Document ompvovODg65jKY5d61j9Mpz9X

Considerations for Evaluating Transformer Askarel By P. G. Benignus, Manager, Technical Service, Aviation and Special Fluids, A/onsart/o Chemical Co., St. Louis 66, Missouri. All photos courtesy Monsanto Chemical Co. Few electrical experts are also c hemiBts. Perhaps that's why so many transformer operators (and buyers) make the mistake of thinking trans former "oil" and "askarel" are alike. They are not--the differences in these two kinds of liquid insulation go be yond the obvious difference that oil burns and askarel does not. The dif ferences also extend to their insulating power, operating characteristics, and manner of maintenance inspection. Due to tiie mistaken belief that trans former askarel and mineral oil are similar, there has been a great deal of confusion concerning maintenance procedures and operating characteris tics of the two fluids. To put an end to misconceptions, our firm is printing an easy-to-follow maintenance guide based on evaluation experience with askarel over 80 years of service. The service history study shows that askarel users generally are confused over several major points; these are discussed here in the hope that they will benefit all operators of askarel transformers. Some of the earliest askarel trans formers were first made around 1930. Many of these are still in service. After so many years of use, the out side appearance of some of these units would not warrant a prize, but their interiors remain clean. The original fill of askarel insulation is still in use and there has been little or no main tenance. They are giving the 30 years of trouble-free service life promised by the manufacturer. However, as with any type of equip ment, over the past quarter century there have been a few instances of unexpected failure. In every case, in vestigation proved that the breakdown could have been avoided. A review of certain facts of opera tion and thorough inspection, without exception in the cases examined, showed the reason for every askarel transformer failure to be the same: the entrance of moisture. However, it is obvious that any transformer can fail from inadequate design, over voltage, damaged insulation, etc. Since askarel (and mineral oil type) transformers are often selected for use in damp areas, sometimes di rectly exposed to the elements, and frequently operated where flooding may be expected--the need to exclude moisture entry calls for optimum transformer design-engineering to assure dryness. Monsanto's interest in transformer askarel is twofold: we manufacture and supply the dielectric fluid and we use a large number of askarel trans formers in dozens of plant locations to benefit from their dependability, economy, and fire s&iety. Since we neither design nor build transformers, specific design detail on new equipment is left to experts in the field. But, the general facts relating to the moisture problem must be con sidered from the viewpoint of the maintenance engineer. There are many thousands of askarel transform ers now installed, many only five or less years old. Recently, an authority on askarel transformers was asked what he con sidered to be the most important point of maintenance. His reply: "Keep 'em dry, then leave 'em alone." Keeping the transformer interior dry is indeed the watchword and not merely an oversimplification of fact. If steps are taken to keep moisture out of askarel, the liquid insulation will undergo no change and the trans former will operate trouble-free for 30 years or more. Compaction of dsfcarof and M/norat Oil Transformer users who are not aware of the marked difference be tween askarel and mineral oil have a tendency to evaluate askarel liquid insulation in terms of the more fa miliar properties of mineral oil. This is the cause of much confusion. The "check point" values of oil and askarel are different; the interpretation of the test values is different. The previously mentioned "Maintenance Guide" at tempts to end the confusion by point ing out the difference in operating characteristics and check tests of the two fluids. However, such literature is necessarily more directive than ex planatory. With a minimum of "rea sons why" the askarel maintenance guide directs the engineer to make just three simple visual inspections of the askarel fluid and determine its dielectric strength. Moisture deter mination by the Karl Fischer method is suggested only if the dielectric strength is found to be less than 20 to 25 KV at 25C. This simplified check ing, which can be easily carried out by any user, comes remarkably close to "leaving 'em alone." Unlike oil. periodic power factor and volume re sistivity determinations are not needed for askarel check testing, and if run. the values obtained are completely un related to what these values would mean when applied to mineral oil. The reason is: oil and askarel are com pletely different chemical substances. The following explains why power factor and resistivity tests are relatively unimportant, while dielectric strength is most important--in inspecting and judging askarel insulalion. evaluating Power Factor The power factor of freshly made Reprinted from Insulation, October 1962. Copyright 1962, Lake Publishing Corporation, Libertyville, Illinois 1 ^ in TOWOLDMONOQ58921 transformer askarel is close to 0.1 per cent at 100C and 1 kc (the condi tions of measurement in our quality control laboratories). This same value level, or even slightly lower, is found when accurately measured in the field, usually at 20C and 60 cycles. This initial low power factor of new askarel is essentially the same level normally found in newly-made, dielec tric grade mineral oils. Such striking similarity of values in both types of liquid insulation misleads many users into assuming that the power factor of these two entirely different fluids should run parallel throughout their respective service lives. But this does not happen. In service, the power fac tor of askarel goes up to a greater extent than that of mineral insulating oil--but the rise has little or no signifi cance. The prime reason why the power factor of askarel and oil do not go hand-in-hand is that unlike mineral oil, askarel is a relatively polar me dium. For example, the dielectric con stant of transformer askarel is about 4, or nearly twice as high as the dielectric constant of mineral oil. In finitesimal amounts of ionic conduct ing impurities in askarel can markedly increase its power factor. The same impurity may not affect the power factor of mineral oil. But despite this sensitivity, askarel's insulating value is unimpaired. Experts know that to correctly measure the low power factor of new askarel requires exceptional cleanli ness of the test cell and strict adher ence to the proper techniques (ASTM method D-924). One leading counselor correctly states that the critical technique re quired to accurately measure the power factor of askarel is impractical for field use. Experience shows that the unskilled or not-too-careful at tempts to determine the power factor of new askarel in the field lead to unreliable results. The testing should he done with the greatest care, exactly as specified by either the ASTM method (or by a similar method ology) to eliminate trace contamina tion and must be carried out in a well-equipped laboratory. With a power factor so sensitive to traces of contamination, it stands to reason that when the askarel is put into a new transformer, the power fac tor of the fluid can be expected to increase immediately on contacting the insulation which contains traces of soluble ionizable compounds. Vari ous transformer manufacturers dictate that the power factor of askarel in a new transformer should not exceed two percent at 20C and t>0 or 70 nycles. While this arbitrary limit i> about 10 times as high as the power factor of newly made askarel, consid erable care in check testing is still required to obtain an approximately correct power factor reading within even this higher level limitation. Authorities with years of experi ence with askarel agree that after years of continuous service, the askarel's power factor will be high, as much as 20 to 80 percent (or morel at 100C and 60 cycles, or about onetenth of this if measured at 20C and 60 cycles. The majority of electrical engi neers and maintenance operators have found that a high power factor in askarel is not important, however much alarm such a high value would warrant if found in mineral oil. The only qualification to be added is this: while high power factor in itself is unimportant, some of the possible causes of high power factor may war rant investigation. However, in the absence of other test data that can indicate the advisability of thorougli inspection, relatively high power fac tor is to be expected. On the other hand, a high power 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. Vofumt Retlttlvlty Rtqulrtmtnft The resistivity of askarel in prop erly constructed new or rebuilt trans formers will not be significantly lower than the 100 x 108 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 108 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. 0^09396 TOWOLDMONOQ58922 no resistivity measurement is even milt'd tor it) the maintenance guide. Dfalactrfc Sfrtngth It th Significant Checkpoint The dielectric strength of askarel is most important. This property can be easily measured by the user or by any competent electrical service shop or nearby utility. Freshly made askarel shows a minimum dielectric strength of 35 kv at 25C (ASTM D-901, D-877). If this value falls below 20 or 25 kv in un operating transformer, the fluid in all probability has been contaminated by water and should be upgraded by lilt ration. Unless excess water enters the trans former, or the unit has arced signifi cantly, the dielectric strength of the askarel should remain above 25 kv. Contaminants that affect the power factor and resistivity of askarel seem to have very little influence on its dielectric strength. Therefore, in the absence of strong arcing, a low dielectric strength read ing calls for inspection of the trans former for excess moisture. The amount of water in the askarel can be determined by using the Karl Fischer method, ASTM I)-1533. But, before outlining the impor tance of moisture, let's consider the inherent thermal and chemical sta bility of askarel. Both are outstand ing. There is nothing the user can do that will alter these fundamental char acteristics. It is the responsibility of the supplier to provide transformer askarel with the expected heat and chemical stability. Askarel liquid insulation, as com monly used, consists of a mixture of chlorinated benzene and chlorinated biphenyl. In both compounds, the highly stable electrical-grade end product is made by chlorine substitu tion in the place of hydrogen on the carbon in the aromatic ring. In the manufacturing process, an interme diate and less stable chlorination product can form by addition of chlorine to hydrogen on the carbon which leads to an aliphatic ring struc ture. It is important to note that the power factor, volume resistivity, and other electrical properties of this less Figure 2, askarel is subjected to a temperature oj 210C for periods of 8 or 16 hours in this apparatus at Monsanto's laboratories in tests to determine its thermal stability. Figure 3, an internal section view of an askarel transformer, shut down for relocation after 18 years of successful operation. The area directly under the cover is shown since, this would reveal surface corrosion at the liquid level if any existed. stable form are just as good initially as these same properties of the more stable form when impurities are re moved by the final, highly efficient earth refinement step. Unless this highly obscure chemistry is known and controlled, a deficiency in inher ent stability would not show up until the askarel had undergone sufficient stresses under normal relatively long use. It would, in fact, not be detected unless the thermal and chemical sta bility are carefully probed when the material is made. Recognizing the responsibility of providing askarel users with insula tion possessing the required maximum thermal and chemical stability, the manufacturer should met iculously check thermal and chemical stability on every lot manufactured so that the askarel meets the strictest of specifica tions. The test for thermal stability consists of heating the askarel at 210C for 16 hours while a stream of purified air is bubbled through the fluid. The chloride ion so liberated in 0509397 3 TOWOLDMONOQ58923 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 (both procedures are documented at ASTM and elsewhere). We consider it essential lo 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 be sure that its chemical and ther mal stability conform with the requiremenls prescribed bv 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 be 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. Mektur*: Atkartl't Grtatcsf 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 bo 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 inspection are highly important. Only about 125 parts per million of water can dissolve in askarel at mom 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 tends lo settle to the bottom ol 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 he "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. jtMctivotlitf O/d TrontferiRtri 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. 0509390 4 Tab/a 7--Comparison of Values Found 7n Askaral Insulation Attar Various Stagas at Use A B CDK F Properties Color New askarel Light, 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) 1 ,ight straw Non-arced hut slightly contaminated askarel in im properly* built units, new or slightly used. Can huve foreign shades, blue, green, red cast. Same lluid in column D after relining with earth and liltcring. Foreign shades remain showing extraction of oil soluble color. Heavily arced askarel. < If tho very few askarel failures, t lie majorit > wen* due to excess water-- remainder due to internal electrical discrepancy Black Condition Moisture. 25C Clear, free from particles Clear, practically Clear, trace of free from par- particles tides <50 ppm <50 ppm 10-70 ppm, all dissolved water Clear, slight amount of particles 20-100 ppm. all dissolved water Clear, free from < 'oiit ains cur particles ium pari icles <30 ppm near or aln\e saturation lex el. 125 ppm. pal lially i mil is. solved water Dielectric strengt h, 25C, 0.1-in. gap 55 kv min 55-18 kv 55-45 kv Volume resistivity, 100C, 500 V, d-r. 0.1 inch gap 100 x 10 ohmcm min, usually 500-1,500 20 x 10 ohm-cm 15-400 x 10 min, usually at ohm-cm least 100 x 10s 50-42 kv about 5x10 ohm-cm 10-15 kv at least 1,500 x 10 ohm-cm 5-15 k\ alsml 3 x I0U ohm-rni Power Factor: 100BC, 60 cycles 25"C, 60 cycles 2-5% 0.05-0.1% 10-25% 0.5-2% 10 to near 80% 0.5 to near 10% 30 to 100% about 7 to over 15% 0.1-0.2% 100% at least 25% A properly built unit is defined as one in which the materials of construction are chemically and electrically compatible with askarel. Out of too many transformers to count over the years, Monsanto plants had two that failed after 13 and 15 years of service, respectively. These were also inadequately sealed--with cork gaskets that had deteriorated after many years of full outdoor ex posure to the elements. Another fac tor--these units were not required to be constantly on line; their inter mittent operation intensified breath ing and moisture condensation. The amount of water that entered and accumulated would have caused fail ure in any type of transformer. These units have since been rebuilt and "modernized" for outdoor service, simply by a welded sealing of top cover, fins, instrument connections, and all other fittings. A blanket of dTy nitrogen over the askarel has also been provided and the units are fit ted with pressure relief devices and pressure gauges to check on pressure fluctuations. New Transformers--"Vaccinated" Against Malstura Transformer makers, aware of the importance of keeping out moisture, now usually offer welded construction at no increased cost over gasketed construction. Where gasket seals are needed for hand-hole openings, they are designed with a recessed fitting to allow only minimum contact with the askarel and with the elements. For gasketing, instead of plain and coarse grain cork--fine grain cork impregnated with nitrile rubber makes the best material. Nitrile rubber alone has been used, but this polymer is slowly eroded over the years by askarel. There are, however, a num ber of askarel resistant polymers; these include "Teflon," "Viton," and most silicone "rubbers." While rela tively costly, these materials warrant consideration for their long term use, at least for the smaller size gaskets. Overseas, a well-known conserva tor construction design is popular: this completely fills the transformers with fluid and thereby eliminates any "air space." This construction greatly minimizes moisture entering units even when located in damp areas. On indoor installations, normal gasketing of the transformer is, of course, entirely satisfactory unless there is excessive dampness in the en vironment and wide temperature fluctuations that accentuate the breath ing. If the user cannot be certain of a dry atmosphere, it is wise to consider fitting the unit with a dehumidifying breather. As an example, highly reliable askarel units are used in transformerrectifier power supply sets, for mili tary communication equipment (trop ospheric scatter) in remote areas. These are kept dry by channeling their breathing through a canister of silica-gel attached to the side of the transformer. The canister is fitted 0509399 with a color indicator plug that warns of moisture pickup in the canister hy a color change from blue to pink. A book entitled "Activated Alumina Maintenance Program" (1955, Alu minum Company of America, Pitts burgh, Penn.), describes procedures to keep mineral oil transformers in good condition and prevent moisture pickup by using a drying-breather containing activated alumina. This drying arrangement works equally well for askarel transformers. Recognizing the moisture problem in mineral oil filled transformers, Westinghouse has developed a small cone that fits onto the hand-hole open ing. Moisture vapor condenses on the inside surface of the cone, trickles down into a trough at the base, then flows away through a small tube lead ing to a container attached to the side of the transformer. This gets us into the area of gadgets or mechanical devices. Consequently, it is important to remember that on new units, the transformer makers know the fundamental and proper de sign approaches required to keep out moisture in transformers. On existing equipment, they can suggest a variety of additional mechanical steps if such guidance is requested. When buying equipment, the user should specify whether new units arc for outdoor installation or will he used in a high moisture area. If so, he should assume any slightly in creased cost for designed-in moisture protection. It will more than pay for itself because askarel units that are kept free of moisture require no other maintenance. Up until now. moisture protective devices have received little attention: doubtless because even without them, the percentage of askarel transformer failures has been very low, regardlosof location or use. To avoid even these few possible failures, Monsanto felt it worthwhile to print an easy-to-follow, easy-lopractice Maintenance Guide. It is available on request. It shows how "check-ups" on the condition of askarel insulation can he done simply and practically without extraneous, elaborate, and meaningless testing. However, since the guide describes such maintenance in a directive sense with only barest explanation, this article provides the interpretive back ground on the significance of the various tests for transformer askarel. Distributed through the courtesy oj MONSANTO CHEMICAL COMPANY 800 North Lindbergh St. Louis 66, Missouri 0609*00