AN INVESTIGATION OF THE TEMPERATURE DISTRIBUTION IN CAST-RESIN TRANSFORMER WINDINGS

Linden W. Pierce, Member

General Electric Company Rome, Georgia

Abstract: The cast-resin dry type transformer which had its development in Europe began to be widely used in the United States in the 1980 's . This type unit was incorporated into the first IEEE standard in 1989. One of the problems facing the IEEE working groups developing product standards and loading guides for the cast-resin transformer was the correct value of the hottest spot temperature allowance used to determine permissible average winding temperature rises so that insulation system temperatures would not be exceeded. Different values of the hottest spot temperature allowance are used in IEEE and IEC standards. Published literature on this subject is minimal,

A comprehensive thermal test program was conducted on a 2000 kVA cast-resin dry type transformer with 300 embedded thermocouples to determine the hottest spot temperature allowance and to obtain data to develop mathematical models. Nineteen thermal test runs were performed at different loads, tap positions, and core excitations under natural and forced air operating conditions. The test data indicated that the hottest spot temperature allowance used in the IEC standard was too low. In the IEEE standard the hottest spot temperature allowance appears conservative up to about 115 OC average winding temperature rise. The locations of the hottest spot for the primary and secondary winding were determined. Conclusions about thermal testing and loading are presented.

Kevwords; Cast-Resin, Dry Type Transformer, Hottest Spot Temperature Allowance, Industry Standards, Loading, Heat Transfer, Thermal Testing.

INTRODUCTION

After the demise of askarel insulated transformers containing polychlorinated biphenols (PCB'S) in the 1970 's , several alternative products appeared. One alternative was a dry type transformer with solid cast windings of epoxy resin. The development and commercial application of this unit began in Europe and this transformer design began to be widely accepted in the United States in the 1980 's . The solid cast and/or resin-encapsulated transformer was incorporated into the first IEEE standard [l] in 1989. Various terminology is used to describe this type unit and it will be referred to as the "cast-resin'' transformer in this paper.

One of the problems faced by the IEEE Working Group which developed the new standard was the correct value of the hottest spot temperature allowance to be used for the cast-resin transformer. The hottest spot temperature allowance is defined as the designated difference between the hottest spot temperature and the observable insulation temperature [2]. ANSI/IEEE Standard 1 states that the value is arbitrary, difficult to determine, and depends on many factors, such as size

and design of the equipment. The observable insulation temperature for dry type transformers is the average winding temperature by resistance. The hottest spot temperature allowance for dry type transformers then is defined as the incremental temperature to be added to the average winding temperature rise plus ambient temperature so that the hottest spot temperature in the transformer windings does not exceed the limiting insulation system temperature. In this paper the hottest spot temperature "increment" is defined as the difference between tested hottest spot temperature rise and tested average winding temperature rise. The hottest spot temperature "allowance" previously defined is the temperature difference used in standards documents. Also in this paper references to "temperature rise" mean "temperature rise over ambient air temperature".

The hottest spot temperature rise usually cannot be measured by tests on production units. The value of the hottest spot temperature allowance to be used for different insulation system temperatures is a major unknown facing the IEEE Working Groups developing standards and loading guides for ventilated dry type and cast-resin transformers. There are major differences between the hottest spot temperature allowance used by the IEEE Standard [l] and the IEC Standard [ 3 ] as shown in Table 1 below.

Table 1 HOTTEST SPOT TEMPERATURE ALLOWANCE USED IN STANDARDS

INSULATION SYSTEM

TEMPERATFE CLASS, c

185

AVE. WFG. HOTTEST SPOT TEMP!RATURE RISE, C ALLOWANCE, C

- - I I n _ - - - 150 30 30

Analytical and experimental investigations of the temperature distributions in ventilated dry type and cast-resin transformers are lacking in the technical literature, especially recent literature. Several papers deal with the mathematical solution of the conduction heat transfer in electrical coils of which those by Jakob [4] and Higgins [5] are typical. Several papers address the heat transfer in miniature transformers such as the paper by Mark [ 61 . The only analytical paper treating heat transfer in large ventilated dry type transformers is the paper by Halacsy [7]. The basic heat transfer mechanisms were considered in a solution of the average winding rise. Core heating effects were not considered and no calculation methods were developed for hottest spot temperature rise prediction.

An excellent experimental study was conducted by Stewart and Whitman [ 8 ] and reported in 1944. They reported laboratory test results of hottest spot and average winding temperature rises for large ventilated dry type coils of various lengths. At that ti!e standard average winding temperature rises were 80 C and the hottest sgot temperature allowance used in AIEE standards was 10 C. Based on the experimental work of Stewart and Whitman and Satterlee [9], the standards were revised so th$t the hotteTt spot temperature allowance was 30 C for 80 C average winding temperature rise. Stewart and Whitman's test results were reported again with additional data by Whitman [ 101 'in 1956. Whitman [ll] presented a paper on loading of ventilated dry type transformers and suggested that the hottest spot temperature rise should be approximately 1.375 times the average winding temperature rise. Even though Whitman suggested that the ratio of hottest spot temperature rise to average winding temperature ri5e was constant, the IEEE standards use a constant 30 C hottest spot temperature allowance for all insulation system temperature classes. IEC standards, although using a variakle hottest spot temperature allowance, still use 10 C at 80 'C average winding temperature rise even though the data of Whitman indicated it should be 30 OC and the AIEE standards were changed. For small transformers the hottest spot temperature allowance may be less as indicated by the data of Antalis and Duncan [ 121 who experimentally determiped a hottest spot temperature increment of 6 to 11 C at average winding temperature rises from 80 to 138 OC for single phase units from . 5 through 15 kVA. Dormer [ 1 3 ] reported test results on a 600 kVA dry type unit.o He reported a hottest spot temperature rise of 153 C at an average winding temperature rise of 118 'C for the high voltage winding and a hottest spot temperature rise of 162 OC at 113 'C average winding temperature rise for the low voltage winding. Thus the tested hottest spot temperature increments were 35 and 49 'C. The ratios of the hottest spot temperature rise to average winding temperature rise were similar to the earlier data of Stewart and Whitman.

Outer [ 1 4 ] stated that based on cast-in thermocouples in 400 and 600 kVA cast-resin units the ratio of hottest spot temperature rise to average winding temperature rise was 1.15. He also stated that the ratio for larger ratings was higher but gave no values. Featheringill [15] presented loading calculations for cast-resin transforme5s using conventional loading guide equations and a 30 C hottest spot temperature allowance. No experimental or analytical results were reported to support the calculations. Although the IEC Loading Guide for Dry Type Transformers [16] applies to cast-resin dry type transformers, this type unit was omitted from coverage in the 1989 edition of the IEEE Loading Guide [ 171 since the IEEE Working Group decided that adequate data was not available.

A review of the dry type transformer standards and published literature indicated that a comprehensive experimental and analytical study was needed to determine the correct value of the hottest spot temperature allowance for both ventilated dry type and cast-resin dry type transformers. This paper reports on the investigation conducted by the author on a 2000 kVA cast-resin dry type transformer with embedded thermocouples in the high voltage and low voltage windings. The test program was conducted to assist in the development of a mathematical model to calculate hottest spot temperature rises in cast-resin transformers and to support IEEE Working Group efforts developing product standards and loading guides.

593

TEST P R O G M

Test Transformer

A prototype cast-resin transformer rated 2000 kVA, three phase, 60 hertz, high voltage of 4160 volts delta connected, and a 480 volt Y connected Low voltage was constructed with embedded thermocouples to determine the hottest spot temperature rises of the high voltage and low voltage windings under various loading conditions. Fans were installed on the unit to evaluate the hottest spot temperature rises under forced cooled conditions.

HV LV

in

Figure 1 . High Voltage and Low Voltage Windings

The winding design used by the author's company and others is illustrated in Figure 1. The high voltage winding consisted of two castings per phase of four sections per casting. The two castings were in intimate contact and clamped with jack screws and rubber at the ends. Each section consisted of 21 to 22 turns of aluminum strip conductor separated by insulation between turns. The low voltage winding was a layer type with 11 layers of full length aluminum sheet conductor separated by layer insulation. Three hundred iron-constantan thermocouples of .005 inch diameter wire were inserted between turns of the high voltage winding and on each layer of the low voltage winding during the winding process. The large number of thermocouples was required to investigate the axial and radial temperature distributions of all the high voltage winding sections and low voltage winding layers at radial positions represented by "in window" and "out of window" locations. The center leg of the three phase assembly was expected to be the hottest and this was the leg that was instrumented with thermocouples. A plan view photograph with the top core yoke removed is shown in figure 2.

Test Procedure

Thermal testing was conducted in the production test facility of the author's company. Additional thermocouples were added to monitor air, casing, and core temperatures. Production thermal tests for dry type transformers are usually made by either the short circuit method or the loading back method in accordance with the ANSI/IEEE Test Code [18]. The short circuit method requires two tests to be performed. A test with rated current is performed with the low voltage windings shorted. During this test core loss is not present. The short is removed and an excitation thermal test is performed with rated voltage which gives core loss but no load on the unit. The average temperature rises are determined for the two tests and combined by means of an equation in the Test Code to determine the corrected average temperature rise with full load currents in the windings and normal excitation on the core. The loading back test method produces rated currents in the windings simultaneously with rated voltage to give core loss. The loading back test method is more representative of actual loading but it requires a greater amount of testing facilities and becomes increasingly difficult to perform as the size of the transformer increases.

Full voltage on the unit presented a possible hazard to test personnel and equipment with the high voltage winding thermocouples connected to the data acquisition equipment. Experience had shown that the core loss has only a small effect on the high voltage winding average temperature rise. Thermal tests were performed first with the low voltage windings shorted and current circulated in the windings. These tests represented the current-run portion of the short circuit test method although the runs were performed at many values of current. The primary objective of the current runs was to obtain temperature data for the high voltage winding. With the current-run thermal test only impedance voltage exists on the high voltage winding and core l o s s is not present.

In dry type transformers the core has a significant effect on the average temperature rise of the low voltage winding. After the current-run thermal tests were completed, the high voltage winding thermocouples were disconnected and each thermocouple lead taped to an insulating board and separated from other thermocouples inside the transformer casing. Test runs were then performed on the unit by the loading-back or "bucking" method in accordance with the ANSI/IEEE Test Code [le]. The primary objective of these tests was to obtain the temperature data for the low voltage windings with core loss present.

Test runs were made on low and high tap positions and different core excitations with self cooled (AA) and forced air operation (FA) to obtain data under different conditions to develop a mathematical model. During run 23 impulse testing in an adjacent test berth damaged the data acquisition equipment. For the last test run the hotter thermocouples were selected based on previous data and measurements taken using alternate equipment with the capability to measure fewer channels. During test run 24 an electrical failure occurred due to the thermocouples in the high voltage winding.

The data acquisition equipment permitted rapid scanning of all the thermocouples throughout the test runs. Transient data was recorded during start-up and cool-down. Comparison of the thermocouple readings taken shortly before and after switch-on and shortly before and after switch-off of the transformer indicated no changes due to electric or magnetic fields. The test runs for which steady state data is reported are summarized in Table 2 . The transient data may be reported in a subsequent paper.

Figure 2. High Voltage and Low Voltage Windings Plan View on Core Leg

The internal assembly of the unit is shown i figure 3 with the unit in the casing with several shee metal panels removed.

Figure 3 . Core and Coil Assembly in Casing

n t

594

Table 2 THERMAL TEST RUNS

CORE EXC. %

0 0 0 0 0 0 0 0 0 0 0 0 0 0

100 100 100 115 100

q-z 1.000

TAP POS.

LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW HIGH HIGH HIGH LOW LOW LOW LOW LOW

1.000

1.406 1.000

1.000

I LEG

6 1 . 6 42.9

7 8 . 3 70 .4 - 52 .0 34 .1 88.8 75 .6 71.8 89.0

8 6 . 6 74.6

64.2

CTR.

66 .2 47.2

83.5 7 7 . 1 66.7 57.3 3 7 . 4 96 .6 82.4 78.3 94.7

9 2 . 1 78 .9

71 .1

COOL.

RUN NO.

1 2 3 4 5 7 8 9 10 11 12 1 4 15 1 6

19 20 22 23

18

AA AA AA AA AA FA FA FA FA FA AA AA AA FA AA AA AA AA FA

HS. RISE

94.2 80.9 54 .8

126 .0 103.2 106.4

9 3 . 1 84.4 46.8

138 .9 95.3 84.5

104 .6 8 9 . 4

- - - - -

yMB. C

AVG. RISE

2 7 . 1

24.5 28.9

24 .8 28 .4 26 .9 25.4 25 .3 24 .4

25.9 25.2

27 .0 31.0 28 .3 30.9 33.3 34 .9

27.8

28.6

HS. INCREMENT RATIO

Hot and cold resistance measurements were taken between terminals of all three phases using equipment patented by Runaldue and Burnett [ 1 9 ] . This equipment was developed by General Electric to accurately measure low resistances. A cooling curve of the hot resistance was extrapolated to the instant of shutdown in accordance with the Test Code, For the delta connected high voltage winding the average temperature rise of the center leg was obtained by calculating the resistance of each coil from the terminal resistance measurements. A similar procedure was used for the Y connected low voltage winding for runs one through six. After run six the average temperature rise of the center leg low voltage winding was obtained by using line to neutral resistance measurements. These procedures .with applicable equations have been proposed by the IEEE Task Force revising the method of performing temperature rise tests for dry type transformers for the next edition of the ANSI/IEEE Test Code [la].

Summarv of Results

Table 3 THERMAL TEST RESULTS

FOR HIGH VOLTAGE WINDING'

21.0 30 .5 22.9 19 .2 1 0 . 5 3 5 . 4 1 7 . 1 1 4 . 0 1 7 . 1 19.7

1 . 2 6 1 . 4 0 1 . 3 3 1 .29 1 . 2 9 1 . 3 4 1.22 1 . 2 0 1 . 2 0 1 .28

' LEG

7 5 . 4 63.3 41.9 98 .7 79.9 74 .6 66 .8 63 .4 3 5 . 3 99.3 75 .8 66.2 83 .6 6 6 . 3 82.7

-12.2 83 .7 84.8

-

6 8 . 1

CTR. 3 LEG

6 6 . 0 1 7 . 6 43.5 1 2 . 9

102.7 i 27.3

77 .5 1 8 . 8

82 .2 23.3 75.9 I 3 1 . 8 70.2 26.3 65.2 21 .0 36 .3 1 1 . 5

103.5 39.6 78.2 1 9 . 5 7 0 . 5 1 8 . 3

69.7 23 .1 87.5 21.0

117 .2 85 .2 8 6 . 8

CTR.JHS'Av.

1 6 . 7 11.22 14 .9 1 . 2 3 1 1 . 3 1 .26 23 .3 I 1 . 2 3

*All temperature rises and differences in 'C.

11 1 2 14 1 5 1 6

20 22

HS . RISE

91 .7 77.4 5 4 . 5

117.7 99.7 95.8 8 4 . 5 7 8 . 1 44.4

122.9 91 .2 86 .6

105.6 90 .8

113 .1 97 .4

146 .6 126 .9 108.1

Table 4 THERMAL TEST RESULTS FOR LOW VOLTAGE WINDING'

110.3 116.2 91.2 I 97 .1 75 .3 I 81.3

IS. INCREMENT

1 1 . 6

1 7 . 8

UTI0 3S/AV.

1 . 1 7 1 . 1 6

1 .19 1.24 1 . 2 7 1 . 3 6 1 . 1 9 1 . 2 7 1.11 1.11 1 .12

1 . 2 3 1 . 2 3 1 . 2 6 1 . 3 1 1 . 3 3

1 . 2 8

All temperature rises and differences in 'C.

A summary of the test results are shown in Tables 3 and 4. The average temperature rises shown were not corrected for ambient temperature. The current run tests were not corrected for core excitation effects. This permitted direct comparison of the tested average temperature rises with the tested hottest spot temperature rises. The hottest spot temperature increment is shown based on an average temperature rise of the three coils (legs) and of the hottest center coil with the thermocouples. The ratio of the hottest spot temperature rise to average temperature rise is based on the center coil.

ANALYSIS OF TEST DATA

TemDerature Distribution

The axial temperature distribution of the high voltage winding is shown in Figure 4 . The plot is for the hottest spot temperature rise of each section between turns 7 and 8 . The hottest spot for the high voltage winding was located in the second section from the top of the winding. The dip in the temperature distribution for run 12 at section 4 is due to turns tapped out and thus not loaded. The temperature of the top section is less than the second section from the top due to additional heat loss from the significant area at the top of the high voltage coil

TEST 1 1 HIGH TAP

TEST 12 LOW TA? + -e-

110,

SECTION FROM BOTTOM

FIGURE 4. HV AXIAL TEMPERATURE DISTRIBUTION

595

The temperature distribution in a radial direction around the high voltage winding showedlittle difference between locations under the core compared with outside the core. Temperature distribution patterns for the high voltage winding were similar for all the test runs differing only in magnitude. The radial temperature distribution of the high voltage winding is shown in Figure 5 for self cooled and forced air operation. As expected the hottest spot was located between turns 7 and 8 for self cooled operation. For fan cooled operation the hottest spot was located between turns 10 and 11. The temperature rises at locations A, D, and E shown in Figure 5 were within insulation under and over the winding. Due to flatness of the curve in the hottest spot region only small measurement errors should result due to thermocouple location.

LOCATION

TEST 12 TEST 7 AA FA * *

TEMP. RISE OVER mBIENT,oCl

"-1 I UNDER INNER LEAD OVER INNER LEAD

LAYER 2 LAYER 4 LAYER 5 LAYER 7 LAYER 10

OVER OUTER LEAD

AVERAGE RISE CTR LEG

A 2 4 6 8 10 12 14 16 18 20 22DE TURN NO.

FIGURE 5. HV RADIAL TEMPERATURE DISTRIBUTION SECTION 7.

CURRENT BUCKING DIFFERENCE RUN RUN 12 18

91.2 1 1 3 . 1 21.9 89 .4 112.4 23 .0 8 4 . 1 105.5 21 .4 84 .2 104.2 20.0 8 4 . 3 103.2 1 8 . 9

_ _ 102 .9 - - 89 .0 103 .8 14 .8 90 .9 1 0 3 . 1 12 .2

92 .1 9 . 7 82 .4

For the low voltage sheet winding a smaller increment of temperature change from the top to bottom was expected. This was confirmed by test as shown in Figure 6 .

JIND

HV

LV

LV

LEAD OPPOSITE SIDE LEAD SIDE

-46- -8-

COOL

H V A A

FA

AA

FA

~

""I 1

IEEE

_ _ _ 30 30 30 30 30

_ _ 30 30 30 30

30 30 30

30

I 0 5 10 15 20 25 30 35 40

DISTANCE FROM BOTTOM, INCHES

FIGURE 6. TEST 18, LV WINDING, LAYER 2 AXiAL TEMPERATURE DISTRIBUTION

IEC

5 5

10 10 10 15

_ _ 5 5 5

15

10 10-15

15

10

The temperature distribution within the low voltage winding was much more complicated than illustrated by figure 6 . The temperature distribution was influenced by the inside and outside leads and core heating. The hottest spot was located near the top of the inside lead next to the core. This would be somewhat expected by observation of Figure 2 which shows the absence of one cooling duct in this region to allow space for the leads without increasing the build of the winding. Additional insulation around the leads inside the winding also contributes to the hottest spot temperature in this region. The temperatures in the low voltage winding varied around the winding in a helical heat flow pattern which is difficult to illustrate.

COMPARISON OF LV WINDING TEMPERATURE RISE TEST RESULTS AT 100 PER CENT LOAD

9 7 . 1 ~ ~

C[S

Table 7 COMPARISON OF TEST RESULTS WITH STANDARDS

RUNI HS IRATIO ITEST NO. RISE HS/AVE HSoINC.

l a c I I c

5 4 . 8 1 . 2 6 1 1 . 3 80.9 1 . 2 3 1 4 . 9 94.2 1 . 2 2 1 6 . 7 95.3 1 .22 1 7 . 1

103.2 1 .26 21.0 t 126 .0 1 . 2 3 23.3

1 0 . 5 1 9 . 2 22.9 30.5 35.4

1 8 . 5 21.0 30.4

596

INSULATION SYSTEM

l'EMPERA:URE CLASS, c

105 120 130 150 155 180 185

HOT. SPOT PERMISSIBLE PERMISSIBLE RISE AVERAGE AVERAGE IN TEMP. RISE TEMP. RISE

40 'c AMB PER TEST DATA* PER STMDARDS AA FA IEEE[l] IEC[3]

65 51.6 48.5 - - 60 80 63 .5 59.7 - - 75 90 71.4 67.2 60 110 87.3 8 2 . 1 80 115 91.3 85.8 - - 100 140 111.1 104.5 - - 1 2 5 145 1 1 5 . 1 108.2 1 1 5

_ - _ _

_ _

TEST

QISCUSSI~

The test program showed that the ratio of hottest spot temperature rise to average temperature rise at different loads only varied slightly. This agreed with the conclusions of Whitman [ll] for ventilated dry type transformers. The tested ratios for the 2000 kVA unit were higher than the ratios reported by Outer [ 1 4 ] for 400 and 600 kVA units indicating that the ratio of hottest spot temperature rise to average temperature rise probably varies with size of unit. The test data indicated that the hottest spot temperature allowance used in the IEC standard is too low resulting in permissible average temperature rises which are too high. This is also confirmed by the data of Stewart and Whitman [ 8 ] for ventilated dry type units. The hottest spot temperature increment was found to increase with increasing hottest spot temperature rise and not the constant 30 '2 allowance used in the IEEE standard. The constant 30 C allowance appears conservative up to about 115 'C average temperature rise for the cast-resin design tested. Hottest spot temperature allowances used in product standards should be conservative and probably based on the largest unit covered by the product standard. Future standards revisions should consider a constant ratio of hottest spot temperature rise to average temperature rise.

Another difference in IEEE and IEC standards involves the method for performing temperature rise tests. The IEEE test code 1181 requires that temperature rise tests be performed on the tap position which gives the maximum winding temperature rise which would usually be the low tap. IEC 726 [3] specifies that the temperature rise test be performed on the rated tap. On low tap the unit would run hotter than tested.

For three phase transformers there is a difference between average temperature rises of the three legs with the center leg usually but not always the hottest. The IEEE Task Force on Dry Type Thermal Tests has suggested that the average temperature rise be calculated for each leg individually and the highest value used to determine if guarantees are met. The test program confirmed that this was the correct approach.

Confirmation of hottest spot temperature rise i s one performance characteristic for which no standard test method exists. Consideration should be given to add this measurement to the IEEE Standards as a design test requirement using prototype transformers to qualify a design family. A standard recommended test procedure should also be developed. The test program indicated that the loading back test method gave higher IOW voltage winding hottest spot temperature rises than expected from the short circuit test method. Core loss raises the temperature of the low voltage winding non- uniformly. The top portion of the winding nearest the core increased the most contributing to the higher hottest spot temperature. A similar effect could occur in the high voltage winding although the effect is thought to be less. Additional investigation is needed to develop correction factors for hottest spot temperature rise for the short circuit test method.

Due to questions raised in the IEEE Working Groups of the Transformers Committee the author decided that this data and initial analysis should be made available to the industry at the present time. Additional mathematical analysis is needed to fully explain the effect of various physical parameters such as size upon the hottest spot temperature increment. Analysis of dry type transformer heat transfer appears to have been neglected as evidenced by the lack of published papers. The data obtained should assist in future product standards and loading guide revisions. The test data reported is for only one transformer design from one manufacturer. Data and analysis from other manufacturers is needed because there are several cast- resin transformer designs available. The test experience reported should assist other manufacturers in planning similar thermal test programs.

TIME CONSTANT, MINUTES

HV LV

LEAD WINDING

RUN 1 START UP INCREASE FROM 75%

TO 114% LOAD

212 110 1 3 3 194 89 129

CONCLUSIONS

1. The hottest spot temperature allowance used in IEC standards for dry type transformers is too low. 2. The hottest spot temperature allowance used in IEEE standards appears conservative up to about 115 C for the cast-resin dry type transformer design tested. 3. A constant ratio of hottest spot temperature rise to average temperature rise should be used in product standards and loading guides. The ratio is different for self cooled and forced air operation. 4. The average temperature rise of each leg of a three phase transformer should be determined for both the primary and secondary winding. The average temperature rise of the hottest winding should be used to determine conformance to standards. 5. Consideration should be given to requiring the measurement of hottest spot temperature rise on a prototype transformer to qualify a design family. Recommended test procedures should be developed. 6. The loading back test method may give a higher hottest spot temperature rise than expected from the short circuit test method. 7. Time constants are different for the high voltage winding and the low voltage winding. Different parts of a winding may have different time constants. The different time constants should be considered in loading equations. 8. Data and analysis from other transformer manufacturers is needed to obtain a consensus agreement on future product standards and loading guide revisions.

REFERENCES

IEEE Standard General Reauirements f or Dry TvDe Distribution and P ower Transf ormers Including n o s e with Sol id Cast and/o r Resin Encawdated m, IEEE C57.12.01-1989, Table 4A p. 16. U E E Standard General PrinciDles for TemDerature Limits in the Rat' of Electric w e n t and foy

Drv TvDe Power Trans formers, IEC .Standard Publication 726. First Edition, 1982, Table IV pp. 19,21,23 and Amendment 1, February 1986. M. Jakob, "Influence of Non Uniform Development of Heat Upon the Temperature Distribution in Electrical Coils and Similar Heat Sources of Simple Form, " Trans. AS ME, Vol. 66, pp. 593-&05, 1944. T. J . Higgins , "Formulas for Calculating the Temperature Distribution in Electrical Coils of General Rectangular Cross Section," Trans. ASME,

M. Mark, "Heat Transfer Techniques for Magnetic Core Components," ElectrotechnolQgy,August1962,

A. A. Halacsy, "Temperature Rise of Dry Type Transformers, " AIEE Tr-, Vol. 77, Part 111,

H. C. Stewart and L. C. hitman, "Hot Spot Temperatures in Dry-Type Transformer Windings," AIEE T r m , Vol. 63, pp. 763-768, 1445-1448, 1944. W. W. Satterlee, "Design and Operating Characteristics of Modern Dry-Type Air Cooled Transformers," AIEE T u , vol. 63, pp. 701- 704, 1445-1448, 1944. L. C. Whitman, "CO-Ordination of Dry Type Transformer Models with Transformer Geometry," AIEE Tr ans,, Vol. 75, Part 111. pp. 328-332, June 1956. L. C. Whitman, "Loading of Ventilated Dry Type Transformers, " AIEE Tran s,, Vol. 76, Part 111, pp, 1077-1084, Dec. 1957.

. .

the Evaluation of Elec trical Insulat i m , MI= STD. 1-1986, p. 8.

Vol. 66, pp. 665-670, 1944.

pp. 87-93.

pp. 456-462, August 1958.

S. J. Antalis and G. I. Duncan, "Temperature Distribution in Insulation Systems for Dry Type Transformers and Their Effect on Design," presented at IEEE Winter Power Meeting, New York, Jan. 30-Feb. 4, 1966, IEEE paper 31CP66-163. R. L. Dormer, "Designing Class C Dry Type Transformers for Maximum Efficiency, AReviewwith Guidance on Predicting Behavior under Various Conditions," Electrical Review, 7 Sept. 1973, pp.

F. R. den Outer, "The Loading of Solid-Insulation Distribution Transformers with Special Reference to the Cast Resin Type," Proc. International Conference on Electricitv Distribution (CIRED 19771 , London, 23-27 Mai-1977. Part I, pp. 75- 79, Discussion Part I1 pp. 31-32,37,41-42, Paper 2.7. W. E. Featheringill, "Power Transformer Loading, " IEEE Trans. on Industrv Amlications, Vol. IA-19, No. 1, pp. 21-27, Jan./Feb. 1983. Loadine Guide for Drv-TvDe Power Transformers, IEC 905, 1987. JEEE Guide for Loadine Drv-TvDe Distribution and Power Trans formers, ANSI/IEEE C57.96-1989. IEEE Stan dard Test Code f or Dry TvDe Distribution and Power Transformers, ANSI/IEEE C57.12.91-1979, Section 11. L. R. Runaldue and J. W. Bumett, "Method and Apparatus for the Measurement of Low Resistance, " U. S, Patent 2.772.395, Nov. 27, 1956.

313-317.

ACKNOWLEDGEMENT

The work described in this paper was financed by the General Electric Company. Many General Electric personnel contributed to the manufacture and testing of the transformer described in the paper. R. E. Gearhart provided valuable technical honsultation to the author. The assistance of the General Electric Medium Transformer Operation Test Department at Rome, Georgia is gratefully acknowledged.

BIOGRAPHY

Linden W. Pierce (M'70) was born in AthenB, Texas on January 4, 1941. He received the B. S . degree in Mechanical Engineering fromThe University of Texas, Austin in 1963, completed the GE Advanced Course in Engineering in 1966, and received the M. S. Degree in Mechanical Engineering from the University of Tennessee, Knoxville in 1973.

In 1963 he joined the General Electric Company and

since 1965 has worked for the Transformer Department at Rome, Georgia with various positions in transformer design, development, and program management. He is currently Senior Engineer, Product Technology. He holds eight patents.

Mr. Pierce is a member of the IEEE Power Engineering, Magnetics, Dielectrics and Electrical Insulation, and Industry Applications Societies, and CIGRE. He is a member of the IEEE Transformers Committee, Chairman of the Working Group on Development of the Loading Guide for Cast-Resin Transformers, and Chairman of the Task Force on Revision of the Test Code for Performing Temperature Rise Tests on Dry Type Transformers. He is a Registered Professional Engineer in the State of Georgia.

598