Automated warehouses

Ocado warehouse fire: The questions that need answers

Avatar photo

Fire-safety consultant

Author Bio ▼

Alan started his fire and safety career with Warwick County Fire Service in 1963 and served as both an Operational and Fire Safety Inspecting Officer. In 1976 he transferred to the West Midlands Fire Service until 1978 when he moved to the NHS as the District Fire Safety Officer for West Birmingham Health Authority where he was also the West Midlands Regional Health Authority Fire Advisor. During his NHS career he worked and studied for six months in the USA looking at different approaches to fire safety. He was also responsible for developing a computerized hospital fire evacuation program that was used in many major hospitals. In 1994 Alan moved to HSBC as its Senior Fire and Safety Officer responsible for the 80 countries in which the bank had a presence. During his career with HSBC he established a global approach to fire safety, organized many international fire and safety conferences, and developed a standardized method of protecting computer areas from fire. In 2005 he set up his own Fire and Safety Consultancy. During his career he has published a number of books on fire safety and made many specialist technical videos on subjects such as hospital evacuation, fire protection of electronic data protection areas, fire doors, and mail room safety. He has been awarded a Brooking NHS Travel Fellowship, Rospa Safety Professional of the Year (twice), FPA Premier Fire Safety Award, and The Prime Minister's Quality Initiative. He also contributes to many fire and safety journals including Fire, IOSH, Fire Surveyor, and Health and Safety Journal. He is a fully qualified Fire Service Inspecting Officer, member of the International Institute of Risk and Safety Managers (MIIRSM), Tech IOSH, and Qualified Fire Investigator. Alan has advised many large companies including the National Trust, Hospital Corporation of America (HCA), Kings College, Cambridge, Briton Hardware, BUPA, British Antarctic Survey Expedition, Chubb, Central Television, BBC, Radisson SAS, and the Falkland Islands Police.
July 25, 2019

Download

Whitepaper: Enhancing security, resilience and efficiency across a range of industries

(Image credit: Techwords under CC BY-SA 4.0)

The blaze that destroyed the Ocado Customer Fulfilment Centre (CFC) in Andover Hampshire is yet another warehouse fire that raises myriad questions.

A full and transparent report is needed if we’re to truly learn from this fire, which cost the online supermarket an estimated £100m, and incorporate better practice into similar projects.

This building was the prototype for Ocado’s robotic dream. Footage released when the building opened in 2016 showed an army of 600 robots swarming over a 3D frame called the ‘grid’ to locate goods, from tins of baked beans to toilet rolls (see video below). Travelling at up to four metres a second, the robots allow a typical order to be filled in around five minutes.

Hampshire Fire & Rescue Authority described the CFC as a refurbished, extended, high bay warehouse of about 500,000m3 and a “large and complex building”. It also reported that the building was fitted with an “air sampling fire detection” system and sprinklers – so how was this building destroyed?

The cause and spread of the fire

Media reports indicate that the fire occurred because of an “electrical fault with one of the first-generation battery charging units at the edge of the ambient storage grid”, which caused the “plastic lid on the top of a grocery-carrying robot to catch alight”.

So why didn’t the air-sampling detection system detect the fire and prompt further action? Why didn’t the sprinkler system control the fire until the arrival of the fire and rescue service (FRS)?

A BBC programme informs us that a report from Hampshire FRS indicated that the fire started at 01:41 GMT on 5 February, but the fire detection system “did not detect the fire as designed” and it was first spotted by an engineer at 02:15.

The sprinkler system started operating 11 minutes later, but was then turned off by Ocado engineers for five minutes, which led to a “significant” growth in the fire. Once staff realised the fire was not being extinguished, they turned the sprinklers back on and finally dialled 999.

The report also found there was an hour’s delay in dialling 999 and staff initially tried to tackle the blaze themselves. If these statements are correct, then surely we need to know the reasons why these things occurred?

A report by the Business Sprinkler Alliance stated: “We understand that an automatic sprinkler system in combination with the fire and rescue service initially contained the fire in the building, which broke out around 2:45am on Tuesday, 5th February. However, the fire escalated late Wednesday afternoon, resulting in substantial damage to the premises.”

This report appears to be at variance with the report from Hampshire FRS.

A report in The Times quoted an unnamed firefighter saying robots were still moving across the grid when they entered the complex, hampering their efforts. However, Ocado maintains that the warehouse robots were immobilised by a safety system as soon as the fire alarm was activated. It’s important to note that Hampshire FRS has indicated they are unaware of this firefighter’s claim and that the article is subject to a legal complaint from Ocado Group Plc.

This was a statement made by Neil Odin, chief officer of Hampshire FRS: “The blaze began high in the warehouse and was so difficult to reach that firefighters had to cut holes in the roof”.

The building and fire protection

Very little is known about the construction products used in the building or the fire detection and protection system. But when I’ve specified air-sampling detection systems in large warehouse type buildings, I’ve made the suppliers conduct practical tests in the building showing how their systems could detect both electrical and paper fires – and the results revealed significant differences in performance.

When I’ve specified air sampling fire detection systems in the past I ‘ve always outlined the fires that I wanted the system to detect and let the installation companies design the system. On commissioning the system, I have replicated the fires that I expect may occur and depending on the air flow, positioned the tests in the least favourable position. It’s surprising how many times systems fail using this method.

Looking at why the sprinklers failed, clearly turning them off for five minutes would have been a major contributory factor.

It’s also interesting that Ocado won a ‘Highly Protected Risk Award’ from FM Global for the extensive sprinkler system they installed in the Andover facility in 2018.

The HPR award states that a building meets the highest possible industry standards for property protection against fire. It’s not known how the sprinklers were positioned in the warehouse or if they were affected by air movements.

The BBC reported that the fire started at 01.41 but was not spotted until 02.15, by an engineer. If they know exactly when the fire started, there must have been some way this was recorded, so perhaps some CCTV footage is available.

Not the first time

Unfortunately, this is not the first time sprinklers have been turned off in Hampshire. This report was published in The Independent on Friday 29 March 1996:

“Hampshire County Council was today facing a compensation bill of up to pounds 12m after a judge held the county’s fire brigade liable for the partial destruction in a blaze of a prestige company headquarters in Basingstoke.

“It is believed to be only the second judgment ever obtained against a fire service in the United Kingdom for alleged negligence in the course of its work. The council is expected to appeal.”

The fire broke out in the roof of the state-of-the-art headquarters of Basingstoke of Digital Equipment Ltd, known as The Crescent, in March 1990. The owners, Capital & Counties plc, had fitted an automatic sprinkler system which, it was claimed, would have limited the damage to a small area of the roof void if it had not been turned off by the fire brigade.

The allegation was that the firefighters stopped the system while the fire was still blazing in the erroneous belief that there were no sprinklers in the roof void; that the firefighting was somehow being hampered by sprinklers at first-floor ceiling level; and that these could not be isolated and turned off without shutting down the entire system.

Judge Richard Havery QC, sitting as a High Court Official Referee, held there was no justification for the fire officer in charge to depart from the principle that sprinklers should be kept running until a fire was completely under control. He said the decision was “a bad blunder”.

The judge rejected an argument by the fire brigade that it owed no legal duty to the owners of the building and, like the police, was immune from being sued as a matter of public policy.”

Insurance

According to another report from BBC News, Ocado has already claimed millions of pounds back from its insurers and eventually expects to reclaim all the cost. As someone recently said to me, “if I left my car outside with the keys in and it was stolen – I wonder if the insurance company would pay up?”

2023 Fire Safety eBook – Grab your free copy!

Download the Fire Safety in 2023 eBook, keeping you up to date with the biggest news and prosecution stories from around the industry. Chapters include important updates such as the Fire Safety (England) Regulations 2022 and an overview of the new British Standard for the digital management of fire safety information.

Plus, we explore the growing risks of lithium-ion battery fires and hear from experts in disability evacuation and social housing.

FireSafetyeBook-CoverPage-23
Subscribe
Notify of
guest
8 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Paul Evans
Paul Evans
July 25, 2019 12:34 pm

I maintain an independant early intervention Pressure and smoke release system would have released all the contributory components of eventual fire spread and collapse. Indeed such a system may have allowed first responders to extinguish the Fire with the assistance of the in situ Sprinkler system.

John Griffin
John Griffin
July 26, 2019 9:54 am

It beggars belief that anyone would turn off the sprinklers when their function is to extinguish fire which, reading the article was still underway. Within the domains of First Aid, if blood comes through the dressed wound one shoves another one on top. The F&R intervention served to supplement the sprinklers. I am not a fire fighter but the action does not seem to make sense. I can well see were the robot trucks dodging about would hamper things but not the sprinklers.

Stephen Penney
Stephen Penney
July 26, 2019 11:59 am

There seems to be an expectation that simply fitting a warehouse with sprinklers and an aspirated smoke detection system will guarantee adequate fire protection – However, as the author quite rightly comments, there can be considerable variation in the effectiveness of a system. Design of an appropriate fire protection system for such a building needs a significant amount of work before the specification phase and, if you have a totally new risk building design or use fire modelling and tests!

Ganesh Nilangekar
Ganesh Nilangekar
July 28, 2019 6:46 am

Pathetic state of affair, Fire Service need to be educated a lot and owners shall not take Fire Systems as show pieces, its functional value shall be evaluated and recorded. Otherwise noting will change. Good judgment by judiciary.

Alan Cox
Alan Cox
July 29, 2019 9:13 am
Reply to  Paul Evans

Paul, The problem is that we do not have enough information about how the building was protected, what it was constructed of and what fire safety measures were considered. It could be that the measures that you suggest would have helped but because of the lack of information we don’t know the full picture so can only guess. This of course is one of the problems with how we investigate serious fires in the UK – much of the investigation is carried out by the FRS and this is not the best solution as they have a “vested interest” in… Read more »

Alan Cox
Alan Cox
July 29, 2019 9:30 am
Reply to  Stephen Penney

Stephen, I could not agree more with your comments. When I have designed such systems I have gone to great lengths to make sure that the level of protection is both effective and suitable for the risk and whilst cost was important – it was not high on my list of priorities. Having done the initial risk assessment I then produce a matrix of how both the fire protection and human intervention (if any) will work and once this is agreed with both the contractors and users the project proceeds on that basis. When the commissioning is carried out I… Read more »

Brian Galsworthy
Brian Galsworthy
August 3, 2019 12:05 pm

WHY THIS REPORT FAILS THE “INDEPENDENCE TEST” • The Review Team is Hampshire Fire & Rescue Service (HF&RS), “the insurer”, and FM Global. Ocado and BRE consulted.(Reuters) • FM Global designed the sprinkler system • FM Global gave the sprinkler system the HPR (Highest Protected Risk Award) • FM Global is “the insurer” (Insurance Times) • FM Global have subsequently demonstrated that ceiling only sprinklers do not adequately control fires in high-density storage of open-top combustible containers https://youtu.be/xzuNVFn0U48 • HF&RS are the Regulator for this facility • HF&RS were the responding service judging their own performance (with potential implications for… Read more »

Nick Wildman
Nick Wildman
March 6, 2020 9:06 am
Reply to  Alan Cox

Totally agree but it can be a real challenge to carry out effective testing in situ. I have witnessed cold smoke tests for extraction systems in basement car parks which achieve nothing due to the obvious differences between hot and cold smoke.
Training and testing has to be as close to reality to be effective but interested to know how you would test a combustible open top container situated at high level in this setting?