An Explaination Of The Flashover Phenomenon Environmental Sciences Essay

The evolution of lifestyles and construction rules led to multiplication materials of synthesis on professional buildings and private houses. Today buildings are more insulated than before, to avoid losses of heat towards outside. But this insulation also contribute to the preservation of energy lost by the fire, when this one break out and so at the rise of temperatures. Moreover the complexity of thermal phenomena which cause by fire break out in compartment is frequently the origin of serious accidents.

In fact, in England the Fire and Rescue Services attended 527,000 fire and false alarm incidents between April 2009 and March 2010. In this same period 328 people died, whose 210 occurred in accidental dwelling fire (Fire Statistics Monitor, 20 August 2010). This data shows uncontrolled fires represent a major risk in everyday life.

However, firefighters are also touch by these fires like shows Blaina tragedy the 1st February 1996. Two firefighters in Wales (Blaina, Gwent) were killed by a flashover when they were searching a person reported missing in smoke-filled house (Baglin, 1996).

Despite of the fire is one of the first conquests of man; he has always fought without asking serious questions about its behaviour. But in United Kingdom and in the world, since forty years, some scientists and firefighters try to explain the flashover phenomena. However, the concept of flashover is very complicated because it is an uncertain phenomenon. That is why it is subject to controversy. In fact the specialists are agree on it general behaviour and different stages which composed it, but concerning some parameters we find bad blood, notably on the temperature range where this phenomena appear.

This report will try in the first stage to make a balance sheet and compare the different theories which are emitted on flashover phenomena. In the second stage we will try to analyze and review these theories based on experiment results.

The main objective of this report is to make a balance sheet on the different flashover theories and make a correlation between experiments data and these theories. To achieve this, some experiments were carried in a small scale compartment. The results and the analysis are provided in this report, which is divided into different part:

Ajoutter les differentes parties du rapport (en gros) et ameliorer le dernier paragraphe notamment l’objectif principal et la cinetique du rapport.

Aims and Objective

Limits

Chapter 2

Literature Review

This part provides elements necessary at the understanding of this project and some technical terms of fire science. The purpose is to offer an overview of significant literature published on this topic.

General Description of Enclosures Fires

This section introduces, in the first part the most dominant of the fire process with a general description of the combustion and in the second part, provides a qualitative description on the fire behaviour in a compartment.

Introduction of Combustion

The fire is (Fire from First Principles, Paul Stollard and John Abrahams, p.161):

A self-supporting combustion characterized by the emission of heat and effluent often accompanied by flame and/or glowing combustion.

A combustion spreading uncontrolled in time or space.

The fire triangle (Principles of Fire Behavior, James G.Quintiere, p.24) is a concept used in order to describe the fire processes. The elements which compose the fire triangle are essential to the existence of a fire. A fuel (liquids, solids or gases) combining with a combustive (e.g: oxygen of the air) and the heat source which warms the combustive. in a chemical reaction to release energy and other chemical products. The heat production allows this reaction to auto-maintain in most of the cases, even to increase in a chain reaction. Flames are the visible manifestation of this reaction between a gaseous fuel and oxygen. If we take away sufficient fuel or combustive, or reduce the energy by extinguishment for example, the fire will not survive.

C:UsersXavierDownloadsfire-triangle.jpg

Figure Fire Triangle

Heat is thermal energy in motion that travels from “hot” to a “cold” region. The “hot” is characterized by a high thermal energy with high temperature and the “cold” by a low thermal energy with low temperature.

During fire combustion there are three forms of heat transfer:

Conduction “heat transfer to another body or within a body by direct contact” (NFPA 921, 2004 ed.). Is the transfer of thermal energy by direct physical contact between regions of matter due to a temperature gradient. Heat spontaneously flows from a region of higher temperature to a region of lower temperature. In the early 1800s Fourier formulated a law relative to the heat conduction:

q = kA(T2-T1)/l

Where k is the thermal conductivity,

A is the area through which the heat is transferred,

T2 and T1 are the respective temperatures of the wall faces,

And l is the wall thickness.

Convection “heat transfer by circulation within a medium such as gas or liquid” (NFPA 921, 2004 ed.). Is the heat transfer from the fluid to a solid surface by the movement of molecules. The law is:

q” = h (T2-T1)

Where l is the distance between the temperatures corresponding to deltaT,

T2 is the air stream temperature,

T1 is the surface temperature,

h is defined as the convective heat transfer coefficient.

Radiation “the emission and propagation of energy through matter or space by means of electromagnetic disturbances that display both wave-like and particle-like behavior” (NFPA 801, 2003 ed.). Is the transfer of heat without an intervening medium between the source and the receivers. In the early 1900s Planck established a theoretical basis for radiation heat transfer with this law:

q” = σT4

Where T is the object’s temperature expressed in Kelvin (K)

Is called the Stefan-Boltzmann constant given as 5.67×10-11 kW/m²-K4

C:UsersXavierAppDataLocalMicrosoftWindowsTemporary Internet FilesContent.WordIMG_0997.jpg

Fire Behaviour in Enclosure

In this part, we present how a fire develops in enclosures, and what are the stages which composed this behavior. Next we will see which factors influence the fire development in an enclosure. (Modify in function of the summary).

According to Walton and Thomas there are five stages in an enclosure fire:

Ignition

Growth

Flashover

Fully developed fire

Decay

Ignition: Ignition can be considered as a process that produces an exothermic reaction characterized by an increase in temperature. It can occur either by piloted ignition (pilot source), or by spontaneous ignition (through accumulation of heat in the fuel).

Growth: After the ignition, the fire may grow at a slow or fast rate. This speed depends on several parameters as the type of combustion, the type of fuel, the interaction with the surroundings, and access to oxygen. However on some fires the growth period may be very long and it may die out before the next stages are reached. At the difference of some fires where the growth stage can also occur very rapidly, especially with flaming combustion, notably if there are enough fuel flammable and sufficient oxygen available for a rapid fire growth.

Flashover: In general, flashover is the transition from the growth period to the fully developed period. Flashover marks a dramatic increase in fire conditions due to the confinement of a room. This stage is complicated and this term demanding more attention in the fire science domain. It is the reason why we are going to define and explain this phenomenon in another part.

Fully developed fire: The energy release in the compartment is very important and is often limited by the availability of oxygen. This is called ventilation-controlled burning, in others terms the rate of burning is determined by the amount of fuel and the level of ventilation. In this stage unburnt gases can collect at the ceiling and when they leave the compartment through an opening that they burned, the gases catch fire. That is why this stage is marked by flames fully encompassing the room, with the flames emerging from windows and doors. The energy release by the fire submits the building’s structure at hard test.

Decay: Is the final period, all the fuel becomes consumed and temperature in the compartment declines.

Flashover Phenomena

Historic of Accidents

Case study 1: Summerland Leisure Complex Fire,

Douglas (Isle of Mann) 2 August 1973, 31 people died and 80 seriously injured.

The fire started around 7.30 pm in a small kiosk adjacent to the centre’s mini-golf course. It is believed that children ignited plastic concessions booth outside the building. This produced considerable heating of the building’s outer wall and started a fire in a small cavity in the wall (limited ventilation) which burnt for about 20minutes before the disaster.

This fire provided lots of rich unburnt gases. It is estimated that 1m thick flames emerged from the cavity into the amusement area delivering a heat flux in excess of 100 kW/m².

The fire spread rapidly through the 32m long arcade and into the other spaces within the complex.

The decorations and gloss painted asbestos-sprayed ceilings contributed to a great deal of flame emerging out of the building.

Case study 2: Bradford City Fire at Valley Parade Stadium (Vinay, 2011),

Bradford (United Kingdom) 11 Mai 1985, 56 people died, between 255 and 270 people injured.

Particularly day for Bradford’s supporters, in fact it is the last game of season, with a title already achieved. In the beginning of the match, 4,000 of 11,076 spectators in the covered grandstand and 144 policemen are present for the security.

Around 3.40 pm the first signs of fire are detected under the grandstand. At 3.43 pm the first alert is given to firefighters by police.

The fire growth is very speed because under the grandstand, ten-kilos of rubbish piled during several years; fatty packing, newspaper, offer a choice source for flames. Indeed, the investigators found a copy of newspaper “Telegraph and Argus” dating on 1968 or seventeen years before disaster.

At 3.44 pm, the spectators trying to evacuate the grandstand, then flashover trapping 56 people.

At 3.47 pm the first fire truck is presenting and at 3.49 pm two others fire truck are presenting. To this date the exactly origin to fire is not known, but the most likely will be a match or a cigarette fallen through floorboard.

After investigations, the violence of the fire will explicate by the presence of a large amount of rubbish accumulated in the grandstand. The vacuum allowed rapid fire growth by providing the hearth starting and accumulate some heat.

Above the floorboard and wooden seats offered a fuel of choice. The roof form permitted to retain gases and allow the flashover in about 4 minutes. The temperature reached was about 900 degrees.

This disaster will radically change the security policy in stadiums.

Case study 3: The King’s Cross Fire,

London 18 November 1987, 30 people and 1 firefighter died and a lot of people injured.

The Fire Brigade was called at 7.36 pm to a fire on an escalator at King’s Cross Underground Station. The fire seemed controllable and in one location. The first Firefighters arrived at 7.42 pm and at 7.45 pm the fire suddenly erupted up into the ticketing hall. It was a Flashover phenomenon. It was described as ‘Within seconds, the area was in total darkness and the conditions had become unbearable’ (Perry, 2003). The fire was likened to a blow-torch and people could not run fast enough to escape the underground. According to simulations and after small-scale studies only a heat flux of 250 kW/m2 was needed. The flashover occurs approximately 10 minutes after ignition.

Case study 4: Blaina Fire

(Wales) 1 February 1996, 2 firefighters and 1 child died.

Catherine H., 24 years wakes up in his living room at 5.48 am and discovered the beginnings of uncontrollable fire in the kitchen. She closes the door and evacuate because the house is already flooded of smoke. A neighbor called rescues at 6.03 am, saying that nobody in the house.

Another call is received by 6.09 am, claiming that children remain indoors. When the firefighters arrived, the door is open allowing the escape in large part higher black smoke, no flames visible.

The first firefighters entered the pavilion at 6.11 am, to try to rescue the occupants. At 6.13 am the first crew spring with the child unconscious. While a large plume of smoke begins to break always at the front door at 6.15 am the partner are re-engages as the residents insist on the presence of another child on the floor. 15 seconds later a big “whoosh” sound. The windows broke, and the whole pavilion suddenly ablaze. In less than 6 minutes after arriving, firefighters are faced with a flashover.

 At 6.29 am the bodies of two firefighters dead are came out on fire.

According to initial assessments, a single glazing window would have failed under the heat to 6.05 am, allowing fire to resume its behaviour until flashover. The temperature in the kitchen would reach 1000 °C and flashover occurs 27 minutes after ignition.

Team Soldats du Feu Magasine, 2008, Feu de cuisine et Flashover. Available at : [ Accessed 13 February 2011].

Balance on these Disasters and the United States of America case

In all the cases studies these disaster began by a simple small fire without “risks appearance” until Flashover. In some cases heat fluxes in excess of 100 kW/m² have been estimated, but the time to flashover was different.

Fire Disaster

Time to Flashover (min)

Heat flux

(kW/m²)

Temperature

(°C)

Summerland Leisure complex

20

100

____

Bradford City

4

____

900

King’s Cross

10

250

____

Blaina

27

____

1000

On the other side of Atlantic, in the United States, according to NFPA’s statistics (Rita F, July, 2002) recorded between 1985 and 1994, it is demonstrated that a total of 47 US firefighters lost their lives to flashover.

We can see by these feedbacks that the Flashover is a significant phenomena during the growth period of fires within confined spaces. It is responsible to a lot of death to firefighters and citizens.

Flashover Definitions

As we have seen with feedbacks, the complexity of this thermal phenomenon which cause by fire break out in compartment, is frequently the origin of serious accidents.

Properly the understanding phenomenon of flashover it’s very complicated because it’s an uncertain phenomenon and depends on several parameters. Of this fact a lot of definitions are established since more 40 years ago and several scientists study this phenomenon, like the British scientist Dr. Philip H. Thomas. He was the first to introduce a serious scientific discussion of the term flashover in the1960’s (Patrick M. Kennedy, August 2003).

“In a compartment fire there can come a stage where the total thermal radiation from the fire plume, hot gases and hot compartment boundaries causes the generation of flammable products of pyrolysis from all exposed combustible surfaces within the compartment. Given a source of ignition, this will result in the sudden and sustained transition of a growing fire to a fully developed fire…This is called “flashover”…” (Forney & McGrattan, 2001).

The initial definition of flashover by Thomas was imprecise; however it was the first time scientific thought was given to this fire progression phenomenon. From that point there has been extensive scientific research and experimentation performed to better understands flashover. At present, it can be acknowledged that there is a solid understanding of the qualitative and quantitative mechanisms that make up this phenomenon due to the extensive research and studies performed by several fire specialists.

Thereby, since 1960’s some definitions of Flashover are born:

According the National Fire Protection Association the Flashover is “a transition phase in the development of a compartment fire in which surfaces exposed to thermal radiation reach ignition temperature more or less simultaneously and fire spreads rapidly throughout the space, resulting in full room involvement or total involvement of the compartment or enclosed space.” (NFPA 921, 2004 ed.)

According to the International Standards Organization the Flashover is a “stage of fire transition to a state of total surface involvement in a fire of combustible materials within an enclosure.” (ISO 13943, 2008).

Additional definitions, by famous scientists, can be found in the literature:

Quintiere: “A dramatic event in a room fire that rapidly leads to full room involvement; an event that can occur at a smoke temperature of 500 to 600 C°.” (Kennedy, 1961).

Drysdale: “the transition from a localized fire to the general conflagration within the compartment when all fuel surfaces are burning.” (Kennedy & Kennedy, 1985).

Walton and Thomas: “Flashover is generally defined as the transition from a growing fire to a fully developed fire in which all combustible items in the compartment are involved in fire.” (Kennedy & Shanley, July 1997)

Babrauskas: “…the full involvement in flames of a room or other enclosed volume.” (Krooren, 1994)

Karlsson and Quintiere: “The transition from the fire growth period to the fully developed stage in the enclosure fire development.” (Karlsson & Quientiere, 2000)

Indeed, this transition is often assumed, by the scientists, to takes place between the growth and fully developed stages. However, neither the ISO nor NFPA definition specifies this. In addition, while the NFPA definition indicates a notion of speed in this transition, the ISO definition does not describe the speed with which the transition to total surface involvement occurs.

Flashover Theories from Experiments

In this section we are drawing up a balance sheet on experimental data that has been found by scientists and that contributes to a working definition of Flashover. Thereafter this review permits to provide theories, found by scientist, on Flashover.

Some experimental studies since 1960 have been performed, notably on full-scale compartments, for quantify the onset of Flashover in terms of measurable physical properties and qualitative aspect. Thus, some calculations for predicting the temperatures, or predicting the Heat Release Rate (HRR) necessary for Flashover occurs, for example, were born. But this work is not easy; because Flashover has been reported as a discrete event and it appearance depends on several parameters like thermal influences, ventilation conditions, or the size of the compartment and also the fuel properties. Indeed, a considerable body of fire test data exists.

Waterman (Drysdale, 1998) studied the flashover in 1968, in a compartment room 3.64 x 3.64 x 2.43 high. He concluded that the heat flux of about 20 [kW/m²] at floor level was required for flashover occurs.

Since this work other experiments have tended to define the minimum conditions at “onset of Flashover” in terms of temperatures and heat fluxes; with the used of flashover’s indicator.

Some of researchers are tried to provide methods in correlation to fire experiments for estimate the minimum release rate necessary to produce Flashover.

In fact Hägglund et al. (Peacock & al, 1999) report that during experiments the Flashover is defined by flames exiting the doorway when the gas temperature below the ceiling reached 600 °C. But concerning the heat flux there is no data recorded.

Hägglund’s Expression (Hägglund, 1980): 3

Where hk [kW/(m².K)] is the heat transfer coefficient;

At [m²] is the wall total surface area of the compartment;

H [m] is the height of the opening;

Aw [m²] is the area of the opening;

And is expressed in [kW]

Babrauskas (Babrauskas, 1977) made a series of full-scale test on mattress fire based on Waterman’s criterion and the Flashover was achieved to 600 °C. Moreover the time and minimum flux to ignition of some commons materials (as newspaper and box cardboard) was tested and it was found respectively 48 [kW/m²] and 43 [kW/m²]. For Flashover occurs it was found a minimum of 20 [kW/m²].

Babrauskas’ Expression (Babrauskas, 1980):

Thereafter Babrauskas et al. (Babrauskas, 2003) are analyzed data from several ISO 9705 room tests. They found that the heat release rates during Flashover are 1975 1060 [kW] and a HRR of 1 [MW] represent the minimum level required for Flashover in an ISO 9705 room.

Fang (Fang, 1975) conducted experiments in a full-scale compartment and he reported an average upper temperature ranging 450 to 650 °C and a heat flux ranging 17 to 33 [kW/m²]. These data provide a sufficient radiation transfer for ignited a crumpled newspaper on the floor level (540 40 °C exactly).

Budnick and Klein (Budnick & Klein, 1979) performed several tests in mobile homes. During tests in the living room, they reported the ignition of crumpled newsprint indicator when the upper room temperatures achieved between 673 and 771 °C (below the ceiling in the centre of the bedroom) and observed Flashover between 634 and 734 °C (Budnick, Klein, & O’Laughlin, 1978), with a minimum of heat flux at 15 [kW/m²] (Budnick & Klein, 1979).

Lee and Breese (Lee & Breese, 1979) performed two experiments in submarine compartments, one in full-scale and the second in a ¼ scale. They reported the ignition of newsprint indicator respectively to 650 and 550 °C, with an average of heat fluxes at floor level of 17-30 [kW/m²] for the full-scale experiment.

Fang and Breese (Fang & Breese, 1980) conducted 16 full-scale tests of residential basement rooms. They observed the ignition of the paper flashover indicators with an average upper room gas temperature of 706 92 °C and a heat flux at 20 [kW/m²] measured at the centre of the floor level.

Quintiere and McCaffrey (Quintiere & McCaffrey, 1980) studied the burning of plastic and wood cribs in a room, in order to assess the fire risk of cellular plastic materials as compared to wood for use in furniture. For that, they divided their experiments into two categories: lower-temperature fires (with the gas temperature below the ceiling less than 450 °C) and high-temperature fires (with the gas temperature below the ceiling greater than 600 °C) which they observed the ignition of cellulose filter paper in 5 cases out 16. Moreover the ignition of this filter paper was achieved at a minimum of 17.7 [kW/m²] and a maximum of 25 [kW/m²].

McCaffrey’s Expression (McCaffrey, Quientiere, & Harkleroad, 1981):

1/2

1/2

Thomas (Thomas, 1981) studied the influence of wall-lining materials and thermal feedback to the burning items within a simple room compartment. Moreover he developed a semi-empirical calculation for to determine the heat release rate required to cause Flashover. This model predicts a temperature rise of 520 °C and a heat flux of 22 [kW/m²] necessary to cause Flashover. Parker and Lee (Parker & Lee, 1973) have proposed using a level of 20 [kW/m²] as the heat flux at floor level to ignite cellulosic fuels in the lower part.

Thomas’ Expression (Thomas, 1981):

Concerning the time to Flashover, in 1988 Anderson (Karlsson & Quientiere, 2000) conducted experiments in small-scale room and he compared the time to Flashover in function on the lining material walls in two cases:

The former case lining material on both walls, with a combustible ceiling, the time to Flashover is 4 minutes.

The latter case lining material on the walls only, with non-combustible ceiling, the time to Flashover is 12 minutes.

Although the time is not the same, Anderson observed that Flashover occurs when the energy release rate is 100 [kW].