Document Jra4MdN17gdo8j08MQ09M5B3r

4 Considerations for Evaluating Transformer Askarel j i By P. G. Benignus, Manager, Technical Service, Aviation and Special Fluids, Monsanto Chemical Co., St. Louis 66, Missouri. All photos courtesy Monsanto Chemical Co. i 1 Few electrical experts are also chemists. Perhaps that's why so many tion and thorough inspection, without exception in the cases examined, Comparison of Askarel and Mineral Oil transformer operators (and buyers) showed the reason for every askarel Transformer users who are not make the mistake of thinking trans transformer failure to be the same: aware of the marked difference be former "oil" and "askarel" are alike. the entrance of moisture. However, it tween askarel and mineral oil have a They are not--the differences in these is obvious that any transformer can tendency to evaluate askarel liquid two kinds of liquid insulation go be fail from inadequate design, over insulation in terms of the more fa yond the obvious difference that oil voltage, damaged insulation, etc. miliar properties of mineral oil. This burns and askarel does not. The dif Since askarel (and mineral nil is the cause of much confusion. The ferences also extend to their insulating type) transformers are often selected "check point" values of oil and askarel power, operating characteristics, and for use in damp areas, sometimes di are different; the interpretation of the manner of maintenance inspection. rectly exposed to the elements, and test values is different. The previously Due to the mistaken belief that trans frequently operated where Hooding mentioned "Maintenance Guide" at former askarel and mineral oil are may be expected---the need to exclude tempts to end the confusion by point similar, there has been a great deal of moisture entry calls for optimum ing out the difference in operating confusion concerning maintenance transformer design-engineering to characteristics and check tests of the procedures and operating characteris assure dryness. two fluids. However, such literature is tics of the two fluids. To put an end Monsanto's interest in transformer necessarily more directive than ex to misconceptions, our firm is printing askarel is twofold: we manufacture planatory. With a minimum of "rea an easy-to-follow maintenance guide and supply the dielectric fluid and we sons why" the askarel maintenance based on evaluation experience with use a large number of askarel trans guide directs the engineer to make askarel over 30 years of service. The formers in dozens of plant locations just three simple visual inspections of service history study shows that to benefit from their dependability, the askarel fluid and determine its askarel users generally arc confused economy, and fire safety. dielectric strength. Moisture deter over several major points; these are Since we neither design nor build mination by the Karl Fischer method discussed here in the hope that they transformers, specific design detail on is suggested only if the dielectric i will benefit all operators of askarel new equipment is left to experts in the strength is found to be less than 20 to \i transformers. field. But, the general facts relating 25 KV at 25C. This simplified check ii Some of the earliest askarel trans to the moisture problem must be con ing, which can be easily carried out formers were first made around 1930. sidered from the viewpoint of the by any user, comes remarkably close Many of these are still in service. maintenance engineer. There are to "leaving 'em alone." Unlike oil, After so many years of use, the out many thousands of askarel transform periodic power factor and volume re side appearance of some of these units ers now installed, many only five or sistivity determinations are not needed would not warrant a prize, but their less years old. for askarel check testing, and if run, interiors remain clean. The original Recently, an authority on askarel the values obtained are completely un fill of askarel insulation is still in use transformers was asked what he con related to what these values would and there has been little or no main sidered to be the most important point mean when applied to mineral oil. tenance. They are giving the 30 years of maintenance. His reply: "Keep 'em The reason is: oil and askarel are com of trouble-free service life promised dry, then leave 'em alone." pletely different chemical substances. by the manufacturer. Keeping the transformer interior The following explains why power However, as with any type of equip dry is indeed the watchword and not factor and resistivity tests are rela ment, over the past quarter century merely an oversimplification of fact. tively unimportant, while dielectric there have been a few instances of If steps are taken to keep moisture strength is most important--in in unexpected failure. In every case, in out of askarel, the liquid insulation specting and judging askarel insula vestigation proved that the breakdown will undergo no change and the trans tion. could have been avoided. former will operate trouble-free for evaluating Power Factor A review of certain facts of opera 30 years or more. The power factor of freshly made Reprinted from Insulation, October 1902. Copyright 1962, Lake Publishing Corporation, Libertyville, Illinois 1 DSW 321753 STLCOPCB4073404 transformer askarel is close to 0.1 per cent at 100C and 1 kc (the condi tions of measurement in our quality factor of mineral oil. But despite this sensitivity, askarel's insulating value is unimpaired. new transformer should not exceedtwo percent at 20C and 60 or 70 cycles. While this arbitrary limit is lk control laboratories). This same value Experts know that to correctly about 10 times as high as the power level, or even slightly lower, is found when accurately measured in the held, measure the low |>ower factor of new askarel requires exceptional cleanli factor of newly made askarel, consid erable care in check testing is still | usually at 20C and 60 cycles. ness of the test cell and strict adher . required to obtain an approximately This initial low power factor of new ence to the proper techniques (ASTM correct power factor reading within askarel is essentially the same level method D-924). even this higher level limitation. normally found in newly-made, dielec One leading counselor correctly Authorities with years of experi tric grade mineral oils. Such striking states that the .critical technique re ence with askarel agree that after similarity of values in both types of . quired to accurately measure the years of continuous service, the liquid insulation misleads many user- power factor of askarel is impractical askarel's power factor will be high, as into assuming that the power factor of these two entirely different fluids for field use. Experience shows that the unskilled or not-too-careful at much as 20 to 80 percent (or more) at 100C and 60 cycles, or about one- \ b should run parallel throughout their tempts to determine the power factor tenth of this if measured at 20C and respective service lives. But this does of new askarel in the field lead to 60 cycles. not happen. In service, the power fac tor of askarel goes up to a greater extent than that of mineral insulating unreliable results. The testing should be done with the greatest care, exactly as specified by either the ASTM The majority of electrical engi neers and maintenance operators have found that a high power factor in uiL oil--but the rise has little or no signifi method (or by a similar method askarel is not important, however cance. ology) to eliminate trace contamina much alarm such a high value would The prime reason why the power factor of askarel and oil do not go tion and must be carried out in a well-equipped laboratory. warrant if found in mineral oil. The only qualification to be added is this: *i hand-in-hand is that unlike mineral With a power factor so sensitive to while high power factor in itself is oil, askarel is a relatively polar me traces of contamination, it stands to unimportant, some of the possible dium. For example, the dielectric con reason that when the askarel is put causes of high power factor may war stant of transformer askarel is about into a new transformer, the power fac rant investigation. However, in the 4, or nearly twice as high as the tor of the fluid can be expected to absence of other test data that can i dielectric constant of mineral oil. In increase immediately on contacting indicate the advisability of thorough 9 finitesimal amounts of ionic conduct the insulation which contains traces inspection, relatively high power fac ing impurities in askarel can markedly of soluble ionizable compounds. Vari tor is to be expected. i< i: increase its power factor. The same ous transformer manufacturers dictate On the other hand, a high power impurity may not affect the power that the power factor of askarel in a factor when found in mineral oil usually indicates chemical and hence electrical deterioration. This is true of askarel only if it has undergone strong arcing, readily detected by darkening and sedimentation. Figure 1, a specially air-conditioned room and test cell shown above is used at Monsanto to measure the power factor and resistivity of askarel. Volume ResistIvify Requirements The resistivity of askarel in prop erly constructed new or rebuilt trans formers will not'be significantly lower than the 100 x 10 ohm-cm, at 100C, 500 v, d-c; 0.1-inch gap, specification limit. However, this value can be ex pected to decline fairly rapidly even in a normally operating transformer. Resistivity should not drop below about 25 x 10 ohm-cm, and ample experience shows it does not fall be low 5 or 10 x 10 ohm-cm, even in the presence of unusually heavy con tamination, other than water contami nation. The reason for this is that askarel, by its chemical nature, is practically completely inert from an electrical standpoint--it simply does not conduct current. In consequence, o ( i s DSW 321754 STLCOPCB4073405 Table J--Comparison of Values Found in Askarel Insulation After Various Stages of Use A B C D i<: K Properties Color New askarel i.igiit straw Askarel in properly* built new or rebuilt units prior to use. Light straw Typical values after normal operation. (Survey of 50 units mostly 2-5 yrs. old-- some 10) Light struw Non-arced hut slightly contaminated askarel in im properly* built units, new or slightly used. Can liuve foreign shades, blue, green, red cast. Same fluid in column D after relining with earth and littering. Foreign shades remain showing extraction of oil soluble color. Heavily urceil askarel. Of the very few uskarel failures, the majority were due to excess water-- remainder due to internal eleelricul discrepancy Black . Condition Moisture, 25C Clear, free from Clear, practically Clear, trace of particles free from par particles ticles <30 ppm <30 ppm 10-70 ppm, all dissolved water Clear, slight amount of particles 20-100 ppm, ull dissolved water Clear, frt*c from Contains car particles bon particles <30 ppm near or aljovc, saturation level, 125 ppm, par tially uiidissolved water Dielectric strength, 25C, 0.1-in. gap Volume resistivitv, 100C, 500 V, d-c, 0.1 inch gap 35 kv min 35-48 kv 35-45 kv 100 x 10s ohm- 20 x 109 ohra-cm 15-400 x 109 cin min, usually min, usually at ohm-cm 500-1,500 least 100 x 109 30-42 kv about 5 x 109 ohm-cm 40-15 kv 5-15 kv at least 1,500 x about3xl09 109 ohm-cm ohm-om Power Factor: 100C, 60 cycles 2-5% 25C, 60 cycles 0.05-0.1% 10-25% 0.5-2% 10 to near 80% 30 to 100% 2-5% 0.5 to near 10% about 7 to over 0.1-0.2% 15% 100% at least 25% * A properly built unit is defined as one in which the materials of construction are chemically and electrically compatible with a.share!. $ Out of too many transformers to New Transformers--'"Vaccinated" tor construction design is popular; a count over the years, Monsanto plants had two that failed after 13 and 15 Against - Moisture Transformer makers, aware of the this completely fills the transformers with fluid and thereby eliminates any years of service, respectively. These importance of keeping out moisture, "air space." This construction greatly were also inadequately sealed--with now usually offer welded construction minimizes moisture entering units cork gaskets that had deteriorated at no increased cost over gasketed even when located in damp areas. after many years of full outdoor ex construction. Where gasket seals are On indoor installations, normal posure to the elements. Another fac needed for hand-hole openings, they gasketing of the transformer is, of tor--these units were not required to are designed with a recessed fitting course, entirely satisfactory unless be constantly on line; their inter to allow only minimum contact with there is excessive dampness in the en mittent operation intensified breath the askarel and with the elements. vironment and wide temperature ing and moisture condensation. The For gasketing, instead of plain and fluctuations that accentuate the breath amount of water that entered and coarse grain cork--fine grain cork ing. If the user cannot be certain of a accumulated would have caused fail impregnated with nitrile rubber makes dry atmosphere, it is wise to consider ure in any type of transformer. These the best material. Nitrile rubber alone fitting the unit with a dehumidifying units have since been rebuilt and has been used, but this polymer is breather. "modernized" for outdoor service, slowly eroded over the years by As an example, highly reliable <r- simply by a welded sealing of top askarel. There are, however, a num askarel units are used in transformer- cover, fins, instrument connections, ber of askarel resistant polymers; rectifier power supply sets, for mili and all other fittings. A blanket of dry these include "Teflon," "Viton," and tary communication equipment (trop nitrogen over the askarel has also most silicone "rubbers." While rela ospheric scatter) in remote areas. been provided and the units are fit tively costly, these materials warrant These are kept dry by channeling ted with pressure relief devices and consideration for their long term use, their breathing through a canister of pressure gauges to check on pressure at least for the smaller size gaskets. silica-gel attached to the side of the fluctuations. Overseas, a well-known conserva transformer. The canister is fitted DSW 321755 5 STLCOPCB4073406 acceptable electrical grade askarel is restricted to only a few tenths of one part per million, even under this severe thermal stress at the finished product's boiling point. The chemical stability check allows no more than the same tiny amount of chloride for mation when the askarel is treated with warm alcoholic caustic solution tboth procedures are documented at ASTM and elsewhere). We consider it essential to carry out this check testing on every lot; to determine all the physical and elec trical values; and to record this data for each shipment. Because of these precautions and knowledge of manu facture, the user of askarel insulation can l>e sure that its chemical and ther mal stability conform with the re quirements prescribed by the electri cal industry. Without this strict check-up and data recording, it would be difficult to lie sure whether any deterioration of the initial electrical values would l>e due to lack of inherent stability, or due to expected pickup of contami nants, of which moisture is by far the most deleterious. Since we check, test, and prove that the new product has maximum stability--we can be sure that any detrition of electrical proper ties must be the result of severe arcing or water contamination. Moisture: Askarel's Greatest Enemy All users of liquid filled trans formers understand that moisture is a major problem, in askarel as well as in mineral oil equipment. The mois ture content of new askarel is not over 30 parts per million. This should not be allowed to increase above about 70 parts per million in a normal operat ing transformer. The proper test for moisture in askarel is the Karl Fischer method. However, this is not a practical field test because it usually requires labora tory facilities. As a service. Monsanto and various transformer makers can run this test for users whenever dielec tric strength loss (in the absence of arcing) indicates moisture contami nation. However, sampling and knowledge able insjtection are highly important. Only aliout 125 parts per million of water can dissolve in askarel at room temperature. Obviously, it is possible to have water condensation on the shell or "globules" of water on the insulation surface that have not dissolved in the askarel. Since con densed moisture may not be uniformly distributed on the liquid surface, it is difficult to assay the transformer's interior condition conclusively, even when carefully drawing a surface sample of the askarel. The problem is much the same with mineral oil except water lends to settle to the bottom of the oil. Askarel, the more dense liquid, holds water on the surface. In the event, of a moisture access into a transformer, the problem of detection centers mainly on detecting the undissolved water that has accu mulated much more so than in detect ing the rise in the limited amount of dissolved water in the dielectric fluid. This rise is gradual, readily detected by dielectric strength checks, and should be considered as the "clue" to when an interior inspection is war ranted. A reliable and suitable method to detect undissolved moisture distrib uted non-uniformly over the environ ment has not been developed for either oil or askarel units. Therefore, a moisture check on the fluid, either by the Karl Fischer direct determina tion or by the indirect check on dielec tric strength, must be relied on as an indicator. This requires a periodic check at regular intervals to establish a possible upward trend. A single, one-time analysis unrecorded and un compared may not reveal the facts. The very few instances of failure of askarel transformers bring out very clearly: moisture is practically the only troublemaker--in the ab sence of arcing. However, the rare instances of failure do not justify saddling the transformer user with an onerous maintenance program that requires complicated, unnecessary testing. In stead. with knowledge and alertness to the indications, the user can zero in on the important moisture factor and thereby simplify maintenance. As the expert said, the best main tenance advice is to "Keep 'em dry, then leave 'em alone." The signifi cance of this statement has been docu mented by 30 years of field experience and by the description of askarel properties presented in this article. For year-after-year of dependable, fire-safe operation with negligible attention, water must be "designedout" or kept out by every practical means. This fact of operating life is equally important to either askarel or mineral oil transformers. By protect ing against the entry of moisture, the user solves the majority of the main tenance problems for both .types of transformers. It is important to note (see table 1, sequence A to E) that the dielectric strength of reasonably dry askarel is not impaired even though there is a characteristic increase of power fac tor and reduction of volume resis tivity. Power factor and volume re sistivity are unimportant criteria in evaluating askarel insulation. With little exception, the very few askarel failures over 30 years showed the same major fault: Improper sealing against moisture caused loss of dielectric strength and insulating values. Moisture con tamination, especially beyond 125 ppm saturation level, caused failure. Reactivating Old Transformers Perhaps some users can benefit from our own experience with a very early vintage askarel transformer. This was an expensive outdoor in stallation--where fire and explosion safety was required to protect workers as well as valuable equipment. The outdoor humidity posed a constant threat primarily because of the trans former's outdated and inefficient seal ing arrangements. Unwilling to take out the trans former for modernization, and faced with a constant pressure loss, we solved the problem by bleeding through the "air space" a very small but continuous (bubble) stream of dry nitrogen from a standard cylin der. A little more than enough nitro gen was introduced to compensate for the normal breathing of the unit due to temperature fluctuations. This "gas drying" has been going on for years; the askarel unit is still operating with the original fill of fluid; DSW 321756 4 STLCOPCB4073407