Following their previous research on car park fires, BRE Global was commissioned to conduct tests into the effectiveness of sprinklers in car parks using stacking machines. David Crowder sets out the main findings.

[Download the full report here]

‘Stackers’ are mechanical devices which hold cars within a car park. Often controlled by a computer, their primary function is to park more cars in a given area than a traditional car park. This is achieved by stacking rows of cars, often with little vertical separation. There is an increasing number of stacker car park systems being installed, particularly in cities where space in scarce. Although the design of individual stacker systems varies widely, they all, by definition, park cars more densely than traditional car parks.

Following a three year Communities and Local Government (CLG) sponsored programme on car parks, concerns were raised regarding the potential damage which may be caused by fires in stacker car parks. The British Automatic Fire Sprinkler Association (BAFSA) believed that sprinkler systems could control the development of such fires. Therefore, to demonstrate the effect a sprinkler system would have on a fire in a stacker car park, BAFSA commissioned BRE Global to carry out a stacker test fire, replicating that used in the CLG programme of work, with the addition of a sprinkler system.

The test rig
To replicate the CLG funded test as closely as possible, the same steel rig was used. The rig supports one car above another, which is typical of a stacker configuration. Cars were sourced and supplied by BAFSA. Each car had 20 litres of fuel in its fuel tank.

Sprinkler system design
The design, installation and maintenance of sprinklers in car parks is covered by BS EN 12845, which classifies traditional car parks as ‘Ordinary Hazard 2’ occupancy. However, due to the wide variety of applications for car stacker systems, guidance is not available. For the purposes of the test, the sprinkler system was designed and supplied by BAFSA and was designed to be as consistent as possible with an ‘Ordinary Hazard’ system. Allowances were made for the vertical distribution of the fuel load but sprinkler heads were not installed pointing towards any particular part of the test rig or vehicles. The system was designed to provide one sprinkler head in the vicinity of each corner of each car.

Elevation of sprinkler layout showing pipe sizes, sprinkler locations and orientations

Plan of sprinkler layout showing sprinkler head positions, pipe sizes and orientation of rig

Prior to the fire test, some preliminary testing of the discharge from the sprinkler system was undertaken. Based on the findings, a ‘conventional’ sprinkler head model was used, with a combination of pendant and upright sprinkler heads. The hydraulic equipment requirement was set to provide 1 bar of pressure at the most hydraulically remote sprinkler location, with all eight heads operating.

During the cold discharge testing it was noticed that the high level sprinklers were wetting the lower level sprinklers which would impact on the operation of the lower level sprinklers. Therefore, a baffle plate was installed above each of the lower level heads.

Findings
The fire was ignited on the driver’s seat of the lower car, replicating the unsprinklered stacker test previously conducted. In the previous CLG tests, a number of windows had been open to simulate an arson attack and thereby ventilate the fire. In the current tests only one window was open, as the level of shielding afforded by the windows was considered to present the more challenging scenario for the sprinkler system to deal with.

The first sprinkler head operated approximately 13 minutes after ignition by which time flaming was visible at the underside of the upper car. Over the next ten minutes, two more sprinkler heads activated. Initially, the sprinklers were allowed to operate for approximately one hour after the activation of the first head. At this point the fire was contained within the vehicles and was no longer representing a challenge to the sprinkler system. The sprinklers were turned off at the water pump to allow the fire to re-establish itself. Finally, the system was reactivated to assess the ability of the sprinklers to suppress the re-established fire. The sprinklers quickly controlled the fire and the test was terminated after a further ten minutes.

Throughout the test, the temperatures in the upper car and external temperatures remained lower than in the CLG test. After the test, the majority of the combustible material in the lower car was found to have been consumed or severely affected by the fire, and the engine bay of the upper car had been significantly damaged. However, the interior of the upper car had not been affected by the fire.

Rig side temperature comparison

Upper car temperature comparison

Conclusions
Due to the rear driver’s side window being closed (this window was open during the CLG test ), the fire in the lower car developed more slowly in this test than in the previous CLG test. However, this also reduced access of the sprinklers heads to the fire, creating a more severe condition for the sprinklers to control once the fire was established.

Once activated, the sprinkler system rapidly controlled and contained the fire. The sprinklers were turned off after one hour, which allowed the fire to grow. This shows that the sprinklers had controlled rather than extinguished the fire. Following reactivation of the sprinkler system, the fire was again rapidly controlled.

Although there was some fire spread from the lower, ignition car to the upper car, it was evident that the sprinklers had controlled the fire. When compared with the CLG test, the sprinklered stacker showed considerably reduced overall fire size and lower temperatures. While the fire spread to the upper vehicle, the upper vehicle did not become fully involved and the risk of fire spread beyond the test geometry to nearby cars was significantly reduced by the presence of a sprinkler system.

The test has demonstrated the potential benefits of installing a sprinkler system into a car stacker, providing the system is designed for the specific risk and with the installation based on appropriate standards and guidance. 

David Crowder is a fire safety consultant at BRE Global. Read the full BRE report here