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 176
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 177
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 178
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 179
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 180
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 181
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 182
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 + FED°rg (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 183
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 184
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 185
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 186
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 187
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 188
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 189
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 190
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 191
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 192
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 193
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 194
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 195
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 196
Representative terms from entire chapter:
respiratory tract
