US1922103A - Process of improving the quality of transformer oils - Google Patents

Description

Aug. 15, 1933. T. E. LAYNG ET AL 1,922,103

PROCESS OF IMPROVING THE QUALITY OF TRANSFORMER OILS Filed on. 26. 1929 EREA/mOW/V POTE/VT/AL //v F aws/m0 VOLT-$ I NVENTORS ATTORNEY Patented Aug. 15, 1933 UNITED STATES PATENT; OFFICE PROCESS OF IMPROVING THE QUALITY OF TRANSFORMER OILS Application October 26, 1929. Serial No. 402,759

7 Claims.

This invention relates to an improved insulating oil and to the process of making same and relates in particular to a transformer oil of enhanced insulating power and to the process of converting ordinary commercial transformer oils and the like into insulating oils of higher dielectric strength.

With the marked increase in the utilization of high potentials, there has been created a notable demand for electrical insulating oils of great dielectric strength. Many efforts have been made to improve the quality of commercial transformer oils by more drastic methods of refining. Improvements in the design of transformer containers have eliminated to a very large extent, if not entirely, the possible oxidation of the oil by maintaining an inert atmosphere in the holder.

In seeking to secure a higher degree of resistance of an oil to electrical breakdown we have borne in mind that three prerequisites for a good insulator are (1) it should resist, or not be readily capable of, electrical dissociation; that is ionize under electrostatic stress; and (2) it should ofier high mechanical resistance to the passage of ions and electrons. In other words, from the standpoint of insulation of oil the medium employed as an insulator should have its molecules in a high state of association. Polymerization therefore is desirable. more, the insulator should consist of highly stable chemical compounds. It should therefore be an inert material, preferably chemically saturated and possessing no free valences. Finally, to impede the migration of ions, the insulating material preferably should be present in a colloidal state.

We have found that the addition of a relatively small proportion of an ethylate of a fixed alkali, such as sodium ethylate, exerts a marked improvement on the dielectric strength of the oil in a breakdown test.

The manner in which tests of the effect of the ethylate was determined'is as follows:

The voltage at which a sample of oil broke down was measured with a gap consisting of two parallel discs, one inch in diameter spaced 0.20 inches apart. The voltage was supplied by a high voltage transformer and the potential applied was controlled through the voltage applied to the primary windings. In the preliminary tests, the voltage was controlled by hand, but since the time during which the sample is subjected to the electrostatic stress is important, in

Furtherthe later tests a motor driven rheostat was used to equalize the time intervals.

Breakdown of the sample opened a circuit breaker so that excessive carbonization of the sample by the spark was avoided. In each test the case containing the gap was filled with oil and then tapped to remove any air bubbles which might have been entrapped.

After each trial the apparatus was again tapped to remove gas bubbles formed by the spark.

The sodium ethylate in a small amount of absolute ethyl alcohol was added to the hot oil (120 C.) in order to quickly remove the alcohol.

The results of preliminary tests are shown in Table 1. In these tests hand control of the voltage was employed.

It was found that the results obtained even on the untreated oil varied from day to day appar ently being greatly influenced by the humidity conditions at the time of the test. It was therefore necessary to always make a comparison of the untreated oil with the treated as a. check in each series of tests.

The effect of varying the amount of the so- 8 dium ethylate is shown in Table 2. From these results, it appeared that a concentration of 0.03 grams of sodium ethylate per 100 c. c. of oil produced the maximum eilect.

A series of very carefully controlled tests were therefore run at the above concentration. The motor driven rheostat was used with a starting voltage of 7,300 volts. The temperature of the oil samples varied from l8-19 C. The results are shown in Table 3. From the data presented it will be noted that the untreated oil has a final average breakdown 6 per cent lower than the average initial breakdown while the treated oil has a final breakdown 32 per cent higher than the average initial breakdown. These results are also shown graphically in Figure 1.

This notable increase in the dielectric strength of the oil when treated with sodium ethylate might be assumed to result from the removal of small traces of water which, of course, are known to have a marked effect on the dielectric strength of an oil. It is supposed that the amount of water which is held in suspension in minute drops in the oil is due to the movement of these drops to the denser portions of the field when voltage is applied. By such movement and localization larger drops or aggregates are formed by collision. After the attraction to 'and contact with a metal part, thus acquiring 110 not found to be the case.

the same potential, repulsion occurs. If the field .is uniform, the drops orient in conducting chains action.

We have noted that the addition of sodium ethylate to ordinary transformer oil has produced an orange colored precipitate. If a suitable concentration of the sodium ethylate was present, the sludge was found not to settle to any great extent. It was found that apparently the presence of this sludge increased the insulating qualities of the oil, or in any event that the oil had acquired improved dielectric strength by the treatment which resulted in such precipitate.

From the foregoing it appears that sodium ethylate causes the polymerization of certain constituents of the oil, causing them to take the form of colloidal aggregates. The improvement in the insulating qualities of the oil, therefore, may be attributed to the removal, through polymerization, of certain undesirable constituents and the resulting colloidal aggregate in a state of non-settling fineness prevents the migration of ions and the formation of conducting chains.

From the tables set forth herein it will be apparent that drying agents such as metallic sodium, calcium chloride and phosphorus pentoxide do not produce the same effect as sodium ethylate which, therefore, from the standpoint of this invention acts not primarily as a drying agent, but as a reagent which removes by polymerization from active conducting participation in the oil certain deleterious components.

While sodium ethylate is preferred on account of its relatively low cost and availability, we do not exclude the employment of potassium ethylate nor other metallo-organic compounds acting in a substantially equivalent manna and therefore embrace the compounds of such metals with various other alcohols such as methyl, propyl, butyl, and the like. The employment of sodium or potassium ethylate or similar metalloorganic compounds capable of polymerizing transformer oil in like manner to eliminate bodies detrimental to high dielectric strength likewise is not precluded.

The tables follow:

TABLE 1 [Breakdown potential in volts] All amounts of sodium are as sodium ethylate.

TABLE 2 Variation in concentration of sodium ethylate [Breakdown potential in volts] Averl 2 5 ago Oil unheated and untreated 30, 000 30, 000 21,000 27, 200 Oil plus 0.017 gram Na per 100 cc.

oil M 26, 500 34, 000 33,000 31,000 Oil plus 0.033 gram Na per 100 cc.

oil 28, 000 44,000 42,000 38, 000

Oil plus 0.066 grain Nu per 100 cc" 34, 000 31,000 32, 500

TABLE 3 Comparison of breakdown potential of untreated oil with oil containing 0.03 gram Na as sodium ethylate per 100 cc.

[Untreated oil] A ver l 2 3 4 0 age 31, 200 24,000 20, 000 if), 000 24, 100 28, 000 2B, 400 17,000 25, 000 25, 700 24, 400 20, 800 20, 000 23, 600 20, 100

Average final breakdown 0% lower than average initial breakdown.

d Average final breakdown 32% higher than average initial breakown.

The oil used in the conduct of the foregoing determinations was ordinary commercial transformer oil of petroleum origin, well refined in the usual manner of refining such hydrocarbon oils for insulating purposes.

The appended drawing is a graph showing the effect of 0.03 grams of sodium in the form of sodium ethylate-per 100 cubic centimeters of oil on the breakdown potential of the transformer oil. The graphs are built from the average of several determinations in each case.

What we claim is:

l. The process of improving the dielectric strength of insulating oil which comprises incorporating therewith a relatively small proportion of sodium ethylate.

2. The process of increasing the dielectric strength of transformer oil which comprises incorporating therewith a relatvely small proportion of an alcoholate of an alkali metal.

3. The process of increasing the dielectric strength of transformer oil which comprises reacting on the oil with sodium ethylate in an amount less than 1 part of sodium ethylate in 1,000 parts of said oil.

4. The process of treating transformer oil to increase its dielectric strength which comprises reacting on a well-refined petroleum oil with a relatively small proportion of sodium ethylate whereby a colloidal aggregate of polymerization which comprises incorporating a transformer oil with sodium ethylate, the concentration of the sodium ethylate being selected to prevent substantial settling of any precipitate formed by the action of the sodium ethylate on the ingredients of the oil.

T. E. LAYNG.

M. A. YOUKER.

Images