*170

THE COLLIERY GUARDIAN.

October 9, 1914.

____________________________________________________________________________________

Honeycomb and Clinker Formation.*

By 5. W. PARR.f

As a general proposition the first step in any attempt
to solve a difficulty should doubtless be taken along the
line of a search for the underlying cause. If we approach
the matter of honeycomb formation on flue sheets from
this standpoint, we are confronted at once with the fact
that the cause must of necessity be more or less obscure
owing to the difficulty of obtaining experimental
evidence in the vital or formative phase of the matter.
We may be justified therefore in resorting to the process
of theorising or setting up of a hypothetical explanation
in the hope that by this means some suggestions may
be developed which will remove or mitigate the evil. If
this should be the result of our hypothetical speculation,
I think you will agree that it does not make any differ-
ence whether our hypothesis is right or wrong, so long
as the desired end is attained. I may say, also, in
passing, that perhaps this theoretical viewpoint of
method in the present instance will be regarded as an
apology why I attempt to discuss a subject concerning
which I have so little experience.

The first suggestion which offers itself to the minds of
everyone who approaches this topic is the question of
the fusibility of the ash. A great deal of activity is
just now being carried on in connection with the study
of ash ■ fusibilities, and it is entirely possible that the
clinkering which occurs in ordinary steam heating
appliances, in which the fusibility of the ash doubtless
plays a large part, may have some relation to the
clinkering on flue sheets, but I am disposed to think
otherwise. For instance, in a recent number of the
Journal of the Society of Chemical Industry! results are
given on an elaborate series of ash determinations, in
which both the fusibility of the various ashes and the
component parts of the ash are set forth, with the con-
clusion that the constituents as determined by chemical
analysis seem to offer no explanation for clinkering on
the grates. The statement is further emphasised that
high sulphur and high iron seem to be wfliolly inadequate
as offering an explanation. I certainly agree that we
get little light in studying either the chemical compo-
sition of the final ash product as it ordinarily comes
from the grates or the composition of the final clinkering
product which runs on the grate bars or fastens itself
to the flue sheets. Take this illustration. I have from
Mr. J. G. Crawford, fuel engineer for the Chicago,
Burlington and Quincy Railroad, under date of
December 12^ 1911, the following results of analyses
covering a sample of honeycomb and the ash of the coal
used at the time.

Analysis of honeycomb and coal ash from engine 2133,
Chicago, Burlington and Quincy Railway, November
1911 :—	Honey- Ash in

comb. coal.
Per cent. Per cent.

Volatile on ignition ..........	0*65	...	0*00

Silica (SiO2).................. 51*82 ... 50*78
Iron oxide (Fe2O3) ............ 12*19 ...	9*53

Aluminium oxide (ALO3) .... 21*34 ... 28*20
Calcium oxide (CaO)..........	8*97	...	7’84

Magnesium oxide (MgO) ......	1*72	...	2*78

Sulphuric anhydride (SO3) ....	0*44	...	0*26

Alkali oxides (K2O and Na20)

by difference ...................__ 2*87	...	0*61

Total............... 10000 ... 100*00

Numerous analyses of honeycomb from other sources
show a substantial agreement with the analyses of the
ash of the coals, which were being used on the engines.
Now, as a rule, these ashes of themselves, or as analysed,
are not of a type to clinker easily. Notwithstanding
that fact, the honeycomb prevails. Some coaling
stations, it is true, will furnish a coat with more fusible
ash than others, but it does not appear from any data
which we have been able to secure that these were the
coals which were giving the trouble.

In studying the conditions of fusibility of the mineral
constituents of coal, the composition of the ultimate ash
product will throw but little light on the subject, if,
indeed, it may not be actually misleading. We must
keep in mind the fact that there are two stages in the
various transformations that are going on. Each stage
has its own fusion temperature, and the components
must be separated; and any study of the final stage will
have value only as it may furnish information as to the
possible conditions existing in the intermediate stage of
the combustion process. I refer here in the main to the
iron pyrites in the coal. The average of this constituent
for most of the coals of the Mississippi valley will be in
excess of 3 per cent, of sulphur, which means substanti-
ally that 6 per cent, of the weight of the coal is in the
form of pyritic iron; and it is not unusual for this con-
tent to reach 10 and even 12 per cent, of the gross
weight of the coal. If this material is allowed to burn
at a leisurely rate until all of this sulphur has burned
out, the resulting product is ferric oxide, Fe2O3, and a
study of the ash residue in which the iron has reached its
final state will not disclose any increased tendency on
the part of the ash to clinker. So far as this sort of
investigation would be concerned, therefore, we would
say that this high content of sulphur and iron in the
coal does not produce a final ash which is any more
fusible than a coal where this constituent is low.

Studying the iron pyrites in the intermediate stage,
we find that it very readily parts with one-half of its
sulphur content, and drops down from an indicated com-
position of FeS2 to the composition of FeS. This action
does not require any oxygen to remove the one atom of

* A paper read before the International Railway
Fuel Association, Chicago, Illinois.

t Professor of Applied Chemistry, University of Illinois.

I Vol. 33, p. 1144.

sulphur, which is purely the result of heat, and takes
place at a temperature between 750 and 900 degs. Fahr.,
400 and 500 degs. Cent. Now the further removal of
the final atom of sulphur does not take place at any
ordinary temperature, and, indeed, requires about
1650 degs. Fahr., 900 degs. Cent., for its rapid elimi-
nation; or, provided air in sufficient quality is avail-
able, it will burn out to a resulting product of oxide of
iron, Fe2O3, which is not only highly infusible, but does
not add anything to the fusibility of its associated
mineral matter in the coal. So I repeat again, a final
ash, with all of the iron burnt to the Fe2O3 stage, may
be a very infusible mixture, and the more ferric oxide
present the more fusible it may be; but in this stage
its fusibility not only bears no relation to the inter-
mediate stage, but it may, and usually does, have a
directly opposite tendency. It is this intermediate
stage in the transformation, therefore, namely, the iron
sulphide in the FeS form, which, it seems to me, is of
vital concern in studying this problem, because it is an
easily fusible constituent, and, being present in an
amount so large, as we have already indicated, it fur-
nishes, when in this stage, magma sufficient in amount
to dominate the mass and give it a pasty character. Now
it must be evident that if conditions are favourable for
the formation of this intermediate product, and these
conditions are maintained for any considerable length
of time, at the high temperatures of the fire box, this
material will run into clinker.

We do not need to go far for evidence that these
clinkers are some sort of a compound of sulphur, for
doubtless everyone is familiar with the pronounced
sulphur smell that arises from these live clinkers when
pulled out of the fire. They still retain the power of
combustion for the reason that all the sulphur retained
in this sulphide of iron form burns readily and continues
to burn for some time after being exposed to the air.
The tendency to form clinker, or, to put it on the basis
fusibility, the fusibility of the ash at this stage, bears
little relation to the ultimate product that comes in the
form of loose ash. We should rather consider the
fusibility of the mass in which this iron sulphide has
reached the half-way stage. Here is a condition of
fusibility which not only bears no relation whatever to
the final product, but it is also a condition in which it
is exceedingly difficult to catch the material in place,
or in a condition to measure its degree of fusibility.
Hence the futility of the average ash analysis as show-
ing anything like the fusion temperature that really
exists in the process of combustion.

It will be seen from the above that we have been
discussing the tendency towards the formation of
clinker on the grate bars. What explanation may that
afford for the formation of clinker on the flue sheets?
To this we would reply that two conditions must exist
in order to furnish any explanation along the lines sug-
gested. First, a chemical condition in which the pyritic
iron is'transformed only to the ferrous sulphide, or FeS
stage. We have already said that this transformation
may be brought about by heat alone; but it is also true
chat to remain in the ferrous sulphide stage it must
have an insufficient supply of air, otherwise it would
continue the process of decomposition until it reached
the oxidised form. Chemically speaking, therefore, we
need only a moderately high heat and either a shortage
in the oxygen supply or such a limitation in the time
element that the oxygen action is incomplete.

We should follow the possibilities a step further at
this point. If the ash in this half-way condition should
be in contact with the higher temperatures of the fire
box, or be moved into such a zone by stirring or by
die movement of the grate, it is entirely possible that
large masses thus exclude the further action of oxygen.
We will by this means prevent the passage of the
material from the easily fusible ferrous iron stage. It
will be trapped, so to speak, in this form.

Second, there are physical conditions which provide
an explanation for the solid material transferring itself
from the fuel bed to the flue sheets. I cannot conceive
it possible that the burning mass on the grate bars can
possibly arrive at a condition of fusion in such a liquid
state that the air draught could lift the molten material
into the combustion current, and thus throw it against
the flue sheets. On the other hand, it is entirely
reasonable to suppose that a condition of this sort in
the burning mass on the grates will cut down the oxygen
supply to such an extent that the combustion chamber
itself is somewhat deficient in oxygen. We are driven,
therefore, to the hypothesis that the coal itself, as it is
thrown into the fire box, has those physical conditions
present which are conducive to the furnishing of
particles which are fine enough to be drawn up into the
draught currents. Such finely divided material is
especially pronounced in the run-of-mine which so
largely predominates in the fuel supply for railroads,
especially of those States where the custom prevails of
putting out coal on the run-of-mine basis. Suppose, for
example, that the coal, as fired, has an undue amount of
fines. This in itself not only has a tendency to cut
down the amount of air which may pass through the
grate bars, but it also makes it possible for a large
amount of fine material to be drawn up by the draught
and into a combustion zone deficient in oxygen and of
sufficiently high temperature to change the pyritic iron
to the FeS stage. In this condition it is thrown against
the front end of the fire box, and will lodge there, especi-
ally if there is a slight leak about the flue opening. Now,
if we examine these incrustations, our first conclusion
would be that on this hypothesis our mass of clinker
would be largely impregnated with the sulphide of iron,
FeS, but a test for such sulphur in these clinker forma-

tions does not reveal any appreciable amount of that
constituent. Our hypothesis is not endangered thereby,
for the reason that, as already mentioned, with a heat
sufficiently high, say about 1600 degs., the sulphur is
dissociated and driven off. This liberation of the
sulphur causes small gas pockets or bubbles throughout
the mass, and this is very characteristic of these clinker
formations. Precisely the same thing is met with in the
burning of shales for brick manufacture, which are
highly impregnated with iron pyrites. At one stage
in the burning, where the ferrous sulphide condition is
reached and one-half of the sulphur has been driven out,
that stage which we have been describing is reached. If
now the heat is raised to the point where the sulphur
dissociates the bricks puff up into a spongy mass and are
worthless. If, on the other hand, a slow heat is con-
tinued over a longer time while the bricks are still in
the porous stage, the oxygen will perform its functions
by having access to the sulphur and burn it all out,
thus yielding dense masses even from shale of this
description. Exactly the same conditions prevail in
the fire box of a locomotive, excepting that usually no
slow moving processes are in evidence. The term
“ honeycombing ” itself is somewhat of an explanation
of the chemical processes which are going on. Particles
of coal which are fine enough to be caught up by the
draught have, in the short distance they may travel,
about the right conditions as to temperature, oxygen
supply, and the time element to bring them to this inter-
mediate or easily fusible stage. They are thrown, there-
fore, against the flue sheet in a semi-pasty condition.
The outer surface glazes over, and no more oxygen may
reach the interior. They are, however, subjected to the
most extreme heat of the fire box, sufficient to dissociate
the remaining sulphur, which thus passes off as a gas,
producing the spongy or honeycomb effect. The iron
remains behind, it is true, but as ferrous iron, and in
this form it readily unites with the silica present to form
an easily fusible slag. The same ratio of iron to silica,
if it were burned to the ferric stage, would represent a
combination with altogether different characteristics,
and a very much higher fusion point.

Some confirmation of this rather hypothetical line of
reasoning is furnished by the study of the ordinary flue
incrustations, and by an examination of the constituents
of the clinker with reference to the form in which the
iron occurs, ferrous iron being prominently in evidence.

A brief discussion of mining methods and mine output
will be suggestive at this point. In the first place the
fine material produced in the ordinary process of mining
is much higher in ash than the lump coal. The amount
of pyritic iron is in a higher ratio also. In a series of
analyses conducted on coals from 75 mines of the State,
each mine being represented by a sample of screened
lump and one of screenings, the results show an almost
uniform ash percentage in the screenings, which is at
least double that of the ash in the lump. Now, in
run-of-mine coal the product is somewhat deceiving, for
the reason* that it has the appearance in the mass of
being very largely lump material. Indeed, it is for
occasional car lots of run-of-mine to be fully equal to
the best screened lump from that mine, but there is a
very considerable amount of fine material which must
come along somewhere in the output. It should be
recalled that after the blast and the breaking down of the
coal at the working face the miners enter and clean up
the rooms by sending out first the coarse or lump
material. At the clean-up w’hich is made before new
drill holes are started that part of the underlying floor
which has been more or less pulverised and loosened in
the various processes which have been going on is
shovelled up and sent out along with the coal. In this
way it is easily seen that the fine material is much
higher in ash, and is. composed of mineral constituents
which are usually in themselves higher in sulphur.
We have, therefore, in the run-of-mine material exactly
those physical conditions of fineness of division and high
content of iron pyrites which, in the processes already
described, are conducive to the furnishing of pasty
particles which can be made to grow by small accretions,
forming, as we have seen, the honeycomb structure on
the flue sheets.

From an examination of numerous specimens in which
a chemical test for iron in the ferrous form has been

made, there is indicated the presence of this material,
and to that extent affording a confirmation of our hypo-
thetical line of reasoning. It will be seen also that the
various physical conditions described are in evidence.

One other phase of ash fusibility ought to be men-
tioned in this connection. With the work of the

Illinois Geological Survey the fact has been developed
that there is rather an unusually large distribution of
coals having a high percentage of calcium carbonate pre-
sent. This material, when it reaches a certain amount,
forms a fluxing agent with the silica present; but these
cases are hardly common enough to explain the every-
day occurrences of honeycomb. However, it is given
here as undoubted evidence, where such quantities of
this material are met with, of a reason for the easy
clinkering property of the ash. Four coals were taken
in which the silica of the ash was approximately the
same in amount. In this series the calcium oxide

show’ed a marked effect on the fusibility of the ash,
while the iron oxide was substantially without influence.
The table below shows four such coals arranged in the
order of fusibility, No. 1 being easily fusible and No. 4

being the least fusible :—

1.	Carterville, Ill., Madison

No. 8, No. 2255........

2.	Cambria, Ill., C. M. B. C.

Co., No. 2254..........

3.	Cambria, Ill., M. C. Co.,

No. 2256...............

4.	Herrin, Ill., C. H. C. Co.,

No. 2257...............

SiO.. JfQ3' CaO. MgO.

Perc. Perc. Perc. Perc.

48*1 ...	27’2	...	23*8	...	0*8

47’5 ...	46*7	...	5*0	...	0*8

56*5 ...	37’6	...	5*0	...	0*9

53*3 ...	41*0	...	4*9	...	0*8