET00053197/9-14} page 9 and 14, paragraphs 37 and 41. at which the Grenfell Tower project was discussed and was sent several emails relating to the project, including those relating to orders for Reynobond 55 PE that were placed between March and October 2015.²⁶⁹²⁶⁹ {MET00053161/28}. Mr Froehlich acknowledged that he had been aware that the product supplied to Grenfell Tower was Reynobond 55 PE.²⁷⁰²⁷⁰ Froehlich {MET00053197/15} page 15, paragraph 45.

¶20.62 Mr Meakins candidly accepted in his oral evidence that he had been aware for a year before the fire at Grenfell Tower that ACM PE should not have been used on tall buildings because it was dangerous.²⁷¹²⁷¹ Meakins {Day88/130:6-11}. Arconic had been well aware of that from the outset, particularly in relation to panels in cassette form, but failed to disclose that information to anyone working on the refurbishment. Mr Meakins himself did not play a part in the supply of ACM PE panels for the Grenfell Tower refurbishment project.

Arconic's reaction to cladding fires overseas: 2015 and 2016

Fire at the King Fahd Medical Centre in Riyadh: October 2015

¶20.63 On 10 October 2015, a fire occurred at the King Fahd Medical Centre in Riyadh, a high-rise building that was clad in Alucobond FR rainscreen panels. Mr Wehrle was told about the fire on 16 October 2015, when he was sent three photographs showing the damage to the building.²⁷²²⁷² {MET00053158_P10/168}. In response he expressed the view that the panels had performed very well and that the fire would have spread over the entire height of the tower if the panels had had an unmodified polyethylene core.

¶20.64 Mr Wehrle clearly understood that ACM PE panels were likely to contribute to the spread of flame when used on high-rise buildings and were much more dangerous than the fire-resistant version. Indeed, his comment reflects the opinion he had voiced in an internal email on 29 June 2015 that ACM PE was dangerous and that Arconic should move to selling only the FR product as a matter of urgency.²⁷³²⁷³ {MET00053158_P05/14}. That warning appears to have fallen on deaf ears among Arconic's senior management.

The Address, Dubai: 31 December 2015

¶20.65 On New Year's Eve 2015, there was a fire at The Address, a hotel in Dubai. A news article and photograph were passed between Claude Wehrle, Guy Scheidecker and others in early January 2016.²⁷⁴²⁷⁴ {MET00053158_P10/174}.

¶20.66 On 4 January 2016 Robert Campbell, the UK sales representative for Reynolux, wrote to Mr Wehrle and others at Arconic telling them that The Address had been clad in Alucobond ACM PE and suggested that architects were wondering whether ACM PE was safe to use.²⁷⁵²⁷⁵ {MET00053158_P10/172}. Mr Wehrle said that he hoped ACM PE would be removed from the facade cladding market because it was reflecting badly on other ACM products.²⁷⁶²⁷⁶ {MET00053158_P10/172}. Guy Scheidecker, a senior executive, expressed the view that ACM PE was not the only component responsible for such a fire.²⁷⁷²⁷⁷ {MET00053158_P10/176}. That may have been so, but there is no evidence that Arconic then undertook any further investigation of that possibility. Rather, the Arconic management were pleased that on this occasion, a competitor's product was involved, not their own.

Place de Hageneau: January 2016

¶20.67 In January 2016, a fire broke out at the Place de Hageneau in Strasbourg, France. The building that caught fire did not feature any ACM, but another building about 10 metres away was clad in Reynobond 55 PE. On 19 January 2016, Mr Wehrle wrote to Alain Flacon and Lionel Marconnet at Arconic expressing the view that Arconic had been very lucky that the wind had not changed.²⁷⁸²⁷⁸ {MET00053158_P10/179}. He went on to say that it was time for Arconic to stop offering ACM PE for construction purposes. It was aware of the danger and should take steps to withdraw it.

¶20.68 However, yet again Mr Wehrle's warnings went unheeded internally at Arconic, and nothing appears to have been done to withdraw ACM PE from the UK market or to warn customers in the UK of the dangers of which Arconic was aware. Arconic's French sales representatives were instructed to offer only Reynobond 55 FR, regardless of the height of the building, but no similar direction was given in respect of the UK.

¶20.69 There is, therefore, clear evidence that senior managers at Arconic had been aware for many years before the Grenfell Tower fire that the statements the company was making about the fire performance of Reynobond 55 PE were inaccurate and, in particular, that the fire performance of cassette-fixed panels was significantly worse than that of riveted panels. It is equally clear that, despite what they knew, they decided not to ask the BBA to withdraw or amend its certificate and did not withdraw Reynobond 55 PE from the UK market. Arconic thus promoted and sold a product knowing that it presented a significant danger to those who might use any buildings on which it was used.