854

______________________________________________________________

duction is from two beds lying wholly under the sea.
The output is about 500 tons per day. Four shafts
were sunk on the shore, each 119 ft. deep, from which
two beds are worked to a distance of about 4,000 ft.
from the coast. The cause of the accident was the
entering of water into the underground workings
through a fault in a bed of sandstone 155-4 ft. thick,
above which there is an alluvial deposit of clay and
sand 82-6 ft. thick.

A small flow of water occurred when the fault was
first reached. The final inrush followed the breaking
of a hole about 4 ft. square in the floor of an entry of
the upper bed, a few feet back from the fault. Through
this the water entered so rapidly that the mine was
completely flooded in two hours. The quantity enter-
ing was estimated at 392,000 cu. yds. The sea bottom
was lowered 60 ft. over a small area, showing that a
considerable amount of solid matter was washed in.
The opening was apparently sealed by the solid
material, and it was planned that the mine should be
re-opened by filling the depression in the sea bottom
with clay and sand, pumping out the water, and build-
ing dams to protect the workings from any future
break.

(To be continued.)

_________________________

ENGINEERING PHASES OF SMOKE
ABATEMENT.*

By Osborn Monnett.

If I were to answer the question, what are the three
most important things in smoke abatement, I would
be tempted to say, first, draught; second, draught;
and third, draught. Draught is necessary to the
smokeless burning of high volatile coal. We must
have oxygen; by the draught we get the oxygen in
over the fire where the volatile is being distilled. In
investigating the smoke problem in Chicago, it was
found that draught was so important as to overshadow
every other consideration, and through a long series
of investigations covering some thousand separate and
distinct studies of boiler settings, a curve was
developed which tells us how much draught we need
for different rates of combustion. Roughly the rule
is this: We need 0-01 in. draught over the fire per
pound of coal burned per square foot of grate surface
per hour. That tells the story almost entirely through-
out the range of the curve. Beginning with 0T5 in.
of water over the fire, we are able to burn 15 lb. of
coal per square foot of grate surface per hour smoke-
lessly. Cases can be cited where 0-15 in. has burned
20 and even 25 lb. of coal, but it did not burn it smoke-
lessly.

The limit rate of successful smokeless operation
for hand-fired units is about 26 lb. of coal per sq. ft. of
grate surface per hour. Above this point reliance is
placed on mechanical stokers. You will note the
draught required follows the rule almost exactly in the
lower rates. When we get to 0-5 in. over the fire, we
can take care of only about 40 lb. satisfactorily. This
relates more especially to Illinois (free burning) coal.

The rate of combustion in any plant is fixed first by
the load. When the load is known, fix the ratio of
grate surface to heating surface. These two factors
decide the rate of combustion. Having this, we are
able from the curve to judge how much draught we
need over the fire.

On return tubular boilers, a ratio of heating surface
to grate surface of 40 to 1 works out very nicely. With
mechanical stokers the ratio of 50 to 1 is standard.
For water tube boilers in Chicago practice we ask
for a damper ratio of 4J to 1; that is, the grate sur-
face* connected with the boilers is 4J times the free
opening in the damper. The ratio of grate surface to
breeching may be as 5 to 1; that is, the actual free
opening in the breeching has a ratio of one-fifth of the
actual connected grate surface. Then the total con-
nected grate surface to stack works out in the same
ratio, 5 to 1; that is, the area of the stack should show
a free opening of one-fifth the total connected grate
surface. This holds good for the average height of
stack. These ratios are more liberal than found in
ordinary practice, but the plant with the liberal ratios
is the plant that is getting away from the smoke and
carrying the load.

In laying out the stack for any plant, it is neces-
sary to know how much draught loss there is in the
particular type of boiler setting selected, so as to know
that the proper amount of draught over the fire will
be provided. This loss runs from 50 to 65 per cent,
of the draught available at the stack side of the
damper.

Allow about 0-1 in. draught per 50 ft. of breeching,
and 0-05 in. draught for every additional boiler over
one connected to the breeching, because of the inrush
of gases having a choking effect on the boiler farther
from the stack; if the boiler happens to be dead, we
have an inrush of cold air which has the same effect.
After getting the draught required in this way at base
of stack, dividing by constant K determines the height
of the stack in feet. This is the standard formula
published some years ago by Stirling, and it has worked
out very nicely: —

D = 0 52 H x P (I _ A

\T T7

in which D = draught in inches of water, H = height
of stack in feet, P = atmospheric pressure, T = atmo-
spheric temperature (absolute), Tx = temperature
stack gases (absolute), K = 0-52 x 14-7

Temperature
stack gases. K.

750	.... 0-0084

700	....	0-0081

650	..... 0-0078....

600	.... 0-0075

• 550	.... 0-0071

Temperature
stack gases. K.

500	.... 0-0067

450	.... 0-0063....

400	.... 0-0058

350	.... 0-0053....

________________________________________________________

* From Proceedings of the Engineers’ Society of Western
Pennsylvania, vol. 32, No. 9.

THE COLLIERY GUARDIAN.

In analysing stack temperature conditions in a great
many plants, it was found that the average gas tem-
perature in the stack is 418 degs. That seems rather
low, but was obtained from. 50 different plants in com-
mercial operation with all kinds of equipment;
500 degs. is more likely to be the breeching tempera-
ture than the stack temperature, as we are losing heat
all the way along.

In studying the smoke problem in Chicago, it was
soon realised that in order to make any kind of a
showing, it would be necessary to develop some type
of hand-fired furnace that could be operated within
the smoke limits; 90 per cent, of all the boilers
installed are horizontal return tubular boilers, and
they make a large percentage of the smoke of the
average city. The old standard Hartford setting
consisted of a low setting with about 24 in. from the
dead plate to the sheet, a plain bridge wall and shallow
combustion chamber, small damper, and everything
unfavourable to smokeless operation. Thousands of
these settings had to be cleaned up.

The first attempt to make such a setting smokeless
was to spring a deflection arch behind the bridge wall,
the idea being to break up the rush of volatile matter
and cause it to go under-the arch and pass into the
combustion chamber. It was nothing but a make-
shift. The next step, one that was developed about
1896, was the old McGinnis arch setting, consisting of
two single spans of arch work, one behind the bridge
wall and the other over the grate, the idea being to
use the coking method of firing, and carry the volatile
matter down under this arch, and impinge on this
deflection arch, at which time it would be pretty well
mixed with air and consumed. This setting, when
used with rates of combustion not higher than 151b.
of coal per sq. ft. of grate surface, worked out very
satisfactorily when carefully attended to. But it
depended on the fireman to such an extent that in
1907, when the engineering department was re-orga-
nised, it was not considered a satisfactory setting.

At this time, the authorities were all agreed that
the Dutch oven was the best thing in smoke abate-
ment. A great many of these Dutch oven furnaces
were installed. The objections to the Dutch oven were
that it took up so much floor space, was excessively
high in first cost, and also high in maintenance. It
took a high head room to get it under the boiler, and
the objections were-so strong that the department
finally took to installing semi-extension Dutch ovens.

Finally, they came to flush front Dutch oven
settings. All these Dutch oven settings were failures.
They made more smoke than the standard Hartford
setting. The reason for this was rather obscure for
some time, but finally it was reasoned out in this way.
When you throw high volatile coal on a fire, there is
a normal distillation of volatile matter due to the heat
of the fire. But when you throw coal on a fire with
a Dutch oven you get not only the normal distillation
of volatile matter, but you also get an artificial distil-
lation due to the high temperature of the heat radi-
ating from the red hot firebrick over the fire, resulting
in a double distillation of volatile matter, and it causes
dense smoke in spite of all efforts to prevent it.

Finally, by taking off the brickwork over the fire, a
smokeless furnace was obtained. This reduced the
rate of volatile distillation, because there was no radi-
ation of heat from the red hot firebrick. It also
increased the steaming capacity of the boiler, because
there was a direct radiation from the grate into the
boiler. Thus we account for three things necessary
for perfect combustion: temperature, in zone of high
temperatuTe back of bridge wall; oxygen, which we
get in through panelled doors; and air siphon steam
jets, and a mixture against the bulkhead wall over the
deflection arch. A setting finally adopted for stan-
dard in the department is the double arch bridge wall
setting. In handling such a setting, the alternate
method of firing is best.

There are some few points regarding operation to
which I would direct your attention, if the best results
are to be expected from hand-fired furnaces. The
ordinary method of firing a new boiler is to build a
fire with kindling, and throw in a lot of fresh coal.
This makes a great deal of smoke. By firing from the
top and covering the clean grates with green coal to a
depth of 4 or 5 in., and then building a fire at the
bridge wall with kindling or live coals, it is possible to
get up steam without violating the smoke ordinance.
The volatile matter from under the fire has to pass up
through a high temperature zone, and becomes
consumed.

Another proposition is cleaning of fires. The ordi-
nary method is to rake out one side, wing over a lot
of live coal, and fill up that side with fresh coal. This
again makes smoke. If one side is cleaned out, and
the grates covered with green coal, and then red hot
coal thrown on top of the green coal, it is possible to
clean a fire without smoke. This method is being used
in Pittsburg in firing locomotives with great success
at the 28th Street yards of the Pennsylvania Railroad,
Lines East.

Another phase of this hand-fired proposition is the
burning of coal at low rates of combustion, for heat-
ing plants. This is really a harder problem than that
pertaining to high-pressure work. We have poorer
firemen, less draught, and everything is against
successful operation.

Studying the progress made in burning coal at low
rates of combustion takes one back to the time of
James Watt in 1769. We started to experiment with
the coking method of firing as we know it to-day, but
discovered right away that it was the human element
in hand-fired units which really decided the proposi-
tion. A good fireman, as soon as he finds out that he
is a good fireman, graduates into a second- or third-
rate engineer. We have lots of engineers, but very
few real firemen.

In 1785, Watt made an improvement on his coking
method of firing by incorporating a magazine feature,
keeping the magazine full and charging the furnace
by shoving through with a slice bar. In 1804, Wolf
put another feature in his system in the dumping grate

May 4, 1917.

_______________	________

bar, which we have even to-day. It made a complete
system of burning coal, with a low rate of combustion,
with a minimum amount of smoke, and incorporated
the magazine feature. This was the first attempt to
do anything of that kind.

The idea was good, and inventors kept trying to
improve on the methods. In 1838, Rodda built a bulk-
head over his fire, causing the volatile to pass through
the fire, but burned out the grates. In 1850, Clark
divided his furnace into two, front and back, with a
bulkhead in the ashpit. This also burned out rapidly,
and was not feasible.

Fairbairn, in England, in 1860, passed the volatile
matter down through the fire. He had the germ of
the down-draught idea as we know it to-day. A water-
cooled grate made the down-draught boiler permanent,
and it is now standard equipment for heating work.
There is danger of making smoke in a properly oper-
ated down-draught furnace, because when sliced it is
through a zone which has already given up its smoke-
forming matter, being all red hot coal from which the
gases have passed away. Charging such a furnace
does not disturb the fire, and, therefore, can be done
without danger of smoke.

Passing to the mechanical stoker, I will characterise
the settings as horizontal and vertical baffle systems.
The horizontally baffled system I will dismiss in a few
words. In a nutshell, the horizontally baffled furnace
is a fool-proof proposition, and any stoker combina-
tion can get smokeless results, provided draught is
available. The only head room required is that neces-
sary to get the particular type of stoker under the
boiler selected. The vertically baffled water tube
boiler is the proposition that requires careful thought.

DISCUSSION.

Mr. O. P. Hood said he had nothing to add, except
perhaps to answer the question that was asked as to
whether these figures, which were probably worked out
with Illinois coal, were just as good for Pittsburg coal.
Some investigations of the Bureau of Mines along this
line might be interesting. Six or eight years ago a
board of engineers was got together to advise what
was then the Geological Survey as to what lines of
investigation the Government should take up that
would be really helpful. The board propounded the
problem, to find out whether it was possible to design
furnaces in a rational manner; rather than by rule of
thumb; that is to say, was it possible to take o als
of different analyses, coming from different parts of
the country, and by proper computation, design a
furnace to fit those particular coals? Mr. Kreisinger
was in the room, and he could say what results he
obtained with three coals, Pocahontas, Pittsburg, and
Illinois, all burned in the same furnace under the
same general conditions at rates varying all the way
from 10 to 15 lb. per sq. ft. per hour up to 60 or 70 lb. ;
this in an effort to discover how much combustion
space is needed to reduce combustible gases down to,
say, 2 per cent.

Mr. H. Kreisinger said he could only in a general
way give the results of the experiments referred to
by Mr. Hood. The complete data relating to these
investigations could be found in the Bureau of Mines’
publications, entitled “ Combustion in Fuel Bed of
Hand-Fired Furnaces ” and “ Combustion of Coal and
Furnace Design.” The combustion investigations
were made in a furnace equipped with a side feed
Murphy stoker, and provided with a combustion space
about 40 ft. long, and, on the average, 3 ft. by 3 ft. in
cross section. At the end of this combustion space
the furnace gases were discharged under a Heine
boiler. Along the path of gases through this long com-
bustion space were inserted water-cooled gas samplers
at space intervals of about 5 ft. In order to reduce the
error, due to stratification of gases, four to nine
samplers were used at each of these cross sections, so
that about 70 gas samples were collected simul-
taneously. The analyses of these samples furnished
the necessary data to study the process of combustion
along the path of gases. The object of these investi-
gations was to determine how much combustion space
is needed to obtain any desired degree of completeness
of combustion, when the kind of coal and the rave of
combustion is given. The results of these investiga-
tions showed that only 6 to 7 lb. of air per pound of
coal burned can be forced through the fuel bed, no
matter how fast the air is supplied, The faster the
air was supplied, the faster the coal in the fuel bed
burned or gasified, the ratio of the weight of air to the
weight of coal gasified remaining very nearly constant
for all rates of combustion. This was true when the
fuel bed was level and free from holes, and about 6 in.
thick. The gases rising from the fuel bed contained
no free oxygen, and 20 to 30 per cent, of combustible
gases, besides large quantities of tar. The combus-
tible content w’as the result of distillation of the vola-
tile matter and of the reduction process of CO2 to CO.
All of this combustible must be burned in the combus-
tion space. If this was not done, smoke was produced,
and a large part of the fuel was wasted. The question
of a proper size of combustion space was, therefore, a
very important part in the problem of designing smoke-
less furnaces. Inasmuch as the gases rising from the
fuel bed contained no free oxygen and a large per-
centage of combustible, a sufficient quantity of air
must be supplied over the fuel to make complete com-
bustion possible. The quantity of air which must be
thus supplied varies from 7 to 12 lb. per pound of coal
fired. It should be introduced into the furnace in
small streams as close to the fuel bed and at as high
a velocity as possible to facilitate its mixing with the
combustible. As the mixture flowed through the com-
bustion space, the combustible burns and the gas
analyses of the samples taken at the successive sections
of the path of the mixture showed diminution of the
percentage of combustible, and an increase in the pro-
ducts of complete combustion. The drop of the per-
centage of combustion is very rapid at first, but
became slower and slower as the mixture flows farther
away from the surface of the fuel bed. For any given
’ fuel, the rate of combustion and air supply, there was
a relation between the completeness of combustion and