OIL TESTING HISTORY Foreword by Corné Dames, Associate Editor in Chief Did you choose the correct oil for your transformer application? This is the one big question I am confronted with after reading through this outstanding article written by the three authors, Lance Lewand, Eileen Finnan, and Melissa Carmine-Zajac. All of them are highly respected in the industry. Interesting to see the long road oil analysis has come and the progress that has been made in the transformer oil industry, to ensure a high-quality product fit for a specific purpose. The criteria are set out to guide us to evaluate the insulating fluid when considering a fit candidate for lifetime optimization and reliability. William Stanley’s invention of the commercial transformer in 1886 and his subsequent AC electrification of Great Barrington, Massachusetts in that same year [1], pioneered the development of commercially viable electrical distribution systems. The first transformers were dry types using insulation made of cambric cloth and other cotton materials. In order to achieve higher voltages a better insulating medium was needed, and as early as 1887 [2] paraffinic mineral oil along with cotton-based materials were used to form the insulation system of early transformers. Although plentiful, early paraffinic oils would often solidify near temperatures of 4°C or lower, causing retention of heat in the windings [3] that eventually led to failure. As a result, naphthenic oils with naturally occurring low pour points became the insulating oil of choice in the early 20th century – and is still the majority of what it is used today in most countries. Oxidation of transformer mineral insulating oils occurs over time, at a rate that is dependent on the composition of the fluid and the conditions to which it is exposed while in service, such as temperature, oxygen and catalysts. Mineral oil filled transformers in the early 20th century were free breathing, which allowed a constant oxygen supply. Coupled with the poor refining practices of the time, which was a simple distillation cut of the crude that resulted in oils that were oxidatively unstable, the mineral oils in service tended to sludge and polymerize. This often resulted in solid terminal byproducts that had to be physically removed from the transformer with shovels. This caused operational headaches for utilities in that many transformer maintenance activities were engaged in to remediate oil issues. As a result, there was a need to develop tests that could determine the quality and oxidative stability of the oil.  In 1898, ASTM was formed as a forum to help industry leaders develop test methods for a variety of products including crude oil. ASTM committee D02 (then called committee N), founded in 1904 [4], was the first committee to take on the responsibility of establishing test methods for crude oil from which transformer mineral oil is refined. For some, oil is an after-thought, just another component that comes with the transformer. But oil plays a vital role as one of two main elements of the insulation system of the transformer. The very first methods did not come into existence until the decade between 1910 and 1920 and were limited to the evaluation of mostly physical characteristics; they did not address oil quality or oxidation stability. It wasn’t until the 1940s through the 1960s that oil quality and oxidation tests were established for use. In the 1950s, the ASTM D27 Committee on Insulating Fluids was established and split out of D02, and serves as the committee that produces standards for specifying, sampling and testing of insulating fluids.   Oil serves as an important heat dissipation medium hence the makeup of the oil is crucial to proper and safe operation. Poor quality or improper application can result in a host of issues, such as corrosive sulfur attack or gasket incompatibility, many of which can result in failure. Nevertheless, the problem of oxidation of transformer oils was recognized early. In 1921, Dr. Van H. Manning authored a paper entitled “The Pioneer’s Field in Petroleum Research”. In that paper, Dr. Manning listed issues to be addressed in the petroleum industry, one of which was the prevention of oxidation of mineral oils in transformers [5]. In June 1923, the Massachusetts Institute of Technology (MIT) began a research study of the mechanism of the oxidation of mineral oils and of the action of negative catalysts (inhibitors) [6]. Inhibitors were known to bind with free radicals thus terminating the propagation process of oxidation. A year later in 1924, 177 substances were tested for anti-oxygenic protection of transformer oil, and 48 appeared to have inhibitory properties. Additional investigations were undertaken both in the United States and other parts of the world which identified other possible materials that could suppress oxidation [6]. Researchers in Russia [7,8], France [9] and England were embarking on a similar path at about the same time. However, those were not the only issues to overcome. Viscosity was a well-known factor that dictated the ability of the oil to dissipate heat from the windings and thus oils with lower viscosities became desirable. Corrosive sulfur was another issue of concern with oils at the time as it caused corrosion of the copper and silver components it was in contact with, leading to failure of the transformer. Probably one of the most overlooked issues, which is still persistent today, is the interaction of the oil with the materials of construction of transformers, including gaskets, coatings, paints and other materials. The effects were a mutual problem with the oil causing the components to degrade, which resulted in leached material from the components contaminating the oil leading to compromised chemical, electrical and physical properties. For some, oil is an after-thought, just another component that comes with the transformer, but it plays a vital role as one of two main elements of the insulation system (the other being the cellulosic insulation) of the apparatus and it further serves as an important heat dissipation medium. Hence the makeup of the oil is crucial to proper and safe operation. Poor quality or improper application can result in a host of issues, such as corrosive sulfur attack or gasket incompatibility, many of which can result in failure, and so it must be selected with care and diligence. And once selected, the oil needs to be monitored over the life of the apparatus, not only to ensure the oil quality is maintained so it can continue to serve its intended purpose, but also as a means of evaluating the operation of the transformer itself, through diagnostic tests such as dissolved gas-in-oil analysis (DGA), and the condition of the other main insulating component – paper, through such chemical markers in the oil as furanic compounds and alcohols. As a result, standards were devised and published to ensure transformer oils meet certain minimum characteristics. These standards and the tests that comprise them can be used to compare products and to evaluate the consistency of a product from year to year. Probably one of the most overlooked issues, which is still persistent today, is the interaction of the oil with the materials of construction of transformers, including gaskets, coatings, paints and other materials. One of the earliest specifications was the Canadian standard CSA-C50 [10] published in 1938, but it was not until 1961 that Doble Engineering company published a truly comprehensive standard called “Transformer Oil Purchase Specification (TOPs)” [11]. Other major specifications for new mineral insulating oils, ASTM D3487 [12] and IEC 60296 [13], were not published until the 1970s. All of these standards undergo regular revisions and are still published today and have many tests in common, but there are distinct and important differences. Thus, the user must be able to navigate these specifications, understand the significance of the tests, and decide which is most suitable for a given application. Detailed oil specifications should be part of new transformer specifications and the product delivered should be tested to verify specifications are indeed satisfied. Any purchase of oil should have the same specifications met to ensure only compatible products that meet standards are introduced into the transformer during its lifetime. A good product and a consistently refined product are of ultimate importance. Evaluating oil to these published standards can help ensure such a product is selected. Detailed oil specifications should be part of new transformer specifications and the product delivered should be tested to verify specifications are indeed satisfied. Any purchase of oil should have the same specifications met to ensure only compatible products that meet standards are introduced into the transformer during its lifetime. A good product provides the right mix of electrical characteristics required by the insulation system, the lack of corrosive sulfur (responsible for attacking the copper and silver in transformers and causing copper sulfide deposition in the paper), good oxidation characteristics so the oil lasts the lifetime of the apparatus without sludging or degrading rapidly, and appropriate flow characteristic (as defined by pour point and viscosity) for operation in the intended climate and to meet heat dissipation needs. In addition, it is important to assess the actual chemical makeup of the oil as this has an impact on compatibility with other materials of construction in the transformer. For example, if a user decides to change out a high aromatic oil with a low aromatic oil, the gaskets will contract, and leaks will develop which can be costly to remedy. Knowledge of additives is another important factor to consider and a good product should have nothing added that has not been mutually agreed on with provider and end user. Most modern specifications include a limit for furanic compounds, oil-soluble compounds used as an indirect measure of the condition of the cellulosic insulation in a transformer. 2-furfual is the main compound used for diagnostic purposes, but it can also be used in refining of transformer oil during a solvent extraction process. If not properly removed prior to use, it may lead to an incorrect evaluation of the cellulose insulation as indications of overheated or aged paper insulation in a transformer. Specifications should also contain any other specific requirements. For example, applications such as cables, bushings and some transformers, may require the use of a negative gassing tendency oil. For those that purchase transformers and other apparatus outside of their home country, knowledge of international standards and products is very important. The apparatus will contain residual oil from factory testing or depending on the size, may come fully filled. Knowing which oil product was used along with the qualities of that oil is essential in making sure no unwanted oil characteristics are transferred over to the new electrical apparatus. For transformers, approximately 7% to 10% of the factory test oil remains in the apparatus when drained after factory testing is completed, and if this retained oil is problematic, it can be enough to contaminate acceptable new oil used to fill the apparatus during commissioning. From 2001 to 2007, this very scenario played out and caused severe, to the point of failure, corrosive sulfur contamination in new apparatus around the world.  In these cases, the original test oil used at the factory was corrosive or in some cases was provided for commissioning on site as part of a turnkey project. These incidents highlight the importance of specifications containing test methods and limits that are rigorous enough to ensure suitable product is provided. Additionally, a detailed and comprehensive oil specification must be a part of the purchase process for transformers and updated on a regular basis. Oils come from different crude sources which are processed using a variety of refining methods, so the chemical makeup of the end products will differ even though all may meet the same functional specification. Identifying changes in delivered product is easier when one with consistent properties is used. Along with satisfying specifications, consistency of product should be considered. Oils come from different crude sources which are processed using a variety of refining methods, so the chemical makeup of the end products will differ even though all may meet the same functional specification. Identifying changes in delivered product is easier when one with consistent properties is used. It should also be noted that the chemical makeup of an oil can also impact gas formation under certain conditions; Dissolved Gas Analysis (DGA) is widely used as a diagnostic tool and knowing how an oil behaves in the equipment on a system from year to year can eliminate an unknown when diagnosing potential issues. Specification testing of oil products can be undertaken periodically to monitor consistency. Oil surveys of transformer oils are available where oil products are compared year-to-year and results tracked. Doble Engineering Company has been performing such surveys either semi-annually or annually since 1953. In the early days, reports were specific for oxidation tests only, as that was the main issue at that time. Current testing is much more comprehensive with 30+ tests being incorporated and provides the user with a detailed analysis for each of the submitted oils. Oils are solicited from refiners around the world and then tested to Doble TOPS and ASTM D3487 specifications and the data along with the analysis of that data is then published. Utilities have used the report as the primary selection criteria of the oils to be considered for purchase. In conclusion, insulating mineral oils should be selected on a quality basis, the ability to withstand oxidation, the characteristics which are needed and overall consistency of a product from year to year. Detailed oil and transformer specifications should be used to ensure suitable quality product is provided. References [1] Laurence A. Hawkins, William Stanley (1858 – 1916) – His Life and Work, The Newcomen Society in North America, New York, 1951 [2] H.R. Sheppard, “Century of Progress in Electrical Insulation 1886 – 1986,” IEEE Electrical Insulation Magazine, Sept. 1986, Vol. 2, Issue 5, pp. 20-30 [3] George A. Von Fuchs, “Performance of Inhibited Transformer Oils,” ASTM Symposium on Insulating Oil, STP95-EB/Oct, 1949 [4] G. Totten, “A Timeline of Highlights from the Histories on ASTM Committee D02 and the Petroleum Industry,” in ASTM Standardization News, June 2004 [5] Van H. Manning, “The Pioneer’s Field in Petroleum Research,” Petroleum Section, American Chemical Society, 1921 [6] “Section 4, Oxidation Inhibitors for Transformer Oils,” in Reference Book on Insulating Liquids, Gases and Materials, A Project of the Doble Client Committee on Insulating Fluids, PN 500-0333, 72A-1971-01 Rev. B 7/04, Watertown, MA [7] D3. B. G. Tychinin and N.A. Butkov, “Neftyanoe I Slantsevoe Khozyaistovo, No. 1”, 1925 [8] D4. B. G. Tychinin and N.A. Butkov, “Neftyanoe I Slantsevoe Khozyaistovo, No. 8”, 1924 [9] D5. C. Moureau and C. Dufraisse, “Sur l’autoxidaton: les antioxigènes,” Compes Rendus, Vol. 174, p. 258, 1922 [10] CSA-C50, “Mineral Insulating Oil, Electrical for Transformers and Switches,” Canadian National Standard, October, 2014 [11] “Transformer Oil Purchase Specification,” Doble Engineering Co, Marlborough, MA, March 2017 [12] ASTM D3487, “Standard Specification for Mineral Insulating Oil used in Electrical Apparatus,” ASTM, West Conshohocken, PA, 2016 [13] IEC 60296, Edition 5, “Fluids for Electrotechnical Application – Mineral Insulating Oils for Electric Apparatus,” International Electrotechnical Commission, Geneva 20, Switzerland, 2020 Lance R. Lewand Lance R. Lewand received his Bachelor of Science degree from St. Mary’s College of Maryland. He is actively involved in professional organizations including the American Chemical Society. He is a representative of the U.S. National Committee for TC10 of the International Electrotechnical Commission (IEC) and ISO TC28, ASTM D-27 since 1989, Chair of ASTM Committee D-27, sub-committee chair 06 on Chemical Tests, secretary of the Doble Committee on Insulating Materials, and a recipient of the ASTM Award of Merit for Committee D-27. Mr. Lewand is the Technical Director for the Doble Insulating Materials Laboratory. Since joining Doble in 1992, he has published over 75 technical papers pertaining to testing and sampling of electrical insulating materials and laboratory diagnostics. Eileen Finnan Eileen Finnan is currently Director of Field Services, part of the Global Professional Services Group at Doble Engineering Company, providing operation management of field testing and remote services to clients. She also participates in cross-functional teams to deliver in-depth condition assessments of critical electrical assets and large-scale fleet assessment services. Her previous roles include Manager at Doble Engineering Material Laboratory which specializes in routine and specialized testing on insulating fluids from electric power equipment. Eileen received her Bachelor of Science degree in Physics and Chemistry from Trinity College in Dublin, Ireland. Share this article

SPECIAL TRANSFORMER TEMPERATURE RISE TESTING

The split-winding arrangement in core-form power transformers requires special thermal design considerations. Presently, industry standards do not provide significant detail on the temperature rise test for these unique arrangements. This paper will demonstrate how to supplement the industry standards for this special transformer construction and some common problems that arise in practice. Introduction
In core-form transformers with a concentric winding arrangement, two or more separate winding sections may be situated one above the other (axially). This unique arrangement is commonly referred to as “axial split-windings” or “axially-stacked windings”. The split-winding arrangement is most commonly used in three-winding rectifier transformers and three-winding unit auxiliary transformers (UAT). Three-winding transformers with axial split-windings are unique because the two axial sections are normally associated with two different terminals. Therefore, unique combinations of varying voltage, impedance and active power can be present, which requires special consideration.
In addition to the various possible terminal characteristics described above; there exists a fundamental difference in the temperature rise distribution model when compared to the traditional full-axial height winding arrangement because the relative top-oil, average-oil and bottom-oil temperature rises are not the same for all windings.
It is therefore necessary to understand the differences in the two temperature rise distribution models and then be able to refine the relative oil temperature rises for each winding in order to appropriately determine the average winding temperature rise and winding hottest-spot temperature rise for each winding.
Presently, industry standards do not explain the modifications that are needed to the temperature rise test for these special winding arrangements. This paper will demonstrate how to supplement the industry standards for this unique transformer construction and some common problems that arise in practice. An application for the axial split-winding arrangement may be to serve large unit auxiliary loads within one three-winding transformer at a generation plant, as opposed to using two separate two-winding transformers.
Winding Arrangement
The traditional and most common winding arrangement for core-form transformers includes an arrangement in which the windings are approximately the same height and are arranged concentrically over one another (radially). Figure 1 depicts the traditional transformer winding arrangement for a three-winding transformer with the primary (designated as “HV”) placed concentrically in-between two secondary windings, “XV” and “YV”. The axial split-winding arrangement includes two or more separate winding sections that are situated one above the other axially. An application for the axial split-winding arrangement may be to serve large unit auxiliary loads within one three-winding transformer at a generation plant, as opposed to using two separate two-winding transformers. The application could allow interconnection of two separate auxiliary buses and possibly with secondary voltages and active power ratings at different levels [1]. Figure 2 depicts an axial split-winding arrangement with one set of secondary windings designated as “YV” placed axially above the other set of secondary “XV” windings.

Figure 1. (left) Traditional transformer winding arrangement    Figure 2. (right) Axial split-winding transformer winding arrangement
 
The importance in noting the difference between the axial split-winding arrangement and the traditional concentric winding arrangement is that the transformer temperature rise distribution is appreciably different between the two arrangements, and therefore greater attention is required in determining the oil temperatures. Namely, in the axial split-winding winding arrangement, the “YV” winding from Figure 2 should not use the average-oil temperature rise of the tank oil, rather it is necessary to determine the average-oil temperature rise for the “YV” winding separately from the tank average-oil rise.
Transformer Temperature Rise Distribution
In the traditional transformer temperature rise distribution model, the winding heights are assumed to be uniformly distributed within the transformer core window. This assumption yields additional assumptions such as the top-oil temperature rise is adjacent to the top of the windings and, similarly, the bottom-oil temperature rise is adjacent to the bottom of the windings. In other words, the temperature rise of the oil inside the windings is assumed to increase linearly with the height of the windings. Additionally, the heat generated from the windings, in the form of losses, is assumed to transfer (on average) to the adjacent oil in proportion to the corresponding winding height. In graphical form, these assumptions result in two parallel lines, with one representing the winding temperature rise relative to height and the other representing the oil temperature rise relative to height. The difference between the two parallel lines is the temperature drop between the winding and the adjacent oil, which is commonly referred to as the average winding gradient.
In the traditional transformer temperature rise distribution model, the average-oil temperature rise (Figure 3 – Point B) is determined by the arithmetic mean of the top-oil temperature rise (Figure 3 – Point A) and bottom-oil temperature rise (Figure 3 – Point C). Also, in the traditional model, the average winding temperature rise (Figure 3 – Point B’) is the sum of the average-oil temperature rise and the average winding gradient. Point K in Figure 3 is the winding hottest-spot (“hotspot”) temperature rise and is determined by the sum of the top-oil temperature rise and hotspot gradient, which is determined by a hotspot factor and the average winding gradient. The hotspot factor is a dimensionless quantity greater than 1.0 and is individually assigned to each winding based on a combination of mathematical analysis and empirical test results from direct measurement [2] .
Figure 3. Traditional transformer temperature rise distribution model This stable reference temperature for each axial split-winding should follow the method for determining the oil temperature rises according to the axial split-winding temperature rise distribution model. A figure similar to Figure 3 can be found in International Standard IEC 60076-2 [3].
The IEC standard uses similar points to help illustrate the terms for bottom-oil, average-oil, and top-oil temperature rise; however, it defines the vertical axis of the chart as “Winding Height (%).”
Although the IEC standard associates the 0%, 50%, and 100% of the winding height with the location of the bottom-oil temperature rise, average-oil temperature rise, and top-oil temperature rise of the tank (respectively), the differentiation in the vertical axis is useful for the axial split-winding transformer temperature rise distribution model.
In the axial split-winding transformer temperature rise distribution model, one of the main differences is the oil temperature rise to be used to determine the average winding rise. It should not use the average-oil temperature rise (Figure 4 – Point B) of the tank oil, rather it is required to determine the average-oil temperature rise across the “YV” winding, which is Point D in Figure 4. Point D is the arithmetic mean of the oil rises from points A and B of Figure 4. Figure 4. Axial split-winding temperature rise distribution model for YV winding
The “XV” winding in Figure 4 will require similar adjustments to the temperature rise distribution model. The “HV” winding in Figure 4 uses the traditional temperature rise distribution model because the windings are not stacked axially and are full height.
Corrections to the Temperature Rise Test
The purpose of the temperature rise test is to establish the top-oil temperature rise, average winding temperature rises, and winding hottest-spot temperature rises. The temperature rise test is primarily composed of four components. The first component is the cold-resistance measurements of the windings. Next is the total loss injection for determination of oil temperature rise. Special attention is given to each component of the temperature rise test for transformers with axial split-windings, but this paper will focus on the cold-resistance measurements and total loss injection because the principles described can easily be applied to the other two remaining components.
Cold-Resistance Measurements
The industry standards have very definite criteria to minimize recording errors in the cold-resistance measurement [4]. One of the major prerequisites is the determination of the cold winding temperature since resistance varies proportionately with temperature. Cold resistance measurements require the windings and oil temperature to be stable. This stable reference temperature for each axial split-winding should follow the method for determining the oil temperature rises according to the axial split-winding temperature rise distribution model.
Total-Loss Injection
The most common method for temperature rise testing power transformers is to utilize the short-circuit method, whereby the load is simulated by the effect of short-circuit current. In the case of total-loss injection, the primary goal is to simulate and sustain total-losses on the transformer that coincide with previously measured no-load and load losses for a given active power rating for the purpose of determining the steady-state oil temperature rises.
One difficulty that can occur during the total-loss injection for three-winding transformers, which may also be prevalent during the rated winding-current injection portion of the temperature rise test, is that differences in impedances and active power ratings can result in under-loading or over-loading of certain terminals under test. Figure 5 helps illustrate the equivalent three-winding impedance network and can be derived from Equations 1 through 3. Equation 4 and Equation 5 can be used to solve for the load-current division if both the “XV” and “YV” terminals were shorted during the short-circuit method of the temperature rise test (i.e. a short across point “X” and point “Y” in Figure 5).
Figure 5. Equivalent three-winding impedance network

This can present difficulties in achieving total-loss injection because the over-loading of one set of terminals may be significant and result in exceeding recommended temperature rise limits or current ratings of ancillary equipment. Therefore, it is recommended that the details of the temperature rise test for transformers that fall within this scenario be presented and agreed to during the tender stage. The transformer manufacturer may present recommendations to inject less than 100% of the total-losses and then apply corrections to the oil temperature rises as outlined in the industry standards [3], [4]. It should be noted that the industry standards typically have a minimum percentage of the total-losses for which the temperature correction equations are valid. Due to the transformer impedance requirements or active power ratings, it is possible that the required total-losses fall outside of the minimum percentage as indicated in the industry standard.
Alternatively, it may be of interest to the end user to meet and not deviate from the industry standards. Therefore, it may be recommended to have the transformer manufacturer design the transformer windings and ancillary equipment to withstand any resulting over-loads, which result from maintaining at least the minimum percentage of the total-losses as indicated in the industry standards. It should be noted that the industry standards also have upper ranges for the percentage of total-losses and rated winding current for which oil temperature rise correction calculations can be made; however, the IEEE and IEC standards are not in agreement in regard to the allowable ranges. Due to the transformer impedance requirements or active power ratings, it is possible that the required total-losses fall outside of the minimum percentage as indicated in the industry standard. There are additional considerations that may need to be considered in determining the required total-loss for unique three-winding transformers such as axial split-winding transformers. The derivation of the load-current division from Equation 4 and Equation 5 and the use of the measured impedances and losses for each two-winding combination can result in some inaccuracies in large transformers where eddy losses in windings and stray flux losses in structural components are considerable. Additional corrections to determine the three-winding (combined) total-losses may be useful and required [5].  
After the top-oil temperature rise has stabilized during the total-loss injection, according to the requirements in industry standards, then the short-circuit current should be reduced to the rated winding currents. The instant where the total-loss injection transitions to the rated winding-current injection is commonly referred to as “cutback.” The oil temperatures should be recorded for the total-loss injection immediately before the cutback. The recorded oil temperatures should be categorized according to the axial split-winding temperature rise distribution model for windings that are axially-stacked. These categorized oil temperature rises, from the moment of cutback, will be used to determine the final average winding temperature rises and winding hottest-spot temperature rises after rated winding-current injection and hot-resistance measurements are completed. Other Special Thermal Considerations
The axial split-winding example presented in this paper illustrated two axially-stacked sections divided into two equal height winding segments. In practice, other winding constructions are possible, such as two unequal height winding segments due to differences in voltage or impedance requirements. Unequal height winding segments present additional challenges in determining the bottom-oil, average-oil, and top-oil temperature rises for each axially-stacked winding.
It may also be advisable to install temporary or permanent temperature measurement probes at the winding heights that coincide with the tank oil adjacent to the bottom, middle, and/or top of the axial split-windings. Additionally, a different temperature rise calculation method can be applied based on the proposed winding heights of each axial split-winding. It is possible that one of the axial split-windings will be under-loaded while the other over-loaded during temperature rise testing. If direct measurement is used to determine the winding hottest-spot temperature rise during the temperature rise test, and there exists unequal loading of the axial split-windings, then the direct measurement may not result in meaningful winding hottest-spot temperature rises at rated load for all of the windings. It should be noted that this scenario exists due to the short-circuit temperature rise test method; however, it is not a limitation that necessarily exists in operation. For reasons already explained in this paper, it is possible that one of the axial split-windings will be under-loaded while the other over-loaded during temperature rise testing, which means that a more refined calculation may be necessary to determine the calculated winding hottest-spot temperature rise from the direct temperature measurements.
Due to the difficulty in obtaining a correct measurement at full load and total-losses, it may not be advisable to use the direct measurement to guarantee the winding hottest-spot temperature rise. Another option may be to determine the hotspot factor from the direct temperature measurement [2] for each axial split-winding at the presumed loading based on the load-current division as determined from Equation 4 and Equation 5. Once the hotspot factor is determined for the applied test current then the temperature corrections based on the industry standards can be applied to the oil temperature rises and winding temperature rises. The hotspot factor can then be applied to the corrected winding gradient and then added to the top-oil temperature rise for each axial split-winding. When determining the hotspot factor, it may be necessary to review the variation in the hotspot factor at different percentages of rated load [6].
Conclusion
Transformers with split axial-windings present unique challenges when performing temperature rise testing. The industry standards generally have enough provisions to determine the oil and winding temperature rises for split axial-windings if some discretion is used in applying various corrections as presented in this paper. This paper provided some considerations for the temperature rise distribution model and temperature rise test via the short-circuit method. This paper emphasizes the necessity to consider transformers with axial split-windings as a unique design, which requires special involvement between the buyer and transformer manufacturer during the tender stage.
 
References

IEEE Guide C57.116, IEEE Guide for Transformers Directly Connected to Generators, IEEE, 3 Park Avenue, New York, NY 10016, USA

IEEE Standard 1538, IEEE Guide for Determination of Maximum Winding Temperature Rise in Liquid-Filled Transformers, IEEE, 3 Park Avenue, New York, NY 10016, USA

IEC 60076-2, Power transformers – Part 2: Temperature rise for liquid-immersed transformers, International Electrotechnical Commission, 3, rue de Varembé, PO Box 131, CH-1211 Geneva 20, Switzerland

IEEE Standard C57.12.90, IEEE Standard Test Code for Liquid-Immersed Distribution, Power, and Regulating Transformers, IEEE, 3 Park Avenue, New York, NY 10016, USA

IEC 60076-8, Power transformers – Part 8: Application guide, International Electrotechnical Commission, 3, rue de Varembé, PO Box 131, CH-1211 Geneva 20, Switzerland

H. Nordman and O. Takala, “Transformer Loadability Based on Directly Measured Hot-Spot Temperature and Loss and Load Current Correction Exponents,” CIGRE Report No. A2_307-2010

Jason Varnell
Jason Varnell is a member of Doble Global Power Services, employed as a Principal Transformer Engineer working on projects that include factory inspections, condition assessment, design reviews, failure analysis and other general consulting projects. He has over 10 years of experience, having most recently served as Lead Design Engineer at SPX Transformer Solutions in Goldsboro, NC where he managed an electrical design group. He has extensive expertise in core-form transformer design and construction having personally designed several hundred power transformers. An active member of IEEE Transformer Committee, Jason has contributed to the revision or creation of a significant number of IEEE transformer standards. Jason Varnell received a bachelor’s degree in Electrical Engineering (Honors) from the North Carolina State University with an emphasis in Power Systems and Power System Protection. Share this article

This article was originally published in the March 2021 issue of the From Specification to Commissioning magazine.

View Magazine