May 31, 1918.

THE COLLIERY GUARDIAN.

1097

GAS FIRING BOILERS.*

By T. M. Hunter, M.A., B.Sc., A.M.I.C.E.

Gas is an ideal boiler fuel, the supply being controlled
by turning a valve, and as there are no clinkers and
dust to be removed, the labour and upkeep charges
are very greatly reduced. The capital cost of the
plant is small, and the cost of conveyors, bunkers,
stokers, and ash-handling plant is saved. Gas is
simple to use, and since it is possible to maintain the
boiler efficiency at a high level when the fuel supply
is regular and uniform, gas firing ought to show in
every case a higher efficiency than coal firing.

Up to the present time the boiler has been built
round the fire-grate, and, in the case especially of
water-tube boilers, a most efficient steam-raising
apparatus has resulted. In most cases where gas
firing is to be used, it is necessary to retain the fire-
grate, so that the existing designs of boilers are
likely still to be used at most works. In some cases,
where coal firing is not required as an alternative,
the designer is free to make a boiler suit the special
needs of gas firing. The Bonecourt boiler is one
example, and many other types will follow as the
demand arises for more economical boilers to use gas
fuel only.

Cleaned gas leaves no deposit on the tubes or flues
of the boiler, and no soot on the tubes of the econo-
miser or the plates of the air pre-heater. Using
softened water and cleaned gas, the boiler should
always work in its most efficient state. The econo-
miser needs no scrapers on its tubes, and as it can
be enclosed in an airtight case, all the cold air leak-
ages caused by the scraper chain holds are avoided,
and the economiser is made more efficient. Gas makes
the separately fired superheater a most convenient
arrangement—an important matter in connection with
that development of the steam turbine which calls for
reheating the steam that has passed through one or
more of the elements.

The defects of industrial gas as a fuel are that its
pressure and composition are liable to variations, and
that most blast-furnace gas is heavily laden with
dust. Using gas varying in pressure and calorific
value, it is difficult to set the combustion arrangements
so as always to burn the gas at the highest efficiency.
Even on the most modern installations it is in prac-
tice found necessary to burn the gas, in its average
state, with 20 per cent, excess of air in order to avoid
the serious loss caused by unburned CO in the waste
gases, when the gas pressure or calorific value
increases.

Dust in gas fuel gets on the boiler tubes and flue
walls, and very quickly reduces the efficiency.
Cleaing involves a loss of sensible heat, which amounts
often to from 10 per cent, to 15 per cent, of the
total calorific value of the gas, so that the saving
in the boiler from the use of cleaned gas must be
more than this, and must also pay for the power used
in the cleaning plant and for the maintenance and
capital charges on that plant. In water-tube boilers
it is quite a question whether cleaning will pay, but
with Lancashire boilers, which can only be cleaned
periodically, cleaning will surely pay. In most boilers
the burning gases are allowed to impinge directly on
the tubes, and thus the dust is burned on to the
tubes and must be removed like scale. In water-tube
boilers, however, it is possible to arrange gas burners
in such a way that a great amount of the dust can
be fused by the flame and left on the floor, and there-
fore only a small amount of the dust in the gas has
to be blown off the tubes by the steam lance.

Tests made by the United States Steel Trust have
shown that the average boiler efficiency of water-tube
boilers fired with uncleaned blast furnace gas was
55 per cent., which could easily be improved to 65 per
cent., but that it was doubtful whether 70 per cent,
efficiency could be maintained for very long at any
of the plants under ordinary working conditions.

Water vapour in the gas in any quantity has a
very serious effect on the efficiency of combustion,
lowering the flame temperature and increasing the
amount of heat carried away to the chimney by the
waste gases. Such moisture should, wherever possible,
be eliminated by cooling the gas to 30 degs. Cent, or
lower.

As regards the recovery of potash from blast furnace
gas, the dry processes of cleaning give the dust in
a condition ready for immediate sale. The wet
cleaning process has the additional disadvantage, from
the gas firing point of view, that all the sensible
heat in the gas is lost, and that the gas is generally
loaded with a fine spray of moisture, which must be
evaporated and raised to the flame temperature.

In the Halberg Beth process the temperature of the
gas as it leaves the cleaning plant is about 70 degs.
Cent. In very many cases the moisture supplied in
the ore and the coke is sufficient to saturate the gas
at this temperature. Any additional moisture present
is condensed when the gas is cooled before the cleaning
plant. The gas at 70 degs. Cent, carries only some
2 B.Th.U. per cu. ft. of sensible heat, which is almost
a negligible amount. If this gas is burned and the
products of combustion pass to the chimney at 250
degs. Cent., the moisture present in the gas is re-
sponsible for a loss in efficiency of 5| per cent, com-
pared with the result to be attained by burning the
same gas supplied at atmospheric temperature and
saturation. This loss rises very rapidly with the
temperature of the saturated gas. At 100 degs. Cent,
the gas carries 3 B.Th.U. of sensible heat per cu. ft.,
whereas the moisture carried at just under that
temperature is three times as much as at 70 degs.
Cent. Thus it will not pay to retain these small
amounts of sensible heat when the gas is loaded with
moisture, further cooling being necessary.

Gas cleaned at a high temperature by the Lodge
electric method contains all the moisture put into
the furnace. On the other hand, the sensible heat
carried by the gas is much larger. At 200 degs. Cent,
it is 7 B.Th.U. per cu. ft., at 250 degs. Cent, it is

* From a paper read before the South Wales Institute of
Engineers, Friday, May 24, 1918.

9 B.Th.U., at 300 degs. Cent, it is 11 B.Th.U.
Further, the temperature of the waste gases leaving
the boiler is little, if any, greater than the entering
temperature of the gas. The loss in this case is that
due to the poorer transfer of heat caused by the lower
flame temperature, which is the result of the dilution
of the products of combustion by the water vapour.
This loss is one of boiler output, not of efficiency,
so long as the amount of water vapour present is
not sufficient to prevent complete combustion of the
gas; and given ample boiler heating surface, the loss
may be neglected. As an example, the blast furnace gas
of 100 B.Th.U. per cu. ft. (at normal temperature and
pressure), supplied to the boiler at a temperature of
250 degs. Cent., carries 9 B.Th.U. per cu. ft. of sen-
sible heat, or, in other words, the calorific value is
109 B.Th.U. per cu. ft. as supplied. Suppose this
gas to be burned in a boiler and passed to the chimney
at a temperature of 250 degs. Cent. Then if there
are 0’125 lb. of water vapour present per cu. ft. of
the gas supplied, and the gas is burned completely
with 20 per cent. C.O.2 in the dry products of com-
bustion, the flame temperature would theoretically be
640 degs. Cent., compared with 1,110 degs. Cent,
if the same gas had been cooled to atmospheric
temperature and burned with the same excess of air.
The flame temperature is so near the ignition tempera-
ture of CO (about 600 degs. Cent.) that great diffi-
culty would be found in burning the gas even with
such a small excess of air; and as it could not be
burned at all with much more air than this, it may be
necessary, if there is a large quantity of moisture in
the cleaned gas, to sacrifice the sensible heat and even
cleaned gas, to sacrifice the sensible heat and even
as much as 10 per cent, of the calorific value, for
the sake of getting combustion at a reasonable tem-
perature and an adequate boiler output.

The losses due to the admission to the boiler of any
air in excess of the minimum amount required for
the combustion of the gas are the most serious, the
useless excess of air reducing the flame temperature
and interfering with the transfer of heat, besides
adding to the volume of the flue gases, thus carrying
to the chimney much heat which should have been
usefully employed in the chimney. For several reasons
the scientific method to follow in burning gas is to
use an arrasgement or burner which gives complete
control over the conditions of combustion.

The ordinary efficiency of gas-fired boilers, under
present conditions, is from 50 per cent, to 60 per cent.,
and many boilers fired with blast furnace gas work
at an efficiency as low as 30 per cent. Figures have
been obtained by careful testing, which show the
following results: 83 per cent, efficiency for a water
tube boiler with superheater but without economiser,
fired with coke oven gas; 80 per cent, for a Lancashire
type boiler with both superheater and economiser,
fired with cleaned blast furnace gas; 79 per cent, for
a Lancashire boiler with both superheater and econo-
miser, fired by producer gas; 65 per cent, to 67 per
cent, for many tests on water tube boilers without
superheaters or economisers, fired with uncleaned
blast furnace gas. In the case of a small ironworks,
where 100,000 lb. of steam are raised per hour by
gas, the cost of the steam, if it had been raised by
coal, would be about £40,000 per annum. A saving
of 25 per cent, in efficiency means a saving of £10,000
per annum.

The essentials for the economical combustion of
gas are that the gas should burn immediately and
completely at the highest flame temperature and with
the smallest excess of air. America and the Conti-
nental countries have proved that the best combustion
takes place when gas and air are intimately mixed
before combustion begins. For this reason one or
other variation of . the Bunsen • burner is almost
universally fitted in modern foreign works, but there
is no such installation in this country so far, and
indeed, the pre-heating of air for the combustion of
gas has hardly been considered here.

Gas can only be burned with a minimum excess of
air in two ways: either in a long brick-lined chamber
like a Dutch oven, or by the intimate mixture of the
gas and air before combustion starts. In the latter
case the ordinary combustion space allowed for coal
firing is sufficient for gas. With the usual method,
when the air supply is cut down to show a small excess
of air in the waste gases, unburned CO immediately
appears, and the same thing happens .when one tries
to increase the output of the boilers by admitting more
gas. Using modern burners with gas or air pressure,
far more gas can be burned in the boiler than is
possible by the older method, while at the same time
the efficiency is raised. If high boiler outputs are
required a high pressure of air or gas will give a
short intense flame and the boiler output will only
be limited by the volume of waste gases with which
the boiler flues can deal.

To attain the intimate mixture of air and gas, either
the gas or the air must be under a pressure of not
less than 2 in. W.G., or induced draught may be
used It is preferable, unless the boiler brickwork
is enclosed in sheet steel, to use pressure at the
burners; this prevents the serious leakage of air into
the boiler through the brickwork. The power re-
quired to supply this pressure is only about 1 per cent,
of the power generated.

The flames formed by the combustion of the indus-
trial gases, when burned with a correct amount of air,
are generally blue and colourless, and consequently
do not give out much heat by radiation. The problem
of taking the heat out of gas flames is very like that
of best utilising the waste heat from coke ovens or
steel furnaces.

The author has made a number of experiments upon
a Lancashire boiler, fired with cleaned gas, in order
to discover whether a brickwork arrangement in the
flues can be designed, which, while not seriously
diminishing the chimney draught, would cause much
of this heat to be transferred as radiant heat. The
conclusion come to was that it was not possible, with
natural draught, to heat the brickwork to such a
degree that the transmission of heat was measurably

improved. The result was practically the same in the
case of brick battles built in the Hues, though such
battles are very useful where it is necessary to mix air
and gas during combustion when working on chimney
draught alone.

As regards the transfer of heat from hot gases to
the water in the boiler, most of the heat must be
transferred by actual contact between the gas and
the tubes. By far the greatest part of the resistance
to this transfer of heat occurs at the surface where
the gases touch the tubes, because a very thin film
of cooled gas sticks to the boiler tubes and prevents
the contact of hot gases. To get over this, either the
surface of tne tubes in contact with the gases may
be greatly increased in relation to the wetted surface
of the tuoes, as is done by the use of corrugated flues
or ribbed tubes, or by using small tube water tube
boilers like the Yarrow boiler; or the gases may be
caused mechanically to scour the surface of the tubes,
sweeping away this layer of cooled gas, and making
close contact all the time between the hot gases and
the tubes. This can be done by applying considerable
power to the bases in the way of forced draught or
suction, as is done in the Bonecourt boiler.

Similar results, with a much more reasonable
temperature gradient through the length of the fire
tubes, could be attained by the use of suitably de-
signed boilers with much smaller suctions than 16 in.
"W.G. The Lancashire boilers made thirty years ago,
with a tube plate about 20 ft. from the front and
small fire tubes running from it to the back of the
boiler, would make excellent boilers for gas firing if
used with induced draught and treated as internally
fired only, giving up the outside heating surface.
Only the smallest suctions cam be applied to boilers
with brick settings, owing to air leaking in through
the walls. If, however, the gas were caused to burn
completely in the wide flues, and the products were
drawn rapidly through the small fire tubes, an
adequate duty from the boiler, at a high efficiency,
would be obtained, and this would cut out all the
loss by air leakage and much of the loss by radiation,
which are so serious in the Lancashire boiler as usually
fired. The marine type of Scottish boiler, with two
or more flues, would be even more suitable for this
treatment, and would beat the best water tube boiler
in efficiency.

The cost of the power used to produce the forced
draught, or induced draught, must be debited against
the boiler before the net efficiency is brought out.
To give a suction of 16 in. W.G. on the boiler, at
plants where 30 lb. of steam are required per horse-
power, in the engine driving the fan, costs 10 per cent,
of the steam raised in the boiler: it costs 4 per cent,
of the steam where 12 lb. of steam per horse-power
are required, as at modern power stations. A 4 in.
W.G. pressure or suction should be enough for the
boilers just described, and this would only cost a
fraction of these amounts, * while the added efficiency
of the boilers would far more than pay for the steam
used. It might, however, be cheaper to instal addi-
tional boilers, and to run them at natural draught,
than to run the boilers with mechanical draught at
higher outputs. The improvement in efficiency attain-
able by the latter method is the factor which will
decide the question.

Pre-heating the air for combustion is even more
important when using gas than when using coal fuel.
At many works no economisers are fitted on the boilers
because the feed water is already heated by the
exhaust steam from the engines. It is quite usual
to find a temperature of 350 degs. Cent., and often
more, in the waste gases from gas-fired boilers if
they are working at a fair output. This could easily
be reduced to about 200 degs. C. by the use of a
good air heater. Pre-heated air also promotes much
better combustion and a higher flame temperature,
and thus improves the heat exchange. Modern air
pre-heaters can be used even with uncleaned blast
furnace gas, as the plates can be brushed or blown
clean at frequent intervals while the heaters are at
work, with a very small amount of labour.

In some cases it has been stated that gas firing
with Mond gas was responsible for damage to the
boiler tubes, and an examination has shown that this
damage was caused by the combustion of the gas,
probably by the improper proportions of gas and air
present at any points when combustion in layers took
place. At one instant there would be a strongly
oxidising flame, and at another instant a strongly
reducing flame. On the other hand, boilers fired for
many years with this gas show no damage at all. If
gas is burned with a small excess of air, so that there is
a minimum of oxygen in the burned gases, and com-
bustion takes place quickly so that it is practically
complete before these gases touch the boiler tubes,
no damage whatever can happen to the tubes. In
firing water tube boilers it is best to arrange the gas
burner, pointing towards the floor, at an angle of,
say, 20 degs. below the horizontal, so that the gases
may be completely burned before rising to the tubes.
This also keeps the brickwork of the boiler, upon
which the flames play, very hot, and assists in the
rapid combustion of the gas. With uncleaned blast
furnace gas, the flame temperature is so high by this
method that much of the dust can be melted, and thus
thrown out of the products of combustion, instead of
being burned on to the tubes.

Readings made in America of the flame tempera-
tures in water tube boilers fired in this way show
temperatures ranging from 1,040 degs. Cent, to 1,260
degs Cent., the average being fully 1,160 degs. Cent.
The CO2 reading in the gas averaged 23 per cent.
The flame temperature would be about 1,300 degs.
Cent.

Similar tests were made on other boilers in to which
the gas and air were admitted by ports in the front
of a combustion chamber. This chamber was large
enough for complete combustion of the gas, so that
here also the CO2 readings averaged about 23 per
cent. The temperatures in the combustion chamber
varied from 760 degs. Cent, to 960 degs. Cent., with
average about 900 degs. Cent., while, of course, the