At Aviva headquarters in London, RISCAuthority held its latest member seminar, and William Roszczyk writes of the diverse subjects under discussion
CHAIR DAVID Williams, technical advisor at AXA Insurance, welcomed members, stating the organisation continues to do ‘good and invaluable work’ and is currently conducting research into best practice. He was quick to ask whether the industry is ‘doing enough about fire’, and hoped attendees would be ‘better informed afterwards on current issues’, while fire was ‘not singled out’.
Chris Hanks, chairman of the FPA, began by showing a video of London firefighters attending Grenfell Tower, noting that many assumed such a fire ‘couldn’t be in London’, and that from an insurer’s perspective, the fire was foreseeable and preventable. Mr Hanks reflected that after use of combustible materials and ‘years of neglect’, there was now ‘widespread culpability’.
Having collected anonymised information from insurers, he started with what they didn’t do: panic or run for cover, change underwriting strategy or decline business, this continuing ‘as usual’ alongside ‘immediate first aid’. A leading thought was likely ‘is it our building?’, while insurers would also have wondered ‘who could be involved’? or ‘who’ll get dragged in?’.
They would have searched to see if they were responsible within a database of high risk, and identified similar characteristics in buildings in their portfolios.
With board pressure, risk assessments of similar buildings would have followed alongside studies of desktop surveys and whether EML (estimated maximum loss) figures were right.
In addition, they would have searched for clues and understanding, in order to give advice to policyholders and brokers. Insurers would have questioned how the fire happened, and advised staff and clients particularly not to panic. In Mr Hank’s opinion, insurers ‘acted responsibly’, and would have found it hard to ‘identify similar risks’ with a lack of ‘detailed construction information’. Additional underwriting action was needed for 25% of risks identified, and some would be outside ‘risk appetite’.
This reinforced a need to ‘update underwriters on guidance for high rise buildings and construction’, many having a ‘lack of technical skills or capability to understand’ risks. There is also a ‘poor understanding of building regulations’, while the testing regime is ‘woefully inadequate’, using ‘perfect build’ tests that bear ‘no relation to real life’. This confirmed his ‘intuitive feeling that UK PLC has, over a long time, neglected property protection’.
Actions to take include to ‘improve training and expertise in construction types’, networks and combustible materials, ‘use data better’ and improve collection. Revision of rules and calculations of EML for high risk buildings, alongside retention limits and reinsurance purchases, should happen alongside realignment of ratings and acceptance rules for buildings with certain characteristics.
He recommended ‘strict controls for property owners and managers’, looking at the effectiveness of fire alarms (as 95% are false), as well as the ‘overriding need to share information’ and working ‘more collectively as an industry’. Referencing the FPA’s testing work funded by the Association of British Insurers (ABI), Mr Hanks stated that the FPA submitted documents to the terms of reference for the Grenfell inquiry, and is participating in both the enquiry and the expert panel.
It identified 14 research themes, and Mr Hanks pointed out that ‘after 15 years of banging our heads against the door’ in relation to combustible materials used in buildings, ‘at last this is our moment for change’, though ‘sadly it takes a tragedy to make changes’. The inquiry ‘should raise questions of all’ involved, as ‘if insurers don’t worry about property losses, who does?’
As ‘custodians’ in this sense, ‘why do we allow politicians to remove protection?’, he asked, and questioned why insurers cut premiums to chase business, as well as dumb down skills and standards. He asked whether more money in the sector ‘meant not enough attention’ was paid to technical aspects, and queried whether insurers have ‘abdicated underwriting’ responsibility.
Insurers ‘knew major loss of life in fire would happen’, and have been ‘non assertive co-pilots’ turning a blind eye to ‘obvious disaster’. All the issues surrounding the fire had been identified, and ‘known about for years’, Mr Hanks stating he was ‘afraid there are things we could have done’.
On ‘what insurers want’ from the inquiry and regulations review, this would include a ‘root and branch review’ of the latter, as well as the importance of property protection, a move to ‘reintroduce’ local acts, which may have meant the tragedy ‘probably wouldn’t have happened’, alongside an ‘independent, competent and reliable’ testing regime, ‘more sprinklers’ and ‘action on false alarms’.
The overall cost is ‘unknown’, but Grenfell Tower was likely worth £25m to £50m, with a ‘ludicrous’ £5,000 premium, a consequential loss of £200m, and a liability of up to £500m. The costs to UK PLC, society and families however were ‘enormous’. On this, he argued ‘everyone is running for cover, particularly politicians’, and argued ‘how could we let this happen in the 21st century? Basic requirements were ignored, and we should resolve never to let us be involved in this ever again’.
Insurers ‘should contribute’ in a ‘better way’ to society, and increased cooperation would ‘make a difference’, as we ‘need all stakeholders’ to work together.
Dr Paul Jowett of Burgoynes next examined factors contributing to severe fires in commercial kitchens, alongside practical measures to prevent or reduce the impact of such incidents, and ways in which improvements can be made. Common causes include human error, loss of thermostatic control, gas equipment failure and fume extraction equipment.
Equipment such as deep fat fryers are typically fitted with two thermostats: a ‘working temperature’ device allowing set point temperatures up to 205°C that cycle heating to maintain said set point, and an ‘overheat limit’ thermostat, typically providing an upper limit of 230°C, and which should be of a non resetting type.
In some cases, the initial fire is confined to the immediate point of origin, with no significant development beyond there, but as such fires are often caused by cooking equipment placed under a fume extraction canopy, flames can easily touch on the canopy’s underside. Guidance found in the Heating and Ventilating Contractors’ Association (HVCA) publication, DW/172: Specification for Kitchen Ventilation Systems, states: ‘Primary filters that retain grease within the filtration matrix until cleaned shall not be used. The mesh type filter shall not be suitable except for secondary filtration.’
Additionally, LPS 1263: Issue 1.1 gives requirements for the LPCB approval and listing of the fire performance of kitchen extract systems. This describes tests to determine grease removal efficiency and flame resistance of grease filters. There is also the HVCA TR/19: Guide to Good Practice: Internal cleanliness of ventilation systems, which notes typical cleaning intervals are weekly for primary filters and accessible parts of the canopy, and every six months for ductwork.
Suitable access to the ductwork is integral, and DW/172 advocates access panels at minimum every 3m. In relation to fire spread, Dr Jowett recommended Approved Document B and Approved Document J: Combustion appliances and fuel storage systems, with the former considering effects of ducting passing into protected means of escape areas, and advocating fire resisting enclosures and ductwork.
However, where ductwork passes through a roof structure, requirements are not so clear, one example being that a fire burning within the canopy and ductwork of a fume extraction system is similar to a flue/chimney fire. In these circumstances, he asked, should we apply the same requirements to ductwork as to flues in Approved Document J? According to DW/172, if a fire originates in or transfers to ductwork, it may spread beyond its original location, so the ductwork route must take account of this and include a minimum separation of 500mm between uninsulated ductwork and any combustible material.
Finally, after taking a look at fire suppression systems, Dr Jowett summarised his recommendations as: the use of ‘fail to safety’ thermostats and ‘baffle’ type primary canopy filters, suitable access openings for cleaning ductwork, and suitable separation between ductwork and combustible material, as dealt with in DW/172 and TR/19.
Graham Brogden, head of technical claims at Aviva, expanded on a similar topic from the Manchester seminar in June on flood resilience, ‘following on from previous talks’ and updating on the organisation’s plan. One in six are at risk from flooding, or five million people and 30,000 businesses, while in 2015, 50,000 properties suffered damage from flooding, and business losses continue.
The frequency of severe storms is increasing, so now is the ‘time for resilience’, as ‘if we don’t address the issues, these numbers will increase’, and this would require a ‘change in behaviour’. The recovery process requires an ‘extreme level of strip out’ – technology should be used to improve this, yet replacement with non resilient materials or unnecessary removal continues, meaning there are the ‘same results each time’.
As standards are few and far between, there are also scenarios where the right materials are used but installed poorly, and the emotional impact of floods encompasses the whole community, from children to local businesses and the vulnerable. Giving the example of Kendal in Cumbria, which saw roads and bridges washed away, he said the impact on local people was huge, with many forced to move away and families broken up, as well as loss of business for small companies.
After the winter 2015/16 floods, DEFRA’s property flood resilience five year action plan (2016) aimed to promote a better understanding of property level resilience and related systems and practices in the insurance, building and finance sectors, as well as equipping those at high flood risk to adapt their properties to limit the flood damage and speed up their recovery.
In essence, the aim was that an environment where it is standard for properties at high flood risk to be made resilient would become available, with this plan delayed after the general election earlier this year. Mr Brogden stated that the situation was ‘not a sprint, but a marathon’, with flood roundtable task groups building community initiatives, a flood ‘front face’ and national portal, standards and certification, communications and behaviour change, and data/evidence, eg case studies.
The idea was to move people on from flood resilience grants to setting up community initiatives and to strengthening learning about resilience in practice.
The structure of the plan was such that it has yearly aims, with the first year’s being a ‘better understanding of property resilience’, and the second’s ‘significant progress on insurance, building and finance systems’. Looking ahead, within three to five years ‘more knowledge and capability’ was expected, with the plan ‘standard practice’ in five years.
Peter Williams, insurance business development manager at VPS Group, gave a short overview of risks facing vacant properties, including that those without regular inspections and maintenance have significantly increased risk of water damage from burst pipes or undetected water leaks. One undetected roof leak and burst pipes in a pub/nightclub caused £250,000 of damage because weekly inspections had not been undertaken.
One in six UK properties is at risk of flooding, with the average flood damage claim from the 2013/14 winter estimated at £24,000. On fire, 60 blazes take place each day in or beside vacant or derelict buildings, while 11 construction sector fires per day are deliberately lit, costing £400 million a year. Vagrants light fires to keep warm, and stolen vehicles are set on fire in or next to sites.
One fire in a vacant retail unit occupied by squatters, which had not been inspected since the previous tenant left, led to a total loss of the unit, while over 40% of vacant commercial premises are subject to vandalism, alongside theft, costing the construction industry £1m per day. Metal thieves target historic buildings, churches, out of town retail units, and vacant or isolated industrial units, with one theft worth £80,000 from vacant offices leaving ‘severe damage’ and a £2m repair bill.
Another example given was extensive damage and fly tipping over two months at an industrial/office unit, which went undetected and cost £200,000 for waste removal plus 24 hour security. With around 20,000 squatters across the UK, since squatting in residential property became illegal in 2012, commercial premises have been targeted, resulting in an estimated doubling of legal proceedings. It costs an average of £7,000 in legal fees to remove squatters, with clean up on average £32,000.
Illegal raves and trading often start as a squat in a large office or industrial warehouse, with events marketed on social media but the location revealed only on the day. Premises may have no fire protection or health and safety provision, and clean up costs can reach over £100,000.
Mr Williams highlighted a Croydon rave that resulted in the death of a reveller, as well as a £500,000 bill for one owner after a rave in a small Hackney warehouse.
Property owners have a duty of care to protect those entering a vacant property under the Defective Premises Act 1972 and Occupiers Liability Act 1984, with these including estate agents, surveyors, members of the emergency services, trespassers, and vandals. Should a defect in a vacant building cause someone injury or damage, a legal claim could be made against the owner.
Mitigating measures, Mr Williams stated, include risk assessments, removing combustibles and instigating security measures, while properties should be regularly inspected and properly maintained, utilities turned off and drained down, and letterboxes secured. Advantages and disadvantages of physical (security guard, steel security screens) or technological (alarms, CCTV) security were shared, while fire safety trained, live in guardians were focused on last.
These are especially popular in city centres, and around 6,000 currently occupy vacant buildings in the UK, with guardians paying a licence fee and subject to referencing and various checks. The buildings they occupy need to be wind and water tight and comply with house in multiple occupation (HMO) standards, while utilities must meet required safety standards.
Internet of things
Ruby Ghunia, digital insurance analyst for Global Data, explored the idea of the internet of things (IoT) in insurance.
As a network of connected devices and sensors, IoT technology is ‘not new’, but has seen dramatic growth through mobile devices and wireless technology.
While commercial insurance has already got into IoT, there was the ‘potential in industrial insurance to dwarf consumer insurance by several magnitudes’, with some already using it for fleet or building management or security. Ms Ghunia also gave a case study of a US property company that utilised the technology to warn residents and tenants of a hurricane.
The implications of the technology on insurance can go both ways, to ‘risk or reward’, with the role of insurers themselves shifting from carriers of risk to ‘orchestrators of prevention’. Rewards include real time risk monitoring and pricing, as well as customer engagement, while one risk is that premiums would see contraction.
Data may ‘reshape the value chain’, and she stated it will disrupt but also ‘drive actionable intelligence’, with predictive analytics changing the way insurers interact with customers. In commercial insurance – where the technology is used in auto telematics for young drivers, healthcare and smart homes – around 67% of organisations are considering entering the market in the next two years.
An incremental introduction with ‘future proofing’ is needed, but ‘opportunities abound’ in advantages, including asset tracking management, predictive maintenance and smart buildings, as well as impacts on industries including property, work compensation and construction, marine, aviation and automotive sectors.
By monitoring risks in real time, companies could offer ‘real time insurance’, though one issue is that most businesses have sensors and technology, so why haven’t insurers been able to get that information thus far?
With commercial insurance, there are ‘tremendous’ opportunities but also challenges in terms of convincing customers of ‘related gains’, as well as securing and sharing data from disparate sources. Infrastructure and technological barriers can also provide problems, as can multiple parties being involved.
Ms Ghunia stated that ‘quite frankly’, the sector was ‘not the best place to use data’, but it was trying to modernise and open up to new sources. She named a series of fundamental points for deployment, noting that insurers should be proactive as there’s ‘no going back’.
First was connectivity, while second was ‘data lakes’ requiring cross collaboration to access, and finally analytics – specifically artificial intelligence – mean correlations could be quickly identified. Building a system is one option alongside outsourcing, both having advantages and disadvantages in processing, construction, security and cost.
As new technology continues to emerge, this will lead to ‘new possibilities’ in data management, processing and more. IoT is ‘here to stay’, and insurers need to ‘remain relevant in an increasingly connected world’.
Mr Williams returned to speak on driverless cars, noting that insurers are involved in the technology as ‘we’re the people that pick up the bill’. Driver error causes 90% of all accidents and costs £16 billion each year, with thousands of people killed. In that sense particularly, reducing human error could bring down these statistics greatly.
The Automated Driving Insurance Group (ADIG) is focused on the technology, with Mr Williams believing the technology ‘will transform the insurance industry’, so ‘we need to be prepared’. A cross industry collaboration was begun, in which AXA wanted more direct involvement, so it and partner organisations obtained government funding.
BAE Systems’ Venturer project in Bristol, alongside Bristol Robotics Laboratory and AXA, is focusing on legal and insurance implications, while Oxford and Cambridge universities are looking at integrating vehicles into urban environments in a separate element. The Venturer project is focused on testing, rather than a production car, and ‘trying to bridge the gap’ or handover between drivers and technology, which insurers are most interested in.
Its test prospectus involves transport, time, testing and trust. Currently, trust is clear: the percentage of consumers who feel fully self driving vehicles will not be safe range from 62% to 81% internationally, though car manufacturers are still investing millions. Companies are also dealing with fictional depictions of hacking, which are likely to heighten concern among sceptics.
Autonomous car technology is expected to be ready by 2018, and the Institute of Electrical and Electronics Engineers anticipates that 70% of road vehicles will be driverless by 2040. Estimated timelines include: driver only or assisted (current), partial (2016), conditional (2018), high (2021) and full (2025), meaning assisted driving continues until 2018/19, with automated taking over from 2021.
On insurance, the Vehicle Technology and Aviation Bill 2016-17 contains new rules to ensure safe and effective insurance, including a compulsory requirement to protect victims in collisions that involve a highly automated vehicle, and a claims process not significantly different from that followed for conventional crashes.
Workshop and review
In the afternoon, FPA experts presented talks on different aspects and ramifications of the Grenfell Tower fire. The topics covered largely echoed those at the FPA’s Tall Buildings Seminar (you can read our report starting on page 46), with topics including legal and political background, managing risk challenges in high rise buildings, the Grenfell inquiry’s 14 main research themes, and cladding system specifics.
Dr Glockling also gave an overview of RISCAuthority’s key initiatives, noting it was ‘healthily’ expanding in terms of membership and well supported in working groups. The aim is to ‘extend it still further’, with 2017 looking at ‘exploiting better what we already have’. He stated that there are lessons to be learned across risk, with the organisation’s resources able to help, but ‘at the end of the day we have not been successful’ in influencing changes to the building regulations, though there is hope that the work the industry has done will be acknowledged during the review.
William Roszczyk is editor of Fire Risk Management. For more information, view page 5