Sei sulla pagina 1di 6

1 INTRODUCTION

If safety measures for self-rescue can not be as-


sessed within a quantitative risk analysis (QRA), the
risk judgement system does not motivate companies
or local authorities to take measures, because their
effects are not visible in a state-of-the-art risk as-
sessment. In most quantitative risk analysis methods,
persons present in the hazardous area are assumed to
be exposed for a fixed amount of time. Assumptions
for fixed exposure times are 30 minutes for a toxic
exposure and 20 seconds for exposure to heat radia-
tion. Furthermore, persons are assumed to stay on
the same place. The reality is different: in case of an
emergency, every person capable of escape will try
to rescue himself. In case of a toxic release it is pos-
sible that a safe place (for example inside a building)
is reached within the prescribed 30 minutes. On the
other hand, in case of fire in crowded places, it can
be expected that people are unable to escape within
20 seconds. Examples of methods are the Dutch
probabilistic QRA methods for transport of danger-
ous goods and for stationary installations (Uijt de
Haag 2006). In QRA calculations where self-rescue
is very important, for instance in tunnel safety stud-
ies, extensive calculations with evacuation models
are performed. Subsequently the output of the
evacuation models is used in the QRA, the evacua-
tion model is not integrated within the QRA.
The model described in this article provides a so-
lution between evacuation modelling, which might
be too time-consuming and state-of-the-art QRA-
calculations, which have insufficient detail to show
the effect of safety measures. The model can be in-
tegrated in a quantitative risk analysis method,
therewith enabling to account for self-rescue and to
show the effect of safety measures by running a sin-
gle QRA model. The model is especially suitable for
QRAs involving risks to densely populated areas,
such as railway stations, stadiums, and offices as
well as for industrial plants and their inhabited sur-
roundings.
2 MODEL OUTLINE
In order to judge the effect of self-rescue improving
measures, a model has been developed to quantify
the effect of self-rescue, depending on exposure to
fire or toxic chemicals. The model distinguishes heat
radiation, smoke and several types of toxic chemi-
cals and quantifies their effect on the walking veloc-
ity and exposure duration (see Figure 1).


Figure 1. Self-rescue in quantitative risk analysis
Self-rescue in quantitative risk analysis
I.J.M. Trijssenaar- Buhre & I.M.E. Raben & T. Wiersma & S.I. Wijnant
TNO, Apeldoorn, The Netherlands

ABSTRACT: In quantitative risk analysis (QRA) methods, the damage of toxic and fire effects to persons is
determined using a fixed exposure time. The model described in this paper enables to include self rescue in
QRA methods. The model can be used 1) for calculations of a more realistic exposure time, 2) to determine
the number of persons incapable of self-rescue, which is important information for the rescue services 3) to
determine the effect of safety measures improving self-rescue. The model provides a solution between
evacuation modelling, which might be too time-consuming and state-of-the-art QRA-calculations, which have
insufficient detail to show the effect of safety measures.

Other factors that are considered in the model are the
distance to a safe location and the presence of a bot-
tleneck in the evacuation route. The model can be
used for the following purposes:
For fast calculations of the exposure time and the
number of persons incapable of self-rescue;
To determine the effect of safety measures im-
proving self-rescue, by calculating the remaining
consequences for a scenario after taking meas-
ures. Subsequently these consequences can be
compared to those of the original scenario in the
QRA.
The following sections show the relations for de-
scription of the toxic effects and their influence on
the mobility or walking velocity. Note that in case of
fire, for instance, there can be interaction between
toxic exposure and heat exposure: if the mobility de-
creases due to toxic exposure (smoke), the exposure
time to heat can increase.
3 TOXIC INJURIES
For toxic injuries a distinction is made between as-
phyxiant and irritant gases. For an asphyxiant prod-
uct, such as carbon monoxide, the most important
criterion for its toxic effects is the concentration in
the blood supply to the brains. On the other hand, for
an irritant product the most important factor is the
concentration in the lining of the nose, throat, or
lung (Purser 2002).
3.1 Asphyxiant fire products
The Fractional Incapacitation Dose (FID) model of
Purser (Purser 2002) has been selected to describe
the effects of asphyxiant fire products for their pres-
ence in fire smoke as well as for the pure substance.
The FID model relates the toxic dose inhaled by a
person to the dose where incapacitation (i.e. loss of
consciousness) occurs. The FIN is the sum of the
FID values for all asphyxiant fire products, which
also accounts for hyperventilation resulting from ex-
posure to carbon dioxide. Incapacitation occurs by a
(cumulated) exposure to carbon monoxide, hydro-
cyanic acid and oxygen deficiency (FIN
total
=1) or by
exposure to high concentrations of carbon dioxide
(FIDCO
2
=1).

( )
2
2
* ) (
O HCN CO total
FID VCO FID FID FIN + + = (1)
where FID
CO
, FID
HCN
and FID
O2
= Fractional Inca-
pacitation Dose for CO, HCN and O
2
respectively;
CO = carbon monoxide; HCN = hydrocyanic acid;
O
2
=oxygen; CO
2
= carbon dioxide; VCO
2
= multi-
plication factor hyperventilation [-].

The Fractional Incapacitation Dose is described for
CO, HCN, O
2
, and CO
2
by equation 2 to 6 (Purser
2002).

PID
t
RMV CO FID
CO
036 . 1 5
] [ 10 317 . 3

= (2)
where [CO] = CO concentration in ppm; RMV
=Respiratory Minute Volume in litre/min, which is
25 litre/min at light activity; t = exposure time in
min; PID = Personal Incapacitation Dose =30%.

The FIDs for oxygen and hydrocyanic acid are de-
scribed by equations 3 and 4. Equation 3 and 4 are
valid for relatively short exposure time (shorter than
1 hour) and a constant concentration.

) % 9 . 20 ( * 54 . 0 13 . 8
2 2
O V O
e
t
FID

= (3)
where V%O
2
= O
2
volume percentage;

] [ * 023 . 0 396 . 5 HCN
HCN
e
t
FID

= (4)
where [HCN] = HCN concentration in ppm.

Carbon dioxide (CO
2
) increases the Respiratory
Minute Volume, resulting in a larger uptake rate of
other gases. The increase of Respiratory Minute
Volume by hyperventilation is described by a multi-
plication factor VCO
2
:

1 . 7
004 . 2 2 % * 1903 . 0
2
+
=
CO V
e
VCO (5)
where V%CO
2
= CO
2
volume percentage.

Besides its effect on the Respiratory Minute Vol-
ume, CO
2
in itself can lead to incapacitation when
present in higher concentrations.

2
2 % * 5189 . 0 1623 . 6 CO V
CO
e
t
FID

= (6)
Using equations 1 to 6, the FIN can be determined
for toxic fire products within fire smoke as well as
for the separate toxic substances.
For asphyxiant fire products the time to incapaci-
tation and its severity usually show a short period of
intoxication that is followed by a relatively sharp
decline into incapacitation (Purser 2002). The rela-
tion between FIN and mobility is shown in figure 2
(University of Greenwich, 2004).

0,0
0,2
0,4
0,6
0,8
1,0
0 0,2 0,4 0,6 0,8 1
FIN or FIC
M
o
b
i
l
i
t
y

f
a
c
t
o
r
FIN
FIC

Figure 2. Mobility as a function of FIN and FIC

3.2 Irritant fire products
Dose calculations are applied for incapacitation ef-
fects of asphyxiant fire products as well as for the le-
thal effects of irritant fire products. However, the in-
capacitation effect of irritant fire products is related
to the concentration instead of the dose. The toxic
effects are described by the Fractional Irritant Con-
centration (FIC) model.

2 2
] [ ] [ ] [ ] [ ] [
2 2
NO SO HF HBr HCl
F
NO
F
SO
F
HF
F
HBr
F
HCl
FIC + + + + = (7)
where [HCl] = Hydrochloric acid concentration in
ppm; F
HCl
= threshold concentration for HCl in ppm;

The threshold values for irritant fire products are
given in table 1. Incapacitation occurs when the FIC
equals unity, for which the toxic effect of different
irritant fire products can be added. The relation be-
tween FIC and mobility is shown in figure 2 (Uni-
versity of Greenwich, 2004).


Table 1. Threshold values for irritant fire products
_____________________________________
Toxic gas Severely irritant concentration
(ppm)
_____________________________________
HBr 200
HCl 200
HF 120
NO2 80
SO2 30
_____________________________________

3.3 Toxic chemicals
In order to be able to translate the presence of toxic
chemicals into mobility, the methods of FID and FIC
are also applied for releases of toxic chemicals. The
general equations for the FID and FIC methods are:


= = =
= =
n
i i
i
n
i
t
t i
i
Ct
t D
t
Ct
C
FID
1 1 0
) (
) (
) (
(8)

=
=
n
i i
i
F
C
FIC
1
(9)
where C
i
= average concentration in ppm of chemi-
cal i over the chosen time increment; t = chosen
time increment in min; D
i
(t) = dose of chemical i at
time t; (Ct)
i
= the threshold dose in ppm*min; F
i
=
threshold concentration for chemical i in ppm.

Because the FID and FIC threshold values are only
known for fire products, other threshold values are
required for the toxic chemicals. The Acute Expo-
sure Guideline Level 2 (AEGL-2) is very suitable
for the definition of ability of self-rescue. AEGL-2 is
the airborne concentration of a substance above
which it is predicted that the general population, in-
cluding susceptible individuals, could experience ir-
reversible or other serious, long-lasting adverse
health effects or an impaired ability to escape.
AEGL-2 values are available for exposure durations
of 10 minutes, 30 minutes and 1 hour, 4 hours and 8
hours. The subdivision of several chemicals in as-
phyxiant and irritant as well as their threshold values
are shown in table 2. The chemicals in table 2 are
representative for various hazard categories as used
in QRAs. The AEGL-2 is not (yet) determined for
acrylonitrile and ethylchloride, the Immediately
Dangerous to Life and Health (IDLH) value is used
for these chemicals. IDLH is defined as a concentra-
tion that an exposure up to 30 minutes does not
cause death, serious or irreversible health effects, or
does not impair or impede the ability to escape.


Table 2. Subdivision and AEGL-2 threshold values of toxic
chemicals.
________________________________________________
Chemical Subdivision Concentration (ppm)
__________________
10 min 30 min
exposure exposure
________________________________________________
Acrylonitrile FID - 85*
Ethyl chloride FID - 3800*
Ammonia FIC 220 220
Chlorine FIC 2.8 2.8
Methylisocyanate FID 0.40 0.13
Nitric acid FIC 43 30
_____________________________________________
* IDLH threshold value

The FID can now be determined by calculating
the threshold dose (Ct) from table 2 and inserting
this value into equation 8. For instance, the threshold
dose (Ct) for acrylonitrile is 255 ppm*min. The FIC
can be determined by taking the threshold concentra-
tion from table 2, for nitric acid the value depends
slightly on the timescale of the calculations (10 min
or 30 min value).
The FID and FIC are translated into a walking ve-
locity using the relations shown in figure 2.
4 THERMAL INJURIES
Exposure to thermal radiation is assumed to have lit-
tle or no effect on the mobility of the victims, until
the injuries become lethal. During the self-rescue pe-
riod, the heat radiation to which the person is ex-
posed will decrease with the distance. This is ac-
counted for by calculating the effective exposure
duration (CPR 16E 1992):

(
(
(

|
|

\
|
+ +

3
5
0
0
1 1
5
3
v r eff
t
x
u
u
x
t t (10)
where t
eff
= effective exposure time (s); x
0
= the ini-
tial distance to the fire source (m); u = the evacua-
tion velocity (m/s); t
r
= reaction time (= 5 s); t
v
= (x
s
-
x
0
)/u = evacuation time (s); x
s
= distance from fire
source where heat radiation is below 1 kW/m
2
(m).
The evacuation velocity, u, can be influenced by
exposure to toxic fire products: if the mobility de-
creases the exposure time to heat is increased:
u = mobility factor * u
0
, where u
0
= maximum
evacuation velocity. The effective exposure time ob-
tained with equation 10 can subsequently be used to
calculate degree of injuries and lethal victims due to
heat exposure.
5 EVACUATION
Two important characteristics of the evacuation
route are included in the model, namely the distance
to a safe location and the presence of bottlenecks on
the evacuation route.
5.1 Distance to safe location
The distance to a safe location can be defined by a
threshold value (concentration, maximum dose or
heat radiation level) or it can be the distance to a
place to shelter, such as a building, where windows
and doors can be closed. The mobility of a person is
influenced by the concentration or toxic dose to
which he is exposed on his way to the safe location.
In case of fire, the distance to a heat resistant shelter
or the distance to 1 kW/m
2
(x
s
) is inserted in equa-
tion 10.
5.2 Bottlenecks on the evacuation route
The evacuation calculations within the model are not
intended to model the evacuation route in detail.
However in cases of important bottlenecks in the
route, such as a door or stairs, it can be very impor-
tant to be able to model the effect of the bottleneck.
The evacuating persons need to pass the bottleneck
before reaching a safe location. The actual exposure
time increases due to the time necessary to pass the
bottleneck. An average value for the extra exposure
time due to the bottleneck is given by:

t
bottleneck
= n/ C
bottleneck
(11)

where n = number of persons that need to pass
through the bottleneck (-); C
bottleneck
= bottleneck ca-
pacity in number of persons per minute.
6 IMPLEMENTATION AND APPLICATION
6.1 Model implementation
The model can be implemented in several ways, this
chapter describes the following examples:
1 Straightforward implementation for QRA
method;
2 Dynamic implementation for scenario analysis or
more extensive QRA calculations.
The difference in implementation is the level of de-
tail of modelling the effect of toxic exposure to the
evacuation velocity.
6.2 Straightforward implementation method
For the QRA method a straightforward implementa-
tion is often sufficient. In this case the evacuation
velocity is based on the concentration profile on the
starting location of the person. For asphyxiant
chemicals, the first step is to estimate the exposure
time with the maximum evacuation velocity, u=u0,
according to:

t
exp
= t
r
+ t
bottleneck
+(x
s
-x
0
)/u; (12)

Subsequently the dose is determined using the con-
centration profiles C
i
(x
0
, t) on the starting location of
the person:

=
t
i i
dt t x C t D
0
0
) , ( ) ( (13)
FID and corresponding mobility factor are calcu-
lated with the equations in sections 3.1 or 3.3 and
figure 2. Subsequently the effective exposure time is
determined with equation 12, this time with u= mo-
bility factor*u
0
:

t
eff
= t
r
+ t
bottleneck
+ (x
s
-x
0
) / (mobility factor*u
0
)

In the final step the effective FID is determined us-
ing the effective exposure time. The effective FID is
used to determine whether a person with a certain
starting location will become incapable of self-
rescue during his attempt to escape (using figure 2).
For irritant chemicals the FIC at the starting posi-
tion is determined with the equations in sections 3.2
or 3.3 and the mobility factor is read from figure 2.
The effective exposure time can then be calculated
with equation 12, assuming u= mobility factor*u
0
.
The effective exposure time is important input for
calculating lethal damage.
6.3 Dynamic implementation method
For a scenario analysis the evacuation velocity will
be calculated dynamically: when a person moves to
another (safer) location, he will be exposed to a
lower concentration. Therefore the toxic dose will
increase less compared to staying on the same loca-
tion. For each time increment t, the FID or FIC and
mobility factor are calculated and used as input for
the next time increment until a person either reached
a safe location or is incapable of self-rescue. For irri-
tant chemicals C
i
(x,t) is input for equation 9. For as-
phyxiant chemicals the dose is calculated with:

=
t
i i
dt t x C t D
0
) , ( ) ( (13)
For both irritant and asphyxiant chemicals:
x = x
0
for t t
r
x = x
0
+ u(t-t
r
) for t > t
r
u= mobility factor*u
0
is determined for each time
increment with FID (as a function of D
i
(t)) or with
FIC and corresponding mobility factors.

6.4 Model applications
When applying the model, a more realistic expo-
sure time and therewith a more realistic estimation
of the (lethal) damage of this exposure can be calcu-
lated. Even more important, is the use of the model
to estimate the number of persons in the hazardous
area, who are not capable of self-rescue. The FID
and FIC can be determined for every exposed person
depending on his starting location. If the FID or FIC
of a person exceeds unity, that person will be inca-
pable of self-rescue. Combining this information on
self-rescue with population data yields the total
number of persons incapable of self-rescue.
The number of persons incapable of self-rescue is
important information for preparation and real-time
activities of rescue services. In order to determine
the rescue capacity necessary to handle the accident,
the number of persons incapable of self-rescue is
more relevant than the number of fatal injuries.
The effects of safety measures improving self-
rescue can be determined with the use of the model.
An example of a safety measure is increasing the ca-
pacity of a bottleneck in the evacuation route or
placing a shelter or safe haven.
7 CONCLUSIONS
With state-of the art QRA methods, safety measures
for self-rescue cannot yet be assessed. The model
described in this paper enables to include self-rescue
in QRA methods. The model distinguishes heat ra-
diation, smoke and several types of toxic chemicals
and quantifies their effect on the walking velocity
and exposure duration. Evacuation route characteris-
tics considered in the model are the distance to a
safe location and the presence of a bottleneck in the
evacuation route. The model can be used to deter-
mine the number of persons incapable of self-rescue,
which is important information for the rescue ser-
vices. Furthermore the more realistic exposure times
obtained with the model can be used for improving
the estimation of lethal damage.
The model can be integrated within a quantitative
risk analysis method to directly show the effect of
self-rescue improving measures. The model is espe-
cially suitable for QRAs involving risks to densely
populated locations, such as railway stations, stadi-
ums, offices as well as industrial plants and their in-
habited surroundings.
NOMENCLATURE
[CO] Carbon monoxide concentration (ppm)
AEGL-2 Acute Exposure Guideline Level 2: airborne
concentration of a substance above which it
is predicted that the general population, in-
cluding susceptible individuals, could ex-
perience irreversible or other serious, long-
lasting adverse health effects or an impaired
ability to escape
C
bottleneck
Bottleneck capacity (persons / min)
C
i
(x,t) Concentration profile of chemical i (ppm)
F
HCl
Threshold concentration for HCl (ppm)
FIC Fractional Irritant Concentration (-)
FID Fractional Incapacitation Dose (-)
FIN Sum of the FID values for asphyxiant fire
products (-)
IDLH Immediately Dangerous to Life and Health
(IDLH) value: concentration that an expo-
sure up to 30 minutes does not cause death,
serious or irreversible health effects, or does
not impair or impede the ability to escape
n Number of persons that need to pass through
the bottleneck (-)
PID Personal Incapacitation Dose (%)
QRA Quantitative Risk Analysis
RMV Respiratory Minute Volume (litre/min)
t Exposure time (min)
t
eff
Effective exposure time (s)
t
r
Reaction time (= 5 s)
t
v
Evacuation time (s)
u Evacuation velocity (m/s)
u
0
Maximum evacuation velocity (m/s)
V%O
2
Oxygen volume percentage (%)
VCO
2
Multiplication factor hyperventilation (-)
x
0
Initial distance to the fire source (m)
x
s
Distance from fire source where heat radia-
tion is below 1 kW/m
2
(m)
REFERENCES
CPR 16E. 1992. Methods for the determination of possible
damage to people and objects resulting from releases of
hazardous materials. First edition. Committee for the Pre-
vention of Disasters caused by dangerous substances. The
Hague: directorate-General of Labour of the Ministry of
Social Affairs and Employment. ISBN 90-5307-052-4.
Purser, P.A. 2002. Toxicity assessment of combustion prod-
ucts, In: The SFPE handbook of fire protection engineering.
3
rd
edition. Quincy, Massachusetts: NFPA. ISBN 087765-
451-4.
Uijt de Haag, P.A.M. 2006. Handleiding Risicoberekeningen
BEVI (Guidelines for quantitative risk analysis, in Dutch).
University of Greenwich. 2004. Building Exodus Manual.

Potrebbero piacerti anche