IN 2006, Communities and Local Government’s (CLG’s) Sustainable Buildings Division commissioned the Building Research Establishment (BRE) to carry out a three-year research programme, Fire Spread in Car Parks. The results were presented at a one-day conference hosted by BRE in October 2009 and was attended by over 100 delegates to enable a forum for their discussion. This article is a summary of the conference and highlights the main issues raised by the research programme.
Conference summary
The first speaker, Brian Martin – a principal construction professional within the Sustainable Buildings Division at CLG – summarised the objectives of the research. He explained that the basis for the existing guidance for car parks, set out in Approved Document B1 (ADB) to the Building Regulations in England and Wales, relates to fires involving cars whose designs are decades old. A key objective of the project was to produce up-to-date experimental data representative of the current stock of road vehicles so that fire safety guidance can be reviewed and updated, where necessary. Martin Shipp, BRE project leader, then explained the background to the research. Mr Shipp said that, while there have been few deaths or injuries recorded in car parks in the UK to date, new research is necessary due to the changes in materials now being used in vehicles – for example, plastic fuel tanks, glass reinforced plastic panelling and alternative fuels.
He discussed a number of issues, including:
• is the fire load well defined?
• is there a ‘low probability’ of fire spread from one storey to another?
• what is the probability of fire spread between vehicles?
• how robust is the current guidance?
Next up, BRE consultant, David Crowder, described the literature review undertaken as part of the research. Currently, new cars are designed around safety, comfort, performance and the environment, but not fire safety, he explained. The use of plastics in vehicles is increasing and alternative fuels are now available to consumers. Liquefied petroleum gas (LPG), biofuels and electric/hybrid vehicles are gaining popularity, as a result of environmental concerns, he asserted. In addition, car park designs are changing and are often integrated into shopping centres, residential flats and office buildings. Mr Crowder went on to explain that in urban areas, where space is a factor, automated stackers are becoming commonplace, although fire development in such systems is not defined. A review of fire incidents in car parks followed.
This included: fire spread between multiple cars; fire spread through interlocking floors; significant structural damage due to fire; and firefighter fatalities. The review included discussion on an incident which occurred in the UK in December 2006. The incident resulted in a fatality in the sprinklered Monica Wills House care home2 which was built above an unsprinklered, partially open-sided car park where 22 cars were destroyed in a fire.
Statistical analysis
The statistical analysis of fires in UK car parks was then presented by BRE consultant, Róisín Cullinan. The data was gathered from fire brigade incident reports (FDR-1 forms) from 1994-2005 (inclusively) and produced the following key findings during the 12-year period:
• there were 3,096 fires in car parks in buildings – of which 1,592 started in a vehicle
• the average number of fires each year in car parks within buildings was 258, with data showing an overall decline of such fires over the period
• in total, two fatalities and 87 non-fatal injuries were reported for the period
• fires in car parks rarely spread to additional cars, however, where this does occur, significant structural damage is possible
• fires in car parks rarely spread to additional floors
• fires in car parks represent a very small percentage of all fires in the UK. For example, in 2006 they represented less than 0.1% of the UK’s total fires
Ms Cullinan then explained that there was a disproportionate number of injuries in car parks fires when associated with flats: fires in car parks in flats accounted for 5.5% of the total fires in car parks reported, but resulted in 26% of the injuries.
BRE carried out fire investigations on car parks over the three-year project, she explained. One fire investigation was of a car park fire which occurred in a busy shopping centre on 26 December 2007. The fire started in the engine compartment of a large vehicle and resulted in fire damage to multiple vehicles, with at least one vehicle being completely destroyed. However, a soft-top vehicle in the adjacent space to the destroyed vehicle was entirely undamaged, apart from soot deposits. While the experimental tests showed a high potential for fire spread, this real incident highlighted the lack of fire spread under seemingly conducive conditions. The low frequency of fire spread is also shown by the statistical data.
Modelling studies
Principal consultant at BRE, Richard Chitty, then reported on the results of computational fluid dynamics (CFD) studies. Mr Chitty explained that following an investigation into the Monica Wills House care home fire in 2006, BRE carried out CFD modelling of the incident. The model assumed that fire spread would occur to an adjacent car when a heat flux of 10kW/m2 was achieved on the exterior of the adjacent car (the value determined by cone calorimeter testing).
The car park in this incident was located beneath the residential care home with ventilation openings on two adjacent sides. The CFD model showed that, following the ignition of a single car due to an electrical fault, the fire spread to 21 other cars in the car park within 31 minutes. The CFD model of the fire ran until the 22nd car ignited, and the heat release curve which was generated showed a peak of just under 60MW.
He explained, however, that the fire was still growing at this point and that this result was consistent with the fire investigation findings.
Full-scale tests
Conference delegates were then shown footage of the full-scale tests by senior BRE consultant, Phil Clark. Mr Clark presented some of the key findings.
Tests 1-3, he said, were carried out in the Burn Hall facility at BRE Garston under a large 10MW calorimeter. The rig design was based on current car park design guides and BRE utilised CFD modelling to determine an appropriate level of ventilation (see Figure 1).
Test 4, which included an LPG car, was carried out in a similar rig without calorimetric measurements, in the open air at the Health and Safety Laboratory (HSL) in Buxton, Derbyshire, because of the associated explosion risk (see Figure 2).
He went on to explain that all the vehicles used were less than five years old or of a current model and in full working order, and that the cars used in the tests were selected solely on the basis of age, size and availability, and not on the basis of make or model.
Test 1: Three car test (unsprinklered) small-to-medium cars
Test 2: Three car test (sprinklered) small-to-large cars
Test 3: Same as Test 1 but with larger cars
Test 4: Four car test (unsprinklered) LPG car test
The objective of Test 4 was to determine the behaviour of fires (and explosions) involving cars fuelled by LPG.
The fire was started on the driver’s seat of Car 1 for all of the above tests.
Hertfordshire Fire and Rescue Service was on hand for Tests 1-3. Test 4 (carried out at HSL) was allowed to burn out.
The key results were:
• fire spread occurred to cars 2 and 3 (across the gap) in both unsprinklered tests (Test 1 and Test 3)
• in Test 3, where the average car size was bigger than Test 1, the time to fire spread to cars 2 and 3 was significantly reduced
• once the second car became involved in Tests 1, 3 and 4, the remaining car (two cars for Test 4) became involved within 3 minutes in all tests
• the peak heat release recorded (Figure 3) was 16MW in Test 1 shortly after car 3 ignited. The test was extinguished shortly after car 3 became involved
• three of the hollow core roof slabs used in the construction of the test rig, which were situated above and near the ignition car, had to be replaced following Test 1 due to the spalling, which occurred as a result of the fire
• in Test 2, the sprinklered test, there was no fire spread outside of the ignition car, only minor heat damage to car 2
• the peak heat release for Test 2 was 7MW, which occurred 52 minutes after ignition
• in Test 4, he pressure release valve on the LPG tank operated successfully. The tank did not rupture or noticeably impact on the fire size or development
Mr Clark went on to briefly discuss the six single car fire tests which were carried out under the calorimeter at BRE Garston, to investigate the potential for fire development with different ignition scenarios (engine fires, compartment fires and radiant panel fires, in a large and a small/medium-sized vehicle).
The final test of the experimental programme was a ‘stacker’ test. The increasing popularity of automated ‘stacker’ car parks in recent years is deemed a significant issue for car park fire safety, since the parking configurations could increase the potential for rapid fire spread. The results of a two-car vertical stacker fire showed that fire spread occurred from car 1 (the bottom car) to car 2 in about five minutes, which was much faster than in Tests 1, 3 and 4.
Research implications
Martin Shipp (technical development director at BRE) then discussed the potential implications of the research, explaining that the research should enable the fire safety community, in particular fire engineers and enforcers, to develop safer designs from a position of knowledge.
The implications of the project from the perspective of the wider fire safety community were also discussed. Howard Morgan, chairman of the British Standard Institution’s Standards Committee FSH/25 asserted the importance of this project work to the committee. FSH/25 has responsibility for keeping smoke control codes up-to-date and acts as a liaison with the European Committee for standardisation for CEN/TC191/SC1: Smoke and heat control systems and components, explained Mr Morgan. Given the outcomes from this project, the codes of practice may need rethinking, he urged, and sprinklers in car parks may need to become more prevalent.
Gordon Dewar of Westminster Council District then spoke on enforcement issues in car parks, specifically in relation to smoke control and ventilation. With the ADB guidance potentially not being updated until 2016, Mr Dewar asked whether we should concentrate on making amendments to BS 7346-7: 2006: Components for smoke and heat control systems. Code of practice on functional recommendations and calculation methods for smoke and heat control systems for covered car parks3 and look at design fire sizes and ventilation recommendations?
Ian Gough of the British Automatic Fire Sprinkler Association (BAFSA) described an additional stacker test, commissioned by BAFSA and carried out at BRE Garston using the existing stacker rig, this time with sprinklers. The fire did spread from the ignition car to the car above, but was controlled by the sprinklers. Mr Gough pointed out that the particular system that was used in the test may not be suitable for car stackers of different configurations and that specific tests may be required for each type of car stacker configuration. Also, the lack of data available on water mist systems suggests that, where such systems are proposed, relevant test data should be requested.
The implications of the project findings for the fire and rescue service were then outlined by group manager (Fire Engineering Group) of London Fire Brigade, Paul Jenkins. Sometimes, explained Mr Jenkins, where sprinklers have been specified in car parks, smoke control systems are used as a replacement.
As such, this research should enable designers to create an informed design in the future. The biggest implication from the research, said Mr Jenkins, is related to fires in stackers, where
the need and benefits of suppression systems in stackers was clearly demonstrated by the test commissioned by BAFSA.
Finally, team leader (Public Sector Property) at Zurich, Stuart Blackie, provided an insurer’s perspective. While the results of the LPG test were comforting, the issue of fire spread between cars is an issue for the insurance industry, explained Mr Blackie. Car parks should not be looked at in isolation, he said, and the adjoining buildings should be taken into consideration when determining the potential destructive effects of a fire.
Concluding comments
The conference offered delegates the chance to discuss the findings and their implications with an open discussion but, with some years before the next revision of ADB, delegates questioned how the fire community could effect change in the near future. Attendees at the conference agreed that those concerned with fire safety must identify areas where action or further research is required, and that the findings of this research should be utilised as a sound basis for future design. A summary of this project has been published3 and the full report will soon be available on the CLG website, www.communities.gov.uk
Róisín Cullinan is a consultant with the Fire Safety Engineering Group at BRE
The author wishes to acknowledge many individuals and organisations for their assistance in the project. Please note that the views expressed in this article are not necessarily those of CLG.
References
1. The Building Regulations 2000, Approved Document B (Fire Safety) Volume 2 – Buildings other than dwelling houses (2006 Edition)
2. ‘A careful attitude’, Fire Prevention and Fire Engineers Journal, July 2007, pp 54-56.
3. Shipp, M., Fraser-Mitchell, J., Chitty, R., Cullinan, R., Crowder, D., Clark, P., Fire Spread in Car Parks, Fire Safety Engineering, pp14-18, June 2009
A DVD showing a range of the full-scale tests conducted in the research is available to purchase. For more details e-mail, mallowsr@bre.co.uk