September 1, 1916.

THE COLLIERY GUARDIAN.

407

CURRENT SCIENCE

Rapid Gas Analysis.

Mr. 0. A. Krone (Journal of Industrial and Engineer-
ing Chemistry) describes a method by which a complete
analysis of a gas can be made in 20 minutes. The
apparatus comprises two 100 c.c. water jacketed burettes
graduated the entire length in 0-1 c.c., connected with
a brass tube containing a platinum combustion tube of
1'5 mm. outside and 0'5 mm. inside diameter, and about
10 to 11 cm. long, bent to form a U, the two legs of
which run parallel about 1 cm. apart. Around the brass
body is soldered a piece of sheet copper to form a cup
for water. The platinum tube is then soldered in, gas-
tight. A Meker high temperature burner is used to
heat the tube.

The reagents used are : A 1:1 solution of purified
stick potash, 150 c.c. of which are poured into a stop-
pered single Hempel pipette filled with clean iron turn-
ings to increase the absorbing surface of the potash;
Nordhausen sulphuric acid, or saturated bromide water;
yellow phosphorus (5 mm. sticks) in a stoppered single
Hempel pipette covered by 150 c.c. of pure distilled
water; a stoppered double Hempel pipette filled com-
pletely with straight T5 mm. copper wires and covered
by 125 c.c. of 1 : 4 solution of ammonium hydroxide. A
large excess of ammonium chloride is then added to
this solution, and a saturated solution is made at 'about
80 degs. Fahr., the excess of ammonium chloride being
allowed ‘to settle and the clear solution decanted into a
clean flask, stoppered, and kept for use.

Of the sample of gas to be analysed, 100 c.c. are
drawn into the water burette. The capillary is then
washed, and to seal the gas in the burette acidified water
is allowed to completely fill all the capillaries and trap.
The potash pipette is then connected up, and the gas in
the burette aspirated back and forth three times to
remove all the carbon dioxide. The residue is measured
and the reading recorded. The difference between
100 c.c. and this reading gives the percentage of carbon
dioxide present in the sample.

The potash pipette is disconnected, the capillary
washed, and the gas in the burette put under slight
pressure. The cock is then closed, and the fuming
sulphuric acid is introduced. The gas is aspirated back
and forth twice. Water from a wash bottle is next
blown into the rubber connection of the potash pipette,
and this pipette is quickly connected up. To take out
the sulphuric fumes, the gas is passed back and forth
four times. The reading is taken and recorded. The
difference between this and the previous reading gives
the percentage of illuminants (unsaturated hydro-
carbons, benzene vapour, etc.). If all the illuminants
are not withdrawn by two passages through the
sulphuric pipette, they should be removed by passing
through once more. The oxygen is next removed with
phosphorus, the gas remaining in contact with the
moist phosphorus until most of the fumes disappear.
The gas is then drawn into the burette, measured, and
the amount recorded. The difference between this and
the previous reading gives the percentage of oxygen in
the mixture.

The residual of gas is transferred to a holder, a
measured excess of oxygen being then added. After
measuring, the oxygen is transferred to the holder, and
the entire gaseous mixture shaken. After thoroughly
mixing the gas and oxygen, 100 c.c. of the mixture or a
close approximation to 100 c.c. are drawn into the water
burette, and the amount accurately measured. The
mercury burette is then adjusted so as to contain 15 or
20 c.c. of air. For the combustion of the gas, the
platinum tube is heated to a bright red, and
the gas passed through slowly, the residual of
gas, after combustion, being measured and the
amount recorded. The potash pipette is next
connected up, and the gas aspirated back and
forth four times, the gas left being measured and
recorded and transferred to the copper pipette to absorb
the oxygen completely. After drawing the residue of
gas into the burette, it is measured, then passed into
a dilute solution of sulphuric acid to remove any
ammonia, and is then re-measured and the amount
recorded. The formula used for calculating the com-
bustion results is as follows. Representing the oxygen
used in the chemical combination by O2, the contrac-
tion by C, we have :—

h2 = c - O2

CJJJ __ 3 O2 — (C + CO2 formed)

3

CO = CO2 formed minus CH4.

The nitrogen is found by difference, or can be calcu-
lated from the residual nitrogen. Both methods give
the same results.

Waste.

In dealing with the question of waste in power pro-
duction, Mr. R. O. Wynne-Roberts (Canadian Engineer)
mentions that, in Canada, as much coal is wasted as is
extracted, owing to the methods adopted, and that this
waste amounts to very many tens of millions of tons,
and advocates that, in addition to taking measures for
checking wasteful methods of mining coal, the Govern-
ment should carry on investigations with a view to deter-
mining the suitability of slack and low’-grade coals for
use in gas producers for generating power, and their
adaptability for the manufacture of briquettes for
domestic use. By utilising these inferior products in
this way, not only would there be less waste, but the
value of the public coal lands would be considerably
increased.

The lignite coal fields of the West are mined to a very
insignificant extent. There are about 20,000 million tons
of lignite lying in the prairie provinces which are not
wasted, but are not adequately utilised. This coal is

AND TECHNOLOGY.

most suitable for gas producers, and some day huge
central power plants will be located at the mine mouths
to generate electrical energy for transmission to many
parts of the provinces. Meanwhile, fuel is transported
from the United States and from distant west Canadian
coal mines at considerable cost, constituting another
form of waste. Whilst many Canadian works have rela-
tively efficient power plants, the majority cannot be
placed in this category. Consequently, central plants
would be productive of great savings.

Another cause of waste is the method of producing
metallurgical coke. There are nearly 3,000 coke ovens
in Canada, but only 730 of them are constructed to save
the valuable by-products, such as gas, tar, ammonia,
etc. The gas from coke ovens should be available for
power and heating, and if used in gas engines there is
sufficient surplus gas from every ton of coal to provide
250 horse-power hours. The quantity of ammonia will
suffice to produce about 201b. of ammonium sulphate,
which is valuable for fertilising the land. The ammonia
may be reduced to concentrated crude ammonia for
refrigerating purposes, or utilised for cleansing. These
are only a few uses to which the ammoniacal liquid may
be applied. The tar is one of the very valuable complex
by-products of coal, which can be distilled for several
hundred uses. Creosote for preserving wood, consti-
tuents for high explosives and dyes are the pressing
needs of the hour, but in 2,000 ovens in Canada the tar
is not recovered, and the Empire has to buy elsewhere.

Flow of Viscous Liquids Through Pipes.

In experiments on the influence of viscosity on the
flow of oil through pipes, carried on by Messrs.
Greenough and Dinsmore, two horizontal lines were
built, one of lin. pipe, 136 ft. long, and the other of
| in. pipe, 63-8 ft. long. In other experiments by
Messrs. Haylett and Lucy, the oil was pumped from a
tank of 400 gals, capacity into an iron drum and then
through four lines, one of 2 in., one of lin., and two of
| in. pipes. Three of the pipes were of wrought iron,
the fourth (Jin.) pipe was of steel; a certain (not
specified) length of the pipes was straight, and the oil
was returned to the starting-point and delivered into
tanks on platform scales. In the second series, in par-
ticular. the rates of flow were observed at considerable
velocities ranging up to 20 ft. per second. Under these
conditions the flow may be turbulent, and the practical
point of the research is that a simple method is deduced
by means of which it can easily be calculated what rate
of flow may be expected in a pipe line of given dimen-
sions under a certain head.

Poiseuille’s formula,it is well known, holds for stream-
line or parallel flow (called also viscous flow by Prof.
Lewis) through capillary tubes, up to 5 mm. in diameter,
as long as the velocity remains below the critical velocity
at which the flow becomes turbulent (or sinuous).
Poiseuille’s formula is p = Sylv/gr2, where p is the
pressure drop in lb. per sq.ft., y the coefficient of
viscosity (in lb. per second per sq. ft.), I the length and
r the radius of the pipe (in ft.), v (or Vw) the mean
velocity of the fluid in ft. per second. For velocities
above the critical speed, various formulae have been
proposed on the strength of experiments with water,
steam, and air; they are all of the form p =
/ slv2/gr ; p is the density of the liquid, and f the
hydraulic coefficient, and this formula is adopted by
Prof. Lewis. The p is a function of I, v, and r, and the
function is influenced by the /, which depends upon the
viscosity. Viscosity measurements were made by the
Gurney viscosimeter, which is a vertical tube (jacketed
with steam or water) through which the liquid is drawn
by a pump; the formula of Gurney was not adopted,
however. The value found for f w’as : f = fw (0-955 +
0-045 z), where fw was the hydraulic coefficient for the
flow of water through a pipe of the respective dimen-
sions (to be looked up in tables) and z the relative
viscosity of the liquid, i.e., the viscosity referred to
water. The z is measured in the usual way, by deter-
mining the relative times of efflux of the liquid and of
water through the same Gurney capillary. This formula
for f shows that for liquids of low viscosity, the carrying
capacity of a pipe will practically be the same as that
for water: for, when z is approximately 1, f becomes
nqual to fw . With higher viscosities the carrying capa-
city rapidly decreases, and the formula, Prof. Lewis
points out, would not hold for liquids (like glycerin and
heavy oils) whose viscosities are more than 20 times as
great as that of water. The relative viscosity z multi-
plied by 0-0671 (which is the viscosity of water at
20 degs. Cent.) gives the absolute viscosity y in pounds
per second per sq. ft.

The practical outcome of the research is the following.
We have two formulae for p, one for stream-line flow,
one for turbulent flow: we thus find two values for the
velocity v. The smaller of these two values, or v2,
will give the answer to the problem. An example will
make this clear. Supposing we wish to find the velocity
of an oil of density p = 0-91 flowing through the length
I = 600 ft. of standard pipe (1-07 in. internal diameter
— 2r) under a head of 30 ft. = h; at 20 degs. Cent, the
oil gave a time of efflux of 108 seconds, water giving
4-90 seconds: then : p = h p = 30 x 0-91 X 62'3 =
1,7011b. per sq.ft.; 2 = 108/49 = 22-05: y = 0*0671:
s = 1'478; r = 1*07/12 X 2 = 0*0446; (from hydrau-
lic tables) 0'0075; Vj, (for stream-line flow) = pgr2/Sy\
I ~ 0*0153 ft. per second; v2 (for turbulent flow) =

V g r/pl p = 2-22 ft. per second. Thus the stream-
line flow requires only a fraction of the pressure needed
for turbulent flowy and vx =0*0153 ft. will be the
answer. Experimentally, the formulae have been sub-
stantiated only for pipes up to 2 in. in diameter and for
relative viscosities of 20. They may roughly be appli-

cable for larger pipes and more viscous liquids. If a
pipe is carrying liquid in viscous motion, increase in size
is always well worth consideration; for with viscous
motion doubling the pipe radius increases the velocity
fourfold, and the discharge 16-fold for the same drop of
pressure.—Engineering-.

Distribution of Pressure by Granular Materials.

Very little is definitely known as to the wray in which
the pressure of a wheel on road metal or a sleeper on its
ballast is transmitted to the underlying foundation. It
is generally held that the intensity of the pressure due
to an isolated load.diminishes as the depth increases, and
the assumption is sometimes made that at any depth the
pressure is uniform, and confined to the base of a cone
having the load at its apex and with its surface inclined
at an angle of 45 degs. to the line of the load. Some
definite information on this head is provided by some
experiments recently carried out in the engineering labo-
ratories of the University of Illinois, a description
of wdiich is contributed by Prof. M. E. Enger to a recent,
issue of the Railway Review. The experiments -were
made by applying a local load to a bed of sand supported
on a concrete floor. An opening in the floor 4jin. in
diameter w7as closed by a plug 4 in. in diameter, which
rested by a knife-edge on a lever, the outer end of w-hicb
was borne by the platform of a weighing machine. The
load wzas applied to the upper surface of the sand by
means of a circular plate, the diameter of wdiich ranged
in different experiments from 9 in. up to 21 in. Tn the
first instance the load was applied centrally over the plug,
and the weight borne by the latter was recorded. The
pressure-plate was then shifted a little to one side, and
the observation repeated. By proceeding in this wav
sufficient data were obtained to make it possible to plot a
curve showing how the pressure on the sand varied with
the distance from the line of load. The sand wras
thoroughly compacted before taking a reading. If, for
example, it was intended to determine the pressure distri-
bution at a depth of 12 in., the plate wTas started on a
layer about 15 in. deep and forced dowm to 12 in. depth.
The load was then released and re-applied several times
in succession, readings of the pressure borne by the plug
being taken during each of these re-applications of the
load. The interesting fact w7as established that at
shallow depths the intensity of the load on the plug
might be very considerably greater than the average
pressure applied at the surface of the sand. Thus in one
series of experiments in wKich the pressures wmre
measured by a diaphragm gauge 6 in. in diameter,
instead of by means of the 4 in. plug aforementioned, the
following figures were observed :—

Depth below surface.	Diameter

f----------------A-------------> of load-

6 in. 12 in. 18 in.	plate.

Per cent.	Per cent.	Per cent. Tn.

72	...	43	...	19	...	9'0

160	...	90	...	40	...	13'5

—	...	195	...	92	21'0

The figures given show the average pressure per square
inch on the 6 in. gauge, as compared with the average
pressure per square inch on the load-plate when this is
concentric with the gauge. Obviously, since the
pressures given are averages, the maximum pressure
along the centre line must be considerably higher. The
diaphragm gauge wras found to be unsuitable for “ out-
of-centre ” observations; but the results obtained with
the wooden plug made it possible to plot curves showing
the distribution of pressure. These were found to be very
similar to the curves of pressure distribution that are
found when a jet of water strikes a flat plate normallv.
One result which appears to be fairly deducible from the
experiments is that the distributing powder of sand is not
as great as wrould be anticipated from the 45 degs. rule.

MIRING AND OTHER NOTES.

Mr. V. K. Ting, Director of the Geological Survey of
China, in an article in the Far Eastern Review, estimates
the coal resources of that empire to be next in importance to
those of the United States and of Canada. Any complete
estimate is largely a matter of conjecture, as is shown by the
fact that 996,612.700.000 tons were assumed by Mr. Drake,
and 39.973.000.000 tons by Mr. Tnouye, in the International
Geological Report on the coal resources of the worM. The
production of China in 1913 amounted onlv to 14.515.000 tons.
The writer points out that with the development of rail-
wavs the consumption of coal throughout China must extra-
ordinarily develop.

Another step in the exploitation of Spitzbergen coal is
reported from Christiania, where a joint stock company was
recently formed, with a capital of between 3 and 5} million
kroner. The promoters have bought for 3 million kroner the
property of the Advent Bav Coa1 Field Company on Spitz-
bergen, which is estimated to contain 400 million tons of
coal, and also the coal fields of the Svalbard Company at
Green Harbour, which are estimated to coniain 200 million,
tons of coal. The yearly production of both coal fields is
estimated at 200.000 tons of coal.

We have referred in previous issues to the extensions
which are being carried out at Port Talbot by Messrs.
Baldwins Limited, and we are now informed that the con-
tract for the coal washing and coking installation has been
placed with the Coppee Company (Great Britain) Limited,
of King’s House, Kingsway. London. W.C. The plant-
consists of a Coppee coal washery to wash 120 tons per
hour and 120 Coppee regenerative coke ovens with
by-product plant for the recovery of tar. ammonia, and
benzol and the manufacture of sulphate of ammonia and
the rectification of the benzol. A 74)0.000 cu ft. gasholder
is also being supplied. The Coppee Company have also
lately commenced w’ork on the foundations of the new
W’ashery which they are installing at the Gedling Colliery
of the Digby Colliery Company Limited. Nottingham. This
plant is to be able to treat 112 tons of coal per hour from
3 in. to nil. The plant will be the most modern and up-to-
date in the district.