Lehigh University Lehigh Preserve Theses and Dissertations 1-1-1980 Overloading power transformers utilizing projected daily ambient temperatures. Daniel Chaply Follow this and additional works at: http://preserve.lehigh.edu/etd Part of the Electrical and Computer Engineering Commons Recommended Citation Chaply, Daniel, "Overloading power transformers utilizing projected daily ambient temperatures." (1980). Theses and Dissertations. Paper 1822. This Thesis is brought to you for free and open access by Lehigh Preserve. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Lehigh Preserve. For more information, please contact preserve@lehigh.edu. OVERLOADING POWER TRANSFORMERS UTILIZING PROJECTED DAILY AMBIENT TEMPERATURES By Daniel Chaply A Thesis Presented to the Graduate Faculty of Lehigh University in Candidacy for the Degree of Master of Science in Electrical Engineering Lehigh University 1980 ProQuest Number: EP76094 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest ProQuest EP76094 Published by ProQuest LLC (2015). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Certificate of Approval This thesis is accepted and approved in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering. (Date) Professor in Charge Chairman of Department XI Acknowledgements The author expresses his appreciation to the Pennsylvania Power and Light Company for providing the opportunity to complete this thesis through its Education Assistance Program and to Mr. J. K. Redmon for his direction and guidance as advisor to this thesis. 111 Table of Contents Page Abstract 1 Chapter 1 Introduction 4 Chapter 2 Background 7 Chapter 3 Insulation Life 10 Chapter 4 Environmental Factors and Effects 15 Chapter 5 Overload Methodologies 22 Chapter 6 Ambient Temperatures and Load Profiles 40 Chapter 7 Results and Discussions 46 Chapter 8 Dielectric Strength of Conductors in Contact With Hot Spots 90 Chapter 9. Summary 128 Chapter 10 Conclusions 138 141 Bibliography Appendix A Listing of Transformers Used in This Study 148 Appendix B Climatological Data 149 Appendix C Oil Temperature Change Equation 150 151 Vita IV List of Tables Page Table 4-1 Wind Velocity Effects on Top Oil Temperature 19 Table 7-1 Per Cent Overloads Using the Day/Night Model For 2 Hour Emergency Period 59 Table 7-2 Per Cent Overloads Using the Day/Night Model For 6 Hour Emergency Period 60 Table 7-3 Per Cent Overloads Using the Day/Night Model For 10 Hour Emergency Period 61 Table 7-4 Per Cent Overloads Using the Varying Ambient Model For 2 Hour Emergency Period 62 Table 7-5 Per Cent Overloads Using the Varying Ambient Model For 6 Hour Emergency Period 63 Table 7-6 Per Cent Overloads Using the Varying Ambient Model For 10 Hour Emergency Period 64 Table 7-7 Per Cent Overloads Using the PJM Method 65 Table 7-8 Tabulation of Temperatures for 10 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 180°C 78 Table 7-9 Tabulation of Temperatures for 10 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 180°C 79 Table 7-10 Tabulation of Temperatures for 10 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 180°C 80 Table 7-11 Tabulation of Temperatures for 15/20/25 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 180°C 81 Page Table 7-12 Tabulation of Temperatures for 15/20/25 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 180°C 82 Table 7-13 Tabulation of Temperatures for 15/20/25 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 180°C 83 Table 7-14 Tabulation of Temperatures for 150 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 180°C 84 Table 7-15 Tabulation of Temperatures for 150 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 180°C 85 Table 7-16 Tabulation of Temperatures for 150 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 180°C 86 Table 7-17 Tabulation of Temperatures for 300 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 180°C 87 Table 7-18 Tabulation of Temperatures for 300 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 180°C 88 Table 7-19 Tabulation of Temperatures for 300 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 180°C 89 Table 8-1 Per Cent Overloads with the Day/Night Model 2 Hour Emergency, 150°C Hot Spot Limit 97 Table 8-2 Per Cent Overloads with the Day/Night Model 6 Hour Emergency, 150°C Hot Spot Limit 98 Table 8-3 Per Cent Overloads with the Day/Night Model 10 Hour Emergency, 150°C Hot Spot Limit 99 VI Page Table 8-4 Per Cent Overloads with the Varying Ambient Model 2 Hour Emergency, ,150°C Hot Spot Limit 100 Table 8-5 Per Cent Overloads with the Varying Ambient Model 6 Hour Emergency, 150°C Hot Spot Limit 101 Table 8-6 Per Cent Overloads with the Varying Ambient Model 10 Hour Emergency, 150°C Hot Spot Limit 102 Table 8-7 Per Cent Overloads Using the.PJM Method with Hot Spot Limit at 150°C 103 Table 8-8 Tabulation of Temperatures for 10 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 150°C 116 Table 8-9 Tabulation of Temperatures for 10 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 150°C 117 Table 8-10 Tabulation of Temperatures for 10 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 150°C 118 Table 8-11 Tabulation of Temperatures for 15/20/25 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 150°C 119 Table 8-12 Tabulation of Temperatures for 15/20/25 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 150°C 120 Table 8-13 Tabulation of Temperatures for 15/20/25 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 150°C 121 Table 8-14 Tabulation of Temperatures for 150 MVA Transformer 2 Hour Emergency, Hot Spot . Limit of 150°C 122 Table 8-15 Tabulation of Temperatures for 150 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 150°C 123 vn Page Table 8-16 Tabulation of Temperatures for 150 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 150°C 124 Table 8-17 Tabulation of Temperatures for 300 MVA Transformer 2 Hour Emergency, Hot Spot Limit of 150°C 125 Table 8-18 Tabulation of Temperatures for 300 MVA Transformer 6 Hour Emergency, Hot Spot Limit of 150°C 126 Table 8-19 Tabulation of Temperatures for 300 MVA Transformer 10 Hour Emergency, Hot Spot Limit of 150°C 127 Vlll List of Figures Page Figure 5-1 Insulation Life - Loss Curves 26 Figure 6-1 Idealized Daily Ambient Temperature Curve 43 Figure 7-1 Overloads for 10 MVA Transformer Two Hour Emergency, 180°C Hot Spot Limit 66 Figure 7-2 Overloads for 15/20/25 MVA Transformer Two Hour Emergency, 180°C Hot Spot Limit 67 Figure 7-3 Overloads for 150 MVA Transformer Two Hour Emergency, 180°C Hot Spot Limit 68 Figure 7-4 Overloads for 300 MVA Transformer Two Hour Emergency, 180°C Hot Spot Limit 69 Figure 7-5 Overloads for 10 MVA Transformer Six Hour Emergency, 180°C Hot Spot Limit 70 Figure 7-6 Overloads for 15/20/25 MVA Transformer Six Hour Emergency, 180°C Hot Spot Limit 71 Figure 7-7 Overloads for 150 MVA Transformer Six Hour Emergency, 180°C Hot Spot Limit 72 Figure 7-8 Overloads for 300 MVA Transformer Six Hour Emergency, 180°C Hot Spot Limit 73 Figure 7-9 Overloads for 10 MVA Transformer Ten Hour Emergency, 180°C Hot Spot Limit 74 Figure 7-10 Overloads for 15/20/25 MVA Transformer Ten Hour Emergency, 180°C Hot Spot Limit 75 Figure 7-11 Overloads for 150 MVA Transformer Ten Hour Emergency, 180°C Hot Spot Limit 76 Figure 7-12 Overloads for 300 MVA Transformer Ten Hour Emergency, 180°C Hot Spot Limit 77 Figure 8-1 Overloads for 10 MVA Transformer Two Hour Emergency, Hot Spot Limit of 150°C IX 104 Page Figure 8-2 Overloads for 15/20/25 MVA Transformer Two Hour Emergency, Hot Spot Limit of 150°C 105 Figure 8-3 Overloads for 150 MVA Transformer Two Hour Emergency, Hot Spot Limit of 150°C 106 Figure 8-4 Overloads for 300 MVA Transformer Two Hour Emergency, Hot Spot Limit of 150°C 107 Figure 8-5 Overloads for 10 MVA Transformer Six Hour Emergency, Hot Spot Limit of 150°C 108 Figure 8-6 Overloads for 15/20/25 MVA Transformer Six Hour Emergency, Hot Spot Limit of 150°C 109 Figure 8-7 Overloads for 150 MVA Transformer Six Hour Emergency, Hot Spot Limit of 150°C 110 Figure 8-8 Overloads for 300 MVA Transformer Six Hour Emergency, Hot Spot Limit of 150°C 111 Figure 8-9 Overloads for 10 MVA Transformer Ten Hour Emergency, Hot Spot Limit of 150°C 112 Figure 8-10 Overloads for 15/20/25 MVA Transformer Ten Hour Emergency, Hot Spot Limit of 150°C 113 Figure 8-11 Overloads for 150 MVA Transformer Ten Hour Emergency, Hot Spot Limit of 150°C 114 Figure 8-12 Overloads for 300 MVA Transformer Ten Hour Emergency, Hot Spot Limit of 150°C 115 Abstract In an era where financial, environmental, and public constraints are placed on additional facilities needed by utilities, a greater demand is placed on the industry to maximize the capability of all the equipment that is used to transfer power from generation sites to usage areas. It has been known for many years that power transformers can be loaded such that the output can be greater than the unit's nameplate rating. The factor that must be determined by each user is the amount of overload that can be tolerated without subjecting the transformer to undue deterioration and possible premature failure. Various methodologies for overloading have been developed over the years. All have taken into account the temperature of the air that surrounds the transformer. Approximation of this ambient temperature have been made so that overloads could be readily computed. This thesis examined the various factors that determine the allowable output of medium and large power units installed in outdoor locations. These factors included transformer design characteristics, paper insulation life, and external environmental effects. Use of projected daily ambient temperatures instead of a presently used two temperature approximation (one temperature for summer and another for winter) was investigated. Short time emergency situations were specifically addressed. Two methods for incorporating a changing daily ambient were devised, and the results obtained with these methods were compared to that obtained with the present method. The comparison was made for a series of short time emergency situations that were assumed to occur randomly and with equal probability throughout the year. Analysis of the results indicated that, in most of the cases, the use of an assumed constant ambient temperature did not handicap the overall potential overload capability of a unit. There were certain periods of the year that the overload capability was greater when calculated with the projected ambient temperature rather than with one of the two constant values. However, when comparisons were made for the year as a whole, the advantage gained by using projected ambient temperatures was overcome. It was concluded that, in the general case, the use of one constant ambient value for summer and another for winter in the short time emergency calculations would yield results that adequately project overload capabilities. Based on internal temperature comparisons, it was determined that additional overload capability can be obtained by performing a separate computation for the month of April. The remaining months of the year would still remain on the two temperature basis. Calculations were made to determine if the relative performance of the methods would change when the operation of the transformer was within the constraints of a lower hot spot temperature. The results indicated that the relative overload capability would remain unchanged. CHAPTER 1 INTRODUCTION Much of the equipment needed to direct the flow of electrical power on a utility's transmission and distribution system is rated to carry a given'amount of ultimate load without exceeding some predetermined constraints. constraints are usually thermal in nature. These Since this equipment does have losses associated with it, a portion of these losses are in the form heat dissipation. When the amount of heat generated is such that material strengths and characteristics become affected, then the thermal constraint is reached. This thermal constraint will not only be influ- enced by the design parameters of the piece of equipment, but also by the system loading profiles and the external environmental conditions. The last two factors generally vary from hour to hour and consequently will have a pronounced effect as to whether or not that thermal constraint is exceeded. For power transformers, the thermal loading capability is based on the concept that the maximum or hot spot temperature, when applied for a given amount of time, will age paper insulation at a certain rate. Hence, one can see that both time and temperature influence the maximum loading that is allowed. To obtain the maximum capability of any piece of electrical equipment, including transformers, requires real time monitoring of the equipment's thermal and electrical characteristics, as well as the environmental conditions external to that piece of equipment. In the past, the installation of monitoring equipment was rare, either because installation was economically unjustifiable or techniques for such monitoring did not exist. Consequently, various methodologies for overloading power transformers have been - A 8»21 created. ' However, with the increasing burden on utilities to keep costs down and still furnish service for customer's needs with previous reliability, equipment rating methods must be scrutinized to insure that components are operated at maximum capability without sacrifice of reliability. In this study, the effect of environmental factors on transformer overload capability was analyzed. Two methods were devised to incorporate those factors that were determined to have the greater effect on that capability. Results obtained from these calculations were compared with those overloads obtained with present methods. Analysis was made so as to determine what benefit inclusion of such climatological data could have on increasing the present overloading limits on transformers. To determine what effect such calculations had on transformers typically used in substations, a selection of medium and large power transformers was picked to cover a range of 10 MVA to 300 MVA (55°C rating). Data from four units, of four different MVA sizes and kV ratings, manufactured by four different companies, were used in the calculations. This study was limited to oil-immersed medium and large power transformers that are installed in outdoor locations. These units are of recent vintage and all use the thermally upgraded 65°C rise insulation system. CHAPTER 2 BACKGROUND On May 1, 1885, K. Zipernowsky, M. Beri, and 0. T. Blathy displayed the first closed magnetic circuit transformer at the Hungarian National Exhibition in Budapest. 18 The following year, George Westinghouse used this new apparatus and, in Great Barrington, Massachusetts, developed the first commercial lighting system. During the same year, oil was introduced as an insulating and cooling media. Such were the beginnings of the electric power distribution system as we know it today. Electric power requirements have skyrocketed. In 1882, Edison's Pearl Street Station had a capacity of 30 kW. In 1980, the total capacity of the U.S. electric power system is over 600 million kW. Nameplate ratings originally were based on standards that assumed continuous full loading at the maximum ambient temperature, usually 40°C, and a hot spot temperature of approximately 105°C. Intensive research into insulation life led to the concept of thermal aging processes. Since most units experience variable loadings, short time overloads were allowed to use up the surplus insulation life that resulted from cyclic operations at a lower rate than normal loading. These short time overloads would be compensated by loadings at lower levels during the remaining time of the load cycle. Later, standards, ' were developed to allow an increase in loss of life allowances for emergency overloads. 24 Various formula and calculation methods for computing insulation degradation and permissible overloads have been introduced. The method described in the report "Determination of Power Transformer Overloads" is one of the latest of these methods. Present day American National Standards Institute (ANSI) - National Electrical Manufacturers Association (NEMA) - Institute of Electrical and Electronics Engineers (IEEE) guides define the power transformer rated output by the power that the secondary winding can continuously deliver while operating in an ambient of•30°C continuously. The maximum hot spot temperature of 180°C must not be exceeded, and normal thermal life must be expected. That maximum hot spot temperature limit applies for transformers whose average winding temperature rise by resistance does not exceed 65°C. The vast majority of transformers that are constructed for use in the United States have this upgraded thermal insulation. However, it should be noted that the vast majority of transformers do not operate in an ambient of 30°C continuously nor do these transformers operate at full load continuously. CHAPTER 3 INSULATION LIFE To gain a very precise knowledge of the expected life of a transformer, functional life tests should be performed. These tests measure the capability of all of the materials in the context of a system. At the present time, no func- tional life testing of medium and large power transformers has proven to be practical and economically justifiable. Full scale life testing is economically unjustifiable so that life tests of a model is necessary. The model, as a whole, must apply the same stresses, in the proper sequence, as the unit itself. The model must also use the same material, and, in the overall sense, accurately simulate the configuration of the system itself. The stresses that are involved are thermal, dielectric, and mechanical in nature. The life of a transformer can become a function of any of these stresses. 19 The limits must be properly defined to adequately correlate with actual experiences. Hence, one is required to rely on material life tests to make projections of expected life for the unit itself. 10 Cellulosic insulation material (paper and press board) has been commonly used in transformers to separate the current carrying parts. With high temperature tests, the degradation of the insulation has been shown to be a gradual and cumulative process. With time, the material exhibits dryness and charring, increased brittleness, and loss of mechanical strength rather than dielectric strength. The dielectric strength retains a relatively high constant value until, after a certain point, it decreases rapidly. The degradation was also found to be quicker when the insulation 3 was oil impregnated. Once the insulation had deteriorated so badly that 1) adequate dielectric strength cannot be maintained against impulse voltages and 2) adequate mechanical strength cannot prevent conductor separation when stressed by short circuits, then the insulation, and consequently the unit itself, is subjected to a very high probability of failure. Different tests can be performed to determine the mechanical strength. These include folding, tearing, stretching, bursting, and tensile tests. yields the longest insulation life. The tensile test It is a simple test to perform, and reproducible results are obtainable. 11 Loading guides have, in one form or another, used the tensile strength criteria as the means of defining the length of functional service life of transformers. Results of early experiments 13 showed that deterioration occurred at some definite rate at all times, but that for approximately every 8°C increase in temperature, the rate of aging doubled. This aging phenomena was also found to be cumulative and progressive. The relationship defining this deterioration of the insulation in oil is based on a chemical reaction theory. Equation 3-1, shown below, defines this 4 21 relationship, known as the Arrhenius Reaction Rate. ' Log1Q L = A + y (eqn. 3-1) where: L = minimum life expectancy in hours A = constant = -13,391 for 65°C rise insulation = -14,133 for 55°C rise insulation B = constant = 6972.15 T = Absolute temperature in °K Equation 3-1 states that the logarithm of time that it takes for the insulation strength to reach a limit is linearly proportional to the reciprocal of temperature. 12 It should be noted that the factor L is the minimum life expectancy. When the tensile strength of the insulation drops to 50% of the initial strength, then it is assumed that the insulation's strength will not withstand thermal and mechanical forces. Consequently the transformer's end-of-life is assumed to be reached. After the concept of a transformer and the idea of parallel operation of transformers was introduced in 1885, improvements in design continued so that by 1913, 55°C average winding rise transformers were available. In 1958, Westinghouse Electric Corporation introduced the Insuldur system to improve thermal endurance. Initially it was amine stabilizer that was added to the transformer oil. But, by 1960, the stabilizer was actually added to the cellulose. By 1964, improvements allowed the insulation impregnated with Insuldur to withstand higher temperatures than was possible with the regular kraft paper. The composition of cellulose is a chain of glucose (CfiH, _(),.) molecules with three hydroxyl (OH) radicals associated with each glucose unit. When the cellulose is soaked in oil and heated, hydrogen ions join with the hydroxyl radicals and form Water. The water remains in the insulation 13 and, by hydrolysis, acts to deteriorate the cellulose. The Insuldur acts as a stabilizer and retards the ability of the hydroxyl radical to dissociate. With the use of Insuldur and its greater thermal capability came an increase in the average winding rise from 55°C to 65°C. With this higher temperature came an increase of approximately 12% in the continuous load rating of the transformer. If a hot spot temperature of 110°C (for 65°C average winding rise at rated power) is assumed, then, by using equation 3-1, page 12, a theoretical minimum transformer life of 65,020 hours can be expected, or approximately 1\ years of operation. This figure is based on the assumption that the unit is operating continuously at either 110PC maximum hot spot temperature for a unit with a 65°C average winding rise or at 95°C hot spot temperature for a unit with a 55°C average winding rise. In practice, the typical transformer on a utility system is not continuously loaded at full load in a 30°C ambient. Actual installations undergo varying duty cycles and changing temperatures. Hence, based 5 on operating experience, unit lives between 30-50 years are expected. 14 CHAPTER 4 ENVIRONMENTAL FACTORS AND EFFECTS Although the modern power transformer is very efficient (ranging from approximately 99 to over 99.5% efficient), there is a small percentage quantity of energy that is lost in the transformation. Most of the loss is in the form of 2 heat, originating either from the transformer winding's I R losses, eddy losses, core losses, or stray losses in the metallic structure of the unit. This heat must be removed or thermal runaway will occur and the unit will fail prematurely. Three means of heat transfer exist in a transformer: convection, conduction, and radiation. The heat from the core and coils is transferred to the oil principally by convection and thermosiphon circulation. 25 The heat is then transferred to the tank and radiators by conduction. Finally, radiation and conduction transfer the heat to the outside air. Convection sets the air molecules in motion, creating air currents which bring in new air molecules to the radiating surface. Of the three modes that account for heat dissipation in a transformer, convection is by far the most important accounting for approximately 75% of the heat dissipation. The heat transfer can be significantly improved by forced 15 13 movement of air, as would be accomplished by wind or with fans. The forced air movement breaks up and decreases the air film next to the radiating surface. This decreases the thickness through which heat must flow and thus allows for quicker removal of heat. Hence, one can see that the ultimate capability of the transformer is not only dependent on the designed parameters of the core and coils, but also on the external environmental factors. The factors that were considered in this study included ambient air temperature, wind, and solar radiation. The scope of the study was limited to outdoor installations only. Wind Wind is defined as moving air. The velocity and duration of wind can determine the heat loss rate from a piece of electrical equipment. However, since wind is difficult to predict, variable, omnidirectional, and nonuniformly distributed in height, its use in the computation of overloads for general study purposes is not easy to incorporate. In actual installations, studies would be required to accurately profile wind characteristics at a particular location. 16 The resulting wind rises would indicate not only speed, but direction and probability of occurrence. Monitoring facilities would also be required to give real time indication of wind conditions at the time that the piece of equipment would be loaded above nameplate. Two points need to be considered when analyzing the value of considering the effects of wind on a transformer's ratings. 1. The vast majority of large MVA transformers are cooled using heat exchangers and fans. These fans force air directly through the cooling coils of the heat exchanger. The effect of any additional movement of air that could be contributed by the wind would probably be negligible. 2. For small transformers that are not cooled with heat exchangers, i.e. such as OA, OA/FA, or OA/FOA types, the cooling surfaces are radiator banks. Substation facilities that contain these smaller transformers tend' to be straightforward installations that are located to serve a relatively small area. Consequently, the number of such facilities is quite large. An intelli- gence hookup with some central operator's office usually 17 does not exist. The costs involved with the installation of monitoring facilities that would indicate wind speed and direction can become a significant percentage of the total cost and would probably prove to be economically unjustifiable when compared with the benefits derived. However, where the loadings on a transformer are such that real time wind monitoring equipment is desirable and justifiable, the additional heat transferred can significantly lower internal temperatures especially for OA units. Heat can be dissipated "at a rate almost linearly with the air velocity that is moving past a radiating surface. This holds true for velocities up to approximately 95 kilometers per hour. 32 A rough approximation of the dissipation magni- tude can be made by realizing that, in ducts, 0.00155 watts/sq. cm./0.3 km/minute/ degree centrigrade above average air temperature can be dissipated. Experiments 2 9 have been performed to determine the effect of wind on the top oil temperature of a transformer. Listed in Table 4-1, page 19, are the results that were obtained with a 30 MVA transformer. 18 TABLE 4-1 Wind Velocity Effects on Top Oil Temperature Top Oil Rise in % of Rise in Still Air Wind Velocity (kilometers per hour) 0.00 100 5.86 80 17.17 50 23.44 30 35.16 25 It must be noted that the effect of the wind will not be the same for all transformers. In addition, oil tempera- ture decreases may not hold when the direction of the wind changes, since most medium and large power transformers are rectangular in shape but do not have the radiating surfaces arranged in a symmetrical fashion around the transformer. In light of the above, the effects of wind on a power transformer's ratings were not incorporated into the calculations, Solar Of the 1,350.9 watts/square meter of solar heater radiation on a plane surface that occurs outside the earth's 19 atmosphere, the peak amount that eventually reaches the ground in the Pennsylvania-New Jersey-Maryland area varies from 870 watts/square meter in July to 964.5 watts/square meter in March. 14 The attenuation and variation is due to absorption and scattering by water vapor, clouds, haze, dusk, and the angle of incidents between direct rays and the exposed surface. On a clear day, the maximum solar intensity that can be expected at sea level is approxmately 940 watts/ square meter. 14 The effect that such solar intensity can have on increasing the top oil temperature of a transformer can be calculated by the equation in Appendix C, page 150. This temperature calculation takes into account the low temperature total emissivities as well as the coefficient of absorption of solar radiation. The absorption coefficient ranges from about 0.07 for highly polished silver to 0.75 for gray paint and 0.97 for black matte. The emissivity coeffi- cient can be as low as 0.02 for highly polished silver to 0.95 for gray paint as well as lamp black. However, it has been found that the temperature of the oil of a transformer is practically independent of color. Tests 14 23 2 ' have indicated that for units (5 kVA) loaded continuously, the maximum difference in the top oil tempera- 20 ture of units painted with various colors (including white and black) was approximately 2°C. For units with little loading and for those with no loads, the top oil temperature difference between the different colored units increased. But, it was less than a 4-5°C difference. These tests were conducted continuously during a three day period and night cooling results were also recorded. The top oil temperature between the various colored tanks was much closer, with 1°C and less differences prominent. Based on these results, the effects of solar irradiation on transformer tanks have been neglected in the overload calculations. Consequently, the overload calculations have been done with ambient temperatures being the most influential environmental factor. 21 CHAPTER 5 OVERLOAD METHODOLOGIES Although a transformer will wear out, it will wear out from thermal deterioration and the resultant lowering of electrical dielectric strengths and short circuit capabilities rather than from frictional degradation of moving parts. Since the thermal deterioration is dependent on a summation of the various temperatures of the transformer, the ambient temperature that is chosen becomes an important factor in determining the allowable insulation deterioration. Various methods for calculating overloads have been introduced over the years. Each model the transformer and its environment by including certain assumptions and simplifications so that calculations could be done without too much difficulty. In 1969, the Transmission and Substation Design Subcommittee of the Pennsylvania-New Jersey-Maryland Interconnection (PJM) published a report entitled "Determination of Power Transformer Ratings" and thereby setforth a standard procedure for establishing overload thermal ratings for old and new power transformers. The method/procedures contained in that report work within the framework of the ANSI and NEMA 22 standards that have been established for power transformers. The calculations in this study used that report as the preliminary basis for overload calculations. The largest loading that a transformer can be expected to carry is proportional to the hottest spot conductor temperature that exists in the unit. This hot spot tempera- ture is calculated by adding the ambient temperature, top oil rise over ambient, and the hottest spot rise over top oil, as shown in equation 5-1 below. 6 HS = T a + 0 o + 9 ^»- g where: 6WC, = Hot spot temperature (°C) T = Ambient temperature (°C) 6 = Top oil rise over ambient temperature (°C) 8 = Hot spot rise over top oil temperature (°C) O Predicting the overload capability of the transformer for some future time period requires a need to approximate the ambient temperature for that time period. According to the NEMA Standards, the average daily temperature for the month involved should be used for loads where normal life expectancy is expected. Where short terra overloads are 23 5 X " ) involved, the average of the maximum daily temperatures for the month involved should be used. The report recommended that, from the viewpoint of ambient temperatures, the year be divided into two parts. The months April through October were defined as "summer" and, for emergency overload periods of less than 24 hours, an ambient of 35°C was assumed. For periods greater than 24 hours, the ambient was assumed to be 30°C. The months of November through March were defined as "winter", and, for all overload calculations, 10°C was assumed. The daily load cycle has been assumed to be two step, with peak loading occurring for 14 hours at one per unit and 0.7 per unit off-peak loading for 10 hours. Short time emergencies would occur in the final hours of the peak period so as to include all of the insulation life deterioration that occurs because of the emergency. The NEMA Guide is the basis for the insulation life curve. The equation 5-2, page 25, is based on the Arrhenius Chemical Reaction Rate Theory. In the report, the 55°C insulation loss of life curve was chosen to be parallel to 24 v the 65°C curve, such that proportional life expectancies between the two insulation types could be obtained. % Loss of Life/Hour = lOOe (K " 6K + T hs a + 273) (eqn * 5 2) " where: 0, = Hot spot rise over ambient temperature °C T = Ambient Temperature °C K = 32.543 for 55°C rise insulation 30.834 for 65°C rise insulation These insulation life curves are shown in Figure 5-1, page 26. Certain limits were imposed so that additional overloading on the transformer would be prevented. For the 65°C rise transformer, these limits included a maximum top oil temperature of 110°C, a maximum hot spot temperature of 180°C, for short time emergencies (less than 24 hours) a maximum loss of life of 1.5% per day, and a maximum overloading of 200%. These suggested limits were based on a concensus of the PJM member companies. Each member utility may modify these limits to best suit their operating conditions, experience, and philosophy. In the case of the Pennsylvania Power & Light Company, Allentown, Pennsylvania, the allowable top oil temperature was set at 10°C higher (120°C), the short 25 FIGURE 5-1 INSULATION LIFE-LOSS CURVES 1.0 ■- RISE INSULATION 0.1 ■- C RISE INSULATION P< o.oi •- s 160 140 120 HOT SPOT TEMPERATURE 26 time emergency loss of life was increased to 5% per day, and a maximum overload was set at 2.25 per unit. In this thesis, the computations involved the following scenario. Only emergency periods were considered, and, for evaluation and comparison purposes, these were assumed to be 2, 6, and 10 hours long. The need to overload a transformer was assumed to occur randomly and unexpectedly. For compar- ison purposes, it was assumed that in any given day throughout the year, there was an equal probability that a condition may exist which would require the transformer to surpass its rated nameplate rating. Most actual operating situations are such that transformers experience a higher probability of having an overload only during certain periods of the year. But, there are also instances, such as transfer of power between utilities, that a need for short time overloads may indeed occur at any time throughout the year rather than being more prevalent during only a certain part of the year. To simplify the calculations, every day of a given month has been assumed to be identical. ambient temperature models were required. Thus, monthly . The values for the temperature obtained from these models have been used in the calculations. An overload computation was necessary for 27 each month and for each of the three emergency periods. Data from four different transformers (shown in Appendix A, page 148) was used in the computations. Although the PJM Report addressed both the 55 and 65°C rise insulations, only the results obtained with transformers with the 65°C insulation were studied here. New transformers with this thermally upgraded insulation are used almost exclusively by every utility in the United States. Units by four different manufacturers were chosen, with sizes ranging from 10 MVA to 300 MVA. The limits used in the overload calculations were the same as used by the Pennsylvania Power & Light Company, i.e. a hot spot limit of 180°C, top oil temperature limit of 120°C, loss of life for the short time 2, 6, and 10 hour emergencies of 5% per day, and a maximum rating of 2.25 per unit. The rating method developed in the PJM Report calculates the allowable overloads by determining the winding hot spot temperature for a given loading. To make this calculation, it is necessary to use two groups of data. The first set includes information on the transformer's thermal and physical/ electrical properties. These include such factors as no-load i f losses, load losses, weight of the core and coils, weight of the tank and fittings, gallons of oil, average conductor 28 rise over ambient at rated load (from a factory test report), top oil rise over ambient at rated load (again from a factory test report), hot spot over top oil temperature values (based on a manufacturer's design), and rated insulation temperature rise (55°C or 65°C).■ The second set of information that is necessary concerns the external factors that influence the environment that the transformer operates in. The load cycle is of most importance. Although 24 different hourly values of loading (with peak load defined as one per unit) can be placed into the program, it is much more convenient to use a two step load curve. The NEMA Standard Publication #TR98 derives the means of converting the continuously varying daily load cycle into the two step cycle. Again, the peak is defined as one per unit and the off-peak is a function thereof. As previously stated, the cycle that was agreed on was one per unit peak for 14 hours and 0.7 per unit off-peak for 10 hours. Finally, it should be noted that the computation technique developed in the report is done in two parts. Long time overload calculations are done for periods of one day and greater. Short time emergency calculations are performed for periods of between 1 through 24 hours. 29 For long time ratings, the entire load cycle is overloaded with an incrementally higher per unit loading until one of the limits is reached. However, even at the higher loading, the relation- ship between the peak and the off-peak values remains as in the normal case (i.e. 1 to 0.7 per unit). For the short time emergency calculations, the load cycle is used as before. However, the day that the emergency occurs, the emergency period is automatically assigned to the final hours of the peak period. Thus, a computation of the overload for a 2 hour emergency period would assume that it had occurred starting during the 22nd hour. It is this two hour period that would be loaded with a higher per unit loading until a limiting constraint is reached. Normal load cycles are assumed for the day before and after the day the emergency occurs. In this manner, the entire loss of life attributable to the emergency is included. Aside from a load cycle, the factor that causes significant variation in allowable overload is the ambient temperature. A quick calculation showed that for a 10/11.2 MVA transformer, the temperature change from -15°C to 45°C yielded overload capability ranging from 223% to 89% respectively on a 10 MVA basis. As stated above, the report assumed that the ambient 30 temperature was a constant value throughout the load cycle. If, instead of using a constant value, an ambient temperature value that more closely approximated the actual value at the time that the emergency occurs could possibly yield greater allowable overloads without decreasing the expected life of the unit. One proposal that would get away from using only one temperature value involves using a two step ambient temperature model. The temperature values would be adjusted at sunrise and sunset. At sunrise, the projected high tempera- ture of the day would be used as the temperature value in the computation of the overloads for the sunrise through sunset period. For the remaining hours of the 24 hour day, the temperature recorded at sunset would be used. In this study, this model was referred to as the Day/Night Model. Additional development of it was done in Chapter 6, page 40. In the development of this study, it was determined that adequate information was available to allow the use of the continuous temperature profiles to formulate a more realistic model of the transformer's operation. This model was referred to as the Varying Ambient Model and it, also, was developed more completely in Chapter 6, page 40. 31 The computation of the short time emergency overloads with the ambients as described in the PJM Report (referred to as the PJM Model), the Day/Night Model, and the Varying Ambient Model followed the same sequence. Using the values of the load cycle as well as the ambient temperature, the top oil temperature was calculated for each hour of the 24 hour period. For the hours prior to the emergency, the normal loading was used. During the time of the emergency, an assumed peak loading was used. The top oil rise over ambient at the end of the 24 hour period was calculated. The assumed peak loading was increased, and an hour by hour determination of the oil temperature was again made. When, at the end of the two successive 24 hour periods, the top oil rise over ambient was within a specified tolerance, then the calculated loss of life was checked against the allowed limit. Should the calculated loss exceed the limit, the loading was decreased until the calculated value was under the limit. The calculated hot spot temperature was then checked to determine if that limit had been surpassed. If needed, the loading was again decreased to bring it below the limit. The top oil temperature was finally checked and the limiting check and adjustment was made here as well. Once the loading was such that the above limits are not exceeded, then the allowable loading above nameplate had 32 been determined. The temperature determination equation used in the calculations are described as follows. The thermal time constant for rated load with an initial temperature rise of 0°C is defined as T . The time constant is the length of time that would be required for the temperature of the oil to change from the initial value to the ultimate value if the initial rate of change were continued until the ultimate temperature was reached. Since transfor- mers not rated OA or OA/FA are designed such that the exponential power of temperature rise versus loss equals unity (i.e. n = 1), then 63% of the temperature change occurs in the length of time equal to the time constant. This occurs regardless of the initial and ultimate temperatures. The equation for the time constant is: T r = JP* r (eqn. 5-3) fl where: T = Rated time constant of the top oil rise (hours) H = Thermal capacity (watt hours/degree) T = Top oil rise over ambient at rated load (°C) Pf, - Totaliwatts loss at rated kVA 33 The thermal capacity H for non-self cooled transformers can be calculated by: H = 0.06 WCC + 0.06 WT + 1.93 GO (0.06 WCC + 0.06 WT + 1.93 GO) (eqn. 5-4) where: WCC = Weight of the core and coils (lbs) WT = Weight of the tank and fittings (lbs) GO = gallons of oil For units that are self cooled (n = 0.8), equation 5-3, page 33, holds only for full rating starting cold. In these units, the time constant becomes a function of both the initial and final temperature rises. Hence, this time constant can be computed using the equation 5-5, below. ou T 0 T = T "e ou T 0_ 1 n 01 To e Ol.' 1 n (eqn. 5-5) _T o where: ou Ultimate top oil rise over ambient temperature (°C) 0 01. = Initial top oil rise over ambient temperature (°C) T = Top oil rise over ambient at rated load (°C) n = Exponential power of temperature rise versus loss 0.8 for self cooled transformers 1.0 for FOA, 0A/F0A, and 0A/FA/FA transformers 34 The thermal capacity H is calculated by changing the coefficients of the tank weight and gallons of oil from 0.06 to 0.04 and 1.93 to 1.33, respectively, in equation 5-4, page 34. A resistance correction factor C for any average winding temperature is: _ 234.5 + AWT r " 234.5 + D , r C _ ,, (eqn - 5_6) where: D = Design average winding temperature of 75°C for 55°C transformers and 85°C for 65°C units (°C). AWT = Average Winding Temperature (°C). The resistance correction factor compensates for the variation in load loss as the winding temperature changes. Therefore, the resistance correction factor, C, modified to take into effect of winding and stray resistance, is as i shown below. CL C = CJ r ££ LL + X I C where: J2L = Rated copper loss (watts) LL = Rated load loss (watts) 35 LL (eqn. 5-7) The steady state top oil rise for constant load can be calculated using equation 5-8, below. 8 ou = To NfTT1] " <«*•• 5"« where: 8 ou = Ultimate top oil rise over ambient temperature (oc) T = Top oil rise over ambient at rated load, as determined by factory tests (°C) C = Resistance correction factor for ultimate winding temperature K = Per unit load R = Ratio of load to no-load losses at rated load on the temperature test tap. An hour by hour tabulation of the transient top oil rise is next made using equation 5-9, below. 6 O = (6 OU - 6 .)(1 - e" Ol /T ) + 6 . 01 (eqn. 5-9) ^ where: 9 = Top oil rise over ambient temperature at the end of time t T = Thermal time constant for the temperature rise interval 6 . to 6 oi ou 36 These calculations are done for whole hour periods, with the previous hour's results impacting the results of the next hour's calculations. Hour-by-hour computations are then made of the average winding temperature. The average winding temperature rise over ambient is first calculated, as shown below. 6 av=0o+ (Tav - To)(CK2)n (eon. 5-10) where: 0 av T = Average winding temperature rise over ambient for a load K (°C) = Average winding temperature rise over ambient, as determined by factory test at rated load (°C) The average winding temperature can then be determined by adding the above result to the ambient temperature, T as shown in equation 5-11, below: AWT = T a + 6 av (eqn. 5-11) ^ Using this calculated value for the average winding temperature, the resistance correction factor can now be 37 recalculated and this new value used to correct the top oil rises and the winding temperature rise. With these recalculated values, the transient hot spot gradient'can be calculated. This rise, 0, , is the hot spot rise over ambient temperature for an amount of time (t) for a load K, and is the sum of the top oil rise and the hot spot rise over top oil per equation 5-12, below. 8 hs' = 6 o +6 g v(eqn. ^ 5-12)' where: 0, = Hot spot rise over ambient temperature (°C) 0 = T 8 T 8 (CK2)n 8 = Hot spot rise over top oil temperature at per unit load K (°C) = Hot spot rise over top oil temperature at rated load on the test tap, as per the design of the transformer (°C) Finally, the hottest spot temperature, 0trq> °f the system can now be computed by equation 5-13, below. 6 HS 38 = 6 hs + T a (e< n * ' 5 13) " The loss of life of the insulation for the computed hot spot can be determined using equation 5-2, page 25. With all 3 temperature-related limiting factors now available, a check is made to determine if the limits have been exceeded on the top oil temperature, hot spot temperature, and the loss of life. If, for the given loading, none of the limits are above their limit, then the per unit loading is increased and the above computations are redone. When the per unit loading is such that an additional incremental increase will produce a violation in a limiting constraint, then the proper loading has been attained. This per unit loading is then translated into a MVA reading and a percentage reading. 39 CHAPTER 6 AMBIENT TEMPERATURES AND LOAD PROFILES The Development of an Ambient Temperature Model To determine the various ambient temperatures for use in subsequent calculations, it was necessary to develop a representation for the temperature variation throughout a given day. The heating of the air during the daytime will be directly dependent upon the rate that the sun can warm the earth's surface and, by convection, heat the surroundings. This can be represented by a sinusoidal function. Nighttime cooling is exponential, but can adequately be represented as a sinusoidal function as well. 9 A harmonic function that represents the changing temperatures throughout a day in a given month can be expressed by 40 an equation of the following form: AMB(t) = M + D sin |^ (t - (P - 6)) (eqn. 6-1) where: AMB(t) = Ambient temperature during any given time t (°C) M = Mean temperature (°C) T = Period duration, expressed as 24:00 hours D = Temperature deviation from the mean (°C) t = Time of day (hour on the basis of 00:00 to 24:00) P = Time that peak temperature occurs (hour expressed on the basis of 00:00 to 24:00) In this study, it was assumed that all days in a given month are identical. To make the overload computations, the idealized ambient temperature values was derived using equation 6-1, above, and the data on ambient temperature averages. These averages are obtained from the US Department of Commerce National Weather Service's tabulation of averages for a 30 year period. various locations. These averages are tabulated for In particular, the data from the Allentown- Bethlehem-Easton Airport location was used in the study. Appendix B, page 149, average maximums, minimums, and mean temperatures for each month are shown. 41 In Studies have indicated that the peak in daily ambient temperature occurs about one hour after the peak in the intensity of solar radiation. Since solar radiation peaks occur around noon, the high ambient temperature of the day can be expected to occur about 1 PM. Substituting P equals 13:00 hours (1 PM on a 24 hour clock basis) in equation 6-1 results in a continuous profile of ambient temperatures for a typical day of the month in question. displayed in Figure 6-1, page 43. This is graphically To determine the ambient temperature data required to make the Day/Night calculations, the average time that sunset and sunrise occurs for a given month was required. These times were obtained from an almanac, and the respective temperatures profiles were determined using equation 6-1, page 41. Shown in Appendix B, page 149,are the average peak and sunset temperatures, as well as the times chosen for sunrise and sunset, that were used for subsequent overload calculations. For the calculations with the Varying Ambient method, the temperature values were obtained directly from equation 6-1, page 41, on an hour-by-hour basis. 42 o O W ^-1 VO O H a n w Q M 43 Correlating Load Curves With Changing Ambient Temperature: The typical daily load cycle on a transformer is usually a curve that is continuously oscillating; varying from a minimum point to a maximum point back to a minimum point within a 24 hour period. As detailed in the NEMA Standards such a curve can be reduced to a two-step rectangular function. This function would consist of a peak and an off-peak period. The values for the periods used here were the same as used in the PJM report "Determination of Power Transformer Rating", i.e. 14 hours at 1 pu peak and 10 hours at 0.7 pu off-peak. The load cycle was assumed to remain constant each day throughout t!he year. For the Day/Night methodology, the peak period was arbitrarily chosen to start at 5 a.m. and continue for 14 hours through 7 p.m. The sunrise and sunset times for a given month, as previously determined, were then located relative to the load cycle. Peak temperatures were allocated to the daytime hours between sunrise and sunset, and sunset temperatures were allocated to the remaining hours. The emergency periods were assumed to occur during the final hours of peak period. Thus, the two hour emergency was assumed to begin at 5 p.m. and continue until 7 p.m. For the Varying Ambient calculations, the correlation of the ambient temperature model with the daily load cycle 44 model required additional areas of consideration. shown 20 It can be that the change in the top oil temperature will lag a given ambient temperature change. In several hours, the oil temperature's change will asymptotically approach the ambient temperature change. Since both the ambient and oil temperatures determine the final internal temperatures of the transformer and, consequently, the total permissible overload, there was a need-to determine at what point the peak ambient temperature and the load cycle produced the most conservative overloads. Calculations were made and the resulting sensitivity models indicated that placing the peak ambient temperature in the second last hour of the peak period resulted in the lowest permissible overloads. Again, it was arbitrarily assumed that the peak period would begin at 5 a.m. and continue for 14 hours until 7 p.m. Since the ambient temperature peak must occur in the second last hour of the peak cycle, the overload calculations with the Varying Ambient models positioned the peak daily temperature at 6 PM. The hourly values of the ambient temperature for each month were determined using equation 6-1, page 41, and P equal to 18:00. These values were then used in the calculations to determine final allowable overloads. 45 CHAPTER 7 RESULTS AND DISCUSSIONS In this chapter the results obtained with the methodologies previously described are analyzed and compared with each other. Because of the large amount of data that was obtained from these calculations, it was necessary to organize it in the form of tables and charts to better convey the various trends of the limits and overloads that will be considered. As previously mentioned, each day of the month was assumed to have an identical temperature and temperature variation. These variations were as defined by equation 6-1, page 41, and the temperature values were as listed in Appendix B, page 149. In addition, the load cycle configur- ation was assumed to be rectangular and identical each day throughout the year. To effectively compare the three methods, the results with the Day/Night and the Varying Ambient methods were compared with the PJM method. To extend the scope of this comparison, a selection of new vintage medium and large 46 power transformers, ranging in size from 10/11.2 MVA to 300/336 MVA (on a 55/65°C rating basis) was chosen, as stated in the introduction. A hypothetical emergency scenario was devised so that overloads can be calculated. It was assumed that, for example, one two-hour emergency would occuri.in each and every month for a year's time period. The permissible overloads that resulted with the Day/Night and the Varying Ambient methods were tabulated and compared with those obtained with the PJM method. Such calculations were repeated for the 6 and 10 hour emergency periods. If, for example, the permissible overload with the Day/Night method was 5% over that allowed with the PJM method, then a +5 is assigned for that month. A negative value, on the other hand, indicates that a permissible overload using the new method was less than what was obtained with present methods. These values are shown after the slash (/) in Tables 7-1 through 7-6, pages 59 through 64. The positive and negative values for each month were then summed for the entire year, and the resulting net positive/negative value was part of the overall analysis and comparison. Using these net values, one can quickly determine which method was allowing a greater net overload and for what emergency overload periods. The method that allows for less emergency overloading than that obtained by the present means was also quickly pointed out. 47 The per cent overloads over nameplate were also shown in the tables mentioned above. preceding the slash. These values were shown These values were plotted and shown in Figures 7-1 through 7-12, pages 66 through 77. Each chart represents one emergency period (either 2, 6, or 10 hour) for each transformer. Hence, one can get a quick represen- tation of the magnitude and time of the deviation of the overloads with the new methods as compared with those obtained with the existing method. One can also get an idea of how the two methods compared with one another. Finally, since most of the time the two limiting factors that prevent additional transformer overloading were the oil temperature and the hot spot temperature, a group of tables were made to show these two temperatures as a function of the different months of the year. Again, only one transformer size and one overload period was listed per table and the temperatures from all three methods were listed in that table. An asterisk (*) before the temperature indicates that the allowable limit has been reached and further overloading was prevented. 48 The following summarizes the grouping of these tables and charts: A) Tables 7-1 through 7-7 inclusive (Pages 59 through 65) list the allowable overloads as determined by the various methods. B) Figures 7-1 through 7-12 inclusive (Pages 66 through 77) graphically display the overloads from the various methods for a given time per transformer. C) Tables 7-8 through 7-19 (Pages 78 through 89) compare the limiting criteria preventing additional overloads for the different methods. The discussion of the results was broken down into two parts. The first covered the results from the time (2, 6, or 10 hour) standpoint, while the second part reviewed the results from an individual transformer basis. First, the analysis was made on the short time emergencies. The Two Hour Emergency Case: In this short time emergency case, the overload calculations with the Day/Night method exceeded the 49 values calculated with the present method, with a difference ranging from 11-24%. The largest gains occurred in the month of April, September and October. Smaller but still pronounced gains occurred in February and March. Less overload than allowed with the present method was found during the summer months of June, July and August. In most cases, the factor that limited additional loading throughout the year was the hot spot temperature. The exception was the small 10 MVA transformer for which the oil temperature during the summer months reached its allowable limit. With the Varying Ambient method, the allowable overloads that resulted usually proved to be less than what would be allowed with the present methods. Overall, one would, for the year, have additional (depending on the MVA size) capability of 27-58% over that allowed with the present method. The only months in which the Varying Ambient method allowed more loading were April and October. The hot spot temperature prevented additional loading throughout the year for the large MVA transfor50 mers and, during the spring, fall and winter months, for the medium size transformers. Oil temperature limits were reached during the summer months thereby preventing additional loading on the latter units. The Six Hour Emergency Case: In this situation, the Day/Night method yielded results which, on the balance, proved to be out performed by the present method by as much as 15%. The majority of the loss occured during the winter and summer months. Very similar results were noted in May and September, but, during April and October, the Day/Night method allowed greater overloads. The limiting factor for all except the small 10 MVA transformer was the 180°C hot spot temperature. For the 10 MVA transformer, additional overloading was stopped because the oil temperature limit was reached. Investigating the 6 hour emergency case with the Varying Ambient method showed the results to be very similar to the 2 hour case! With the present method, the allowable overloads exceeded by anywhere from 51 23.7% to 38% what was calculated with the Varying Ambient method. Except for April, May and October, the allowable loading above nameplate was less throughout the remainder of the year. Again, the hot spot temper- ature limited loadings for the two larger transformers, t while a combination of oil/hot spot and just oil temperature limits were noted for the two smaller units. The Ten Hour Emergency Case: The 10 hour case gave the most pessimistic results for the Day/Night method. The present method allowed a transformer to exceed nameplate ratings by 10-38% for the 12 month period. As with the above cases, the results with the Day/Night method exceeded the present method only in April and October while in the remaining months of the year, the results were noticeably the opposite. The limiting constraints for this case varied widely. For the small 10 MVA unit, oil temperature limited overloads during all of the months. A combin- ation of oil temperature (for the summer months) and hot spot temperatures (for the remainder of the year) 52 prevented additional loading on the 15/20/25 MVA unit. On the large 150 MVA unit, neither the oil nor hot spot temperatures were the limiting constraint. Rather, the 5% loss of life limit prevented more loading. For the 300 MVA unit, overloads during all of the months were limited by hot spot temperature. Results similar to the above were obtained with the Varying Ambient temperature model. For the 12 month period, the present method allowed up to 33% more loading above nameplate than what the calculations with the Varying Ambient method allowed. The April, May and October months were again the only months that the present method did not allow more MVA output than what was shown with the Varying Ambient method. The limiting constraints for this method for all except the 150 MVA transformer proved to be the same as with the Day/Night method. For the 150 MVA unit, the hot spot temperature instead of the loss of life limit was reached. It should be pointed out that the values mentioned above are not that extremely out of line. 53 A 12% difference in favor of one method for a 12 month period yields an average difference of 1% per month in that method's favor. Although the magnitude of that percentage is important, it is felt that it is of greater value to determine which method is actually working out for the better during different times of the year. By analyzing the results on an individual transformer size basis, one can determine how the limiting constraints changed with the different methods as well as with the length of time of the overload, as represented by the emergency period. For the 10 MVA unit, all three methods showed, from winter to summer, increasingly higher oil temperatures. The oil temperature activated the overload cutoff in the summer for the short 2 hour emergencies. But, in the longer 6 and 10 hour emergency periods, the oil temperature limit was reached for the Day/Night and the Varying Ambient methods throughout the year. With the PJM method, the limiting constraints changed from oil temperature in the summer to the hot spot limit for the winter. It should be noted that the temperatures obtained for both oil and hot spot at the allowable overload were very close with all three methods. 54 Many times, there was a difference of less than 5°C in any given month when the oil or hot spot temperatures were compared between methods. In the case of the 15/20/25 MVA unit, the hot spot temperature remained very close to the 180°C limit. In most instances, it was this temperature that limited additional overloads. The oil temperature remained relatively cool in the winter, but got increasingly warmer in the summer. Occasionally, it got hot enough in the summer to reach its limit, thereby cutting off additional overloading. Again, the difference in temperatures obtained for the oil and hot spot with these three methods, for any given month, was very close. Whereas the two smaller transformers above have both the oil and hot spot temperature limits reached, the limiting constraint for the two larger units, except' for two instances, was the hot spot limit only. The oil temperature, although varying between the cooler and warmer ambients, remained relatively cool as compared with the smaller units above. As mentioned previously, in two time periods of the year, the overload capability, as calculated with the PJM 55 method, gave results that were less than what was calculated with the other two methods. and September/October. These two periods were April/May All of these months lie within the "summer" range as defined in the PJM methodology. When looking at the value of the oil temperature as computed in the PJM method and comparing it with that obtained from the other two methods, acceptable differences were shown for the months of May, September and October. the values obtained for April are noteworthy. However, The oil temperature for April in the PJM method was closer to that obtained with the Day/Night and the Varying Ambient method for the warmer month of May. This implies that the assumed ambient of 29°C used in the PJM program may be unduly high, and therefore unnecessarily overstate the temperature increases inside the transformer. However in most instances, to include April as a "winter" month and use the 4°C ambient may prove to be too liberal. Since the average temperature for April is approximately 10 °C (with a deviation of +/-11°C), consideration should be given to making a separate overload computation for that month with an assumed ambient in the 10-20°C range. 56 The resultant greater transformer overloads will then decrease the large differential that was seen when results were compared between the three methods. It was of interest to note how the results vary as the length of time of the emergency changes. For all four transformers, the calculations with the Day/Night method yielded overloads that were very favorable, as compared with the PJM method, for the short time periods, i.e., two hours. As the time periods began to get longer, the overloads went in favor of the PJM method and became increasingly so with longer times. This was irrespective of the transformer MVA size: On the other hand, a different trend was detected with the Varying Ambient method. For short time periods, the PJM method was very much ahead in total overload, as per the scenario. However, as the length of time of the emergency continued, the advantage began to decrease. Although the advantage never switched to the Varying Ambient method (at least as far as these calculations could determine), the fact that this trend existed is more significant. First it should be recalled that, in developing the Varying Ambient method, it was decided to position the peak temperature at a 57 location opposite the load cycle that would produce the most conservative results, i.e. in the 23rd hour of the peak of the load cycle. The two hour emergency would begin at exactly that instant. However, with increasing longer emergency periods, the start of these periods, while still occurring during the peak portion of the load cycle, was also occurring at a time period where the ambient temperature was significantly lower than the peak temperature. Hence the transformer began to bear the emergency overload during a period when the ambient was still cool, as compared to the peak temperature. With the Day/Night method, the peak temperature was either in existance or proceeded all overloads. With longer emergency periods, the effect of the higher projected peak ambient temperature became increasingly more important. Consequently, the allowable overload decreased. seen for all transformer sizes in the study. 58 This was TABLE 7-1 PER CENT OVERLOADS USING THE DAY/NIGHT MODEL FOR 2 HOUR EMERGENCY PERIOD Size of Transformer 10 MVA Month 15/20/25 MVA 150 MVA 300 MVA January 197/0 219/1 167/0 178/0 February 200/3 221/3 169/2 180/2 March 195/-2 220/2 167/0 178/0 April 195/15 217/13 164/9 176/11 May 181/1 205/1 156/1 167/1 June 176/-4 202/-2 154/-1 164/-2 July 172/-8 200/-4 152/-3 163/-3 August 174/-6 201/-3 153/-2 164/-2 September 187/7 212/8 160/5 172/6" October 190/10 213/9 162/7 173/7 November 193/-4 215/-3 164/-3 175/-3 December 196/-1 217/-1 166/-1 177/-1 Net overload gain (+)/loss (-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: +11% +24% +13% +16% The net overload gain or loss was determined by summing the monthly per cent differentials when results from the above method are compared to those of the PJM method. This difference is shown after the slash (/) above. 59 TABLE 7-2 PER CENT OVERLOADS USING THE DAY/NIGHT MODEL FOR 6 HOUR EMERGENCY PERIOD Month 10 MVA Size of Transformer 150 MVA 15/20/25 MVA 300 MVA January 182/-2 201/0 159/-2 166/0 February 182/-2 202/1 159/-2 167/1 March 175/-9 197/-4 156/-5 162/-4 April 169/10 193/7 153/5 159/5 May 161/2 187/1 149/1 155/1 June 156/-3 184/-2 146.1/-1.9 152/-2 July 153/-6 182/-4 145/-3 151/-3 August 154/-5 183/-3 146/-2 151/-3 September 160/1 188/2 148.2/0.2 155/1 October 170/11 194/8 153.4/5.4 160/6 November 175/-9 197/-4 156/-5 163/-3 December 181/-3 200/-1 158.2/-2.8 165/-1 Net overload gain (+)/loss (-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -15% +1% . -12.1% -2% 60 TABLE 7-3 PER CENT OVERLOADS USING THE DAY/NIGHT MODEL FOR 10 HOUR EMERGENCY PERIOD Size of Transformer 10 MVA Month 15/20/25 MVA 150 MVA 300 MVA January 176/-5 193/-2 155.8/-2.2 161/-2 February 176/-5 193/-2 155.8/-2.2 161/-2 March 171/-10 189/-6 153.2/-4.8 157.2/-5.8 April 165/8 185/6 150.3/5.3 154.2/4.2 May 159/2 180/1 146.8/1.8 151/1 June 153/-4 176/-3 144/-1 148/-2 July 151/-6 173/-6 142.5/-2.5 147/-3 August 152/-5 175/-4 143.4/-1.6 147.1/-2.9 September 156/-1 180/1 145.8/0.8 150/0 October 163/6 185/6 149.4/4.4 154/4 November 169/-12 189/-6 152.2/-5.8 157/-6 December 175/-6 192/-3 155/-3 160/-3 Net overload gain (+)/loss (-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: +38% -18% -10.8% -17.5% 61 TABLE 7-4 PER CENT OVERLOADS USING THE VARYING AMBIENT MODEL FOR 2 HOUR EMERGENCY PERIOD 10 MVA Month Size of Transformer 150 MVA 15/20/25 MVA 300 MVA January 191/-6 212/-6 164/-3 173/-5 February 191/-6 212/-6 163/-4 173/-5 March 188/-9 210/-8 161/-6 171/-7 April 185/5 207/3 158.1/3.1 169/3 May 180/0 203/-1 156/1 166/0 June 174/-6 200/-4 153/-2 163/-3 July 170/-10 199/-5 152/-3 162/-4 August 172/-8 200/-4 152.1/-2.9 163/-3 September 176/-4 202/-2 154/-1 165/-1 October 183/3 205/1 157/2 168/2 November 187/-10 208/-10 160/-7 170/-8 December 190/-7 211/-7 163/-4 173/-5 Net overload gain (+)/loss (-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -58% -49% -26.8% -36% 62 TABLE 7-5 PER CENT OVERLOADS USING THE VARYING AMBIENT MODEL FOR 6 HOUR EMERGENCY PERIOD 10 MVA Month Size of Transformer 150 MVA 15/20/25 MVA 300 MVA January 178/-6 197/-4 157.1/-3.9 163/-3 February 178/-6 196/-5 157/-4 163/-3 March 173/-11 194/-7 155/-6 160.1/5.9 . 168/9 190/4 152/4 158/4 May 162/3 187/1 149/1 155/1 June 156/-3 183/-3 146.2/-1.8 152/-2 July 153/-6 182/-4 145/-3 150/-4 August 155/-4 183/-3 146/-2 151/-3 185/-1 148/0 153/-1 April September ,159/0 i October 165/6 189/3 151/3 156.1/2.1 November , 177/-13 192/-9 154/-7 159/-7 December 177/-7 196/-5 157/-4 162/-4 Net overload gain (+)/loss (-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -38% -34% -23.7% -25.8% 63 TABLE 7-6 PER CENT OVERLOADS USING THE VARYING AMBIENT MODEL FOR 10 HOUR EMERGENCY PERIOD Month 10 MVA Size of Transformer 15/20/25 MVA 150 MVA 300 MVA January 176/-5 191/-4 156/-2 160/-3 February 176/-5 191/-4 156/-2 159/-4 March 171/-10 188/-7 154/-4 . 157/-6 April 166/9 184/5 151/6 154/4 May 160/3 181/2 148/3 151/1 June 154/-3 177/-2 145/0 July 151/-6 174/-5 144/-1 147/-3 August 153/-4 176/-3 145/0 148/-2 September 157/0 179/0 147/2 150/0 October , 163/6 183/4 150/5 153/3 November 169/-12 186/-9 152.3/-5.7 156/-7 December 175/-6 190/-5 155.2/-2.8 159/-4 ' 148.1/-1.9 Net Overload gain (+)/loss (-) for. the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -33% -28% -1.5% -22.9% 64 TABLE 7-7 PER CENT OVERLOADS USING THE PJM METHOD Size of Transformer: 10 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 180 159 157 197 184 181 Size of Transformer: 15/20/25 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 204 186 179 218 201 195 Size of Transformer: 150 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 155 148 145 167 161 158 Size of Transformer: 300 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 166 154 150 178 166 163 65 <3 m o o a) •Q -I > o -L -55 I I I 4J a •o CO ■co 60 3 < H O PM CO b o o o 3 e 3 JJ CO o H XI & 4-iS s g £ CO OS s wCO u CO § +> c o u o a. 04 w 3 4J en CO u cd 3 U |j _ E3 T § M cd IF* eg m a» iH o CT> i-l E CO CO e U «H CO J3 „ CO u CO 3 J-> JJ c cd co JJ e D- S "H BOA co B u u id U 60 60 L o o <u co cd p. a. u C -rt C CO IS -H •H ■»» >» •O >, M O 00 rH IT) CO rH m r^ rH o t-» rH S cd cd < Q > I I I <Hiau CO pBox^a^o ^uso JC3d 66 CO CO > > > u u o u3 os o < pq o i CD ' 'Q ■L L. 1 1 -1 1 > • u. * o ■•53 1 , ■u « u ■o L; a. CU •co E-H O PL) CO |1 k 1 ■ • . H O m o CM a u o o O 3 • t-i i •>-3 3 c ft 3 *-} 00 1 It 1 o w r --1 ; 1 i• s 1 -i 1 , .c c . o M $ J. 1 8 <U co a; a. 0) AJ B P a> 1 I o co <N m CN CM i ^T I o e 3 CM CM o rH CM h*4 1^* f* 6 co £ a) B u u nt AJ iH CM 1 i in o CM < « CJ 1 O CM o ■ <U PEOIJS^O J3 00 00 CD 53 -H 'i m OT CD J-4 3 4J £ (U I M M 3 4J CO M ai ai d —». : *~i (U ^ r 1 o cu a) e2 S to - Ambi - Day/ - Vary S CO v r-».~ iri H 00 3 '< m ratu lent nt t aM H> ^U3D asd 0) CO > > > 1-1 Lj (J 3 3 3 CJ o o 67 <J PQ o u CO o I > o ■ss u o CO to (SO H o PM CO 3 H o w u o o o m 4J CO .1 iH oi o Pn 55 CO o J3 «?— / i C 0 s u o a. a) 2 • •-1 PU CO CO u 3 J-> cd CO co m 3 u u - H I) 01 14 p. p. M' B CO n ii a OWE M CO 3 4J 4J O a •8 JJ 3 ca co w M -H P CO X> CO a- g -H B to ^ co e u u *-» id *J C I) •H •O "5) -H 2 -v. >> 60 C -H >, U ■ 6 ca io < o > o vo i i i CJ CO CO PBOIJ3A0 ^-USO 03d > > > u u u 3 3 CJ o o 68 <J M CJ I a cu Q •I > o '23 4J O a, co CO s 00 H 1% E-l O •-I CO 3 EH O W — .' cj o o o o "?. i.J J5 -p 4"«s o u CO I Q. CO Pi a w • co >-i u (0 3 CU n 3 •U tl j< _ _J-cd co co cd ft a M B co 0) Ct] a B c o u u o 4^ CO CO P. CO -V u C CO CO - J3 — C CO a. B -H B co ^ co B u u id .e 4-1 CO ■CTJ CO 55 •H — J* >> bO (J «H >» U E ca rt < a > o i i ■< m cj pecrpraAO 3-USD ae<j 69 co co cu > > > VJ >J K 3 33 CJ CJ CJ M u a o > ._ o I, o o •» I T> 0) CO H tJ 00 3 H O fU w 3. H o W o o o J-l l-l 00 x: 4-.S pi o w 1 S3 Pi c o s o Q. eg <£ S oo cu •-} u 3 w cu Q) p. co cfl M a) a, M 3. .M M « 4J __ w u sm •» _ _ o w X H CO CU H 3 a) a. E cu 4J c cd cu U «H cu a, 6 cu u xi B rt u .fi a cu _ -H Xj B nj W 00 60 CMC CU 55 -H •H >> B co a < a > m co o 00 m r^ o m vo t>« O vo peotaeAo ^u^D zsd 70 >n m i i i <J « u cu cu cu > > > u u u 3 3 S3 uoo m <J o in / Xi o m +1 c o S g u u o cu <u erf 3 (0 CU •-3 H 3 P-l 4J . u _ cu cu cd CU Cu U CO o CO 0) 3 W ct) a <u u w c cu »-i -H 0) J3 2* s B to cu cu § ci cu ja S U U It .A AJ 60 CO CTI (! CU 55 TH •H *v. >% ,0 >> M E ft) rt < Q > o H <N m o <N o o CM IT) O o» CT> rH H m 00 f-i o 00 rH I CU PBOXJ9AQ 1-USD J3<3 I I <J « O CU CU > > > (-IMP 3 3 ^ oo o 71 < pq u CO •Q > o •S5 o •o P. CO 'CO to 3 H H O i CO 3 EH O c3- CJ o o o 00 H 4J Pi o J.J9 B CO 9 P4 o W CO ■■ i• i H CO C o u u o o. CO a! S w i ■ . i _ 3 co CO M 3' (Q M M 4J CO CO Cfl a. a. M 0) B CO CO CO V u .i i 0) i e -0J 3 4J AJ C CO U co 0. CO ■u -H ja co B «H c 6 co .n co s u 4J d j= 4J 60 60 C 2j C i CO !S iH •H ^ S>s J3 >> U .Lj e «j ta o o m < Q > I I I <3 « c_> CO CO peoiaaAO H-USO J3d CO > > > M M U 3 3 2 ooo 72 <J PQ JJ u o a- o o 0) OS en ai o S CD a) • ^ M (0 3 CU cu JJ M CTJ 3 n U 4J a) <u rt a. a. m cu cu <u ■V s a) n 3 u « U c CO a> u Vi •H c 0) -a ai a. 6 •H s CO .O ai S u ■u CO CO o S3 4J e <u •H .O 73 z >* >> u S co a -»J M CJ <U peoxoaAO V&D JSd 00 CO •H c •H (U CO > CU > > > u M u 3 9 Si CJ O CJ o a; a O pq / t > o a o a) CO 60 3 <S H O "Hfl CO W o o o JJ 00 Lm Pi o C O o a. <u ai 3 to <u CO • vcjj o ^ M 3 CU 4J . * a) a) B to a) o) v n 3 (0 3 4J cd a> a. e a) ft. ft. M ■8 If If* BJ a) w <U J3 QJ ft. B -H G tfl ^ 0) B u « a x •U 00 C «H 0J Z •H-^ J3 >> 00 B -H >. M H a * £ < > m CO H o m o m O T-l i-» rH rH rH 00 r>« r» vo \o peoxjcaAQ V*3D J9d 74 m m iH i i i <! M U n o m > > > M M M 3 3 J O O O < < I I > O •25 l_ •ui u •o ■OT 60 3 *>-3 3 4J J3 O CM +> C O O. 0) o m « »-> U 3 r J 3 U • 4J CD P. cd M CO a. i a) n J ■3 4^ CO n 3 •u cfl u B (U 4J c CD c O. S «H 6 «i ,o u<u u a6 .fl AJ 60 GO m <J\ rH 11 o ON rH C <U •H U3 -H 55 -«» >. C iH >» H i i i e « « < a > m oo rH m r^ O OO iH rH o r^ \£> H I-I PBOXJ3A0 ^USD 393 m <J « o <u > u 3 a) a) > > u u 3 u O o 75 < CJ a n 0) 1 1 1 1 > o 1 1 "— -^ ^— ._ 1 . 4J i U •o i 1 0) ■to 1 1 H O CM W 00 ■3 1 rH 3 1 o J 1 CO IT) o 1— B CO § o w 3 C 3 i J3 +> C O 2 i U 1 i u CO S i o i •8 1 1 i c 3 i J , o , m m VO rH i o m rH • M O CXi cu S ■^ CO 0J . - Ambient temperaltures p>er P J - Day/Night ambieint temp»eratui - Varying ambient terapeiratures w o o o <J « o rH iJEO [J9AQ V&D X3d Q) Q) 0) > S> r> H h M a 3 a ouo 7 76 <l «o o ! K > o 55 a Q U CO i, to o o o 4J 00 4-* & .C -P C O S 9 Pd i-i a. •K w f— - o w 2; W H (U u o p. & s 0} <D • ^ u co 3 a) CU 4J n cO 3 )-i u 4J a> QJ rt a. O. )-l E cu ro a) ex a) +» E 01 )-i 3 •u 4J iJ c tfl 0) 4J U •H e <u .O ai a B •H e CO XI 4J •u CO 4J 00 bO •H c Z •H 0) c a> •H m \o o \o o m ir> m peoxasAO 3-uso m «* JSZ o •* e •a S^ tfl Q 1 1 <! « QJ <u > > >» 1-4 1 > U o a) > n M CO 1 3 3 J3 o o o 77 TABLE 7-8 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER? 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Hot Spot Oil °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 117.0 *178.5 119.7 *178.3 117.1 *178.6 Feb 116.4 *179.5 120.2 *178.8 117.1 *178.6 Mar 117.2 *178.7 121.3 *178.5 117.1 *178.6 Apr 118.9 *179.5 123.6 *179.4 *124.7 178.1 May 123.9 *177.8 *124.4 177.8 *124.7 178.1 Jun *124.4 175.9 *124.3 174.9 *124.7 178.1 Jul *123.8 173.4 *123.8 172.5 *124.7 178.1 Aug *123.9 174.5 *123.9 173.5 *124.7 178.1 Sep 121.5 *178.2 *123.8 175.3 *124.7 178.1 Oct 120.1 *178.3 123.9 *178.7 *124.7 178.1 Nov 119.0 *178.6 122.6 *179.3 117.1 *178.6 Dec 118.0 *179.0 120.2 *178.4 117.1 *178.6 indicates Limiting Constraint 78 TABLE 7-9 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Oil Hot Spol °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan *123.9 178.2 *124.1 176.5 123.4 *178.7 Feb *124.5 178.9 *124.8 177.2 123.4 *178.7 Mar *124.6 175.6 *124.0 174.0 123.4 *178.7. Apr *124.4 172.6 *124.8 172.6 *124.1 167.1 May *123.9 168.5 *124.6 169.7 *124.1 167.1 Jun *i*24.5 166.9 *124.2 168.6 *124.1 167.1 Jul *124.4 165.5 *124.3 165.6 *124.1 167.1 Aug *123.8 165.3 *124.5 166.5 *124.1 167.1 Sep *124.2 168.4 *124.6 168.3 *124.1 167.1 Oct *124.8 173.5 *124.0 170.4 *124.1 167.1 Nov *124.2 175.2 *124.5 173.6 123.4 *178.7 Dec *124.5 178.3 *124.7 176.6 123.4 *178.7 ^Indicates Limiting Constraint 79 TABLE 7-10 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Oil Hot Spoi °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan *124.0 175.4 *124.1 175.6 123.0 *177.0 Feb *124.7 176.2 *124.8 176.3 123.0 *177.0 Mar. *124.3 173.5 *124.0 173.1 123.0 *177.0 Apr *124.1 170.4 *124.9 171.7 *124.4 167.3 May *124.3 168.0 *124.7 168.8 *124.4 167.3 Jun *123.8 164.8 *124.2 165.8 *124.4 167.3 Jul *124.8 165.0 *124.3 164.5 *124.4 167.3 Aug *124.1 164.7 *124.6 165.7 *124.4 167.3 Sep *123.9 166.2 *124.6 167.5 *124.4 167.3 Oct *123.9 169.4 *124.1 169.6 *124.4 167.3 Nov *124.3 172.5 *124.5 172.7 123.0 *177.0 Dec *124.7 175.7 *124.7 175.7 123.0 *177.0 ^Indicates Limiting Constraint 80 TABLE 7-11 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 96.5 *179.2 119.7 *178.3 117.1 *178.6 Feb 94.5 *178.6 120.2 *178.8 117.1 *178.6 Mar 95.9 *179.3 121.3 *178.5 117.1 *178.6 Apr 97.4 *178.5 123.6 *179.4 *124.7 178.1 May 106.3 *178.9 *124.4 177.8 *124.7 178.1 Jun 108.3 /*178.8 *124.3 174.9 *124.7 178.1 Jul 109.1 *178.3 *123.8 172.5 *124.7 178.1 Aug 108.6 *178.4 *123.9 173.5 *124.7 178.1 Sep 101.8 *179.3 *123.8 175.3 *124.7 178.1 Oct 100.2 *178.3 123.9 *178.7 *124.7 178.1 Nov 98.6 *178.3 122.6 *179.3 117.1 *178.6 Dec 97.1 *178.3 120.2 *178.4 117.1 *178.6 'Indicates Limiting Constraint 81 TABLE 7-12 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Oil Hot Spot °C °C Method Varying; Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 109.8 *179.6 112.1 *179.2 109.1 *178.9 Feb 108.0 *178.5 111.9 *178.3 109.1 *178.9 Mar 112.3 *179.4 114.1 *179.2 109.1 *178.9 Apr 114.2 *178.6 116.0 *178.4 119.3 *179.3 May 118.4 *179.0 118.6 *179.1 119.3 *179.3 Jun 120.6 *179.3 120.2 *178.2 119.3 *179.3 Jul 121.5 *178.9 121.9 *179.3 119.3 *179.3 Aug 120.9 *179.0 121.2 *179.3 119.3 *179.3 Sep 117.6 *178.7 119.5 *178.8 119.3 *179.3 Oct 113.7 *178.7 117.2 *179.0 119.3 *179.3 Nov 112.2 *179.3 114.7 *178.5 109.1 *178.9 Dec 110.8 *180.0 112.7 *179.1 109.1 *178.9 ^Indicates Limiting Constraint 82 TABLE 7-13 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Jan 113.9 *178.3 115.7 *178.8 113.6 *179.3 Feb 114.5 *179.0 116,3 *179.5 113.6 *179.3 Mar 117.5 *179.4 117.8 *179.0 113.6 *179.3 Apr 119.4 *178.7 119.7 *178.4 123.5 *179.1 May 122.4 *178.6 122.6 *179.4 123.5 *179.1 Jun *123.9 177.6 124.1 *178.5 123.5 *179.1 Jul *123.8 175.7 *124.2 176.7 123.5 *179.1 Aug *124.2 177.2 *124.4 178.1 123.5 *179.1 Sep 123.1 *179.2 123.5 *179.1 123.5 *179.1 Oct 119.2 *178.3 121.1 *179.1 123.5 *179.1 Nov 117.4 *17?.2 118.3 *178.2 113.6 *179.3 Dec 114.5 *178.3 116.2 *178.5 113.6 *179.3 Month PJM Oil °C Hot Spot °C c ^Indicates Limiting Constraint 83 TABLE 7-14 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/:Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C PJM Hot Spot °C Oil °C Jan 77.5 *178.5 82.3 *180.0 77.3 *178.2 Feb 76.1 *179.5 82.1 *178.5 77.3 *178.2 Mar 77.9 *179.0 84.4 *178.6 77.3 *178.2 Apr 80.9 *178.5 87.4 *178.3 91.0 *178.6 May 90.0 *178.6 90.8 *179.6 91.0 *178.6 Jun 93.1 *179.7 93.4 *178.8 91.0 *178.6 Jul 94.2 *178.5 95.2 *179.6 91.0 *178.6 Aug 93.5 *179.0 93.9 *178.4 91.0 *178.6 Sep 86.0 *179.1 92.0 *178.4 91.0 *178.6 Oct 83.4 *178.7 88.6 *178.3 91.0 *178.6 Nov 81.0 *178.6 85.9 *179.0 77.3 *178.2 Dec 78.7 *178.5 83.0 *179.5 77.3 *178.2 ^Indicates Limiting Constraint 84 TABLE 7-15 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Oil Hot Spot °C °C Method Varying; Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 86.5 *178.4 88.4 *178.2 85.5 *179.7 Feb 87.2 *179.2 89.8 *178.7 85.5 *179.7 Mar 91.1 *179.8 91.5 *179.0 85.5 *179.7 Apr 94.0 *179.5 94.4 *178.7 98.8 *178.9 May 97.7 *178.8 97.3 *178.4 98.8 *178.9 Jun 100.2 *178.3 100.0 *178.3 98.8 *178.9 Jul 102.0 *179.1 101.9 *178.9 98.8 *178.9 Aug 101.4 *179.5 101.2 *179.2 98.8 *178.9 Sep 98.1 *178.3 99.3 *179.4 98.8 *178.9 Oct 93.8 *179.2 95.7 *179.0 98.8 *178.9 Nov 91.1 *179.8 93.0 *179.5 85.5 *179.7 Dec 87.4 *178.4 90.0 *179.8 85.5 *179.7 'Indicates Limiting Constraint 85 TABLE 7-16 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night (1) Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM (1) Hot Spot °C Jan 89.2 177.5 89.8 *178.3 85.7 177.1 Feb 89.9 178.2 90.5 *179.1 85.7 177.1 Mar 92.3 177.8 93.0 *179.5 85.7 177.1 Apr 95.3 177.7 95.9 *179.2 99.3 176.8 May 98.2 176.9 98.8 *178.9 99.3 176.8 Jun 100.9 176.7 101.4 *178.3 99.3 176.8 Jul 102.4 176.7 103.4 *179.5 99.3 176.8 Aug 101.6 176.8 102.7 *179.8 99.3 176.8 Sep 99.7 177.4 100.8 *i80.0 99.3 176.8 Oct 96.2 177.7 97.3 *179.5 99.3 176.8 Nov 93.2 177.6 93.7 *178.3 85.7 177.1 Dec 90.1 177.4 90.5 *178.2 85.7 177.1 ■^Indicates Limiting Constraint (1) The limiting contraint for all months was the 5% Loss of Life Limit. 86 TABLE 7-17 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Month Day/Night Hot Spot Oil °C °C Method Varying Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 90.2 *178.9 94.2 *178.2 90.8 *179.7 Feb 88.5 *179.2 94.6 *178.7 90.8 *1,79.7 Mar 89.6 *178.3 96.5 *178.6 90.8 *179.7 Apr 91.4 *178.3 99.2 *179.6 101.5 *179.1 May 100.7 . *179.3 101.5 *179.1 101.5 *179.1 Jun 102.6 *178.4 103.5 *178.4 101.5 *179.1 Jul 103.9 *178.9 105.1 *179.2 101.5 *179.1 Aug 103.5 *179.3 104.5 *179.5 101.5 *179.1 Sep 96.4 *179.6 102.9 *179.6 101.5 *179.1 Oct 94.3 *178.3 100.2 *179.7 101.5 *179.1 Nov 92.7 *178.6 97.6 *178.9 90.8 *179.7 Dec 91.1 *178.9 95.4 *179.4 . 90.8 *179.7 *Indicates Limiting Constraint 87 TABLE 7-18 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Day/Night Oil Hot Spot °C °C. Method Varying Ambient Oil Hot Spot °C °C Oil °C Jan 102.0 *179.6 104.2 *179.1 100.8 *178.4 Feb 101.1 *179.6 104.8 *179.7 100.8 *178.4 Mar 104.6 *178.6 105.9 *178.2 100.8 *178.4 Apr 106.9 *178.3 109.2 *179.8 112.5 *179.7 May 111.6 *17.9..6." 111.5 *179.5 112.5 *179.7 Jun 113.5 *179.0 113.5 *179.0 112.5 *179.7 Jul 115.1 *179.8 114.5 *178.2 112.5 *179.7 Aug 113.7 *178.3 113.9 *178.5 112.5 *179.7 Sep 110.8 *178.7 112.2 *178.5 112.5 *179.7 Oct 106.1 *178.3 109.4 *178.3 112.5 *179.7 Nov ]05.0 *180.0 107.1 *178.5 100.8 *178.4 Dec 102.9 *179.6 104.6 *178.7 100.8 *178.4 Month ^Indicates Limiting Constraint 88 u PJM Hot Spot °C TABLE 7-19 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 180°C Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C Jan 106.0 *179.2 107.5 *179.8 104.8 *179.8 Feb 106.7 *179.9 106.9 *178.2 104.8 *179.8 Mar 108.6 *178.4 109.1^ *178.7 104.8 *179.8 Apr 111.0 *178.3 111.5 *178.6 101.5 *179.1 May 114.8 . *179.5 113.9 *178.5 101.5 *179.1 Jun 116.7 *178.9 116.1 *178.3 101.5 *179.1 Jul 118.5 *179.9 118.0 *179.3 101.5 *179.1 Aug 117.0 *178.4 117.4 *179.6 101.5 *179.1 Sep 114.9 *178.7 115.7 *179.6 101.5 *179.1 Oct 111.1 *178.2 112.7 *179.0 101.5 *179.1 Nov 108.9 *178.6 110.5 *179.3 104.8 *179.8 Dec 106.5 *178.8 107.9 *179.3 104.8 *179.8 Month *Indicates Limiting Constraint 89 PJM Hot Spot °C CHAPTER 8 r DIELECTRIC STRENGTH OF CONDUCTORS IN CONTACT WITH HOT SPOTS Test results from recent investigations 17 22 ' have uncovered a concern about the generation of visible gas bubbles from the thermal degeneration of cellulose insulation material that is in contact with hot metal. These tests have shown that the gas generated can reduce the dielectric strength of the insulating oil and therefore possibly lead to premature electrical failure of the transformer. When the transformer oil is heated, gases, such as hydrogen, methane, acetylene, ethylene, and ethane are generated. 27 33 ' Evidence that these gases are generated can be seen at temperature/ of 350°C and above. However, the investigations found that when cellulose (winding insulation) was immersed in transformer oil and heated, carbon dioxide, water, and carbon monoxide were detected at significantly lower temperatures. These tests also indicated that the dielectric strength decreased 5% for every 25°C increase in conductor temperature for temperatures up to 15D°C. Above 150°C, the loss of dielectric strength became significantly greater in one case dropping from approximately 22 kV to 13 kV for a hot spot temperature increase from 150 to 175°C. 90 The implication from those results is that the probability of experiencing a dielectric failure becomes increasingly greater as the temperature of the conductor is increased. Undoubtedly additional research will be performed to determine what effect this thermal deterioration of cellulose material at overload temperatures will have on the life of transformers. Should this research point the way toward greater concern with this decreasing dielectric strength, one may find it to one's advantage, on certain transformers where a very high degree of reliability is justifiable, to accept a more conservative philosophy on overloading and lower the hot spot limit to 150°C. As was seen in the previous calculations, the vast majority of the transformer overloads were limited by the 180°C limit imposed on the hot spot temperature. A decrease of this limit by 30°C will obviously decrease overload capability. However, what needs to be determined is 1) what a decrease in the allowable hot spot temperature will do to the overload capability when calculated by either the Day/Night or the Varying Ambient method and 2) if there is any advantage in using any one method in this particular case. 91 Consequently previously performed calculations were redone incorporating the lower hot spot limit of 150°C. Again, to cover the range of medium and large power transformers, the selection of the four units was used as before. Results and discussions are presented as was done in Chapter 7, page 46. All assumptions made previously, including load cycling, ambient temperature, etc. were used here as well. The following summarizes the tables and charts that are used in this discussion: A. Tables 8-1 through 8-7 inclusive (Pages 97 through 103) list the allowable overloads as determined by the various methods. B. Figures 8-8 through 8-12 inclusive (Pages 104 through 115) graphically display the overloads from the various transformers for a given time per transformer. C. Tables 8-8 through 8-19 inclusive (Pages 116 through 127) compare the limiting criteria preventing additional overloads for the various methods. The discussion of the results were again broken down into two parts. The first covered the results from the time 92 standpoint, while the second reviewed the results on a transformer size basis. For all transformer sizes, big gains with the Day/Night method were found in April, September and October. Overall, for the year as a whole, the Day/Night method resulted in a 1-20% more overload capacity than would have been expected with the PJM method. As anticipated, winter months showed token differences in favor of the Day/Night method, while the summer months showed a similar difference in favor of the PJM method. The vast majority of the difference was found in the spring and fall periods of the year. All limits to additional overloads were as a result of the 150°C hot spot limit being reached. The oil temperature, cooler in the winter than in the summer, varied within a small band width and never approached its limit. For the smaller of the four units, the oil varied from about 90-ll0°C while for the larger units the oil remained significantly cooler, usually varying in the 60-90°C range. The results obtained with the Varying Ambient method showed that a more significant gain in overload capacity can i be realized by using the PJM method. For the two smaller units the difference was about 55% in favor of the PJM 93 method, while for the two larger units, the difference was approximately 37%. Again, there were two months - namely April and October - where these differences were reversed by a small percentage. The 150°C limit was reached for all cases, thereby preventing additional overloads. The temperature variations for the oil and hot spot were very similar to that obtained with the Day/Night method and comments made above to that effect hold here as well. For the six hour case, very close correlation of results was seen when the overloads obtained with the Day/Night and the Varying Ambient methods were compared. However, both methods were outperformed by the PJM method, with the Day/Night results lower by 11-26% and the Varying Ambient results lower by about 27-37%. The decreases in the summer months were expected, but the consistent decreases in the fall and winter months were not. As before, the only positive months were April and October. However, the difference between the PJM method and either of the two new methods was relatively small. Allowing a longer time period for the emergency to persist, the temperature of the oil was higher than that 94 which was calculated in the two hour case. At no time, with any of the methods, did the temperature come close to reaching its prescribed limit. the hot spot limit. All overload cutoffs were because of As with the two hour case, the oil temperatures for two smaller units were higher than for the two larger units. For the Day/Night method, the pattern setforth in the six hour case was continued in the ten hour emergency case. The overload capability was less here than in the six hour case for all transformers, with a range of 20.8-37% in favor of the PJM method. Again, the three months of April, May and October reversed the.trend. However, the remaining nine months were so negative so as to overwhelmingly cancel out these gains. As in the six hour case, the Day/Night and the Varying Ambient methods yielded results that are very close. Oil temperatures in this case were higher than in the six hour case. However, in all instances, the hot spot limit was reached before the oil temperature was even close to its limit. The oil temperature for the larger units again remained cooler than for the smaller units. 95 In the discussions on the trend of the overload capabilities (with the hot spot limit set at 180°C) it was noted that the two new methods exhibited opposite performance. The same was also found to occur when the hot spot limit was set at 150°C. With the Day/Night method, overloads were more favorable for short time periods while with the Varying Ambient method, overloads were changing such that, while still indicating less allowable loading than was obtained with the PJM method, nonetheless showed decreasing differences as the time period increased. The same reasons apply here as before. In addition, the repetition of the large margin of difference during April and October was the same as was shown for the 180°C case. Similar comments apply here as well. 96 TABLE 8-1 PER CENT OVERLOADS WITH THE DAY/NIGHT MODEL 2 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Month 10 MVA Transformer Size 150 MVA 15/20/25 MVA 300 MVA January 173/0 190/0 151.1/-0.9 158/1 February 175/2 193/3 153/1 160/3 March 172/-1 187/-3 148/-4 155/-2 April 168/14 183/10 145/7 152/8 May 156/2 174/1 139/1 145/1 June 152/-1 171/-2 136/-2 142/-2 July 150/-4 170/-3 135/-3 141/-3 August 151/-3 170/-3 135/-3 141/-3 September 162/8 178/5 141/3 148/4 October 165/11 184/11 146/8 153/9 November 168/-5 186/-4 148/-4 155/-2 December 171/-2 189/-1 150/-2 157/0 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: +20% . +14% +1.1% +14% The net overload gain or loss was determined by summing the monthly per cent differentials when results from the above method are compared to those of the PJM method. This difference is shown after the slash (/) above. 97 TABLE 8-2 PER CENT OVERLOADS WITH THE DAY/NIGHT MODEL 6 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Transformer Size 150 MVA 15/20/25 MVA Month 10 MVA January 162/-3 179/-1 145/-3 February 161/-4 179/-1 145/-3 March 157/-8 174/-6 142/-6 146/-5 April 153/7 170/7 138.3/4.3 142.1/5.1 May 147/1 164/1 135/1 138/1 June 143/-3 160/-3 132/-2 135/-2 July 141/-5 158/-5 130/-4 131/-6 August 142/-4 160/-3 131/-3 134/-3 September 146/0 164/1 134/0 138/1 October 152/6 170/7 138/4 . 143/6 November 156/-9 174/-6 141/-7 146/-5 December 161/-4 178/-2 144/-4 149/-2 . 300 MVA 150/-1 149.1/-1.9 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -26% -11% -18.7% -12.8% 98 TABLE 8-3 PER CENT OVERLOADS WITH THE DAY/NIGHT MODEL 10 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Transformer Size 150 MVA 15/20/25 MVA Month 10 MVA 300 MVA January 159/-5 174/-3 143/-4 February 159/-5 173/-4 143/-4 146/-3 March 155/-9 169/-8 141/-6 143/-6 April 151/6 165/6 137.2/4.2 140/5 May 146/1 160/1 134/1 136/1 June 142/-3 157/-2 131/-2 133/-2 July 140/-5 158/-4 129.1/-3.9 131/-4 August 141/-4 156/-3 130/-3 132/-3 September 144/-3 159/0 132.3/-0.7 135/0 October 149/4 165/6 136.1/3.1 139/4 November 154/-10 169/-8 December 158/-6 173/-4 146.2/-2.8 139.2/-7.JB 142/-7 143/-4 146/-3 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -37% -23% -27.1% -20.8% 99 TABLE 8-4 PER CENT OVERLOADS WITH THE VARYING AMBIENT MODEL 2 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Transformer Size 150 MVA 15/20/25 MVA Month 10 MVA 300 MVA January 166/-7 183/-7 147/-5 ,*- 153/-4 February 166/-7 183/-7 147/-5 152/-5 March 163/-10 180/-10 145/-7 150/-7 April 159/4 176/3 142/4 147/3 May 154/0 172/-1 138.1/0.1 144/0 June 150/-4 170/-3 135/-3 141/-3 July 148/-6 167/-6 134/-4 140/-4 August 150/-4 168/-5 135/-3 140/-4 September 152/-2 171/-2 137/-1 142/-2 October 157/3 175/2 140/2 146/2 November 161/-12 178/-12 143/-9 149/-8 December 165/-8 182/-8 146.1/-5.9 152/-5 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -53% -56% -36.8% -37% 100 TABLE 8-5 PER CENT OVERLOADS WITH THE VARYING AMBIENT MODEL 6 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Transformer Size 150 MVA 15/20/25 MVA Month 10 MVA January 160/-5 176/-4 144/-4 February 160/-5 175/-5 144/-4 147/-4 March 156/-9 172/-8 141.2/-6.8 145/-6 April 152/6 168/5 138/4 141.1/4.1 May 148/2 164/1 135/1 138/1 June 143/-3 160/-3 132/-2 135/-2 July 141/-5 158/-5 130/-4 133/-4 August 142/-4 159/-4 131/-3 134/-3 September 145/-1 162/-1 133.1/-0.9 137/0 October 150/4 166/3 137/3 140/3 November 155/-10 170/-10 140/-8 143/-8 December 159/-6 174/-6 143.1/-4.9 147/-4 300 MVA 147.1/-3.9 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -36% -37% -29.7% -26.8% 101 TABLE 8-6 PER CENT OVERLOADS WITH THE VARYING AMBIENT MODEL 10 HOUR EMERGENCY, 150°C HOT SPOT LIMIT Transformer Size Month 10 MVA 150 MVA 300 MVA January 159/-5 173/-4 143.3/-3.7 146/-3 February 159/-5 172/-5 143/-4 145/-4 March 156/-8 169/-8 141/-6 143/-6 April 151/6 165/6 138/5 140/5 May 147/2 161/2 134.1/1.1 136/1 June 142/-3 157/-2 131/-2 133/-2 July 140/-5 155/-4 130/-3 August 141/-4 156/-3 September 144/-1 159/-0 133/0 October 149/4 163/4 136.1/3.1 138.1/3.1 November 154/-10 167/-10 139.2/-7.8 142/-7 December 158/-6 171/-6 143/-4 145/-2 15/20/25 MVA 131.2/-3.8 130.2/-2.8 132.1/-2.9 135/0 Net overload gain(+)/loss(-) for the 12 months: 10 MVA: 15/20/25 MVA: 150 MVA: 300 MVA: -35% -30% -24.1% -21.6% 102 TABLE 8-7 PER CENT OVERLOADS USING THE PJM METHOD WITH HOT SPOT LIMIT AT 150°C Size of Transformer: 10 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 15*% 146% 145% 173% 165% 164% Size of Transformer: 15/20/25 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 173% 163% 159% 190% 180% 177% Size of Transformer: 150 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 138% 134% 133% 152% 148% 147% Size of Transformer: 300 MVA Emergency Period Summer Winter 2 hours 6 hours 10 hours 144% 137% 135% 157% 151% 149% 103 rU L u a 1 > o .1.—, u a o o o o m 60 tH W < O 3 s" §M W H 5i § O O1 P- W H 43 4— W 05 o en i +> C O 4-1 M O C oi as • (0 01 • u. CO 3 a 0- u u CO 3 u M JJ o> 01 cd c a. M »-j a. § s w CO 01 § 03 <U M 3 JJ B 01 4J 01 Q. B 0) JJ JJ e « 0) P •rl 01 03 JJ c a. E s a 01 •H 03 u CO 01 4J B 4J CO CO •H c 01 2 •H ■H *s^ >% jQ S-. l-l CO Q c 1 o CO m r^ o i»» m o NO VO in m o m < a 01 PBOXO3A0 ^U9D JSd 104 a 0) > CJ 01 > > > )-i u 3 3 ■s O U u IJ u <u <J PQ Q > o •55 a o P. 00 3 '1 3 n o & ai s r* £ _l .a a) ►-J .H a) a n <D H 3 to a) M 9 « M 4J a) a) ah a <u u a •w 0) M 3 . 4-1 B '1 <g a> 4J u -H c _ d) .n o) Q. B «H S «} J3 a) B u u d U 60 to C 2j 0J 55 •H^» .O !*. C «H >•, H Emm < Q > o o CVJ m o> rH O a\ fH m o 00 ■H CO iH m i^ r-i peoijraAO 3-uso JSd 105 o f*. T-l I I I <5 m u ai ai m > > > l-i H JJ 33a U U CJ u u <5 « '1 "Q • ■ — 1i i i > o i I ■ a •o i 1 , a. u 'CO o o 60 ' i1 a i Pn o i o ■ < j r-4 1 CO s '1-J 1 1 J H 1 CO o § 3 >* .S w i' 8 W _ ]"*"■" i ii i ■ 1 l | ' 1i h 0. , ' ■•< u 1 U3 , ! I . 1 Ll * ^""l 1 i O 35 1, 1' t f* c CO cu G O ai CK •-3 u tn 3 0) P< *J u CO 9 U >-i -U CU (!) CO p. a. M S CD 10 oi a 0) -W E P CU 3 4J 4J JJ s 3 11 U W "H c C -H C •H — >, .fl Ss M i m u P. 6 -H 1 L o • u o .1 H g CO c 3 ■—i i Pu t tem ght a g amb en l oo i ©mom m sjsr en iH H iH H ' o 3 1 S CO CO i i i < a >. urve A urve B Urve C PA UUU 106 M <J o O (U Q 11, I > I I I— AJ o o CO 00 a Jj O O XI -p +•3 *C4 B. CO 1 u s . -1 w u cr o s AJ l-l O o. (U ptf S co • cu u CO a cu *i pti AJ H <U +• 3 AJ AJ B a) cd 3 vi M AJ <U cu CO a. a. u a ai CO eg a. u § •S I I to AJ I » on C •H ai 55 •H XI J_4- o VO e « a> AJ u •H a cu J2 cu a. e •H s CO J3 <u 6 AJ AJ I 1 AJ o IT) 3 psoxjaAO ^-uao J3<3 107 o CO M c •H >% >> 1a Ul CO > <3 PQ o ai o CU > u 3 CJ CO > > M n 3 3 o o m u u CU Q > o 52 4J o o CO o o o m 60 3 rH < o J H rJ g 3 4J CO H o Pi o o s B o 9 tf S ° § CO c o PC rl o a. cu OS- S to cu ^ n - 3 CO u rt 3 <u co nj E CU CO CU D. OWE l-l CU J3 a) lfX4 H CO 3 *J *J 4J B « 0) u U -rl C cu .n cu a. E iH E CO £i cu E u u d *■* c iH c <U 53 tH •H -^. >, o r»» rH XL «T) VO rH O VO rH .O m o m rH lO rH »n sf rH o sf r^ S>> rl Bum < o > t I I <! MU cu CU PBOTTOAO 108 V&D aad CU > > > Ul M M 3 3 3 U O O u eg a .1 > "1* 1 a o •1 '1 a. eg co j • o o ■ i i o i• a H ►J i H O (U o eg H O a 1 u 1 >. +£ 1 <-i Pi 1• S5 PS 1S3 3 : i CO evj 3 < 1 1 * !*■ H CO r I CO CO u 0. 4J Ct) H s co co >-i 3 cd ■ g u u 3 ! o w M O P. eg PJ 1 ■8 1 j-i cd M eg O. S eg 4J CO § C CO ,n B co J-l 4-1 c CD •H •§Ct) # U 60 60 1_ •H-^ >■> C (D 53 iH CTJ s « to < Q > o CO rH m r-. H O r» rH m O <£> rH VO rH m m iH o a I I I <J PQ U 0)0)0) PBOIO9A0 V&D ^9d 109 > > > H M rl 3 3 2 O O O pq O u 0) Q > o 'S3 n 4J a o a. ai to 60 1% ^ a JJ VI J5 ■P •C o in o S Pi O u CO I u CO S o a. a> PSIS to 0) M • ca 3 a) CU 4J M CO 3 t-i M u QJ a) CO ex a. M B CD CO QJ a. CD AJ B a) H >-J 3 4J 4J -3 IF* « co U a) a. C ai 4J d .n a) E •H B 0) 4-I U Ti CO ja & 00 4J C -H B cfl to c a) z: •H 15 > |I •H ^ >. J3 >> M 1 o 1 1 1 cfl en peoxaeAO 3-U3D asa a) a) a> > > > M U >-i 3 3 .3 CJ O U 110 <! pq O CJ ai a > o +55 T\ 4J a +o a. a) CO o o m 60 3 < o H 3 H O P-, 4J U CO J3 o o f> Pi o CO a *S B o s o n. co cd 2 co cu •-3 a S CU U 3 4J 3 u ctf M u CO ex, o a) a. B CO a) g o CO 3 CO M CO CO CO U •U CO M CO co D. E •rl S 03 f CO 4J 00 bO C 2j C CO 25 -H •H-v. >, 1 o SO rH m m t-H o m iH .a >* u E td cd < a > m «* o sr iH i-H m en rH PBOXO9A0 ^U3D ^3d o co rH i I i < « o 0J CO CO > > > )-i M u 3 3 3 uuu 111 a O pq CO il 1 11 > o •23 i 1 i i U •o i i . 1 p. <u 1 cj o o m "CO (SO 3 L o r H H O W o i i 1 S3 w c >% *-> , u 9* •< (0 s i o ■8 J* o I-■H m VO i-H c 3 ■■ i " O vo rH s p. • 53 CO P^ 4J at M 3 4J c Q) M •H O in r-l psoi: IT) siH O sr TH 4J c CO •H § •8 4J 4J 4J CO 00 S >% u cd > 1 1 1 •4 PQ CJ a) CD 0) > > > 3 3 .3 o 112 AJ a) .fl 1 <=>cd IT1 ir r- 3 s a) J3 i CO CD M u u 4J CD a) cd p. a. u B CO (0 ai o. P< 6 cd 1 1 1 +> c o H i 1 • u u o 3 1j 1 i 3 ■*->- i i g CO iH '^ i CO o i * PM '< CJ CJ ■"">;>* • ,,'"■, ,v pq j o u a) i •Q 1 1 > o i •» I" • i. 4-1 a ■o a, 1l i 1II o o m rH | O 1 H O CM o cs m CO ■■§ 1 11 I-l 3 s iT • fi ll >* •iS «C4 I 1 1 J-l a. • § "< 1 w o ! 1 i i i u CO s 0) J* i I —l C • A' i -ww 1 m 00 o 00 J 1 1. m r» o r«» .C -P £7 O 4J M O 01 0) • ^ K K 3 01 P-4 *J U cfl 3 M 4J a) a> cd p. a M B 0) co a) a. M U ri c tljQ 01 a. e *H E co .a ai s co *J *J U C 0) •H 60 -H 25 ^ M C «H >, ,o >-. W B co cd 1 1 • 3 ■■*-> i ■ _ H O W O 3 00 • 1 PL, CO H o 1 ', i M hJ m ai "CO ature ent t t tern < 1 ir\ *£> o vo yeoixa/iQ iv&o ja& < Q > ' m m 1 1 1 -*« PO r^ a> a> a) > > > n u u ^ 3 =1 ooo 113 u O pq <! Q > O Li. 23 4-1 o O P. 0) CO 00 4J J3 o m >1 H Pi o rl CO o ct) a o •a 0) IF* +> C O U O P. a) ai X CO at *-> M CO tu rl m m H o m rH m rH m en rH PBOIJ9A0 ^Uao J3d 114 o n rH c 4-1 4J s a) .O O e (U 4-) 0) •H 0) •H sr sjrH 3 n •H a) ^3 a. E G vO tH CO CU M 3 4-1 4J 4J trt 4J o 4-1 • ai u u 4J ■a 0) cd p. a. M e 0) CO cu a, a> •p e 6 at I 3 CO J3 to 00 •H z C •H >* U >% 1 Q > 1 1 1 CO CO <J « u <u 0) 0) > > > u ^ u 3 3 s CJ O o u CU 'O I > o -23 'HI U •O a, CU •co o o o 60 13 3 L'i p* o 3 t ,1 H O C 3 CO o o co a o PH CO o H *>-> 4-J? § 6 3 x: +) c o o cu a) 05 S3 w •-> u co 3 cu u rt 3 . M M _ u w CO s co a co cu u 43 0) cu a B a) -u co n cu a, 6 CU 3 4J u 4J C d ai w CU 43 CU a e ri B c .CO o to rH rH VO m H a cu CO 43 B ■u JJ co 4J "eo so C iH C cu 2: in •H >» 43 i*. U S (0 CO < Q > m -* O -* rH rH m PI rH PBOXOSAO ^U3D J3d 115 O III rH <5 m CJ CO cu > u 3 U cu > TABLE 8-8 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying; Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 99.7 *149.2 102.3 *148.7 99.8 *149.4 Feb 98.2 *148.6 102.7 *149•2 99.8 *149.4 Mar 99.7 *148.8 104.2 *149.3 99.8 *149.4 Apr 101.5 *148.9 106.1 *149.5 107.7 *148.9 May 107.4 *149.5 . 107.4 *148.6 107.7 *148.9 Jun 108.9 *149.3 109.1 *148.6 107.7 *148.9 Jul 109.9 *149.4 110.1 *148.8 107.7 *148.9 Aug 109.3 *149.3 110.0 *149.6 107.7 *148.9 Sep 104.3 *149.0 108.4 *148.8 107.7 *148.9 Oct 102.8 *148.8 106.6 *149.1 , 107.7 *148.9 Nov 101.4 *148.7 104.9 *149.1 99.8 *149.4 Dec 100.1 *148.7 102.9 *148.9 99.8 *149.4 ^Indicates Limiting Constraint 116 TABLE 8-9 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C ; Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C Jan 104.7 *149.4 105.4 *148.9 102.7 *148.7 Feb 104.5 *148.7 105.7 *149.0 102.7 *148.7 Mar 106.7 *149.2 106.5 *148.5 102.7 *148.7 Apr 108.9 *149.7 108.7 *i49.1 111.6 *149.5 May 110.4 *148.7 110.9 *149.7 111.6 *149.5 Jun 112.2 *148.8 111.9 *148.6 111.6 *149.5 Jul 113.2 *149.1 113.1 *149.0 111.6 *149.5 Aug 112.5 *148.8 112.3 *i48.6 111.6 *149.5 Sep 111.3 *149.2 111.2 *148.7 111.6 *149.5 Oct 108.6 *149.0 109.2 *148.8 111.6 *149.5 Nov 106.7 *148.8 108.1 *149.8 102.7 *148.7 Dec 105.4 *149.7 105.8 *149.1 102.7 *148.7 Month ^Indicates Limiting Constraint 117 PJM Hot Spot °C TABLE 8-10 TABULATION OF TEMPERATURES FOR 10 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C Jan 105.4 *148.7 105.4 *148.7 103.1 *148.7 Feb 106.0 *149.4 106.0 *149.4 103.1 *148.7 Mar 107.0 *148.7 107.8 *149.9 103.1 *148.7 Apr 109.3 *149.3 109.1 *149.0 112.2 *149.7 May 110.9 *148.8 111.4 *149.7 112.2 *149.7 Jun 112.8 *149.1 112.4 *148.6 112.2 *149.7 Jul 113.9 *149.4 113.5 *149.0 112.2 *149.7 Aug 113.2 *149.1 112.8 *148.6 112.2 *149.7 Sep 111.8 *148.9 111.6 *i48.7 112.2 *149.7 Oct 109.5 *148.7 109.6 *148.7 112.2 *149.7 Nov 108.4 *149.7 108.5 *149.7 103.1 *148.7 Dec 106.1 *149.1 106.0 *149.0 103.1 *148.7 Month ^Indicates Limiting Constraint 118 PJM Hot Spot °C TABLE 8-11 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spoi °C °C Method Varying Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 86.9 *148.9 91.1 *148.7 87.9 *149.9 Feb 85.1 *149.1 91.4 *149.0 87.9 *149.9 Mar 89.1 *149.2 93.4 *149.1 87.9 *149.9 Apr 91.2 *148.8 95.4 *148.8 97.3 *148.9 May 96.7 *148.8 97.6 *148.5 97.3 *148.9 Jun 98.8 *149.2 100.1 *149.9 97.3 *148.9 Jul 100.0 *149.8 100.9 *149.0 97.3 *148.9 Aug 99.2 *149.2 100.2 *148.9 97.3 *148.9 Sep 94.4 *148.9 98.9 *149.3 97.3 *i48.9 Oct 90.7 *148.9 96.5 *149.2 97.3 *148.9 Nov 89.1 *148.6 94.2 *148.7 87.9 *149.9 Dec 87.9 *149.3 92.1 *149.1 87.9 *149.9 ^Indicates Limiting Constraint 119 TABLE 8-12 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night . Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 93.9 *149.4 96.3 *149.6 93.5 *149.2 Feb 94.5 *149.6 96.2 *148.9 93.5 *149.2 Mar 97.4 *149.6 98.4 *149.2 93.5 *149.2 Apr 99.7 *149.5 100.6 *149.3 103.7 *149.5 May 102.7 *149.1 102.8 *149.2 103.7 *149.5 Jun 104.5 *148.7 104.8 *149.0 103.7 *149.5 Jul 105.6 *148.7 106.0 *149.1 103.7 *149.5 Aug 105.5 *149.7 105.2 *148.9 133.7 *149.5 Sep 102.9 *149.3 104.0 *149.3 103.7 *149.5 Oct 98.9 *148.6 101.3 *148.7 103.7 *149.5 Nov 96.8 *148.9 99.2 *149.0 93.5 *149.2 Dec 94.6 *149.1 96.8 *149.0 93.5 *149.2 ^Indicates Limiting Constraint 120 TABLE 8-13 TABULATION OF TEMPERATURES FOR 15/20/25 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 97.3 *149.4 97.9 *149.5 95.8 *149.7 Feb 97.1 *i48.6 97.7 *148.6 95.8 *149.7 Mar 99.5 *148.7 100.0 *149.2 95.8 *149.7 Apr 101.8 *148.7 102.4 *149.3 105.7 *149.4 May 104.6 *148.7 104.7 *149.4 105.7 *149.4 Jun 107.3 *149.8 106.7 *149.3 105.7 *149.4 Jul 108.4 *149.8 107.9 *149.4 105.7 *149.4 Aug 107.7 *i49.7 107.1 *149.1 105.7 *149.4 Sep 105.2 *148.8 105.9 *149.5 105.7 *149.4 Oct 102.5 *149.4 103.0 *148.8 105.7 *149.4 Nov 100.3 *149.5 ,100.8 *148.8 95.8 *149.7 Dec 98.0 *149.6 98.3 *148.7 95.8 *149.7 ^Indicates Limiting Constraint 121 TABLE 8-14 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying; Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 67.8 *148.5 72.1 *148.7 68.1 *149.8 Feb 66.1 *148.8 72.6 *149.2 68.1 *149.8 Mar 71.2 *148.8 75.1 *149.7 68.1 *149.8 Apr 74.3 *148.9 78.1 *149.9 81.6 *149.4 May 80.4 *149.2 80.8 *148.6 81.6 *149.4 Jun 83.2 *149.1 83.6 *148.5 81.6 *149.4 Jul 84.9 *150.0 85.4 *149.5 81.6 *149.4 Aug 83.7 *148.6 84.6 *149.6 81.6 *149.4 Sep 78.9 *149.7 82.6 *149.4 81.6 *149.4 Oct 73.9 *149.6 79.0 . *148.7 81.6 *149.4 Nov 71.4 *149.0 76.0 *148.6 68.1 *149.8 Dec 69.0 *148.6 72.9 *148.6 68.1 *149.8 ■^Indicates Limiting Constraint 122 TABLE 8-15 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 74.3 *148.9 75.2 *148.8 71.9 *149.5 Feb 74.9 *149.6 75.8 *149.5 71.9 *149.5 Mar 78.0 *149.6 77.8 *148.6 71.9 *149.5 Apr 80.6 *148.7 80.9 *148.6 85.7 *149.8 May 84.5 *149.5 84.1 *149.1 85.7 *149.8 Jun 87.3 *149.5 87.0 *149.3 85.7 *149.8 Jul 88.5 *148.8 88.3 *148.7 85.7 *149.8 Aug 87.7 *149.0 87.5 *148.7 85.7 *149.8 Sep 85.6 *149.7 85.4 *148.6 85.7 *149.8 Oct 81.4 *149.2 82.3 *149.2 85.7 *149.8 Nov 78.3 *149.0 79.3 *149.0 71.9 *149.5 Dec 75.1 *148.7 76.0 *148.7 71.9 *149.5 indicates Limiting Constraint 123 TABLE 8-16 TABULATION OF TEMPERATURES FOR 150 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C PJW[ Oil °C Hot Spot °C Jan 75.7 *148.6 75.7 *148.6 72.3 *148.9 Feb 76.1 *148.7 76.1 *148.7 72.3 *148.9 Mar 79.1 *149.9 78.8 *149.5 72.3 *148.9 Apr 81.6 *148.6 82.1 *149-5 86.3 *149.4 May 85.0 *149.1 84.5 *148,6 86.3 *149.4 Jun i 87.9 *149.2 87.4 *148.6 86.3 *149.4 Jul 89.1 *148.7 89.5 *149.9 86.3 *149.4 Aug 88.3 *148.6 88.0 *148.5 86.3 *149.4 Sep 86.2 *148.. 6 86.6 *149.7 86.3 *149.4 Oct 82.7 *148.6 82.7 *148.7 86.3 *149.4 Nov 79.7 *148.6 79.7 *148.6 72.3 *148.9 Dec 77.2 *149.8 77.1 *149.8 72.3 *148.9 *Indicates Limiting Constraint c 124 ' TABLE 8-17 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 2 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Method Month Day/Night Oil Hot Spot °C °C Varying; Ambient Oil Hot Spot °C °c PJW[ Oil °C Hot Spot °C Jan 80.5 *149.3 84.9 *149.5 80.8 *148.7 Feb 78.7 *149.2 84.9 *148.6 80.8 *148.7 Mar 82.8 *149.1 86.8 *149.0 80.8 *148.7 Apr 85.2 *149.0 89.3 *149.1 91.7 *149.2 May 90.9 *149.1 91.8 *149.2 91.7 *149.2 Jun 93.1 *149.0 94.0 *149.1 91.7 *149.2 Jul 94.4 *149.6 95.5 *149.9 91.7 *149.2 Aug 93.5 *148.6 94.4 *148.8 91.7 *149.2 Sep 88.8 *149.4 92.8 *148.7 91.7 *149.2 Oct 85.0 *149.7 90.3 *149.3 91.7 *149.2 Nov 83.3 *149.5 88.0 *149.4 80.8 *148.7 Dec 81.5 *149.5 85.6 *149.4 80.8 *148.7 ^Indicates Limiting Constraint 125 TABLE 8-18 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 6 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Oil °C Day/Night Hot Spoi °'C Method Varying Ambient Oil Hot Spot °C °C Oil °C PJM Hot Spot °C Jan 87.2 *149.4 88.8 *148.6 85.9 *148.8 Feb 87.2 *148.6 89.3 *149.0 85.9 *148.8 Mar 90.7 *149.7 91.5 *149.8 85.9 *148.8 Apr 92.6 *148.6 93.5 *148.7 97.2 *L49.3 May 96.1 *149.0 96.1 *148.9 97.2 *149.3 Jun 98.4 *149.1 98.4 *149.1 97.2 *149.3 Jul 99.4 *148.6 99.4 *148.6 97.2 *149.3 Aug 98.7 *148.6 98.8 *i48.7 97.2 *149.3 Sep 96.9 *149.8 97.8 *150.0 97.2 *149.3 Oct 93.0 *149.7 94.6 *149.0 97.2 *149.3 Nov 90.5 *149.'5 92.1 *148.7 85.9 *148.8 Dec 87.8 *149.2 90.2 *150.0 85.9 *148.8 "Indicates Limiting Constraint 126 TABLE 8-19 TABULATION OF TEMPERATURES FOR 300 MVA TRANSFORMER 10 HOUR EMERGENCY, HOT SPOT LIMIT OF 150°C Month Day/Night Oil Hot Spot °C °C Method Varying Ambient Oil Hot Spot °C °C PJM Oil °C Hot Spot °C Jan 89.5 *148.6 90.6 *149.6 87.4 *148.7 Feb 89.9 *148.9 90.2 *148.5 87.4 *148.7 Mar 92.4 *149.0 92.5 *149.2 87.4 *148.7 Apr 95.2 *149.6 95.4 *149.8 99.0 *149.7 May 97.9 *149.3 97.1 *148.5 99.0 *149.7 Jun 100.3 *149.5 99.6 *148.8 99.0 *149.7 Jul 101.1 *148.9 100.7 *148.6 99.0 *149.7 Aug 100.5 *149.0 100.0 *148.5 99.0 *149.7 Sep 98.8 *149.5 99.1 *149.7 99.0 *149.7 Oct 95.3 *148.9 95.8 *148.7 99.0 *149.7 Nov 92.7 *148.5 94.1 *150.0 87.4 *148.7 Dec 90.9 *149.9 90.2 *150.0 87.4 *148.7 ^Indicates Limiting Constraint 127 CHAPTER.9 SUMMARY After reviewing the results of the computations from the three methods, one can see that, in certain instances, a particular method may give an increased overload capability over that which can be presently obtained. However, as can be seen quite vividly, these gains are offset by lower than presently used overloads during other periods. These periods can be length-of-time periods or time-of-year periods. In the Day/Night method, the length of time that the emergency period persisted determined whether the method gave an overall net increase or decrease in the final allowable overload. For short periods of time, the total allowable overload for one year of the portrayed scenario was greater than with the existing PJM method. Increasingly longer periods of time reversed the result so that overloads became increasingly in favor of.the PJM method. When the results of the calculations during the various months of the year are reviewed, it is seen that there are certain times of the year when use of the Day/Night method results in a greater allowable overload. 128 It was shown that, with the Varying Ambient method, set so as to follow a conservative policy and yield safe results, the overloads that were computed for all three emergency periods proved to be lower than what would be allowed with the PJM method. Although the difference was quite large for the short time periods, the difference began to decrease as longer emergency periods were used. Although the Varying Ambient method gave the lowest overloads of the three methods, it must be remembered that the peak ambient temperature was adjusted within the load cycle so as to give the most conservative results. A calcu- lation was made to determine what change would occur if the peak ambient temperature would be placed more in the middle of the 1 per unit part of the load cycle. Calculations were made for both the 150°C and 180°C hot spot limited cases. The change in overload capability for the 15/20/25 MVA unit during the month of March can be seen on page 130. The calculations assume that the peak temperature occurs in the sixteenth hour of the load cycle. Hence, it can be seen that such a change in the positioning of the peak ambient temperature within the load cycle can substantially modify the overloads that are allowed. 129 Hot Spot Limit Set at 180°C Emergency Period Hours Overload % Change % 2 223 13 218 6 206 12 201 10 Ove:trload From PJM Method % 7.8 195.8 195 Hot Spot Limit Set at 150°C Emergency Period Hours Overload % Change % 2 199 19 190 6 186 14 180 10 178 9 177 Overload From PJM Method % The time of year that an emergency occurred produced different results as far as overload capability. The.PJM method showed that, the lower ambient of the "winter" months, as per the definition of winter in the methodology, gave greater overloads than the "summer" months. The other two methods used in this text also gave overloads that were greater in the months near seasonal winter than in the seasonal summer months. But not always was the sum total of all of the overloads for all of the months greater with the new methods. Rather, in most cases, the total of all the overloads proved to be less. 130 It was seen that there was a consistent repetition of several months throughout the year when the present method always yielded results that were far below that obtained with the other two methods. Most of these deviations were seen during April, followed closely by October, and, somewhat less consistently, in September and May. From the temperature calculations it was noted that, at the overloaded state, the temperature of the oil calculated with the PJM method during April, was similar to the oil temperature that, with the other two methods, was calculated during May. Yet, the average temperature during May is more than 5°C higher than in April. This was found to be true for both the 180°C and the 150°C hot spot limited cases. The implication is that, with the PJM method, the assumed ambient of 29°C could be unnecessarily high, (as compared to the actual ambient of 10°C) and therefore understate the allowable overload by too great of a value. It is believed that defining April as a "winter" month will have just the opposite extreme. It may be best to define a new category in the existing PJM program to handle this case so as to obtain the most optimal loading. An assumed temperature for this month of, say, 10-20°C would provide additional overload capacity for emergencies occurring that time period. 131 Although a similar adjustment could be made for October, the lack of any outstanding temperature deviations between the three methods would lead one to conclude that the heating values calculated by the PJM program are not out of line. Therefore the October calculations should remain as is. As was seen from the various tables, the most common cutoff in overloading was the hot spot limit of 180°C. For the smaller MVA units, the oil temperature unit was occasionally reached. It seems that because of the design of these units as well as the smaller quantity of oil, the temperature of the oil increased to a higher value for all months and during all three emergency periods considered. With the larger units, the oil temperature was lower and consequently the hot spot limit became the limiting constraint in most cases. When the lower 150°C hot spot limit was imposed, the limiting constraint became that limit. The effect of this lower limit was to impose a 15-25% across the board drop in overload capability of a unit. This drop was rather consistent for all units, all emergency 132 periods, and all months throughout the year. The oil temper- ature dropped as well, with the coolest temperature seen for the short two hour emergency, and warmer temperatures seen for the longer overload periods. The lower hot spot limit may have lowered the percent rating above nameplate, but, from the graphs, the relative position of the results for each month, by each of the three methods remain the same. In addition, the position of the curve for each method relative to one another also remained almost the same as in the 180°C case. The imposition of a new hot spot limit shifts the position of the curves along the vertical axis, but the position of the curves relative to one another changed little. Hence, no method gains much advant3ge or becomes more desirable just because of hot spot limit changes. With the hot spot limit being such a consistent factor in determining the maximum transformer rating, it was seen that real time monitoring of this temperature is highly desirable. At the present time., the hot spot is calculated by adding a margin over and above the average copper temperature. 12 What is needed is detection devices that can actually measure the hot spot temperature while the trans- 133 former is in service. Whatever is used must not, in any way, act as a foreign body that could allow the reliability of the unit to decrease. By installing such devices, the local hot spots that exist in the transformer can be carefully measured. The operator can then monitor the hot spot temper- ature on a real time basis and, knowing the actual temperature rather than some supposed derated safe temperature, could overload the unit to the highest thermal capacity that results when the actual hot spot temperature of the unit is closer to the chosen critical value. The problem up to this point has not been developing a technique for monitoring the temperature of the hot spot. thermocouple suffice. A installed at the hot spot location will However, with the hot spots located in high magnetic and electrical stress areas, the transfer of the thermocouple information from such areas to the monitoring points outside the transformer is difficult. Research is presently underway to develop reliable and economical techniques for determining and continuously monitoring this hot spot temperature. A number of attempts have been tried to relay the information from the monitoring point to the reading point outside the transformer. 134 Experi- ments with acoustic data links much success. 29 have been tried without too These attempts involved converting an electrical signal from the thermistors located on the hot spots into acoustic waves and transmitting the acoustic waves through the oil. External detection and conversion devices then yield electrical/temperature^readings. Another technique that holds greater promise is the spectral analysis of the acoustic variations of sensors that are implemented in the hot spot locations. 28 In this tech- nique, the sensor is imbedded in the turn of the coil, and is acoustically excited via fiber wave guides. Transducers outside the transformer convert the received acoustic signal into an electrical signal which, in turn, corresponds to a temperature. Laboratory model tests of this system have shown that the direct measurement that is taken is accurate to 0.01°C. Should such hot spot detection methods become practical for production units, the operator can then safely overload the unit to a higher capacity without fear of exceeding the hottest spot temperature. 135 At this point, several factors need to be mentioned. The analysis that has been performed has strictly pertained to just the core and coils (i.e. windings) of the transformer, immersed in oil and in a tank. To consider this as the complete transformer and overload according to the calculated ratings may cause some of auxiliary items to become dangerously overloaded and prematurely fail. Components such as bushings, cables, load and no-load tap changers, and current transformers need to be included in the overall analysis as to whether the entire unit can be overloaded as indicated by the calculations. In addition, the larger the MVA size of the unit, such as the 300 MVA unit in this study, the more important becomes the effect of stray flux. heating of metallic parts. Stray flux causes localized Large units that are exposed to high MVA loads may have stray flux heating levels that are so high that the insulation can be damaged and oil can be decomposed into gas. Therefore, picking the emergency loading that a transformer can be expected to carry must include a review of the capability of all of the components in the unit. 136 It should also be mentioned that once a rating method is determined, switching to another method for the sake of obtaining a higher rating is not permissible. Such a practice can result in excessive loss of insulation life, thereby possibly shortening the life of the unit. 137 CHAPTER 10 CONCLUSIONS The desirability of incorporating environmental factors into the methodologies for calculating permissible loadings above the transformer nameplate ratings was analyzed. The ambient temperature was determined to have the most pronounced effect on the overall rating. Two methods for incorporating the ambient temperatures into the calculations were devised the Day/Night method and the Varying Ambient temperature method. Computations were made and the results of those methods were compared with the results obtained with the present PJM method. The study also discussed the. results of recent investigations into the decrease of dielectric strength when the hot spot temperature approaches 180°C. Calculations were made to determine what effect a 30°C lower hot spot limit would have on the relative advantage of any of the three methods. As expected, certain time periods throughout the year yielded results that were in favor of one method over another. 138 For the Day/Night method, what was not expected was 1) the minimal total amount of overloading in favor of the Day/Night method over the PJM method, 2) the few times during the year that such overloading exceeded that presently allowed, and 3) the large deficit when longer emergency periods were involved. Although the Varying Ambient temperature method was adjusted to yield conservative results, the amount that these loadings underran those obtained with the PJM method was rather large. More advantageous results can be obtained by positioning the peak ambient temperature with respect to the load cycle that would more accurately reflect actual operating conditions. These results were true for both the 180°C and the 150°C hot spot limited cases. By comparing the temperature results from all three methods, the values obtained for the month of April with the PJM method seemed out of line with those obtained with the other two methods. It is suggested that consideration be given to placing April into a separate category so as to yield perhaps a more realistic value for the limit the 139 transformer can actually carry during a short time emergency condition. In all, the present PJM methodology portrays quite advantageous transformer loadings above nameplate. It is felt that, if necessary, it can be improved by considering the month of April as a separate temperature period of the year. Although the other two methods can be implemented to give greater overloads for any one particular time of the year, the individual characteristics of the transformer, the loading cycles, and the projected probability of. the occurrence of emergency situations at that particular time period at that individual substation, must be taken into account and analyzed on an individual basis before a decision can be made as to the merits of any one method. 140 BIBLIOGRAPHY Books 1. Transformers for the Electric Power Industry, Bean, R. L.; Chackan, N.; Moore, H. R.; and Wentz, E. C., McGraw Hill Book Co., New York, 1959. 2. Transformer Engineering, Blume, L. F.; Camilli, G.; Boyajian, A.; and Montsinger, V. M., J. Wiley & Sons, Inc., New York, 1951. 3. Magnetic Circuits and Transformers, Massachusetts Institute of Technology Department of Electrical Engineering Staff, J. Wiley & Sons, Inc., New York, 1962. 4. The J & P Transformer Book, Stigant, S. A.; and Franklin, A. C, Butterworth, Inc., Boston, Mass., 1973. Standards 5. American National Standards Institute, Guide for Loading Oil - Immersed Distribution and Power Transformers, Appendix C57.92, New York, 1962. 141 6. National Electrical Manufacturers Association, Guide for Loading Oil Immersed Power Transformers with 65°C Average Winding Rise, Publication Number TR98, New York, 19.71. Articles and Reports 7. "Determination of Power Transformer Ratings", Blake, J. H.; Bast, R. R.; Donaldson Jr., R. J.; Downs, C. L.; Mills, H. E.; Moberg, C. V.; Studeny, B. C; Wilker, S. J. ; Wood, D. C., Pennsylvania-New Jersey-Maryland Interconnection, 1969. 8. "Oil-Immersed Power Transformer Overload Calculations by Computer", Blake, J. H.; Kelley, E. J., IEEE Transactions on Power Apparatus and Systems, Vol. PAS-88, No. 8, pp. 1205-1212, August, 1969. 9. "The Loading of Transformer Units in Power Plants and Networks with Regard to Ambient Temperature", Nejedly, J., CIGRE, (12-01), August 30 - September 7, 1978. 10. "A Proposed Functional Life Test Model for Power Transformers", McNutt, W. J., Paper F-77-055-7 presented at IEEE PES Winter Meeting, New York, January 30*February 4, 1977. 142 11. "Loading of Substation Electrical Equipment with Emphasis on Thermal Capability", Conway, B. J.; McMullen, D. W.; Peak, A. J.; Scofield, J. M., Papers A-77-656-2 and A-77-649-7 presented at IEEE PES Summer Meeting, Mexico City, Mexico, July 17-22, 1977. 12. "Operation of Transformers at Loads in Excess of Nameplate Ratings", Chew, 0., Paper presented at IEEE PES Transformer Seminar, New Orleans, Louisiana, March 9, 1978. 13. "Loading Transformers by Temperature", Montsinger, V. M., AIEE Trans., Vol. 49, p. 776, 1930; 14. "Effect of Color of Tank on the Temperature of SelfCooled Transformers Under Service Conditions", Montsinger, V. M.; Wetherill, L., AIEE Trans., Vol. 49, p. 41, 1930. 15. "New Insuldur: A Proved Insulation System", Westinghouse Electric Corp., Publication #SA-9025, 1964. 16. "Measuring Transformer Winding Temperatures Continuously with the Transformer Carrying Load", Lockie, A. M.; Stein, G. M., AIEE Trans., Vol. 67, p. 1608, 1948. 143 17. "The Effect of Gassing During Overloads on the Impulse Strength of Transformer Insulation", Heinrichs, F. W.; Truax, D. E.; Phillips, J. D., Paper A-79-434-2 presented at IEEE PES Summer Meeting, Vancouver, British Columbia, Canada, July 15-20, 1979. 18. "Transformer Invented 75 Years Ago", Halacsy, A. A.; Von Fuchs, G. H., AIEE Trans., Vol. 80, p. 121, 1961. 19. "The Combined Effects of Thermal Aging and Short-Circuit Stresses on Transformer Life", McNutt, W. J.; Patel, M. R., IEEE Transactions on Power Apparatus and Systems, Vol. 95, No. 4, p. 1275, July/August 1976. 20. "Measurement of Ambient-Air Temperature During Temperature Test on Transformers", Beavers, M. F., AIEE Trans., Vol. 79, p. 1021, 1959. 21. "Relative Transformer Aging Obtained Under ASA Load-Time Recommendations When Using Different Estimating Procedures and Aging Curves", Beavers, M. F., IEEE Trans. Power Apparatus and Systems, Vol. 83, p. 909, 1964. 144 22. "Short Time Failure Mode Considerations Associated With Power Transformer Overloading", McNutt, W. J.; Kaufmann, G. H.; Vitols, A. P.; MacDonald, J. D., Paper F-79-695-8 presented at the IEEE PES Summer Meeting, Vancouver, British Columbia, Canada, July 15-20, 1979. 23. "Does Color Really Affect Transformer Temperature?", Power, McGraw Hill Co., New York, December, 1957. 24. "Loading of Power Transformers", Norris, E. T., IEEE Proceedings, Vol. 114, p. 228, 1967. 25. "Thermal Ratings of Transformers", Norris, E. T., IEEE Proceedings, Vol. 118, p. 1625, 1971. 26. "Growth of the Electric Power Industry", Webler, R. M., The Line, Power System Division, McGraw Edison Co., 3rd Quarter, 1979. 27. "Transformer Gas Analysis Helps Trace Degradation", Oomen, T. V., Electric Light and Power, Vol. 57, No. 11, Nov. 1979. 145 28. "Hot Spot Detector Helps Predict Transformer Life", Yannucci, D. A.; Thompson, J. H., Electric Light and Power, Vol. 57, No. 9, September, 1979. 29. "Transformer Hot Spot Detector", Bell, R.; Bertin, M., Nucleohic Data Systems Report on EPRI Research Project 752, EL-573, October, 1977. 30. "Monthly Normals of Temperature, Precipitation, and Heating and Cooling Degree Days 1941-70", U.S. Dept. of Commerce, National Oceanic and Atmospheric Adm., Environmental Data Service, August, 1973. 31. "Determination of Ratings for Tubular Bus", Bethke, T.; Davidson, G.; Dyraek, V.; Geiger, R.; McElhaney, R.; Nabet, G., Pennsylvania-New Jersey-Maryland Interconnection, August, 1979. 32. "Thermal Consideration for Outdoor Bus Design", Prager, M.; Pemberton, D. L.; Craig, Jr., A. G.; Bleshman, N. A.; IEEE Transactions on Power Apparatus and Systems, Vol. PAS-95, No. 4, p. 1361, July/August 1976. 146 33. "Gas Detection - A Key to Transformer Health", Transmission and Distribution, Vol. 32, No. 1, January, 1980. s- 147 APPENDIX A Listing of Transformers Used in this Study A. RTE-ASEA Corporation, 10 MVA, 67-13.2 kV, OA, 55/65°C, 3 phase, Serial Number A4140. B. Westinghouse Electric Corporation, 15/20/25 MVA, 67-13.2 kV, OA/FA/FA, 55/65°C, 3 phase, Serial Number RCP1653. C. McGraw Edison Electric Corporation, 150 MVA, 230-69 kV, FOA, 55/65°C, 3 phase, Serial Number C-05043-5. D. General Electric Company, 300 MVA, 230-138 kV, FOA, 55/65°C, 3 phase, Serial Number M101504. 148 CU C/3 0J -- p CO « CJ CD U O 00 CU m fM a& 0) CO n vo vo o rH i-H m O o CM CM CM CM rH CM in vo o m CM CU P CU 03 o aa o m o o o o CO H PL, -* m VO VO CD • CO CO o r~- o o CO CO r~ r^ o o r»- o o o vo m CO in »* *tf o CO vr xi P d o a d 01 > •H 00 a w •H u co OJ • aa O CM H < r-~ o m vo rH o co »tf VO m •<f m m o CO m •■* i—i m vf o o m <f «* ** m m vo vo m o o r- )-i O S f3 P-I 23 •rl 00 o l-l o ■P cO a d u TJ O W X 00 d o3 n) o CO •H n) +J iH >? d 01 W <U U PM d P CU co 00 H o eO 0) o U P< CU a > cu <6 H cu a o r^ 00 >* o rH 00 rH ■H rH rH m rH CM r- i—i CM CO -* CM CO CO r- vo CM vo CO CO en CO CM 00 rH CM rH CM •H P d CO •H T3 CU a •H rH CJ Cfl CO P d o d OJ u o XI OJ ooo d co o o n •H OJ 1 P > ^> (0 •rl < + CO CO rH rH CO rH v* CM rH «* CO rH r-t rH rH CO rH rH •<!■ rH rH CJ vo m CO i—i rH rH rH rH rH m CO CM rH CO rH OJ > CU « u CU s w OJ a •H p p cu cu CO u 0) d 00 P co (0 1-1 l-l CJ OJ 0 a) 5 1* CO Vf •<f O VO 00 «■* 1^. CM CO r» r- CM rH CO O rH m o CO CN rH CM 00 rH CM rH in o rH CM 303 T3 d CO OJ CU H & E? CO a u fl XI Xl (U J-I CO >-j PH xl CO d o P •a l-l CU d CJ CO a rH •H l-l Ck < 149 P >> CO a cu 3 »-) >> rH d >-> C/l d 00 d < X> a OJ p Q* CU CO u U CU X> O P CJ O OJ xi a cu > o £5 w •H rl I-I cu X< a cu o cu n 03 X) P o « APPENDIX C Oil Temperature Change Equation To calculate the increase in the transformer's top oil temperature caused by solar radiation, the following equation T, used. A 14 can be Sr _ __a A T ° + L - 6.50xl(f4 4 C + 3.88xl0~ F1,19 (T . - Tn y)1,19 1-19 0 19 v(T ol. - T,la') ' 01 F la Where: 2 C = Thermal capacity - (joules/cm'/°C)/3600 F = L = Ratio of the average tank surface temperature rise over ambient to top oil rise over ambient. 2 Transformer loss - watts/cm r = Average value of the effective intensity of solar radiation over an hour - watts/cm S = Coefficient of solar absorption T . = Value of top oil temperature at the beginning of an hour - °K AT = Change in top oil temperature in one hour - °K T, = Average temperature of the surface absorbing heat radiation over a one hour period - °K 01 It is assumed that the transformer is painted gray or green giving an emissivity value of 0.95. 150 VITA The author was born in Allentown, Pennsylvania, on July 27, 1950. He is the son of Mr. and Mrs. Wasyl Chaply. He is a graduate of the Class of 1968 from Northampton Area High School. In 1972, he received a Bachelor of Science Degree in Electrical Engineering from Lehigh University. He joined Pennsylvania Power and Light Company during the summer of 1971 as a Cadet Engineer: Upon graduation, he accepted a position with Pennsylvania Power and Light Company in the Substation Engineering Section. His present position is as Project Engineer in that section. ' The author currently resides in Northampton, Pennsylvania. 151