986

- THE COLLIERY GUARDIAN.	May 25> iyiZ

MM	r'"' ^CTgnSMy MIMI IIIWIIIIIIIIIMIIMIIMII1H1IIMMMHIM	l«	I I—B

Any suitable apparatus may be used, such as steel
plate fans, turbine- or engine-driven, or multi-blade
fans similarly operated. The fuel is fed to a hopper
extending across the front end of the stoker above the
grate, from which it is deposited on the grate, the
thickness of the fuel on the grate being regulated by
the adjustable coal gate. Ignition takes place imme-
diately, and combustion is supported by forced draught
under the grates. There are three, or sometimes four,
air compartments, each extending crosswise of the fur-
nace; in each of these the air pressure may be inde-
pendently regulated. It is thus possible to vary the
rate of combustion over each air compartment as
desired. At the rear end of the grate, where it turns
over the sprocket wheels, the ash is discharged into the
ash pit.

The three compartments may, for convenience, be
called the ignition, combustion, and burning out com-
partments. Ignition takes place above the front com-
partment, the bulk of the combustion over the second,
while the third compartment serves for burning out
the remaining carbon at ordinary ratings. The func-
tion of these compartments is to enable the operator
to vary the pressure under the grate in accordance with
the thickness of fuel immediately above that section.
The rear end compartment, or burning out area, is
subdivided laterally into three distinct compartments,
in each of which the air pressure may be separately
controlled. Thus, if there is a tendency for the fuel
to burn out on the sides, at the rear end, the side com-
partments may be closed and the air pressure retained
in the centre. This grate is sold by George J. Hagan
Company, Pittsburg, Pennsylvania.

Under regular operating conditions, in connection
with a 500 horse-power boiler, one of these stokers is
stated to develop from 150 to 200 per cent, of boiler
rating, with efficiencies ranging between 65 and 70 per
cent. The ratio of water heating surface to grate sur-
face in this installation is 42 to 1. The breeze had the
following composition:—Moisture 8 to 12 per cent.;
volatile matter, 2 to 4 per cent. ; fixed carbon, 76 to
80 per cent. ; ash, 18 to 20 per cent. ; calorific power,
11,500 to 12,000 British thermal units.—Coal Age.

SOUTH WALES COAL AND THE
CALORIFIC STANDARD.

At the general meeting of the Wales and Mon-
mouthshire District Institution of Gas Engineers and
Managers at Merthyr Tydfil, Mr. R. G. Clarry
(Swansea) drew attention to the poorer class of gas
coal obtained in South Wales, as compared with the
Midlands or Northern districts. The new calorific
standard of 500 British thermal units gross would bear
unduly on gas undertakings using a large proportion of
South Wales coal, for while the figure might leave scope
in other parts of the country, in South Wales there
was practically no margin after oil washing. To be
just to gas managers in South Wales, the standard
should be at least 25 British thermal units gross less
than 500. Moreover, the benzol recovered, per ton,
from South Wales coal was much lower than that
obtained from the coalln general use throughout Great
Britain. It seemed possible that the coal requirements
of certain gas undertakings might be cut down, and
that they would then be required to use a proportion of
“ blue ” water gas to make up the reduction. At
Swansea they were experimenting, and he did not
believe they could add more than 5 per cent, of
“blue” water gas, if as much, and maintain the
calorific standard of 500 British thermal units. At
present in South Wales they had to pay at least 2s. 6d.
per ton more for their coal, owing to the addition to
the standard amount in the Price of Coal (Limitation)
Act, but large quantities of bituminous coal would no
doubt be exported from South Wales after the war,

and tend to maintain this inflated price. They would
also have to reckon, in a few years time, with the grow-
ing scarcity of suitable gas coal, even of this poorer
quality, in South Wales. That difficulty might be
overcome by some restriction of the export of bitu-
minous coal from South Wales until the coal required
for home consumption had been provided. He pro-
posed that the attention of the National Gas Council
and of the Institution of Gas Engineers should be
drawn to the poor quality of gas coal obtainable in
South Wales and Monmouthshire, as compared with
the Midland and Northern districts, and, in addition,
to the fact that the gas undertakings of that district
had to pay 2s. 6d. per ton more for their coal under
the Price of Coal (Limitation) Act, with a view to
giving due consideration and allowance to South Wales
in comparing results or dealing with any general
instruction to be issued to the gas industry, more par-
ticularly in reference to the calorific standard and the
use of “ blue ” water gas, as suggested, as a means of
reducing the consumption of coal.

Mr. J. Mogford (Briton Ferry) said if the standard
for South Wales was fixed at 500, the same as for the
rest of the country, they would always be compelled to
sell gas at a higher rate than in other parts of the
country. The price of the local bituminous coal had
always been higher than the Midland coal, although
inferior, owing to the high price Welsh steam coal com-
manded at all times. Few collieries were working
bituminous coal, and in the future they were likely to
get less than in the past; coke ovens were being put
down in connection with almost every bituminous coal
colliery. As to the combination of small works to pur-
chase cargoes of Lancashire or South Staffordshire
coal, he hardly thought that would help from the
national point of view, and tend to relieve the rail-
ways. To his mind, they were confined to South
Wales coal, and from the economical point of view it
was the best in the long run.

Mr. E. Rees (Margam) said the price paid in South
Wales for local coals had always been out of proportion
to their value. The price had a very important bear-
ing on the calorific value. The main question was
what would be the result of a lower standard of 475

British thermal units from the point of view of effici-
ency to the public? As gas suppliers, they had always
to bear in mind that the public demanded the most
efficient service.

The President said he enquired some years ago as to
getting coal from Garston, and to his surprise the cost
loaded at Garston and freight to Penarth Dock was less
than the Taff Vale Railway wanted for bringing it
from Cardiff to Merthyr.

NOTES ON TURBINE PUMP
CONSTRUCTION.*

By A. E L. Chorlton.

The individual impellers of a turbine pump revolver
in chambers or cells containing the outward flow guide
passages and the return conduits for the water.* These
chambers are of two main types: Osborne Reynolds or
divided (“ring”) type, and the Sulzer, integral or
one-piece (“ cylindrical ”) type. Recent American
“ practice provides a variation of this type, with the
housing in halves divided on the horizontal centre
line. An examination of the present day practice of
various makers proves that the “ ring ” type ha& ulti-
mately proved the preferred one.

Before dealing with either type or variations arising
out of them, it is advantageous to consider the essen-
tial functions of the chamber or housing of a turbine
pump. Primarily, each cell consists of: (1) The out-
ward flow guide passages, in which the kinetic energy
of discharge from the turbine impeller is converted
into static head; and (2) the return water passages
back to the centre for conduction to the next impeller.
A complete housing is a collection of such cells.

Obviously, in the design, commercial considerations
must have a material guidance on theoretical claims.
Against the requirements for best theoretical conver-
sion of kinetic energy must be matched the allowable
limits of dimensions conformable with commercial pos-
sibilities ; and, in the interests of efficiency, special
attention must be paid to the arrangement and dimen-
sions of the divergent channels or guide passages,
which must be disposed in some form to lie conveni-
ently in the desired casing. The general character will
be either a simple outward flow type in one plane, or a
mixed type outward and axial in two planes.

The divergent angle of guide vane for best efficiency
has been shown by Prof. Gibson to be 10 to 11 degs.
Often such a small divergent angle leads to a large
over-all diameter guide vane, in order to give a suffi-
ciently reduced speed of water to permit reversing its
radial direction, and the result is a very heavy casing
in consequence. For this reason, divergent angles of
15 degs. are commonly found in practice. Evidently a
more efficient pump will sometimes be heavier and
more expensive than one less efficient, and commercial
considerations must provide the final deciding factor
between efficiency and weight. There is, of course, a
school of design which believes in dealing with a pro-
portion of the velocity conversion in the wheel itself,
thus leaving less to be dealt with in the guide passage;
the extent, however, to which this method can be used
for weight saving is very small, if any.

The various assemblies of passages may be grouped
into : —(n) Tangential and radial with return radial;
(b) tangential and spiral; (c) combinations.

The Osborne Reynolds pump (1887 and 1875 type)
employed an early form of the (b) assembly, and has
adhered to this type up to the present day, various
improvements being embodied from time to time, in
some of which the author was concerned. On the
Continent, Messrs. Sulzer introduced in 1896 design
(a), and the author believes they have made little
departure from the type beyond a considerable simpli-
fication of their early arrangement of passages.

Speaking generally, combination designs (c) are not
so efficient as the simpler types (a) and (5), owing pro-
bably to the hydraulic loss through changing the radial
direction of the water at high speed. From a works
construction point of view, “ring casings” are the
most economical, and in practice give high efficiency.
For “cylindrical” casings, a complete pattern is
required for each size and variation in number of
chambers. Its accessibility, however, and ease of dis-
mantling, is sometimes considered to be greater than
with the “ring” type, though this is, to a certain
extent, a matter of opinion. A great drawback to
casings containing separate cells is that, on account
of the sliding fit between the intermediate pieces and
casing, an unknown amount of leakage constantly
takes place between the cells. With a “ring” type
of pump, leakage is instantly detected, and can be
remedied. As a commercial proposition, the author
favours the divided or “ring” type of casing, and
when properly carried out, has met no difficulty with
it in practice.

Impellers.

Impellers are either (1) single entrant, or (2) double
entrant, the first being almost universally in use for
multi-cellular pumps, and the second almost exclu-
sively for single-chamber pumps. Of the several
types, the unbalanced form, with unequal side area
and rubbing shoulder, may be said to be the one now
generally used, that is, the preferable design.

The internal design of all impellers is governed by
the same controlling features: (1) The entrance or
inlet angle of vane; and (2) the delivery or exit angle
of vane.

The entire design must, while based on these con-
siderations, consult the convenience of the workshop
to the utmost degree possible, without departing from
required dimensional accuracy. The standing diffi-
culty with turbine pumps, from the manufacturing
point of view, is their constant variation to meet the
infinite number of conditions of varying head, speed,
and quantity encountered in practice. Whatever
efforts are made, it seems impossible to keep to a small

* From a paper read before the Institution of Mechanical
Engineers.

number of standard impellers if the highest efficiency
is to be reasonably well reached each time. Efficiency,
and not apparent first cost, is really the prime factor in
design, for the pump is very often driven by an elec-
tric motor of greater value. If, therefore, by the use
of a pump of higher efficiency, a reduction is effected ,
in the necessary power and size of motor, the combi-
nation will generally come out cheaper.

The inlet angle depends on the resultant of the
(supposed) radial flow of the entering water, and the
peripheral speed of the inlet tip itself, and1 it usually
varies between 15 and 30 degs. If we select to use
always 15 degs., we might suffer in some cases to the
extent of 3 or 4 per cent. If we take two sizes, 15 and
25 degs., and make a liberal provision of inlet width
(or “ weir coefficient ”) at entrance, we shall only drop
perhaps 1 per cent, in exceptional cases; so it seems
possible to do something in the way of standardising
the inlet angle.

The discharge angle depends on the resultant of the
peripheral speed of the wheel and the radial outflow,
the variation of which is so great it seems impossible
to devise any standardisation. To meet quantity
variation, two widths of impeller may be used, and to
meet required speeds of revolution it is usual to allow
a small percentage variation in the impeller diameter
io suit special demands.

The surfaces of the impeller, both inside and cut,
play an important part in the efficiency of a turbine
pump. The smoother the outside surface, the less the
power lost in disc friction, and the less the power
wasted in revolving idle “ dead water.” An interest-
ing point to note is that the greater the speed of revo-
lution of the “ dead water ” (that is, the greater the
power absorbed in this way), the less the leakage from
the periphery of the impeller; but a little considera-
tion will show that greater over-all economy is gained
by reducing the wasted power in idle revolution to a
minimum.

Balancing Appliance.

The experienced designer knows, from practical
knowledge, that even if a group of impellers, or even
a single impeller, is theoretically balanced by equiva-
lent areas subject to pressure, it will not in actual prac-
tice be free from all end thrust. Apart from variable
leakage at the two sides of an impeller, which, by the
way, is the principal cause of end thrust, we often have
variable side surface both of the impeller and the cell
chamber. If the conditions of capacity or surface of
the two clearance chambers vary, the resulting pres-
sures will vary, and an axial thrust is set up. The
experiments of Nedden have shown that the speed of
the revolving “ dead ” water at the two impeller sides
sets up pressure in opposition to the leakage pressure
from the impeller tip ; the faster the water revolves,
the greater the resistance to leakage, and the less the
resultant pressure due to that leakage. Hence, the
effect of the chamber at one side of a wheel having*
greater capacity than the other, or having more
obstructions in the way of ribs, pockets, or exceptional
roughness, etc., plays some part in the determination
of end thrust.

All the earlier turbine pumps, employed some form
of thrust bearing, either of the collar or the ball type,
to keep the impellers in correct alignment with the
guide passages, and it was not until much trouble had
been experienced with these that hydraulic control was
ultimately adopted. Experience showed that much
higher end pressures were set up than were ever anti-
cipated, but for a period mechanical devices were per-
sisted in, improvements being made attempting to
withstand the excessive loads, and scant notice, it
would appear, being taken of water turbine practice
where for many years it had been the custom to relieve
axial thrust hydraulically. It should be borne in mind
that the relation between the calculable axial thrust
and the thrust realised in a water turbine is much
closer than in a centrifugal pump of the multistage
type, the reason of this being that the disturbing
factor in a high-lift pump is the leakage from stage to
stage.

The automatic hydraulic end balance devices now in
use are of two basic forms: single acting (simple), and
double acting (differential). The general effect of
wear on single plate balancers is increased leakage, the
outcome of which is further end movement in one direc-
tion, until finally the impeller discharges do not match
their guide vane entrances, but become offset. This
tendency in hydraulic balancers led the author to devise
the double-action control, with which such difficulties
are overcome.

It is usual to make hydraulic balancers with renew-
able faces at the points of close running, so that the
appliance can be readily repaired after wear has taken
place. As regards the best material to be used for the
renewable part, the present practice usually is to fit
hard bronze. The ideal material is one which is hard
and “ short,” and has a very low coefficient of friction
when working in water; a material which “ drags ” is
quite unsuitable, and apparently an incorrodible iron,
one containing a high percentage of silicon, or a nickel
steel, seems to best fit the requirements. Non-metallic
substances, such as red fibre, woodite, or dexine, have
been considered, but any material which becomes
slightly absorbent after long immersion is useless.

Sometimes difficulty is experienced in starting up
large turbine pumps on account of the large diameter
balance rings being held together in metallic contact
instead of being apart in their usual running position;
the result is an abnormally high starting torque for
the motor, and probable damage to the faces of the
balancing device. When a turbine pump with differ-
ential balancer is “ stopped,” the rotor will always take
up a position with the balancing faces apart if the
waste water outlet is taken to an open drain. How-
ever, if the spent water is led back to the suction pipe
of the pump, and a foot valve is provided at the end
of the suction pipe, the whole rotor is forced towards
the suction end of the pump as soon as the foot valve
closes and the head pressure comes on. The reason for
this is that the area of the balancer on the outlet side
is greater than on the inlet side by an amount equal