MAGED/ LT .

0 Y

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

' Hr
N. F. Cowed' and O . W. Jones•
FLAMMABLE mixture of gases, such a s
A methane and air, may be diluted wit h
one of its constituents or with other gases
until it is no longer flammable . The dilution
limit of flammability, or simply the limit of flam-
mability, is the borderline composition ; a sligh t
change in one ;erection produces a flammabl e
mixture, in the other direction a nonflammabl e
mixture .
There are clearly two limits of flammabilit y
a lower and a higher, for each pair of so-called
combustible gases and supporters of combustion.
The lower limit corresponds to the minimu m
amount of combustible gas and the higher o r
upper limit to the maximum amount of com-
bustible gas capable of conferring flammabilit y
on the mixture . For example, for methane -
air mixtures these limits under normal condi-
tions are approximately 5 and 14 percen t
methane, respectively . Mixtures within these
limits liberate enough energy on combustio n
of any one layer to ignite the neighboring Isle'' = .amay
of unburned gas an are therefore capable' of
1,,-
self-propagation of flame ; others are not .
Mixtures richer than 14 percent methane
however, may burn on contact with external
air, for mixtures that contain less than 14 per-
; methane are formed in the zone where th e
gases mingle.
The experimental determination of limits o f
~• flammability is more difficult than may be
%`. expected, as is shown by the contradictory
s, . figures reported from tame to time . This bully.
t work on maouserlot entvpkted November IPSO .
' Late Director of Safety to Mitt Rowel+, Ministry of Fuel an d of the source of energy to iitiate flame, it i s
Poser . Great Arltsln .
t Cblef, Osseous EiPloslona Draneb . Central Erpsrimeot Motion.
Duran of Minos, Pllutwr>tb . Pa.
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
tin presents the results of a critical review of al l
figures published on the limits of flammability
of combustible gases and vapors when admixe d
with air, oxygen, or other "atmosphere. " Sus-
pended dusts and liquid mists are not consid-
ered, except in one or two instance" in whic h
direct comparison can be made with vapors .
CONDITIONS FOR PROPAGATION OF FLAME
IN MIXTURES OF GASES
SOURCE OT IGNITIO N
When a source of heat of sufficient size and
intensity is introduced into a weak mixture ,
some combustion, occurs, even when the mix-
ture is incapable of selfpropagation of flame.
This is often visible as a "cap" of flame, whic h
may be large if the source of heat is ample.
The flame cap may be fixed in relative positio n
to the source of ignition, as in a miners flam e
lamp burning in a gassy atmosphere, or ma y
become detached from thea9ut4e and float, for
limited distance in s moving atmosphere, or
travel away 2or 3 feet frotn an uutiatuti
spark or flame in a still atmosphere (236) .
Such flames aril b7t"b'rllepcopagating, as the y
art extinguished when the influence of the source
of ignition is lost . .
When a weak source of ignition is employed,
some flammable mixtures, especially those near
the limits, may not inflame . The source o f
ignition is not strong enough to be satisfactory
for the test.
As the test concerns the capability of th e
mixture to propagate flame / not the capacity

axiomatic that the limits are unaffected b y
variations in the nature and strength of th e
source of ignition . When statements are made
that limits vary according to the means o f
ignition, it is clear that the observers use d
either such strong sources of ignition that•the
caps of flame gave the appearance of general
inflammation or such weak sources that flam e
was not started in mixtures which were, in fact,
flammable. Under these conditions they wer e
determining the limits of ignitibility by th e
particular sources of ignition they used, not th e
limits of flammability of the mixture itsel f
(63, 227) .
DIRECTION OF FLAME PROPAGATION
When a source of ignition, such as an electric
spark or a flame, is introduced into a flammabl e
mixture, flame tends to travel away from th e
source in all directions . In a very large volum e
of mixture the form of the zone of combustio n
would be a spherical shell of increasing radius
were it not that the hot expanded products o f
combustion tend to rise and hence to introduc e
convection currents . Flame cannot trave l
downward when the upward movement of th e
oases due to convection, is faster than th e
speed of flame in a still mixture, as happens i n
weak mixtures near the limits of flammability .
(fence, near each limit there is a range of mix-
tures that will propagate flame upward but no t
downward . These may correctly be terme d
'flammable mixtures," as it is only necessar y
to ignite them near their lower confines t o
observe self-propagating flame traveling to th e
higher confines . The gentle convection curren t
set up by the flame increases the apparen t
:.peed of flame but, as far as is known, does no t
enable flame to travel when in the absence o f
convection effects it would not be propagate d
(63) . It seems correct, therefore, to observ e
upward propagating flames when defining th e
limits of flammability of ►as mixtures ; but, for
some purposes, it is desirable to know the limit s
of flammability for propagation of flame . in
other directions (33) . Such limits when de-
termined are included in the experimenta l
results given in this bulletin .
For safety in industrial operations it i s
generally wisest to consider the limits for up-
ward Propagation as the danger line, sinc e
these limits are wider than those for horizonta l
or downward propagation of flame .
DIAMETER Or VESSEL
The propagation of-flame depends upon th e
t.m ensfcr of energy from the burned to th e
neighboring unburned gas, and in a limit mix-
ture the amount of energy available for transfe r
is only just enough to maintain flame propa-
gation ; therefore, anything that reduces the
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
available energy will affect the limits . Hence,
it is necessary to make observations in vessels -
, wide enough that the effect of cooling by their
walls is negligible .
Tbo observed limits of flammability are.
almost always widened as tube diameter i s
increased, rapidly at first and more slowly
afterward, so that increase of- diameter above
5 cm. rarely shows more than a few tenths of
1 percent increase in the range of flammability ;
many examples of this may be found in th e
tabulated results in the 'following pages, bu t
there are a few notable exceptions . For
example, while the. higher limit (downward or
horizontal propagation of flume) of acetylene -
air mixtures is much increased by enlarging
the diameter of the vessel beyond 5 cm ., the
higher limits (upward propagation of flame )
of hexane and ethylene are much reduced .
Moreover, the abnormal "cool flame" in higher-
limit acetone-air mixtures appears to be abl e
to travel upward in certain mixtures in a
2 .5-cm. tube but not in 5- or 7 .5-cm . tubes.
With few exceptions, however, the general
rule holds .
LENGTH OF VESSEL
When flame travels from the open end of a
tube toward the closed end, its speed is unifor m
over a distance that depends on the composi-
tion of the mixture and the dimensions of the
tube ; the inflammation of one layer of gas re-
peats the inflammation of any other layer in
the "uniform movement" of flame . Sooner o r
later vibratory movement of the flame may re-
place the uniform movement ; but this is rarely
observed in limit mixtures . When it has hap-
pened, however, the simple expedient of bold-
ing a pad of cotton wool loosely against the
open end of the tube has suppressed any tend-
ency to vibration, without sensibly hinderin g
maintenance of constant pressure conditions .
The flame then travels throughout the tube at
uniform speed ; and variatyv, .pt length of th e
tuba provided it is long enough (say ; 4'feet)
for the initial itiipulse of the source of ignition
to be dissipated, has no effect . . on the limits
observed therein.— ,
•ln experiments with' cJdaL4l 'tubes, however
t,itsunrehalgoftubdesn
affect the results. The longer the tube the
smaller must be the pressure attained durin g
propagation of a limit flame, because in longe r
tubes the gases behind the flame have time t o
cool more before the flame reaches the end of
its journey. Schiltzenberfler (303) showed long .
that the observed limits in closed tubes are
Meted by the length of tube . In experiments
with mixtures of 10 percent hydrogen and 9 0
percent oxygen he found that there was " a
maximum length of column of gas, beyond


DETERMINLTION OF LIMIT S

which flame is propagated only a short way small changes in the oxygen content of the air .
from the spark, whilst it can be propagated for The higher limit is 'noticeably depressed by a
lesser lengths" (of column of gas) . An elab- small reduction h oxygen content, because a
orate experimental study has confirmed and correspondingly smaller amount of the corn -
extended these observations (322) . bustible gas can burn. Thus, reduction of
the oxygen content of the stir from 20.9 to 20. 6
EFFECT OF SMALL CHANGES IN ATMOSPHERIC percent depressed the higher limit ' of methane
COMPOSITION about 0 .3 percent . (See fig: 20 .)
Humidity .—Some gas mixtures are exceed -
ingly difficult to i g nite if they have. been dried ._ PRESSUR E
by long-continued contact . with phosphorus The normal variations of atmospheric press
pentoxide . Such a degree of dryness is peculiar .Sire do not appreciably affect the limits o f
to the laboratory and is . not considered in the 'flammability, as has been shown both by
present review. direct observation and by deduction from th e
Limits have sometimes been determined with Course of curves showing the variation of limit s
roughly dried mixtures in which the partial 'over much wider variations of pressure tha n
pressure of water vapor is less than 1 mm ., those of the atmosphere (14, 17, 95, 154, U5, t
sometimes with mixtures saturated with water 235, 309) .
vapor at the temperature of the laboratory. The effect of larger variations in pressure i s
The condition of humidity has not always been neither simple nor uniform but is specific fo r
stated, perhaps because it has been assumed t o each flammable mixture . So far as is known ,
be negligible . Probably the difference is a- reduction in pressure below 760 mm . generall y
preciable for most gases only in exact wor k narrows the range of flammability by raisin g
Thus, the lower limit of methane in air, in cer- the lower limit and decreasing the higher
tain comparative experiments, was found to b e limit . This change is often imperceptible fo r
5 .24 percent for a mixture dried by calciu m the first few hundred millimeters' fall in pressure
chloride and 5 .22 percent for one saturated below atmospheric, but thereafter the effect
with water vapor at, laboratory temperature . increases until at a suitably low pressure the
These are actual percentages and are equal limits coincide ; below this point no mixture
within experimental error ; but, as usually re- can propagate Dame (65, 171a, 180) .
ported from analyses calculated on a dry basis , The limiting pressure is somewhat difficul t
they appear as 5 .24 and 5 .33 percent, respec- to find, because often it is so low that th e
tively . For methane therefore, the true lowe r difficulty of insuring a powerful enough source
limit is not appreciably affected by the replace- of ignition has not certainly been overcome .
ment of about 2 percent of air by an equa l For example, in electrolytic gas a flame tha t +ii
volume of water vapor . The higher limit, how - filled a 570-cc . globe has been produced at
ever, is reduced by such a replacement, becaus e 5 mm . pressure, and with the same mixtt:e
the oxygen content, which is less than th e flame traveled through a cylinder 2 meter s
amount required to burn the methane com- long at 8 mm . pressure (66) . At such pressure s
pletely, is thereby reduced . For example, th e the electric discharge used to test the flam-
corresponding higher limits for methane in dry mability was diffuse and if made stronge r
and in saturated air are 14 .02 and 13.54 per- might have produced , self, propagating flames
cent, respectively ; expressed on a dry basis (a s at still lower pressures . For this reason the
reported analytically) these figures become rate at which the lower- and ' higher-limit
14 .02 and 13 .80 percent, respectively (69) . curves approach one another as the pressur e
Carbon monoxide shows greater differences is decreased will appear to dtpend on th e
than methane . Tire lower limit of carho u strength of the source of ignition ; unless the
monoxide in air saturated with water vapor;, source Is carefully madefli strong enough to
at laboratory temperature was 13 .1 percent ' insure ignition and ' the vessel used is largo -
for upward propagation of flame in a 2-inch enough for the flame to travel such a distance
tube ; if the gases were dried by passage over from the source as will enable 'observers to see
calcium chloride, the lower limit became 15.9 whether the Inixtui'irw .capable of self-propa -
percent (70) . Such a large difference is escep- gation of flame . This has not been done, so
tional, perhaps unique, as moderate drying the course of limit curves at very low pressures
of carbon monoxide-air mixtures affects their is unknown . In attempting to discover th e
ignition temperatures and flame speeds more course of such curves, the difficulty of main-
than it affects these properties of any other taining constant pressure during the inflam -
gas hitherto examined . Illation will have to be surmounted . However,
Oxygen Content—The lower limit of meth- it is almost certain that whatever may be th e
ane-air mixtures and of some, probably all, exact course of the curves they do approac h
other mixtures is not, appreciably affected by and ultimrtely meet as the pressure is decreased .

t'

et:
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


• Increase of pressure above that of the at- prop ation of flame, The argument was sup- '
moapbere does not always widen the limits, ported by the discovery that for horizontal• .
On the contrary,, the range of flanunability of propagation, in which the flames could not b e
some mixtures is narrowed by increase of • retarded by convection ; and. might even Ile ' i
pressure, so that a mixture that can propagate - assisted, the lower limit of Methane in air re-'
Game at atmospheric pressure may not be able mained unaltered between 760 -and 5,000 mm .
to do so at higher pressures . For such mix- . pressure, and the higher limit was increase d
lures the lower limit may pass through a mini- even more than for downward propagatio n
mum and the higher limit through a maximum (t77) . '
at pressures (not necessarily the same) equal With only one exception known to the writers •
to or greater than that of the atmosphere (S77, • (104), all measurements of limits nit pressures
324) .. Details are given in the paragraphs on above or below atmospheric have been made in
the effect of pressure on the individual mix- closed vessels, hence the results relate to pres-
tures, more especially those of air with hydrogen , sures that increased at an unknown rate and
carbon monoxide, the simpler paraffin hydro- to an unknown amount during the experiment .
carbons, natural gas, and ethylene . The rate and amount would vary with the size
Attention may be called to the observation s and shape of the container.
that the range of flammability of hydrogen-ai r The same criticism applies to experiments i n
and carbon monoxide-air mixtures (downwar d closed vessels when the pressure is initially
propagation of flame) is narrowed at both atmospheric ; the results obtained are a functio n
limby moderato increases of pressure abov e of the length of the vessels used, as this governs ,
atmospheric and that under the same condition s in part, the rate and amount of pressure
the range of each of the gases of the paraffi n development (63) .
series, with air, is narrowed at the lower-limi t
side (by moderate increase of pressure beyon d TEMPERATURE
a certain critical value) but widened at th e
higher-limit side . It has been suggested that To propagate flame, the layer of unburned
the loss of beat by convection is greater at hig h gas next to the burning layer must be brough t
pressures than at low, as density differences ar e to such a temperature that it will "burst int o
greater at high pressures, therefore that th e flame" rapidly . If the unburned gas is alread y
range should be narrowed at both limits whe n at a temperature above that of the laboratory ,
flame is traveling downward . This would ex - less beet has to be supplied from the burnin g
plain the results for hydrogen (both limits , layer ; therefore the lower limit should be de-
moderate pressures), carbon monoxide (both creased by increase of initial temperature an d
limits, . all pressures) and the paraffin gase s the higher limit should be increased . In other
(low mit only, moderate pressures) . Super - words, the range of flammability should be
,- . p. .,ed on this effect is another, however . Ac- widened at both limits by increase of temper-
cording to the law of mass action, and assumin g ature .
complete oxidation in one stage, the rate o f Experiment has shown that, for most mixtures ,
reaction at constant temperature is greatest i o there is a straight-line relation between the limi t
methane-air mixtures when 33 .3 percent meth- of flammability and the initial temperature o f
ane is present . The mass-action effect be- the mixture (358) . This relation was obscured ,
comes a more important factor as the pressure in earlier experiments by errors due to slo w
is raised and hence tends to raise the highe r combustion, which changed the composition o f
limit in methane-air mixtures with increase of the mixture while it was in the heated vesse l
pressure . A similar explanation may be given and before it. was tested by sparking . The .
for higher-limit curves of the other paraffin same error presumably explains .,tl►oeibresults ..
hydrocarbons. For hydrogen and carbon mat- +in whicb•a widening of the-range of norms . .
oxide, however, the'mass-action law indicates batty was not observed . v
that the 66 .7-percent ' mixture has the greatest Ordinary variations of laboratory tempera-
rate of reaction at constant temperature . As ture have no . spprecfablti'setEeckon limits of
the higher limit of each of these gases is above flammability, . "
66.7 percent, the effect. of mass action would Npg U gt .
not tend to reverse the effect of convectio n
(277) . The more recently discovered rise in Few observations have been made on th e
the higher limit of hydrogen at pressures from effect of turbulence on limits of flammabilit y
10 to 220 atmospheres calls for explanation but it has been shown that the lower limits o f
along other lines (14, 17) . methane and ethane in air are reduced some -
The foregoing argument on the influenco of what by a suitable amount of turbulence pro -
convection on the limits of flammability was duced either by a fan or by stream movemen t
applied to results of experiments on downward of the mixture and that the range of flamma-

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

bility of ether air mixtures is somewhat widene d
• by stream movements .
Turbulence caused by fans or by the detona-
tion of explosives affects the lower limit o f
natural gas somewhat . (See the paragraphs
that deal with these mixtures . )
LIMITS IN AIR COMPARED WITH =TS IN OXYGE N
In general, the lower limit of a gas is nearly
the same in oxygen as in air, but very few . . of oxygen that will support inflammation o f
strictly comparable results are available t o ' the nuxture or the maximum,. permissible i f
show whether small differences exist. Th e
lower limit of methane, observed with centra l
ignition in a globe or with downward propaga-
tion of flame in a tube; is rather higher in oxyge n
than in air (267, 271), presumably because th e
mean molecular heat of oxygen is higher than
that of nitrogen between room temperatur e
and that of a limit flame, say, 1,2000 to 1,400 0
C . For upward or horizontal propagation th e
lower limit of methane is slightly less in oxyge n
than in air ; the ammonia limit is markedly less
(271) . The higher limits of all flammable gases
are much greater in oxygen than in air ; henc e
the range of flammability is always greater i n
oxygen .
EFFECT OF CHEMICALLY INERT SUBSTANCES
The addition of increasing amounts of a
chemically inert substance to the atmosphere
causes the limits of a gas to approach and ulti-
mately to meet. Many examples can be' foun d
in the following pages, but the most complet e
series available is for methane limits, as illus-
trated in figures 24 and 29 . The extinctive
efficacies of the five diluents, carbon tetra-
chloride, carbon dioxide, nitrogen, helium, an d
argon, decrease in the order given, that is
CCl.>CO,>N 1>He>A.
The experiments giving these results wer e ; limit
carried out in tubes of 5 cm . diameter. In tubes
of 2 .2 cm . diameter the order of the last four
diluents, for several combustibles, is (341) :
In still narrower tubes, 1 .7 and 1 .6 cim . diameter, mixtures, if mixed in any proportions, give rise
the order is. (194, 286) :
fle>CO,>N,>A.
It will be seen that the order CO,>N,>A i s
the same in all experiments, but that heliu m
tends to rise in efficacy as the diameter of th e
tubs. in which the experiments are made de -
creases. The effects of carbon tetrachloride
carbon dioxide, water vapor, nitrogen, an d
argon correspond closely with theirs specific.
beats, but the very high thermal conductivity
of helium tends to give it an abnormal positio n
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
in the series ; this effect is enhanced in the
narrower tubes because of the greater readines s
with which helium conducts heat from th e
gases to the walls of the tube thus cool ing the
flame in this way more than do the other'gasea
named (94) . '
From curves showing the effect of an inert
diluent on the limits of flammable gas in air, -
it is easy to deduce (I) the minimum percentage .
it is desired to make a nonflammable mixture,
(2) the minimum percentage .of inert diluent '
that must be mixed with a combustible gas
if it is desired that none of its possible mixture s
with air can propagate flame . Many examples
are given later in the text. A graphical metho d
of deducing such results has been put forwar d
by Burgoyne and Williams-Lein (39), wit h
tabulated results . From these, with the add-
itional knowledge of the limits in air, it i s
possible to make a more or less accurate re -
construction of the flammability-limit curves .
MATURES OF FLAMMABLE GASES AND YAPOR S
A simple formula, of additive character ,
was advanced by Le Chatelier (218) to connec t
the lower limits of two gases with the lowe r
limit of any mixture of them . It is
N,+
in which N, and N, are the lower limits in air
for each combustible gas separately and n ,
and n are the percentages of each gas in any
lower-limit mixture of the two in air .
The formula indicates that, for example, s -
mixture of air, carbon monoxide, and hydroge n
which contains one-quarter of the amount ok
carbon monoxide and three-quarters of th e
amount of hydrogen necessary to form a . lower-
mixture, will be a lower-limit mixture . '
it the formula is true, the lower limits of flam-
mability form a series of flanunability equiv-
alents for the individual gases of a .mixture
(64) .
-The formula 'also indicates that lower-I' '
- to mixtures that are also at their lower limits •
or, vice versa, the ^fbte .s4J~a,, mdy badeduced
from the latter statement as a postulate (84).
The formula may be generalized to appl y
to any number of combustible gases, thus ;
n, n ,
and, so far as it expresses experimental results
truly, may be applied to higher-limit mixtures ,
with the appropriate rewording of the defi-

A small algebraic transformation gives a
more useful formula for calculating the limit s
of any mixture of combustible gases that obey s
it r64), as follows :
l0U and carbon monoxide, in a wide range of
L"
. mixtures of air, nitrogen, and carbon dioxide ,
in which L is the limit of the mixture of com-
bustible gases and pt, pa, pp1 • are the pro-
portions of each combustible gas present in the
original mixture, free from sir and inert gases,
so that
pr'1' p t4- p i 1 . . . .'~ 1110. '
An example of the use of the formula will
make its application clear : The lower limit o f
a "natural gas" of the compositio n
"rifle pee*
Mel:bane 80 (loner limit, 5.3 )
Ethane __________"l5 (lower limit, 3 .22 )
Propane 4 (lower limit, 2.37)
Butane 1 (lower limit, 1 .86 )
is given by
The accuracy of the formula has been tested
carefully for many mixtures . The results are
discussed separately in the appropriate section s
later. In general, it may be sail that, whil e
the formula is often correct or very nearly so ,
there are some marked exceptions . It seem s
thee the limits (lower and hi g her) of mixtures
of hydrogen, carbon monoxide, and methane ,
taken two at a time or all together, and of wate r
L gas and coal gas may be calculated with
' approximate accuracy (3.f) . The same is true
for mixtures of the simpler paraffin hydro -
carbons, including "natural gas" (74) . Some -
times, however, the ditlerences between calcu -
lated and observed values are very large ; for
examples, see figures 56 and 57 . Many of th e
greater discrepancies are found with upward -
propagating flames, especially when one of th e
,constituents is a vapor, such as ether or acetone ,
i papeble of giving rise to the phenomenon know n
as a "cool flame" (355) . Le Cliatelier•'s law is
useful when its applicability lies been proved ,
' I but it should not be applied indiscriminately.
Limits of Industrial Mixtures Containing
Hydrogen, Carbon Monoxide, Methane, Nitro-
gen, Carbon Dioxide, and Perhaps Air.—An ex-
tension of the law to apply to other atmosphere s
than air (95, y7!, 272) is that, when limi t
mixtures are mixed, the result is a limit mixture,
provided that all constituent mixtures are of th e
same- type ; that is, all are lower-limit mixtures
or all . are higher-limit mixtures . This law holds,
for example, for methane in a tango of oxygen-
nitrogen mixtures and in ar-carbon dioxide,
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
air-argon, and air-helium mixtures, except nea r
the point at which lower and bighe ; limits
meet where the proportion of inert gas is large .
It holds also for mixtures of hydrogen, methane ,
and may therefore be used to calculate th e
limits of flammability of mine-fire gas mixtures
and of the atmospheres after a mine explosion ,
of blast-furnace gas, of automobile-exhaust as,
and of the gases from solid explosives (133, 3
A brief account of the method of calculating
limits' of complex industrial gases, such a s
those just mentioned, follows. Greater detail
will be found in the original account 133) .
The chief gases in these mixtures are hydrogen ,
carbon monoxide, methane, nitrogen, carbo n
dioxide, and oxygen. The procedure is as
follows :
1. The composition of the mixture is firs t
recalculated on an air-free basis ; the amoun t
of each gas is expressed as a percentage of th e
total air-free mixture.
2. A somewhat arbitrary dissection of the
air-free mixture is made into simpler mixtures ,
each of which contains only one flammable
gas and part or all of the nitrogen or carbo n
dioxide .
3. The limits of each mixture thus dissecte d
are read from tables or curves. (See figs. 1
and 2, )
4. The limits of the air-free mixture ar e
calculated from the figures for the dissecte d
mixtures obtained in (3), by means of the
equation :
loo
.+ } . . .
Lam,
NiP~ t N1
..?! }Ns
where p,, p,, p, - . . are the proportions of th e
dissected mixtures, in percentages and NI, Ns ,
and Ns . . . are their respective limits.
5 . From the limits of the air-free complex
mixture thus obtained the limits of the 9rigina l
complex mixture are deduced .
The following iss an example of the calcula-
tion applied to a mine-fire atmosphere . I t
contained : '
• .~i„tndem ha Air frea,
Caoatltawt
M- siri+i:
tent percent amt pewees
Carbon sinsW& . . . : r1 0 1S 0 Mdhana- 10 .la
Ot reea 1 a m ♦0 1 r
Cot boo mermaids- a) a0 ti _ .. ._« 7a0 0la
1 . The composition on an air-free basis, also
given above, is found thus : '
The •amount of air in the mixture is 2 .8 X
100/20,9==13 .4 percent. - The air-free mixture
is therefore 86.6 percent of the whole, Whe n
the original amounts of carbon dioxide, carbo n
monoxide, methane, and hydrogen are divide d


". .
DETERMINATION OF LEWI S

80
MMMIMMIIMMMIOMMHNMIMMMMHIIMMIIMIMMMMMIMM M 5 tF ; . t
IM-M.'EtIMIEMMMMMEMSM1111MIMIMOMOMMMP-2
M 1 .i
g 9l.,OMMMIMO AMMOMIMMOMMMIMMIMMMIMMIMiMMI M 6
72

64
/hisie.wrAMEMMIMMMMMIMMMIMMMIMMMMUMME 7
11IMMOMMIMMMM H MIlmililCEMMMMUMPLUMMMMMM 8
2-+CO2
MMMMIUYAMMM 1HIMMINRAIFIMMIPT'dMIMMMIMM11P 9 irtol
MMMMMMMMIMMMMUMP2MMO r iaMMOMMMMIMMM 10 wra.z
56

MIMMMMMMIMPIdIMMIed MMMMOMMMMIIM M
MIM AV MIMMMMIMMINAMIMMMMIMMIUMMINHIMIMI M 11 al
48

40 M MIM0 122NIMUSIMMMMIMMMMIMMMMIMMM M
MI X4 MP' N1110M21IMMIMMIMIMMMMMMMIMMI M 13 3
MI MR rapMEMMUMMIMMMOMMMMMUMMIM P 14
32
M1MM P''Pwdll MOMMMIMMMUMIMMIMMMMM I 15 6
24 overim
wAa / praptd,lig m mmmmmummsomm m
prammmummonumninuum s 16 Q .1 t
16 , AtIMMOMIMIMIMMIMMMMIMMUMMUMMMMMI R
,:4sivommommommosmmmmmmmm m 18 .. . .. Z1
8 19
%iAMMIMIRIMMMMMMUMMMIMMMMMIMMMIMMMMIU M 20 ,.
MMMIMMIMIMMMIMIMMMMMIMMIIMMIMHMKMMMMMIMMHIM ,
0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
INERT GAS
RATIO:
FLAMMABLE GAS
FlattaE I .—Limits of Flammability of Hydrogen, Carbon Monoxide, and Methane Containing Various Amount s
of Carbon Dioxide and Nitrogen.

,
it •ti 1i
EEMEMENNUEEWNURNMEMP m roMM&MNMRaMENUMN WM. ,,
=MEMMRMMWAIOMMRM mmwmn s
==ZONNUIP I =M11111IMImmmi
immmmommommummusowommommmomom=awrUimmmim am
mimmoomm 0% m g mPsae .iommmmowwwwommommm mmm
MAI/AaMEWWW 0.011019MENN.WMEEN .I. MEMMI
dilmown-avplarvommommas r stismR
ak, wimatinapAs ammmirminaHmis w tv
UIOEW11U M r ' Wi
UNMWEIPOPW AMMUrSIO A
EINIUM:iW8165011TAMMEWWWNEREWEIIINIMMErMEMMMEW EI
UNEMMIM UWMUMw .98pWEMENIMI.M.E.I.11WEEMMIENOMENWIM
r4mmmm rwislommormwommmo m IIMEIOIMMOEVIUUOI
ommmawrAmmumpmArommmommummmo
111INWN11OWWNMM&IIINEWMENAMMOR AE r eNIMMUMM W
.Ol A.WUMUMENUMMMIEWE
UNWAWAMWEEwUMEMEMECIMMMMOMAMENNUMMMMMUNEI M
NIIAMIMLYAMEN.WUM180. MMEEPMAN .E .MUNUMNEWIMEMOME
wamnswoismomilmmmamolmommotmmommommmmom .
onsm''alsemmumisize g uemmmirmimm
swAmorimmmommammerommommommommmmimsmommmommi
FAIINK181119IM
m
mndwmww mmmemisr samdommoommormommmimommmmommo m
ooN AN iSMewimMUMENNUENNEERN M E l e w m E U M EWMEMIRRUMN
MOM E WMwIllEllImMEON.W .WEIR
11WW
WUmWOM;6MIIINEENIONEWNIIIENNUMMEBMWMEIN.MMIEMO11MOUNNE NWO W WOOMMINEUmEENUIR
,aNEUMMEMENIBEW .E .E.W.MWEEWEEN .mENMEmNIME . r
mmimmommmmmmmommosmarommummmimmommmmmormomo 19 19 20
q 1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 1 7 21
'INERT GAS
RAT ' FLAMMABLE GAS
Floats 2.—Limits of FlimmsbUity of Ethane,Dioxide
Ethylene,
andand Benzene Contsinlng Various Amounts of Carbo n
Nitrogen,

--

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


8 LLM1TS OF FLAMMABILITY OF GASES AND VAPORS

by 86 .6 and multiplied by 100, the "air-free" 2. The flammable gases are paired off with
percentages are obtained . The nitrogen figure the inert gases separately to give a series of 1
dissected mixtures, as shown In the followin g -
is the difference between 100 and the sum of
these percentages . table.

Ratio or Limits from fit. 1 si

Ni. 'total inert to
Combustible Percent CO, , '
percent percent percent combat-
. 4•
Obi* Lower Higher

CO 5.0 ---------- 17.5 225 .3.5 61

CH 20 9 24.7 5. 5 36 . 73,
41 . 03
3- 8 ---------- - 7& 0
b 7 {2.3 07 . .- - 31 . 2 3t 2 10. 4 50
8, .___ . . 15 9 18.6 5.9 . 32 64. 0
, r. f
Total 1t 5 15. 9 69.6 100
.0 ---_------ . . . . . . . .. . .^ »I I
.i• n

Some discrimination is needed to choose ap-
propriate quantities, but a fair latitude of
choice is usually available .
3. The limits of the dissected mixtures fro m
figure 1, are shown in the last two column s
above . For example, the fist mixture contains
5 .0 percent of carbon monoxide and 17 .5 per-
cent of nitrogen ; the ratio between its nitroge n
and carbon monoxide is 17 .5 :5 .0=3 .5 ; and th e
limits from the curve for carbon monoxide-
nitrogen mixtures are 61 percent (lower) an d
73 percent (higher) .
4. The values in the last two columns and i n
the column ""Total percent," substituted in th e
equation, Five the two limits of the air-fre e
complex fixture, calculated to 0.5 percent :
100
Lower limit=* 22 43 percent,
.5 24 .7 34 .2 18 .
61 + 36 + 50 } 32
100 ethylene, and benzene 'are available (fig. 2) ,
Higher limitee2
., .5 + 2l .7~34 .2 * 18ti =*61 percent . and these three gases have been included in
73 41 .5 ,6 64
The range of flammability of the air-free com-
plex mixture is therefore 43 to 61 percent.
5. As the air-free mixture is 86 .6 percent of- . • Safety of Industrial ?4 xtures of Gases.—If a.
the whole, the limits, in air, of the mine-fire
atmosphere are 43X100+86 .6, and 61X100 +
86 .6, or 50 and 70 percent, respectively .
The novice's difficulty .with such calcula-
tione is in stage 2, where an appropriate amoun t
of inert gas has to be chosen to pair with eac h
combustible gas in turn . The ratio of inert to
flammable gas must not be so high that the
mixture falls outside the _extreme right of th e
corresponding curve in figure I . A little prac-
tice will soon overcome-thus difficulty .
It need only be added that, if the amount of
inert gas is so great that a complete series o f
flammable mixtures cannot be dissected, th e
air-free mixture is not flammable . Dloreover ,
the air-free mixture may be flammable, but
when its limits are multiplied by the appro-
priate factor in the final stage of the calcula-
tion the result may be greater than 100 for tt
each limit ; the original mixture is then inca-
pable of forming an explosive mixture with air
because it contains too much air already .
Finally, if the lower limit of the original mixture
is less than 100 and the higher limit greater tha n
100, the mixture is flammable per ee and woul d
explode if a source of ignition were present .
About 20 examples, which cover a wide rang e
of industrial cases, have been tested (133) by
experiment . The calculated and observed lim- i3
its agree within 2 or 3 percent, excepting on e
higher-limit figure for a mixture that containe d
an unusually large amount (nearly 24 percent)
of carbon dioxide .
Extension of Mixture Law.—The combustibl e i 1
gases covered by the examples just given are r.I
hydrogen, carbon monoxide and methane .
'I he necessary data for the limits of ethane,
calculations of the limits of complex industrial
combustibles, such as coke-oven gas, coal gas ,
carbureted water gas, oil gas, and produce r
gas (136) .
complex mixture . of combustible and incom-
bustible gases, such as any of those just dis-
cussed, is fla+nmehle fer se or is capable o f .ar 1l'
forniing flammable mlxtui ti with air, the mini -
mum amount of incombustible gas that wil l
render it nonflammable and also incapable of
forming a flammable mixture with air can be
cnleislated by a small (and. justifiable)
j ustifiable) extensio n
of Le Chatelier's law procedure also
gives the maximum percentage of oxygen belo w
'which all possible mixtures are nonflammabl e
1(62), elaborated in (1)l . A somewhat elaborate
graphic method of ascertaining) the flammability
of triple mixtures of combustile gases has been
given (15) .
The purging of gas-manufacturing plants an d
distribution mains by the use of carbon dioxide

• -- er'',s.c: lee. . .. . :, *
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

DETERMINATION OF LIMITS
or automobile exhaust gas or specially prepare d
inert mixtures of low oxygen content can b e
controlled by applying similarly deduced data .
A full description is given in reference $29 .
Data for purging with steam are given in
reference 988.
SUPPRESSION OF FLAM*UABILtT T
A flammable mixture may be rendered non-
flanunable by (a) a suitable increase in th e
amount of either constituent, (b) the addition
of a suitable amount of chemically inert sub-
stance, and (c) the addition of a flammabl e
substance in sufficient amount to exceed th e
higher limit of the resultant mixture . It has
been suggested (129) that the last-named may
be advantageous and technically feasible i n
certain circumstances when the added flamma-
ble substance has a low higher limit of flamma-
bility. For example, the mixture of hydroge n
and 'air containing just enough oxygen to bur n
the h diogen completely (29 .5 percent hydro-
gen) becomes nonflammable in final mixture s
containing 86 percent of air, or 65 percent o f
:l►ydrogen, or 67 percent of nitrogen, or 10 per-
cent of methane .
CHOICE OF EXPERIMENTAL CONDITION S
In the li ght of the preceding discussion, it
seems that limits of flammability are physico-
t chemical constants (at defined temperature s
and pressures) of gases and vapors that can be
determined when observations are made with
c uiescent mixtures in vessels of large enoug h
dimensions, with ignition from below (and, i f
desired, at other points) and with maintenanc e
of constant pressure during the experiment .
• A somewhat wide experience has shown that ,
if observations are made in a vertical tube 2
inches in diameter and 4 to 6 feet in height, th e
n results are nearly : but not quite, the same a s
those obtained in much larger apparatus.
Limits observed in smaller apparatus;for .
example, tubes of 1 inch diameter or less—ar e
usually sigitifiantly narrower. Results ob-
'' lamed in small, closed tubes often differ s o
much from normal results, and even from on o
- -another, that they may be very misleading .
Mixtures for test are made in gasholders over
watec mercury, or other suitable liquid befor e
introduction into the test apparatus ; or they
are made in the apparatus itself by introducin g
the components separately and using some
mixing device ; or they are made by supplyin g
constant metered streams of the components
via some suitable mixing device.
Effective ignition usually can be obtaine d
equally well by passing an electric spark from
an induction coil (say, from "2-inch" to "12 -
. inch" as convenient) across a gap several
{. S .. http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
Source:
9
millimeters long or by drawing the flame of a
small spirit lamp or a je : of burning hydrogen e
across an aperture in the observation vesse L
This aperture is conveniently made at the
moment preceding ignition by gently sliding
away a ground-glass plate that previously ha d
sealed the vessel . When gases of small solu-
bility are tested, a water seal may be used fo r
this purpose. For a few gases a small tuf t
of guncotton fired by a spark or heated platinu m
wire is a more certain means of ignition (353,
171a) ; for a few others, an electric spark suc -
ceeds in ruing the mixture when a flame fails t'
(132) . Ignition attempted by an electricall y
heated wire may not always be .reliable with
mixtures near the limits .
Bureau of Mines Apparatus.—An apparatu s
used for many of the determinations made in
the Bureau of Mines laboratory and recorded
in the following pages is illustrated in figure 3 . i . .,
It is specially designed for determining limits
of vapors of liquids that are sufciently, vola-
tile at laboratory temperatures ; it can be simpli-
fied for gases, but must be elaborated for the
vapors of less-volatile liquids .
In figure 3, a is the glass tube in which th e
mixture is tested . Its lower end is closed by a
lightly lubricated ground-glass plate b, sealed
with mercury e . It is evacuated by a pum p
through the tube j . The vapor under test is
drawn from its liquid in the container p, i n
amount measured by the manometer k . Air
or other "atmosphere" is then admitted
through the drying tube g until atmospheric :I :1
pressure is reached . The air and vapor are the n
thoroughly mixed by circulation, by suitabl y
raising and lowering the mercury vessel g
repeatedly for 10 to 30 minutes, dependin g
upon the density of the added combustibl e
vapor. The mercury seal is then removed ; the
glass plate b is slid off the tube,, and the flam-
mability is tested almost at the same momen t :t
by sparking at y or by passing a small flam e
across the open end of the-0u bb
Special Conditions .-'---An examination of th e • :f ep
effects of temperature on limits of flammabilit y it N
requires special a,rrugementa for heating (or
cooling) the test vessZR :":41iese are not diffi-
cult to provide . To examine the effect of '
pressure is, however, much more difficul t
because of the problem of avoiding lame in-
creases of pressure during (and as a result of)
the inflammation . A few experiments have
however, been made with the "open" end of I.
the test vessel in connection with a very
much larger vessel full of air at the same pres- .
sure as that of the mixture ` under test (104) .
Most of our information about pressure effects
on limits relates only to the pressure of the
mixture before inflammation .
The limits of turbulent . mixtures have us-
?{J (J


LIMITS OF FLAMMABILITY OF GASES AND VAPORS

Induction coil
X

Motor Vacuum prying

pump. fewer

Fume 3 .-Apparatus for Determining Lindta of Flammability of Gases and Vapors .

wally been determined in small vessels fitted
with fans, and the criterion of continuer! prop-
o
r tion of flame has not been applied, *l-
ough for testing the effect of turbulence du e
to the detonation of solid explosives short
sections of a cylindrical metal gallery of 0 .3
feet diameter have been used .
Stationary Flame in a Flowing Mixture .
q uito recently (286) an apparatus has been do-
vlsed from which a cylindrical column of gaseous
1 '•1
'*
- -mixture rises into, the air in a uniform stream,
"jacketed" by a layer of nitrogen which
to maintain the strict uniformity of motion of
.t s .[l
r f ee
the mixture. Tho 001sees,tAf this device i s
shown by the fact that flat, stationary, hori-
zontal flames have been maintained in mixture s
of composition slightly below the lower limits
determined in tubes. Only two figures of limits
determined by means of this device are ye t
available (see under Propane and Butane) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


PART IV . LIMITS OF MIXED FLAMMABLE GASES AND VAPORS

Part III has covered the limits of flammabilit y Other observers obtained similar results in a
of individual flammable gases •and vapors . I n Bnnte burette 19 mm. in diameter, with down -
part IV the. limits of mixtures of two or more ward propagation of flame, using a mixture
flammable gases or vapors are given . Th e of equal volumes of hydrogen and carbon
chief question is the possibility of using Le monoxide (95, 270, 323) .
Chatcher's law of mixtures (p . 5) to calculat e The higher limit in air of a 50 :50 mixture
reliable figures for the various mixtures fro m of hydrogen and carbon monoxide, with upwar d
the ascertained figures for the individual com- propagation of flame in a tube 2 inches in
pponents . This is dealt with in detail in th e diameter and 5 feet in len gth, open at the (ring
following pa g es, but it. may be said here that, end, was 71 .8 percent, which is 0.7 percent less
in general, the law is fairly closely followed b y than the calculated value (64).
mixtures such as the common fuel gases, an d For both limits the differences between th e
that gross exceptions are observed in man y observed figures and those calculated from
mixtures that contain hcdrogcu sulfide,carbo n Le Chatelier's law just exceed the experimental
disulfide, or vapors such as ether which ma y errors ,
give rise to "cool flames. " The limits with downward propagation o f
flume in a tube 1 .6 cm . in diameter have bee n
HYDROGEN AND CARBON MONOXIDE recorded (217) .
Influence of Pressure .—The limits of a mix-
ture containing 51 percent carbon monoxid e
The lower limits of various mixtures o f and 46 to 47 percent hydrogen, in air, were
hydrogen and carbon monoxide in air, wit h 10 .4 and 6a percent, at . atmospheric pressure
upward propagation of flame in a vessel 6 fee t and 11 and 78 percent at. 800 atmospheres (10) .
high and 12 inches square in section, open a t HYDROGEN AND CARBON MONOXIDE IN OIBE R
the firing end (64), were as follows ; ATMOSPHERES
Lower limits of flammability in a large vessel Atmospheres of Composition Between Air
and Pure Oxygen .—The limits of flammability
lower Ilmit, percent . of almost equal volumes of hydroge n
Hydrgea Cerbon
In000tlde carbon monoxide in atmospheres of nitroge n
Observed Calculated DtRereaee and oxygen rangingg from air to almost pur e
I00 0 4.10 .. .. .... ...... .
oxygen have been determined with downward
73 23 4 .70 4.0 -020 propagation in a burette 19 mm . in width .
b0 30 0 03 02 — . 13
23 7$ 0 .20 63 —.10 l'he lower limit rose slightly from 12 .4 percen t
10 40 10 .00 10.4 „ +_40 of the mixture in air to 12 .6 percent in oxygen ;
0 100 1230 . .. .....
the higher limit rose from 66 .1 percent in ai r
to 92 percent in 98 per.ebtWxvgen .(3 9) .
In a small vessel 4 cm . in diameter and 25 rm . Atmospheres of Nitrous and Nitric Oxides .—
in length presumably with downward propa- The limits of mixtures of hydrogen and carbon
g
ation of flame, the lower limits were a s monoxide In nitrous oxide and in mixtures o f
follows (220) : nitrous and nil rT n ottidgs. with downward propa -
gntion of flame inn 15-num . burette (339), ar e
Lower limits of flammability in a small vessel plotted in triangular and tetrahedral diagrams ,
respectively .
r.ar.w limit. broccoli Atmosphere of Chlorine .—The limits of cer-
$rdrobea Croton
taotwxlde tain mixtures of hydrogen and carbon monoxid e
Observed Cakuleted DlRerenoe in chlorine are plotted in a triangular diagram
100
(228) .
0 ` 10 . . . . . .. . ...... .
- 63 0 36 I I10 11.1 . +0 4 WATER GA S
41 .6 0 13.23 12 0 +43
P`411 .0 14 1 14 . 4 + 3
04 IS• 0 ... .. .. ....... The limits of water gas in air, calculated fro m
its chemical composition with the aid of L e

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LLM1[TS OF SIDLED FLAMMABLE OASES AND VAPOR S
Chatelier's formula for mixtures (p. 5), are i n
fair agreement with experimental results (367) .
In some old experiments the limits in air,
with upward propagation of flame in a 3-inch-
diameter tube, open at its lower end, were 9 and
55 percent (56) .
The limits of another sample of water gas ,
with downward propagation of flame in a small
tube, were 11 .9 and 65.4 percent (30) .
Still another sample bad limits of 8 .9 an d
69 .5 percent in a small bulb compared with 6 . 1
and 65 .4 percent, calculated by Le Chatelier's
formula but based on limits of the individual
gases obtained in large apparatus . A carbu-
reted water gas under similar conditions ha d
limits of 6 .4 and 37 .7 percent (367) .
The limits of various samples of semi-wate r
gas and Mond gas have been determined in a
unte burette with both upward and downwar d
propagation of flame . Analyses and results
are given in the original papers (269, 270) .
HYDROGEN AND AMMONIA
The limits of mixtures of hydrogen and am-
monia in air and in oxygen, with downward
propugation of flame , have been determined i n
small tubes . Apparently
e results
theinhihigher-limi
gher-limitt mix-
tures in oxygen giv agreemen t
with Le Chatelier's law (209) .
HYDROGEN AND HYDROGEN SULFID E
Both the lower and higher limits of mixture s
of hydrogen and hydrogen sulfide in air diverge
widely from La Chamber's law throughout the
whole range of mixtures . The limits are nar-
rower than those calculated, hence limi t
mixtures of the individual gases when blended
produce nonflammable mixtures . The results ,
obtained in dosed tubes 5 cm . in diameter, are
plotted in figure 5G for both limits and with Op-'
ward and downward propagation of flame ; the
curves calculated from the law are also show n
(367) .
The results that would he oh :ained if experi-
nsents were so conducted that atmospheric pres-
sure was maintained throughout are not known ,
but it seems unlikely that the wide difference s
between observed and calculated results woul d
disappear .
HYDROGEN' AND METHANE
The lower limits of various mixtures of hydro-
gen and methane in air, With upward propsa-
tion of flame, in n vessel 6 feet high and 12 inchs e
square in section (64), were as follows :
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
Lower OWL Pa rotr
Bremen MetbMne
obeenoit Coleuteted DLQerao
s
100 0 t1
10
_ l l i --.0
M
75 d0 to 4.7 4 —.r
it :e 4. t e. 0 —. to
e 100 5e . . .__ ___-_
The higher limit of a mixture of nearly equa l
volumes of hydrogen and methane in air, with
upward propagation of flame in a tube 2 inche s
in diameter and 5 feet in length, open at th e
firing end, was 22 .6 percent (64), which is 0 . 1
percent less than the calculated value based on
a corrected higher-limit value (13 .8 percent)
for methane in the same vessel :
The differences between the observed values
and those calculated from Le ' Chatelier's
law are just beyond the experimental error i n
two instances . Rather greater differences were
observed in dosed tubes 5 cm . in diameter
with upward and downward propagation of
flame (857) .
The approximate limits for mixtures of
hydrogen and methane have been determined i n
a 5-titer bomb with central ignition (347) and ,
for mixtures of equal volumes only, in narrow
tubes 0 .9 to 0 .2 cm . in diameter (276) .
Influence of Pressure.—The higher limits of
mixtures of hydrogen and methane have been
determined up to about 50 atmospheres pres-
sure in narrow tubes . The results are un-
doubtedly low, as those for the two gase s
separately are low (51) .
HYDROGEN, CARBON MONOXIDE, AND
METHANE
The limits of mixturetinAf'I'iydrogen, carbon
monoxide, and methanein air can be calculate d
- as accurately as those for mixtures of any two '
of these combu ti le gases (64) . •
HYDROGEN, CARBON MONOXIDE, AN D
ETHYLENE
The limits of mixtures of hydrogen, carbo n
monoxide, and ethylene in air, with downwar d
propagation of flame in a tube 1 .6 cm. in
diameter, have been determined (217) .
HYDROGEN, METHANE, AND ETHANE
The limits of all mixtures of hydrogen ,
methane, and ethane in air can be calculate d
with approximate accuracy by means of L e
Chatelier's law (146) .


■■■r■■■rr■■rr■■■r■r■■■■
■r■raarrrr■r■r■rrrrr■■ ■■
■rrrrrrr■r■ r ■rr■■ ■
i
■■ ■
.■■■rr■rr■■r■ ' rr■■■
■re~■■r ■■a■r■■a■■rrrr■
■ . ■r
~~ir■rrr■C■...■■■■■■■■■■■ r
A a■■■rE■■rra■rr■■rrr■ ■ ■
RiME■rrr~craar■r ■■rrrr ■
■ir nn rr■■raa a ■ rrr■
rma■
o
r
i
u rrn , Crr■rr
■■a.■■■
n■
r INNr N
, r nn
nn NAM= r ■
■
nn +~C■C
C 'ME WM
nn a■a te rr UpM
gi
u
rrrr■ rra ■r
nDaervad

' r

C■rrr■■
' nn r■■■■Ca■rr ■ r

. prop,B'lion
-,Q
,./ a la m a.,.
_ . ~ ~ /aye

■ar■■ ■a Cr ■■r
rr ■
. -.
°°'P--
ro
■I■ ■ ■ .S.. r
r■ rir ■ aaauaaaa
■r■■ a ■■r■■rr■■■■ ■
~■r■■r■
■
~~:L n and ropat•~ r r■~~
wril = Lorerlimita

~• SiSiamieriam p
UP•'aM ProPaFation
- • __ _,-_ . _ abxrved
,
a

i~ i`iiiliil'iirMiiE(iifiW
■■r■ ■■■■WNrr
'US a 20 10
Hs 100 8o 60
PERCENT
Ftaune 50.-Limits of Flammability of Alixtures of Hydrogen Hydrogeo Sulfide in. Mr .

Atmospheres of Air and Nitrogen (Air De-

ficient in Oxygen) .—The limits of all . mixture s
of these three constituents in air deficient in
oxygen or of mixtures of the three constituent s
with nitrogen in air (or in air deficient in
oxygen) can be calculated with approximat e
accuracy from curves given in figures 1 and 2
or from tabulated results (146) .
HYDROGEN AND ETHANE IN OXYGE N
According to an old observation, the highe r
limit of a mixture of equal volumes of hydroge n
and ethane in oxygen, with downward propaga-
tion of flame in a cudiometer tube 2 cm . to
diameter, is between 56 and 57 percent (243) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 LLMITS OF MIXED FLAALMABLE GASES AND VAPOR S
HYDRQGEN AND PENTAN E
The limits for a mixture of 3 volumes o f
pentane and 2 volumes of hydrogen in air,
m a horizontal glass tube 2 .5 cm . to diamete r
with the firing end open, are 2 .4 and 8 .7 per-
cent (278) .
HYDROGEN AND ETHYLENE
HYDROGEN AND ETHYLENE IN AI R
The limits of the entire range of mixtures o f
hydrogen and ethylene in air, with upwar d
and downward propagation of flame in close d
tubes 5 cm . in diameter, were rather narrower
than those calculated by Le Chatclier's law ;
the differences were a few tenths of I percen t
at the lower limits and a few percent at th e
higher limits (357) .
The limits with downward propagation o f
flame in a tube 1 .6 cut . in diameter have bee n
determined (217) .
HYDROGEN AND ETHYLENE IN OEYGD1
The limits of mixtures of hydrogen an d
ethylene in oxygen have been determined i n
small apparatus (202) .
HYDROGEN AND ETHYLENE IN OTHER ATMOSPHERES
Atmospheres of Air and Carbon Dioxide .--
The limits of hydrogen and ethylene in variou s
mixtures of air and carbon dioxide have bee n
determined with downward propagation o f
flame in a tube 1 .6 cm . in diameter (217) .
HYDROGEN AND ACETYLEN E
Lower limits of mixtures of hydrogen an d
acetylene in air have been determined wit h
upward and downward propngation , of flame
in closed gloss tubes 5 and 7 .5 cm . in diameter.
With upward propagation end up to about 4 3
percent acetylene in the hydrogen-acetylene
mixture, the lower limits were markedly greate r
than those calculated ; with 43 to 50 percen t
acetylene the lower limits decreased suddenl y
and thereafter the limits almost coincide d
with the calculated values. With downward
propagation the lower limits were consistentl y
a little greater than those calculated (357) .
"AMMONIA CONTACT GAS "
The higher limit id oxygen of a mixture pre -
pared for the commercial synthesis of ammoni a
is 92 .2 percent at atmospheric pressure and
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
94 .5 percent at 400 atmospheres . When 5
percent methane is added the limits are 83 .7 an d
92 .8 percent, respectively . No analysis of th e
contact as is given (10) .
The limits of the mixture 3H,+N 1 in a
500-cc . flask with central ignition are given a s
6 and Si percent in air and 5 and 90 percent in
oxygen (323) . These figures do not agree with
other results on the flammability of mixtures of
hydrogen, nitrogen, and oxygen . (.See p . 20 . )
HYDROGEN AND OTHER GASES OR VAPORS
The limits of a series of binary mixtures o f
hydrogen and various gases and vapors hav e
been determined with downward propagatio n
of flame in a tube 5 cm . in diameter and 65 cm .
in length . The results for each pair of mixture s
were of the same general type as those for mix-
tures of hydrogen and ethyl bromide, whic h
have been quoted on page 22,. where their
significance is discussed .
The gases and vapors used were ethyl bro-
mide (311) ; methyl iodide, methylene bromide,
bromoform, ethyl iodide, and ethylene bromid e
(312); hydrogen selenide and diethyl selenid e
(313); din►cthyl selenide and din ethyl tellurid e
(314); tin tetramethyl and lead tetramethy l
(315); ethyl alcohol, ether, acetone, benzene,
pentane, cyclohexane, methyl cyclohexane, an a
a hydrocarbon mixture (316) .
From a study of the dew point, density, an d
range of flammability of such mixtures of hydro-
gen and small quantities of other gases an d
vapors, it was concluded that tin tetramethyl
is the best suppressor of explosions for hydroge n
to be used in balloons or airships, followed b y
dimetl►yl selenide and ethyl bromide ($17) .
However, the •only practical importance o f
these experiments seems to be the proof that
none of the additions is effective in destroyin g
the flammability of hyfirt en when it is mixed
with air ; with 0 .90 percent of lead tetramethyl
the range of flammability is 9 to 54 percent corn -
nred with a to 71 percent for pure hydrogen .
K 1oreover, the jtrrnianbility of a balloon fabric)
to the 'explosion suppressors is not considered .
AMMONIA AND ETHYL BROMID E
The limits of mixtures of ammonia and ethy l
bromide in mixtures of oxygen and nitroge n
have been determined in very small tubes. A
diagram in the original paper shows the limits
for ammonia and ethyl bromide vapor singl y
and mixed ; neither gas propagates flame down-
ward when mixed with air, but each will bur n
in oxygen (204) .


112 UMITS OF FLAMMABILITY OF GASES AND VAPORS

MMMMMM MiMWMMUMMMM .
1MM RUM WMMMMMMM
WMM MMMIMMMMMMMMMM MEE NM
mum ammo manamma n
nammmum mmunmammumm
MWIMMMM MMMMM UM MU MM1
10

'gam n
MMNM MUMMMMMMM M NM MUM
MMMMMMMMMM M
mommummmonum
I
mummmmmmmmommm m
M EMMMMMM WM
mm lams m
IIIIIIIIIIN.

in
1R
3
II"
mm . MIMESIS. Ma ss
Ammmam
amm o
mm

'OlIMMPmESss ss
pw -===MUM MMM
MMMEVE-
im mmmmmmmmmmmmmm am
:111'
alillIlIO' ~
:Wm
1
1

m -r- u
r
ulmlu ■
Em.samm i
MMMIMMa°,,v EMMMEWlir ME
coft,-,,11hJUIMMMMMkIMMMM am
M
-ampmummmsms
mbmggg N
Downwardpropagation

mmammambftkim,m r
MMEMMUMMMMM MWMM===EM!9I
MMMMMWWM MMMMIMAIM
MMMMEMMMME . MMMM
MI MMMMMMMMUMMMMMMMEMMM M
1
MMEMMI WMEMMEMMMMMM6Z2 ME
MMMMMMMMM MUMS MMM
owmed MAIM=
-n4swOMPsla mmmum s
lummm.
-fo p
D-n—l ProPeati- MI .

ff7a t m
L.5.osimmimmmoswi-
5 P,L
AMWm--- - BM W MUM 1-- umi w fil
MIM u - . H P-Para tion MINNIMIM IMM

m mom mom mum mo m

mr mu Namur
mmmmmmmmmnm
Methane 0 40 50
His IOO 60 40 20
PERCENT
Fin vat: ST.—Limits of flammability of hlIxtures of Hydrogen Sulfide and Methane to Mr.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 HYDROGEN SULFIDE AND METHAN E
!loth the lower and higher limits of mixture s
of hydrogen sulfide and methane in air, wit h
upward and downward propagation of flame i n
closed glass tubes 5 cm. in diameter, deviat e
widely from those calculated . The lower limits
are markedly high . The higher limits, on the
other hand, are very high for mixtures rich. in
hydrogen sulfide and low for other mixtures o f
the series, so that the curves (with both direc-
tions of propagation) for experimental an d
calculated figures cut across one another .
These curves are reproduced in figure 57 (357) .
HYDROGEN SULFIDE AND ACETYLEN E
The lower limits of mixtures of hydrogen
sulfide and acetylene in air have been deter -
mined with upward and downward propagatio n
of flame in closed glass tubes 5 cm . to diameter .
The maximum deviation from the calculate d
limit was 0 .3 percent (857) .
CARBON DISULFIDE AND VARIOUS GASE S
AND VAPOR S
Neither lower nor higher limits of mixtures
of carbon disulfide with ether, benzene, acetone ,
and acetaldehyde, with downward propagation
of flame in closed tubes 5 cm. in diameter and
150 cm . in length, agreed with Le Chatelier ' s
law . It has been suggested that the propaga-
tion of flame in mixtures of carbon disulfide an d
air may be catalyzed by some product of its
combustion and that the catalytic effect may be
inhibited by ether, benzene, acetone, and
acetaldehyde (355) .
Mixtures of carbon disulfide with ethy l
bromide, diethyl selenide, tin tetramethvl, len d
tetramethyl, pentane, and a gasoline (`faction
boiling at 91 .6° C . give abnormal results . Th e
lower limit is raised by the firs small addition s
of these substances, then decreased by larger
amounts . The higher limits are at first rapidl y
reduced, but further additions of some of thes e
suhxtnnres cause the higher limit to rise (250) .
The abnormal effect, of small quantities o f
certain flammable substances on the lowe r
limit of carbon disulfide has already been dis-
cussed (p . 29) ..
CARBON MONOXIDE AND METHANE
The lower limits of various mixtures of carbo n
monoxide and methane in air in a vessel 6 fee t
high and 12 inches square in section, wit h
upward propagation of flame at atmospheri c
pressure (64), were as follows :
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
Lower limits of flammability of carbon monoxide
and methane in air
Le .ee amt!, permit
Carbon
tneawtdo Menem
Observed 4h°lated D2•erwos
t0O
0O
•
10
tt 5
11
7$ 11O. S0 *.10 _at
w d0 7. r 7. r i
03 7, 1
2
. L!
75 —. 1
• 100 aOi. ._
The higher limit of a mixture of equal volumes
of carbon monoxide and methane in air, with
upward propagation of flame in s tube 2 nches
in diameter and 5 feet in length ; open at the
firing end, was 22 .8 percent (64), which is 0 .4
percent less than the calculated value based on
a corrected higher-limit value (13 .8 percent) for
methane in the same vessel .
Experiments with downward propagation o f
(lame in a Hempel pipette showed fair agree-
ment with the calculated values for the lower
limits of mixtures of carbon monoxide and
methane, but for the higher limits the observe d
values were always low, sometimes as much a s
10 percent (46) .
Atmospheres of Nitrous Oxide .---The limits of
mixtures of carbon monoxide and methane in
nitrous oxide, with downward propagation of
flame in a 15-mm . burette, have been plotte d
in a triangular diagram (890) .
CARBON MONOXIDE AND ETHYLENE
The limits of mixtures of carbon monoxid e
and ethylene in air, with downward propaga-
tion of flame in a tube 1 .6 cm . in diameter, have
been determined (217) .
CARBON MONOXIDE AND ACETYLEN E
The lower limit of pd,tgixture containing 84 '
ecrcent carbon mnnnAide and 16 percent acet7-
Lne has been determined in a vessel 4 cm . in
diameter and 25 cm . in' height, presumably
with downwaMa egation of ffarne'from an
open end . It was :t-percent in air, agreein g
closely with the figure calculated from the limit s
of carbon monoxide and acetyleno observed
individually under the same conditions (222).
METHANE AND HIGHER-PARAFFIN
HYDROCARBONS
METHANE AND ETHANE; METHANE AND PROPAN E
METHANE AND BUTANE, AND ETHANE AND SUTANE
Bt AIR
The lower limits of methane-ethane, methane -
propane, methane-butane, and ethane-butan e
mixtures in air, with upward propagation of


114 LLUITS OF FLAMMABILITY OF CASES AND 'VAPOR S

flame in a tube 5 cm . in diameter and open a t
the firing end, agreed closely 'with the values
calculated from I .e Chntelier's law ; the higher
limits were slightly less than those calculate d
r (74) .
The regularity of the lower limits of suc h
mixtures is discussed on page 115.
METHANE AND ETHANE IN MIXTURES OF AIR .
NITROGEN, AND CARBON DIOXIDE
The limits of mixtures of methane an d
ethane in mixtures of air, nitrogen, and carbo n
dioxide can be calculated closely by the metho d
given on pages 5 to 8, with the aid of figure s
1 and 2 (147) .
METHANE AND PENTANE IN AIR
Throughout the range of mixtures of methan e
and pentane the lodg er limits with upward or
downward propagation of flame in a tube 5 cm .
in diameter agree with the limits calculated b y
to Chateliers formula from the limits of
methane and pentane separately (357) . • The
lower limit of a 50 :50 mixture, with horizontal
propagation of flame in a tube 2 .5 cm. in diam-
eter, open at the firing 'end, was close to the
calculated value (272) .
The higher limits in the S-cm . tube were
usually less by a few tenths of 1 percent than
those calculated (357) . ,
MIXED PARAFFIN HYDROCARBONS IN
GENERAL
The lower limits, but not the higher limits ,
of the simpler paraffin hydrocarbons and thei r
mixtures exhibit an interesting and usefu l
regularity (74) ; they are approximately a
function of the analytical ratio C:44, C being
the contraction observed on exploding the mix-
ture with excess air and .1 the volume of carbon

6
■r■■r■ua■■aa■ar■■■r■rr■riaa■ ■■
■r■rr■aar■r■■r■s■■■ ■■■r .ar■ m
mmommmmumpummmmummo
5 111111111111111111111111111 1■
■rrOrr■Err■■M■■MEr■ ■r■r■tlI IRr
_■■■IIaa■■■■I■aru■■■■■■Ip ■■
■■r■~■~r>,r,■■■■~~r■■r■c .li~w■
t%H%.uetru.aa..amu~Mraaa.N a
■r■■■aaaarri■r■r.■r;Ma■■■aa■■ ■
■■■Mr■■M■■■■M■■rMvtr■r■■■■■■r ■■
■rrar■ra■ r■rr■rrM■r■■■■■M■r
■•MMUMMMUMURAMMM■■■>r■■M>•■ ■■■
■■■M■■■■■Mr■■■rW■■■■■■■■■M■
■■u■■■r■=■■rr■■m■r■I■■r■m■■
r■■r■■rm■ mummmumm ■r■>•■■ ■■
rr■M■r•EMM■r■■■Or■■■m■■■■■■
,■■~~l•■1•~/rr~~■
■M)•U■/Ml'" MMMM>•/ • Pure byd a, . ■■

r■■m■■cn■■rm■ +o Other
x Methane
ethane mm
nOxtures ■ ■

■r■~■em■■■r■mr binary
Natural ga .
mixtures

■■■MM*MW■*M*MMr■■■M*r■■■r■I■r *
roMEWMEI•MM■ nnnn ■M>.r■■■■EMMM M
r■~l,■a■urr■r■■■C$rr■rrrr■■■r■
I

0.5
VEMMEraiaaai%ra iar araaaauuuu~
1 .0 1 .5 2.0 .:
C :A RATI O
1
uieoae 5&—Variation of Lower Limit of Paraffin Hydrocarbons with CA Ratio.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LLMITS OF MIXED FLAMMABLE GASES AND VAPOR S
dioxide produced. The lower limit of any
mixture of paraffin hydrocarbons can therefor e
be deduced without exact knowledge of it s
composition ; all that is required is a determina-
tion of the 6:A ratio of the mixture and refer-
ence to figure 58, the curve of which is draw n
through experimental points . The broken par t
of the curve may be used to supply the lower
limits•of heavy paraffin hydrocarbons and o f
their mixtures .
Recent figures for the higher hydrocarbon s
confirm the curve, but figures for ethane an d
propane tend to flatten it somewhat .
NATURAL' GAS
NATURAL GAS tN AIR
. The lower limit of a natural gas in air, wit h
upward propagation of flame in a tube 7 feet i n
height and 12 inches in diameter and with th e
firing end open, was 4 .91 percent . The mixtur e
contained about 2 percent eater vapor . The
percentage composition of tho gas was :
A!ethane, 87 .4 ; ethane, 6 .8 ; propane, 1 .55 ;
butane, 0.81 ; pentane, trace ; nitrogen, . 3 .2 ;
oxygen, 0 .1 ; and carbon dioxide, 0 .0 . The limi t
figure represents, for convenience, the propor-
tion of the flammable constituents of th e
natural gas in the limit mixture with sir . I f
the nitrog en and oxygen were included, th e
limit would be nearly 5.1 percent. For propa-
gation of flame upward from closed to open en d
in the same tube, the lower limit was 4 .74 per-
cent (flammable constituents) .
The limits of the same natural gas wit h
upward propagation of flan g e in a tube 2 inches
in diameter, open at the firing cod, very nearly
obey Le Chatelier's law, For dry gases th e
limits of the flammable part of the natural ga s
were 4 .80 and 13 .46 percent. The figures cal-
culated from the corresponding limits for th e
individual constituents—methane 5 .24 and
14 .02 percent ; ethane 3 .22 and 12 .45 .percent;
propane 2 .37 and 9 .50 percent ; and butane 1 .86
and 8 .41 percent—were 4 .85 and 13 .72 percent .
The lower limit is not appreciably affected by
the nitrogen in the natural gas, but the highe r
limit is (compare fig . 20, showing the influenc e
of nitrogen on the limits of methane in air) ,
An estimate of the influence of nitrogen on th e
higher limit may be marie thus: If 0 .4 percen t
nitrogen is deducted (which corresponds wit h
0.1 percent oxygen found in the natural gas )
there remains 2 .8 percent nitrogen, which is
brought to the mixture with the natural gas .
As this amount is present with the 13 .7 percen t
hydrocarbons in the whole limit mixture, it
represents about 0 .4 percent "additional nitro-
gen" in the original "atmosphere ." Figure 20
shows that this depresses the higher limit o f
methane about 0 .1 percent, and doubtless its
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
effect on the higher limit of natural gas is abou t
the same (74) .
Further evidence that the limits of natural
gas may bo calculated fairly , accurately from
the limits of its various constituents is provided
by data showing that Le Chatelier's law is fol-
lowed when certain pairs of the constituents are
taken together. (See the sections on methane-
ethane, methane-propane, methane-butane ,
methane-pentane, and ethane-butane mixtures . )
The limits of 22 samples of "natural gas," o f
widely varying compositions from many States, .
have been calculated on the basis of analytical
figures expressing the flammable constituents as
methane and ethane only (154) .
Influence of Pressure.—In experiments in a
closed tube 2 inches in diameter and 12 inches
in length, with upward propagation of flame
and with initial pressures from 1 to about 24
atmospheres, the lower limit of a natural ga s
(annl yysis : Methane, 79.6 ; ethane, 19.2 percent)
remained nearly constant ; the higher limit ros e
from 14 .0 to 37 .5 percent (154) . In an ex-
tension of thew experiments, but with a tub e
15 inches long and a natural gas giving by
analysis methane around 85 percent and ethane
around 15 percent, the following figures may b e
read from plotted results (160) :
Initial pressure . Atmos 100 1 .000 3.000
lb. per U,.' photo 3p00
Lem*: Molt x1 0 i ti 1 .0 0 7.A 11 3
Higbee 1Lmlt , ._ . .- 14 .3 44.3 S2 0 10 .0 000
)
In narrow tubes (3 or 5 mm . in diameter)
with downward propagation of flame the highe r
limit of a "natural methane" (methane, 80 .3 ;
higher paraffins, 11 .1 percent) rose from 14 .2 5
percent "methane " at atmospheric pressure to
39 percent at. 12 atmospheres and 52 percent a t
50 atmospheres (E81) .
At pressures below spheric, in a close d
tube 2 inches in diameter and 6 feet . % length,
with upward propagation of flame, the limits
remained nearly constant as the initial pressur e
was rcdueed .- kwJ,,,,,atmosphcric to about 20 0
tnm . At lower pressures the limits converge d
and propagation ' was not obtained belo w
50 mm . (15 4 ) .
Influence of Turbulence Caused by Fans or
by the Detonation of Explosives :—Experiments
have been made to determine whether the lower
limit of natural gas in air is affected by turbu-
lence such as that produced by fans or by th e
detonation of explosives (75) . The use of
explosives necessitated experiments on a rathe r
large scale, therefore a 20-foot section of a
100-foot steel tube 6 ;f feet in diameter was used .
The explosives were fired near one end of the
horizontal axis of the tube . The tube ha d
paper-covered relief vents along the top, an d


116 LIMITS OF FLAMMABILITY OF GASES AND VAPORS

the end opposite the source of ignit ion was close d
by a paper cover . The appearance of flame a t
r the various orifices enabled the extent of a n
inflammation to be roughly judged . • The
L observations were aided by noting the effec t
of an experiment on tufts of guncotton placed
at various points in the tube .
The results were conveniently classed as (1)
explosions, when the paper covers were blow n
out , violently and flame appeared from th e
vents ; (2) inflammations, when it *as eviden t
that some gas had been ignited, but no great
violence was observed ; and (3) nonignittons .
A small electric igniter of black blasting
powder gave a lower limit of explosion of abou t
5 .6 percent natural gas and a lower limit o f
inflammation of about 4 .6 percent .
Black blasting powder (50 or 200 grams),
burning with a long flnme, gave limits of 5 . 1
percent and less than 4 .7 percent, respectively.
A straight nitroglycerin dynamite (50 o r
200 grams) gave a lower limit of explosion of 5 . 6
percent ; "inflammation" could not be observed
with certainty at lower percentages on accoun t
of the violent effects of the detonation on th e
paper covers.
Turbulence induced by a fan run at appropri-
ate speeds reduced the limit of explosion to 5 .0
percent . The limit of inflammation was reduced
slightly at moderate speeds ; at higher speeds
inflammation, apart from explosion, could no t
be observed.
The term "inflammation", as used in . the
preceding paragraphs, apparently. covered th e
range of mixtures capable of propagatin gflam e
from the source of ignition for some distance
upward, spreading more or less laterally ; the
term "explosion" was used to designate enough
inflammation at sufficient speed to cause the
effects described . The reduction of the limi t
of explosion by the long flames of gunpowder
was due to the greater burning induced by
them ; a similar effect was produced by turbu-
lence.
The natural gas used in these experiments
contained 87 .8 percent methane, 6.9 percent
ethane, 2 .6 percent propane, 0 .8 percent butane,
1 .9 percent nitrogen, and no oxygen or carbo n
dioxide. The percentages of natural gas in th e
limit mixtures refer to "nitrogen-free" gas (75) .
NATURAL GAS III OTHER ATMOSPHERES
Atmospheres of Air, Nitrogen, and Carbon
Dioxide .—An extensive series of experiments in
a 100-cc . Hempel pipette and in a short steel
cylinder of 2 .8-liter capacity has been made
with one sample of natural gas (Pittsburgh )

■■■ii■ssa■sst■s■a■E
Ii'!' IMMMMMMMMEMM■$■■
■■■uu■■ ■■
■r■■ll■■
20
■~1omm R■l.l■■■ ■rl■l■■ll■■ ■
■UM ■ . IMMMIM■l■■■ll■l■■ll■■ ■■
n I!■l
4l1 '~Ill■■r■■■■■■■■■ll■/
16 i■MramllIU$I
,mmmmmasmmmr
l hq
nn Impossible mixtures

■ Ammmu R■■m..U■■l■ ■■l■■1■■

■ ►l■l■■■■■■■■■1Mi:R■■lMOMNMO■ ■
nn ■■■l■l■■ll■■■$■R 5l■■m ll■
X ■■1■■■m■■■01■■■
12
sm. •ws■■■■ p

11111111111■1!! .:ima ammo Ip,_

irrmmo~m
W L000 lb.
a.
z itaiaaarr
115sawaam mmmeemm a
°r m 2.000 lb.
m

0 mmmmwmmummmumummmsmsum i

■■■■■■l■■■■Ull■■ l■■l■■■■■■l
■■■llllll■■l■ l■r■ll■l■■■■■l ■

■
■■l■■

■■■■■l■mmmuml■■r■R=■l■■l■■■ ■
■■l■ ■ ■■~■ m■r
4

l mmmm ■■l■UU■R■■■■■■ ■ .l .

0
■■■ mmmmmR$U mmmmi■ii■■i■■■■
8 16 24 32 40 48 56
NATURAL GAS, PERCENT
Fiouni : 59 .-Oxygen Content of Flammable Natural Gen-Air-Nitrogen Mixtures at High Pressures .
C..Tl

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LIMITS OF MIXED FLAMMABLE OASES AND VAPORS
and atmospheres that contained oxygen, nitro-
gen, and carbon dioxide . The composition of
the mixtures tested, including the natural ga s
ranged from 19 .9 to 14 percent omen an d
from 0 to 61 percent carbon dioxide . The
results are plotted in comprehensive curve s
in the original paper (64) .
The limits of any sample of natural ga s
containing methane, ethane, propane, butane,
nitrogen, and carbon dioxide can be calculate d
with reasonable accuracy from the limits o f
the individual combustible constitutents i n
atmospheres of different oxygen and carbon
dioxide content (147) . The experiments were
made in a tube 2 inches in . diameter and 6 fee t
in length, with upward propagation of flam e
at atmospheric pressure ,
Atmospheres of Air and Nitrogen at Elevate d
Pressures .—The limits of a natural gas (meth-
ane, 85 percent ; ethane, 15 percent) in air-
nitrogen mixtures at elevated pressures hav e
been determined, with upward propagation o f
flame in a tube 2 inches in diameter and 1 5
inches in length. The results are summarize d
in figure 59, from which may be ascertaine d
whether any such mixture, of known natura l
gas and oxygen content, is flammable at th e
various pressures indicated (160) . Interpola-
tion, or reference to other diagrams in th e
original paper, will give the same information
for other pressures.
METHANE AND ETHYLENE
The limits of mixtures of methane and ethyl-
ene in air, with upward and downward propa-
gation of flame in closed tubes 5 cm . in diameter,
were generally slightly narrower than thos e
calculated (367) .
METHANE AND ACETYLENE
The limits of mixtures of methane an d
acetylene in air with upward and downward
propagation of flame have been determined i n
closed glass tubes 5 cm . in diameter. Th e
lower limits with both directions of propaga-
tion were slightly higher than those calculated.
The higher limits with downward propagatio n
were a little less than those calculated, bu t
with upward propagation they were much
greater. The greatest difference was shown b y
the 40 :60 acetylene-methane mixture, fo r
which the observed higher limit, was 47 percen t
compared with 21- percent for the calculate d
higher limit (357) .,
METHANE AND CERTAIN HALOGENATE D
HYDROCARBON S
For the limits of mixtures of methane wit h
certain halogenated hydrocarbons which ar e
either flammable separately, or become flam -
sssaao•--as—a
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
11 7
nimble when mixed with methane, se pp . 51 to
54 .
METHANE AND VARIOUS VAPOR S
The complete ranges of flammability o f
methane and the vapors of various substances
have been determined individually in air wit h
downward propagation of flame in narro w
tubes. These substances are : lsobutyl chloride ,
propyl bromide, ethyl iodide, sulfuryl chloride,
silicon tetrachloride and silicochloroform (195) ,
and isoamyl bromide (196) .
Atmosphere' of Oxygen .—Representative ae-
ries of experiments on the limits of mixtures o f
methane and methyl chloride in oxygen have .
been made at pressures from 600 mm . dow n
to the point of coincidence of the limits (89) .
BUTANE AND ETHYL CHLORID E
The limits of mixtures of butane and ethy l
chloride in air, with downward propagation o f
flame in a I,6-cm .-diameter tube, are given in
curves (340) .
BENZINE IN AM
The lower limit of benzine in air with upward
propagation of flame in a tube 6 .2 cm . in
diameter, open at the fi ring end, was 1 .1 per-
cent . The limits for propagation downwar d
in a closed pipette 1 .9 cm . in diameter .^-er e
2 .4 and 4 .9 percent . All the benzine distilled
below 105° C. (95) . In similar experiments to
the last named, a second observer found limit s
of 1 .9 and 5 .1 ercent for benzine having a, boil-
ing range of 67 to 94° C. (323) .
BENZINE IN OTHER ATMOSPHERES
Atmospheres of Co sition Between Ai r
and Pure Oxygen .—'!71b Ibwer limit-of benzin e
in all atmospheres Of composition between ai r
and pure oxygen, with downward propagation
of flame in a .H,q t_o burette 19 mm . In diameter,
was about 2 .0 p2'I•etnt ; the higher limit rose
from 5 .1 percent in air to 19 percent in 6 0
percent oxygen and 28 .6 percent in 94 percent
oxygen (323) .
Atmospheres of Air and ,Carbon Tetra-
chloride .--Large amounts of carbon
must be added to benzine In orde r
that the vapors arising from the liquid shall
be incapable of forming a flammable' mixture
with air (7) . ,
BENZINE AND BENZEN E
Influence of Pressure .--Curves showing the
influence of pressure up to 500 atmospheres o n
the limits of a 60 :40 mixture of benzine

MITTS OF FLAM.L&BILTTT OF GASES AND VAPOR S
(boiling range, 78°-82°) and benzene in ai r
have been given (11), but the range of flamma-
bility seems impossibly wide.
GASOLINE (PETROL)
GASOLINE 1N AI R
The limits of three gasolines, with upwar d
propagation of flame in a 2-inch-diameter tube
at approximately atmospheric pressure, have
been determined (141). The gasolines had th e
following properties :
epeelfle os~
flue& Oela+» ?attar
ley at tales r. suet . tam . Rs
1 77 0 me
2 A2 . :ro1
S .7;61
Tests of the flammability of mixtures were
made alternatively by applying an alcoho l
flame at the open lower end of the tube, or b y
passing a spark just within the lower end of th e
tube without, removing the ground-glass cover-
plate . When the latter procedure was used ,
the mixture was originally at a reduced pressur e
but the passage of flame raised the pressure t o
approximately atmospheric . The results ob-
tained by the two methods were nearly th e
same, as is shnwa by the following figures .
Limits e1 flammability In air, percr°t by volum e
sunpie !i
t(—~
r
1'utr ef n. name
Im+Una
v_ I1
Lower nlaber
I
1 f 1 .SO 1 .0
2 I. 50 7. 0
7 1 .0 7.4
The limits in sir may therefore be put a s
1 .4 and 7 .6 percent (141) .
Other determinations of the limits of gasolin e
v apor in air are as follows :
1'ho liniita with upward propngntinn of flam e
in a 2 .8-liter glass jar were 1 .45 and 6 .2 percen t
gasoline vapor (by volume) . The limits with
downward propagation in a Hempel pipett e
of 100-ce. capacity were 1 .95 and 5.25 percent ;
with upward propagation the lower limit was
1 .55 percent . The . gasoline had a specific
gravity of 0.689 at 15° C . and a boiling range
of 50 to I40° C . ; ht 140° C. there was 2 . 7
percent residue (41, 48) .
The limits in a cylindrical bomb of 1 .2 liters
capacity were 1 .4 and 6 .0 percent (856).
The lower limits of five gasolines in a smal l
test tube ranged from 1 .24 to 1 .64 percent ; the
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
higher limits from 5 .42 to 7 .88 percent (334) .
The lower limits, by weight, of the more -
volatile parts of a series of cracked and straight-
run gasolines are equal (896) . This is explaine d
by the observation that the limits of n exane ,
n-beptane, n-octane, cyclohexane, and benzene
are equal, 57 mg. per liter of air (36) .
The limits of gasolines obtained by the dis-
tillation of raw oils up to 180° C . have been
determined in a 2 .3-liter bottle. The oils are
described as Irak, Pechelbronn, Kettiema n
Hills, Equateur, Saxet, and Santa-FE . Th e
results are expressed in terms of cubic centi s
meters of liquid per liter of air at 100° C ., an d
• r
Vapor P e,- the maximum differences were 4t percent i n
the lower limit, 14 percent in the higher limi t
ure a
(231) .
700 Influence of Pressure .—At pressures below
in
10 atmospheric, in a closed tube 2 inches in
diameter and 6 feet in length, with upwar d
propagation of flame, the limits , of a gasoline
(100-octane) remained nearly constant as th e
initial pressure was reduced from atmospheric
to about 300 nun . At lower pressures the,
limits converged, and propagation was not
obtained below 35 mm. pressure (180) . In a
tank of 125 cubic feet capacity, the limitin g
pressure, at whieb no propagation of flame was
obtained in any mixture of gasoline and air ,
was about 33 mm . (257) .
In another series of experiments, which see m
to have been tests of the uicendivity of various
sparks rather than determinations of limits o f
flammability, ignition was not obtained, in. -
Tubesates ,
Itoitfon
spec k
4 .19-liter spherical vessel, below 100 mm. (852; .
Influence of Temperature .—At 400° C . th e
lower limit with upward propagation of flam e
tower Higher
in a Hempel pipette of 100 cc . capacity wa s
1 .1 percent compared with 1 .55 percent at
la 7.73 room temperature (41) .
1. 43 7. i
1.40 7. 1
Other experiments indicate that temperatur e
has a greater effect ; btit4. results .scera unre-
liable because in a parallel series of tests th e
lower limit of alcohol at room temperature we*
undoubtedly too high (20) .
The lower limits-eCome French gasolines,
with downward propagation of flame in a
2i4-liter bottle, at temperatures from 100° t o
2500 C ., are given in table 41 (p . 126) .
Experiments in a 4 .19-liter spherical vessel ,
at 1,000 mm . pressure, gave the limits of a n
87-octane gasoline as 3 .1 and 7.6 percent a t
0° C., 1 .35 and 7 .1 percent at ._ 46 .5 0 C.
Other results are represented in a series o f
die~ ggrems (259) .
Tlie lower limits of a gasoline (distillatio n
range, 58° to 131° C .) with upward propagatio n
of flame in a cylinder 30 .6 cm . in diameter and
39 cm. in length, vented at its upper end, are
1 .07 percent at, 21° C ., 0.94 at 100° C., and
0 .77 at 200° C. (899) .


LIMITS OF MIXED FLAMMABLE GASES AND VAPORS 119

GASOLINE IN OTHER ATMOSPHERES
Atmospheres of Air and Nitrogen (Air De-
ficient in Oxygen) .—The limits of the three gas -
alines defined above (see Gasoline in Air ,
samples 1, 2, and 3), with upward propagatio n
of flame in a 2-inch tube at approximately
atmospheric pressure, are given by a singl e
curve (Ne) in figure 60. No mixture of gasolin e
vapor, air, and nitrogen is flammable if it
Atmospheres of Air and (a) Trichloromono-
Iluoromethane, (b) Dichlorodifluoromethane ,
and (c) Dichloromonofluoromethane .—Figur e
60 also shows the results of similar experiment s
with these three halogen-substituted methanes .
Their extinctive effects are considerably greate r
than those of nitrogen and carbon dioxide, an d
no mixture of gasoline vapor, air, and any of
them which contains less than 17 .2 percent o f
oxygen is flammable (141).

•MMMMMMMMMMMMMMMMMMM■■MMMMMMUMMMMMMMMaaaiM ■
►tA//a/M//aM/MMMMOMM/MMMs/MSMMMaamuss ■MSM■
WA&ZWMmRMMMMMMMMMMMMMSMMMMMMMMMMMMMMM MIME .
7
■IANNMOMMMMMMMMMMMMMMMMMMMMMMMMMmMMM■ ■MMM ■
OMCUUMIMMMIMMMMMMMEMMMMMMMMMUMMMMIN M
MeMOOOPASSMUMMMMUMMMMMIMMMIMMMMMMMUMM ■MMM ■
6 MMMM►`'1iM1~' :4.''k~IMMMMMMMMMMMMMMMMMuuuuuuss ■MMM ■
MMMMI►'M M►.MeRMMMMnuMMMuM/MMMMMMMusR■ uuM/ M
■MM ■uuuuus uuu ■Mum■
'AIM MMMMMMMMMMMMMMMM■ MMMM ■
MMM N ~MMMiini
z 5
0
L1
M m' ` l. iM~
.
w ■MMMMMM 11MM■MMMMt"
a
.Ms~r"S MMMM/MMM_MmMMMMMMMMMMM M
4
a MMMMMMMM■MM MMMMSMw'MMI°'~ .`MMM■ MMMMM MM MMMM M
O /mmMMmM~:° M :- ■M■ M■ .MMM/■
NOMMMMM
4 MMm■MiM'l►!'~M■
w MMMMMMMMM `mMMMMMMmommmmmir mcb uuiuimiiumm
Z MMMMMMMMMM/.MIIM/MMSMMMMMMmwsM hdo is sus
M MMM i ■ `MM - Se:Pa"r~a1?ti !MMMui
Matt'

0 3 MMMMMM MMUMMMUM
3 ti
M■MMMMM /IMMMMMM=MMM M MMM ►MM■ M►r.M/rrt .!IMMM
SMM

MMmmm■ Ia ■MMMMM■mMMMMMMMMaM/MMMM►MMMMi:s■M
MMMMMM .. MMM■HMMMMMMMMMMMM/1MMMM■IMIMSOW $MMS►t
■u./~/$sr/Muumuu/iUSMSPAsl$i
U
2
MMMstiui sammummmmmses m pe tt,mMMMMMMMM1l
6waltsitttum .m .m . .-trr.uaauu. um•MUMM MUMMEMaI1 ■
MMMMM■MMMMM■MMMMMMMMMMMMMMMMMMMMMMMMMMMMMM ■■
■■■MsM$auuMMMmmuMMM .MMUMEMMMMu/i*MMU/mMMIMM
MMMM MM MMMMMMMMMMMMMMMMMMMMMMMUMM SM rM MMN MMMM ► R
m■■isvauisssisMMSMgg aMMMMMMMMMM■9u1MMS 4 u1M M
0 4 8 12 16 20 Zit 32
24 36 0
INERT GAS IN ATMOSPHERE, PERCENT ,
Fiouea 60: -IJrnite of Flammability of Gasoline Vapor ha Various Air-lnerf T3a *tmoapbcres

contains less than 11 .6 percent of oxygen (141) .
Atmospheres of Air and Carbon Dioxide . —
Figure 60 also shows the limits of gasoline vapo r
in . all mixtures ()Lair and carbon dioxide ,
determined as in the previous paragraph (141) .
Atmospheres of Air and Exhaust Gas .—Fig-
ure 60 ulso• shows the limits of gasoline vapo r
t :. in all mixtures of air with a mixture containin g
85 percent of nitrogen and 15 percent of carbo n
dioxide, that is, an "exhaust gas" of a petro l
engine with the fuej :air ratio adjusted to
give a maximum of carbon dioxide and a
minimum of carbon monoxide and oxygen (141) .
Atmospheres of Air and Carbon Tetra-
chloride .--See table 20.
NAPH7TTIA
The limits of naphtha (distillation range , i)
146° to 171° C.) iu air, with upward propaga-
tion of flame in a tube 1 inch in diameter an d
18 inches in length, open at. the top and at a
temperature sullicicut to vaporize. the naphtha,
are 0 .83 and 4 .85 percent (138) ,
Influence of Temperatures .—The lower limits
of three types of naphtha in air at various

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


temperatures, with upward propagation of

flame in a steel cylinder 30.6 inches in diamete r
and 39 cm . in length, vented st, its upper end, .
are given below (239) : Caaspatllie,. tweeze by velum' Composition. peneetby values

Et yI a. Oqp° p !t6Plms Oe7p m

aipet niI um Na 10 mhxml 'Ma r
eepbtlos eptrlu
10.0 0 400 10.0 sa0 70. 0
IS 4 Is I M. s ILO el 0 37. 0
r51,t1IletWO rsnio 104' is 137• C . JSS• le tee C. Sr to 141. 0. 71 3 b.s 7ls 0 70.0 . 90. 1
Specter 'mitt* 0 .5170 0.7st0 0 .7N0.
"Lab point 25• C ' -_ . 40A' C C.
Lower emits st th e
leeowlns temper .
tun*. • C. . CTCLOPROPANE AND ETHYLENE IN OTHER
II 1 .00 0.02
100 0.90 0.77 0.70. ATMOSPHERES
200 0 74 . . . . . . . . . . . 0.17.
Me— . . . . .. . . . . Preesmemm- Protein» ow .
eu..lq, u hustles The addition of IA percent of helium to cyclo-
230 0.t° 4o nn . propane-ethylene-oxtgen mixtures affects th e
minimum oxygen content for flame propaga-
tion only slightly . Similarly, the addition of 1 0
percent of hydrogen had little effect (163) . Th e
Atmospheres of Air and Carbon Tetrochldr- addition of nitrogen, up to 20 percent., "is of no
ide.---See p. b8 . particular advantage in reducing flammability ,
nor does such addition permit mixtures of a
higher oxygen content to be used in anesthesi a
if nontlnnunnbility is required" (163).
The limits of kerosine in air, with upwar d
propagation of flame in a tube 1 inch in diamete r DIMETHYL CYCLOPENTANE AN D
and 18 inches in length, open at the top and a t ISOHEPTANE
a temperature sufficient to vaporize the kero-
sine, are 0 .70 and 4 .80 percent (138) . The lower limit of a mixture of 43 percen t
dimethyl cyclopenteoe and 57 percent isohep v
tane, with downward propagation of flame in a
tube 5 cm . in diameter and 65 cm. in length ,
The limits of the vapors arising from a crud e open nt the firing end, is 1 .12 percent (250) . -
petroleum (flash point -21° C ., boiling rang e
61° to 313° C.) were 1 .3 and 4 .9 percent wit h TURPENTINE .
dewaward propagation of flame in a small
vessel . Above S C. air saturr.ted with the The loiter limit of turpentine in air, condi-
vapors was above the higher limit of flamma- tions not specified, is 0 .8 percent (101) .
bility (229) . Influence of Temperature .—The lower limits
The limits for the higher fractions of petro- of turpentine (distillation range, 153° to 197 °
leum (that is, Diesel oils) lie in a range up to C .) with upward- propagation of name in a .
about 56 mg . per liter even when the oil is cylinder 30.6 cm . in diameter and 39 .0 cm. in
suspended as a fine mist at ordinary„tout .., length, vented at its upperfnd; are 0.6L Ra man '
pentanes (36) . at 100° C. and 0 .54 at 200° C . (239).
Atmospheres of Air and Carbon Dioxide .
The addition of carbon dioxide to the atmosphere COAL ;GAS AND COKE OVEN GAS
narrowed the limits until they coincided ; at COAL GAS ANDCOit: VEN GAS IN AM *
5° C., 61 percent carbon dioxide was necessary ;
at 20°, 51 percent ; at 30°, 44 percent ; and at The lower limit of a "town gas" (64), wit h
40°, 37 percent (889) . upward propagation of flame in a vessel 6 fee t
high and 12 inches square in section, open at the .
CYCLOPROPANE AND ETHYLENE firing end, was 5 .35 percent. The figure calcu-
lated from an analysis given below in conjune -
CYCLOPROPANE AND ETHYLENE IN OXYGEN Lion with the limits of the individual .constitu-
ente of the gas is 5 .36 percent. -
The higher limits df mixtures of cyclopropane The higher limit of a coal gas (64), with up-
and ethylene in oxygen, with upward propaga- ward propagation of flame in a tube 5 feet to •
Lion of flame in a 2-inch-diameter tube, open height and 2 inches in diameter, open at th e
at the lower end, are given in table 39 (165) . bring end, was 30.9 percent. The calculated

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LIMITS OF MIXED FLA.vrMABLE GASES AND VAPOR S
figure is 28 .8 percent, uncorrected for inert con -
stituents of the gas .
The limits of coke-oven gas, with upwar d
propagation of flame in a 6-cm -diameter tube ,
loosely stoppered, were 4 .4 and 34.0 percen t
(280) .
The compositions of the town gas p coal gas ,
and coke-oven gas used in the foregoing experi-
ments were given as follows, in percentages :
Town w Cad P* Cotters,
Rearmed fell fu) ~.+ Wo l
Beaten.etc ° 0.$ L2 . .. . . . . . .. . .
Carbon dtoeWe 10 .1 . .. . . . . . . . . .
Orates s .1
Ctb)Ine, etc 2.9 29
Carbon manmlde 14 1 7.9
RTdroern M9 79.9
11tthsne 19 .4 79.1
Ethou /.9 12 . . . . . . . . . . .
Wtroten 0.2 t.•
The recorded limits of town and coal gas i n
smaller vessels and for other directions of propa-
gation arc such as might be expected fro m
mixtures containing a large proportion of hydro-
gen and different amounts of heavy hydro-
carbons (9, 24, 28, 30, 43, 55, 95, 218, 241 ,
270, 276, 281, 297, 323, 367, 368) . Thus, fo r
the lower limit the figures range from 5 percen t
for upward to 10 percent for downward propa-
gation ; the higher-limit figures range from 18 . 5
percent for downward to 31 percent for upwar d
propagation . One higher limit given is 53 . 8
percent ; the composition of the gas is no t
stated (9) .
An old set of comparative experiments showe d
an increase from 8 .1 to 8 .7 percent in the lowe r
limit of a. coal gas as the diameter of the experi-
mental tube was decreased from 30 to 10 mm .
(218) . More recent experiments showed tha t
flame would still travel, apparently indefinitely ,
in a tube 3 mm . in diameter when one of th e
most explosive mixtures was used but woul d
not travel in a tube 2 mm . in diameter (276)
.—The higher limit of.InfluecoPrs a
French town gas rose rapidly with increasing
ppressure . At atmospheric pressure the higher
limit was 18 .5, at. 10 atmospheres 37 .5, and a t
22 atmospheres 51 ]ercent . Further increase i n
had less effect, as at 54 atmospheres th e
Lit was 56 .5 percent gas . The experiments
limit
.were made with gas compressed into tubes of
3- or 5-mm . diameter, with downward props-
gation of flame, and reasons are given for con-
cluding that the results would be the same i n
wider vessels (281) . However, as similar
experiments (51) gave low results for hydrogen
and for methane these results . also are probably
too low . Moreover, for upward propagation o f
flame the results would probably be higher .
The observations are of importance in relatio n
to the use of compressed gas (281) .
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
Influence of Temperature .—In a 35-cc. bulb ,
wit.h side ignition, the lower limit fell gradually
from 6 .9 to 6 .3 percent, and the higher lima
rose from 22 .8 to 28.7 percent as the tempera-
ture was increased from 15° to 300° C . The
composition of the gas used is given under th e
bead ing Coal Gas in Oxygen (297) .
Influence of Impurities (Nitrogen and Carbo n
Dioxide) .—Two curves in figure 61 show th e
limits of a coal gas polluted with nitrogen and
carbon dioxide, with downward propagation o f
flame in a Bunts burette . The abscissa s
represent the composition of the gas in term s
of its calorific value (30) . These can readily
0 2 be converted to percentage composition, as the
1.i
31 calorific value of each mixture is proportional
96 2 to its coal-gas content. ' -
ti 9
14 COAL GAS IN OXYGEt4
The limits of coal gas in oxygen in a 35-cc .
bulb, with side ignition, were 7.4 and 69 .7
percent (297) . The percentage compositio n
of the coal gas was : Heavy hydrocarbons, 5 .3 ;
carbon dioxide, 2 .0 ; carbon monoxide, 6.4 ;
methane, 34 .5 ; hydrogen, 49 .4 ; and nitrogen,
2 .4 .
Influence of Temperature .—In the same series
of experiments it. was found that, as the temper-
ature was raised to 300° C ., the lower limi t
fell gradually to 6 .9 percent and the higher
rose to 72 percent .
Influence of Pressure .—The higher limit of a
French town gas in oxygen was 78 .6 percent at
atmospheric pressure, 85 percent at 10 atmos-
pheres, and 89 .1 percent at 43 atmosph eres .
The experiments were performed as describes ;
under Coal Gas in Air . At 74 atmospheres
pressure a mixture of 89 .6 percent gas and 10.4
percent oxygen could not be exploded . It was
concluded that mixtures of town gas and oxygen ,
compressed in tanks up to 200 atmosphere s
pressure, could not. explode if they containe d
less than 10 percenCbxtgen but that it woul d
be prudent not to exceed 5 or 8 percent oxygen .
Moreover, such mixtures must be compresse d
in such. a w as to .avoid undue rise in -thei r
temperature, Melt-would widen' the limits. '
Up . to 80° C. however, the higher limit is no t
appreciably altered (50, 281) .
COAL GAS IN OTHER ATMOSPHERES
Atmospheres ' of Composition Between Ai r
and Pure OxYgen .--The limits of a synthetic il-
luminating gas in air in a Bunts burette wer e
9 .7 and 25.1 percent; they rose to 9 .9 and 73 .7
in a mixture of 96 percent oxygen and 4 percent
nitrogen (323) . The percentage compositio n
of the gas used was : Carbon dioxide, 2 .0 ; ethyl-
ene 3 .8 ; carbon monoxide, 9 .0 ; methane, 30.2 ;
hydrogen, 51 .0 ; and nitrogen, 4 .0.


LIMITS OF FL MMABZL.tTY OF GASES AND VAPORS

ry I
Water tat

eo

IM O
,"lr
.
L
r

%MI N•
Coal sits and COI
A Coal gas and N:

sa g
Tows

Coal
JuJiL r 7arlII

Coal gas and water sae

seer X00 '+ 100

CALORIFIC VALUE OF THE GAS, 8 . T. U. PER CUBIC FOOT

Frown. el .—Limlts of Flammability of Coal Gas With Various Admixtures.

Atmospheres of Air and Nitrogen (Air De- of water vapor is present the limits coincide a t
ficient in Oxygen) and Air and Carbon Di- about 8 .4 percent coal gas (368) .
oxide.—From the data in figure 61 calculations Atmospheres of Air and Methyl Bromide.--
may be made of the composition of the atmos- The addition of increasing amounts of methy l
pbere which, in a small burette, is just inca- bromide to the air causes the limits of a coa l
pable of forming i flammable mixture with coal gas (composition given in the original) to ap-
gas . If the air contains less than 11 .5 percent proach and, in a 2-inch-diameter tube, to mee t
oxygen it cannot form a flammable mixture when 11 .0 percent of the mixture is methy l
with coal gas ; if the oxygen content is reduced bromide (38) . For a . comment compare th e
by the addition of carbon dioxide, then when corresponding paragraph on hydrogen (p . 22) .
the mixture contains more than 31 percent car- The limits of coal gas in mixtures of air an d
bon dioxide, and, consequently, less than 14 .4 methyl bromide have been tedl by the
percent oxygen it cannot form a flammable . ,, - method given on , jlige .5 Igo the limits of
mixture with coal gas . These results are in its components in air and in mixtures of ai r
dose agreement with the analytical figures of and methyl bromide . The agreement with th e
an old observation (90) . These conclusions experimental results '4 '+g~u d,for lower limits,
may be fairly representative of a normal coal moderately so for higberlimifii (38) .
gee, but they are based on observations of flame Atmospheres of Oxygen and Carbon Di -
traveling downward in a narrow burette ; in oxide.—The limits of coal gas in an atmosphere
wider vessels, especially when ignition occurs composed of 21 percent oxygen and 79 percen t
near the. lower pert of the vessel, the oxygen carbon dioxide, in a 35-ac. . bulb with side !pi-
figures would doubtless be several percent less . Lion, were 7 .6 and 25.2 percent. At 300 C.
Atmospheres of Air and Water Vapor .—Tbe the limits were 9 .3 and 182 percent . The
limits of mixtures of a typical coal gas and air composition of the gas used is given under th e
standing over water in a 350-cc . spherical ves- heading Coal Gas in Oxygen (297) .
eel, and ignited near the 'water surface, have
been determined at various temperatures . As COAL GAS AND METHANE
the temperature rises, and consequently the The lower limits of mixtures of coal gas an d
water-vapor content also, the lower limit rises methane, with downward propagation Of flame r
slowly and the higher limit falls rapidly, as can be calculated front the limits of coal gas
with other diluents. When about 42 percent and methane separately (218) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF MIXED FLAMMABLE OASES AND VAPORS

COAL GAS AND WATER GAS
Figure 61 shows the limits in air of all mix-
tures of a typical coal gas and a water gas, with
downward propagation of flame in a Bunt e
burette (S0) . The abscissas represent the corn ..
position of the gas in terms of its calorific value .
The addition of water gas to coal gas widen s
the range of flammability and consequently in -
creases the chance of an explosion. With up-
ward propagation of (lame the range woul d
doubtless be wider at each limit by severa l
percent.
MINE-FIRE GASES AND GASES FROM MIN E
EXPLOSION S
Although mine-fire gases and atmosphere s
after explosions vary greatly in compositio n
their limits of flammability may be calculate d
with approximate accuracy by the method give n
page 5. Examples (133) are given in tabl e
40,
If the atmospheres contain methane and
negligible quantities of hydrogen and carbo n
monoxide, approximate limits can be obtaine d
from figure 22 (p. 47) .

TABLE 40 .--Typical samples of mine-fire atmospheres . in coal mines

Composition of atmosphere, percent by volume Limits of flaintnability

Sample No .
Lower Higher

13 8 2. 8 L. 3 3. 3 70. 9 45. 1 70. 4

I. 8 It 1 1. 0 7. 2 7a 1 69. 2 100. 0
.5 17 . 4 0 12. 6 69. 5 40. 1 100. 0
1. 3 17. 0 1. 9 4. 0 75, 8 (9)30. 5
6. 4 I. 4 1. 1 16. 7 74. 2
5. 9 1. 4 I. 4 22 3 69 0 23. 1 034103
26 8. 8 0 a1 80. 5 (e)
Noneammable.

AUTOMOBILE EXHAUST GAS propagation and 43 .9 and 67 .8 percent with

downward propagation of flame (269) .
The usual adjustment of an automobile car- The composition of the gases used in the
buretor gives an air:fuel ratio of about 12 .5: 1 preceding experiments was, in percentages :
by weight, and the exhaust is nonflammable .
If, however, the air :fuel ratio is reduced belo w a-Menoe _ (111) ia) is11)
about 11 .7 :1 the exhaust is flammable, an d
there is danger of an explosion as well as o f (d (11
carbon monoxide poisoning when the engin e
is run in a poorly ventilated place. The limits Carbon Morldo I& M „d 1Ja is • U
n . raen :1
of flammability of the exhaust gas may. be eel- Carbon monoelde A 70 Xi s6 , . x7.1 1s •
culated with approximate accuracy by th e Methane . .30 ~ 10 1 •
H 'Amgen— . .. .. . 4 70 i ai 17 L4
method given on page 5 (131, 183) . Nitrogen_ UAW 17.11 i.1 01 .0

BLAST-FURNACE GAS PRODUCER GA S

The limits of two typical blast-furnace gases ,
with upward propagation of flame in a glas s
tube 2 inches in diameter and 6 feet in length ,
open at the firing end, were 36 and 72 percen t
and 35 and 73 .5 percent . These figures agree
approximately with tho90 calculated from th e
quantities and limits of the component gases
by the method given on p . 5 (133) .
The limits of a blast-furnace gas in a close d
tube 3 inches in diameter and 81i inches i n
length, with central ignition, are 45 and 6 5
percent (.43) . In a Bunte burette the limits
were 35 .8 and 71 .9 percent with upward
The limits of producer gas in a Bunte
burette were 35 .5 and 80 .0 percent with up -
ward propagation of flame and 40.6 and 76 . 5
percent with downward propagation (269) .
The limits of another sample of producer ga s
were 20,7 and 73 .7 percent in a small bulb com-
pared with 20 .2 and 71 .8 percent calculated by
the method given on page 5 but based on limits
of the individual gases in larger apparatu s
(367) . A sample of nearly the same composi-
tion, tested at a later date in a similar apparatus ,
gave similar figures : 19 .9 and 72 .5 percent
f368) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


Ix<

MITTS OF FLAMMABILITT OF GASES AND VAPORS

The limits of a "theoretical producer gas, "

with downward propagation of Came in a Coapo *Uon cwt C.a .t wee
(►1
wate r du:;w
no-
M (il pelt) (11 pa 01 M 1
' Bunts burette, were 39 .8 and 69 .6 percent
(270). Carbon 0io gds- _. 1.0 .2
0 t• a0
t1 a0
The compositions of the various samples . of ae7.tut .w bfdm-
producer gas used were : waves 4 .4 a0 a6 ~
7.0 .0 . 4 4 .4 t
Carbon mm~wW ._ . .
Carbon .0
0 t0 17.0 4Q 4 14.1 16. 1
n rdr*Req 47. 6 42 4 43 .4 43.0 K4 .0
10
aelere°w (t6/ 1 (x17) nu t (510 ) Alttbene 77 .0 77.0 Ito . . .. .. . . . . . .. . . . . . . . . . . .
1 Nampa 7. 6 6 II 11.0 120 7x6 66. 1

Carbon dtoddt -- 26 13 6.0 . . . . . . .. . . too or limn . . . t6 6.6 la 4 13 . 6 It 0 31 .0

Otlten 0 .. .. . . . .. Hl5btr0atlt .. .. $0 10. 0 at II LL6 74
.0 74
.0
Hydrogen : 4 .6 17:4 14.6 :..
Carbon aln°oslde 30 .0 77.1 37.6 34 . 0
Methane . .7 True .. . . . . . .
Nltrcgea 6t. 7 S7.4 51 .1 H.3 COKE-OVEN GAS AND AMMONI A

The very marked differences in the lowe r
limits are partly due to the differences in
experimental' methods, but mainly to th e
variation of hydrogen content between th e
samples.
PRODUCER GAS IN OTHER ATMOSPHERE S
Atmospheres of Air and Water Vapor .-Tb c
limits of mixtures of producer gas (compositio n
given in previous table under reference 368)
and air, standing over water in a 350-cc . spher-
ical vessel and ignited near the water surface ,
have been determined at. various temperatures .
As the temperature rises, and consequently th e
water-vapor content also, the lower limit rise s
slowly and the higher limit falls rapidly .
When about 43 percent of water vapor i s
present the limits coincide at about 27 percen t
of producer gas (368) .
The limits of a gas of unspecified composition ,
perhaps in a audiometer tube, are given as 6
and 13 .4 percent (120) .
. carbon" (isolicptape and dititethyl cyclohesane )
The limits of a mixture of 75 percent oil gas
and 25 percent acetylene' in air, observed per -
haps in a. audiometer tube, are given as 4 an d
15 .5 percent (120) .
VARIOUS FUEL GASES
The limits of "explosihility" of various fuel
;bases in air were determined in a tube 4 cm .
^n diameter and 79 .6 cm . in length, one en d
being permanently closed and the other sealed .
with a diaphragm of varnished paper . Ignitio n
was obtained b a coil of iron wire that was
caused to glow by a 10- to t5-ampero current .
The results were (103) :
The limits of some mixtures of coke-oven gas
and ammonia were found to be nearly in ac-
cordance with Le Chatelier's law ; but the
mixture used as synthesis gas for the production
of IICN had a somewhat greater higher limi t
than the calculated figure (280) .
ETHYLENE AND (1) ETHYLENE DICHLORIDE,
(2) ETHYLENE DIBROMIDE, (3) n-BUTY L
BROMIDE
The limits of these pairs of mixtures in air,
for downward propagation of flame in a tube
1 .6 cm . in diameter, have been determine d
(217) .
SOME HYDROCARBONS : ETHER ; ETHYL
CHLORIDE AND BROMIDE ; DIETHYL SEL -
NI .DE; TIN TETRAMETHYL; LEAD TETRA -
METHY L
Limits of the following pairs of mixtures
in various proportions, have been determineJ
with downward propagation of flame in a tub e
5 cm . wide endp 65 cm . long : Methyl cyclo-
bexane and ethyl chloride (246) ; "hydro-
and ethyl bromide (246)- "hydrocarbon" (iso- -
heptane and dimothyl cycfopcntane) and diethy l
sclenide, also with tin tetramotbyl and lea d
tetrametbyl (24fT'A ebh?I-- ether and ethy l
bromide (246) ; ethyl ether and methyl iodide
(246) ; ethyl ether and tin tetramethyl (247) ;
and ethyl ether and lead totramethyl (247) .
The results are of the general type shown i n
figure 4 (p. 13) end were interpreted similarly .
CYCLOHEXANE; BENZENE ; ETHYL ALCOHOL
The lower limits of mixtures of cyclohexane,
benzene, and ethyl alcohol (iA pairs or together)
agree fairly well with Le Chatelier's law at
temperatures of 100° to 250° C . (21, 23) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 TOLUENE, ETHYL ALCOHOL; ETHYL
ACETATE
The lower limits of mixtures of toluene ,
ethyl alcohol, and ethyl acetate (in pairs or
together), with upward propagation of flam e
in a closed tube 10.2 cm . in diameter and 96
cm . in length, agree fairly well with the figure s
calculated by Le Chateher's law (139) .
Influence of Temperature.—Under the same
conditions, the lower limit, of a mixture contain -
ing 39 percent toluene, 6 percent ethyl alcohol ,
and 55 percent ethyl acetate fell from 1 .76 per-
cent at 24° C . to 1 .44 percent at. 200° (139).
METHYL CYCLOH CANE; ALCOHOL ; ETHER
The lower limits of mixtures of methyl
cyclohexene, ethyl alcohol, and ethyl ether
in pairs and together, have been determined
with downward propagation of flame in a tub e
5 cm . in diameter and 70 cm. in length . Th e
results agree with those calculated by L e
Chateher 's law (319) .
GASOLINE; ALCOHOLS; ETHER
Series of observations with ethyl alcohol-
gasoline-air and ethyl alcohol-gasoline-ether-
air mixtures in a 2,300-ce . vessel at 50° and
90° C . showed fair agreement with L e
Chat.elier's law. The results, expressed i n
cubic centimeters of liquid vaporized in 1,00 0
cc . of air, are probably too high throu ghout,
as the result for pure alcohol was undoubtedl y
too high (20) .
In the same apparatus the limits of isopropy l
alcohol-gasoline mixtures nearly obeyed Le
Chatclier's law (231) .
BENZENE AND TOLUEN E
The lower and higher limits of mixtures o f
benzene and toluene in air in a closed tube .
5 cm . in diameter and 150 cm . in length, with
upward, horizontal, and downward propagatio n
of flame, agreed closely with the values cal-
culated by Le Chatelier s formula (355) .
BENZENE AND ETHYL ALCOHO L
The lower limits of mixtures of benzen e
and ethyl alcohol in air, with upward propaga-
tion of flame in a tube 5 cm . rn diameter and
150 cm . in length, .open at .the firing end ,
deviate somewhat from the values calculated b y
Le Chatelier's formula (164) .
Influence of Temperature .—Tho lower limits
of mixtures of benzene and ethyl alcohol agre e
fairly well with Le Cbatelier's law at. tempera-
tures from 100 0 to 250° C. (21, 23) .
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
"ESSENCE TOURISME"
Curves A and B in figure 62 show the lower
limits of these mixtures, respectively, wit h
downward propagation of flame in a 2g-lite r
bottle at various temperatures ($66) . The ab-
scissas represent cubic centimeters of liqui d
vaporized per liter of gaseous mixture at th e
temperatures of the experiments. From th e
density of the liquid mixtures it would be,possi-
ble to calculate the volumetric composition i n
the gaseous state . The curves for limits draw n
from the calculated figures would be more nearly
vertical ; that is, the lower limit would be les s
affected by change of temperature than th e
curves might suggest at first sight . Thes e
curves help to explain why it is more difficul t
to start a cold internal-combustion engine whe n
the fuel is mainly alcoholic than when it i s
mainly a light hydrocarbon mixture . .
The benzoic used in the above experiment s
distilled between 78° and 110° C . (86 percent
below 90°) ; 87 percent was benzene. The es-
sence distilled between 58° and 179° C . ; 92 .3 5
percent was saturated hydrocarbons and the
remainder aromatic hydrocarbons ..
The lower limits of a mixture of benzene ,
methyl alcohol, and ethyl alcohol at 100° to
250° C., with downward ;nropagation of flam e
in a 2);-liter bottle, agreed with the limits cal-
culated by Le Chatelier'a law within 11 per -
cent (81) .
The lower limits (expressed as cubic centi
meters of liquid evaporated in 100 liters of th e
mixture) of various French gasolines and mix-
tures, with downward propagation of flame in
a 2)4-liter bottle (21), are given in table 41 .
METHYL ALCOHOL AND ETHYL ALCOHO L
Lower and higher limiAotmixtures of methy l
and ethyl alcohol in downward propa-
gation of flame in a niter cylinder at 75° C ., ..
agreed with the values calculated by Le Chate-
lier's formula (.t'1~ h~owever, all the values ar e
probably too hig6:"' 'Anather observer ' (365 )
found that the lower limits of a 50 :50 mixture
of the two alcohols were 0 .7 to 0 .3 less than
those calculated over the range 50° to 250° C.
under similar conditions.
Influence of Water.—The lower limits of mix-
tures of water and the two alcohols rise steadil y
as the quantity of water increases from 0 to 6 0
percent by weight, but the quantity of th e
mixed alcohols is approximately constant in the
limit mixture. With 80 percent of water it wa s
difficult to inflame any mixture of the vaporize d
liquid and air at 105 C . and with 85 percen t
inflammation was virtually impossible (1t80) .


LLM1TS OF FLAMMABILITY OF GASES AND VAPOR S

TABLE 41 .-Lower limits of carious French gasolinis and mixtures

Gasoline, Tourisme" gasoline

Aviation
Gasoline and bensole mixtures gasoline,
etbylated , 7 h ol.
alco
25 " Esso" "Amu"'
Tempera-
Lure, • C.
Prortions
po . ._ . 80 20 60 : 40 10 : 60 . . : ^ _
d --r 0. 750 0. 779 0. 808 0. 725 0. 738 0. 741 0. 756
Octane no— 73 82 93 80 91 78. 5 80

100 4. 35 4. 15 3 95 4. 55 4. 85 4 . 55 4. 60
150 ________________ 3. 55 3. 45 3 35 1 80 t 15 3 . 70 1 80
200 3. 00 2. 95 2 75 3. 15 3. 50 3 . 00 . 3. 1 5
250 ________________ 2. 55 2. 40 2 25 2, 55 2.75 .50
2 •2. 55

rum al
Dal

Na
Iliumm
am
II Iwrr
Ink
man.
mum mama
s mrmnhm
A ! , I mam ,

0 0.02 0 .04 0.06 t1.0i8 0.10

LIQUID PER LITER Of GASEOUS MIXTURE . CC.
Fracas 62.—luflucnce of Temperature on Lower Lamita of Certain Liquid Fuels .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LIMITS OF MIXED FLAMMABLE GASES AND VAPOR S
Influence of Temperature.—Curve C (fig. 62 )
shows the influence of temperature on the lowe r
limit of a mixture of equal parts (presumably
by weight) of methyl and ethyl alcohols (366) .
The curve would be less steep if the results wer e
based on abscissas representing the volumetri c
composition of the gas mixture .
METHYL ALCOHOL . AND ETHER; METHYL
ALCOHOL AND ACETONE; ETHYL ACE-
TATE AND BENZENE; ACETALDEHYDE AND
TOLUENE; ETHYL NITRITE AND ETHE R
The lower limits of methyl alcohol-ether ,
methyl alcohol-acetone, ethyl acetate-benzene ,
acetahlchyde-toluene, and ethyl nitrite-ether
mixtures in air were determined in closed tube s
5 cm . in diameter and 150 cm . in length . Th e
results agreed closely with Lc Chatclier's law .
No figures are quoted in the original communi-
cation (353) .
For mixtures of methyl alcohol, acetone, an d
air the law has been shown to apply to highe r
as well as lower limits by experiments in a
frlrr~
111 high 'te
MKIMIMut
Delayed Combustion
1IME
'' '
Strong combustion
Violentt
combustion's
`)MIR
011111111. ~NM
/'"
ER
1) ■
~ Aldehydic combustion
t B 16
CARBON DIOXIDE. PERCENT
Fluent a3.-Limits of Flammability of Eolvicet (Ether : Alcohol= 3 : I) is Mixtures of Mr and Carboa Dioxide.
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
closed 2-liter spherical flask with ignition in the
lower part of the flask (76) .
ETHYL ALCOHOL AND ETHER; ACETONE
AND ETHER
The lower limits of ethyl alcohol-ether-air
and acetone--ether-air mixtures in closed tube s
5 cm . in diameter and 150 cm . in length agreed
well with the values calculated by Le Chatelier'a
formula with "upward, horizontal, and down-
ward propagation of flame . The higher limit s
agreed closely with downward propagation, bu t
with upward or horizontal propagat oi n they
differed greatly from the calculated values ;
these differences are ascribed to the irregula r
intervention of the "cool flame" of ether .
Details of the results are given in the original
refe rences (319, 356, 361) .
Large-scale experiments have been made with
a mixture of ethyl alcohol and ether in th e
proportion of 1 : 3 by weight, vaporized into
air nod into mixtures of air with carbon dioxid e
and with nitrogen . The experimental vessel
er II
a
~+d,~~
1► . ■~
Weak combustio n
No combustion, with inflammation
observed
Weak combustion
24 32
.
 was a section, 3.8 meters in length, at the en d
of a horizontal tubular steel gallery 20 meter s
in length and 1 .68 meters in diameter . Various
types of ignition were employed, all at th e
center of the experimental section . One en d
of the section (the end of the gallery) wa s
closed permanently ; the other was closed by a
paper diaphragm . The nature of an ignitio n
was judged by the sound, appearance of flame ,
bursting and burning of the paper diaphragm ,
pressure recorded on a manometer, and aide-
hydic odor that was sometimes developed .
The nature of the results recorded when gun-
cotton was used as the source of ignition of th e
vapor in mixtures of air and carbon dioxide i s
shown in figure 63 . When nitrogen was used
in place of carbon dioxide, the limit curve was
less steep and met the axis at about 38 percen t
nitrogen . With other means of ignition (ful-
minate, induction-coil spark, and fusion of a
tungsten filament) the limits for weak combus-
tions differed considerably from those in th e
figure . It is evident that, in spite of the lar e
volumes of mixture used, the limits for self -
propagation of flame were not obtained, doubt -
less because the local effects of the igniters coul d
not be distinguished, by any available means ,
from self-propagating inIlemi nation . J low ever ,
as a general description of the phenomena i n
relation to the experimental conditions, th e
results are of value (40) .
The lower limit of a 1 : 2 mixture of ethyl
alcohol and ethyl ether, with horizontal propa-
gation of flame throughout a pipeline abou t
100 meters in length and 45 cm . in diameter ,
is 43 grams per cubic meter (289) ,
ALCOHOL, CHLOROFORM, AND ETHE R
The mixture of alcohol, chloroform, an d
ether commonly used to induce anesthesia give s
vapors that form explosive mixtures wit h
suitable proportions of air (60) ,
The lower limits of two mixtures (3 : 1 and
1 : 3) of ethyl alcohol and furfurel in air, wit h
upward propagation of flame in a tube 5 cm . i n
diameter and 150 cm . in length, open at th e
firing end, deviate somewhat from the values
calculated by La Chatelier's formula (164).
ETHYL ALCOHOL AND ACETONE
The lower limit of a 4 : 1 mixture of ethyl
alcohol and acetone in air, with upward propa-
gation of flame in a tube 5 cm . in diameter an d
150 cm . in length, open at the firing end, wa s
nearly in agreement with the value calculated
by Le Chatelier's formula (164) .
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
ETHER AND ACETALDEHYD E
The lower limits of mixtures of ether and
acetaldehyde in air in closed tubes 5 cm . in
diameter and 150 cm. in length showed goo d
agreement with the values calculated by L e
Chatelier's formula, with upward and down -
ward propagation of flame ; also the hi&e r
limits, with downward propagation . The
higher limits with upward propagation ar e
definitely lower than those calculated. Both
ether and acetaldehyde give rise to "cool
flames" in the richer mixtures, and the higher
limits reported are those of the "cool flames . "
The propagation of the "cool flame" by eithe r
constituent is not assisted by that of the other
to the extent that an additive law would indicate
(956) .
ACETONE AND METHYL ETHYL KETON E
Observations with mixtures of acetone an d
methyl ethyl ketone in air in closed tubes 5
cm . in diameter and 150 cm . in length showed
close agreement with the values calculated b y
Le Chatelier's formula for both limits, wit h
horizontal and downward propagation of flame .
l:'ith upward propagation, the lower limits were
:lightly higher than those calculated ; th e
'.riglier limits were several tenths of 1 percen t
higher than those calculated (956) .
PARAFFIN-HYDROCARBON HALIDE S
The limits of the following mixtures have bee n
determined in air in a 25s-liter bell jar, with
electric-spark ignition : Methyl chloride an d
ethyl chloride, methyl chloride and methy l
bromide, and methyl bromide and ethyl chlo-
ride . All mixturest of each pair are capable o f
violent explosion when mixed with appropriat e
amounts of air . The observed limits differ ap-
ppreciably from those cal c,,,ii .d by La Chats-
lier'a formula (192) . .
SOME MDIED SOLVENTS FOR LACQUER S
Table 42 gives lower limits, in grams pe r
liter, of mixtures of hydrocarbons, esters (active
nitrocellulose solvents), and alcohols . The
limits were determined with upward propaga-
tion of flame in a dosed tuba 25 cm . in length
and 2 .5 cm . in diameter, with central ignitio n
(894)
MISCELLANEOUS MIXTURES CONTAININ G
FLAMMABLE GASES
Table 43 includes mixtures not mentione d
elsewhere in this bulletin and containing various


LIMITS OF I'fLXED FLAMMABLE OASES AND VAPORS 12 9

proportions of flammable constituents . Their whether they can form flammable mixture s
compositions are stated, together with the with air, and if so in what proportions the mix-
results of calculations (p . 5) which show tures are flammable (137).

TABLE 42 .-Compaction and limits of mined solvents

Composition I 2 3 4 5 6 7 8

Benzene
Toluene
50 60 20
40
........ .....
60
_- .
60
20
40
........ .......
30 30
V. M . P. naphtha ----° --------------- - ---------------- --------------- - 30 . . . . . . ..
Ethyl acetate SO 30 15 15 . . ..... 10 . . . . .... .......-
Butyl acetate 15 . . . . . . . . 15 10 . . :
Butyl propionate
Ethyl alcohol
Butyl alcohol
-------- 10 5
5
15
5
5
........ ........ .. ...... ........
15
10
10
10
30
10
40
30
Loner limit, grams per liter - 0. 070 0. 057 0. 071 Q. 098 0. 082 O. 062 0. 069 0. 077

TABLE 43 .-Composition and limits of flammability of miscellaneous gase s

Limits of
Composition, percent flammability,
Sam- percen t
plc Type

CO: 0, H, CO CH, CM, Lower Higbes

I Soil Ras 10 .6 0.3 1 .0 0.2 0 3. 2 0 78. 7 (I) (I)

2 do ^ 9. 6 0 5 .9 0 0 .2 0 8t3 (7 (' )
3 do i 6. 8 .7 6.0 0 0 10. 5 0 76.0 49. 0 55. 0
4 Soil+natursl gas ._ . 8. 0 .2 5. 9 0 0 IS . 4 5. 0 62. 5 20. 0 39. 0
5 - . .- .do .6 0 6. 7 0 0 52 .5 12. 1 28. 1 7.0 22. 0
6 Sewage gas 25. 2 0 .1 0 0 67. 8 0 6. 0 7. 5 19. 0
7 Fermentation gas, cellulose_-_ .- 79. 1 0 0 6. 2 0 It 7 0 0 26. 5 38. 5
8 Fermentation gas. grain 65. 4 0 0 32. 5 0 .1 0 2.0 13.0 60. 0
9 Gases from heated insulation .- I. 9 17. 4 0 28. 5 2. 2 45. 3 .3 t4 4. 5 2l . 5
10 Gases from burning wood . . 6. 4 .6 9. 4 0 3. 5 3 0 79. 8 (t) (t )
11 Gases from burning paper 7. 6 1. 2 7. 7 0 6. 2 0 77. 3 (') P)
12 Soil+coal gas 3 .0 1 .2 12. 2 7. 1 4.4 1I0.8 2.2 58. 1 18.0 66 5
13 Gases from explosives, TNT 6. 4 0 0 26.9 53.2 2 0 13.3 9.5 • 72. 0
14 do 1. 5 0 0 19. 9 60. 7 .3
2 0 •~1 10. 0 07.0

NocAei sb1s .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 Tables 44 and 45 contain selected values fo r
the limits of flammability of single gases and
vapors and of some industrial mixtures in ai r
and in oxygen at ordinary temperatures an d
pressures. Table 46 contains selected limits
in nitrous and nitric oxides
The last two columns of tables 44 and 45
ive the oxygen percentage, in mixtures o f
5 nmmablo gas (or vapor) and air with (a )
nitrogen and (6) carbon dioxide, below whic h
no mixture is flammable .
The values in boldface observed with upward
propagation of flume in large vessels, open at thei r
lower ends, are chosen as those most useful for
reasons given earlier . The other figures represen t
experiments with mixtures contained in closed
or small vessels ; they are therefore not appli-
cable with certainty to conditions in whic h
normal pressure is maintained during the
passage of flame . Reference to the text wil l
define more exactly the conditions of observa-
tion .
Many of the higher limits . and a few of the
Time 44 .-Summand of limits of flammability of individual gases and rapers in air and in oxygen
Limits In air, percen t
Gas or vapor
I .o .rer Higher
INORGANI C
Hydrueen 4. 0 75
Deuterium 5 75
Ammonia 15 28
Ifydraxine 4. 7 100
Hydrogen sulfide 4. 3 45
Hydrogen cyanide 41
Cysnogen 32
Carbon disulfide 1 .25 . . . . . . . SO
('arbors oxyaulade 29
Carbon monoxide 12 .5 74
Chlorine monoxide 23.8 100
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
lower limits, of the vapors of liquids could only
be determined by raising the temperatur e
sufficiently to raise the vapor pressure to th e
extent necessary to reach the limit composition .
The limit given, therefore applies to the experi-
mental temperature, which is quoted in the
text if it is stated in the original source of th e
information .
Further information about the limits o f
gases and vapors will be found in the text ; the
data are not suitable for inclusion in the table s
hut may be useful if the conditions of determina-
tion are kept in mind .
Information on limits in other atmosphere s
and on the influence of pressure and tempera-
ture and the effect of turbulence will also be
found in the text.
The limit figures in table 45 apply only to
particular samples ; analytical data will he
found in the text. By the use of Le Chateliers
law the limits of similar mixtures can b e
calculated .
Limits in oxygen, percen t
Carbo
Nitro- dioxiden
Highe r gen sa as dilu-
diluen t
of air ent
of air
 10934953 9

'so

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


SUMMARY OF LIMITS OF FLAMMABILITY 13 1

Ts.w~E 44 .-Summary of limits offlammabil of individual gases and vapors in air and in oxygen-
ontinued

Oxygen per-
Limits in air. percent centage below
Limits in oxygen, percen t which no
mixture Is
flammable
Gas or vapor

IS
15

13. w

---' GO 13. w
--- ' --- '
--- ' ~~~~~ ~~~ ~
--- - ---'
--- -

.' Methyl alcohol — 7. m mv nm ._—_ . . .__ .--_ ._ ._] .nm m

~ 4. * ~u/~ .---~_—_~— _
2.x xs
~m .1 .__ . .~-- _
vs-Butyl''l
Amy
alcohol .—_ .— . . . -
. . . . .. .. . .... . . ... ....
~v .J
1. 2
.__.` . ._- -
__ _
l Furfuryl alcohol
Allyl alcohol 23 1~0 3
Propylene glycol ~w
. .l mx
. . . r. . . . . . . . . . . . . .
Trlethyleoe glycol

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


132 LIMITS OF FLA .JMABJLITY OF GASES AND VAPORS

TAat,E 44 .-Summary of limits offlamrnabilil t of individual gases and vapors in air and in oxygen-
--Continued

Oxygen per-
eentap below
Limits in air, percen t Limits In oxygen, percent which no
mixture Is
flammable
Gas or vapor
Carbo n
dioxid e
Lower Higher as dilu-
en t
of ai r

ZTBEtta

Methyl ether 18 27
Ethyl ether 1. 7 18
Ethyl n-propylether__ 24
Isopropyl ether 1.3 21 23 -- -
Vinyl ether 1. 7 27
Ethylene oxide 3. 6 80 10 0
Propylene oxide 2 0 21 . 5 --
Dioxins 2- 2-
Trioxane 3. 6 20
Acetal . .. .. . . 10.4
Methyl cellosolve 19 . 8
Ethyl oellosolve 1. 8 I4. 0
Butyl cellosolve 1. 1 10. 6
Diethyl peroxide 2.3
AUDEUTDE 5

Acetaldehyde
Paraldehyde
Buty rtdehyde
Acroleio
Croton aldehyde
Furfurei
an-one s
Acetone 2 5
Methyl ethyl ketone . I. 8
Methyl propyl ketone 1. 5 8
Methyl butyl ketone 1. 3 8
Methyl Isobutyl ketone 1. 4 7. 5
Cyclobexanooe 1. 1
lsophorooe .8 3. 8
ACID ; ANHYDRIDE S

Acetic acid 5. 4
Acetic anhydride 2. 7
Phthalic anhydride .. .. . . . 1. 7
ESTEaa

Methyl formate 5. 9 5. 0 23
Ethyl formate 2. 7 16. 4
Butyl formate .. .. ... L7 8
Methyl acetate 3. 1 16
Ethyl acetate 2. 8 2. 2 31
Vinyl acetate Z, 4 ..
Prepyl acetate 2. 0 1.8
Isopropyl acetate_ ._-__ .-__ 1. 8
Butyl acetate 1. 7 1. 4
Amyl acetate 1. 1
Methyl cellosolve acetate _ 1. 7
Methyl propionate 2.4 5
Ethyl propionate 1.85
Methyl lactate 2. 2
Ethyl lactate 1. 5
Ethyl nitrate 4. 0
Ethyl nitrite 4.1 3. 0

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


SUMMARY OF xIAUTs OF FumuMuRumzz

133
TABLE 44 .—Summary of units of flarnmabilitj of individual gases and vapors in air and in *xygen_
onttnued

Oxygen per-
multiage belo w
Limits in air, percent Limits in oxygen. percent 'which no
Ift ixture is
flammabl e
Gas or vapor

cm~v~ ._°o~wu~__-- . --- ' Lu ____ ____ ___ _

LMIPCE3 AND MIME S

Methylamine 4.9 20. 7 -_-- .

Dimethylarnioe 2. u 14. w ___ _
Trimethylamine 2. 0 w ___ _
Ethylamine 3. m 14 . 0 ___ _
Diethylamine ,. o 10. 1
Triethylamine 1. m ' 8. u
- ---- - 2. 0 —~~[~ --- ' ___ _
n-Butyl amine 1. 7 m a ___ _ --- '
All y 2.2 22 --- ' . ......
Ethylene iml~~~~~~~~~~~~~ ~ 3. m 46 --- ' .. . ...
OTHER NITROGEN COM-
POUND S

... .. . . ... . ... . 3 . 0 ___ _ ___ _ 17 '-- - --- - --- ' '-- - . . . . . . . --- -
___ _ Lm --- ' cz o . .. . . .. --- ' . .. . . . . . ..... . ...... .......
___ _ .7 --- ' 4. m ___ _ ---' . . . .... '-- - . . . . . . . ___ _
HALOGEN DER1VAT1VR S

8 68
--- '
Methylene chloride *n

Ethylene dichloride

10 ___ _
--- ' ___ _ --- -
--' ^ . .... . .
Propylene dichloride ____ _
___ _

...... .
. .. . . . .
. .... . . ..... ..
- '--- ' . .... ..
- - -- -
s[~^ ! ___ _
so
... ... .. . . . . .
.... .. . -- - ___ _ . .... ..
---' --- - '-- - ___ _
.. ..... --- - . . . . . ___ _
wpSCIELLANXwn°
.
Dimethyl sulfide ._ —'- - '
_ ...... . --~— . ... .. .
Diethyl aelenide . 2. m ..... . .
Dimethyldichlorosilane . ___ _ ___ _
Methyltrichlorosilane __ _ --- ' ___ _
____ ~ p --- - . ..... .
—'--' Lw --'—' '- -

w18110'—EI--10

/
/

-.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 LLMI[TS OF FLAMMABILITY OF GASES AND VAPORS

TABLE 45 —Approrirnate limits of flammability of some complex industrial mixtures of gases an d

capors in air and in oxyge n

Oxygen percentage be -
Limits in air, percen t Limits in oxygen, percent low which no mix-
. turn is flammable

Gas or vapor
Nitroge n Carbon di-
'Lower Higher Lower Higher as diluen t oxide asdil -
ot air cent of air

Water gas 7.0 -^--•-^ 72 . . . . . . . . . . . . . . . . . . . . . . . . . . . . •

5. 5 . . . ... . 36 ....... .... .. . . . ... .. .... . .. .. . . ... .. .. . .... . .. ... ... '
Pittsburgh natural gas, 4.8 _______ 13.5 _______ _______ _______ _______ _______ 12. 0 14. 4
Other natural gases . 3 .8-6.5 __,._ .__ 13-17 _______ ___^•_ __-____ ___ .^_ __-•_-_ __^-_^ .. . .
Benzine 1 .1 _ - - - ---•-__ - - - --------- -
Gasoline 1. 4 _______ ------- 7. 6 _______ , . __-__-_ _--__-- IL 6 14. 4
Naphtha 0 .8 . . . . . . . 5
Kerosine - 7 ------- 5 -------------- ------- • ••• -
Coal gas 5. 3 .. . . . .. 32 . . .. . . . . . .... . 7 . . .. ... 70 11 . 5 14 4
Coke-oven gaa 4. 4 . . . . . .. 34 _______ _______ ______ _ _______ _______ ____ . . . .. .. . ..
Blast furnace gas 35 --- ------------------------------------------------------ -
Producer gas____, 17 20-35 70 70-80 ______________ _____ ___________________ _
------- ------- ----- . . . . . . . . . . . . . . . . . . . . . . . .
Oil gas 4- 7 33
------- ------- -

TABLE 46 .-Limits of flammability in nitrous and nitric oxides

Nitrous oxide Nitric oxid e

Gas or vapor
Lower limit Higher limit Lower limit Higher limit

Hydrogen 5 .2 . . .... .. 80 .... . .. . 13 .5 __

Ammonia 2 .2 72 ___---__ --• --------------- -
--------
Hydrogen sulfide -------- --•►---^ . . .... .. . .. . ..
disul5de__~ . ^_-_ _--___-- 1 . . .... . . 52 _—_____ 4 .5 . . .. .. .. 59
Carbon monoxide
. . . .. . . . .... . . ..... ..... . . .... . . . 2. 2 . ... .. . .
Butane 2.6 ________ 20 _ .____ . 7.5 12. 5
Ethylene 1. 9 5 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . ^•- -
Propylene 1 .45 . . . . . . . . 29 -_- . _-__-___________ :-
Cvclopropane•---•--•-• °---- 1 . 6 30 --------------------------------------- -
--------
Ethyl ether 1. S 24 ________ ^_ ___ ----- • •-- -
°---°---
Vinylether I. 4 25 ^__ ° . . . . . ... . . .. .. .. ..... ...
•--
Methyl chloride _______ _-______ ________ ______-_ _______ _
Ethyl chloride 2. 0 ... . .... 3.3 _--A .__ _ . _---____ .,_ - .__ . .- -

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF INDIVIDUAL GASES AND VAPORS 21

e = fie■■m
lam
llama
siiiasr Inimsmsa
t iz

' II
11111=1111INI ME
MI 11111111111WIMMI
-unaimEmiam
II na
f::bzt.===tts,z

amummIme GO
NYDPACIIt, PERCEN T
1W

Flamm 8.--Relation Between Composition end Flammability of Mixtures of Hydrogen, Oxygen. sod Nitrogen.

percent greater than those of figure 7 . The mixtures of air and carbon dioxide are shown in
"nose" of the cmve is at the same percentag e figure 7 . The determinations were made in t bi j r
: of additional nitrogen . The addition of 0 . 5 tube 6 feet in length and 2 inches in diameter,
cent of tin tetramethyl reduces the highe r with upward propagation of flame at atmoe-
Lit and retracts the "nose" of the curve plieric pressure during propagation (133,18?) .
considerably (318) . The limits for downward The limits with downward propagation o f
propagation in a closed tube 2 .2 cm. in diamete r flame in the same series of mixtures have bee n
have also been determined (341) . determined in a closed tube 3 cm . in diameter
' Atmospheres of Air and Water Vapor.--Th e and 65 cm . in length . The lower limits are 3 to 6
limits of hydrogen-air mixtures standing over percent greater and the higher limits 1 to 4
water in a 350-cc . spherical vessel, and ignited ercent less those m figure 7. The
near the water surface, have been determined at p'nose" of thethancurve is at. SB percent carbon
4'
various temperatures. As the temperature dioxide in the atmosphere . The addition of
rises, and consequently the water-vapor content 0 .5 percent tin tetramethyhreduces the..highe r
Also ; the lower limit ruses slowly and the h her limit and retracts the .!.'riose" of the curve -
limit falls rapidly, u with other t~tluenta . Whe n considerably (318) . The limit, with downward - ;
60 percent of water vapor is present (86 0 C. ) propagation in closed tubes 2 .2 .and 1 .6 nun, in
the limits coincide at about 10 percent hydroge n diameter have also besw.d termined (217, 341) .
(368) . Some earlier observations (95) show, as migh t
Earlier experiments, made in a Bunts burette, ' be expected, a more rapid narrowing of the
vets . show similar effects but the range of flamma- limits in a Bute burette. Others (1) maybe
. bility is smaller (95) . mentioned, but they can hardly be accepted
Atmospheres of Air and Carbon Dioxide . without confirmation because they intimate
1'he limits of flammability of hydrogen in al l several improbable conclusions--for example,
i snseo~-as-- a
f'

v4=

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


WtITS OF FLAMMABILITY OF GASES A,XE) VAPORS

that the lower limit of hydrogen is reduced fro m
6 .5 percent in air to 3 percent in an atmosphere
composed of air and 3 or 4 percent of carbon
dioxide .
Atmospheres of Air and Helium.—When thi n
• rubber balloons 2.5 inches in diameter wer e
filled with various mixtures of hydrogen and
helium and a lighted match or a white-ho t
platinum spiral was used to burn a hole in the
Fa"uric, the hydrogen could be raised to 26
percent before the mixture became flammable ,
but if the hydrogen exceeded 28 percent th e
mixture .woilfd burn . Hence, it was concluded
that to dilute the helium used for airships wit h
more than about. 26 percent of hydrogen (299)
would not insure safety against tnllanimation
of the gas .
Experiments by the present writers sho w
that homogeneous mixtures of helium, hydro-
gen, and air would propagate flame when th e
proportion of hydrogen to helium is much less
than that indicated above . When as little as
8 .7 percent hydrogen is present in admixture
with helium it is possible to nuke, with this
mixture, a blend with air that will propagate a
weal: flnnic up through the central part of a
tube 2 inches in diameter and 6 feet in length .
The complete sell of observations is shown i n
table 3.
TSOLE 3 .-- .Limits offfammobility of mixtures of
I< .gdrogcr. 'sad helium in air, upward propa-
gation- of lfame in tube 2 inches in diamete r
and 6 feet in length, open at its lower en d
Orlgtnet as mature, recent Lingua aemmebinir,pnant 2-inch-diameter tube, to meet when 35 percen t
)lydrogtn vellum Low., never
iceae o, oits er s s y drogen in this atmosphere, with u war d pro
ids
rs ,sea?r sss nt
as t+• ss
par ss) ai.s sea sus
r l s1 °
The mixture that contained 19 .3 percent
hydrogen and the rest helium had very wide '' lower limit is 13 .1 percent as compared with 9 .0
limits of flammability . Flame traveled rapidly
,and ,cith some violence through mixtures with,
. air that lay well within these limits, although
the balloon experiments had seemed to show and 70 .3 percent ; the tango is narrower in
that these mixtures were nonflammable . No
mixture would be safe for use in an airship
unless it contained less than 8 .7 percent limits were 11 .9 and 68 .2 percent hydroge n
hydrogen in admixture with helium (72) . .
with downward propagation of flame in a
closed tube 2.2 cm . in diameter the weakest
mixture of hydrogen and helium that can form
a flarunable mixture with air contains about
12 percent hydrogen (951), '
The maximum amount of hydrogen that
could be added to helium without, enabling the
mixture to burn when issuing into air from a je t
15 nun. in diameter, ignition by a gas flame, wa s
14 percent (223) .
Atmospheres of Air and Argon .—The limits o f
hydrogen in mixtures of air and argon, with
downward propagation of Home in at closed tube
2 .2 cm . in diameter, have been dcterrnined-
(34 1 )
Atmospheres of Air and Certain Halogenated
Hydrocarbons .--A series of results shoring the .
lower and higher limits of hydrogen in air con-
taining iurcreasmg amounts of trichlorocthylene
at 14', 25°, and 35° C. have been reported .
They were observed in small burettes 15-mm .
in diameter, so are of limited value (SPX)) . Sim-
ilar experiments with the vapors of other chlore-
derivatives at laboratory temperatures were
reported in an earlier communication (109) .
The addition of increasing amounts of methy l
bromide to the air causes the limits of hydrogen
to approach, and in e. 2-inch-diameter tube, to
meet when 13.7 percent of the mixture is methyl
bromide . The lower limit of hydrogen is, how -
ever, appreciably reduced (down to 2.4 percen t
with 12 percent of methyl bromide), showin g
that the bromide takes some pert in the com-
bustion (38) . The results are of doubtful appli-
cability to wide spices, for with a sufficiently t
powerful source of igustion seine mixtures o f
methyl bromide alone and air can be inflamed -
(p. 101) .
The addition of increasing amounts of di-
chlorodifluorometbane ("Pree n ") to the air causes
the limits of hydrogen to approach and, in a
of the mixture is keen (88) .
Atmosphere In Which Nitrogen of Air Is Re -
laced by Carbon Dioxide .--The limit,! of
nation of flame in a 5.3-em .-diameter tubeopen
at the lower end, are 5 .31 and 69 .8 percep t
compared with 4 .19 and 74 .6 percent, respect-
tively, in air. For downward rgpagaLion i n
the same tuba, the upper end being open diet
percent in air (94) . '
In a dosed 2 .5-ctn . .dialnetcrtubee with up-
ward propagation of flame, the limits are 4 .7 . i
narrower tubes (98) .
In a closed 35-cc . globe with side ignition, the
compared with 9 .4 and 64 .8 percent, respec- i
tively, in air in the same apparatus (297) . The
higher-limit result is unexpectedly greater in ,
the artificial atmosphere than in air ; a similar /
result, however, had been obtained earlier (846) .
Heimann have already been given (footnote 4
table 1, p . 17) for thinking that the figur
64 .8, for the higher limit in air, is too low.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


the same circumstances, but at room tempera- (65) . With a stronger spark the limits did no t
ture, the limit is 6 .1 percent. coincide until the pressure was reduced t o
Influence of Pressure at Various Tempera- 10 mm . (80) . A curve has been obtained
twes.—The curves of figure 17 show the limit s for results in a burette (56) .
of flammability of methane-air mixtures a t The higher limit is increased by an increas e
20 " , 250°, and 500° C. at all pressures belo w of pressure above atmospheric. One obser-
atmospheric . The limits were observed in a vation (330), incidental to other work, is that,
closed tube 2 cm . in diameter and 50 cm . i n at 10 atmospheres pressure, a mixture contain-
length (235) . ing 71 percent methane slowly propagate d
A mixture of 2 .1 percent methane in air was flame . In a small bomb (58) the higher limit
ignited by sudden compression to 80 atmosphere s rose rapidly from 58 .4 percent at 1 atmospher e
pressure, which produced a temperature of to 81 .7 percent at 60 atmospheres, then slowly
705° C., and a mixture of 55 percent methan e to 84 percent at about 145 atmospheres ;
at 118 atmospheres and 540° C . (87) . however, the mixtures contained about 4 .5
Influence of Impurities .—The lower limit of percent nitrogen, and the limits would be some -
methane in air, with downward propagatio n what, higher to pure oxygen . Combustion was
of flame was raised about 1 .3 percent by iro n far from complete in such mixtures under
carbonyl, (0.03 cc . liquid vaporized per liter) . moderately high pressure .
The higher limit was reduced from 13 .0 to Influence of Temperature .—In a 35-cc. closed
10 .5 percent by the same quantity of iro n bulb the limits were 6.2 and 57 .1 at 15° C.
carbonyl (325) . and 5 .1 and 57 .8 at 300° C. (297).
Influence of Temperature at High Pres-
sures .--As the temperature is raised, the pres-
sure required to make certain mixtures of meth =
The limits of methane in oxygen, wit h ane and oxygen flammable decreases .. For
upward propagation of flame in a 2-inch- example, the pressure limit of a mixture con-
diameter tube open at its lower end, are 5 .1 5 taining 81 .7 percent methane at atmospheri c
and 60 .5 percent (138) , temperature was 60 atmospheres, but at 332 °
'.Cable 15 gives other determinations of th e C. it was only about 21 atmospheres (58).
observed limits of methane in oxygen . Curves that show the higher limit at elevate d
temperatures and pressures are given to the
TABLE 15 .-Summary of other determinations of original paper .
limits of flammability of methane in oxygen
Upward Propagation of Flam m METHANE D4 OTHER ATMOSPHERES
Atmospheres of Composition Between Ai r
maebo eion or LIMO, pereen l
serrn os
and Pure Oxygen .—The limits of methane i n
?trt°o .od _ mixtures of nitrogen and oxygen richer in oxyge n
Na
Doe.Nr Length Lower Weer than ordinary air have been found as follows :
(1) Ina closed globe 2 .5 liters in capacity th e
1$ h o oven- IL 4 . .. .. .. . rn lower limit rose regularly from 5.8 percent in
air to 6 .0 percent in oxygen (267) ; (2) in a
Horizontal Propagation of flam e horizontal glass tube 2 .5 cm. in diameter, ope n
at the firing end, the lower limit fell from 5 .8
s. • n open Snail da in percent in air to 5 .7 percent in oxygen, and th e
o j s l
higher limit rose linearly from 13 .3 to air to
Downward Propagotiea of flame
59,2 in oxygen (189) ; (3) in a closed tube 1 . 9
lea o e at .. .. ... e
n
cm . in diameter the lower limit with downwar d
as
aa3 s ... .. ... . . . . . Closed . .. .. . . . a $W propagation of flame rose regullrl ,r from 6.1 5
. . . .. .. . . a4 _ M
f.a
E$ _. .w
... ..
.. . aN as percent in air to•6 .45 percent Mr oxygen .'- The
Z.7 _ ab ao . &7.2 . iv higher limit rose from 12 portent in air to 3 8
. . .do oa 1p
percent in a 62-percent oxygen mixture and 5 2
percent in a 95-percent ..Q Qx., yyggen mixture (323) .
Atmospheres of Air an3'Ni#rogen (Air De-
so mets c. q eased a o wr ficient in Oxygen) .—Large-scale experiments .
Use. ugwr a d o rar with mixtures of methane, air, and nitrogen hav e
le oa apsKttr da j. ._ n Sal so
— been made in a tube 7 feet in length and 1 0
inches in diameter, with upward propagatio n
Influence of Pressure .—The limits of methane of flame from the open end of the tube ; the
in oxygen were not appreciably narrowed unti l mixtures throughout were at atmospheric pres-
the pressure was reduced below 150 mm . A sure and were saturated with water vapor .
moderately strong igniting spark was use d The range of observations shown in figure 20

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF INDIVIDUAL OASES AND VAPOR S

111111111111
1111111 M
MI
111 ME II 111
■ rrMe
MIME
II UM
NM MEE
~1 ADDITIONAL HffROGEN DI ATTMOSPHERP. PERCENT
l
Fiooas 20.-Limits of Flammability of Methane In Mixtures of Air and Nitrogen (Experiments In Large Vessels) .

covers all compositions from air to mixtures that show narrower limits represent expert -
in which the amount of oxygen is too small for meats in various types of apparatus with
flame propagation, regardless of the amount of downward propagation of flame . One curve
methane present. The abscissas represent the shows greater values over part of the higher -
"atmosphere" in each mixture of air and limit range ; the values were obtained in a
nitrogen ; for example, 25 percent "additional closed vessel in which the pressure rose con-
nitrogen" means that the "atmosphere" used eiderabiy during the inflammation, and increas e
for the observations was composed of 75 percent of pressure is known to Wrease'markedly the
air and 25 percent nitrogen. Along the top of . . higher limit of methane. " The limits of thes e
the diagram the corresponding percentages of maxtures with downward propagation of flame
oxygen in the atmosphere may be read . It is in a closed tube 2 .2 cm . in diameter (341) and
evident that no mixture of methane is flammabl e in a closed glssarbal} .5,cm. in diameter (209)
at ordinary temperatures and pressures when have also been determinlil :
the atmosphere contains less than 12 .8 percen t For some purposes the results are more usefu l
oxygen and the .remainder is nitrogen (68). when expressed (62) as in figure 22. For e x
Ile results of observations made in smaller maple, it cannot be deduced from figure 2 0
apparatus are shown in figure 21, In a 5-cm . without calculations, that the mixture
.r
tube, with upward propagation of flame in a ~d
dry mixture, the limits nearly coincide wit h Methane
Oxygen 2
those plotted in figure 20. The three curves Nitrogen_ $a

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


I`

46 LIMITS OF FLAMLMABILUTT OF GASES AND VAPORS

cannot form an explosive mixture with air,
whatever the proportions used, whereas th e
mixture
Imam
Methane 9
Oxygen • . 12
Nitrogen 79
although not itself explosive, map form a serie s
of explosive mixtures with air. Figure 22 give s
this information at a glance .
Explanation of Figure 23.-Figure 23 explain s
figure 22 . The straight line AD (fig. 23) repre-
sents the composition of all mixtures of methan e
and pure air that contain up to 20 percen t
methane . No mixture of methane and air ca n
fall above this line, and all mixtures of methane,
air, and nitrogen must fall below it . The lin e
BE is the line of lower limits of flammability of
methane and CE the line of higher limits. A s
the oxygen content falls, BE and CE approach
each other until they meet at E. No mixture
which contains less oxygen than that corre-
sponding with Eis explosive per se, but all mix-
tures in the area BEC are within the limits of
flammability and are therefore explosive .
Next consider any mixture to the right of th e
line CEF; for example, the mixture represented
by the point G . Join CA . Then GA repre-
sents the mixtures formed, in succession, as C
is diluted with air . Because CA passes through
the area BEC the mixture, as it is diluted wit h
air, becomes explosive and remains so as lon g

O ; IN ORIGINAL ATMOSPHERE . PERCENT

' 20 19 19

~
17 15

~
15

1
14

I 1
13

4
L ,
I. Hempel pipette . 100 cr. capacit y 54
2. Tuba, 2.7 cm . diem. 225
S. Bomb. 3 .5 titers capacity 5 4
4 . Sphere, 2 liters capacity 32
3 . Tube, 5.0 cm. diem., upward propagation 59

1 EaMhUMM
MEGRI
Mna
alaMaTil
i

t
INIMMENAl
.
~r
am

0 5 10 15 20 25 30 —i3 40
ADDITIONAL NITROGEN IN ORIGINA
ORIGINAL L ATMOSPHERE. PERCENT .
a» 21 .-Limits of Flammability of Methane in Mixtures of Air and Nitrogen ; Comparison of Results Obtained Ii
la Smaller Vessels.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF INDIVIDUAL GASES AND VAPOR S

20
Mixtures which can not be
produced from methane and air
S■
38 ~

16

14 IiP __~

12 ~im ■

REIM
Capable of forming flammable with ai r
contain to much explode per se)

$ Not capable of
formin g flammable
6 mixtures with ai r

i1 1
0 2 4 6 8 10 12 14 16 18 20 i~
METHANE, PERCEN T
Ftoaaa 22.-Relation Beteeea Quantitative Composition and Flammability of Mistime* of Methane, Air, and .1 5
Nitrogen.

as its composition is represented by any poin t
on HJ.
It will now be clear that the position of FE " the greater—then they can be found from th e
is exactly defined by drawing a tangent from
A to the curve BEC and extending the tangen t
to meet the axis of abscissas in F, because th e
line joining any point above and to the righ t
of FE to A must pass through 11EC, while th e
line joining any point below and to the left o f
FE to A urust fail to pass through the explosiv e
region REC. FE is therefore the boundary o f
those mixtures capable of forming explosiv e
mixtures with air .
Figure 22 shows clearly that any mixturo of
methane and nitrogen that contains more tha n
about 14 .3 percent methane can form explosiv e
mixtures with air. If oxygen is present , a
correspondingly smaller amount of methan e
suffices
. Mixtures of methane, nitrogen, and oxyge n and the ratio AJ : JO is the ratio for the lower -
that are represented by any point in the are a
DCEF of figure 23 can form explosive mixture s
when mixed with air in suitable proportions .
If it is of interest to know what these propor-
tions are--end if there is a wide range of pos-
sible explosive mixtures the danger is so much
following considerations :
Suppose the mixture of methane, nitroggen ,
and oxygen is represented by the point Gi n
the area DCEF, then the straight line GA repre-
sents all possible mixtures of the original mix-
ture and air. As the original mixture is dilute d
step by etcp with air, the composition of th e
new' mixture is represented by points farthe r
and farther along GA. . T,he . air mixture firs t
becomes explosive per se SC tte point H,wher e
GA crosses EC'. The hi per limit of the original -
mixture is defined by this point ; the lower limit
of the original m uro Is defined by the poin t
J, nt which the GA ri ttrF.li'.
The ratio All : IIG is the ratio of original
mixture to air in the upper-limit. mixture,
limit mixture . Hence the limits of flamma-
bility of the original mixture are given (in per-
centages) by 100 AJ/AG (lower lumit) and 100
All/AG (higher limit) (62).

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF FLAMMABILITY OF CASES AND VAPOR S

\~ C

• ♦

I5 1 '
. ..

•
A
.\
1
%

ti Is

All Atmospheres of Oxygen 'and Nitrogen .— observations were made at temperatures above
Limits of methane in these atmospheres hive normal . Had it been possible to experimen t
been determined with downward propagation at normal temperature the' curve probabl y
of flame in a 1 .7-cm .-diameter tube (346). would have been a little to the right of th e
Atmospheres of Air and Water Vapor,- Ob- carbon dioxide curve over the lower-limit rang e
nervations that show the small difference in the and at the nose, but the two curves would have
limits of methane in dry air and in air saturated coincided over most of the higher-limit range.
with water vapor at laboratory temperatures Similar experiments hates been made- .in a
are quoted under Effect of Small Changes in - dosed 350-cc : spherical vessel with a "natural --
Atmospher c :Composition (p . 3) . gas" containing 07 percent methane, 3 percent.
The effect of large amounts of water vapor ethane. Similar results were obtained, with
on the limits of methane in air is shown in somewhat smaller-itnsite, which met at abou t
gure U . The determinations were made in a 6 .3 percent gas in a mixture containing abou t
tube 3 feet in length and 2 inches in diameter, 30 percent of water vapor (368).
faith upward propagation of flame at stint's- Atmospheres of Au and Carbon Dioxide .—
Aerie pressure during propagation (67) . For Figure 25 shows the limits of methane is mix--
each experiment the tube was, heated to the tares of airr and carbon dioxide saturated with
temperatur e necessary
.'nxoue t of wate r
the reured water
Hence, most of I the
'w . r The tests were made in a tube
7 feet in length and 10 inches in diameter, with

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


." .5

LIMITS OF INDIVIDUAL GASES . AND VAPOR S . 49

••iuiiIIii i
OXYGEN IN ORIGINAL ATMOSPHERE, PERCENT

, 0.11111111111111111111111111M II
malumomainnuo
:1111IMM1111111111111111111 1
:mmmommmum m
1 mmmoommumm m
mma ■
0

mmmmmmmums m
i-
* e

Immah iciMM 1g
M .i/'M r
M
s 10 IS -
OASES IN ORIGINAL ATMOSPHERE. PERMIT'
a s
1
Flow's 24.-Limits of Flammability of Methane in Separate Mixtures of Air with Carbon Dioxide, Water Vapor,
Nitrogen, Helium, &nil Argon-
.r

upward propagation of flame from the ope n
end of the tube . In these circumstances n o
mixture of methane with an atmosphere com-
posed of air and 24percent .or more carbon
dioxide is capable of pro agating flame far
from the source of ignition (70) .
Observations made in smaller apparatus are
shown in figure 26 . The results obtained in a
5-cm . tuba with dry gases show, in general, a
slightly wider range of flammability than thos e
obtained in al5-em . tube with gases saturate d
with water vapor (fig. 25) ; for example, th e
completely extinctive atmosphere in the 5-cm .
tube contains 25 percent carbon dioxide com-
pared with 24 percent in the wider vessel . The
four curves showing narrower limits represen t
experiments in various vessels with downward .
propagation of flame .
The limits of these mixtures with downwar d
propagation of flame in a tube 2.2 cm . i n
diameter (341) and in a closed glass bulb 6 . 5
an . in diameter (f09) have also been deter-
mined .
Influence of Temperature: The limits of fire-
damp in atmospheres of air and carbon dioxid e
have been determined from 100* to 600* C .
The firedamp used contained 73 .8 percent
methane and 26.2 percent nitrogen. The
observations were made in a horizontal glas s
tube 2 cm . in diameter and 100 cm . in length ; ' . 'I t
The results' are expressed in curves which aha w
the limits of methane in the various atmos-
pheres of air, carbon dioxide, and such addi-
tional nitrogen as was due to the firedamp itself .
The curves are of the same type as hose o f
figure 25 and show an increase in the range o f ?I1 ; 1
flammability of all mixtures as the tempera-
ture is increased (18).
Atmospheres of Air, Nitrogen, and Carbon
Dioxide (Including Mixtures of Air and Black-
damp) .—The limits of methane in' atmospheres - '
of air, nitrogen, and ca rbon dioxide are plotte d
in figure 27 ; they were obtained with roughl y
dried gases in a tube 2 ti ohes in diameter, wit h
upward . propagation Sf dime front•xhe ope n
end of the tube . The boundary at the righ t
of the series of curves gives the limits in all
atmospheres at rdinary air and carbon dioxid e
mixed in any pro' sons . The straight-line
boundary at the' eft of the aeries of curves I:
gives, with the oxygen figures inserted thereon ,
the limits in all atmospheres of air and nitroge n
nixed in any proportions . The whole area
between represents the limits in atmosphere s
which are deficient in oxygen and also contai n
carbon dioxide. When the deficiency of oxygen

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

50 LLMITS OF FLAMMABILITY OF GASES AND VAPORS
Os IN ORIGINAL ATMOSPHERE, PERCEN T
20 iS IS iT 16 le
14
4 r5 20 T.
CARBON DIOXIDE IN ATMOSPHERE, PERCEN T
FrecRS 23.-limits of Flammability of Methane i n
Mixtures of Air and Carbon Dioxide (Experiments
in Large Vessels) ,
is replaced by an equal amount of carbon diox-
ide, as would occur very nearly in tho com-
bustion of coke, these atmospheres and the cor-
responding methane limits are all represente d
by points on the hrnken curve .' 'rho regio n
right of the broken curve-represents limits i n
atmospheres formed by the replacement of
t oxygen with more than an equal volume of car -
bon dioxide . The region left of the broke n
curve represents limits in atmospheres forme d
by the replacement of oxygen with less tha n
an equal volume of carbon dioxide . Atmos-
pheres of this type are produced by ordinary
combustion, by respiration, and by mixin g
blackdamp and air (68) .
As the composition of blackdamp is variable,
its effect on the limits of methane is shown i n
figure 28 for a range of variations in composi -
- ' Sea Mirme, as for earlier oMenstkua, midi ins Beta Eames
iv°
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
Lion . These curves show that the lower limi t
of firedamp is increased only slightly by the
presence of blackdamp and that the danger of
an explosion of firedamp is not removed by th e
presence of blackdamp, unless the amount o f
blaekdanip is so high that the atmosphere i s
not fit to work in even if gas masks are worn .
An artificial supply of oxygen or air would be
necessary for work in an atmosphere of black-
damp and air incapable of propagating a n
explosion of firedamp (68) .
l"iguro 22 may be used to determine whether
any giren mixture of firedamp, blackdamp, and
air is explosive or can explode if mixed with a
suitable amount of air. The mixture is anal-
yzed to determine the proportions of methane ,
oxygen, nitrogen, and carbon dioxide present .
The carbon dioxide is assumed to be replace d
by an equal volume of nitrogen, and the infor-
mation required is read from the figure, as ex-
plained on page 46 ; in a coal mine the amoun t
of carbon dioxide in such mixtures is about 5
or 6 percent (often less, rarely more), and th e
error introduced by considering this as nitroge n
is small and on the side of safety (68) .
Experiments with the some gases in smal l
apparatus, with downward propagation of r
flame, have been reported, but on account o f
the experimental conditions the limits through -
out are narrower and the extinctive amount s
of inert gas are much less tlm those quote d
above (809) .
Atmospheres of Air and Argon or Helium—
Figire 24 shows the influence of nitrogen, water ,
vapor, and carbon dioxide on the limits of
flammability of methane in air, with upward
propagation of flame in a tube 2 inches in diam-
eter, with the firing end open . The differen t
effects of the three gases are ascribed to their
different heat capacities ; as carbon dioxide has
the greatest heat capacity, it has the greatest
extinctive effect on flame. The corresponding e
a
curv e for argon in the same figure agrees with Q v.
this supposition, as argon has a smaller heat
capacity than nitrogen . The curve for helium,
a gas of heat capacity equal to argon, shows
that this is not the only factor determining th e
extinctive effect of an inert gas ; apparently the
high thermal conductivity of helium makes it a
more efficient flame extinguisher than argon .
It " seems, however, that t1i efflct of difkren t
thermal conductivities is insignificant unless the —
difference is great (69) .
The order of ,increasing efficacy_ of diluents,
according to figure 24,-''ill # Ho, ,, CO,. Ex-
periments with similar series of mixtures i n
narrow tubes (1 .6 and 2 .2 cm . in diameter), for
downward propagation of flame, showed th e
same order of efficacy, except that the ppsitio n
of helium in the series varied with the diameter
of the tube and, perhaps, with the strength or
(r`


'7-
.S I l

LOUTS OF LVDIVIDUAL GASES AND VAPOR S 51

02 IN ORIGINAL ATMOSPHERE, PERCEN T

20 19 18 17 I6 16

I 1 I I
1 . liampel pipette„ 100 e .cespacity 6W '
S. Tube, 6.2 =a diem 98
l3 1
S. 2.7 cm . dim. 225
1 Bomb. .0 m capacity
., st
LS.
.

12 `' s, 5 .0 an., diem.

Tabs, amm.,. upward ptapaKatiaa '69

NEIL
'

It

IINAMM
to

BrI.„gW
a
S., .'

'
. I

sr Q S l0 1s 20 23
COs IN ORIGINAL ATMOSPHERE, PERCENT
Fumes 26.-Limits of Flammability of Metb*ne in Mixtures of Air'and Carbon Dioxide ; Comparison of Results
Obtained in Smaller Vessels. • .

;= r
the igniting spark . (For a discussion of this
important point, nee p . 5 . )
The broken line in figure 24 is the locus o f
mixtures in which the ratio of methane to
oxygen . is that required fur complete combus-
tion, CH, : 20,.
Atmospheres of Air and Sulfuryl Chloride,
Silicon Tetrachloride, Silicochloroform, and
Phosphorus Oxychloride .—Tbe limits of meth-
ane to air with various amounts of sulfuryl
chloride, silicon tetrachloride, silicochlorofonn ,
and phosphorus oxychloride have been deter-
mined in a burette 16 or 17 mm . in diameter
could propagate flame upward in a 5-an . tub e
(195) .
Atmospheres of Air end Certain Halogenate d
Hydrocarbons.— Figure 29 shows the influence
of certain chlorinated hvdroearbons, mixed in
the stated amounts wits air, on the limits of
flammability of methane in a tube 2 inches in
diameter with upwar4 .p wpagation of flam e
from tho'open end. Elul. of the carrion dioxid e
'curve is inserted for comparison (71). The
carbon tetrachloride curve * has recently been
redetermined indapa}sdently, with nearly th e
same results (2'8) . . 5

(186,195) . The curves for trieldoroethylene and tetra-

No mixture of methane and air containin g
more than 1 .2 percent of phosphorus oxycblo-
• ride was flammable in a 14 .4-mm .-diamete r
tube, by a spark passed at the upper end of th e
tube, but mixtures containing up to 3 .9 percent
chloroethylene are incomplete ; the experiments
were earned to the point at which the atmos-
phere was saturated with these vapors at labo-
ratory temperature . Tetrachloroethana an d
pentachloroetbane, up to the saturation point .

e. c

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
 t(c4f7cAn
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


54 LIMITS OF FLAMMABILITY OF GASES AND VAPORS

"mum ism
Ness= mis
'moinssemml
llllllli giilllMIIMII
unnummul m
mum um
m r-odallIMIS1.111111
,ft ma our

..e:~lC ~ 1 DILUENT IN A11406PSERL PER‘23CP

Flamm 29.--Limits of Flammability of Methane in Mixturca of Air with Certain Chlorinated Hydrocarbons an d
with Carbon Dioxide.

behave like inert diluents' whose effect is to be bromide to the air causes the limits of methan e
attributed to their thermal capacities . The to approach and, in a 2-inch-diameter tube ,
ethylene derivatives, however, contribute to the to meet at about 6 percent methane when 4 .7
flammability of the mixture, hence the lower percent of the mixture is methyl bromide (88) ,
limit of methane falls with an increase in the For a comment, compare the correspondin g
proportion of vapor ; the higher limit also falls paragraph on hydrogen (p . 22) . .. An olde r
rapidly for the same reason . The order of in- aeries of experiments (200), made in almos t
creasing combustibility is C,Clr-- eCrllClr---•) identie*l circumstances, suggested higher re -
C,H,CIi ; the feat vapor forms flammable mix- suits for both limits •tj ioughout . tlle_, whole
turea with air without the help of any methane . , range of methyl bromide boncentrations ; these
The extinctive effect of carbon tetrachloride' are partly ex{ilained by the criterion of "ffam-
on methane flames is due apparently entirely inability' adopted for the older experiments ,
to its high thermal capacity . Volume for which was propagajjpn ,of flame for at leas t
volume, carbon tetrachloride vapor is twice as 18 inches, not for the wilbfe lea tit of the tube .
eittiinctive as carbon dioxide, as the presence o f The extinctive effect of dichlorodifluoro-
l2 .5 percent carbon tetrachloride in sir renders methane on methane flames (168) is almos t
'' the mixture incapable of propagating flame, exactly equal to that of carbon tetrachlorid e
whereas 25 percent carbon dioxide is necessar y vapor, volume for volume.
for the same effect. Equal volumes of the tw o . Series of experiments with halogen deriva-
liquids, however, have approximately equa l tives, in narrow vessels with downward propa-
extinctive effect . gation of flame, have been reported (195, 197,
The addition of increasing amounts of methyl 198,1 99, 200) and discussed (71) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LIMITS 07 INDIV1DOAL GASES AND VAPORS
Atmospheres in Which the Nitrogen of th e
Air Is Replaced by Carbon Dioxide, Argon, o r
Helium.--The limits of methane in atmosphere s
in which the nitrogen of the air is replaced b y
(a) carbon dioxide, (b) argon, (c) helium, wit h
upward propagation of flame in a 5.3-cm .-
diameter tube open at the lower end, are (a )
coincident at 9 .0 percent, (b) 4 .01 and 17. 3
percent, and (c) 4 .83 and 16 .1 percent, respec-
tively, In comparison with 5 .26 and 14 .3 per-
rent in air. When the diameter of the tube
is increased to 10 cm ., the lower limit in th e
argon mixture is unaffected, but in the heliu m
mixture it is reduced to 4 .5 percent, showin g
that with helium mixtures in 5-cm . tubes th e
effect of the loss of heat to the walls is not
negligible (94) .
In a horizontal tube, 2 .5 cm . in diameter,
open at the firing end, the corresponding results
are (a) 4 .40 and 15 .80 percent and (6) 5 .5 5
and 14 .25 percent (73) .
In a dosed globe of 35-cc . capacity, with sid e
ignition, a corresponding result is (c) 8 .9 and
11 .7 percent methane (297) .
Influence of Temperature .—In an atmos-
phere of 20 .9 percent oxygen and 79 .1 percen t
carbon dioxide, the limits in a closed 35-cc .
-e globe were slightly widened by increase o f
temperature to 300° C. (297) .
Atmospheres of Oxygen and Carbon Dioxide ,
and of Oxygen and Argon .—Limits of methane
in these atmospheres have been determined
with downward propngation of flame in a
1 .7-cm .-diameter tube (345) .
Atmospheres of Oxygen and Carbon Dioxide ,
Oxygen and Sulfur Dioxide, Nitrous Oxide an d
Carbon Dioxide, Nitrous Oxide and Sulfu r
Dioxide, and Oxygen and Nitrous Oxide . —
Limits of methane in these atmospheres hav e
been determined at pressures ranging fro m
600 :iim . down to the point of coincidence of th e
Simi' (89) .
. A' ieospheres of Nitrous Oxide and of Argo n
Mix,. 'i With Nitrous Oxide.—The limits of
methane in nitrous oxide, with downward propa -
ptton of (lame in a ctoeed tube about 16 mm .
m diameter, are 2 2 and 36.6 percent (206), 3 .9
and 36 .3 percent (339), and 4 .1 and 41 .2 percent
at 600 mm . pressure .(89) . In a mixture of 2 upward propagation of flame in a tubei ms. .
volumes of nitrogen and 1 of oxygen, equivalen t
to nitrous oxide the limits under similar condi-
tions are 4 .3 and 22 .9 perent . " The progressiv e (74), 3 .1 and 12 .5 (146), and 2 .90 andl2l
addition of argon narrows the range, and n o
mixture with nitrous oxide is flnrnmable unde r
the conditions stated if it contains more tha n
about 60 percent argon, or with the mixture o f
2N,+0, if it contains more than about 6 7
percent argon (206) .
Atmosphere of Nitric Oxide .—The limits of
methane in nitric oxide, with downward props.-
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
S
gation of flame in a closed tube 15 tuft
diameter, are 8.6 and 21 .7 percent (339).
Atmospheres of Nitrous and Nitric Oak -
The region of flammability of mixttaeat l
methane with mixtures of nitrous andtiiia :
oxides is plotted in a triangular diagramMI L
Influence of Small Amounts of '!liar•
moters .--The addition of about 0.5
of various possible "promoters" (diethy l
ide ethyl nitrate, ozone, ether, ethyl a
had little more effect on the limits of mim e
in air (upward propagation) than that i s
the thermal effect of their reaction . lllllk
nitrogen peroxide, 0 .5 percent reducible
lower limit by 0.33 percent (93) and 0 .2 post
by 0 .26 percent (70). The lower limit
downward propagation was not a prix r i
effected by up to 1 .0 percent of nitrgst
peroxide (70) . With methyl iodide, 0 .5 peat
reduced the range from 5.26 to 14.3 perdd
6 .29 to 12 .3 percent (93) .
Dilution of CH !'+20 3 With Gases, Ina =
Otherwise.—The following results were olxliw d
with downward propagation of flame in An
burette 1 .9 cm . in diameter.
Efectsofdiluentr upon flammability of 0114 -*
Amount o/ Cll,+aGaM
■ilk Mien : camel iomid
of t at a/pommotNUIimt
Diluent :
Oxygen Il
Nitrogen
Carbon dioxide Af t
Nitrogen has a greater extinctive actionf m
oxygen, Although it has less heat capnl r
carbon dioxide which has much grcaterl d
capacity, also has a greater extinctivesttie ii
(95) .
With upward propagation of flame insta l
5 cm . in diameter, the limits were found twif.
writers to be 11 .85 percent CII,+20, i
diluted with argon and 15 .0 percentci
diluted with helium . The effect of theiiihs-
ence in thermal conductivity is notewortk
ETHAN E
ETHANE
.
The limits of ethane it1 air, nearly dryaalk
in diameter, open..iz5 the firing end, wereimi
at various times to '~'br ''342• and 12 .45 ~p.~eattt
percent (138).
Table 16 summarizes other detorminatimalt
the limits of ethane in air .
The higher limit varies more than tholes '
with the direction of propagation of flame I.
closed tubes the higher limit is apperentlymah
greeted by the pressure developed taxis
inflammation .
9C if7cen i


OXYGEN IN ORIGINAL ATMOSPHERE . PERCEN T

19 18 17 16 15 14 13
■■0a/■■■■■■■■/■■■■■■■■■■■mi=■■■■■■■■■■N■■■■
■■■vino■■■■■■■■r■■■■■■■■■■ ■■■■■■■■11■ ■■■■ ■■
■■■■Nimm■■■■■■■r■■■■■■■■■■ ■■■■■■■■■■m■■■■■
.r;\■a■■■■■■/■■/■■/■/■s■■■■■■/■/■/■/■ ■
■■■■INo il■■/■■■■■■■■■■■■■/■ ■■■■■■■■■■/■■■■ ■
■/■/■■
nnnnnnn .~Ir, nnnnnnnnnnnnnnnn nnnnnnnnnn nnn n
■■■■■■■ICC■II:H■■■■■■■■r6■■■■M//■■■■///■/OM■■■ ■
nnnnnnnnn r:S■oU■■■■■■■■■■■I■■■■■■■■■■■■ ■■■■ ■
■■■■■■■■■■E■■biRtU■■■■■/■■■■ ■■/■ ■ ■■//■■■ ■
nnnnnnnnnnn 'm■■..a■■■■■_■■■■l■■■■1e.■.■■_■■■■■
nnnnnnnnnn c' I+i■Iif■■A■■ ■■/ UM , n nn mums
■■s■■■■//■■ d% ' /■/ 4"r■ ■■a /■/■ //■■■/■/■
■■■■■//■/■■ 4 `N■■/ri . 1/■■■■■■■■■■ ./■■■ ■
■//■■■/■■■/■■ 7■///s br'o /■■■/ ■■■/_■■■/
■s■■■■
umwmim mmorm '''~s1uab .OMUmm nnnn
C::::::: :::iiUNIMMI ■IC1`i ~. iiii ■::CMMMENE::::
nnnn Immommo

■■■■■■■■■■/■■/■■■■■/■/ ./.■■:rCW//■//■////■
::::S:aDMMUMEN=:::3 : ::= =::0.':~ORIMMoMMEE .
■■■■■■■■ ■■■■■■ / ■/ ■■■ a ■■■■■■►:t■■ ■■■/ ■
Tsp ■■■■■■■■■/■ ■■■■■■/i&!■ ■ ■■■/M1•■■■ ■
■■■■■■■t■/■■■■ a■■■■///■i►!■// /
■■■/■■■_5lS~
s
ii■ iil■iii~:!~■i■i
/ r■lllg=N* iiiiiii iii■
luaaa■m■/■■■■=/',tl■
■■■■■■■ *a . . sal ma ■/■■■/■■
+Atat ~ r~•--■A■■■
iiii/■/■ rl~r

■l■■■■ :.;. irr'ai..

msm -~._■a/4t
a■■■r1i1imi■mii-■i l!tr y.r~l//■■
.t.r. ..r . ./■■/■/r///Itliliiic
r.= .giarsR■■■■■■■■■■■orscemmem
.EMIN/I■■■/I■/■/■■■///■///■N/ ■
■■■■■■■■■ ■ ■ ■ ■ ■■■■■■■
/ ■ ■■ ■■. ■■ ■■■■ ■■■■■■■n ■■■
.■■■■C■■■■■■■■■S■■■■■ ■■■■■
■■■■■■■ ■■■/ ■■■■■ ■ ■/////■■//■■/■■//■■//■
■■■■■S■■■■■■■■■■■■■■/■■■■
■■■■■■■■■■■■■■■■■■■■■■■■ ■■■■■■■■■■s■■■■ ■■
S■■■u■■■■■■■■■■■■
iiii:iiiEWMI iiisii 11INEEMEN88WEEWE .M.MONiiiii
5 10 15 20 25 30 35
ADDED INERT GAS IN ORIGINAL ATMOSPHERE . PERCENT
Flaoaz 34,-Limits of Flammability of Butane in Mixtures of An and Nitrogen, and of Air and Carbon Dioxide .

butane in atmospheres of air and nitrogen, andl

ISOBUTANE IN AIR ofwithairtheandcorresponding
carbon dioxide,limits are almost
of butane Identica
(175) .
The limits of with upward-
isobutane in 2airinches PENTANE -. .-
propagation of flame in a tube in diamy
eter and
atmospheric 6 feet in length at approximatel
pressure are 1 .83 and 8.43 percent PENTANE IN AI R
(174) . propagation of flamepentane
The limits of in a tube in air,
5 cm.42 with upwar d
. in diameter
ISOBUTANE IN OXYGEN and open (138) at theandfiring 1 5and 7.80in'
The limits of isobutane in oxygen with up n-
ward propagation oftheflame in aend,
tubeare2 inches od percent in a end, similar are tube .3 cm.
diameter, open at firing 1 .80 an diameter the lower limit is 1 .62 percent (94).
48.0 percent. I *nu
The p.ntane
OwClrretor. used in MeInstitute.
rimeLlmltePetrnkum
tenoned
drrwminal ionl(rrstrb
fur a brsanr
el ',wasmolest
cycbber ane ethyl x . Boorb
beats4&.was C.supplied
ryebbutane.
ISOBUTANE IN OTHER ATMOSPHERE S ethyl
hrtane eyrluprntane
3.3tantent% . n•heptene2.2..3.metbyl
dlrtbytprnlann. brnoens
3 (rlr. bobutyl (blu .w).metbyl
metbyl p•ntant, sibylbrnama
cycle,.
a-butyl
brnrrnr. see. butyl brewer bentent, and
by Murex of %l uw (ref. ISa) Is airbus 'actions of ads bulletin . tat-butyl
Air and CarbonofDioxide
Atmospheres Air and.—TheNitrogen,
limits ofandiso-
of
t.nitste.obtained wltb wanks .1 blab purity supplied by the oboe .
were

—w wows. -aw■■wswr wt..."...".01.191111111pIPPI rIf''Ir'

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF INDIVIDUAL GASES AND VAPORS

TABLE 19 .-Summary of other determinations of limits of pentane in air

Upward Propogatiou of Haa s

Dimensions of
tube , em- Limit+ . percent
Firin g cod
Firing Content of aqueous Referenc e
- vapor No.
Diameter Length Lower Higher

7.5 150 Closed 1 . 42 8.0 Haif-ssturated ~ 363

5. 0 150 ___- .do 1.43 8.0 de 168

7. 5 150 Closed 1 .44 7. 45 Half-saturated ___ sae

50 150 _- _- .do 1 . 46 Q 70 ._ __ .do
2. 5 150 Open 1 .6 5.4 Saturated __ t71

7.5 150 Closed I . 48 4.64 Half-saturated 33 d

62 33 Open l.3 . . . . . . . .. . Saturated 85
5.0 150 Closed 1 .49 4 .58 Half-saturated 360
5.0 65 do- 1 . 43 4.6 Dry 310
5.0 65 Open . . . 1 .50 4.75 . . . . . . . . . . . . . . . . . . . . . . . . t.,
2. 0 40 Closed 1 . 75 4. 68 . . . . . . . . . . . . . . . . . . . . . . 37 7
1. 9 40 .-___do 2.4 4.9 Saturated .95

Capacity :
Globe, 14.5 liters_ ________________________ 1 . 2 .-__ ._- ._ . . .._ 96
Large- 1. 4 4.5 __ 17 1
Bottle, 2liters .__ . Open 1. 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .._ Std
Globe, 2 liters---- .
1 37 ----------------------------------- 31

Tables 19 summarizes other determination s

of the limits of pentane in air.
Influence of Pressure .-Observations (277) in
a closed tube 2 em . in diameter and 40 cm. in
length, with downward propagation of flam e
are plotted in figure 30 . (For a discussion of " .
these curves, see p. 3. )
The effect of reduced pressure on the limits o f
ignitibility of pentane by a standard spark ,
rather than its limits of flammability, have bee n
reported (13).
Influence of Temperature .-In a closed tube
2 .5 cm . in diameter and . 150 cm . in length, with
downward propagation of flame, the lower limi t
decreased linearly from 1 .53 percent at abou t
17° C. to 1 .22 percent at 300° C . ; the higher
limit increased linearly from 4 .50 percent a t
about. 17° C. to 5 .35 percent at. 300C . ($o8) .
Atmospheres of Air and Nitrogen (Air De-
ficient in Oxygen) .-Tile limits of pentane in all
mixtures of air and uitt•ugeyy et~e tthowu iu figur e
35. The determinations tfere made in's tube
2 inches in ditusteter and 6 feet in length, with ' _
upward propagation of flame. St atmospheric
pressure during .propagation . From the ordi-
nates of the ' nose' iif-lhe curve it may be
calculated that no mixture of pentane, nitrogen ,
and our at atmospheric pressure and tempera-
ture can props ate•flame if it contains leas than
12 .1 percent of oxygen (138) .
Atmospheres of Air and Carbon Dioxide .-
The limits of pentane in all mixtures of air and
carbon dioxide are shown in figure 35 . The
determinations were made as described in the
previous paragraph (138) .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


OXYGEN IN ORIGINAL ATMOSPHERE, PERCEN T

17 15 13
m■rmlirmmil~■■r■■mi■■Im■■ .Iim■■■II■■mm■maRII■■mmllmi■mI ■m■
Simi■cirri■■■■■■ir■■■■umm■■f•■mm■ii■mm■■■■■■■■■ri ■
■11CSS■■i■■■i■■■r■■i■■■■■■■i■■■r■r/■■i■!■i■■■■■■■!■ i
rroust■m■■■i■rrr■■■■■■■■mmmmmrmm■nmimm■■■■m■■■■■ m
■■■>1► .■■■■/■■■i!■//■■IIQ■i■Rir/i■■■iii■■■i■■■■■■■ ■
/■>•!■1.i./iI//■Ir,/IUIIni/RI~*//IlIIIII//!!/■ii/ i
■■■■r■iIIi•i. .■■■■■■.U■■■IR■■■■$■■■■■■■■■■■■■■■■■■■■
■/■■■■/■>a► ■.u■as//■■n■■■~■■/■■■■■■!■■■■/■■■■■RR■
■■■■■■■i■i■ ./►•Ui■■a■■I>•■■■■■■/■■/■■!■i/■■■■■■■■
■/!ii■i■iii■ c,~nMI■■i■ii■i■i■■iii■i■iiri■nri■ r
■■m■■■■■■r■■■ 604
■■mm■■im■■■■ri■ °'ar;oP•~■■
a■■
RIMM ■■■■■r■irii■■i■■t■/rii■
ii■rmr■i■■ii■■r■■i■i
.~ 4 , ■immim■■m■■■imm■■r■m
■■mm■■mi■r■rmmii■ ~r~■■~°g~,, m
■■■mm■mi■r■r■m■■m■ ■azU■mmU im■■m■■■i■■m■■mm■mr ■
■■■mm■IIIIIIIamm■mm■■mi■mmm■ .ININEI t N11111 mi■im■immmmmi i
ir■■■■i/i/■■/i/■■/■ia■/11SMISi/il .MR■■i/■■i■■ri■■i■
■■■t.Ilf/■/■■/■//■ii■!i■■a■■a!~■■■ni■■-~■■■m■■■i■m■■■
■■■■■■m■mm■m mmi■■■■rri■■/!a■ii■iiiim~ir■■■■■ri■ m
■■m1~■mmm ■■■■■■i■■■■■■rm■i■■m■■~►tri■■■m■■■r►Lmmmi■■mm m
~l~■■■■ Theore#tcai m i■i■■■i■■iirtir■n■mi■i■l1Amm■■■■■
ixtures /i■rrii■ii1!'I■ii/iiiiiililli/iri■ ■
■■■7■■m■■■■■mm rC1Z,lIm■■,Ci1imr■■■■ii■i■■■mlim■■ m ■ i n■
■■■t>iiii

■■t ■m■mr■■■■■•■/■i■w!~:~~iG~~ll~■■/ii/iN~i/■iii■■
■tit■gawwN~~~~~~wst■ssi_•~_w~_w_w_"~` ~iiiiGfi3~~~iA■■i■■■■TP■
iai ne .■■■■m■mmmmm■m■■m■mmM■■■■■■■mi■r■iria=
■r■■■le~ii■■■■r■■i■■ii■■i■/■i■rr■■■■■■i■i■/■i■rii■ i
mmmm■m■■mm■■■■m■m■■■■mmum■mmmmm■mmmii■■memmmm■■m■ m
■■■■■■in■■ii■■■ii■■■■■i>tiiii■i■■r/i■i■■rm■m■i■i■ i
■■.■■■■■■■■■■■■■■■■■■■■■m■mmmm■m■■■■■■■■■■■■■■■■■ ■
5 10 15 . 20 25 30 . 35 40 45 50
ADDED INERT GAS IN ORIGINAL ATMOSPHERE, PERCENT
Fmu* 3S .-IJmits of Flammability of Pentane In Mixtures of Air and Nitrogen, and of . Air and Carbon Dio:lda

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


pressure (353) . (See below under Influence of

Extraordinarily large differences are to b e
found between the various figures recorded fo r
the higher limit of ethyl ether in air. These are
to be explained by the phenomenon of "coo l
flames," by which rich mixtures suitably ic -
e can propagate partial combustion slowly ,
with a comparatively small rise of temperatur e
but with the normal appearance of a flame, i n
the upward or horizontal direction . In a 2-inch-
diameter horizontal tube the higher limit of the
"ordinary " flame merges with the lower limit of
the "cool" flame at atmospheric pressure bu t
becomes separate at lower pressures (359) . I n
a 1-inch-diameter tube the ranges for the ordi-
nary and cool flames are separate at atmospheric
ressure . )
Two observations of the limits of ether in ai r
have been made with upward propagation o f
flame in a 2-inch-diameter tube, open at th e
firing end . These gave 1 .92 and 43 .5 percent
(31), and 1 .85 and 25 .9 percent (183) ; in th e
latter tests the source of ignition was presum-
ably unfavorable to the initiation of a ""cool "
Same .
Table 31 summarizes other determinations of
the limits of ethyl ether in air .
The limits in a small vessel have been state d
to be 1 .2 and 51 .0 percent (9) ; and, with a o
obviously weak spark, 3 .14 and 9 .5 percen t
(327) . A lower limit of 1 .67 has also been re -
ported (30.6) .
The lower limits in a 5-cm . glass tube closed
at both ends differed only in the second decim a

TABLE 31 .-Summary of other determinations of limits of flammability of ethyl ether is ai r

Upward Propagation of Flam e

Dimensions or tube, cm . Limits, percen t

Firing end Far end Content of ague- Refer-
ous vapor cote No .
Diameter Length Lower Higher
11

15 300 (iron) Closed Open 1 .73 23. 30 Dry 38 1

10. 2 96 (iron) do_ Closed I . 95 30.5 do 183
7. 5 150 (glass) . - . - .---do do 1 . 71 48 do 353
6.0 120 (glass) do 2. 1 ........ Partly dried- .._ $35
5 150 (ion) . . ._- ._ Closed Open 2.24 15. 45 _ _ $61
5 150 (glass) . do--- do- . .. 1 .93 15.75 do 361
5 ISO (glass) do Closed I . 84 48 do 35 3
1 5 91 (glass)_- do do .__ . .95 15. 60 _ . ._do_ .____ .._- 351
2. 5 150 (glass) do do 2.00 47 do 353
2.5 25 (glass) Central ignition- -__-_do 1. 25 10.0 Undried 39 4

IS 300 (iron) Closed Open 1 .80 22. 30 Dry _- 381

7.5 150 (glass)_ _ do Closed .- . .-- .-- l : 75 40 do 355
5 150 (Iron) . . . . . . . . . Open 2.29 7 95 do 861
5 150 (glass) do_ . do 2. 0S 8 do 86 1
5 150 (glass) do Closed 1 . 88 33 do __ 855
5 91 (glass) do do 2. OS 8 do $5 1
2.5 150 (glw) 1. .98 56.,25
d o d o do $55

15 300 (iron)
Closed_ On I . 93 a 50 Dry 36 1
7. 5 150 (glans) do_ . .___ ._ . Closed 1 . 85 0.40 --do 355
6.2 33 (glass) -- Open do I. 6 Saturated 95
5 ISO (iroe) - close(' open 2. 34 8.50--D $81
6 150 (glass)- do do- 2. 1S a 15 - $81
5 91 (glass) ._ __ - .__ .do Cloned 2.15 615 do $5 1
1 5 (glass) . . . . . . . . . . . . do do . 89 6. 7 do 31 6
5 65 (glass) Partly open .--- . Partly open 1 . 93 6. 66 34 6
2 .5 150 (glees) Closed Closed 1.97 6 15 do 853
1.9 33 (glass)--- -- .
Open do-_ ._- .__ . 2. 75 7. 70 Saturated ._- . ._ 95

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

LIMITS OF INDIVIDUAL GASES AND VAPORS
place from those obtained when a small stop-
cock was opened at one end or the other ; the
higher limits differed by not more than 1 per-
cent, Similar differences were observed when
the length of the tube was varied from 100 to
250 cm . and when the point of ignition was
brought forward about 6 cm . (353) .
Influence of Pressure .--As the pressure is re-
duced below normal, the range of flammable
mixtures in a horizontal . tube 5 cm . in diameter
divides into two. At 500 mm . pressure the
range is 1 .88 to 9 .25 percent for the ordinary
flame and 13 to 33 percent for the cool flame .
As the pressure is reduced further, each range i s
contracted, and a little below 400 mm . the cool
flame is no longer propn gated . At 90 mm. th e
range for the ordinary flame has contracted t o
2 .3.2 to 6 .1 percent (359) . Similar observa-
tions have been made with a tube 2 .5 cm . i n
diameter (331).
The effect of reduced pressure on the limits of
ignitibility of ethyl ether in air by a standar d
spark, rather than its limits of flammability, ha s
been examined (8, 13) .
Increase of pressure above atmospheri c
widened the range at both ends when the tem-
perature was raised to maintain enough ether
vapor for the test . The effect may be due mor e
to the change of temperature than to the chang e
of pressure (14) .
Influence of Temperature .—The higher limi t
of ether in air is appreciably increased by a ris e
of 40° C . (361) . (See also next paragraph . )
Influence of Pressure and Temperature To-
gether .--The limits of ether in air have been ob-
served, with horizontal propagation of flame in
a closed tube 2 .5 cm . in diameter, at pressures
up to 4,000 mm . and temperatures of 20°, 50° ,
100°, and 150° C. The limits of both ordinary
and cool flames are widened by an increase i n
temperature ; therefore, the two ranges meet a t
lower pressures as the temperature is raised .
A third type of flame . green in color, is observ-
able in certain circumstances (125, 331, 332) .
The relation between cool flames and norma l
flames is shown in diagrams and discussed, an d
conditions under which a cool flame may becom e
a normal flame are given (126) .
The limits in a cylindrical bomb of 700 cc.
capacity, after heating under pressure for 30 t o
45 minutes to allow prelim-no combustion to
occur, are shown in table 32 . The limits are
expressed in percentages by weight ; the rang e
widens at first with increa4o of temperature bu t
narrows when prcflamo combustion become s
evident . The bomb was filled at atmospheri c
temperature to an initial pressure of 5 .8
atmospheres (245) .
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
89
TABLE 32 .—Limits of ethyl ether in air at in-
creased pressures and temperature s
Umita. plrat by we1.t t
Twnptn r
tore.• c. Pre:own
Lower a+ce~
12S Not aalM 7. t
140 do."-'_ ^' asst
mar
170
do_
. . .do 4 .7
172S
t77 dodo 4.a
L7
1 apan4a ous ledomaatlos
Influence of Streaming Movement of Mix-
ture.--The limits were widened several tenth s
of 1 percent when velocities up to 9 cm . a
second were imparted to the mixtures (361) .
Influence of Impurities.—The presence of
diethyl peroxide and of ethyl hydrogen peroxide
scarcely affected the lower limit, but any large
quantity raised the higher limit (361) . The
laser limit was unaffected by small additions
of diethyl selenide or lead tetraethyl . Th e
effect of pyridine (one experiment) accorded
with Le Chatelier's formula (313) .
ETHYL ETHER IN OXYGEN I
The lower limit of ethyl ether in oxygen, wit h
upward propagation of flame in a 2-inch-
diameter tube, open at the lower end, is 2 . 0
percent (162) ; in a closed tube 10 .2 cm . in
dinmeter and 96 cm . in length, 2 .10 percent.
The higher limit with upward propagation in
a tube 4 .4 cm . in diameter and 60 cm . In length,
open at the upper end, is 82 .0 percent (183) .
The limits with downward propagation of
flame in a narrow burette are 1 .7 and 39 . 5
percent ether (205) .' The higher limit with
downward propagation of flame in a closed
tube 4 cm . in diameter is 65 percent ; in th e
presence of 5 percent of carbon dioxide, 60 per-
cent (124) . The limits in a small vessel ,
obviously with a weak spark, are given as 2 .98
and 45 .5 percent (327) .
ETHYL ETHER Dl OTHER ATMOSPHERE S
:. . C .
' Atmospheres of Air and " Nitrogen (Air De-
ficient in Oxygen) .—Tho limits of ethyl ether i n
air deficient in oxygen have been determine d
in a small blilb . Its-the oxygen was reduced ,
the lower limit remained` nearly constant, but
the higher limit fell rapidly ; the l imits coincide d
at 10.7 percent oxygen (13) .
Atmospheres of Oxygen and Helium .—The
limits of ethyl ether in all mixtures of oxyge n
and helium, with upward propagation of flam e 1
in a 2-inch-diameter tube open at the lower '
end, are shown in figure 48 (162, 163) .


1' t
90 LIMITS OF FLAMMABILITT OF GASES AND VAPORS

100
oamanomilemmienmammanamranviommoni m
■lwairs /piunsai minis R/Rw//w .//R..R■R■RR■R■■R R
■■ . .►VR■.■./.aff1.R .Ra Helium = 100 — ether + oxygen) Ra■R.■
..uaa,ammu .mrrm main s women ■■ .Rai■RR■~ ~ui
90

■■■a/.~.w..am. ■ . ■R■ /■■R■ ..RR■RaR■/R*R/ R

lauim.R ■R ■■r.uRR■R.R■Ra.RR.aaauaii■. ■
■,a■■RIRINWO .■Ru/ ■aR.a~■m/ .Rm .■■/■■R■■ mm
00 NUUU■■■■■■►■■■ ..R■.■R■■■ .■■■■■~~■Rt■■■■■■ ■■■
■ w~~w~wwt#w ♦~war~taw~s .rM.~lw~waaw ■wRwRw .www EMI■
g Ia./R .■■ .Rs.. /R.R.#.R_■R■R.■ .R■am.ra/RR/ai■w■
.m■R/.■■■■■R■■■■■■Rm■m .
gI,. .a.■■.■R.m'.aa.■■ R
70 I umUmrrrrrr ; U =wsoms IIM$ I mpossible mixtures
.■
■ rr
ia.■■■ .a.a■a.a■■..■■■rR■■a■ .■/■a.vm■■Riau■■
Itimansimmmumimiowinnamununns p immma l mon
W 60 murnimummanimmiamolnmesponnammerarmus ua m1 a ■/
.■m p ■aau■ma■■■m■■aamiSRa■
111111 .111111111111111111116111111111111111111111
`■■■mamma ■imam■■r
iii
.suaa■■■■aa■■■
a i i~~r~.
c)
c
w IIauR■r
■■■■ Explosive area v.m.muu,a.a■.■
■/.a■ ■ Vummuua.R■aa/a ./I.■.R
~IIrNIM%~■■~~
a' 50
aa .m iii'. ' i i.ii u UMv .. UaMa
r~r1ta■rrII
z t t
w
I Mr~11111 a'alma■/a
' ■rwMI
r/■ / /
C7

0 4 0 R■Ra .a =/R/a■■/ . a■ a
au
■.■■r■mmR■u
•m
/mal.■.m. ■■■. ma■■■.■■■ !Ulm■■maUC ■a=■mm
:Ram ml umrimiRum'a'll'. siC
Nina. ' aNlaii■'
I0■ r-
30
Cool dames ~
rr.■ a' anima rll a. ■ RR■amm kit !lam■■
mail as■■ma■r/
..a■a . ■ R■■■ .m.■mm■a wa.a■1~.~1~■ww■ a~ k ■ maim t
20
■/
arsdi: . r Ra 11111=1111111116honli n

■ ENIENIN
II 111. II 1111I M

mess' a
ONO m■ M
a mama■■
.as. a ■a~'il°iiii■mm■■ .■.■ Im o NE W

~'■
10
.a■Ram■ ■ ::omr.a r~Crmmrmso
. ■■ ' a
m
aue .uaur'1'llramu.Ur
a rr:rrrwwrrr :r■
■r■.ua.a.rrma■./■a■mar■■m■ .R■a■.■■m/■■aamm■
0 10 20 30 40 50 60 70 80 - . 90
ETHYL ETHER, PERCENT it
Frounc 48.-Flammability of Ethyl Ether-Oxygen-Vellum Mixtures.
i
of Nitrous Oxide 30 percent ether when oxygen forms 10 percen t
ethyAtmosphere
l ether in nitrous oxide, with.—The
upwardlimits
propaof- of•the whole mixture 104)and
; Nitrous-0=de -
gation Atmospheres of Oxygen
1.50 andof24.2flamepercent
in a 2-inch-di4metor tube, ar de
(163) . With downwar The composition of all flammable mixtures .o-f I
propagation of flame in a narrow burette they ethyl ether,49.oxyggen, and nitrous oxide, is shown
are 3.8 and 25 .7 percent ether (205) ; in a vessel inwithfigure 1'tie~•determinations were mad e
320 co. in capacity, 1.5 and 16 percent (60) . upward propagation of flame ins 2-inch-
Atmospheres of Air and Nitrous Oxide,—Tb of diameter tube open at the lower endin(163, 181).
limits ofnitrous
ethyl ether inhave
it series of mixturesfor o For downward propagation of. flame a narrow
air and oxide been determined burette, see reference (205)Oxide and Helium.— ii
downward
burette (205)propagation of flame in a narro w Atmospheres of Nitrous
. The compositions of all flammable mixtures o f
of The
flamehigher lima, withtubedownward propagation ethyl ether, nitrous oxide, and helium are show n
in a closed 4 ctn . in diameter, is in figure 50 (163),

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


s'x

LIMITS OF INDIVIDUAL GASES AND VAPORS 91

100
Il r
MI 6
1111111C.
Nitrous oxide 100 — (ether + oxygen )

IL

IIIIIiil
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIk Impoisible mixture s

70
•1111111111IM111MIM•1111• 6
IN 1111111NIIIIHIIIMIII•IIIMll n

011•111011111111M II MIN MO h,
Z 60 MUM 111 MIL

w
CC 111111.1111•111111111111111MM11111111i.
If
MBE ■l MEN •11IM M•IIIIIIHIIL.
z -5 0
w
(.1
1
11 .111111111111111111111INIHMINIM19111111111111IIl ,
04 Flammable per se ll""Imm ''

il1~N%IIIIESIIS%EIMS11
n. .■■/■ .~% .~~ almn
.1MOI••111M•111BONNIIIIIIIINIIIIIIIIIHIM11•1111111111IM1111 L
Ell n

30

20
1.81ININNINIONNIERSEMISIMMINIISI11111MOHOOMEM 'Ziull

1
10 11111111MIMINIHIM1111111PMaillI1111111111MI1•111111111111111111111•11
11111111••11111IMMIlIVAIIIIIIIIIIIINIIIIININIM11•1111111NIM11111 1
rminmonset umama sanasualmsanaaaarmasimsma .

11111W11111111111111'i Mill•1111111IM111111111111IIIIIIIIIIIIN1111111IMMI l
10 20 30 40 50 60 " 70
ETHER, PERCENT -
Roost of Ethyl Ether-Oxygen-Nitrous Oxide Mixtures .

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


92 LIMITS OF FLAMMABILITY OF . CASES AND VAPORS

100 EMU .ENNUII II IN III MN MUWI

110111111111111111 .1101111111MENIIIIIIIIIIIII ME EMU
1111101111111111111111I N IIIIIIIIIIIINIII 1111111111 IIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIII
IIINIIIIIIIIIIN Ill II Mall II NUM 1111111111MIM 11111I ENMU M
90 11111011111 :11
1111111111•1110111111111111 N
Helium = 100 — (ether + nitrous oxide emir
MIR II EMINI
SUM.
II 1111111111111111011111111111111111101111111111•11111111111111111111111111111 N IIENEIIII
ME 11111111111101110 MUM UM
80
MUM MINIIMI 111111MIIINIE IIIIIIISM111111111111111111 IIIIII II
N NI WIN MIR 01111111111111111111111111
I MIME111 111 N. 01111111111MIl NM IIMU 111 NNEII
raWaRlMiraNIMEMENE
IlrllllllllllllllllllllllllWA WM IN N EN Impossible mixtures
111 .1111111MIll II
- :, 70
NI namma le per se T/11111111MMIM11111111111111111111111111MINIIIIIIIIIIIIIIIIIIIIIIIIII I
II 41111111 .1111101 .11HII •11111111OMMIN El MEIIMIN
WU .MUMII11111111111111MIMMKIIIIIIIIIIIIII I
III MIA IHIHIIRMIII 111111011111111•111 IIMUIIIIIIIIINIIIMIIIRNII
a 11111111111111111111IM11111111111 .1111111111110111111111111111111111IN UM . II MN 111

F IMIMI NIEW8111111111111MI N
111.11111MM MU 111101111111111111 IIENIII
•111111111111111 NIIIMNIIIII R

111111.11WAIIIIIIIIIIIIIIIIIIIIINIS1111111111111101INKINENIIIIIIIIIIIIIIIIIIIIIIII I
mmmsommmm.mmummmmmmeo.1mmmmommmumm .m.I
I IIIIIIIIINVAMMIS11111111111111111INIIIIIIIIIIINIIIIIHM1111111111SINNIIIIIIIIIIIIIIII
III

E
IMMMINUERMENNUMMENUMEMENEE .M.UNEUM U
MAIWANINERIME. .E.I.M.NMENNEIRUNEMNMEMMM M
ElpENEIINIMEEMMMENHI NUMEEVIENIINXIOUN I N
IN''ElINNIMENIIEUMUNUMMENUENNUMENZENUIEU E
30 1 II
rmimmmmummmmummummimmmummmmmalm s
1 1WAIMNSINIIIIIIIIIIIIIIIESIM11111111111111111 1111110111111111111111111*11E011 1
IMIN11111111111I
20 NIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIINIHN 111I MENU ER 1111011111111111111111111111 .11h:
IMIIINIIMIIN MINIIIIIIIIII NM II NM
111111111111111 NUM II MIEN IN
111111111111111@1111111111N MUM
10 11111HIMM111111WIIIII 111111111MNIIIIIII I
IIIIIIIIIIIIIIIIIIIIIIIIIIII IINMI111111111111111111IIIIIH I
*111111111111111111111■11111110 MIMEINMEN IIII111111111101111111
111111111IMNII IIIIIIINIM11111N UMW NMI= 111111111111111 1
1111111111111 .11111111111111111111INUIN NEE ME NMI 111I MM.NIIIIIIIINMIll
II N II P1411111111111111111111IIIIIIIIII
1'iZ 0 . lo 20 30 40 . 50 . -- * 60 70 '
. .-
EDI ER, PERCEN T
novae SO.—Flammability of Ethyl Ether-Nitrous-Oxida- .9elium Itfixturva.

a ETHYL PROPYL ETHER PROPYL ETHER

t r. . The limits of ethyl n-propyl ether in (a) air an d
(b) oxygen, with upward propagation of flame
In a 2-Inch-diameter tube, open at its lower end ,
p are (a) 1 .9 and 24 percent and (b) 2 .0 and 78
percent. (834) .
PROPYL rrtwt or AIR
The limits of isopropyl ether in air, wit h
upward propagation of flame in a 2-inch-
diameter tube, open at iti lower end, have bee n
given as 1 .38 and 7 .90 percent (138) and 2 .1

. ,.

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF FLASLMSABIL1TT Or GASES AND VAPOR S

stated as 1 .6 and 15 .3 percent (9) . The agree-

ment of the figures in table 34 is poor.
'ACETONE IN AI R Unusual difficulties have arisen in the inter-
pretation• of observations of the nature an d
The limits of acetone in air, with upward progress of flame in acetone-air mixtures .
propagation of flame in a tube 5 cm . in diam- Thus, one observer writes (353 )
eter, open at the firing end, are 3 .0 and 10 .80
The greatest difficulty was found with acetone.
percent (138) or 3.1 and 11 .15 percent (31) . • • • The (higher-limit) figures finally taken were
Table 34 summarizes other determinations o f the highest %slues obtained in any of a great numbe r
the limits of acetone in air. - . of trials . Below the values given an Iguitioa woul d
often occur and a flame only go halfway up the tube
The limits in a , 2-liter flask and a 13-lite r • • + The fart that a flame goes only halfway or
flask, with ignition near the base, are 2 .5 and . less along a tube is no proof that the mixture is above
10 .4 percent (70) . Two older figures for the the limit In that tube • * • In a 7.5-cm. tube
near the upper limit upward a mixture which onl y
lower limit, probably for downward propaga- propagated flame 50 cm. or less would often propagate
tion of flame, are 2 .9 and 2 .7 percent (213, 221) . • . flame much farther at the second trial and at the thir d
The limits in a very small vessel have been all the way to the top
TABLE 34 .-Summary of other determinations of limits of flammability of acetone in ai r
Upward Propagation of Flam e

Dimensions of tube, cm . Limits, percent

Firing end Far end Content o f Refer -
aqueous vapor eerie No .
Diameter Length Lower highe r

30. 6 39 (iron) Closed Open 2 .67 . . . . . . . .Undried 339

15 300 (iron) do _-__-- .-do 2 .88 12 40 Dry 36 1
10. 2 96 (iron) du Closed 2.55 i2 80 ._ .__do 14 3
10.2 96 (iron) do..__ . .__-- do- . . . 2 .64 . . . . . . . .85 percent sat- 143
nested.
10 75 (Rese) do do 2. 15 9 .7 Dry 35 t
7. 5 150 (Klass) do do 2.89 11 .8 do_ 353
5 150 (iron) do -- Open_ 3. 80 do 36 7
5 150 (glass) .- . do .- .-_da 2. 89 12 20 - do 36 1
5 150 (glass) . do Closed 2. 90 12 6 do- 35 3
5 150 (glass) . . .. . . . . . . . . do do 2. 20 9.5 do- 356
2.5 150 (glass) do do 3. 12 12 95 __ _ .do_ 35 3
2.5 60 (glass) do do 2 30 7. 5 do_ 356
25 2S (glass) (Central ignition) --do 2. 5 9.0 Uadried 694

15 300 (Iron) Cloned Open _ . . 2. 89 12 40 Dry 361

10 75 (glass) do Csed . .__ . ._ ._ 2. 20 9.3 do_ 351
7 .5 150 (glssi) do do 2. 92 11 .9 do 36 3
5 150 (iron)
150 glass)
do
do
Open 3. 90 ... .. . .. . . . . . do_ 361
5 _ - .-do__ .---- .- 3. 04 9. 15 do 361
5 150 (glass) . . . . 7 do Closed 2. 96 9.9 do_ 35 3
5 150 (glass)-__ .- . . . do do_ _ 2 23 9.3 do 36 1
2. S 150 (glass) . . . do do_ 3 10 8 25 __ _do_ 33$
2. b 60 (glass) do :~ ._ .. .. . . ... . 2. 40 8 7-" . 361

22 5 33 .5 (Iron) Open Cloned 25 'j-"hl' ___ 106

15 300 Oren) - Clewed Open 3 . II la 90 Dry 16 1
10 75 (lass) do Cloned 2 35 a5 do- 36 8
7. 5 150 (glans)_- . . . ._ . . do . . . .__ ._- - .---do .-- 2. 93 8.6 do 353
5 150 (iron) do Open 4-00 do,. __ ._ .__ 36 f
5 150 (glass) do do---_ . .-__' 3. 15 8 35 36 1
5 150 (81555) . d°-- ._-_ - Closed 2 99 8 40 -- -- --do- $51
b 150 glans)- do .-___do_ 2. 40 8.3 do_ 351
5 65 (glass) do do 3.00 . . . . . . . . . . . . do 31 6
2J 150 (glaw~)_ do . . . . . . . . . . . . . . do 3.15 8 25 do 353
25 (10 (glass) do do 2 75 6. 6 .__ ._do_ 351

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204


LIMITS OF INDIVIDUAL OASES AND VAPORS 97

Influence of Pressure .-The effect of reduce d
pressure on the limits of ignitibility of mixture s
of acetone and air by a standard spark, rathe r
than their limits of flammability, have bee n
examined (13) .
Influence of Temperature .-The lower limi t
of acetone in air, with upward propagation of
flame in a tube 2 inches In diameter, is 3 .00 a t
laboratory temperature (138) and 2 .92 at 125°
C . (164) .
The limits with upward propagation of flam e
in a closed tube 10 .2 cm . in diameter and 9 8
em . in length widened from 2 .55 and 12 .8 0
I
percent at room temperature to 2 .17 and 13 .0 5
percent at 175° C . (142) .
In a standard machine for coating wire wit h
cellulose acetate, the lower limit with down -
ward propagation of flame fell from 3 .0 percen t
at 65° C. to 2.8 at 150° C . ; with middle ignition ,
horizontal and upward propagation, from 2 . 9
at 65° to 2 .5 at 150° C . ; and with propagation
against a current produced by a fan, from 3 . 5
at 65° to 3 .0 at 150° C . (142) .
The lower limit with upward propagation o f
flame in a cylinder 30.6 cm . in diameter an d
39 cm . in length, mixtures undried, vented a t
its upper end, are 2 .67 percent at 21° C ., 2 .4 0
at 100° C ., and 2 .00 at 200° C. (239) .
The limits with downward propagation of
flame in a 2%-liter bottle widened from 2 .78
and 8 .70 percent at. 100° C. to 2.33 and 9 .7 5
percent at 250° C. (21, 23) .
ACETONE 1H OTHER ATMOSPHERE S
For atmospheres of nitrogen and oxygen ,
carbon dioxide and oxygen, and equal volumes
of nitrogen and carbon dioxide mixed with 20 . 9
percent or less of oxygen, the limits of acetone
have been determined in a 2-liter flask with
ignition near the base . Curves are given i n
the original paper (76) for the whole region o f
explosibility . It appears that, when the oxyge n
content of a nitrogen-oxygen atmosphere is
reduced below 13,5 percent, . no mixture wit h
acetone,sa propagate flame (76) .
1 ' . urea of acetone and sulfur dioxide ,
evaporated into the air, can form flammabl e
mixtures only when the ratio of sulfur dioxid e
to acetone is less than 1 .9 by volume (liquid)
(308) .
METHYL ETHYL KETONE
Table 35 gives the observed limits of flamma-
bility of methyl ethyl ketone in air .
The lower-limit figures in this table are
moderately consistent. In a large space, a t
atmospheric pressure, the lower limit is probabl y
about 2 percent and the higher not greate r
than 12 percent .

TABLE 35 .-Limns of flammability of methyl ethyl ketone in air

Upward Propagation of Dam.

Dimensions of tube, Limits, percent

cm .
Firing end Content of aqueous vapor Reference
No.
1 •
Diameter , Length Lower Highe r

30 .0 39 Closed (top vented) - 1 .83 . . . Vndrted , 339

10.2 96 Closed 181 9.5 Dry . ._ . . ._ . ._ . ._ . 16 0
7.5 150 do _ 1 .97 10. 0 do 36 $
5.0 1S0 ____ .do 2. 05 9.9 _____ . .- .- .- • •363
5.0 91 do 2.15 11 . 5 .__ .do ' 161
2.5 150 do 2.12 10. 1 ___ ._ do __ . ._ .,_-_ .__ : _ . 061

Hartzovtal Propogatio Ram. .st

--i

7.a ISO Closed 1.97. 10.1 Dry 363

5.0 150 dos ._ . .._- ._ . . 2. 05 8. 5 -_ ._ .do- - : 353
5.0 . 91 do. : 2.25 '10.-3 do x61
sa 150 --= ao 1 12 a e . . _ .4d_ : 163
f

Downward Propagation d flame

7.5 . 150 Closed 2.05 7.8 Dry 36.3

30 150 _do 2.10 7.4 -- .-, do_, _-- 365
6.0 91 do 2.40 a.8 _ .___do 36 1
2.5 150 , do__ . ._ . .___ ._ .____ 2 1T 0.8 .___ .do J6d

Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204

Influence of Temperature.—The following
limits were obtained for upward propagatio n
of name in a tube 3 inches in diameter, with a
loosely fitting cap at the top (238) :
I°IW1 I.Spe
rmit Riche r
~R parotn$ lR
ao 0
. 1.$
7a1 1. O. i
150 I .6 R6
!00 I .7 O.
The lower limit with upward propagation o f
flame in a cylinder 30 .6 cm . in diameter and 39
cm. in length, vented at its upper end, is 1 .83
percent at 21° C., 1 .70 at 100° C., and 1 .33
at 200° C. (239) .
METHYL PROPYL KETONE; METHYL BUTYL
KETONE
The limits of methyl propyl ketone an d
methyl butyl ketone in air, with upward prop-
agation of flame in a closed tube 4 inches i n
diameter and 38 inches in length, are 1 .55 an d
8.15 percent and 1 .35 and 8 .0 percent, re-
spectively (138) .
The limits of methyl isobutyl ketone in air ,
with upward propagation of flame in a tube 1
inch in diameter and 18 inches in length, ope n
at the top and at a sufficient temperature to
vaporize the substance, are 1 .40 and 7 .50
percent (138) .
CYCLOHEXANONE
The lower limit of c clohexenone in air wit h
upward propagation of flame in a cylinder 30 . 6
cm . in diameter and 39 .0 cm . in length, vented
at. its upper end, is 1 .11 percent at 100° C . ,
0 .96 at 200° C ., 0.94 at 225° C ., and 0.91 at
250° C.(239) .
ISOPHORON E
The limits of isophorone in air, with upwar d
propagation of flame in a tube 1 inch in diamete r
and 18 inches in length, open at the top and a t
a sufficient temperature to vaporize the iso - .
pborono, are 0 .84 and 3 .80 percent (138) .
ACETIC ACID
The lower limit of acetic acid in air, with
upward propagation of flame in a closed bom b
4 inches in diameter and 38 inches in length, at
a sufficient temperature to vaporize the acid ,
is 5 .40 percent (138) . An old determination
gave the lower limit in air in a 500-cc . flask,
apparently with downward propagation o f
flame, as 4 .05 percent at 36° C. (221).
Source: http://industrydocuments.library.ucsf.edu/tobacco/docs/gxkw0204
ACETIC ANHYDRIDE _
The limits of acetic anhydride in air, with
upward propagation of flame in -a 1-inch-
diameter tube, open at the upper end, are:
Lower, 2 .67 percent (at 47 .3° C .) ; higher, 10.13
percent (at 74 .4° C .) (172) .
PHTHAL!C ANHYDRID E
The limits of phthalic anhydride in air, with
upward propagation of flame in a tuba 1 inc h
in diameter, open at its upper end, are : Lower,
1 .7 percent (at 140° C.) ; higher, 10 .5 percent
(at 193° C.) (184) .
METHYL FORMATE
METHYL FORMATE Ili AIR
The limits of methyl formate in air, wit h
upward propagation of flame in a tube 6 cm.
in diameter, open at the firing end, are 5 .9 and
20 .4 percent (31) . Slightly .narrower limits
have been found in similar circumstances (185) .
In a closed tube 4 inches in diameter and 3 8
inches in length, the limits with upward
propagation of flame are 5 .05 and 22 .7 percen t
(165) .
The lower limit in a closed tube 2y inches in
diameter and 4 feet in length, is 4 .4 percent with
upward propagation n of flame and 4 .5 percen t
with horizontal propagation (256) .
METHYL FORMATE IN OTHER ATMOSPHERE S
Atmospheres of Mr and Carbon Dioxide.—
One volume of methyl formate needs 2 .3
volumes of carbon dioxide to wake a nurtur e
that is nonflammable in air (185) .
ETHYL FORMAT E
ETHYL FORMATE III AIR
The limits of ethyl formate in air, with up-
ward propagation of flame in it tube 6 cm . us
diameter, open at the firing end, are about 2 .7 '
and 13 .5 percent ("125) :
Other determinations, made in a closed tub e
2 inches in diameter and 36 inches in lengt h
are : \Yit4 ,uyward propagation, 3.5 and 16 .1
percent ; h rizontal, 3 .7 and 14 .0 percent; and '-
downward, 3 .9 and 11 .8 percent (351) .
The limits with upward propagation in a
closed iron tube 4 inches' in diameter and 3 8
inches in length are 2.75 and 16 .40 percent
(138) . • . • -
ETHYL FORMATE IN OTHER ATMOSPHERES .
Atmospheres of Air and Carbon Dioxide.— C)
One volume of ethyl formate requires 6 .0
volumes of carbon dioxide to make a mixture t.;.i
that is nonflammable in air (185).
\-o
ON