Critical Ignition Temperatures

of Wood Sawdusts

TAKASHI KOTOYORI
Research Institute of Industrial Safety, Ministry of Labor
5-31-1 Shiba, Minatoku, Tokyo, 108 Japan

ABSTRACT

Critical ignition temperatures were estimated for fifteen species of wood

sawdust slabs, being stacked in air at normal pressure, with an adiabatic self-
ignition testing apparatus following a procedure. Assuming each stack to be in
the form of an infinite slab of a thickness of 60.96 cm (2 feet), the temperature
is estimated to range from 118 for Zelkova to 142 °c for Sitka spruce. These
values are in reasonable agreement with a few real data observed on wood sawdust
stacks of similar sizes. Zelkova and Western red cedar are relatively easy to
ignite, meanwhile Sitka spruce and Western hemlock are relatively hard to ignite.
Douglas fir and Port Orford cedar are medium in ignitability.

KEY WORDS: Critical ignition temperature; wood sawdust; self-ignition; adiabatic

oxidation reaction; thermal diffusivity; relative ignitability.

INTRODUCTION

Many research reports have appeared on various aspects of pyrolysis and

combustion phenomena of wood or woody materials. Instances, however, have been
'only a few so far where the critical ignition temperatures of bulky stacks of wood
sawdusts were actually measured. As for that point, the works by Gross et al.[l],
by Akita[2], and lately by Anthony et al. [3], by John[4] and by Schliemann[5] are
quite precious.

In the present work, critical ignition temperatures of fifteen species of

wood sawdust slabs, assuming each to be stacked in hot air surroundings at normal
pressure, were estimated with an adiabatic self-ignition testing apparatus
following a procedure.

A detailed description of the apparatus was given elsewhere[6]. So the

procedure to estimate the critical ignition temperature of any stack of self-
heating substance and results when the procedure was applied to wood sawdust slabs
are mainly described in this report.

Wood species tested are as follows: Telaling (a Southeast Asian lumber),

Meranti (ditto), Sawara cedar, Japanese cedar, Japanese red pine, Japanese
cypress, Zelkova, Paulownia, Western red cedar, Douglas fir, Port Orford cedar,
Alaska yellow cedar, Western hemlock, Sitka spruce and Eli ayanskya (a Siberian
lumber) •

AN EQUATION HOLDING FOR THE SELF-HEATING PROCESS UNDER AN ADIABATIC ZEROTH ORDER

FIRE SAFETY SCIENCE-PROCEEDINGS OF THE FIRST INTERNATIONAL SYMPOSIUM 463

Copyright © International Association for Fire Safety Science

ASSUMPTION

When a sample heats spontaneously at a temperature T, eqn. (1) holds

generally under an adiabatic zeroth order assumption, based on the principle of
energy conservation.
dT E
cPcit = 6H·A exp[- RT] (1)

If we assume that the temperature rise rate, dT/dt, remains effectively

constant within small temperature ranges near T, we can easily integrate eqn. (1)
and get the following equation after taking natural logarithms on both sides of an
equation obtained by the definite integral of eqn. (1).

(2)

where 6t is the time required for the sample temperature to rise from T to
(T + 6T). Equation (2) can be applied to the adiabatic self-heating curve
recorded precisely, by an adiabatic self-heating process recorder such as the
apparatus used here, in temperature ranges of a few degrees above T. Then the
gradient, a, and intercept, b, of an empirical formula, In 6t = (a/T) + b, can be
expressed as,
E
a="R (3)

b = 1 [6T'c P] (4)
n 6H.A
respectively.

A RELATION DERIVED FROM FRANK-KAMENETSKII'S CRITICAL CONDITION FOR THERMAL

IGNITION AND EQUATION (2)

Frank-Kamenetskii's critical parameter 0c for thermal ignition is expressed

as,
2A
6H'Er .!L.]
oc 2
exp [-
RT
(5)
ART
c
Frank-Kamenetskii's critical condition for thermal ignition holds on a
balance between the heating process, again based on a zeroth order reaction, and
the heat loss process under non-adiabatic conditions. Incidentally eqn. (2) is
concerned with the heating process alone, since it stands on the adiabatic
conditions. So far as the heating process is concerned, however, both equations
deal with the identical slow heating process in the subcritical state prior to
ignition. Therefore it can be thought that factors such as E, A and 6H
appearing in eqn. (2) and corresponding those in eqn. (5) represent the same
physical quantities, respectively.

Taking natural logarithms on both sides of eqn. (5) and rearranging them
gives,

In T + (E/R)
2T
= In 11 [(6H'A)(E/R)]
r + "2 n 0'1.
(6)
c
c c
substituting eqns. (3) and (4) in eqn. (6) for (E/R) and In(6H·A) gives,

In T + ~ = In r + i[ln(a'~T:~p) - b] (7)
c 2T
c c

464
since thermal conductivity is expressed in terms of thermal diffusivity and heat
capacity per unit volume, as in eqn. (8),

A= Cl'Cp (8)

eventually we obtain,
a
In Tc + 2T (9)
c
We see from eqn. (9) that we may dispense with direct knowledge of the
individual values of 6H, E, A, A and c of the substance, so far as the estimation
of T is in question. The only data required to estimate T value is the thermal
diffhsivity associated with the material, besides data, obt~ined with the
apparatus, such as a, band 6T. values of rand 0 can be assigned as desired.
Methods of determining thermal diffusivity for mat~rials are well established[7].
As a matter of course the packing density p of the substance in the sample cell
of the self-ignition testing apparatus and of the thermal diffusivity measuring
apparatus should be as equal as possible to that in the practical conditions.

STARTING PROCEDURE

A block diagram of the apparatus used here is shown in Fig. 1. Measuring

procedure for oxidatively heating substances has been described elsewhere[6], but
a part of it was modified in this work because of the nature of wood sawdust.

300 mg of wood sawdust sample is packed in the cell that has a hole at the
bottom. Next the thermocouple for measuring sample temperature is inserted into

THERMOCOUPLE

OUT

INERT
GAS

OXYG~9 iI
I

: TLIMIT SW
i r:----i
: 'PRIN·'
-------1TER
I.-
!
...J

INERT
GAS TI f--J2<J--'--'--' AIR BATH
CONTROLLER
~ _ - ~ ~

FIGURE 1. Block diagram of the adiabatic self-ignition testing apparatus. 1,

Thermometer; 2, glass wool; 3, sample cell assembly; 4, silica tube; 5, glass
wool; 6, wind guide plate; 7, fan; 8, heater.

465
the sample. Then the sample cell assembly is set in the adiabatic jacket, and
nitrogen gas is supplied at a rate of 2[ml/min]. The initial starting temperature
is selected as desired by setting a temperature dial on the air bath. Then the
power is switched on, that is, bath heating is started. After that the apparatus
is left for 270 minutes at the starting temperature in nitrogen atmosphere, to
remove all of the moisture from wood sawdust and to establish thermal equilibrium.
While moisture-removal process lasts, the measurement can not be started, since
the endothermic effect due to vaporization of moisture masks the exothermic
phenomenon due to oxidation reaction of the substrate. Then, zero-suppression
procedure is carried out, and immediately after that, the supply of air at a rate
of 2[ml/min], adiabatic control and recording of the heating process are
commenced.

ESTIMATION PROCEDURE OF CRITICAL IGNITION TEMPERATURE OF WOOD SAWDUST

Wood strips are first prepared by cutting some 2.5 mm thick slabs from
visually uniform pieces of wood block of each species using a band saw. The
strips are next cut into some 2.5 rom by 2.5 rom by 2.5 mm chips with a nipper.
Chips are then pulverized using an ultra centrifugal mill and size-graded to 35 to
60 mesh by sieving.

Adiabatic oxidation heating curves of each species of wood sawdust are

recorded at starting temperatures ranging from 150 to 178°C, following the
starting procedure stated above. The curves for Port Orford cedar sawdust are
shown in Fig. 2 as an example. In these cases it is seen that the rate of
adiabatic oxidation reaction is gradually accelerated, as nitrogen atmosphere

178

-
~ 176
~
oS 174
E
"ol
~172
~

o 50 100
time (min)
FIGURE 2. Adiabatic oxidation heating curves of Port Orford cedar sawdust.

466
 200

.- , ,:
e190 .
• e"
~ .:~
oS ....,..
p ,.
,.... .....
~ 180 soc .
~ . ---: \1':>' ,
;;;: f;"-- •...
~ /, .... . .
~ .' ,

~ 170-
~

50 100 200 300 400 500

time (min)
FIGURE 3. Adiabatic oxidation heating curves of Western red cedar sawdust
with wider temperature increments.

around and in the sample cell is replaced by air for about 30 minutes after the
start.

Adiabatic oxidation heating curves of wood sawdusts recorded with this

apparatus appear in most cases approximately linear. For instance, a curve is
almost linear at temperature levels of 160 to 175 °e in the case of Western red
cedar sawdust, though the temperature rise rate begins to accelerate gradually at
levels in excess of 180 °e, as shown in Fig. 3. The same thing has been reported
also by Akita[2], by Anthony et al.[3] and by Schliemann[5]. Therefore a decision
was made in this work that the gradient, ~T/~t, of the linearly rising part of a
curve was used as a physical quantity corresponding to the rate of the oxidatively
heating reaction of wood sawdust.

The time, ~t min, required for the sample temperature to rise by a ~T value
of 1.25 K is determined on each heating curve. Then In ~t vs. liT plot is made
for each species. The entire plots for fifteen species of wood sawdust are given
in Fig. 4. We can see roughly from Fig. 4 the relative liability of these wood
sawduststo heat oxidatively, or to ignite ultimately.

Values of coefficients, a and b, of eqn. (2) are determined from these plots
by the least-squares method with a computer and are listed in TABLE 1.

467
 175 170 165 160°C
I
mm

if *"//;!Jj;~ II. -.
50
~~0 ~ 7'1
/0 I
3.sl /f ~

~
iVe7
,-
*0
~40

3.6~ II I o'1}; /1 / I f J/ /
f
"'~

II ",~ .~(/ W
f'-.,~

-+-00

/-
<i)

/ 11!1: // ~.fj
,j::oo
al

/
CO '::]

.....= /
EO -130

/
EO

2.22 2.24 2.26

I
/
I

2.28
/
» I

2.30 2.32
3
10 / T
FIGURE 4. The entire In 6t vs. liT plots for fifteen species of wood sawdust.
 TABLE 1 Adiabatic experimental parameters of slow oxidation reaction and critical ignition
temperatures of 15 species of wood sawdust, when each stacked in a slab, 60.96 cm thick.

Wood species a b a E p T
c

3
Zelkova 11 ,899 -23.838 0.14 99 0.279 g/ cm 118°c
Telaling 13,736 -28.379 0.15 114 0.214 119
Western red cedar 12,487 -24.936 0.16 104 0.153 124
Sawara cedar 12,536 -24.994 0.16 104 0.173 125
Douglas fir 12,430 -24.474 0.14 103 0.197 127
Meranti 15,998 -32.881 0.15 133 0.185 130
~ Port Orford cedar 13,906 -27.950 0.17 116 0.138 131
0)
CD
Japanese cedar 13,944 -27 .994 0.18 116 0.145 132
Paulownia 13,110 -25.887 0.17 109 0.173 133
Japanese red pine 16,483 -33.656 0.13 137 0.253 133
Eli ayanskya 14,478 -28.882 0.16 120 0.138 135
Alaska yellow cedar 16,124 -32.765 0.17 134 0.138 136
Japanese cypress 17,644 -36.123 0.16 147 0.131 138
Western hemlock 16,505 -33.267 0.17 137 0.131 141
Sitka spruce 17,092 -34.517 0.17 142 0.131 142

Reference data
Mixed hardwood sawdust[3], when stacked in a 61 cm cube, but in cool surroundings, 0.178 135 (ob s)
Kiefernholz (German pine) sawdust [4, 5], when stacked in a sphere, 1 m diameter, 0.125 130 (obs)
 The a value of wood sawdust was measured at room temperature and at 100°C
with the stepwise heating method developed by Araki et al.[7]. Values at both
temperatures were much the same for each species, so the value is given in TABLE
as an arithmetical mean, respectively.

The form of the stack was assumed to be an infinite slab, which has a 0c
value of 0.88, of a thickness of 60.96 cm (2 feet), which is equal to 2r.

Substituting these data in eqn. (9), T values are obtained and are presented
in TABLE 1. Wood species are arranged in o~der, wood sawdust of high critical
ignition temperature being listed near the bottom.

RESULTS AND DISCUSSION

A possibility of heat loss due to air flow in the sample cell could not be
ignored, though it would be almost perfectly covered by the adiabatic control. If
that effect were serious, the true T value might be several degrees lower than
c
the one determined by the procedure. Furthermore there might be some uncertainty
in In bt vs. liT plots, though it would be considerably compensated by the least-
squares method. As for that problem, to increase the number of the point on each
plot from 4 or 5 to 10, e.g., may be helpful to increase the certainty of the
plot. It is also not very pleasant that we can not make the packing density
constant for each species because of the nature of wood sawdust.

For all those problems, T values of wood sawdust slabs estimated by

calculation, following the progedure introduced above, agreed in the result
reasonably with a few real data observed on stacks of similar sizes, though the
species are different from those used in this work[3, 4, 5]. It thus appears that
those problems anticipated above are practically negligible with wood sawdusts, at
least. Then we can say that the procedure is applicable to estimate relatively
and quantitatively the ignitability of any stacks of wood sawdusts.

Assuming wood sawdust to be stacked in a form of infinite slab at a thickness

of 60.96 cm (2 feet), the T value is estimated to range from 118 for Zelkova to
142°C for Sitka spruce. I~ the slab is 20.32 cm (8 inch) thick, the value is
estimated to range between 147 for Telaling to 167°C for Sitka spruce. Zelkova
and Western red cedar are relatively easy to ignite, meanwhile Sitka spruce and
Western hemlock are relatively hard to ignite. Paulownia may also belong under a
category that is hard to ignite. Douglas fir and Port Orford cedar are medium in
ignitability.

The oxidatively heating process of wood sawdusts proceeds showing an almost

linear temperature rise rate in temperature ranges up to about 180°C under
adiabatic conditions.

The relative liability of wood species to heat oxidatively is scarcely

recognized by thermal analysis such as TG-DTA, meanwhile it can be clearly
distinguished from one species to another under adiabatic conditions.

Wood is a natural product and its composition is not homogeneous, so T

c the
values listed in TABLE 1 are by no means of absolute character. Nevertheless
author thinks that, although there may exist still some uncertain factors such as
those stated at the beginning of this chapter, a procedure to estimate T values
of wood sawdust stacks has been fundamentally established. The ignitabi!ity of
any self-heating substances can be relatively and quantitatively determined by
applying the critical ignition temperature as the criterion. So it would be
interesting to see if the procedure is applicable also to other self-heating

470
substances.

SYMBOLS USED
3
A frequency factor [mol/cm min]
a coefficient expressed by eqn. (3)
b coefficient expressed by eqn. (4)
c molar heat capacity [J/mol K]
p molar density [mol/cm 3]
3
cp heat capacity per unit volume [J/cm K]
E apparent activation energy [J/mol]
6H molar heat of reaction [J/mol]
R universal gas constant [J/mol K]
r size factor [cm]
T initial starting temperature [K]
T critical ignition temperature [K]
c
t time [min]
<5 Frank-Kamenetskii's critical parameter for thermal ignition
c
thermal diffusivity [cm2/min]
thermal conductivity [J/cm min K]

REFERENCES

1. Gross, D. and Robertson, A.F.: "Self-Ignition Temperatures of Materials from

Kinetic-Reaction Data," J. Research Natl. Bur. Standards, il: 413-417, 1958.

2. 'Akita, K.: "Studies on the Mechanism of Ignition of Wood," Rep. Fire Research
Inst. Japan,~: 1-106, 1959.

3. Anthony, E. J. and Greaney, D.: "The Safety of Hot Self-Heating Materials,"

Combustion Science and Technology, ~: 79-85, 1979.

4. John, R.: "Zur SelbstentzUndung von Holz. Aufheizphase und freiwerdender

Warmestrom," in Origin and Spread of Fire (latest state of research) - No.1,
4th Internatl. Fire Protection Seminar - ZUrich, 59-84, Brand-VerhUtungs-
Dienst fUr Industrie und Gewerbe, ZUrich, 1973.

5. Schliemann, H.: "Die Anwendung der Differential-Thermoanalyse (DTA) zur

Einschatzung des SelbstentzUndungsverhaltens von Holz," Holzindustrie, 30: 5,
140-142, 1977.

6. Kotoyori, T. and Maruta, M.: "An Adiabatic Self-Ignition Testing Apparatus,"

Thermochimica Acta, ~: 35-44, 1983.

7. Araki, N.: "Measurements of Thermophysical Properties by a Stepwise Heating

Method," Internatl. J. of Thermophysics, 5: 1,53-71,1984.

471