Part 7: Replacement of the gas riser

Chapter 68: Replacement of the gas riser

Introduction

¶68.1 In his Phase 1 report the Chairman found that the supply of gas to the tower did not play a significant part in the outbreak or development of the fire on 14 June 2017 and that the employees of Cadent Gas Limited who attended to shut off the gas supply made appropriate decisions and executed their task admirably.¹¹ Phase 1 Report Volume IV, paragraph 31.14. In this phase of the Inquiry we have investigated the replacement of one of the gas risers between 2016 and 2017. The works had not been finished at the time the fire occurred and we therefore thought it appropriate to consider whether they might have contributed to the spread of smoke or fire within the tower.

The supply of gas to the tower

¶68.2 On completion of the refurbishment work to the tower in 2016 gas was supplied to residents by two pipelines which entered the basement below ground level on the east side: a ten-inch steel pipe that fed the boilers supplying hot water to residents²² Hancox, Gas Engineering Report {RHX00000012/15} paragraphs 30-31 and figure 6. and a four‑inch steel pipe that supplied gas to the flats.³³ Hancox, Gas Engineering Report {RHX00000012/13-14} paragraph 29 and figure 4.

¶68.3 The pipe providing gas to the flats entered the basement at a high level. Four pipes, known as "risers", ran vertically up into the tower; two of them were then split into two, making six risers in total.⁴⁴ Hancox, Gas Engineering Report {RHX00000012/23} paragraph 44. These passed through the fabric of the building and directly into a cupboard in each flat. One of the risers supplied gas to all the "Flat 2s".

¶68.4 The gas supply had been installed at the time the tower was built and by the time of the fire the pipework was almost 50 years old. It did not comply in a number of respects with current regulations.

Supply to the tower following works commencing October 2016

¶68.5 On 1 October 2016, corrosion of the riser serving the "Flat 2s" made it necessary to cut off the supply of gas to all the "Flat 2s".

Cadent

¶68.6 Cadent Gas Limited ("Cadent") is a gas transporter which owns and operates the pipes and apparatus that transport gas. Cadent does not own the gas itself but is paid by suppliers to deliver it through a network of pipes in a particular area. In the case of Grenfell Tower the area was that served by Cadent's North London gas distribution network.⁵⁵ Mason {CAD00003005/2} page 2, paragraphs 7-8. Cadent was responsible for the safety of the service pipework up to the point of the emergency control valve located at each customer meter; the installation pipework beyond the meter was the responsibility of RBKC as the owner of the building.⁶⁶ Harrison {Day159/16:4-14}.

¶68.7 The Health and Safety Executive monitors and oversees the health and safety of onshore gas distribution and supply⁷⁷ Harrison {Day159/17:4-8}; https://www.hse.gov.uk/regulating-major-hazards/energy-division.htm. and Ofgem determines the funding and ensures the competitiveness of gas pricing.⁸⁸ Harrison {Day159/18:7-16}. Since the 1990s, in common with all gas transporters, Cadent is under certain statutory obligations, including obligations to: a. Connect and maintain a connection for customers within 23 metres of a main (measured horizontally),⁹⁹ Gas Act 1986 {INQ00014770/52-53} section 10; Mason {CAD00000004/2} page 2, paragraph 5. unless such a connection endangers the public;¹⁰¹⁰ Gas Act 1986 {INQ00014770/53} section 10(9)(b); Harrison {Day159/19:16}-{Day159/20:8}. and

¶b. Maintain service pipelines and apparatus, such as shut-off valves, in efficient working order and in good repair.¹¹¹¹ Pipelines Safety Regulations 1996 {INQ00014790/7} regulation 13; Pipelines Safety Regulations 1996 {INQ00014790/3} regulation 3(1)-(2).

Inspection on 30 September 2016

¶68.8 Part of Cadent's strategy to maintain service pipework for buildings in multiple occupation was to conduct a rolling programme of surveys.¹²¹² Harrison {Day159/21:12-21}. On 30 September 2016 Cadent carried out a survey on Grenfell Tower¹³¹³ {CAD00000031}. and identified three significant problems: a. There was a gas leak between Flat 22 and the flat above, Flat 32, on floors 5 and

¶6 respectively;¹⁴¹⁴ {CAD00000031} sheet 3: Riser Information, row 12, column AE; {CAD00000031} sheet 4: Lateral Information, row 9, column U; Hancox, Gas Engineering Report {RHX00000012/135} paragraph 310. b. The pipeline isolation valves on the east side of the tower could not be found;¹⁵¹⁵ {CAD00000031} sheet 2: Supply Pipe Information, rows 11-12, column L; Hancox, Gas Engineering Report, {RHX00000012/68} paragraph 158. and c. There was severe corrosion of some laterals.¹⁶¹⁶ {CAD00000031} sheet 4: Lateral Information, rows 7-12, column O; Hancox, Gas Engineering Report {RHX00000012/114-115} paragraph 248 and figures 65-66.

¶68.9 In response to the discovery of the leak Cadent immediately cut and capped the riser which supplied "Flat 2s". That was completed in the early hours of 1 October 2016.¹⁷¹⁷ {CAD00000031} sheet 4: Lateral Information, row 9, column W; Hancox, Gas Engineering Report {RHX00000012/23- 24} paragraph 44 and figure 17.

¶68.10 Cadent used this and other information from the survey to calculate a building priority score¹⁸¹⁸ Harrison {CAD00002985/8} page 8, paragraphs 30-32. which indicated how often a building should be surveyed. The score for Grenfell was such that Grenfell Tower was placed on the list for discussion at the regular Cadent hazard and operability meetings, known as "HAZOP" meetings. The inability to find the missing pipeline isolation valves and the severe corrosion was noted on a schedule presented at the HAZOP meetings of 5 December 2016 and 23 February 2017.¹⁹¹⁹ {CAD00002984}.

Lost pipeline isolation valves

¶68.11 Pipeline isolation valves, by which the supply of gas to a building can be shut off quickly, are the main emergency control mechanism for buildings in multiple occupation.²⁰²⁰ Harrison {Day159/40:11-16}. They are mandatory on service pipelines to such buildings²¹²¹ {RHX00000005/64} paragraph 7.3.1; Pipelines Safety Regulations 1996 {INQ00014790/7} regulation 13. and are located outside the premises so that access to them can be obtained easily through valve chambers with marked covers visible at pavement level. For obvious safety reasons, as part of making appropriate arrangements for responding to incidents and emergencies, gas transporters have a duty to maintain access to pipeline isolation valves.²²²² Pipelines Safety Regulations 1996 {INQ00014790/7} regulation 12; see also the duty to maintain the pipeline (including all its related apparatus) in an efficient state, efficient working order and in good repair: Pipelines Safety Regulations 1996 {INQ00014790/7} regulation 13. Notwithstanding that, we were told that it was well known in the gas industry that pipeline isolation valves are often lost, for example, as a result of being covered over by landscaping or road works.²³²³ Hancox, Gas Engineering Report {RHX00000012/220} paragraph 470; Harrison {CAD00002985/22} page 22, paragraph 82; Harrison {Day159/108:18}-{Day159/109:1}. The survey of Grenfell Tower carried out by Cadent in 2008 showed that "fire valves" (another term for pipeline isolation valves) had been installed on the pipes supplying gas to the tower.²⁴²⁴ {CAD00002989/5}. The fact that Cadent could not find them in 2016 was almost certainly attributable to the landscaping works completed as part of the refurbishment.²⁵²⁵ Hancox {Day161/41:10-20}.

¶68.12 Having discovered in October 2016 that the pipeline isolation valves could not be found, Cadent should have taken steps to reinstate them immediately but it appears that the appropriate team in Cadent was not notified about the need for that to be done. Mr Harrison, one of Cadent's network directors, accepted that that had been a failure on its part which should not have occurred.²⁶²⁶ Harrison {Day159/127:19-20}.

¶68.13 The failure to reinstate the pipeline isolation valves did not affect the course of events surrounding the fire because burning debris falling on the east side of the tower would have prevented anyone from obtaining access to them.²⁷²⁷ Phase 1 Report Volume IV, paragraph 31.4. However, on another occasion access to them might be of critical importance and we therefore think that all gas transporters should have a legal duty to inspect these emergency valves at intervals to ensure that they are accessible and to reinstate them if they are not.

Corrosion survey

¶68.14 Although steel pipes are very strong and have a long service life under normal conditions, they are susceptible to corrosion, particularly if exposed to damp. Because corrosion causes the steel to deteriorate and lose its integrity, keeping it within acceptable levels is an important part of maintaining the safety of pipelines.

¶68.15 Cadent commissioned a corrosion survey for Grenfell Tower on 5 December 2016.²⁸²⁸ {CAD00002984/1}. Mr Harrison explained that the survey was not completed because of problems over obtaining access to the tower. A Cadent engineer attended on two occasions but could not get into the relevant parts of the building.²⁹²⁹ Harrison {CAD00002985/10-11} pages 10-11, paragraphs 37-39. Although a decision was made in February 2017 to try again, the full corrosion survey had not been completed by the time of the fire, some eight months after the corrosion had been discovered.³⁰³⁰ Harrison {CAD00002985/11} page 11, paragraph 41; Harrison {Day159/137:22}-{Day159/138:7}. As Mr Harrison accepted, that was unacceptable.³¹³¹ Harrison {Day159/137:22}-{Day159/138:7}; {Day159/138:22}-{Day159/139:5}; {Day159/140:8-14}; {CAD00002987/14} paragraph 5.6.3; {CAD00003001/13-15} paragraphs 6 and 9.

¶68.16 After the fire it was found that a riser in Flat 115 had ruptured. It is possible that that was a result of a failure to keep corrosion levels within acceptable limits and occurred during the fire or later as the building continued to deform from its effects,³²³² Harrison {CAD00003043/13-14} pages 13-14, paragraph 43; Cadent Module 3 submissions {CAD00003047/23} paragraphs 64-65. but we do not think it is necessary for us to reach a decision on that question and in any event do not have sufficient evidence to do so with any confidence. All we can say is that several factors may have contributed to the rupture of that particular pipe, one of which may have been corrosion that weakened it at one of its threaded joints.³³³³ Hancox, Gas Engineering Report {RHX00000012/110-111} figures 60-61 and paragraph 242.

Inaccurate maps

¶68.17 By June 2017 Cadent's map of the gas pipes running into the tower should have shown all three systems: the pipe supplying the boilers, the original pipe supplying gas to the flats and the new pipe supplying gas to the "Flat 2s". However, the maps available to the Cadent engineers who attended the fire only showed the pipe supplying the boiler.³⁴³⁴ {CAD00002174}; {CAD00002241}. Cadent should have known that its maps were inaccurate because those available when it carried out the survey in September 2016 also showed only the pipe supplying the boilers.³⁵³⁵ {CAD00000031} Supporting Photos sheet: row 3, column B.

¶68.18 Cadent maintained that its failure to correct the maps did not, in fact, have any material effect on the ability of its engineers to identify and isolate the gas supply pipes on the night of the fire, since they were able to obtain the correct information from other sources.³⁶³⁶ Cadent Module 3 submissions {CAD00003047/18-19} paragraphs 50-52; Harrison {MET00018804/8} page 8, paragraph 31. That is true, but it was attributable more to luck than good organisation that the engineers concerned happened to know the true layout of the pipework and had accurate information readily available to them³⁷³⁷ Cadent employee Jason Allday and surveyor Simon Boygle. as well as the time and the means to provide it to those who needed it.³⁸³⁸ {MET00018309/2-4}; Phase 1 Report Volume IV, paragraph 31.3.

¶68.19 It is important that all gas transporters have accurate maps of the pipelines serving buildings in multiple occupation.³⁹³⁹ Hancox {Day161/58:22}-{Day161/59:3}. Although Cadent has a process for revising its maps, it was not implemented in good time in the case of Grenfell Tower.⁴⁰⁴⁰ {CAD00002957}; Hancox, Gas Engineering Report {RHX00000012/74} paragraph 174 and figure 41; Harrison {Day159/145:5-14}. Accurate maps are essential in order that operatives can locate the pipeline isolation valves and may be of critical importance in emergencies. There appears to be no good reason why the map was not revised after the 2016 survey, or, failing that, after the installation and commissioning of the new pipe in early 2017.

Replacement gas supply: 2016-2017

¶68.20 UK designers of gas supply systems pay particular attention to the standards published by the Institution of Gas Engineers and Managers (IGEM) when planning and executing work of that kind.⁴¹⁴¹ Hancox {Day161/8:23}-{Day161/9:1}; Harrison {Day159/33:8-23}; Dolan {Day160/17:20}-{Day160/18:11}. Although gas engineers should be aware of the Building Regulations,⁴²⁴² Dolan {Day160/17:2-18}. in 2016 there was a strong emphasis and reliance on IGEM publications and there existed within the gas industry a general understanding that if an engineer complied with that guidance, all relevant regulatory standards, including the Building Regulations, would be met.⁴³⁴³ Hancox {Day161/20:24}-{Day161/21:16}. The IGEM publication applicable to the work to be carried out at Grenfell Tower in 2016 was IGEM/G/5, second edition.⁴⁴⁴⁴ {RHX00000005}.

¶68.21 Having cut and capped the riser serving the "Flat 2s" on 1 October 2016, Cadent had to consider whether and, if so, how, to reinstate the supply of gas to them.

¶68.22 For that purpose it engaged tRIIO, a consortium formed by Morrison Utility Services Limited and Skanska Construction UK Limited, to design and install replacement gas pipes and connections.⁴⁵⁴⁵ Dolan {MET00012711/2} page 2, paragraphs 7-10. tRIIO entered into subcontracts with other organisations and individuals for some of the work.⁴⁶⁴⁶ Dolan {MET00012711/4-7} pages 4-7, paragraphs 23-28. There was plainly some urgency and it is apparent that Cadent sought to reinstate customers' gas supplies as quickly as possible.⁴⁷⁴⁷ Dolan {Day160/29:21}-{Day160/30:2}.

¶68.23 Two options were available to it: it could either find a way to install a new riser to reconnect the "Flat 2s", or it could pay customers to replace their gas appliances with electric appliances. On 17 November 2016, Cadent rejected the latter option because only part of the building was affected.⁴⁸⁴⁸ {CAD00000059}; Mason {CAD00003005/8} page 8, paragraph 23. It therefore decided to restore the supply of gas, provided it could do so safely.⁴⁹⁴⁹ Gas Act 1986 {INQ00014770/53} section 10(9).

The first design

¶68.24 IGEM/G/5 provides a hierarchy of options for siting gas pipes. It states that the best option is an external network of pipes so as to give them direct ventilation to natural air.⁵⁰⁵⁰ {RHX00000005/29-31} paragraph 4.2.1.4. In the case of Grenfell Tower however, the TMO and tRIIO agreed that installing pipes on the outside of the tower was not practicable because they would pierce the newly installed rainscreen cladding, would compromise the integrity of the facade and potentially invalidate any warranty for the cladding system.⁵¹⁵¹ Dolan {MET00012711/8-9} pages 8-9, paragraph 34; Dolan {Day160/53:4-10}. In our view, that was a reasonable position for the TMO to take.

¶68.25 tRIIO commissioned Simon Boygle, a surveyor trading under the name of London Ops Gas, to find an alternative route for a replacement riser. He and tRIIO considered several options, including running the new riser in the same place as the original riser, that is, vertically through the floors and ceilings of the flats and directly into the kitchens. That idea was quickly discounted as impracticable, however, because of the difficulties involved in obtaining access to the pipes embedded in the fabric of the building.⁵²⁵² Dolan {MET00012711/8} page 8, paragraph 31; Dolan {Day160/62:11-23}; Hancox {Day161/112:7}-{Day161/113:7}. They also considered running the new pipe through the existing utilities shaft,⁵³⁵³ Dolan {MET00012711/9} page 9, paragraph 35; Mason {CAD00003005/11} page 11, paragraph 39; Dolan {Day160/64:20}-{Day160/65:12}. but that option was rejected, principally because the ventilation was considered to be inadequate.⁵⁴⁵⁴ Hancox {Day161/113:9-25}.

¶68.26 A key consideration was that any gas pipework must be adequately ventilated so that if a leak occurs it does not cause the atmosphere to become unsafe, with the consequent risk of an explosion. While external ventilation directly to fresh air is preferred,⁵⁵⁵⁵ {RHX00000005/52} paragraph 6.4.3.1. IGEM/G/5 allowed for any steel pipeline that had welded or screwed joints to be ventilated indirectly to outside air through an area that was "normally-occupied", meaning occupied as an individual dwelling or a common corridor or lobby.⁵⁶⁵⁶ {RHX00000005/52} paragraph 6.4.3.5. However, IGEM/G/5 stated that mechanical ventilation should not be used to achieve the required ventilation level,⁵⁷⁵⁷ {RHX00000005/52} paragraph 6.4.3.2; {CAD00000054/2-3}; Dolan Day160/66:21}-{Day160/67:8}. which in the case of Grenfell Tower prevented siting gas pipes in the lobbies, since they relied on mechanical ventilation. IGEM/G/5 also required ventilation to be provided at the top and bottom of every fire compartment,⁵⁸⁵⁸ {RHX00000005/54-55} figures 18-19. consistently with the guidance in Approved Document B relating to ventilation in protected shafts.⁵⁹⁵⁹ {CLG00000224/81} paragraph 8.41. Approved Document B also provided further guidance on the ventilation of shafts, directing the reader to BS 8313:1997 on the size of ventilation openings.⁶⁰⁶⁰ {CLG00000224/81} paragraph 8.41.

¶68.27 It is clear that there were two requirements that had to be taken into account: ventilation for gas safety and compartmentation for fire safety. Giving appropriate recognition to each of them is a challenge for anyone seeking to supply gas to any complex building in multiple occupation.

¶68.28 On 12 October 2016 Simon Boygle proposed that a new riser should be installed in the stairwell of Grenfell Tower.⁶¹⁶¹ {CAD00000054}; updated on 1 November 2016 {CAD00000038}. That was the only means by which to satisfy the twin requirements of ventilation and compartmentation.⁶²⁶² Dolan {Day160/55:24}-{Day160/56:14}. The new riser would travel from the basement, through the lower floors, through the stairs, and then by means of laterals restore the supply of gas to the "Flat 2s". It was envisaged that at a later date the pipework would be extended to serve all the flats in the tower and the original pipework would be decommissioned.

Firefighting stairs

¶68.29 What nobody involved in the project understood at the time was that although both IGEM/G/5 and Approved Document B allowed gas risers to be placed in protected stairs,⁶³⁶³ {RHX00000005/52} paragraph 6.4.3.3; {CLG00000224/32} paragraph 2.42. if the stairs were also firefighting stairs, additional protections were required.⁶⁴⁶⁴ {CLG00000224/32} paragraph 2.35. Diagram 52 in Approved Document B advised that firefighting stairs should be constructed in accordance with BS 5588-5:2004,⁶⁵⁶⁵ {CLG00000224/116} diagram 52. which in turn advised that such a shaft should contain only services associated with fire-fighting. In other words, the guidance in Approved Document B and the British Standard to which it referred was that gas supply equipment should not pass through the same compartment as services associated with firefighting.⁶⁶⁶⁶ Fire Precautions in the design, construction and use of buildings – Part 5: Access and facilities for firefighting {BSI00000087/23} paragraph 7.1.4. While BS 5588-5 was superseded by BS9999:2008 (and ADB therefore referred to an out of date standard), identical provisions appeared in BS9999:2008 at paragraph 27.2.7.2, to the effect that only services associated with firefighting should pass through the firefighting shaft {BSI00000064/123}.Show full note Although it is arguable that appropriate standards of safety could have been achieved by separating the riser from the firefighting stairs with fire-resisting construction, as was explicitly permitted for gas pipework in protected stairs,⁶⁷⁶⁷ {CLG00000224/81} Note to paragraph 8.40. Neither BS 5580-5:2004 nor BS9999:2008 make any such provision. the position is not clear in the guidance document or standards in relation specifically to firefighting stairs. We note that the first design did not have any boxing protecting the riser in the stairs.

¶68.30 That special consideration had to be given to firefighting stairs does not appear to have been widely understood in the gas industry before the Grenfell Tower fire⁶⁸⁶⁸ Dolan {Day160/91:19}-{1Day160/92:14}; Hancox {Day161/117:7-18}. and there is no evidence that any other professional expressly considered the point.⁶⁹⁶⁹ {TMO10015305}; {TMO10016548}; {CST00001240}.

Ventilation and compartmentation challenges

¶68.31 tRIIO's first design did not provide much detail of how the ventilation of the pipework would be achieved.⁷⁰⁷⁰ {TRI000000263}. The riser was to be constructed of steel with welded joints, so indirect ventilation was available as a design choice pursuant to the guidance in IGEM/ G/5.⁷¹⁷¹ {RHX00000005/52} paragraph 6.4.3.5. On that basis the design included the following features: a. Ventilation for the basement was to be achieved by way of pre-existing ducts that

vented to the outside air and were suitable for the gas installations that were already in the basement.⁷²⁷² Dolan {Day160/101:6-10}; {Day160/104:4-9}.

¶b. Between the basement and level 2, the riser passed through the south-east utilities shaft and into a storeroom adjacent to the community meeting room. The utilities shaft, which rose more than one storey, had a louvered door at low level which would have provided some ventilation⁷³⁷³ Hancox, Gas Engineering Report {RHX00000012/30} figure 22; Thompson {TRI00002923/2} page 2, paragraph 8; Images of louvred doors {TRI00002921}; {TRI00002922}; Dolan {Day160/105:3-15}. but there was no obvious ventilation at the top of the compartment.⁷⁴⁷⁴ Hancox, Gas Engineering Second Supplementary Report {RHX00000027/21} paragraph 44. It was not clear what ventilation strategy applied to the storeroom.⁷⁵⁷⁵ Hancox, Gas Engineering Second Supplementary Report {RHX00000027/22} paragraph 50; Dolan {Day160/105:15- 23}; Hancox {Day161/135:5-18}. c. Through the stairs, the riser was ventilated across the whole shaft.

¶d. At each floor, the lateral pipes passed through the compartment walls between the stairwell and the communal lobbies through oversized holes. On the lobby side, the lateral pipework was to be boxed in with fire-rated materials.⁷⁶⁷⁶ {TRI000000392}. The boxing in was necessary because the lobbies had mechanical ventilation, so the gas pipework could not ventilate directly into that space. The holes in the wall between the stairs and the lobby were left open so that any gas leaking from the pipework in the lobby would dissipate into the stairs, but the holes in the walls between the lobbies and the flats were to be fully fire‑stopped. e. At the top of the stairwell the riser vented directly to the air through the rooftop vents.

¶68.32 No analysis was made of the ventilation available for each space through which the gas pipes were to pass.⁷⁷⁷⁷ Dolan {Day160/102:3}-{Day160/103:9}; {Day160/103:16-22}; {Day160/148:2-14}. In our view, such a complicated design required detailed consideration of the ventilation arrangements in order to demonstrate that the pipework would be ventilated adequately and safely.⁷⁸⁷⁸ Hancox, Gas Engineering Report {RHX00000012/191} paragraph 385; Hancox, Gas Engineering Report {RHX00000012/195} paragraph 394; Hancox {Day161/139:15-17}. tRIIO did not carry out a detailed analysis of that kind.

¶68.33 A detailed risk assessment should also have been carried out.⁷⁹⁷⁹ Hancox {Day161/142:8-11}. According to IGEM/G/5, designers of gas pipework in buildings in multiple occupation should carry out a risk assessment to identify and mitigate hazards and maintain a written record of the results.⁸⁰⁸⁰ {RHX00000005/28} paragraph 4.2.1.1; {RHX00000005/94-102} Appendix 3. In late 2016 tRIIO's procedure was to record risks in its CDM Design Risk Register, which took the form of a spreadsheet that generated an initial risk score. If that score was 18 or more, tRIIO would carry out a fuller assessment by responding to a list of more probing questions.⁸¹⁸¹ {TRI000000369}. tRIIO completed the CDM Design Risk Register in relation to Grenfell Tower on 15 November 2016. It generated a score of 17.⁸²⁸² {TRI000000369}. Although the designers could have manually over-ridden the process and carried out a fuller assessment of risk, they did not do so.⁸³⁸³ Dolan {Day160/85:14-16}.

¶68.34 We do not think that tRIIO's risk assessment process was adequate for buildings in multiple occupation.⁸⁴⁸⁴ Hancox {Day161/150:10}-{Day161/151:9}; {Day161/172:3-16}. We note that tRIIO's own CDM Design Risk Register shows that any building of that kind would be "high risk" in and of itself⁸⁵⁸⁵ {TRI000000369} at "MOBS" where a red "high-risk" categorisation is shown in the final column. and accordingly, the process should have required that it be the subject of a fuller assessment. That was particularly so in the case of an old building such as Grenfell Tower in circumstances where the proposed design was relatively complicated. tRIIO should therefore have overridden the scoring system and considered the risks posed by the installation in more detail. Mr Dolan, tRIIO's director of operations, accepted that it had been at fault in failing to do so.⁸⁶⁸⁶ Dolan {Day160/89:11-23}.

¶68.35 Cadent's contract with tRIIO did not require it to check design decisions and it did not expect to do so. However, once work began, Cadent was under a statutory duty to notify the Health and Safety Executive (HSE) using the form "F10".⁸⁷⁸⁷ The Construction (Design and Management) Regulations 2015 {INQ00011316/7} regulation 6; The Construction (Design and Management) Regulations 2015 {INQ00011316/29} regulation 6, schedule 1. tRIIO had produced an F10 covering all the work it intended to carry out during the period 2016-2017,⁸⁸⁸⁸ {TRI000001784}. but that did not constitute an adequate notification of the work at Grenfell Tower. Cadent should have produced an F10 to notify the HSE of that work.⁸⁹⁸⁹ Lane, Phase 2 Report Version 2 {BLARP20000016/148-149} paragraphs 11.3.1-11.3.9. A generic F10 of the kind produced by tRIIO was inappropriate for a project affecting a large building in multiple occupation.⁹⁰⁹⁰ Harrison {Day159/84:5-14}; Dolan {CAD00003044/13} page 13, paragraphs 51-52.

¶68.36 Over the course of some months the original riser was replaced using a new pipe which entered the basement below ground level on the east side.⁹¹⁹¹ Hancox, Gas Engineering Report {RHX00000012/14} paragraph 29 (c) and figure 4. From there it passed up through the lower floors and into the stairwell where it passed through holes cut in the floors of the landings.⁹²⁹² Hancox, Gas Engineering Report {RHX00000012/31} figure 24. If the supply of gas was to be restored to a flat, a horizontal pipe, or 'lateral' was led from the riser through the stairwell wall and into the lobby.⁹³⁹³ {CAD00001643}; {CAD00001640}. The lateral then passed round the perimeter of the lobby just below the ceiling and a pipe from it entered the relevant flat through a hole in the wall between the flat and the lobby.⁹⁴⁹⁴ {TRI000000263}; {TRI000001431}. The pipe was connected to a meter which had to be moved from its original position in the kitchen to the hallway near the front door.⁹⁵⁹⁵ Hancox, Gas Engineering Report {RHX00000012/26} figure 18. Further pipework was then installed between the meter and any kitchen appliances that burned gas.

¶68.37 After the gas pipework had been installed into the tower it was commissioned on 10 March 2017,⁹⁶⁹⁶ Dolan {Day160/125:7-9}. even though the work in the lobbies of boxing it in had not been completed. At that stage the work did not comply with the guidance in IGEM/G/5 because a leak in the lobby would have been ventilated by mechanical ventilation, which was not permitted. It appears that tRIIO took a calculated risk and reconnected the supply of gas to customers before all the work had been completed. We understand the desire to reconnect occupants as quickly as possible, but the fact remains that tRIIO should not have introduced gas into a pipeline without all the ventilation requirements being in place.⁹⁷⁹⁷ Dolan {Day160/125:7}-{Day160/126:10}.

A change of design

¶68.38 On 21 March 2017, tRIIO's design manager inspected the work and noted that flanged joints had been installed rather than welded joints.⁹⁸⁹⁸ Dolan {MET00012711/11} page 11, paragraph 47. Flanged joints have a compression fitting which seals the connection between the two pipes with a gasket, as opposed to screwing or welding the pipes together.⁹⁹⁹⁹ Dolan {Day160/140:16}-{Day160/141:10}. As a result, flanged joints are more prone to leaking and could not be used if the stairwell was to provide the necessary ventilation.¹⁰⁰¹⁰⁰ Dolan {Day160/139:23}-{Day160/140:10}. The installation of the flanged joints compromised the ventilation strategy and the error was not noticed by tRIIO's design team before gas was reintroduced into the pipeline.¹⁰¹¹⁰¹ Dolan {Day160/141:20}-{Day160/142:8}.

¶68.39 On 24 March 2017, in response to that discovery, tRIIO introduced a modification to the arrangement, under which the vertical riser (as well as the laterals) was to be fully boxed-in, both in the stairwell and also throughout the lower levels and the basement, using materials with a two-hour fire rating.¹⁰²¹⁰² Dolan {Day160/144:13-25}; {Day160/145:15}-{Day160/146:3}; {TRI000001223}. Intumescent vents were to be installed in the boxing in the stairwell, which would close in the event of a fire but allow for some ventilation in normal circumstances.¹⁰³¹⁰³ {TRI000001223}.

¶68.40 The description of the modified arrangement did not show how the boxing-in was to be achieved at the lower levels, i.e. through the utility duct and in the storeroom and basement. Nor did it show how the boxing-in would be connected to the vent in the roof.¹⁰⁴¹⁰⁴ Dolan {Day160/157:16-23}; Hancox, Gas Engineering Report {RHX00000012/194} paragraph 391. Moreover, the new design created other problems in relation to ventilation. It assumed that air would circulate within the boxing and across the pipework through the holes in the landings and the walls between the stairwell and the lobbies, but the holes through the landings had not been enlarged to allow for a flow of air within the new boxing.¹⁰⁵¹⁰⁵ Hancox, Gas Engineering Report {RHX00000012/195-196} paragraphs 395-396; Hancox {Day161/158:3-16}. There was no evidence that calculations had been carried out to determine the size of the holes that would be required nor was it clear how an increase in the size of the holes and the boxing could be accommodated given the narrowness of the stairs.¹⁰⁶¹⁰⁶ Hancox {Day161/161:3-13}. The use of intumescent vents was not consistent with the requirements of IGEM/G/5.¹⁰⁷¹⁰⁷ Hancox {Day161/154:15}-{Day161/155:4}.

¶68.41 In this case tRIIO manually over-rode its CDM Design Risk Register process and carried out a full design risk assessment that was recorded on a four-page spreadsheet identifying factors for consideration.¹⁰⁸¹⁰⁸ {TRI000001218}. Three of those factors related specifically to buildings in multiple occupation, namely, breach of compartmentation, failure due to thermal expansion of south-facing pipes, and ventilation.¹⁰⁹¹⁰⁹ {TRI000001218/4}. They were appropriate matters but they did not cover everything that a designer ought to consider, such as, for example, means of escape in fire, corrosion, inspection and maintenance, and valve access and security.¹¹⁰¹¹⁰ Hancox, Gas Engineering Report {RHX00000012/156} paragraph 354; Dolan {Day160/171:1-15}; Hancox {Day161/173:3-9}.

¶68.42 In addition, the information tRIIO added to the Design Risk Register lacked sufficient detail. On the important question about fire compartmentation, the action recorded was simply to follow the Building Regulations and Fire Safety Order and seal the compartments accordingly. A gas engineer was to review the arrangements for ventilation, although they were not described in any detail in the design. Mr Dolan accepted that the assessment process fell below the standard required and that a more detailed consideration of matters affecting buildings in multiple occupation was needed.¹¹¹¹¹¹ Dolan {Day160/162:10-19}; {Day160/170:15-25}.

¶68.43 Although tRIIO adopted a more formal approach to the modified arrangement, there is no evidence that anyone calculated the amount of ventilation needed for the pipework.¹¹²¹¹² Dolan {Day160/186:9}-{Day160/187:9}; Hancox {Day161/156:14-25}. That was particularly important, given that the ventilation strategy had changed and that the degree of ventilation had been reduced by the boxing-in of the entire pipeline.

Construction of the modified arrangement

¶68.44 Regrettably, tRIIO and its sub-contractors continued to make mistakes during the construction of the work. As Mr Dolan and tRIIO accepted,¹¹³¹¹³ Dolan {Day160/189:21-25}. although the boxing-in of the pipework should have been carried out using two-hour fire-rated materials¹¹⁴¹¹⁴ {TRI000001218}. (and although the TMO was told that it would be),¹¹⁵¹¹⁵ {TMO10016548}. the sub-contractors, Express Building Contractors Ltd, in fact used materials that were rated as having a fire integrity of only 88 minutes.¹¹⁶¹¹⁶ {TRI000001377}; Dolan {Day160/189:14-25}; Hancox, Gas Engineering Report {RHX00000012/196} paragraph 397.

¶68.45 A further problem also arose from the need to move the gas meters from the kitchens closer to the front doors because tRIIO's specifications required that there be no more than two metres of service pipework inside a residential property.¹¹⁷¹¹⁷ Dolan {Day160/72:24}-{Day160/73:10}. As a result, a sub-contractor of K&S Pipe Contractors, Holland Gas Engineers Ltd, was obliged to place meters in or adjacent to the entrance hall, which was the only escape route.¹¹⁸¹¹⁸ Hancox, Gas Engineering Report {RHX00000012/33} paragraph 61. Meter points can present a particular fire risk because if there is a fire, gas can burn freely from open meter ends. IGEM/G/5 suggests relatively easy mitigations, including installing the meter inside a 30-minute fire-rated cupboard with a self-closer or installing a thermal cut off device¹¹⁹¹¹⁹ {RHX00000005/38} paragraph 5.2.3. to stop the flow of gas either before or after the emergency control valve.¹²⁰¹²⁰ Hancox, Gas Engineering Report {RHX00000012/58-61} paragraph 133(f). However, neither of those steps were considered at Grenfell Tower.¹²¹¹²¹ Hancox, Gas Engineering Report {RHX00000012/161} paragraph 364.

¶68.46 tRIIO maintained that the meters were often sited in an alcove or cupboard adjacent to the entrance hall and that the mitigation measures mentioned above were unnecessary.¹²²¹²² Dolan {Day160/190:18}-{Day160/191:1}. However, we have seen examples of meters being located behind curtains¹²³¹²³ {MET00016651}. or behind doors that had rising hinges and were not self-closing¹²⁴¹²⁴ {MET00016558}; Hancox, Gas Engineering Report {RHX00000012/159} paragraph 362. and it is plain that in some cases the meter was sited close to the escape route from the flat. Although there is no evidence that any of the meters played a role on the night of the fire, we do not consider those arrangements to have been satisfactory.

The night of the fire

¶68.47 There is no evidence that gas leaked from any part of the new installation during the fire on 14 June 2017.¹²⁵¹²⁵ Hancox, Gas Engineering Report {RHX00000012/117} paragraph 252. However, the works had not been completed by that time. The boxing-in in the stairs had almost been completed, with only the top floor and the connection to the roof vent outstanding, but the boxing-in of the laterals in the lobbies had not been completed, except on floor 5. Battens had been installed for the boxing-in in the lobbies on floors 4, 8, 9, 10 and 11 only. Mr Dolan estimated that the remaining work would have taken about five weeks.¹²⁶¹²⁶ Dolan {TRI00001797/2-3} pages 2-3 paragraph 7; Mason {CAD00003005/11} page 11, paragraph 34.

¶68.48 While the work was being carried out the oversized holes between the stairwell and the lobbies were not temporarily fire-stopped. It was possible, therefore, for smoke to move from one compartment space to another. Heavy smoke staining on the internal faces of the boxing in the stairs indicates that some smoke must have travelled from the lobbies into that boxing.¹²⁷¹²⁷ {MET00016722}; {MET00016516}. There was also evidence of staining between adjacent floors which appears to indicate that smoke had travelled between floors through the boxing of the riser.¹²⁸¹²⁸ {MET00016516} showing smoke staining at the top of the boxed portion of gas riser on floor 17; {MET00016533} showing staining at the bottom of boxed portion of gas riser on floor 16; Hancox, Gas Engineering Report {RHX00000012/178-179}. It is possible, therefore, that during the fire smoke may have travelled from one lobby into the boxed-in riser and out into lobbies on other floors higher up the tower, thereby compromising the compartmentation between the lobbies. It is not possible, however, to know how much smoke may have travelled in that way or at what stage in the fire. We think it is likely to have been a much less significant route of smoke spread than open fire doors between the lobbies and the stairs.

¶68.49 In March 2017 Cadent expressed concern to tRIIO about the oversized holes between the lobby and the stairwell, which it thought might create an unsafe situation in the event of a fire.¹²⁹¹²⁹ {TRI000000985/3}. In those circumstances we think that tRIIO should have made sure that those holes were temporarily fire-stopped while the work remained incomplete,¹³⁰¹³⁰ Hancox {Day161/166:8}-{Day161/170:6}. as Mr Dolan accepted.¹³¹¹³¹ Dolan {Day160/175:15}-{Day160/176:20}. We are unable to tell whether the failure to do so had any adverse consequences on the night of the fire, but we think it important to record that the risks should have been identified and mitigated during the work.

The involvement of building control

¶68.50 One question we thought it appropriate to consider is whether the work involved in the replacement of the gas riser required building control approval. In 2016, the gas industry did not routinely consult building control in relation to reinstatement work.¹³²¹³² Hancox, Gas Engineering Report {RHX00000012/187} paragraph 375. In this instance, although Carl Stokes advised that the work was notifiable,¹³³¹³³ Stokes {CST00003063/47} page 47, paragraph 138; {TRI000000405}. tRIIO did not consult building control because it thought that the work did not require approval, despite the material changes to compartmentation that the riser design entailed.¹³⁴¹³⁴ Dolan {Day160/110:20}-{Day160/111:8}. In March 2017 Janice Wray told John Allen, the head of RBKC's building control department, that the new riser had been installed in the staircase and invited him to send a team to the site to look at the work.¹³⁵¹³⁵ {TMO10016428/1}. However, he told her that the riser works were regarded as a repair and were a matter for a fire risk assessment.¹³⁶¹³⁶ {TMO10016546}. tRIIO, for its part, appeared to believe that, if carried out in accordance with IGEM/G/5, the work would comply with the Building Regulations but that the TMO would make any necessary application for building control approval.¹³⁷¹³⁷ {TRI000000405}; Dolan {Day160/111:1-8}.

¶68.51 Whether the work of replacing the gas riser required building control approval is a difficult question on which the evidence does not enable us to express a clear conclusion. However, we think that in cases where the structure of a building on which effective compartmentation depends is affected by the replacement of existing services, careful consideration should be given to the need to obtain building control approval as well as complying with any relevant industry guidance.