Document Ky4v2geRbOr6jL4rggR8dGXo

ENVIRONMENTALLY ACCEPTABLE INSULATING FLUIDS MAY REPLACE ASKAREL A A r^^ciu R.7'Gr BY: D. A. DUCKETT CHIEF PRODUCT ENGINEER RTE Corporation PRESENTED AT: The General Meeting of the Edison Electric Institute Transmission and Distribution Committee Minneapolis, Minnesota May 8, 1975 MONS 051475 Envlronmcntolly Acceptable Insulating Fluids Mny Replace Askarel The transformation oT electrical energy poses a fire hazard In high- rise buildings, large industrial plants, and other locations. Utilities and contractors have four choices In supplying power to these locations, none of which have been economically practical but have met the fire codes and non-flammable conditions set forth by insurance underwriters and other regulatory bodies. These four choices include: - 1. Remote transformers with long secondary runs at utilization voltage (occasionally at intermediate voltages which employ additional dry type transformers at key locations) 2. Dry type or gas cooled transformers 3. Fire-proof vaults Askarel cooled transformers The disadvantages of remote transformers are Immediately apparent. High cost and system complexity are difficult to Justify. Voltage reg ulation problems exist. Gas cooled and dry type transformers have limitations In physical size and voltage. High cost, low overload capacity, and poor reliability make these a difficult choice. Fire-proof vaults with mineral oil transformers also have high Installation costs. Space limitations tend to make this choice unaccept able for many installations. This leaves the latter choice of askarel cooled transformers. Askarel cooled transformers have the advantages of lower initial unit cost than the previously mentioned choices, higher kV and kVA ratings, and better reliabil ity, physical size, and weight. The disadvantages of askarel come from Its fire-proof characteristics. Askarel, a polychlorinated biphenyl, is extremely stable in the environment. It does not blo-degrade and builds In the food chain affecting fish and certain types of birds. This has led to criticism by environmentalists. Costly precautions must be taken to either reclaim or dispose of askarels when their useful life Is expended. Askarel Is a powerful solvent and requires special sealing and Insulation systems which add to Its cost. Under arcing conditions, askarel produces hydrogen chloride, which is corrosive and toxic, and carbon, which prevents the use of submerged fuses and switching devices. MOMS OS1476 Environmentally Acceptable Insulating Fluids Hay Replace Askarel The purpose of this paper is to acquaint you with the problems associated with askarel, and to share with you our attempts to develop a substitute insul ating fluid that is environmentally acceptable and will still do all the things necessary to provide successful transformer performance. The concern over askarel began in the mid-1960's with the search for the presence of DOT In the oceans. Refinements led to Improved test equipment and monitoring techniques. Biphenyls, closely related to DDT, were also found to be prevalent in the environment. This led to restrictions and limitations on the amounts of biphenyls that could be allowed to enter the waterways. Open and semi-open systems utilizing askarels were abandoned and their use outlawed. In Japan, during the early t970's, a food processing company was using askarel as a heat transfer fluid. A leak occurred, and many people were poisoned with excessive amounts of biphenyls. Because of this, end several other political Issues of the time, Japan outlawed the use of askarels. They have since approved silicones for use in electrical devices on the National Rail System, where a replacement was necessary. . These events led Dow Corning Into research toward developing a.silicone fluid for use as an alternate transformer coolant. Dow Corning reviewed their alternative fluid with Commonwealth Edison who suggested a joint venture with RTE, who could produce additional tests and sample transformers. During the 1960's, RTE was studying dielectric fluids for use in a new high temperature transformer. This design, the Power Seed, Is a small kVA, porcelain enclosed. Class H, direct burled transformer. It was produced in the early 70's as a pilot project for development of a total underground system and Is presently undergoing field testing at various utility sites. When approached by Commonwealth Edison and Dow Corning, RTE chose to also Include the fluid used In the Power Seed, RTEHP, and leave the door open for additional choices to provide an atmosphere for development. Fault Tests , The main question was, "How do the fluids react when subjected to a high energy arc?" A fluid which continues to burn after a catastrophic fault has occurred could expand the explosion into a building fire. A fluid which does not continue to burn would limit the explosion to the immediate vicinity of the device. Would these new fluids, mainly because of their high flash points, put themselves out after an explosion, or would they continue to burn? A series of high current fault tests were performed. These tests compared transformer oil, Silicone DC-200, RTEHP, and askarel under high current submerged arcing conditions. Approximately four gallons of each fluid were placed in separate cylindrical containers and pre-heated to 150C. Each container had fused Internal electrodes HONS OSl**1 Page Three Environmentally Acceptable Insulating Fluids Hay Replace Askarel Fault Tests (cont'd): mounted on the end of SBT Bushings. The electrodes were shaped upward with an expanding gap; this produced an effect which forced the arc upward and into the gas space, where It was present In the gasses and vaporized fluids as they were blown out of the test container (Figure 1). The container was sealed before the test to contain the hot gasses in the air space prior to the explosion. A cable was attached to the cover to prevent It from leaving the test site. Barriers (Figure 2) were placed around the test container to simulate types of wall surfaces which might collect the flaming liquids and promote burning. CONCRETE BLOCK b1 x b' STEEL PLATE b* x V To protect the surrounding environment from contamination, a nylon reinforced plastic sheet was laid down approximately 30 feet In the direction of the wind (less than 5 mph at test time). This was covered with a coating of sand to collect any unburnt fluids. After the tests were completed, the entire package was burled In a dry land fill area. HONS 051478 Page Four Environmentally Acceptable Insulat Ing Fluids Hay Replace Askarel Test Circuit The test circuit utilized Is shown In Figure 3SOURCE FIGURE 3 TEST CIRCUIT HONS 091*7* Environmentally Acceptable Insulating Fluids May Replace Askarel Comments on each test follow with the visual and electrical data summarized: Test Number One - Transformer Oil This test was expected to be violent In nature and was placed In the series as a control sample. Results met expectations. The test current applied was 4820 amperes at 4800 volts. The back-up fuse cleared the fault after 10J cycles. - The explosion was very violent with an initial fireball, orange and yellow In color, approximately 20 feet high by 15 feet in diameter. This mushroomed Into a cloud of flame and smoke approximately 55 feet high by 40 feet in diameter. The resulting smoke cloud produced was voluminous and black to dark grey In color. . The test container and surrounding area were covered with burning liquid which was manually extinguished quickly to avoid damage to the test cable and connectors. Test Number Two - Silicone Fluid DC-200 (50 CS) The test current applied was 4760 amperes at 4800 volts. The back-up fuse did not blow. The fault self-cleared after 4-i cycles; this will be discussed later. Compared to Test One, the explosion and the noise produced were mild. A fireball, orange and yellow in color, approximately 30 feet high by 15 feet in diameter was noted. The flash appeared to be very bright in comparison to the other tests. The smoke produced was white in color and of less volume than Test One. Black flakes were seen in the cloud. White particles. Identified as silica, were noted floating In the air after the explosion. Films show a flame, low In magnitude and quiet in nature, to burn for a few seconds in the test container after the explosion then self-extingulsh. Test Number Three - RTE High Temperature liquid (RTEMP) The test current applied was 4700 amperes at 4800 volts. The fault self-cleared after 4 cycles. The explosion was mild in comparison to Test One and resulted In a fireball approximately 15 feet high by 10 feet In diameter. A quantity of non-burning fluid was noted preceding the fireball upward; this can be seen falling back to earth In the films. The smoke was grey-white In color and similar in volume to the smoke cloud of Test Number Two. Noise was mild compared to Test Ono. HONS 0S1480 Envl ronnienlol ly Acceptable Insulating Fluids Hay Replace Askarel Tcs-t Humber Three (cont'd) A re-strike occurred approximately 115 cycles after the initial fault. The loss of fluid In the test container was due to the placement of the arc and the physical dimensions of the container. The fluid self-cleared and the electrodes were energized as the remaining fluid began to settle. The current tracked across the contaminated surface and was cleared by the back-up fuse. This effect was noted in the rehearsal tests for both Silicone and RTEHP and should not be considered a negative point toward the fluids, but a criticism of the physical parameters selected for the containers and electrodes. Comments were made that possibly the secondary flashover extinguished the flame; however, films show RTEHP burning in the test container after the re-strike and seif-extingulsh. Included with this report Is a series of sequential photographs for each test. The machine driven camera speed was 12 frames/11 seconds, or approximately 0.92 seconds (55 cycles) between frames. By this timing, the re-strike would occur between frame two end three of the sequence. From the remaining frames, It can be seen that the flame was present for a few seconds after the re-strike. High speed movies also verify this. Test Number Four - Askarel The askarel used In the test was Inerteen* 70-30. The test current was 1)660 amperes at 4800 volts. The back-up fuse cleared the circuit after 11-i cycles. - A fireball of bright orange flame, approximately 25 feet high by 15 feet in diameter, rose In a pitch black smoke cloud for l-i seconds after the explosion; black stringers were noted falling from the cloud. There was no fire in or around the test sample. The smoke cloud was voluminous, approximately 25 feet high by 40 feet in dlamotar, and remained In the air about 20 feet above the ground for approxi mately five to ten minutes before dissipating. The test site and equipment were covered with black fluid and the area had a very noxious odor. Dielectric tests show the fluid to retain little dielectric strength after the fault. Before the test, the fluid was good for 34 kV; after the test. It was reduced to only 7 kV. It would be an understatement to say that the observers were shocked at the poor performance of this approved "safe" fluid. Although the fluid Itself did not burn, flame was present In the smoke cloud (experts from Monsanto have Identified this as possibly "free carbon" Igniting). The dense smoke was considered a dangerous by-product. ,, *Trademark of Wcstlnghouse Electric Corporation HONS 051481 Environmentally Acceptable.Insulating Fluids Hay Replace Askarel Test Conclusions Both RTEMP o.nd Silicone DC-200 (50 CS) passed the test requirements and deserve further investigation toward approval as an alternate for the environ mentally hazardous askarels. - . In Test One, transformer oil, all liquid thrown out of the container would have been consumed by fire If It had not been extinguished. This could have led to a disastrous .but Iding fire. ' As the temperature of the liquids were below the fire point in the remaining tests, Silicone, RTEMP, and askarel,the fluids self-extinguished. This would limit the after effects of an explosion and, with reasonable precautions, would allow electrical equipment with these fluids to be mounted Inside building structures. The gasses produced by the askarel are considered extremely dangerous to the well-being of anyone In the area. Askarels are non-biodcgradable in the environment and dangerous as they build in the food chain. They were established at a time when little or no concern was felt for these objections, and It's doubtful that any testing was done to research these areas. A simple non-flammablllty test would meet the requirements for Indoor applications; historical usage has gained universal approval. Silicone fluid, although also non-biodegradable, is not known to be dangorous in the environment. The only questionable by-product of an explosion Is silica particles, of which only minimal exposure is possible. The RTE High Temperature Fluid, like transformer oil, Is biodegradable. The gasses produced during an explosion are essentially hydrocarbons. Both Silicone and RTEHP limited the duration of the fault and self-cleared; this Is a property which adds to the value of these liquids and is a desirable feature for designs requiring submerged high voltage fuses, breakers, and switching devices. Table 1 shows the fluid properties before and after the arcing tests-. This data reflects the remaining dielectric Integrity of the fluids after high current arcing and shows RTEMP to surpass all tested fluids. The electrodes in all cases showed signs of arcing along the entire length and burning at the top. This Indicates that the arcs followed theory and were present In the gasses and vaporized fluids as they were thrown upward. Approximately one gallon of fluid was left In each of the test containers after each test. HONS OS1482 Environmentally Acceptable Insulating Fluids Hay Replace Asknrel Test Conclusions ( cont'd.) The level of fault currents applied during the tests, 4,000 to 5,000 amperes, and the voltago, 4.8 kV, could be considered low under today's conditions of higher available fault energy and increasing voltage levels. These values were selected to extend the possible fault time to 8-12 cycles with available back-up fusing. Higher values of energy, l2t, could be released Into, the test containers If the current were Increased. Higher voltages might promote arc expansion to the walls of the container and other ground points. The physical placement of the arc and the moss of the fluid In the container would also effect the results. It has been discussed that this should suggest further testing to study the explosive potential of the various liquids under these varied conditions; however, the purpose of these tests was to determine the flammability of the liquids after an explosion. Both Silicone and RTEMP self-extinguished and passed the test. 2 Higher values of I t might create a greater explosion in any of the fluids but would not effect the self-extinguishing feature unless the mass of the fluid was so small as to have Its surface temperature raised to values In.excess of the fire point of the liquid. Although the temperature at the point of arcing It several thousand degrees, this can last only for a few cycles and cannot significantly effect the temperature of a container the size of even a small distribution transformer. Greater values of l2t would create a more spectacular test, but the flammability results would remain the same. MOHS 051483 TRANSFORMER OIL SILICONE FLUID DC-200 RTE HIGH TEMPERATURE FLUID (RTEMP) AS KAREL BEFORE ' TEST AFTER TEST BEFORE TEST AFTER TEST BEFORE TEST AFTER TEST BEFORE TEST AFTER TEST DIELECTRIC STRENGTH (kV) ASTM 087 32.0 20. <1 42.0 15.5 34.0 24. 35.0 7.0 1 FT (dyne/cm) ASTM D971 VISCOSITY (SSU) ASTM 088 49.0 58 @ 25C *5.5 20.8 22.2 25.5 50.0 50.0 54.5 77.4 e 25C 50 8 25C 65 e 25C 4000 e 25C 1900 8 25C 54 8 25C 66 8 25C TABLE I FLUID PROPERTIES BEFORE AND AFTER FAULT TESTS HONS 051484 S IV T S O SNOW DIELECTRIC FLUIDS - EXPLOSION TESTS NOV. 9,1974 TEST No. 3 RTE HIGH TEMP (RTEMP) TEST No. 4 HONS OS 1486 PCBIASKAREU RTE CORPORATION VlmWnltm. 'Nneorafcn HOMS OS1487 MOMS O S i*B * I HONS 0 5 1 4 *9 CHEMICAL BONDS AND GASES PRODUCCD DURING ARCING Askareis Askarcls are mixtures of polychlorinated biphenyls. Depending on the manufacturer, they may be such a mixture as Pyranol* 1470 which is 44:4:l{ chlorinated diphenyl, trichlorobenzlne and tctrachlorobcnzine, with 0.125$ by weight, amount of.tIn-totraphenyI added as a "getter" to react with the hydrochloric acid produced during aging. The material used in the tests was 70$ pentachlorobipheno) and 30$ trichlorobenzine. These have the following chemical bonds: PENTACHLOROBI PHENOL CL C CL \ / \/ cc II cc / X /\ H CH I CL C CL \/ c I c /\ c %/ c I c \ I CL *Trademark of General Electric Company HOMS 051491 TRI CHIOROP.ENZ INE H I CL C ' \\ cc II 1 cc \/ c 11 CL \ CL During breakdown, the products of decomposition are: ABSENCE OF OXYGEN Hydrochloric Acid (HCL) Free Carbon (C) PRESENCE OF OXYGEN Hydrochloric Acid (HCL) Carbon (C) Water (H.O) Carbon Monoxide (CO) Carbon Dioxide (COj) Further, with arcing In the presence of paper, phosgene gas (C CL. 0) Iso possible. HONS OS1492 Silicone DC-200 The silicone fluid used In the test was DC-200, 50 CS. This is 8 dimethyl silicone fluid which has the chemical bonds shown below: DIMETHYL SILICONE FLUIDS CH 3 CH, -- SI -- 0 *i CH 3 CH, CH 3 SI - 0 SI -- CH, i * CH3 /,, CH3 Some of,the gasses produced during arcing are: ABSENCE OF OXYGEN Silicon Dioxide (SI 0.) Hydrogen (H.) Hydrocarbons (CHn)' PRESENCE OF OXYGEN Silicon Dioxide (SI 0.) Hydrogen (H ) z Water (HO) Carbon Monoxide (CO) Carbon Dioxide (CO,) Hydrocarbons (CHnP HONS 051493 RTfMP RTEMP is essentially a mineral oil. It is refined from a paraffinic type base oil. A special two-stage hydro-treating process is used in its manufacture which results in a highly refined stable oil of food grade. It lias been used in food products and associated equipment, and would be considered safe In the environment. The chemical bonds of this type of oil would be similar to the paraffinic group shown below: TYPI CAL PARAFFI NIC TYPE OILS H-C-H I H H-C-H I H As can be seen, the paraffins re characterized by long straight chain segments with relatively few off-shoots or attachments. Under arcing conditions these chains would be broken and the products of decomposition would be those released during the breakdown of any hydro-carbon of similar structure. Shown below are some of the typical products of decomposition which would be found during this process: ABSENCE OF OXYGEN Hydrogen (H.) Carbon (C) 4 Hydrocarbons (CH0) PRESENCE Of OXYCEN Water (HjO) Carbon Monoxide (CO) Carbon Dioxide (C02) Hydrocarbons (CHp) HONS OS1*94 FLUID PROPERTIES: TRAHSFORHER OIL RTEHP SILICONE DC-200 ASKAKCL DIELECTRIC Dielectric Strength (ASTM D-877) kV (2$C-fluids as received from vendor) 31 37 34 *10 Dielectric Constant Power Factor 50C 100 C . I50C Dissipation Factor (ASTH D-150) Volume Resistivity (ASTH D-1169) Ohm-cm 2-2.5 2.2 2.74 4.5 <01 1.0 2.5 0.6 2.2 10.5 0.6 0.9 1.5 .. .. -- .0004 1.0 x 10 12 <05 1.1 x lo'3 .0002 5.6 x 1011* 03 5.0 x 10 12 THERMAL Flash Point C Fire Point C Pour Point C Thermal Conductivity 25C cal/(sec-cm-C)/cm Specific Heat (cal/gm/C) 25C Coefficient of Expans Ion(cc/cc-C) 150 162 <-57 .000318 296 321 -21 .000297 .393 .00063 .650 .0008 304 360 -55 .000360 .340 .00104 -- -37 .000262 .264 .00067 PHYSICAL Specific Gravity (ASTM D-1810) 25C Inter facial Tension (dyne/cm) Neutralization Number (mgKOH/gram) Viscosity (cent 1 stokes) 25C 50C 100C I50c .883 49. A <.02 16 8 3 "" GO GO 25-5 .011 800 150 24 8 .961 20.8 <01 50 30 16 12 1.545 50.0 <01 18 10 4 <3 HONS 051495 MATERIAL COMPATIBILITY STUDY Transformer oil, RTEMP, Silicone DC-200, and Askarei have been compared In sealed bomb accelerated aping tests. This is a standard material compatibility test at 150C for 6 weeks. The sealed bombs contain paper, epoxies t varnishes, various coated wire, samples, and the insulating fluid. Physical and electrical character istics are compared at the end of the test period. RESULTS: . TRANSFORMER OIL RTEMP SILICONE DC-200 ASKAREL INITIAL 6 WKS AGING 8150C INITIAL 6 WKS AGING ei50C INITIAL 6 WKS AGING: ei50C INITIAL 6 WKS AGING ei5oc FLUID PROPERTIES Dielectric Strength (kV) IFT (dyne/cm) NEUT NUMBER (mgKOH/gr) 37.1 48.0 .0015 21.5 40.4 .078 39.8 55.7 .011 24.7 35.0 47-7 .099 30.8 .0025 (29)* 38.5 29.0 24.2 .103 35.9 .005 432.2 .095 PAPER- Tensile Strength (kpsl) Dielectric Strength (v/ml1) 14.9 1480 11.6 1540 14.9 1480 12.0 1610 14.9 1480 12.2 1530 14.9 1480 11.7 1460 WIRE Dielectric Strength (v/ml1) 3000 3400 3000 3500 3000 3330 3000 3600 Average of best three rcadlngs(discussed further in test conclusions) MONS 051496