The Megger as a Means of Testing Transformer Oil Citation Richards, Roy Thomas (1917) The Megger as a Means of Testing Transformer Oil. Bachelor's thesis, California Institute of Technology. doi:10.7907/e3st-7a56. https://resolver.caltech.edu/CaltechTHESIS:07082024-155851933 Abstract No abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: (Electrical Engineering) Degree Grantor: California Institute of Technology Division: Engineering and Applied Science Major Option: Electrical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1917 Record Number: CaltechTHESIS:07082024-155851933 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07082024-155851933 DOI: 10.7907/e3st-7a56 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16538 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 11 Jul 2024 19:10 Last Modified: 11 Jul 2024 19:13 Thesis Files PDF - Final Version See Usage Policy. 8MB Repository Staff Only: item control page

TBESIS 'T'h e Megger as a 1": eans of Testing Transformer Oil. by ROY 1'. RIC:F.ARD8. Class of Nineteen Pundred and Seventeen Department of Electrical Engineering. 'J'i-1ROO'P ~OLLEGF. OF 'l'EGT-1NOLOGY Pasadena, California 1917. .l.' It is the purpose of tt.is thesis to discuss the importance of, and methods commonly used in, testing transformer oil, to give the results and methods of some experirr,ents on the insulation resistance of transformer oil using the me gg er as the measuring apparatus, to show the relation of dielectric tests on oils to insulation resistance tests, and to give conclusions derived from the experiments. Modern transformer design tends toward greater dependence on oil for insulation and less on mica and other mechanically strong types of insulation. At the present time the use of oil for the higher-voltage transformers is considered absolutely essentia.l, and a verv large proportion of the low-voltage transformers of large size are also oil-insulated. The oils in use at the present time for insulating purposes are obtained by fractional distillation of crude petroleum. To be suitable for transformer insulation and cooling, the oil should have a low viscosity over the range of temperature through which the transformer rr:ay have to operate, because the oil must readily carry the heat from the transformer to the case or cooling coils, it must be free from moisture, acid, alkali, sulphur, or other Katerials which might impair the insulation of the 2. trans former, and it must also have as high a flas h point a s is consistent wit h low viscosity. The oil must not on l y b e of such consistency th a t it will circulate freely and so carr y heat from t h e sources where it is generated to externa l cooling surfaces, but it must also be able to stand voltage stresses. Not only is a high disruptive strength es s ential, but a high insulation resistance is also desirable. The failure t o meet these cond i t ions means t h e failure of t h e transformer in wh ich the oil is used. The method comffionl y employed in testing the insulation resistance of transformer oil is one of substitution. A ~tandard resistance connect~d in series wi th a D'Arsonval galvanometer is connected across a direct-cu rrent s ource of e. m.f. of about 500 volts and t t e deflection of t h e galvano meter noted. From tte de f lect io n o f th e galvanometer, with the unknown resistance connected in pl a ce of the standard resistance, t h e amount of unknown resistance is deterrr ined. This method ~t the best is long a nd te d ious a nd requires apparatus not rea dily portable. The usual test th a t is given transformer oil before it is put into th e trans former, and at frequent intervals thereafter, is one for dielectric strength. The test con- sists in i mmersing a spark gap (usu ally spheres or fla'.t disks) in the oil, t h e gap bein g set at a known distance, usually 0.15 or 0.2 of a n inch, a nd gradually raising the potential until rupture of the oil occurs. 3. To the writer's kno~ledge no tests on th e insulation resistance qualities of transformer oil , using tte megger as the resistance measuring instrument, have ever been carried on. The meg ger is a rea d ily portable instrument of the galvanometer type. It is used principally in measur- ing high resistances where great accuracy is not required. The machine consists essentially of a srr,all d-c generator and a special form of pivoted galvanometer, consisting of a perrr,anent magnet, a soft iron core, a current coil and two pressure coils. These three coils are rigidly attached at a fixed angular distance apart to the sha ft which carries the pointer. The generator and galvanometer are mounted in a single box provided with two binding posts for connecting the unknown resistance and a handle for turning the arrr.ature of the generator by hand. To determine the resis- tance of a circuit it is only necessary to set the megger down on a fairly level base, connect the circuit wires to the line and earth terminals, and give the generator h andle half a dozen rapid turns, when the pointer comes to rest and indicates the resistance of the circuit in megohms; whence the name - in egger. The auxiliary apparatus used in these tests were two spark gaps, one consisting of two flat disks 0.5 of an inch in diameter, set at a distance of 0.2 of an inch for testing the disruptive strength, and t h e other consisting of two hollow spheres 3 inches in diameter for testing the 4. insulation resistance of the oil. 'J'he oil sample bottles wer·e given a rigorous treatment to be absolutely sure that ttey contained no moisture or dirt before they were filled wit h the oil to be· tested. The bottles were first cleaned with a weak solution of hydrofluoric acid and t h e residue left by t h e acid rinsed out with high grade gasoline. They were then placed on drying racks, and heated air, which had been passed through sodium-carbonate to remove all moisture, ~Has forced into them. When thoroughly dry they were placed on racks, and air at room temperature,and free from moisture, was forced. into them. When the bottl~s reached room temperature they were tightly corked and dipped into melted parafine, making an air tight seal. Care must be taken to have t h e electrodes free from dust or particles of cotton fibre before performin g the tests. The importance of this became evident when it was found that t h e method of polishing the electrodes had a great effect both on the dielectric strength and the insulation resistance. After polishing wit h a cotton cloth, which usually lec1.ves particles of cotton fibre on the electrodes, the pu~ure voltage was reduced to about onehalf the true value of t h e dielectric strength of the oil. The insulation resistance was still more affected, due to the shorter distance between the electrodes. In order to be sure that the vessel and the electrodes were perfectly 5. clean the following method was used: The polished elec~ trodes were washed while immersed in the oil; they were th~n. taken out and the vessel was washed by stirring up the oil in it. This process was repeated several times, using clean oil for each operation. The oil used in these tests was what is commonly known as 11 high flash transformer oil." The follmving character- istics were found upon analysis: specific gravite - .855 . at 22°~; viscosity - 3.44 at 24°C; flash point - 165.5 ° C; burning point 184.5 °C. Effect of Temperature. The insulation resistance of oil varies greatly with changes in temperature. A slight increase in temperature produces a comparatively large decrease in the insulation resistance. At ordinary temperatures tte specific resis- tance is so high as to render its determination exceedingly difficult. A test was made on the oil with varying degrees of ~eat and a curve plotted from the results obtained. This curve shows clearly the decided effect of heat on the insulation resistance. The temperature must be ~aken into account in any test on the insulation resistance of transformer oil. A testms also made of the di- electric strength of oil with increasing degrees of heat. As is shown in the results (curve sheet No. 2) the dielectric strength of oil does not vary grecctly with slight changes in temperature. 6. Effect of Moisture. Transformer oil, as far as the dielectric strength and especially the insulation resistance is concerned, is extremely sensiti;re to moisture. Shipment of oil in either tank cars or steel drums introduces t h e possibility of moisture or other foreign substances entering t h e oil. Some of the moisture may settle to the bottom, where it can be gotten rid of by simply drawing off a limited quantity of oil. A certain percentage will however be retained and kept in circulation. From the curves it is seen that even 0.04 of one percent reduces the dielectric strength to onehalf and the insulation resistance to one-seventh of the original value. It is evident from these tests that the importance of dry oil cannot be overestimated. The following methods were used in testing for the effect of moistur~: As has been shown in the tests on the effect of temperature on the insulation resistance of transformer oil, a slight change in temperature makes quite an appreciable change in the insulation resistance, therefore it was t hought advisable to make one teat holding the temperature constant at room temperature, or 20 ° C., and another holdin g the temperature constant at 55°C., the ordinary working temperature of most transformers. Oil which was known to be perfectly dry was given a dielectric test, withstanding~ potential of 65,000 volts bet ween two polished brass balls set at a distance of 0.2 of an inch apart, 7. before breaking down. The test sample was then placed in the receptacle containing t h e hollow sph eres electrodes and the meg ger connected to the gap. The gap was then adjusted until the megger indicated 5000 megohms, the h ighest reading for which it was calibrated. Water was then added in the proportion of one part of water to 10,000 paits of oil by volume. The mixture of oil and water was thoroughly emulsified and then allowed to stand until all bubbles - had disappeared. It was then again subjected to the two tests as outlined above. Tests were continued with increasing proportions of moisture until fairly constant results were obtained, indicating that the oil was completely saturated. The results of these tests indicate that the meg ger detects moisture in the oil much more readily than the dielectric tester. Conclusions. The field for which the meg ger test is best a dapted is power plants, sub-stations, etc., where an oil laboratory containing apparatus for the dielectric test is not at hand. A fair indication o~the condition of the trans- former oil could be obtained by subjecting it to a megger test. If the oil "meggered" over 3000 megohms, at 55° C between 3 inch sphere electrodes set at 0.04 of an inch apart, corresponding to a nun~ure voltage of over 45 kilovolts, the oil would be suitable f or use in the transformers. Oil which came under this value would either have to be Vo passed through the filter press or taken to the laboratory for a complete analysis. As to tte relation between insulation resistance and dielectric strength, the results of these experiments show that the insulation resistance of an oil is not an indication of its dielectric strength unless the temperature is also taken into consideration. Owing to the number of variables which may creep into the work, the difficulties in carrying on a study of this kind are great. It was found very difficult to duplicate results even with apparently identical conditions. rhis applied to both the dielectric and insulation resistance tests. Great care must be used in measuring the distance between the electrodes in the insulation resistance tests, as this distance is very small and is one of the controlling factors in the test. A great number of tests were necessary to obtain re- liable figures, and curves as reproduced in this thesis were plotted from average values. 9. Effect of Temperature On the Insulation Resistance Of Transformer Oil, Oil tested: Specific gravity - 0.855 at 22°C. Viscosity 3.44 at 24°C. Flash point - 165.5°C. Burning point - 184.5 ° G. Test No. 1. Observation Temperature Resistance Avera ge Resistance No. oc Megohms. Megohms. 1. 50 5000 2. 50 5000 3. 55 4600 4. 55 4400 5. 60 3100 6. 60 3300 7. 70 2350 8. 70 23 00 9. 80 1600 10. 80 1600 11. 90 900 12. 90 1000 13. 100 600 14. 100 700 15. 110 425 16. 110 425 17. 120 300 18. 120 250 5000 4500 3200 2325 1600 950 650 425 2 75 Rffect of Temperature On the Dielectric Strength Of Trans former Oil. Oil tested:Dry oi 1, same as in 'T'est No. l; Test No. 2. Observation Temperature Kilovolts No. oC 1. 20 60 2. 20 60 3. 30 60. 5_ 4. 30 60.5 5. 40 60.6 6. 40 61.4 7. 50 61.4 8. 50 61.6 9. 60 62 10. 60 62 11. 70 62.5 12. 70 62.5 13. 80 63 14. 80 63 15. 90 63.2 16. 90 63.6 17. 100 64. 18. 100 64. Aver. Kilovolts 60 60 .5 · 61 61.5 82 62 . 5 63 63 .4 64 1. Test No. 2. continued. Observation Temperature Kilovolts No. or-. 19. 110 64.6 20. 110 64.4 21. 120 65 22. 120 65 A~re r. Kilo~rol ts . ,J. 64.5 65 Effect of Moisture On the Insulation Resistance of Transformer Oil. Oil tested same as in test No. 1. Temperature of Oil- 20°C. Test No. 3. Observation Wat er-p2,rts in Resistance Av. Resistance No. 10,000 by vol. Megohms Megohms 1. 0- 00 5000 2. o.oc 5000 3. 0.05 3000 4. 0.05 3000 5. 0.25 1400 6. 0,25 1300 7. Q.5 600 8. 0 ,5 550 9. 7 . 280 10. 1. 32 0 li. 1.5 250 12. 1.5 270 13. 2 260 14. 2 260 15. 2.5 235 16. 2.5 245 17. 3. 225 18. 3' 225 -'· 5000 3000 1350 575 300 260 260 240 225 Test No. 3' continued. Observation Wat er-pa rt s in Resistance .Av. Resistance No. 10, 00 0 by vol. Megohms Megohms 19. 3.5 200 20. 3.5 200 21. 4.0 190 22. 4.0 170 23. 4.5 185 24. 4. 5 18 5 25. 5.0 150 26. 5.0 150 200 185 185 150 14 Effect of Moisture On the Dielectric 8trength of Transformer Oil. Oil tested same as in Test No. 1. Temperature of oil - ao 0 c. Test No. 4. Observation Water parts in No. 10,000 by vol. 1. ().0 65 2. o.o 65 3. 0 .05 50 4. o.o5 50 5. 0 .17 35 6. O, 17 35 0, 42 25 8. 0 ,42 23 9. (' .6 22 10. o.6 22 11. c.s 20 12. 0.8 19.4 13. 1 19.0 14. ·1 17. 4 15. r.) <..J 15.5 16. 2 15.9 17. 3 14. 18. 3 14 ,'50 7. - Kilovolts Av . Ki 1 ovo 1 ts . 65 50 35 24 22 19.7 18.2 15.7 14.25 l! Test No. 4 continued. Observation Water parts in No. 10,000 by vol. Kilovolts 19. 4 13 20. 4 13 21. 5 12 22. 5 12 Av. Kilovolts 13 12 1 Effect of Moisture On the Insulation Resistance of Transformer Oil. Oi 1 tested same as in Test No. 1. Temperature of oi 1 - 55 °G. Test No. 5. Observation Wate·r . parts No. Resistance in 10,000 by vol. Megohms 1. 0 5000 2. 0 5000 3. 0.1 3100 4. 0.1 3100 5. o.25 1000 6. Q.25 11000 7. Q.4 500 8. 0 .4 500 9. 1. 260 10. 1 240 11. 2 230 12. 2 220 13. 3 165 14. 3 175 15. 4 150 16. 4 150 17. 5 125 18. 5 125 Aver. Resistance Megohms 5000 3100 1000 500 250 225 170 150 125