1478

THE COLLIERY GUARDIAN.

June 26, 1914

Thermal Phenomena in Carbonisation.

By HAROLD HOLLINGS and JOHN W. COBB.

From a paper read before the Institution of Gas Engineers.

A number of chemical studies have been made of the
distillation of coal, such as those of Burgess and Wheeler
in this country, and of Vignon in France, but the
thermal phenomena during distillation have received
very much less attention so that it seemed to us that a
research in this direction was more than justified. The
thermal phenomena during distillation might be defined
as those temperature changes occurring during distilla-
tion which are different from the changes that would
occur during the heating of an inert substance, such as
coke or fireclay, through the same temperature range.

We have examined in some detail the work of Mahler,
Constam, Euchene, Barnum, and others on the thermo-
chemistry of carbonisation; and from the work that has
been done on the relationship between the calorific
values of coal and its distillation products it appears that
the distillation is exothermic when the calorific values
are determined at ordinary temperatures (a convention
in thermo-chemical tables). But we consider that the
question is still open so far as concerns the heat evolved
or absorbed during the actual process of distillation at
the temperatures used. Our own work, was not particu-
larly directed towards the determination of the net quan-
tities of heat involved in distillation, but, having
regard to the complicated sequence of reactions which
inevitably occurs, was an attempt to see how far it was
possible to divide the process into thermal stages—heat
absorbing, heat evolving, and thermally neutral.

The method employed consisted in heating, side by
side under the same conditions, a coal which underwent
distillation and a coke (which could be regarded as
chemically and thermally inert), and noting, by means
of the differential arrangement of thermo-couples devised
by Roberts-Austen for the examination of metallic alloys,
the extent to which the temperature of the coal became
higher or lower than that of the coke at various stages of
the distillation. Complications arising from the secon-
dary decomposition of distillation products were mini-
mised by using a small quantity of coal and sweeping
away the products by a stream of inactive nitrogen.
These precautions would modify such thermal changes
in the interior of a large mass of coal undergoing car-
bonisation.

We do not consider that the heating curves obtained
afford any quantitative data as to the net thermal result
of carbonisation; but it would seem that, in the case of
all the typical coals examined, the net quantity of heat
involved is very small, since there are both exothermic
and endothermic reactions taking place. At certain
temperatures it is not improbable that both exothermic
and endothermic reactions take place simultaneously,
thus considerably increasing the difficulty of detecting
such changes by the experimental method we employed.

One interesting fact in connection with all the curves
for coals so far obtained is that in no case have we found
any evidence of a sudden decomposition at any definite
temperature, such as the exothermic decomposition of
pure cellulose. The decomposition of coal appears to
consist of a very complicated sequence of reactions
which extend over a wide temperature range. In all
cases the distillation, so far as it is revealed by thermal
changes, commences at 250 degs. Cent., and between
this temperature and 1100 degs. Cent, we have detected
a number of fairly well defined stages during which
exothermic or endothermic reactions predominate.

In attempting to explain the various phenomena
observed, we have encountered considerably difficulty,
owing to the lack of chemical data respecting the pro-
ducts evolved from coal at different temperatures. Such
fractional distillations under experimental conditions
comparable with ours as have hitherto been carried out
have been between quite arbitrary temperature limits,
and the most attention has been given to the permanent
gases evolved, whereas those phenomena, chemical or
physical, which gave rise to the expulsion of vapours
condensing to liquids on cooling are probably accom-
panied by thermal effects at least quite as large as those
caused by the expulsion of permanent gases. We have
attempted to correlate the results obtained by Burgess
and Wheeler and Vignon with the thermal change we
observed at 750 degs. Cent., and we hope that, as more
chemical data respecting the sequence of chemical
changes which take place during carbonisation become
available, a much more complete explanation of the
results of our research will be possible. Further, the
critical points noted in the heating curves may suggest
temperatures at which it would be more profitable to
carry out fractional distillations of coal. As a further
attempt to elucidate the phenomena observed, we intend,
to obtain the heating curves of those portions of coal
soluble and insoluble in pyridine or other solvents, and
also to examine the dehydrated cellulose in further
detail.

The coal which has been examined in most detail is a
sample of picked nuts from the Monckton Main seam,
the composition of which is as follows :—

^Volatile matter
Ash

Carbon
Hydrogen
Sulphur
Nitrogen
Oxygen

Per cent.

...	40'2

5’43

...	72’36

4’93

1’64

1’53

...	14’11

* Determined in a platinum crucible by the 1899 method
of the American Chemical Society.

The heating curves of coal obtained in this research
do not afford any quantitative data as to the net thermal

result of carbonisation; but it would seem that the net
quantity of heat involved is very small, since there are
balancing exothermic and endothermic reactions taking
place. At certain temperatures it is not improbable that
both exothermic and endothermic reactions take place
simultaneously, thus considerably increasing the diffi-
culty of detecting such thermal changes by the experi-
mental methods we employed. One interesting fact in
connection with all the curves so far obtained is that in
no case have we found any evidence of a sudden decom-
position at any definite temperature, such as occurs in
the exothermic decomposition of cellulose. The decom-
position of coal appears to consist of a very complicated
sequence of reactions which extend over a wide tempera-
ture range. In the case of the Monckton coal used, the
distillation, so far as it is revealed by thermal changes,
commences at 250 degs. Cent., and endothermic
reactions predominate up to 410 degs. Cent. This is
followed by a short exothermic stage betwe'en 410 degs.
Cent, and 470 degs. Cent., and a second endothermic
period between 470 degs. Cent, and 610 degs. Cent. The
exothermic stage above 610 degs. Cent, is always very
pronounced, and is interrupted between 750 degs. Cent,
and 800 degs. Cent. The probability that exothermic
reactions continue to some higher temperature has
already been discussed.

We are continuing our experiments to the examination
of different types of coal, and, so far as we have gone,
it may be said, in general terms, that the same stages
are in evidence, modified, as is not surprising, by the
nature of the particular coal under examination.

In attempting to explain the various phenomena
observed, we have encountered considerable difficulty
owing to the lack of chemical data respecting the pro-
ducts evolved from coal at different temperatures. Such
fractional distillations under experimental conditions
comparable with ours as have hitherto been carried out
have been between quite arbitrary temperature limits,
and most attention has been given to the permanent
gases evolved, whereas those phenomena, chemical or
physical, which give rise to the expulsion of vapours con-
densing to liquids on cooling may be accompanied by
thermal effects at least as large as those caused by the
expulsion of permanent gases.

Burgess and Wheeler in their study of the volatile con-
stituents of coal (Journal, Chemical Society, 1910 (ii.)
1917 and 1911 (i.) 649) have shown that there is a
critical point in the distillation of coal at some tempera-
ture between 750 degs. Cent, and-800 degs. Cent., and
that when that point is reached there is a large and
sudden increase in the quantity of hydrogen evolved,
while methane is more prominent among the products
evolved at lower temperatures. This observation has
been fully confirmed by Vignon (Compt. Rend., 1912,
155, 1514). Although methane is an exothermic body
—i.e., it is formed synthetically with an evolution of
heat—it by no means follows that all the reactions which
give rise to its formation are exothermic.

It is an interesting point that the figures of Euchene,
Mahler, and Constam, and Kolbe for the heat quantities
evolved in the carbonisation of coal are of the same
order of magnitude as the heat quantities which can be
calculated as likely to be involved in the formation or
decomposition of methane. In the interpretation of
results obtained in this research, it should be remem-
bered that complications arising from the secondary
decomposition of volatile products after leaving the mass
of coal have been eliminated, but it is impossible to say
how much secondary change takes place within the pores
of the newly formed coke.

It seems to us quite probable that the heat evolution
noticed above 610 degs. Cent, in our heating curves may
be connected with those reactions which give rise to the
production of methane. It has already been pointed out
that this heat evolution is not due to one single reaction,
but to a sequence of reactions. These reactions are
interrupted at about 750 degs. Cent, by others which give
rise to the formation of large quantities of hydrogen. It
is hoped that as more chemical data respecting the
sequence of chemical changes which take place during
carbonisation become available, a more complete expla-
nation of the results of this research will be possible.
Further, the critical points noted in the heating curves
may suggest temperatures at which it would be more
profitable to carry out fractional distillations. •

As a further attempt to elucidate the phenomena
observed, we propose to obtain the heating curves of
those portions of coal soluble and insoluble in pyridine
or other solvents, and of substances of known compo-
sition.

Red-Hot Coke and Radiant Heat.

In connection with modern developments in carboni-
sation, many statements have been made respecting the
breaking down of volatile products as a result of contact
with the surface of hot coke, or of the action of heat
radiated from the walls of the retort, or both; but most
of the observations have been made on the large scale
carbonisation processes, in which, owing to the com-
plexity of the phenomena studied, it is very difficult to
isolate cause and effect with any certainty; The need
has been widely felt of experimental work under labora-
tory conditions, in which the various factors would be,
as far as possible, under control, and could be varied
singly at will. The experiments of Burgess and "Wheeler
on distillation of coal are very useful as indicating the
nature of the products evolved at various temperatures,
but in their experiments the gases were removed as
quickly as possible, in order to prevent subsequent

changes; it is these changes with which our experiments
are primarily concerned.

Coal distillations were, in the first place, carried out
under various conditions, so that satisfactory prelimi-
nary evidence might be obtained as to the influence of
red-hot coke and radiant heat on the whole of the
gaseous products of distillation. A Nottinghamshire
(New Hucknall) gas coal was used, the composition of
which was as follows :—

Per cent.

^Volatile matter ................. 37'3

Ash ...	...	...	...	...	3’77

Carbon	...	...	...	...	77’36

Hvdrogen ......................... 5’66

Sulphur .......................... 1’65

Nitrogen	...	...	...	...	1’34

Oxygen	...	...	...	...	10’22

* The volatile matter was determined in a platinum
crucible by the 1899 method of the American Chemical
Society.

The result of subjecting the immediate products of
distillation to the action of radiant heat was an increased
yield of gas equivalent to 1,100 cu. ft. per ton of coal.
It was evident from the results that by passing the pro-
ducts through a column of hot coke a further increase
in the volume of gas was obtained equivalent to 1,250
cu. ft..per ton. It should be noted that these increased
volumes of gas were obtained without heating coal or
coke above 800 degs. Cent. The gas analyses showed
that the influence of the heat had been felt most by the
unsaturated hydrocarbons, and that there has been a
marked increase in the volume of hydrogen. It should
be noted that this extra hydrogen is not obtained at the
expense of the methane, but is most probably a product
of the decomposition of the higher hydrocarbon vapours
and tarry matter generally. It is shown later that the
decomposition of methane under gas retort conditions
is only appreciable when temperatures higher than 800
degs. Cent, are reached.

Attention was turned to the behaviour of individual
volatile products of carbonisation. The problem to
which we addressed ourselves was to determine the
nature and the amount of decomposition suffered by the
volatile products of carbonisation under conditions. of
temperature and time of heating comparable with those
existing in large scale carbonising processes such as in
gas retorts and coke ovens. The time during which the
volatile products are heated varies considerably at dif-
ferent stages, and under different conditions of carboni-
sation. With horizontal retorts and heavy charges it
probably averages 1J to two minutes during the first
half of the carbonising period. The scheme of our
research wTas to take a mixture of equal parts of hydrogen
and methane as the basis, and after noting the influence
of heat on this mixture to add to it small percentages of
other hydrocarbons. Temperatures of 800 degs. Cent,
and 1100 degs. Cent, have been the only ones at which
the decompositions have as yet been studied, and only
experiments in which the gases were passed through a
tube packed with coke are recorded.

It is seen that at 800 degs. Cent, only about 2 per
cent, of the methane is decomposed in one minute,
although the quantity of methane present is consider-
ably in excess of the equilibrium ratio. The decompo-
sition of methane is very rapid at 1100 degs. Cent., as
the above shows. As a result of the decomposition there
was an increase in volume of about 44 per cent.; but
there is a considerably loss of potential heat or calorific
value, as the carbon of the decomposed methane is lost
so far as the gas is concerned.

The decomposition of ethane under these conditions is
rapid, but not complete in 46 seconds, at 800 degs.
Cent. The chief products of the decomposition are
ethylene and methane. The decomposition was more
extensive at 1100 degs. Cent., 88 per cent, of the
ethane being decomposed. At 800 degs. Cent, the
decomposition of ethylene is by no means complete in
45 seconds. But at 1100 degs. Cent, heating for 35
seconds completed the decomposition, the only gaseous
products being methane and hydrogen. The decompo-
sition of benzene at 800 degs. Cent, was negligible, but
no trace of benzene could be detected after heating the
mixture at 1100 degs. Cent. The solid and liquid pro-
ducts, if formed, were too small in bulk to be detected.
This study is only in its first stage, particularly since,
except in the somewhat rough preliminary experiments
made on the whole of the gaseous products from coal,
we have only obtained results for the influence of red-
hot coke. The experiments are being continued on the
action of radiant heat, and on the decomposition of such
substances as the higher hydrocarbons, of ammonia, and
of gases containing oxygen, which has so far been an
excluded element.

Compulsory Co-Partnership.—The text has been issued of
a Bill introduced by Mr. O’Shea, M.P., to make co-partner-
ship compulsory. The Bill proposes to empower the Board
of Trade on receiving a requisition signed by not less than
half the persons employed by a statutory or other company,
to appoint an arbitrator to reconstruct the company or under-
taking on a co-partnership basis. Shareholders are to
receive preference dividend equal to the average return dur-
ing the three preceding years, and are to be entitled to
elect two directors for every one which the employees are
to be authorised to appoint. The arbitrator is further
to determine a maximum dividend for one or more classes
of shareholders, and also a minimum wage for employees
over 21 years of age, not being learners. After provision
for reserve and depreciation, one-third of the net profits are
to go to shareholders, and two-thirds to employees in either
IT ordinary shares carrying not more than 4 per cent,
interest, or in cash, or both in a specified manner. In any
year in which the preference shareholders receive the maxi-
mum dividend the whole of the net profits after such pay-
ment and after payment of a dividend not exceeding 4 per
cent, on the ordinary shares shall be distributed among
employees. Holders of ordinary shares are to be entitled
to vote with preference shareholders in the election of
directors.