Click for next page ( 176


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 175
Smoke Toxicity* David Purser SlJMMARY ** When aircraft cabin occupants are exposed to fire effluent the first hazard encountered is usually smoke, containing particulates and toxic gases that cause immediate visual obscuration and painful irritation of the eyes and respiratory tract. This may be followed by incapacitation due to pain or asphyxia if exposure continues. In smouldering or small, confined, in-flight fires, where the yields of organic irritants and acid gases are likely to be high and exposure times long, the distressing effects of irritants, lung inflammation, and asphyxia induced by carbon monoxide are likely to be the main hazards. For post-crash fires, which tend to develop rapidly to flashover, the time available for escape is often limited to a few minutes before conditions become lethal due to the effects of toxic smoke and heat, so that survival depends upon a rapid egress. Visual obscuration and smoke irritancy are important during the early stages in that they may reduce the speed and efficiency of escape. People have been shown to be reluctant to enter smoke-Iogged areas if these are between them and an exit, and movement speeds are greatly reduced at optical densities above 0.5, OD/m and even more when the smoke is irritant. Once cabin lining and seating materials become heavily involved in the area opposite a cabin breach, then the concentrations of toxic gases, especially carbon monoxide (CO) and hydrogen cyanide (HCN), can increase rapidly further clown the cabin, causing rapid incapacitation of any remaining cabin occupants. This is followed or accompanied by extreme heat, so that deaths result from asphyxia and/or heat shock. For in-flight fires, it is recommended that consideration should be given to reducing the hazard from irritants. For post-crash fires, measures aimed at delaying the involvement of cabin contents (such as spray mist systems) should be considered. INTRODUCTION The majority of fat and nonfatal casualties from fires result from exposure to toxic smoke, but there can be considerable differences between different types of fires in terms of the smoke composition and the ways in which it affects people in different fire scenarios. Fires in aircraft may be classified into two very different major categories, the in-flight fire and the post- crash fire. Aircraft occupants in flight may be many hours from possible landing and disembarkation, so that any fire that grows rapidly and penetrates the cabin space is likely to be fatal, due to asphyxiation of the occupants and loss of the aircraft. Fortunately such occurrences are rare, but of equal concern must be the small nonflaming or flaming Ore, particularly in a *Copyright, British Crown **Building Research Establishment, Fire Research Station, Garston, Watford, U.K. i75

OCR for page 175
176 Imp roved Fire- aru] Smoke-Resistant Mate rz als confined space. This may result in contamination of the cabin atmosphere with a low concentration of toxic smoke that may have to be endured for a number of hours. In such situations the major concerns must be initially the psychological and physiological effects on passengers and crew of exposure to an irritant and optically obscure smoke, and then the asphyxiation hazards presented by lung inflammation and graclual intoxication by asphyxiant gases such as carbon monoxide, both of which may lead to long-term respiratory tract and neural ~ c amage in survivors. Post-crash fire scenarios are very different because they often involve large, rapidly growing fires resulting from fuel involvement, which may enter the cabin through breaches and result in rapid cabin flashover. Also, because the aircraft is on the ground, rapid passenger egress or rescue may be possible. In this situation the key factors are the time within which the passengers can disembark compared with the rate of fire growth and particularly the time within which toxic smoke and heat impair or prevent egress. This paper examines fire scenarios in terms of toxic smoke and heat profiles and presents methods for estimating behavioural impairment and incapacitation. PRACTICAL METHODS FOR THE ASSESSMENT OF FACTORS DETERMINING HAZARD TO LIFE IN IVIES From the point of view of parameters directly related to the fire process, the development of behavioral impairment and hazard to life depends upon two major parameters: I. The time-concentration (or intensity) curves for the major toxic products, optically dense smoke, and heat in the fire at the breathing zone of the occupants, which in turn depend upon: the fire-growth curve in terms of the mass-Ioss rate of the fuel (kg/s) and the volume into which it is dispersed (kg/m3~; and the yield of toxic-products smoke and heat in the fire (e.g., kgCO/kg of material burned). 2. The toxic potency of the products (the exposure concentration [kg/m3], or exposure dose [kg min/m3 or ppm mind required to cause toxic effects (and the equivalent effects of heat and smoke obscuration), the assessment of which requires consideration of three aspects: exposure concentrations or doses likely to impair or reduce the efficiency of egress due to psychological and/or physiological effects; exposure concentrations or doses likely to produce incapacitation or prevent egress due to psychological and/or physiological effects; and lethal exposure concentrations or doses.

OCR for page 175
David Purser 177 There are essentially two practical methods assessing these factors: I. from large-scale fire tests including measurements of the concentration-time profiles of the major toxic gases, heat, and smoke optical density, and from existing knowledge of the effects of exposure to these agents; and 2. from a battery of small-scale tests and mathematical models, or simple large-scale tests, where the essential elements are: the toxic potency data for the materials (lethal mass-Ioss exposure dose [am min/m31) obtained from small-scale tests; and the mass-Ioss/concentration curve for the fire. From the point of view of understanding effects on people in fires, full-scale fire tests are the most valuable, since they enable the concentration-time profiles of the heat and toxic smoke to be measured directly at different levels and positions within the fire. In practice, however, elements of both methods are useful in making an overall assessment. The second method is used mainly for evaluating materials with regard to the lethal toxic potency of potential fire atmospheres. In order to enable the effects of exposure of people to fire hazards to be calculated, a series of algorithms has been developed for calculating time to incapacitation or death, which is published in The SFPE Handbook of Fire Protection Engineering (Purser, 1988), in a NATO-AGARD (Advisory Group for Aerospace Research and Development) paper on aircraft fires (Purser, 1989), and in BST (British Standards Institution) and TSO (International Standards Organization) technical reports (BS1:, 1989; TSO, 1994~. Physiological Hazards In Fires The physiological effects of exposure to toxic smoke and heat in fires result in varying degrees of incapacitation, which may also lead to death or permanent injury (Purser, 19881. Incapacitating effects include: a. Impaired vision resulting from the optical opacity of smoke and from the painful effects of irritant smoke products and heat on the eyes. b. Respiratory tract pain and breathing difficulties or even respiratory tract injury resulting from the inhalation of irritant smoke, which may be very hot. In extreme cases this can lead to collapse within a few minutes from asphyxia due to laryngeal spasm and/or bronchoconstriction. Lung inflammation may also occur, usually after some hours, which can also lead to varying degrees of respiratory distress. Narcosis from the inhalation of toxic gases, resulting in confusion and loss of consciousness. d. Pain to exposed skin and the upper respiratory tract followed by burns, or hyperthermia, due to He effects of heat, preventing escape and leading to collapse. All of these effects can lead to permanent injury, and all except (a) can be fatal if the degree of exposure is sufficient.

OCR for page 175
178 Improved Fire- aru] Smoke-Resistant Materials With regard to hazard assessment the major considerations are: I. the time when partially incapacitating effects that might delay escape are likely to occur; 2. the time when incapacitating effects that might prevent escape are likely to occur, compared with the time required for escape; and 3. whether exposure is likely to result in permanent injury or death. Up to a certain level of severity, the hazards listed in (a)-~) above cause a partial incapacitation by reducing the efficiency and speed of escape. These effects lie on a continuum from tilde or no effect at low levels to relatively severe incapacitation at high levels, with a variable response from different individuals. ye is important to make some estimate of effects that are likely to delay escape; these effects may decrease the number of occupants able to escape during the short time before conditions become so bad that escape is no longer possible. Most important in this context is exposure to optically dense and irritant smoke, which tends to be the first hazard confronting fire victims. For more severe exposures, a point may be reached where incapacitation will prevent escape. For some forms of incapacitation, such as the point where narcosis leads to a rapid change from near normality through a brief period of intoxication, to loss of consciousness, this point is relatively easy to define. For other effects an endpoint is less easily defined; examples are the point where smoke becomes so irritant that pain and breathing difficulties lead to the cessation of effective escape attempts, or the point where pain and burns prevent movement. Nevertheless it is considered important to attempt some estimate of the point where conditions become so severe in terms of these hazards that effective escape attempts are likely to cease, and where occupants are likely to suffer severe incapacitation or injuries. In addition to the physiological effects of exposure to toxic smoke, there are psychological factors to be considered in relation to smoke exposure. This aspect is largely beyond the scope of this paper but some aspects are mentioned in the next section. Evaluation of the Effects of Optically Dense, Irritant Smoke on the Eyes and Respiratory Tract Optically dense smoke affects way-f~nding ability and the speed of movement of occupants, and a smoke barrier may be perceived as being impenetrable. These effects depend upon the concentration topical density) of the smoke and its irritancy to the eyes and respiratory tract. In experiments where people were asked to walk down a smoke-Iogged corridor, fin (1976) found that for nonirritant smoke, walking speed decreased with smoke density, and that at an optical density of 0.5 OD/m (extinction coefficient 1 . 15) walking speed decreased from approximately I.2 m/s (no smoke) to 0.3 m/s. Under these conditions people behaved as if they , . . ~ ~ ~ ~ ~ ~ .~ ~ .a a~ ~~ ~ ~ ~ . . . were in total darkness, teelmg their way along the walls. when people were exposed to lrrltaIlt smoke, made by heating wood chippings, movement speed was reduced to that in darkness at a much lower optical density (0.2 OD/m, extinction coefficient 0.5), and the experience was found to be more distressing.

OCR for page 175
David Purser 179 In addition to these effects upon movement speed, there is the problem of deciding whether people will move at all. In a number of studies of fires in buildings, a proportion of people (approximately 30 percent) were found to {urn back rather than continue through smoke- logged areas (Wood, 1972; Bryan, 19771. The average density at which people turned back was at a "visibility" distance of 3 m (0.33 OD/m), and women were more likely to turn back than men. A difficulty with this kind of statistic is that in many fires in buildings there is a choice between passing through smoke to an exit or turning back to take refuge in a place of relative safety, such as a closed room. In some situations people have moved through very dense smoke when the fire was behind them, while in other cases people have failed to move at all. In a post- crash (or grounded) aircraft fire it seems likely that the majority of people will attempt to move through even dense smoke towards an exit, especially if layering permits them to crouch down to levels where the smoke density is lower and if low-level lighting is used to improve viability. However, it is likely that some people will not move through dense smoke. Smoke irritants consist of inorganic acid gases (such as hydrogen chlorides and organic compounds, particularly low molecular weight aldehydes (formaldehyde and acrolein). More than 20 irritant substances have been detected in smoke, and it is considered that others remain to be identified (Purser, 1988~. The first effect of exposure to smoke irritants is sensory irritation, which consists of painful stimulation of the eyes, nose, throat, and lungs. Sensory irritation depends upon the immediate concentration of irritants to which the subject is exposed rather than a dose acquired over a period of time, the effects {yin" on a con cinuum from mild eye irritation to severe eye and respiratory tract pain. In evaluating this aspect of irritancy the aim is to predict what concentration of mixed irritant products is likely to cause such pain and difficulty in breathing that escape attempts would be stowed or rendered less efficient, and what concentration is likely to seriously disrupt or prevent escape (a degree of incapacitation approximately equivalent to that at the point of collapse resulting from exposure to narcotic gases). For example, with regard to hydrogen chloride it is considered that concentrations of approximately 100-500 ppm would be painfully irritant, and that the effects might slow escape but probably not prevent it. However, at approximately 1,000 ppm and above it is suggested that the effects might be so severe as to prevent escape (Purser, 1988, 1989~. In the absence of detailed information on irritant mixtures it is assumed that all irritants would be additive in their effects, since they are all capable of causing damage to lung tissue. In large-scale fire tests it is possible to measure inorganic irritants directly, but it is difficult to assess the degree of irritancy from organic products, which form a very important component. In general the effects of organic irritants depend on the concentration of partially oxidised organic species in the smoke. For example, smokes from smouldering wood or polyolefins have a high organic content and are highly irritant; they are characterized by low CO2/CO ratios and high smoke yields. Under well-ventilated flaming conditions, by contrast, the organic content of the effluents is low and irntancy is low. In general, it is predicted that smoke from a mixed fuel source with an optical den sity/metre of 0.5 would be strongly irritant to the eyes and respiratory tract Olin, 1976; Purser, 1988~. However, for a given smoke density there are differences between different types of fires, since some people report that smoke from some fires, while dense optically, is relatively tow in irntancy, while that from other fires is extremely irritant. It is difficult to quantify these irritant effects because the database on the effects of individual irritants or irritant mixtures on escape behaviour in humans is poor and because the

OCR for page 175
180 Improved Fire- arm Smoke-Resistant Materials effects lie on a continuum of seventy where there are no precise endpoints. Assessment has to be based upon a small number of human expenmental exposures (usually at relatively low concentrations), accidental exposures, and He results of bioassay studies. The most useful bioassay method for sensory incitation has been the mouse RDso test, in which the concentration causing a 50 percent decrease in respiration rate following a short exposure is measured. A reasonably good relationship has been found between the mouse RDso concentrations for a range of irritant vapours and the concentrations reported as being painfully irritant to humans (Alane, 19811. The test has been applied to a wide range of irritant substances, many of which occur in fire atmospheres, and also to mixed combustion product atmospheres. In order to assess the combined effects of irritants, a concept of fractional irritant concentration (FIC) has been developed (Purser, 1993), whereby the concentration of each irritant present is expressed as a fraction of the concentration considered to be severely irritant to humans. The irritant concentration for each gas has been set on the basis of data from both human and animal studies. The FICs for each irritant are then summed to give a total FIC. If the total FIC reaches unity, then it is predicted that the smoke atmosphere would be highly irritant, sufficient to slow down escape attempts. If the total greatly exceeds unity then it is likely that escape would be prevented, and it is possible that collapse might occur due to static hypoxia from bronchoconstriction or laryngeal spasm. On the basis of available data, current estimates of the concentrations of each gas likely to be highly irritant are as follows: Toxic gas HC1 HBr HF so2 NO2 Total organics Concentration 200 ppm 200 ppm 120 ppm 30 ppm 80 ppm (5 min) 25 ppm (30 min) 0.5 OD/m On the basis of the assumption that all irritants capable of damaging lung tissue are additive in their effects, the overall irritant concentration FIC~ is then given by F}C,rr = FICHE + FICHE + FICHE + FICS02 + FICNO2 + FICorg Another way of expressing the sensory irritancy of fire effluent is in terms of the RDso of the material decomposed (expressed as the mass-loss concentration), rather than in terms of individual irritant products. Table ~ shows the mouse RDso for a number of materials, some of which are used in aircraft, when decomposed under the thermal decomposition conditions indicated, using the FRS tube furnace method (Purser et al., 19941. The majority of experiments were conducted under nonflaming oxidative decomposition conditions, but a small number of experiments were conducted under flaming decomposition conditions. The results show that the

OCR for page 175
David Purser TABLE 1 Mass-Loss Concentration of Thermal Decomposition Products Predicted to be Painfully Irritant (mouse RDSo g/m3) 95% Temperature RDSo Confidence Material (o C) NF/Fa g/m3 Limits General Materials Acrylonitrile butadiene styrene 500NF 0.11 0.07-0.17 Acrylonitrilebutadiene styrene 600F ~1 Low density polyethylene 500NF 0.05 0.03-0.07 Nylon-6 480NF 0.47 0.29-1.10 Nylon-6 600F ~ 20 Polyvinylchloride (PVC) (rigid) 400NF 0.17 0.12-0.25 PVC (plasticised) 380NF 0.19 0.09-0.28 PVC (plasticised) 600NF 0.17 0.12-0.22 PVC (plasticised) 650F ~2.6 Thermoplastic polyurethane 425NF 0.20 0.14-0.96 Thermoplastic polyurethane 600F ~ 3 Cable Materials PVC insulation (plasticised) 550NF 0.56 0.39-1.00 PVC jacket (plasticised) 550NF 0.34 0.27-0.47 Cross-linked polyethylene (insulation) 550NF 0.12 0.09-0.17 Cross-linked polyethylene (jacket) 550NF 0.32 0.20-0.32 Aircraft Materials Phenolic fibreglass 600NF ~ 9.1 PVC decorative laminate 600NF 0.10 Polycarbonate 600NF 0.25 Phenolic oil fibreglass insulation 600NF 0.05 Aluminised PVF/paper covering 600NF 0.37 Redux adhesive 600NF 0.10 0.06-0.16 Silicone rubber 600NF 0.06 0.01-0.29 Joining compound JCSV 600NF 0.18 0.07-0.32 Viton sealant 600NF 0.21 0.15-0.27 Berger elastomer 600NF 1.38 1.12-1.80 aNF = nonflaming, F = flaming. majority of materials have RDso values lying between 0.05 and 0.5 g/m3 under nonflaming oxidative decomposition conditions. This means that if the products of decomposition of between 0.05 and 0.5 grams of material are clispersed into each cubic metre of air, then the resultant atmosphere is predicted to be painfully irntant to the eyes and respiratory tract. However, under flaming decomposition conditions the smoke irritancy decreases by a factor of 10 or more. The other important effect of irntants is that a proportion of those inhaled penetrate into the deep

OCR for page 175
182 Improved Fire- aru] Smoke-Resistant Materials lung. If a sufficient dose is inhaled over a period of time a lung inflammatory response can occur, usually some hours after exposure. This may cause respiratory failure and death, or permanent lung damage in survivors. The 30-minute exposure concentrations likely to be lethal used for each irritant gas (based upon rat LCso data) are as follows: Toxic gas Concentration HC1 HBr HE so2 NO2 3,800 ppm 3,800 ppm 2,900 ppm 400 ppm 170 ppm (30 min), 375 ppm (5 min) Total organics 3 OD/m The effects depend upon the exposure dose, which can be quantified approximately in terms of the product of concentration (c) and exposure lime (I) to give the ct product exposure dose (ppm/min). During a fire, when the concentrations of the toxic products vary with time, it is possible to predict when an incapacitating or lethal dose has been received by using the fractional effective dose (FED) method. For this method the ct product doses for small periods of time during the fire are expressed as a fraction of the dose causing a toxic effect, and these FEDs are summed until the fraction reaches unity, when the toxic effect is predicted. The fraction of a lethal dose (FED) for each irritant is calculated as the ct product exposure dose during a period in the fire (e.g., in ppm/min) expressed as a fraction of the lethal exposure dose. The lethal effects of the different irritants are assumed to be additive on the same basis as the irritant effects, so that the total FEDirr for each time period is given by FED,rr = FEDHC~ + FEDHBr + FEDHP + FEDS02 + FEDN02 + FEDrg (2) Calculation of Time to Incapacitation Due to Effects of Narcotic Gases Narcotic gases (carbon monoxide, hydrogen cyanide, carbon dioxide, and reduced oxygen) affect the nervous and cardiovascular systems, causing confusion followed by loss of consciousness, followed ultimately by death from asphyxiation (Purser, 19881. As narcotic gases are inhaled during a fire, an increasing dose builds up in the body. There is lithe effect initially, but when a critical threshold dose level is reached severe effects occur suddenly. These consist of a brief period of intoxication (similar to severe alcohol intoxication), followed by a collapse into unconsciousness (Purser, 19881. Where several narcotic gases are present the effects have been found to be additive, with carbon dioxide mainly causing an increase in the rate of uptake of the other narcotic gases. It has also been shown that the effects of irritant gases such as hydrogen chloride (MCI) are additive with those of carbon monoxide (Hartzell et al., 1985~.

OCR for page 175
David Purser ~3 Based upon these findings, an FED equation to predict time to incapacitation (loss of consciousness from the effects of narcotic gases for humans) has been developed as follows: FEDIN-(FEDS + FEDS + FED,rr) x VCO2 + FEDS or FIN, (3) where FEDS fraction of an incapacitating dose of all narcotic gases; FEDS fraction of an incapacitating dose of CO; FEDS = fraction of an incapacitating dose of HCN; FEDS-fraction of an irritant dose contributing to hypoxia; VCO2-multiplication factor for CO2-induced hyperventilation; FEDS = fraction of an incapacitating dose of low oxygen hypoxia; and FED~Co2 - fraction of an incapacitating dose of CO2. Each individual term in the FED equation is itself the result of the following equations, which give the FED for incapacitation for each gas and the multiplication factor for CO2, where t is the exposure time at a particular concentration in minutes. The FED s acquired over each period of time during the fire are summed until the total FEDIN reaches unity, at which point incapacitation floss of consciousness) is predicted. Death is predicted at approximately 2-3 times the incapacitating dose. FEDS (~.2925 |04 ppm COME) t/30 FEDS-t/Lexp (5.396 0.023 ppm HCN)] FEDS - calculated in Equation 2 above, VCO2 - exp (0.1903 %CO2 + 2.0004~/7.1 , FEDS - t/exp [~.13 - 0.54~20.9 - %o21], and FED~Co2 = t/exp (6.1623 - 0.5189 %CO2) . (4) (5) (6) (7) Table 2 shows a simplified lookup table of FED s for incapacitation for each gas for use in approximate calculations.

OCR for page 175
184 ~5 5~100 100 125 125-150 150~200 200+ Improved Fire- and Smoke-Resistant Materials 0.05 2-3 0.10 3-4 0.15 4-5 0.50 5-6 0.100 6-7 7-8 8-10 o 0.02 12-11 0.05 11-10 0.08 10-9 0.15 9-8 0.20 8-7 0.40 7-6 0.70 TABLE 2 Simplified Lookup Table for Solutions to Individual Toxic Gas FED Equations FEDS = CO ppm/25~000 ppm HCN FEDS Scot VCO2 %O2 FED'02 %CO2 FED'co2 ~2 1.0 21-13 1.5 13-12 2.0 2.5 3.0 3.5 4.5 4.8 0-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 0.01 0.01 0.02 0.03 0.05 0.08 0.13 0.22 0.38 Calculation of Time to Incapacitation Due to the Effects of Convected and Radiant Heat Another dose-related hazard is exposure to convected heat, where skin pain followed by burns or hyperthermia occurs depending upon the air temperature and exposure time. A FED equation for heat has been developed similar to that for a narcotic gas, as follows: FEDIh ~ I/exp(5. 1849 - 0.0273 temp C) . (9) For radiant heat, skin pain and burns occur rapidly at intensities above 0.25 W/cm2 (Purser, 1988~. Details of the calculation methods are given in Purser, 1988 and 1989. APPLICATION OF THE INCAPACITATION MODEL TO DIFFERENT FIRE SCENARIOS In terms of the basic scenarios, the hazards in terms of the yields of heat, smoke, and toxic products, and the effects on escape behaviour, fires can be considered in several different categories. The main categories are 1. nonflaming or smoulder~ng fires where the victim is in the compartment of fire origin or in a remote location; 2. early-flaming, well-ventilated fires where the victim is in the compartment of fire origin or in a remote location; small, restricted-ventilation fires where the victim is in the compartment of fire - ong1n; and

OCR for page 175
David Purser ~5 4. ventilation-controlled post-flashover fires with low and high ventilation, where the victim is remote from the fire, either inside a very large compartment or outside the fire compartment. In the context of aircraft fires, the first three categories are of concern mainly in the in- flight situation, where cabin occupants and crew may be exposed to fire effluent for several hours. Although fires in categories 2 and 3 may also present a hazard on the ground, it is to be expected that the aircraft would be cleared before such fires become a serious threat. The main hazard on the ground is therefore from fires, usually involving fuel, that rapidly grow very large and cause rapid flashover inside the cabin. Nonflam~ng or Smoulder~ng Fires Nonflaming or smouldering decomposition results in high yields of organic irritants from materials, and high yields of inorganic irritants from materials containing the appropriate elements. These smoke irritants are predicted to cause distress upon initial exposure, due to the painful effects upon the eyes and lungs. As Table ~ shows, very small amounts of material can produce severely irritant atmospheres if decomposed under these conditions. Cases have occurred where the decomposition of a few tens of grams of cable insulation material have rendered large buildings uninhabitable for a number of hours. Prolonged exposure to these irritants over periods of an hour or more may lead to lung inflammation, while carbon monoxide intoxication may also occur. A number of deaths each year result from fires in buildings that are considered to have undergone a prolonged period of smoulder~ng. This may occur in items of furniture, or in structural items such as flooring, or in concealed cavities, sometimes initially involving shouldering cables. A recent case involving a number of deaths resulted from smouldering floorboards and joists. Nonflaming decomposition is slow, so that a relatively long time (approximately an hour or more) is required for the development of hazardous conditions. However, although small masses of material may be decomposed, the yields of carbon monoxide can be high, and these conditions generally provide the highest yields of irritant organic products. The major hazard is to a sleeping or otherwise incapacitated occupant of a closet! room, or to aircraft occupants in flight who may be overcome by carbon monoxide and lung- damaging irritant smoke. A good example of such a situation is presented by a series of tests carried out at the National Institute of Standards and Technology (Braun et al., iL987; Purser, 1990) where two types of armchair made from a standard and a fire-retarded (FR) polyurethane foam with cotton covers (combustible mass 5.7 kg) (see Tables 3 and 4) were burned in a simulated small apartment (volume 101 m3) consisting of a burn room (~.8 m3) connected via a corridor 12 m long to a target room (volume 12.08 my. The chairs were tested by flaming ignition of the seat back, and also by smouldering caused by one or two cigarettes placed in the seat angle for approximately ~ hour, followed either by spontaneous flaming or ignition from a flaming source. Under smouldering conditions, approximately ~ kg of foam was decomposed in just over ~ hour.

OCR for page 175
186 Improved Fire- aM Smoke-Resistant Materials TABLE 3 Concentrations of Toxic Gases and FEDs in Burn Room for Smouldering Followed by Flaming Ignition of Standard Foam Armchairs versus Time (in minutes) 0-13 13-27 27-40 40-53 53-67 67-75 75-76 Gas Concentrations CO (ppm) 180 300 360 700 700 1,000 10,000 HCN (ppm) 0 0 0 0 0 0 1,320 CO2 (%) 0.11 0.16 0.18 0.30 0.30 0.40 15.0 O2 (%) 21 21 21 21 21 21 3 FEDs for Incapacitation FEDCo 0.006 0.010 0.012 0.024 0.24 0.024 Immediately fatal FEDHCN O 0 0 0 0 0 VCO2 1.019 1.032 1.037 1.069 1.069 1.096 FEDo O O O O O O FED/min 0.006 0.010 0.013 0.026 0.026 0.039 FED 0.078 0.218 0.387 0.725 1.089a 1.401 aBy 71 minutes the mass-loss exposure dose of irritants was 600 g me min. which is likely to cause fatal lung damage. TABLE 4 Concentrations of Toxic Gases and FEDs in Target Room for Smouldering Followed by Flaming Ignition of Standard Foam Armchairs versus Time (in minutes) 0-13 13-27 27-40 40-53 53-67 67-75 75-76 76-77 Gas Concentrations COppm 0 0100270 5508002,7002,000 HCN ppm 0 000 00125120 CO2 ~0.04 0.04 0.08 0.15 0.20 0.30 9.00 8.50 O2 % 21 21 21 21 21 21 13 14 Doses of Incapacitation FEDco 0.003 0.009 0.019 0.028 0.099 0.073 FEDHCN 0.080 0.072 VCO2 0.012 1.030 1.042 1.069 5.772 5.248 FEDo 0 0 0 0 0.021 0.012 FED/m~n 0 0 0.003 0.009 0.020 0.030 1.054 0.773 FED 0 0 0.039 0.156 0.436 0.676 1.730a 2.503 aBy 71 minutes the mass-loss exposure dose of irritants was approximately 300 g m3 min. which may cause some lung damage.

OCR for page 175
David Purser 187 The smoke layer had reached the floor after an hour, but there was a concentration gradient for smoke and toxic gases between the burn room and the target room. The major narcotic gas present was CO, which gradually increased in concentration in the burn room from 180 ppm during the first 13 minutes to 1,000 ppm at 67-75 minutes. This was sufficient to have caused incapacitation floss of consciousness) in just over 1 hour in the burn room but probably not in the target room, where the concentration was lower. When flaming ignition occurred, the chair burned very rapidly and produced high concentrations of narcotic gases, which would have been almost immediately fatal in the burn room. Within the target room an occupant would have become unconscious within less than 1 minute and received a fatal dose within 2 minutes. The smoke in the system was also very irritant, and it is likely that anyone spending more than 1 hour in the burn room would have suffered serious and possibly fatal lung damage, even if they had been rescued. This example illustrates the dangers of smouldering, which can continue for several hours and spread lethal products throughout a building. It is therefore dangerous to a sleeping, trapped, or otherwise incapacitated occupant. Since such fires often change to flaming before they are discovered, it is difficult to know the true incidence of incapacitation and death occurring during the nonflaming phase of fires. For this example both the standard and FR chairs would have caused incapacitation after ~ hour in the burn room, but due to its higher yield of CO and irritants, the FR chair would also cause incapacitation in the target room soon after, and death in the burn room after I.5 hours of smouldenng. For the standard foam, death at both locations would occur within ~ minute of the spontaneous transition to flaming after 75 minutes. Early-Flannog, WeN-Ventilated Fires In a Room with an Open Doorway In early-flaming fires, the decomposition conditions, particularly the air/fuel ratio, and the type of fuel are major determinants of the yields of toxic gases. In early-flaming, well- ventilated fires involving non-fire-retarde`d materials, combustion is usually efficient, so that the main products are heat, carbon dioxide, and water, the yields of toxic products and smoke being relatively low. However, when inefficiently burning materials are present, or when the fire ~ or _ ,~ , grows to a large size relative to that of the compartment, so that the ventilation may become restricted even with an open doorway, then toxic products may become important as well as heat. The example shown in Table 5 gives the results obtained during the first 5 minutes of a burn with flaming ignition of a rather bad single armchair (polystyrene with polyurethane cushions and covers) in a room with an open doorway. For an occupant of the room during the fire the following effects are predicted: I. Towards the end of the second minute and the beginning of the third minute, the smoke optical density and mass loss would sufficiently exceed the escape limitation thresholds for visual obscuration and sensory irritation to severely inhibit escape from the room. 2. During the fourth minute, the average temperature is 220 C, and sufficient heat would be accumulated in the skin surface to cause skin burns resulting in incapacitation.

OCR for page 175
188 Improved Fire- and Smoke-Resistant Materials TABLE 5 FED Analysis of Toxic and Physical Hazards from a Chair Burning in an Open Room 1 mill 2 min 3 min 4 mitt 5 min Average Gas Concentrations CO (ppm) 0 0 500 2,0003,500 HCN (ppm) 0 0 0 75125 CO2 (%) 0 0 1.5 3.56 O2 (%) 20.9 20.9 19 17.5 15 Fractional Incapacitating Doses FEDCo 0~00 0.00 0.017 0.074 0.130 FEDHCN 0.00 0.00 0.000 0.025 0.080 VCO2 1.00 1.00 1.442 2.376 4.434 FEDO2 0.00 0.00 0.001 0.002 0.007 FED/min 0.00 0.00 0.026 0.235 0.938 FED 0.00 0.00 0.026 0.261 1.199 Doses of Convected Heat Temp C 20 65 125 220 405 FED/min 0 0.033 0.170 2.273 355 FED 0 0.033 0.203 2.476 355 Radiant Heat Flux W/cm2 0 0.01 0.04 0.10 0.25 3. Dunng the fifth minute, a victim is likely to lose consciousness due to the combined effects of the accumulated doses of narcotic gases. The threshold limit for pain from radiant heat is also reached. 4. An occupant escaping or rescued after the fourth minute would suffer severe post- exposure effects due to skin burns, possible laryngeal burns with accompanying oedema and danger of obstructive asphyxia, and also lung oedema and inflammation which might well be fatal (due to the combined effects of inhaled hot gases, chemical irritants, and the pulmonary secondary effects of skin burns). After the sixth minute, it is likely that a victim would die at some time between a few minutes and ~ hour due to the effects of narcosis, circulatory shock, and possibly hyperthermia. In an aircraft in flight, a growing flaming fire is obviously a very serious hazard, and is likely to cause incapacitation within minutes unless it is contained within an unoccupied, closed compartment or rapidly extinguished. Fires within small, concealed spaces are likely to be subject to restricted ventilation, which results in additional hazards, as described in the next

OCR for page 175
David Purser ~9 section. For flaming fires on the ground, He main consideration is the fire-growth rate and the time required for the occupants to disembark. Although aircraft materials have a high standard of fire resistance, it is possible to conceive of a number of situations in which rapid fire growth could occur. With regard to people's perception of the smoke from these and other fires, there seems to be a considerable variation in the effects, depending upon what is burning. For example, in a recent fire a cotton-covered coconut-matting sofa burned in an open room in a two-story apartment. The apartment and the corridor outside it became logged with a thick smoke, but a number of persons were able to move through the smoke without difficulty (although the occupants perceived the smoke to be too dangerous to escape through and retreated to a bedroom, from which they were rescued). In some situations it is possible to account for effects on fire victims in terms of the classical toxicology models described, but in others the smoke appears to be much worse than would be predicted. For example, a fire fighter reported that he inhaled a whiff of smoke from the burning interior of a car and immediately almost fainted. In some major fire disasters, such as the Woolworth fire in Manchester (England) in May 1979 and the Dupont Plaza Hotel fire, victims have been found dead, sitting at tables with food, in the latter case only a few feet from a safe exit. Such individuals have apparently been overcome by smoke inhalation so quickly that they have been unable to move, possibly by a single breath of smoke. It would seem that this would be most likely to occur to a victim remote from the site of a pre- or post-flashover fire that has become vitiated, who is suddenly exposed to smoke containing a very low oxygen concentration and very high concentrations of toxic effluents. Small, Restr~cted-Ventilation fires In Closed Compartments Another very hazardous situation is that of a fire in a closed room. In this situation a smouldering or especially a flaming fire quickly uses up the available oxygen, and as the oxygen concentration falls after a minute or so the combustion becomes inefficient, producing a dense smoke rich in carbon monoxide and other toxic products. These, together with the lowered oxygen concentration in a room, can produce a rapidly lethal atmosphere. An example is a recent fire involving an adult and a 4-year-old child. Both were in a small bedroom for a short time during which the adult went to sleep and the child is thought to have ignited a small piece of foam using a cigarette lighter. The fire was discovered after a few minutes, when the door was opened by a family member, who extinguished the very small fire with a bucket of water. Both the adult and child were dead, with blood carboxyhaemogIobin concentrations of about half a lethal level. Based upon the dimensions of the room, it is calculated that the decomposition of approximately 0.5 kg of material would be sufficient to lower the oxygen concentration towards 10 percent and give carbon monoxide concentrations of approximately ~ percent or more, which together with other toxic products would cause incapacitation and death within a few minutes. In an aircraft in flight, a small, poorly ventilated, flaming fire could therefore present a serious hazard from the high yields of carbon monoxide and hydrogen cyanide and the consumption of oxygen.

OCR for page 175
190 Improved Fire- arid Smoke-Resistant Materials IbUy Develope~Post-Flashover Fires The final situation to consider is that where the occupants are remote from the site of a large fire, either in a large compartment or at a location remote from the fire compartment. This is the situation leading to the major multiple-death disasters, and against which most regulatory requirements for passive and active fire protection are designed to provide protection. Once a fire has reached a large size, the rate of burning and of the evolution of heat and toxic smoke is very great, so that even a large building can be rapidly fitted with smoke. The decomposition conditions in such fires depend upon the ventilation, but most fully developed fires in buildings tend to be hot and oxygen vitiated. These conditions favour inefficient combustion, with high yields of dense smoke containing high concentrations of carbon monoxide, hydrogen cyanide, and low oxygen. An example of such a fire is the penetration of a large external fuel fire into the cabin of an aircraft, as happened in the Manchester Airtours fire (King, 1989~. Table 6 shows the results obtained inside the cabin of a Boeing 707 containing a few rows of seats opposite an open doorway, outside which were 200 litres (50 U.S. gallons) of burning aviation fuel (Avtur) (Fardell and Purser, 19911. The rapid involvement of the cabin contents gave rise to a dense smoke containing large amounts of carbon monoxide and hydrogen cyanide at a measurement point halfway down the fuselage. It is predicted that escape capability is likely to be severely inhibited after approximately I-~.5 minutes due to the effects of exposure to the dense smoke containing high concentrations of acid gases and organic irritants. Severe incapacitation floss of consciousness) is predicted afteriust 2 minutes, followed rapidly by death, mainly from the effects of hydrogen cyanide (high concentrations of which were found in the blood of the Manchester victims). Although in many large fires the original fuel and the major source of heat and toxic products may be the contents, a significant contribution may be made by construction products. Of great importance in some cases are surface coverings or components with a large surface area such as doors or partitions. Surface coverings may contribute to flashover spread (as in the Dublin Stardust disco fire) and may release a bolus of toxic products very quickly, which may have a serious incapacitating effect on victims. An example would be viny} wall coverings or the viny} laminates used in aircraft cabins. Polyviny~chIoride releases all its hydrogen chloride at a low temperature (approximately 250-300 C), so that as a fire develops and the hot layer reaches this temperature, hydrogen chloride (MCI) may be suddenly released. In another aircraft fire test conducted by the Federal Aviation Administration, high concentrations of HC! and hydrogen fluoride (HF) occurred in the cabin atmosphere before other gases reached toxic levels (see Purser, 19891. In general, although in some cases fire and heat may eventually kill victims, this is usually preceded by dense, highly toxic smoke that can spread rapidly throughout a space or a building, and it is this that is usually responsible for the initial incapacitation of occupants, as well as being the cause of many deaths.

OCR for page 175
David Purser 191 TABLE 6 Average Concentrations of Toxic and Physical Hazards over 30-Second Penods Dunug Aircraft and Fractional Capacitating Doses External Fuel/Cabm Fire - Mid Cabin, 1.7 Metres 0.5 mill 1.0min 1.5m~n 2.0 mitt 2.5 min 3.0min 3.5 mitt Gas Concentration CO (pm) 8 HCN (ppm) 0 CO2 (%) 0.0 O2 (%) 21 34 10 0.0 21 282 1,157 3,3268,410 19,490 38 143 340740 1,380 0.4 1.2 2.84.1 6.0 21 20 1816 13 Fractional Idcapacitating Doses FEDCo 0.00 0.00 0.00 0.02 0.06 0.16 0.38 FEDHcN 0.00 0.00 0.01 0.06 5.65 > 10 > 10 VCO2 1.00 1.00 1.12 1.31 1.77 2.27 3.26 FEDo 0.00 0.00 0.00 0.00 0.00 0.00 0.01 FED/30s 0.00 0.00 0.00 0.10 10.10 > 10 > 10 FED 0.00 0.00 0.00 0.11 11.10 > 10 > 10 Radiant Heat Flux W/cm2 0.10 0.12 0.14 0.18 0.23 0.28 0.57 NOTES: Time to exceed smolce tenability limit: 1 minute 40 seconds. Time to incapacitation by narcotic gases: 2 minutes 15 seconds. Time to incapacitation by convected heat: 2 minutes 4S seconds. Time to tenability limit for radiant heat: 2 minutes 45 seconds. Effects of irritants: Over period between 1 and 4 minutes: average respirable particulates 6.7 mg/1, average total particulates 11.6 mg/1, average HCI concentration 1,027 ppm, average HBr concentration 1,228 ppm. It is considered that the oily, organic-rich, particulate collected, with its very high acid gas content, would be highly irritant and extremely painful to eyes and breathing, causing incapacitation and impairing escape attempts. It is considered likely that these irritants reached high concentrations (approaching 1,000 ppm total acid gases) early in the fire at approximately 1-1.5 minutes, from which time escape capability would be significantly impaired. It is likely Hat sufficient irritants would be inhaled up to 4 minutes to cause life-threatening post-exposure lung damage. Table 7 shows the results of another Boeing 707 fuselage test, which was conducted under the same conditions as those used for the test whose results are shown in Table 6, except that a fine water spray mist was applied to the zone of the cabin interior opposite the fuel fire. In contrast to the previous fire, the water mist prevented significant decomposition of the cabin interior materials, so that the cabin remained tenable for up to approximately 7 minutes.

OCR for page 175
192 Improved Fire- aru] Smoke-Resistant Materials TABLE 7 Average Concentrations of Toxic and Physical Hazards and Fractional Incapacitating Doses over 30 Second Periods During External Fuel and Aircraft Cabin Fire, - Cabin Fire Zone Sprayed with Water Mist - Mid Cabin, 1.7 Metres 3.0 mitt 3.5 min 4.0 mill 4.5 min 5.0 mitt 5.5 min 6.0 min Gas Concentrations CO (ppm)213239248 260 316383535 HCN (ppm)469 9 888 CO2 (%)0.10.20.2 0.2 0.30.30.4 O2 (%)20.720.720.7 20.7 20.620.620.3 Fractional Incapacitating Doses FEDCo 0.004 0.004 0.004 0.004 0.005 0.007 0.009 FEDHCN 0.003 0.003 0.003 0.003 0.003 0.003 0.003 VCO2 1.02 1.04 1.04 1.04 1.07 1.07 1.1 FEDo 0.000 0.000 0.000 0.000 0.000 0.000 0.000 FED/30s 0.007 0.007 0.007 0.007 0.009 0.011 0.013 E FED 0.007 0.014 0.021 0.028 0.037 0.048 0.061 FEDs of Convected Heat Temp C 20 20 20 23 26 31 39 FED/30s 0.000 0.000 0.000 0.005 0.001 0.007 0.008 FED 0.000 0.000 0.000 0.005 0.011 0.018 0.026 Radiant Heat Flux W/cm2 0.9 0.9 1.0 1.0 1.0 1.0 1.0 NOTES: Time to exceed smoke tenability limit: 2 minutes. Time to incapacitation by narcotic gases: no incapacitation. Time to incapacitation by convected heat: based upon the algorithm for dry or slightly humid air incapacitation would not occur. Based upon work with saturated air, it is considered that conditions would be tenable for up to 7 minutes (end of test). Since this was a zone-sprayed test, there would be no spray droplets falling on the skin. Time to tenability limit for radiant heat: limit not exceeded, maximum 1.2 kW/m2 (at 7 minutes). Effects of irritants: Over period between 1 and 4 minutes: average respirable paniculates 0.24 mg/1, average total particulate 0.65 mg/1, average HCI concentration 23 ppm, average HBr concentration 0 ppm. It is considered that the combined concentration of these and other irritants would have some irritant effect on the eyes and respiratory tract, but probably not sufficient to cause serious incapacitation or seriously impair escape attempts, or cause serious post-exposure lung damage. The total particulate concentration is much lower than observed in the nonsprayed and fully sprayed tests. It is considered likely that these irritants reached high concentrations (approaching 1,000 ppm total acid gases) early in the fire at approximately 1-1.5 minutes, from which time escape capability would be significantly impaired. It is likely that sufficient irritants would be inhaled up to 4 minutes to cause life-threatening post- exposure lung damage.

OCR for page 175
David Purser 193 RECOMMENI)ATIONS TO ENHANCE SURVIVABILITY IN AIRCRAFT FIRE ACCIDENTS The development of hazardous situations in a fire involves a wide range of factors. These include fire development, from ignition to He post-flashover spread of fire and smoke; toxicity; the interaction of the fire with the structure and with passive and active fire protection; and escape-related factors including deeechon, warnings, the provision of escape routes, way-finding, physiological and behavioural impairment, and escape movements or rescue. In designing a system to be safe in fire, all these factors should be considered, and the ultimate evaluation of safety depends upon whether it is possible to ensure, by performing a life-threat hazard and risk assessment, that the occupants can reasonably be expected to have escaped before they are exposed to levels of heat and smoke that may endanger health and threaten life. In the context of aircraft fires, there are a number of strategies that could be adopted to enhance survivability in aircraft fire accidents. The considerations differ somewhat between in- flight fires and post-crash fires. In-Flight Fires The foremost considerations with regard to in-flight fires must be prevention, limitation, detection, suppression, and compartmentation. Aircraft materials, design, and construction are all planned with fire prevention in mind. Materials, particularly cabin lining and seating materials, are currency selected with a high fire-performance specification in terms of ignitability. Ignitability, flame spread, and heat release characteristics are all very important in terms of fire initiation and growth. Toxic potency criteria may also be a consideration in terms of materials selection. For fires in concealed spaces, early detection and the provision of suppression systems are very important. Containment and fire stopping is doubly important. First, if the fire occurs in a sealed space it will self-extinguish, and second, penetration of toxic smoke generated in the early stages of the fire into occupied areas is less likely to occur. Fires involving wire and cable can present a problem in smoldering or small flaming fires due to the organic irritants and acid gases that evolve. Materials with an improved flame propagation and higher decomposition temperature performance may be advantageous, provided that they do not evolve high toxicity products when they are overheated (Purser et al., 19941. In situations where the cabin or cockpit atmosphere becomes contaminated with smoke from small in-flight fires, it may be important to minimize the inhalation of irritant smoke. This might be done by increasing the rate of cabin air change to reduce smoke concentration, or in extreme cases by using oxygen masks Provided that the fire is not in the cabin). Consideration should also be given to the psychological effects of being trapped for some hours in an irritant-smoke atmosphere.

OCR for page 175
194 Improved Fire- aru] Smoke-Resistant Materials Fires on the Ground if the aircraft is on the ground the main hazard is from rapidly growing fires, of which a post-crash fire involving fuel is the most serious. As shown in Table 6, such fires can render the cabin untenable within a few minutes. The main methods for improving survivability must therefore be to delay the spread of fire into the cabin, and then the fire involvement of the cabin interior, while achieving rapid evacuation. Improved ignitability and fire-growth performance of materials have gone some way towards improving the fire performance of seats and other cabin materials, but large external fuel fires are still capable of igniting cabin materials in a short time. Low-volume water mist systems have been shown to provide a significant benefit in some situations by delaying cabin interior involvement in the fire. Other problems in this scenario impeding evacuation may be the presence of dense, irritant smoke and high concentrations of hydrogen cyanide and carbon monoxide. The evolution of irritants may be reduced by choosing cabin lining materials that do not evolve high HC! and smoke yields at relatively low temperatures. It might also be an advantage to avoid or protect materials likely to evolve high yields of HCN, such as polyurethanes. Other aspects to consider are ways of improving passenger egress. Better protection to avoid physical injury in crashes, less cabin baggage in upper stowage bins, wider aisles and more exits, good low-level lighting and signage, and similar methods might improve movement ability. Improved warnings in case of emergency, such as encouraging passengers to make an escape game plan in advance, could also be beneficial. REFERENCES Alane, Y. 1981. Bioassay for evaluating the potency of airborne sensory irritants and predicting acceptable levels of exposure in man. Food and Cosmetics Toxicology 19:623-626. Braun, E., B.C. Levin, M. Paabo, I. Gurman, T. Holl, and I.S. Steel. 1987. Fire Toxicity Scaling. NBSIR 87-3510. Ga~thersburg, Maryland: National Bureau of Standards. Bryan, I.~. 1977. Smoke as a Determinant of Human Behavior in Fire Situations (Project People). NBS-CGR-77-94. Gaithersburg, Maryland: National Bureau of Standards. Fardell, P.~., and D.A. Purser. 1991. Toxicological and Respirable Threats. Presented at Cabin Water Spray Systems, Civil Aviation Authority, Industry Consultative Conference, Hilton International Hotel, Gatwick, United Kingdom, May 29-30. BSI. 1989. Guide for the Assessment of Toxic Hazard of Materials in Fire in Buildings and Transport. BST DD 180. London: British Standards Institution. Hartzell, G.E., W.G. Swi~czer, and D.N. Priest. 1985. Modeling of toxicological effects of fire gases. Journal of Fire Sciences 3:330-342. ISO. 1994. Toxicity Testing of Fire Effluents - Part 5, Prediction of Toxic Effects of Fire Effluents. ISO/lEC TR 9122-5. Geneva, Switzerland: International Standards Organization. fin, T. 1976. Visibility through fire smoke, Part 5. Allowable smoke density for escape from fire. Report of Fire Research Institute of Japan No. 42: 12.

OCR for page 175
David Purser 195 King, D.F. 1989. Aircraft Accident Report 8/~S Air Accidents Investigation Branch. HMSO, London. Purser, D.A. 1988. Toxicity assessment of combustion products. Pp. 1-200 to 1-245 in The SFPE Handbook of Fire Protection Engineering. C.L. Beyler, ed. Quincy, Massachusetts: National Fire Protection Association. Purser, D.A. 1989. Modelling time to incapacitation and death from toxic and physical hazards in aircraft fires. Pp. 41-1 to 41-13 in AGARD Conference Proceedings No. 467, Aircraft Fire Safety. Neuilly-Sur-Seine, France: North Atlantic Treaty Organization, Advisory Group for Aerospace Research and Development. Purser, D.A. 1990. The development of toxic hazard in fires from polyurethane foams and the effects of fire retardants. Pp. 206-221 in Proceedings of Flame Retardants 90, British Plastics Federation. London: Elsevier. Purser, D.A. 1993. Interactions Between Behaviour Patterns and Physiological Impairment in Escape from Fire. Presented at Interflame '93 Conference, Oxford, United Kingdom, March 30-April 1. Purser, D.A., P.~. Fardell, and G.E. Scott. 1994. Fire Safety of PTFE-based Materials Used in Buildings. Report 274. Garston, U.K.: Building Research Establishment. Sime, I. 1992. Crowd safety design, communications and management: The psychology of escape behaviour. Pp. 16-29 in Easingwold Papers No. 4. York, U.K.: Home Office Emergency Planning College. Wood, P.G. 1972. The Behaviour of People in Fires. Fire Research Note 953. Garston, U.K.: Fire Research Station.

OCR for page 175