THE COLLIERY GUARDIAN

AND

JOURNAL OF THE COAL AND IRON TRADES.

Vol. CXVI.

FRIDAY, OCTOBER 18, 1918.

No. 3016*1

Primary Considerations in Hydraulic Stowing.*

By C. A. JOHN HENDRY, F.R.G.S., A.M.LM.E.

The conditions prevailing at various mines render
it an advantage to deal with certain factors, to be
taken into consideration before drawing up a definite
scheme, and these notes are only a very rough sugges-
tion of lines on which certain preliminaries should be
dealt with.

Looking at the matter first from a purely hydraulic
standpoint, the following points are quite elementary :
(a) The loss in friction is proportional to the length
of the pipe; (b) it varies nearly as the square of the
velocity; (c) it varies inversely as the diameter of the
pipe; (d) it increases with the roughness of the pipe;
(e) it is independent of the pressure in the pipe.

It may be mentioned that the question of intro-
ducing some lubricating element equivalent to the
nodules of clay mentioned in the above report is pos-
sibly worthy of close consideration in its effect of
reducing wear.

The value of coefficient of friction naturally varies
according to velocity and diameter of pipe, and this
may, for smooth cast iron or wrought iron pipes, be
taken as follows when considering the flow of water
only; the retarding effect caused by the introduction
of sand is dealt with further on.

If a 6 in. pipe is assumed and a velocity of flow
10 ft. per second, the coefficient for this would be
0-022, and with a pipe 3,000 ft. long the loss of head
due to friction is 205 ft. In the case of a compara-
tively short pipe it may be necessary to consider
losses due to the sudden enlargement or contraction

so that the velocity (and therefore the rate of supply)
can be determined, or the maximum length to a
certain head be derived and the cost of an installation
and its capabilities worked out with some exactness
prior to the commencement of the work.

Certain experiments carried out by the writer may
be of interest; two only are quoted, but they may
suggest the lines on which more accurate results may
be derived.

With a 4 in. pipe, head 13 93, length of 167-60 ft.,the
velocity of water only equalled 7-23 per second. With
one bend the coefficient of friction was 0-0316. Upon
introducing sand in the proportion of 1 to 13, the
velocity decreased to 5-1 ft. per second, giving the
coefficient 0-0661. Further, with a 4 in. pipe, head
10-32 ft., one bend at entrance, length 84 ft., the
velocity of the water was 9-8 ft. per .second, and the
coefficient 0-0227. Introducing sand in the propor-
tion 1 to 7-68, the . velocity decreased to 7 ft. per
second, the coefficient being 0’0488. These rough
experiments show that the ratio of sand to water
varied in a direct proportion to the head and length,
while the frictional coefficient was doubled when sand
w;is introduced to the maximum carrying capacity of
the water.

The results, however, would need to be checked with
pipes of a greater head and length, as the effect of
the bend at entrance will be less evident and the flow
steadier. Roughly, where the head to length is 1 to 5,
the sand to water will be about 1 to 3, or where head

result in proportionately heavy pumping (to remow
water after it had been used for stowing), an*^ jt
would remain to be determined which would be r^ofe
economical.

Having determined upon the output and velocity, it
will be a simple matter to gauge the head required
and to work out the most convenient feeding point
at the surface and to judge the correctness of the
dimensions adopted. Thus where the head is incon-
siderable it would possibly be an advantage to put
down two or more boreholes" or to supplement the
head by the introduction of a pumping unit.

The disposal of the flush water and its clarification
are important. The horse-power necessary to elevate
the waste water back to the surface would in this
preliminary investigation be:

W x H

HP—	v-- where W = weight of water elevated

OO,VVv

per minute, and H — height in feet.

Assume, ignoring water friction, the water equals
90 cu. ft. 90 x 62 = 5,580 lb; at a height of 207-2 ft.
Therefore

HP - - 5,682,x 207- = 35 HP.

33,000

In certain mines where the seam is steep enough, it
may be cheaper to flush the sand to the workings
through a flume or trough. This can be done when
the angle is is 15 degs., or even less if the ratio of sand
to water is high.

It may be necessary to bring the flushing water
from a distance, especially as the sand would be more
readily available in the dry seasons. If the distance
is considerable and the grade favourable, the open
flump may be more economical than a pipe line.

Figs. 1 and 2.—Method of Raising Sand from River.

Feed Conveyoi

Rxver "bcid



So -taar 'wiMpons. ft*' Tarei
fieojZ te>-ra. p>ot>\Tiorx. "fo I

Viaulapp.	j

deMfQ'Of biQ. •How j
po&SavbiV-iy op	J

Fig. 3.—Arrangement of Supplementary Conveyor.



of pipes, also elbows or bends, but where, as is usually
the case, the length is greater than 1,000 times the
diameter, the velocity head and loss of head at
entrance will be so small, in comparison to frictional
losses, as to be negligible. If a head of 250 ft., pipe 6 in.
diameter, and length 3,000 ft., be considered, the first
requirement will be velocity, and this in the case under
review will be 11 ft.

Generally speaking, we may, in long pipes, ignore
losses due to’ entrance bends variation in pipe sec-
tions, but usually the sum of all this may be taken
at 0’5. To tfijs may be added the coefficient due to
friction at bends required to take the pipe to its goaf
position.

Let us take a case as before, pipe 6 in. diameter,
length 3,000 ft., velocity 10 ft. per second. For these
conditions a head of 207-2 is necessary, assuming a
perfectly straight pipe. If we introduce 12 bends
each set at an angle of 80 degs. with a radius of 12
in., the head is found by calculation to be 208-78, and
thus the loss in head due to the introduction of 12
bends is only 208-78 — 207-2 = 1-58, so that in long
pipes the effect due to bends is of small importance as
compared to other frictional losses.

It is, however, with these other frictional losses
that we are chiefly concerned, and if sand packing is
to be carried out on definite lines it will be neces-
sary to find out the relation of head to length of pipe,
the ratio of sand to water, and the velocity to the size
of the pipe. All these things will have a defined rela-
tion one to the other, and if the system is to be
carried out on a large scale, then it is desirable to
collect information and experiment, so that a basis
may be established for common use.

While it may be an exhaustive matter to derive
suitable coefficients for all conditions, there is no
reason why the behaviour of certain mixtures, such as
one part of sand to three of water, should not have
certain coefficients of friction worked but for them,

* From paper read before the Geological and Mining
Society of India.

to length is 1 to 3, the sand to water will be about
1 to 1-5, but the capacity of installation will depend
upon the velocity and size of the pipes, and it would
be an advantage if experiments on the lines suggested
above were carried out where the system was in use
so as to determine the proportionate results with
some accuracy.

To determine the size of the pipe it will be necessary
to decide upon the rate at which stowing is to be
done. When the diameter of the pipe is assumed in
feet we get the discharge in cubic feet = -7854D2V,
equivalent in the above case, with a velocity'of 10 ft.
per second and a 6 in. pipe, to 1-963 cu. ft. per second.

The velocity of discharge will bear a definite rela-
tion to the diameter of the pipe and its length.
Hence the first.problem will be. the exact economical
ratio of the head to the underground lead, whether it
would be more economical to put down a series of
boreholes direct to the workings or to establish only
one or two feeding points at the surface with long
underleads.

This will depend on: (1) Rate of stowage required;
(2) nature of strata for boring; (3) quantity of water
available; (4) grade of slope to goaf; (5) velocity
of mixture.

It should be remembered there will be a limit to
the minimum velocity of the mixture, otherwise the
sand held in suspension will settle and gradually in-
crease frictional losses until movement practically
ceases. In like manner high velocities will result in
abnormal and costly wear.

Pipes being generally made in even dimensions,
should calculations call for a fractional size, it would
be better to adopt a larger size in even inches; should
riveted pipes be used, a considerable allowance might
be made for the reduction of carrying capacity due
to rivet heads.	t

The relative specific gravity of water and sand
would be 1-0:1-7. Thus the wear due to friction on
the pipe would be less the more water used. It
would, however, have to be considered that this would

To ascertain the dimensions of a flume to carry a
given quantity of water involves the solution of an
equation of the sixth degree, but assuming that 2,500
gallons of water are required per minute (roughly
7 cu. ft. per second) to be furnished by a flume,
rectangular in shape and having a depth equal to
about half the width, with a grade equal to, say, 1 ft. ,
per mile, the depth required will be about 1-61 ft.,
and this will give a width of 3-22 ft.

Regarding the pipes to be employed, we may con-
sider : (a) Cast iron;; (b) wrought iron; (c) wood;
(d) terra cotta; (e) porcelain-lined. The shape woul^l,
in comparatively small installations, be circular, but
ovoid pipes may be considered in special cases.

It will prove economical to employ thicker cast iron
pipes for this purpose than is usual for water only,
and in no case is it advisable to employ them under
| in. thick, as the tensile strength is low and uncertain.

If cast iron pipes for economic reasons must be
employed, it is advisable to have them as thick as
possible; for instance, assume outer diameter 7 in.,
inner diameter 6 in., then area of effective section
= 38-484 — 28-274 = 10-210. If we increase the outer
diameter to 7-5 in., the extra metal over the above
would be 44-178 - 38-484 = 5-694 sq. in. section. Thus
for scarcely more than half as much more metal the
life of the thicker pipe will be double as long as in
the first case, provided we assume the pipe will
possibly give trouble and have to be discarded when
its thickness gets below, say, | in.	,

As a basis in gauging the thickness of a pipe for
water, assume—

H = head in feet.

P = pressure per sq. in.

W = weight of a column of liquid! ft. long, 1 sq. in.
in cross section.

Then P = W x H.

For water we may take W = -434.

Therefore P = -434 H.

It is usual, in figuring accurately the size of the pipes,
to make an allowance for what is commonly known