Transactions

DANTSIZEN—LUBRICATION OF GENERAL ELEC TRIC STEAM TURBINES 495
A possible explanation for this may be that, although the
protective molecules on the iron are packed closely enough to
keep water in fine globules in suspension in oil away from the iron,
large and heavy globules not in rapid motion will tend to break
through the film.
Another point to be seriously considered is that too high a content
of free organic acids in a turbine oil is detrimental, in that it
may give rise to or be associated with:
1 The formation of metallic soaps.
2 Formation of permanent water emulsions.
3 Sludging of the oil.
4 Destruction of oxidation inhibitors.
The maximum permissible amount of free organic acid in an oil
depends upon the type of oil used.
With oils having antioxidants present, the amount of acid
permissible should probably not exceed that which corresponds
to a neutralization number of 0.15. With no antioxidants present
and depending upon the oil used, an amount equivalent to a
neutralization number of 2.5 might be permitted.
The carboxyl groups which, in an organic acid are responsible
for inhibition of rusting, are also responsible for attack on metals
and oxides of metals forming soaps which are almost insoluble in
both oil and water. Most of these metallic soaps act as active
catalyzers for the further oxidation of the oil.
A high percentage of organic acids and associated oxidation
products give rise to permanent emulsions of water in oil because
there is a sufficient number of the hydrophilic groups
present to take care of the large contact area between the
water globules, which constitute the discontinuous phase of
such an emulsion, and the oil which constitutes the continuous
phase.
High acid values in oils are associated with sludging of such
oils. Such sludges are insoluble oxidation products of the oil.
F a b m e r T e s t f o r O x i d a t i o n C h a r a c t e r i s t i c s o p O i l
One of the best tests for the oxidation characteristics of an oil
has been developed by Harold Farmer* of the Philadelphia
Electric Company. He has pointed out that iron, brass, and
copper, particularly the last, are active catalyzers in speeding the
oxidation of turbine oil.
Because the American Society for Testing Materials has no
test method for finding the relative oxidation rates of various
turbine oils, in the Schenectady Works laboratory, we are using
a modification of the Farmer test; 400 ml of oil are placed in a
pyrex tube with three Vs-m-diam rods, 10 in. long, of three different
metals, i.e., iron, brass, and copper. In the same tube at the
start of the test is placed 12 cc of water. Examination of the
tubes is made daily to assure that approximately that amount of
water is present. The oil and water are maintained at 210 F
and, through the oil and water, air is bubbled at the rate of 10 1
per hr. Samples of oil are tested once a week for neutralization
number, viscosity, and precipitation number. The viscosity
sample is returned to the tube and oil lost in the determination
of the neutralization number is replaced by the addition of new
oil to that under test in the tube. The apparatus is shown in
Figs. 9 and 10.
While we find that wide variations exist in the oxidation rate
of different oils, at present there is no inclination to condemn
some oils which do not prove to be among those giving the best
8 “Copper Catalysis Accelerates Turbine Oil Oxidation,” by Harold
Farmer, Electrical World, vol. I ll, May 20, 1939, pp. 1452-1453 and
1519.
T. H. Rogers and B. H. Shoemaker, Industrial and Engineering
Chemistry, Analytical edition, 1934, vol. 6, p. 419.
results in this apparatus. The oil, for example, which we are
using successfully in the Schenectady Works turbines cannot be
graded as excellent from the oxidation standpoint. On the other
hand, we appreciate that the oxidation-inhibited turbine oils now
on the market are a valuable contribution to the turbine industry.
Discussion
M. D. B a k e r . 7 The difficulty in writing specifications for the
purchase of turbine*oils can be fully appreciated. As the author
points out, specifications of a very general nature can be made,
and generally these do not give information that will predict the
service life of the oil to be purchased.
During one of our investigations, twelve light turbine oils of
different brands were studied in the laboratory. All twelve oils
were found to be within the limits of the specifications which included
tests for gravity, flash, fire, viscosity, and steam emulsion.
No life tests were made. Six of these oils were picked
for an accelerated life test in a small unit. These accelerated
tests gave service records varying from 50 hr on the poorest
oil to 1800 hr on the best oil. The life of the oil was determined
by the length of time required to reach a neutral number of
0.8.
Refining methods have been changed to improve the quality
of the oils delivered. In conjunction with this improvement,
antioxidants have been added to the oils to retard the rate of
oxidation and prolong their service life. An example of this increased
life can be illustrated by the service record of one oil.
Without the antioxidant after 13,000 service hr, the neutral
number was 0.85. This same oil with the antioxidant added now
has 37,000 service hr and a neutral number of 0.08. The addition
of the antioxidant has decreased the rate of sludge precipitation
in the oil coolers, so that now 15,000 hr elapse between
cleanings where previously 5000 hr was considered a long period
of time. This lengthening of time between cleanings has given
longer periods of metal passivity in the coolers and has reduced
the rate of metal poisoning.
The detrimental side to the addition of antioxidants must also
be considered. The author mentions that in the older types of
oils the neutral number of 2.5 could be carried with safety, but
that in the newer types of oils one company has set a limit of
0.15 as the maximum neutral number that can be carried with
safety. The older type of oils increased in neutral number at a
more or less accelerated but definite rate with no sudden or rapid
increases. Knowing the condition of the oil, its life could be predicted
with some degree of accuracy when the rate of rise of the
neutral number was plotted against the service hours.
The accelerated-service test on one oil containing an antioxidant
gave the following results: At 1250 service hr, the neutral
number was 0.08; at 1394 service hr, the neutral number was
17.4. This rise in neutral number when it occurred was very
sudden and very rapid. So, unless some test other than neutral
number is devised and used, the life of an oil containing an antioxidant
cannot be accurately predicted.
The neutral number of 0.8 was set for the top limit of the oils at
Springdale Station, as experience has shown that, when an oil
reached this value and under normal operation with no water
present in the oil, any water contamination would create a heavy
sludge. Units having constant moisture infiltration into the oil
system have been operated safely with neutral numbers as high
as 4.0 but this is too high for a unit that does not have regular
or frequent infiltration of water.
Corrosion in turbine systems has been experienced for years,
7 Springdale Power Station, West Penn Power Company, Pittsburgh,
Pa.