The section of the inquiry covering products installed on the tower contained many shocking revelations. Now that it is over, Tom Lowe explains who said what

Grenfell Tower wrapped


Grenfell Tower

Disclosures from current and former staff at materials giants Celotex, Kingspan and Arconic, as well as certification body the British Board of Agrément (BBA) and testing house BRE, have raised alarming questions about weaknesses in the testing and certification of construction products.

Module two of the second phase of the Grenfell inquiry has heard how some products which have been installed on buildings across the UK for well over a decade have been marketed as safe based on dubious and often deliberately misleading test evidence. 

The second phase of the inquiry’s second module wrapped up at the end of last month. This module, which focused on the production, testing and sale of the materials that were installed on Grenfell Tower prior to the June 2017 fire which claimed 72 lives, has had a profound impact on the wider construction industry.

Some of the headlines the inquiry has generated since this module began last November have become common knowledge among construction professionals. Others, which have surfaced after the initial shock of Celotex and Kingspan’s disclosures last year, may be less well known. To put the whole module into context, here is a summary of the evidence.

Celotex rebranded combustible insulation product as safe for use on buildings above 18m by rigging fire test

The inquiry heard in November how, in 2014, Celotex had changed the name of one of its insulation products which had failed the fire test needed for use above 18m and rigged a new test to ensure that it passed. It did this by adding fire-resisting magnesium oxide boards to the test rig and reducing the size of ventilation gaps between panels to restrict the spread of flames. 

Former Celotex product manager Jonathan Roper, who was given the job of developing the “new” product, was 23 at the time, had no technical experience, received no training in building regulations, and was working in his first job since finishing university. He admitted that he had realised at the time that rigging the test was a “fraud on the market” and said the actions of Celotex had been “completely unethical”.

The hearings were told how the decision to rig the test was made by senior managers to drive up the firm’s share price to achieve a lucrative sale of the business. The firm had also been set a target to increase profit by 15% after it was purchased by French multinational Saint Gobain in 2012.

Following the test, Roper gave a presentation on how to market the insulation to a board meeting in which he said that “nobody understands the test requirements” and that building control had “hugely differing levels of understanding” on fire safety requirements.

He then told the inquiry how he had deliberately not corrected a mistake by a building control officer from the Local Authority Building Control (LABC) who had confused the Class 0 fire rating with the concept of “limited combustibility”. Roper then emailed some suggested wording for a certificate for the insulation to the LABC, which appeared to have been copied and pasted directly into the final certificate complete with a typo from Roper’s email.

Kingspan marketed insulation product as safe for use above 18m for nearly 15 years based on test report of different product

Next to give evidence was Kingspan, which admitted prior to the hearings in October last year in a letter to BRE that it had been marketing its K15 insulation using a 2005 test report of an older, chemically different version of the product.

The hearings were told that this version had been changed in 2006 to a newer version which was “quicker and more cost-effective to produce”.

Despite the firm claiming in its letter that it was “now of the view” that there were differences between the product that was tested and what had been sold on the market, former technical project manager Ivor Meredith told the inquiry that the differences had been “common knowledge” at the firm since the product was altered.

After a 2007 test of the new version had turned the rig into a “raging inferno” within 17 minutes which had to be extinguished because of fears that it would set fire to the laboratory, Meredith said he had sent a “very animated” report to his superiors. Despite the test’s dramatic failure, he said that he could not recall “any real sort of shock”.

The inquiry then heard how technical manager Philip Heath had said facade engineering consultant Wintech could “go f’#ck themselves, and if they are not careful we’ll sue the a’#se of [sic] them” after the firm had raised concerns about the product. 

The hearings were also told that Kingspan staff had reassured worried contractors by sending them test results of a trial product containing different chemicals to what was being sold on the market. 

And the inquiry was shown texts between technical staff discussing the testing of K15 in which one stated that the product’s marketing literature was “all lies”, adding: “Alls [sic] we do is lie in here.”

Arconic knew ACM cladding panels would burn 10 years before the Fire but did not withdraw them due to “cost implications”

Hearings resumed in February after a gap of nearly two months because of a positive covid-19 test on the inquiry team and the imposition of a third national lockdown in January.

The focus turned to Arconic, which had supplied the Grenfell Tower refurbishment project team with 3,000m² of the ACM cladding panels which the first phase of the inquiry had found to be the “primary cause” of the rapid spread of flames up the side of the building.

Despite being warned as early as 2007 that 5,000m² of ACM would have the same “fuel power” as a 19,000 litre truck of oil in a fire, the firm continued selling the panels until 2017. It only halted supply on 26 June of that year, 12 days after the Grenfell Tower fire.

Five days before the panels were withdrawn, the firm’s former sales manager Deborah French admitted in a secretly recorded phone conversation that Arconic had continued selling them because of the “cost implications” of removing them from the market.

The firm had been selling two versions of the product, one designed to be folded into cassettes and another designed to be installed with rivets. While the riveted form had achieved a Euroclass B rating, allowing it to be used above 18m, the cassette form – the type used on Grenfell Tower – had failed the test “disastrously”.

But Arconic only supplied the test results of the riveted form to the BBA, which then issued a certificate giving both types a Euroclass B rating. 

The firm’s president denied that this had been “deliberate concealment”, insisting that the BBA “could have found out” about the test on the cassettes in an audit.


  • 29 June 2017: Prime minister Theresa May announces inquiry
  • 14 September 2017: Hearing opens under chairmanship of retired judge Sir Martin Moore-Bick
  • June-December 2018: Hearings for phase 1 which focuses on the night of the fire
  • 30 October 2019: Findings of phase 1 published
  • January 2020 - March 2021: Hearings begin for phase 2 covering events leading up to the fire. Module 2 of phase 2, covering cladding products, runs from November 2020 to March 2021 
  • April 2021 – February 2022: Five more modules in phase 2 covering aspects such as fire safety measures, further expert evidence and the aftermath of the fire 
  • February 2022 (approx): Inquiry due to complete with findings to follow

BRE was warned about Kingspan’s misuse of test reports four years before the fire

Next in the spotlight was former government-run testing house, BRE, which was privatised in 1997.

The organisation’s current director of fire testing and certification Stephen Howard was asked why he had failed to take any action after being warned in 2013 about Kingspan’s misuse of a 2005 fire test to sell its K15 insulation for high rise buildings.

Asked by then Celotex product manager Jonathan Roper how Kingspan had been able to market K15 for a wide variety of applications not covered by its test report, Howard replied that the onus was on the building owner and building control to ensure the system was compliant, signing off by telling Roper: “I am not sure that I have much more to add.” 

BRE former burn hall manager Philip Clark also claimed that manufacturers of combustible materials could “sneak” extra components onto test rigs without inspectors knowing.

Clark, who told the inquiry that “the reliance was very much on the honesty of the client” and said there was a “large element of trust in everything we do”, denied being aware that Celotex had rigged its 2014 fire test, blaming the oversight on the burn hall being a “busy place”.

And former project manager Tony Baker admitted to a “very basic error” by not noticing the magnesium boards which Celotex staff fixed to the rig to reduce the spread of flames.

The BBA based certificate for ACM panels installed on tower on information on Arconic’s website

Leading UK certifier the BBA used incomplete test data and information on Arconic’s website to renew a certificate clearing the firm’s highly combustible cladding panels for use above 18m.

Former BBA project manager Valentina Amoroso admitted making the decision after Arconic had avoided sending her the required test data for 15 months, emailing the firm that she would go ahead with the data that was available because of “severe delays in closing this review job”.

The body’s former deputy chief executive also told the inquiry that it may have based the fire rating for Kingspan’s K15 insulation on tests of different products. Brian Moore said that it was “more likely than not” that the classification had been “extrapolated” from test reports on other phenolic foam products dating as far back as 1991. He added that despite the practice being “not unusual” at the BBA, he couldn’t be certain whether it is permitted under UK fire regulations.

Shake-up of testing practices

In February, the BBA unveiled proposals for a major shake-up of its testing practices including unannounced, random testing of materials already on the market, observation of tests filed for assessment, and early involvement with manufacturers during a product’s development process. It said the proposals intended to address the “specific and unique challenges” of buildings above 18m.