THE COLLIERY GUARDIAN

AND

JOURNAL OF THE COAL AND IRON TRADES.

Vol. CXV.

FRIDAY, APRIL 12, 1918.

No. 2989.

Notes on the Maintenance of Turbo Alternators.

By L. FOKES.

The introduction of steam turbo plant for the
generation of electrical power brought about a
revolution in the design of electrical generators.
Owing to the high speeds at which the plant had
to operate, centrifugal forces hitherto unknown in
connection with electrical machinery, ‘ together with
other limitations which had to be provided for,
entirely altered the design of the old slow speed
alternators.

In those machines driven by reciprocating engines,
no serious difficulties due to centrifugal forces were
encountered, as large generators were designed to run
at low speeds, so that suitable adjustments could
always be made to meet each case. However, in
connection with turbo plant, the speed must neces-
sarily be high, whether the plant be small or one of
many thousands of kilowatts capacity.

The frequency of the supply was readily obtained
in slow speed sets by adjusting the number of poles
in the rotor to suit the speed decided upon; but
the scope in this direction is very limited with turbo
alternators, because the diameter of the machine is
bound to be reduced to a minimum, owing to the
centrifugal forces. Hence turbo alternators do not
have more than four poles, and in many cases only
two. The frequency per second of the supply from
an alternating current generator depends on the
speed and the number of poles provided on the rotor.

Given the number of pairs of poles and the revo-
lutions per minute of a machine, the frequency of
the supply is furnished by the formula

_ N x R.P.M.

--------

where N = number of pairs of poles on the rotor, and
R.P.M. — revolutions per minute.

For instance, a generator running at only 100
revolutions per minute in order to provide a frequency
of 50 cycles must have	— 30 pairs of poles,

or sixty pole pieces.

As already stated, turbo generators do not exceed
four poles on the rotor, and to obtain a supply of

i	t	i	j -n u 50 X 60

50 cycles with such a machine the speed will be-------—

= 1,500 revolutions per minute.

If the rotor has only two poles, the speed will be
50 *	— 3;000 revolutions per minute. In both

these examples it should be noted that the reference
is to pairs of poles, i.e., two pairs in the case of the
four pole machine and one pair in that of the two
pole generator.

One of the chief difficulties with which the designer
of high speed alternators is faced lies in this limita-
tion to the possible diameter of the machine, so that
as the capacity increases, instead of being able to
make a machine of larger diameter, the extra
capacity has to be obtained by making the generator
longer. This does not present much difficulty from
the mechanical standpoint, but the provision of the
necessary ventilation for carrying off the heat
developed, calls for considerable skill and ingenuity.

In the case of slow speed generators, the heat was
readily dissipated by the large surfaces presented for
radiation, but owing to the large amount of heat
developed in a comparatively small space, turbo
generators require special consideration, and much
attention has been devoted to the problem.

An idea of the difficulty will be formed by con- .
sidering the amount of heat to be dissipated in
large turbo alternators. Take, as an instance, a
generator of 10,000 k.v.a. capacity with an efficiency
of 96 per cent. The heat generated in the machine
and which has to be got rid of is equal to 400 kw.,
and when it is remembered that 1 kw. will raise 100
cu. ft. *of air 18 degs. Cent, in 1 min., such a machine
would raise the temperature of 400 x 100 = 40,000
cu. ft. of air per minute to that extent. To force
these large quantities of air through such a limited
space, calls for very high velocities, as the airways
are long and narrow, and velocities up to 5,000 ft.
and 6,000 ft. per minute are used, or even more in
some exceptional cases.

The chief difficulty consists in getting the air to
the hottest parts of the machine, which are in many
cases so confined that very little air reaches them
unless considerable pressure is used. There are at
present two methods of supplying the air for cooling
high-speed alternators. The first consists of suitable
fans or air scoops attached to the rotor ends. This
method is widely used for moderately sized machines,
although the efficiency of the fans is very low, and
considerable power is wasted, owing to the churning
of the air when running at very high speeds.

In very large sets, separate ventilation is often
adopted, and this appears to have much to recommend
it, and to be more efficient. Besides, the amount
of air delivered to and forced through the generator

can be adjusted to suit the load, and more economical
working of the ventilation results.

Provided the temperature of the air at the outlet
does not differ greatly from that at the inlet, no
benefit will result from increasing the quantity of
air, as the power required to produce the ventilation
lowers the over-all efficiency of the plant, so that it
should be kept as low as possible consistent with
ample ventilation.

Systems of Ventilation.

To appreciate thoroughly the necessity and also the
difficulties in the proper maintenance of turbo
alternators it is essential that the general construc-
tion should be understood, and more especially is
this desirable in connection with the ventilation, as
when cleaning operations are undertaken, a know-
ledge of the principles of ventilating a generator is
of assistance in getting the work properly carried
out, and the air ducts, etc., kept in proper order.
There are several methods of conducting the heat
from the machine, but only three of the better known
and widely used systems will be referred to, i.e.,
the radial system, circumferential, and axial.

Radial System.

The Radial system of ventilation provides for the
air being drawn in at each end of the rotor and
discharged radially through the stator air ducts to
the surrounding atmosphere or into a conduit for
carrying the heated air to the outside of the building.

As the heating of the rotor is only a small per-
centage of that of the stator, its ventilation does
not require the same consideration, and therefore

Fig. 1.	Fig. 2.	Fig. 3.

Stqtor
Winding

S tator
Windir

in one form of radial ventilation, shown in fig. 1, air
is admitted to the air gap and forced out through
the stator core. Owing to the restricted path for
the air, the gap requires to be somewhat larger than
otherwise necessary, so that a modification of fig. 1
is shown in fig. 2. It will be noticed that in addition
to the air being received through the air gap,
provision is made for passing some of it through
ducts in the rotor running from end to end, and
provided with outlets at right angles, to enable the
air to pass across the air gap and out through the
stator air ducts. This provides cooling air for the
rotor and also allows of the air gap being reduced
to a minimum for electrical purposes without having
to allow the question of ventilation to influence its
size. In these two figures, and the succeeding ones,
the winged arrows indicate the inlet air, and those
without wings represent the direction of the air
leaving the machine.

Circumferential Method.

This is applied by admitting the air supply at one
or more points in the outer casing of the stator,
carrying it round a certain distance through the air
ducts, and discharging it again through another
opening in the casing.

Sometimes four openings are provided, i.e., one
at each quarter of the machine, in which case air
may be admitted at each side and discharged top
and bottom. However, two openings are usually
sufficient.

A section of such a machine is shown in fig. 3,
the air inlet being indicated at the bottom, while
the outlet appears at the top. The objection put
forward to this method lies in the long and restricted
path through which the air has to travel.

Axial Flow.

The methods just explained have one common
disadvantage, i.e., that the heat generated in the
core of the machine has to be conducted across the
laminations to the air ducts.

It is hardly necessary to explain that in all alternat-
ing current machinery, the magnetic circuit is made
up of thin sheets of iron or soft steel separated by
sheets of prepared paper, the whole being held
together in a solid mass by suitable bolts and clamps.
The bolts are insulated from the core, so that contact
cannot occur along the edges of the plates to connect
them all together through the bolts. This construc-
tion is necessary in order to reduce the eddy currents
set up in the iron by the rapidly varying magnetic
flux, which quickly causes the iron core to attain
a very high temperature unless some provision is made
to reduce these eddy-current Josses. Referring to
fig. 4, a section of core is represented showing two
radial air ducts. Taking a point B and striking a
line upwards, it will be noticed that, in order for
heat to get from the centre of the core above point
B to either radial air duct, the heat must cross
through numerous layers of paper and iron, from
B to C, or from B to A. Paper being a poor heat
conductor, renders the conduction path in that
direction very poor, so that much local heating is
liable to take place inside the core, and may possibly
destroy the paper insulating sheets between the plates
and be the means of raising the temperature of the
core still higher.

This heating may go still further and cause damage
to the windings themselves, as the maximum heating
of the stator iron of an alternator occurs just behind
the teeth, or just at the back of the tubes carrying
the windings. Referring to fig. 4, it will be seen
that the sheets of iron or steel run from A to D,
and that heat generated at A would be more readily
conducted to D than it would be to C, even though
the distance is shorter in the latter case. Hence if
the ducts are so arranged that the air current passes
across the plates instead of parallel with them, the
heat can be carried off much more rapidly. It is
found that heat conduction in the direction of the

core plates is often more than ten times greater than
it is from plate to plate through the paper insulation.

This fact accounts for the introduction of the axial
flow system, shown in fig. 5. The air enters at
various points at right angles to the upright machine
and after passing through the air ducts, enters a
main duct in the centre of the machine, through
which it is discharged.

Dust Problems.

From the figs, referred to, it is easy to appreciate
the difficulty of removing dust from the various air
ducts when once they are allowed to become blocked
through infrequent cleaning and lack of attention to
the condition of the air supply to the generator.

Open Type Generators.

These machines are fast being replaced or end
covers added. Open type machines are objectionable
in three ways, i.e., (1) they become very dirty; (2)
the noise is excessive; (3) the heated air leaving the
generator is objectionable.

(1)	When the air ducts of a machine become dirty,
the ventilation is restricted, and if the machine has
no margin of temperature rise before reaching a
dangerous point, trouble is likely to occur.

The dust usually encountered in collieries, though
readily held in suspension in the surrounding air,
very soon becomes caked and difficult to move once
it settles down. This is specially so where the
atmosphere is humid and oil vapour rises from the
bearings. This dust often defies all the efforts of
hand blowers and the like, and even compressed air
fails to unstop the ducts once they are allowed to
become blocked. The question of cleaning, however,
will be dealt with later, but it is urged that all
existing open type generators should be enclosed
without delay, and suitable fans fitted to the rotor
for increasing the ventilation.

(2)	The noise is very objectionable, especially when
switchboard attendants have to remain within a short
distance of the machines, as the constant noise drowns
everything else, and no opportunity exists of hearing