A smarter path to transit fire safety
By Kurt Schebel
July 10, 2013
Fire engineering for transportation facilities takes into account how both vehicles and the built structures surrounding them will perform in a fire.
Designers carefully select the materials and systems used for trains, subways, buses, submarines, airplanes, cars, trucks, and more according to fire performance criteria.
The same is true of transportation facilities. Issues such as material selection and the size, quantity, and power of emergency ventilation systems, sprinklers, and other systems are often determined according to the amount of fire risk present in a particular space.
The greater the likelihood of a large fire in a public space, the more fire protection equipment may be needed to manage the risk to a tolerable level.
Understanding as much as possible about the likelihood and extent of fire is therefore highly important. However, up to now, no overall methodology to predict the fire’s spread and overall hazard for compartments has existed. A common way to scientifically determine the fire size to which a design should be targeted has been to actually burn vehicles in a medium- to large-scale test.
Multiple tests are necessary to adequately explore the full variety of possible fire situations for each configuration, materials, initiating fire, and combination thereof. Therefore, clients might need to burn multiple trains.
Due to the great expense involved, the limited number of testing facilities, and the harmful environmental impacts (e.g., pollutants released into the air), many clients have not been able to carry out these tests. Consequently, they have had to make educated guesses with limited information. This can lead to vehicles and transportation facilities being under-protected, which is clearly a problem from a safety perspective, or over-protected, which can cause clients to spend significantly more money than necessary.
Arup’s fire engineering team, in collaboration with Worcester Polytechnic Institute, has been working to address this issue by developing a process to predict how transit fires will develop and spread without burning vehicles. The process: test small amounts of material to generate data that is used to predict how fires will affect both vehicles and facilities.
For example: an urban rail system needs to replace aging vehicles. The stations in the system have been designed based on the assumption of a certain fire size. The client wants to know which of two material choices for the interior train walls, glass reinforced plastic A or glass reinforced plastic B, will perform better in a fire, and whether one of them will have sufficient fire-resistant/retardant properties to make retrofits to the stations’ fire protection systems unnecessary.
To provide a more practical means of scientifically determining the amount of fire protection that will be required for a given vehicle and/or containing structure, and the relative merits of different materials from a fire safety perspective, our fire engineers worked with Worcester Polytechnic Institute to develop a new, three-level process.
Before starting the testing process, engineers examine various data sources to determine the potential sources of ignition and initial fuel sources.
They also compile data about the vehicle and/or stations: dimensions, materials, compartmentation (the size of the room: ceiling height, room dimensions, width), shape (rectilinear, spherical, etc.), overall volume, seating arrangements, and more.
As an initial screening tool, they conduct a high-level material comparison. A carefully calibrated testing system using a cone calorimeter allows them to burn small samples of a given material, then extrapolate the results to show how the material would perform in a fire at a larger scale. Sometimes this level of data is sufficient to answer the sample question: which of two materials is best for the lining of a particular train?
A second layer of information can be determined via a flame spread model. This type of model shows how a fire of a particular size and material will travel upward across walls and ceilings.
This information provides the basis for determining the amount of energy released by burning the material.
Should a client want to know more, engineers can employ computational fluid dynamics analysis. The material testing data is entered into an algorithm that runs tens of thousands of trials based on the parameters of fires possible under different conditions. This can be used to inform the material choice and help understand the vehicle’s overall design fire size, which will directly impact many other fire protection-related systems, both active and passive.
For more detailed information, see the report in the Journal of Fire Protection Engineering.
(Special thanks to my fellow research team members Brian Meacham, Nicholas Dembsey, Matthew Johann, Jeffrey Tubbs, and Jarrod Alston.)