Sie sind auf Seite 1von 4

Calculation of a Toxicity Index for Materials,

Based on a Biological Evaluation Method


Claire Herpol and Paul Vandevelde
State University of Gent, Belgium

An attempt has been made to present toxicological results obtained by a biological evaluation method in a simple and
comprehensive way. The calculation of a toxicity index which expresses mortality rates as a function of the very
important time factor is proposed. Additionally it is shown how such an index could be used to rank a series of
twenty materials according to their relative toxicity.

INTRODUCTION

Papers published earlier112 have provided results on the


toxicity of products of pyrolysis and more or less incomplete combustion of materials obtained by a biological
testing method. From these results, which at first sight
appeared to be very dependent upon testing conditions
(temperature, air flows, etc.), it was clear that even in a
quite definite combustion situation it would be very
misleading to classify materials using the mortality rates
obtained at the end (30 min) of the biological experiment
without taking into account the change of mortality
rates with time.
In assuming escape possibilities during a fire, the time
factor is obviously very important. There are, in fact,
at least two very good reasons for evaluating the
toxicity of a material on the basis of its giving rise sooner
or later to lethal evolved products during the test
duration.
First, it is statistically recognized394 that the majority
of fatalities in real fires-and particularly fires in the
home-are reported as being overcome by smoke and
toxic fumes. Moreover Chandler et al.4 have shown that
two-thirds of the fatalities in the home were found in the
room where the fire started, which suggests that the
subject fainted or otherwise failed to respond to the fire
situation and use the existing escape possibilities, if they
were known.
It remains debatable whether to use the moment of
incapacitation as a criterion (as referred to for instance
in Time of Useful Function (TUF) by Gaume et aL5)
or the moment of death as in our test method. As the
incapacitation time is certainly shorter, methods using
mortality rates to classify materials according to the
hazard they represent to human beings are less severe for
the materials, but nevertheless give valuable information
as to the relative toxicity of a series of materials.
The second reason is that 30 min is in fact a very long
exposure time. When a fire starts, providing detection
and alarm systems are working and escape routes are in
existence and known, the occupants should be evacuated
not only from the compartment where the fire occurs,
but also from the building, in considerably less than
30 min. Of course, toxicity is not only a potential danger

0Heyden & Son Ltd, 1978

for occupants but also for the firemen. However, if the


fire fighting tactics are correct, they should not be
exposed individually for such long periods.
For these reasons it is clear that materials giving high
mortality rates at the end of a 30 min exposure (see Fig. 1
(a and b, line 1)) are less dangerous, from the point of
view of escape, than materials which cause a higher
percentage of very early deaths (see Fig. 1 (a and b, line
2)).

Time( m i n )
1001

Figure 1. Theoretical examples of mortality rates with time scale


0308-0501/78/0002-0007 $02.00
FIRE AND MATERIALS, VOL. 2,

NO. 1,1978 7

C. HERPOL AND P. VANDEVELDE

In order to classify materials a Toxicity Index would


be most valuable. This has been proposed by other
researchers such as Tsuchiya et
who have suggested
the use of a maximum toxicity index. Unfortunately,
however, their proposed calculation is based only on the
chemical analysis of the evolved fire products.
Since it is now generally recognized that chemical
analysis of the evolved gases cannot alone give a sufficiently complete idea of the toxicity hazard, and that the
toxicity of materials should be evaluated by biological
means, with chemical analysis of the evolved products
being considered as a useful tool for the interpretation of
biological results, it seems that the calculation of a
toxicity index should be based on biological test results.
Such an index has been proposed for instance by
Truhaut et al.7 Their global toxicity index results from
physiological parameters reflecting the activities of the
cardio-vascular and nervous systems but it only expresses
the degree of intoxication of a single animal at a given
time.
The toxicity index proposed in this paper is also based
on biological test results but it takes time into account
as well as mortality rates.

kz (after 12min)=e$=2.50
k3 (after 18 min) = $$ = 1.66
kq (after 24 min) = $$ = 1.25
k5 (aftei 30 min) = %$= 1.
However, calculations made with this and other possible
ka values have shown that the best results are obtained
by simply using ks equal to 5 after 6 min, 4 after 12 min,
3 after 18 min, 2 after 24 min and 1 after 30 min.
The theoretical examples of Fig. 1 calculated by this
method give the following results.
Fig. l(a):
5 x 5 + 4 x 10+3 x 35+2 x loo+ 1x 100
~
_
_
_
~
5+4+3+2+1
=31.33

T (line I)=

T (line 2) =

~~~

Fig. l(b):
~

0=6.66
15

5 x 4 + 4 x 8 + 3 x 12+2x16+1 x 2 0
___ ______
15
=9.33

T(1ine 2)=-

In the biological experiment1,2 mortality rates are


obtained at regular intervals (6, 12, 18, 24 and 30min)
during the 30 rnin exposure period.
A first possibility would be to use the surface under
the lines giving the mortality Iates as a function of time
(Fig. l),
as is done for other fire reaction parameters
such as density of smoke, flammability, etc.8 Unfortunately this will not serve the present purpose since in the
two sets of examples given in Fig. 1 lines (I) and (2) cover
the same surface and it has already been recognized that
materials represented by lines (2) present a higher hazard.
Thus, it is clear that the surface under the lines cannot be
used as a valuable toxicity index.
This suggests that materials giving rise to early deaths
should be penalized. To do so a similar type of formula
as used for other criteria of fire reactions was tried and
the following equation was selected

+ 1 x 60

15

=40.33

T (line 1)=
PROPOSED TOXICITY INDEX

5 x 25 + 4 x 40+ 3 x 50+ 2 x 55

A toxicity index of this type could be used to classify


materials after the biological testing of their relative
toxicity. As an example of the many possibilities, Fig. 2
shows a way of establishing such classes: Class I : not
very dangerous-T between 0 and 6.66; Class 11: relatively dangerous-T between 6.66 and 20.00; Class 111:
dangerous--T between 20.00 and 40.00; Class IV: very
dangerous-T between 40.00 and 66.66; Class V:
extremely dangerous-T between 66.66 and 100.
Such a toxicity index must be expressed according to
testing conditions. Two conditions of prime importance
are temperature and dilution. As all kinds of temperatures (up to 1000C) and all kinds of dilutions can occur
in real fires, the aim is to find out the potentialtoxicity of
a material, i.e. the worst reaction of a material tested in
several conditions. This would finally give something

Ckimi

T(testing condition) = -

ZkT

where T(te8ting condition) =the toxicity index proposed,


mi = the total mortality at time i and ki = penalization
factor, decreasing as i increases.
With this formula T= 0 in the best possible situation
when all animals are alive after 30 min, and T= 100 in the
worst possible case when all animals are dead after 6 rnin
exposure.
Several possibilities for ki values have been tried on
a series of results. As the toxicity hazard is closely connected with time the kr values were used in correlation
with the times at which the mortality rates are measured
during the experiment. Since the duration of the test is
30 min, the first 6 rnin being the most important, it was
decided to take kt = = 5.
Pursuing the same correlation of moment of measurement with total test duration the other ki values would be:
:
&
Q

8 FIRE AND MATERIALS, VOL. 2. NO. 1,1978

1 Time ( m m )
k , Values

Figure 2. An example of the possibility of establishing classes


using the proposed toxicity index.

CALCULATION OF A TOXICITY INDEX FOR MATERIALS, BASED ON A BIOLOGICAL EVALUATION METHOD

like the maximum toxicity index proposed by Tsuchiya


et al., but on the basis of chemical analysis.6

CALCULATED TOXICITY INDICES FOR


MATERIALS
In this section actual results and corresponding calculated
toxicity indices are presented.
Twenty materials were tested, consisting of plastics
(synthetic), five textiles (both natural and synthetic)
and seven wood products. These materials are listed in
Table 1.

Table 1. List of Materials


Identification
No.
1

2
3
4

5
6

8
9
10
11

12
13
14

15
16
17

18
19
20

Specification of material
(d=thickness in mm: D=density in kg m-3)
Normal glass fibre reinforced polyester (d= 1 . 4 ;
D= 1429)
Wall cover: flexible PVC on asbestos felt
Polyester sheet
Tufted carpet, polypropylene backed with latex
Natural beech wood
Furniture tissue, 100% clevyl type reps350 g m-2
Polyester sheet with mineral filling-treated
to
become fire retardant
Self-extinguishing glass fibre reinforced polyester (d= 1 . 6 ; D= 1312)
Plywood, light tropical wood (d=18.0)
Wall cover: 100% flax tissue-325 g m-2,
coated with acryl to facilitate application
Wood fibre board glued with ureum-formol
glue, covered with an asbestos sheet (1-2 mm)
also glued with ureum-formol glue (d= 1 8 . 0 :
D= 600)
Plywood poplar, impregnated (d= 5-6)
Electrical tubing : polyethylene, treated t o
become fire retardant
Impregnated wood fibre board glued with
ureum-formol glue, oak veneer (d= 18.0 ;
D = 600)
Red fir ( d = 1 8 . 0 )
Wall cover: 100% flax tissue-325 g m-2
Impregnated wood fibre board glued with
ureum-formol glue, covered with a glued sheet
of paper as support for paint (d=18.0: D=600)
Polystyrene sheet, treated to become fire
retardant
Plastified PVC cables (d= 3 . 5 ; D= 1400)
Furniture tissue : 100% acryl type jacquard270 g m-2

Figure 3 represents the ranking of the materials by


their increasing calculated toxicity index and within the
classes proposed previously.
In previous experiments1. 2, the dilution conditions
were kept constant and the temperatures varied (approx.
400,600 and 800C). This research work has shown2 that
approx. 600C is about the most dangerous temperature
for the majority of materials. There are exceptions,
however, and thus it is indicated in Fig. 3 when a temperature other than approx. 600C, giving a greater

Material No

Figure 3. Ranking and classification of twenty materials by their


increasing calculated toxicity index.

toxicity index, has been taken into account. It must be


emphasized that this highest toxicity is not necessarily
a maximumvalue. Taking material No. 13 as an example,
the toxicity indices obtained are as follows:
T(approx. 400C) = 0
T(approx. 600C) = 30.93
T(approx.

S0O0C)=24.20

1
Time ( m i n )

Figure 4. Mortality rates with time scale of the materials (materials


no. 1 and 2 did not give rise to any deaths i n the three experimental conditions depicted).

FIRE AND MATERIALS, VOL. 2, NO. 1,1978 9

C. HERPOL AND P. VANDEVELDE

The highest value was obtained at approx. 600C but it


is still possible that an even higher value could be found
in a different combustion condition as, for instance,
temperatures of approx. 500 or 700C. This again
emphasizes the great difficulty faced by researchers in
this field and the limitations of each approach.
Figure 4 gives the mortality rate vs time values corresponding to the calculated toxicity indices of the 20
materials. It is clear that the proposed procedure effectively penalizes materials giving rise to earlier deaths as
compared with materials giving higher mortality rates at
later intervals during the experiment : thus materials
No. 6 and 7 are penalized as against material No. 5 and
materials No. 9 and 10 as against material No. 8.
CONCLUSION

The problem of the toxicity of materials is now generally


recognized to be not only very important, but also very
complex and very difficult to approach.

Bearing in mind the inevitable limitations of the smallscale approach to fire problems generally, it is thought
that the toxicity index proposed in this paper could be a
valuable method of ranking materials from the point of
view of the relative potential toxicity of their evolved
fire products.
It is obvious that the use of more or less dangerous
materials in buildings cannot be avoided and that safety
solutions will depend greatly upon the use made of the
materials and on the design of the buildings as far as
prevention, detection, fire-fighting and escape possibilities
are concerned.
Moreover, it is again clear from the results presented
that no category of materials should apriori be considered
more toxic than others, as is sometimes the case with
public opinion where new-to
be understood as
synthetic-materials
are concerned. The materials
which gave rise to no mortality at all in the tests reported
here were all synthetic. By contrast, however, one of the
most dangerous materials was also synthetic.

REFERENCES
1. C. Herpol and R. Minne, Combust, Sci. Technol. 12, 229
(1976).
2. C. Herpol, Fire Mat. 1, 29 (1976).
3. P. C. Bowes, Casualties Attributed to Toxic Gas and Smoke
at Fires. A Survey of Statistics, Fire Research Note No. 1025.
Fire Research Station Borehamwood, England (1975).
4. S. E. Chandler and R. Baldwin, Fire Mater. 1, 76 (1976).
5. J. G. Gaume, P. Bartek and H. J. Rostami, Aerosp. Med.
September, 987 (1971).
6. Y. Tsuchiya and K. Sumi, J. Fire Flammability 3, 46 (1972).
7. R. Truhaut, C. Boudene et J. M. Jouany. Arch. Mal. Prof.
Hyg. Toxicol. lnd. 36 (12), 707 (1975).

10 FIRE AND MATERIALS, VOL. 2, NO. 1,1978

8. R. Minne, The Belgian View on Reaction to Fire of Materials,


ASTM S. T. P. 502. American Society for Testing and Materials,
Philadelphia (1971).
9. C. Herpol, R. Minne and E. Van Outryve, Combust Sci.
Technol. 12, 217 (1976).

Received 8 April 1977

0 Heyden t Son Ltd, 1978

Das könnte Ihnen auch gefallen