October 1983

Engineering and Research Center

'Department of the Interior u of Reclamation

Synopsis of Insulation Problems at Yellowtail Powerplant

October 1983 6. P E R F O R M I N G O R G A N I Z A T I O N C O D E

7. AUTHOR(S) 8. P E R F O R M I N G O R G A N I Z A T I O N R E P O R T NO.

Richard C. Arbour REC-ERC-84-4 I

9. P E R F O R M I N G O R G A N I Z A T I O N N A M E A N D ADDRESS 10. WORK U N I T NO.

Bureau of Reclamation Engineering and Research Center 11. C O N T R A C T OR G R A N T NO.

Denver, colorado 80225 13. T Y P E O F R E P O R T A N D P E R I O D

C O V E R E D 12. SPONSORING A G E N C Y N A M E A N D ADDRESS

Same I 14. SPONSORING A G E N C Y C O D E

DlBR IS. S U P P L E M E N T A R Y N O T E S

Microfiche and/or hard copy available at the Engineering and Research Center, Denver, Colo.

Ed: RDM

6 . A B S T R A C T

A synopsis of insulation problems at Yellowtail Powerplant is presented along with the dissection results of a faulted portion of a coil from the powerplant.

7. K E Y WORDS A N D D O C U M E N T A N A L Y S I S

I. DESCRIPTORS--/ rotating machines/ stators/ insulation/ hydroelectric generators/ groundwall insulation

,. IDENTIFIERS--/ Yellowtail Powerplant, Montana/ Pick-Sloan, Missouri River Basin Program

:. COSA T I F ie ld /Group 0% COWRR: 0903 SRIM:

8. D I S T R I B U T I O N S T A T E M E N T 19. S E C U R I T Y C L A S S 21. NO. O F P A G E . ITHIS REPORT)

A v a i l a b l e from the N o t i o n a l T e c h n i c a l In format ion Serv ice, Opera t ions D iv rs lon . 5285 Por t R o y a l Rood. Spr ingf ie ld. V ~ r g i n i a 2 2 / 6 1 . UNCLASSIFIED 13

20. S E C U R I T Y C L A S S 2 2 . P R I C E

(Microfiche and/or hard copy available from NTIS) ( T H I S PAGE)

U N C L A S S I F I E D

SYNOPSIS OF INSULATION PROBLEMS AT YELLOWTAIL POWERPLANT

by Richard C. Arbour

October 1983

Power and Instrumentation Branch Division of Research and Laboratory Services Engineering and Research Center Denver, Colorado

UNITED STATES DEPARTMENT OF THE INTERIOR * BUREAU OF RECLAMATION

As the Nation's principal conservation agency, the Department of theInterior has responsibility for most of our nationally owned publiclands and natural resources. This includes fostering the wisest use ofour land and water resources, protecting our fish and wildlife, preserv-iny. ti!i::.environmental and cultural values of our national parks andhistorical places, and providing for the enjoyment of life through out-door recreation. The Department assesses our energy and mineralresources and works to assure that their development is in the bestinterests of all our people. The Department also has a major respon-sibility for American Indian reservation communities and for peoplewho live in Island Territories under U.S. Administration.

The information contained in th is report regarding commercial prod-ucts or firms may not be used for advertising or promotional purposesand is not to be construed as an endorsement of any product or firmby the Bureau of Reclamation.

ii

CONTENTS

Introduction and conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Original failure and dissection...........................................

Theory of failure """"""""""""""""""""

Previous testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Subsequent testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Bibliography. . . . . .. . . . . . . . . . . .. . ... .. . . .. . . .. . ... .. . . . . .. . . .. . .. ."

. . . ..

TABLES

12

Results of corona probe tests on unit 3. . . . . . . . . . . . . . . . . . . . . . . . . . .Comparison of readings from two corona probe

tests on unit 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Results of corona probe tests on unit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . .Results of corona probe tests on unit 2. . . . . . . . . . . . . . . . . . . . . . . . . . .

34

FIGURES

123

Ramped d-c absorption test V-I curve, B phase. . . . . . . . . . . . . . . . . . . .Ramped d-c absorption test V-I curve, C phase. . . . . . . . . . . . . . . . . . . .Ramped d-c absorption test V-I curve, A phase

(flashover). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ramped d-c absorption test V-I curve, A phase

retest (flashover). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ramped d-c absorporation test V-I curve, A phase

(retest 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ramped d-c absorption test V-I curve, A phase

(retest 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ramped d-c absorption test V-I curve, A phase

after repairs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dissected coil prior to removal of all groundwall

and turn insulation.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Closeup view of dissected coil prior to removal

of all groundwall and turn insulation. . . . . . . . . . . . . . . . . . . . . . . . . . .View showing discoloration of binder material. . . . . . . . . . . . . . . . . . . .Groundwall and turn insulation removed from coil. .. . . .. .. . . .. . . .Portion of unused spare coil showing normal

color of bi nder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4

5

6

7

8

9

101112

iii

Page

2

4

2

334

56

7

8

9

10

11

12

121213

13

INTRODUCTION AND CONCLUSIONS

On November 19,1980, unit 3 at Yellowtail Pow-erplant "relayed off" (shut down) with a line-to-ground fault on phase A. Through subsequenttesting, the fault was located on the line end coilof one of the parallel groups. The faulted portionof the coil was removed and a half-coil splice wasmade by Westinghouse personnel.

Dissection of the faulted portion of the coil andsubsequent testing revealed deterioration ofstrand binder, a condition that was probably pre-valent throughout all of the machines in the pow-erplant due to the manufacturing process usedwhen these coils were made.

ORIGINAL FAILURE ANDDISSECTION

The faulted portion of the coil was sent to thePower and Instrumentation Branch, Division ofResearch and Laboratory Services, E&R Center,Denver, Colo. for dissection.

This was the first Bureau line-to-ground insula-tion failure of a hard insulation system where theunit had provided a relatively long and successfulperiod of service. The coils in this machine hadnot become loose and had not been subjected toexternal mechanical abrasion. This failure wasunusual because it did not result from an internalturn-to-turn fault, or an external mechanicalwedging problem typical of many "hard" insula-tion systems.

Examination of the coil prior to dissectionrevealed the following:

1. When the coil was pulled from the slot, apronounced bend occurred at the point offault.

2. The fault was located about 300 mm (12 in)down from the top of the steel on the left sideof the boreside turn.

3. The coil sounded hollow when the outside ofthe insulation was tapped with a piece ofmetal.

The coil was dissected by using a sharp knife topeel off the layers of mica one at a time.

The following conditions were found during re-moval of the groundwall and turn insulation (figs.1 and 2):

1. The binder used to hold the mica togetherwas dry and dark yellow to brown in color (fig.3). It had apparently deteriorated, because thelayers of mica were peeled apart very easily, ascompared to the force required to peel the

comparative sections of a portion of the sparecoil (fig. 4).

2. The final two strands of the boreside turnwere distorted and misplaced about 3 mm(1/8 in). The distorted strands had formed anangle at their interface which had created asharp knife-like edge. The outer edge hadworn through the turn insulation and four tofive layers of the groundwall insulation. In thearea of the fault, there was evidence of exten-sive corona activity under the groundwall andthroughout the badly deteriorated turn insula-tion. None of the binder was intact, and onlybrown dust was found. Through the first twoturns, the strand insulation had completelydecomposed into dust (fig. 5). In the third turn,the strand insulation was intact but no bindercould be found.

THEORY OF FAILURE

The failure mode was primarily mechanical. Theouter strands of the bores ide turn are subject totangential mechanical forces created by the fluxentering the front of the slot. If these strands arenot well bonded to the remainder of the strands inthe first turn, they can vibrate and move withrespect to the rest of the strands in the turn. Thisvibration could cause strand-to-strand faultsresulting in higher temperatures and accelerateddeterioration of the binder that holds the micainsulation in place. The dissection of the unit 3fault revealed conditions which coincide closelywith this explanation of the cause of failure. Thevibrating strands wore through the turn andgroundwall insulation, and eventually resulted inelectrical failure. The misplaced strands wereexamined and found to have lost 20 percent oftheir original weight due to vibration-inducedabrasion.

The dry and discolored binder found throughoutthe coil sample was probably due to overheatingcaused by the strand-to-strand fault anddegraded thermal conduction properties. Thiscondition probably exists throughout the coil,including the back half of the coil which is still inthe machine. The bend which occurred when thecoil was removed was due to lack of rigidity of thegroundwall insulation in the vicinity of the fault.

PREVIOUIS TESTING

Routine "ramped" (gradually increasing) d-cabsorption tests [1]' were run on unit 3 ofthe fourgenerators at Yellowtail Powerplant on Sep-tember 17, 1980. The easiest way to disconnect

'Numbers in brackets refer to items in the Bibliography.

these units from the isolated phase bus and gen-erator switchgear for testing was to unbolt theflexible braid connectors on the isolated phasebus between the generator and the generatorswitchgear. This left about 3 m (10ft) of isolatedphase bus which had to remain connected to thegenerator during the tests. The connector thatwas originally bolted to the flexible braids waswrapped with plastic for corona suppression. Thegenerator neutral leads were also electrically iso-lated and corona shielded in a similar manner.

The results of these tests showed that, on phasesBand C, the V-I curve did not have any disconti-nuities or sudden current increases (figs. 6 and7). This indicated good quality insulation. Onphase A the V-I curve (fig. 8) was linear up to 23kV, at which point the current suddenly increasedand went off scale. The quality of the coronasuppression was questioned, more plastic tapeadded, and the phase retested. This time the cur-rent increased and went off scale at 23.7 kV (fig.9). We now suspect that arc-overs occurred in theisolated phase bus during these tests. Theremaining portion of the isolated phase bus wasdisconnected from the generator leads, the gen-erator leads insulated from the isolated phasebus by use of micarta blocks, and the unitretested on September 22, 1980. The V-I curvefor phase A showed a moderate tip-up2 starting at21.5 kV (fig. 10). Another test was made with thesame results, except the tip-up started at 22 kV(fig. 11). For this test. it appeared that the insula-tion was serviceable and that the moderate kneeobserved on phase A was apparently due toinsufficient corona shielding on the winding ter-minals. Therefore, it was assumed that the V-Icurves indicated acceptable quality insulationand the unit was returned to service.

We now conclude that the arc-overs thatoccurred during testing on September 17, 1980,may have overstressed the already weakenedgroundwall insulation on the coil that eventuallyfailed. This is further reinforced by the fact thatthe ramp test run after the faulted half-coil wasreplaced (fig. 12) did not have the tip-up that waspresent on the last ramp test run on September22, 1980 (fig. 11).

SUBSEQUENT TESTING

After the dissection revealed a problem thatcould have been caused by the coil manufactur-ing process and could affect the coils in the rest ofthe machine, it was decided that corona probe

'Tip-up is a gradually increasing rate of rise in the measuredcurrent.

tests be conducted on unit 3. At this time, the unitwas disassembled and being rewedged. Thispresented an opportune time to perform the cor-ona probe tests.

Corona probe tests were conducted on unit 3 onDecember 14, 1981 . Table 1 is a Iist of the higherreadings recorded and the specific locations:

Table. 1-Results of corona probe tests on unit 3

Slot No.

101102103104108109146149155

Coil(In top of slot)

C3-3C3-2C3-1

B2-17C2-17C2-18A2-3C2-4C2-3

Readi ng

80150280

60705575

27095

Wedges were removed from slots 101, 102, 103,146, 149, and 155, and a close physical inspec-tion was conducted. Coils were checked forsoundness by tapping with a coin. Hollow-sounding areas were found on the coils in slots101, 102, 103, 146, and 149. The high readingsin slots 108, 109, and 155 were probably due tothe discharges being generated in slots 102, 103,and 149, respectively.

Arrangements were made to have the front half-coils in slots 102, 103, 146, and 149 removedand dissected. A detailed description of the dis-section of these coils is given in a report byRichard Dye of the Regional Office in Billings,Montana. The dissection of the coils removedfrom slots 103 and 149 revealed extremely dete-riorated binder, voids, evidence of overheating,lose strands, and internal insulation damage dueto vibrating strands. The coils removed from slots102 and 146 had voids, deteriorated binder, andshowed signs of overheating. Spare coils wereused to replace the coils removed. The contractorencountered great difficulty making the half-coilsplices because of the deteriorated condition ofthe insulation on the back-half of the coils thathad remained in the machine.

Since the dissection confirmed our diagnosis thatthere would be more coils in the machine withinternal insulation deterioration and that wecould detect these deteriorated coils with the cor-ona probe, we suggested that corona probe testsbe conducted on the other three units at thepowerplant. We also decided to corona probe testunit 3 again to determine if the repairs had

2

reduced the corona level in the slots that wesuspected were giving high readings due to crosscoupling from the bad coils.

Corona probe tests were conducted on unit 4with the rotor in the bore on April21, 1982. Testresults indicated only two slots with moderatelyhigh readings, slot 194 with a reading of 64 andslot 200 with a reading of 58. Since these wereonly moderate readings, it was concluded that nomajor deterioration had occurred in unit 4.

Corona probe tests were conducted on unit 3 onApril 26, 1982. The areas where the half-coilsplices were made showed a marked decrease incorona probe readings; however, several areasthat were not involved in the repairs hadincreases, although none were as high as inDecember 1981. Table 2 shows a comparison ofthe readi ngs of both tests.

Table 2.-Comparison of readings from twocorona probe tests on unit 3

Slot No.

43477887889395

119133140141146147148159160167173

Coil(In top of slot)

84-2C4-383-9A3-5C3-7A3-2C3-4A2-12A2-7A2-4A2-5A2-3A2-2A2-181-18A1-16A1-13A1-12

Readings

12/81 4/82

8 5010 4610 5250 13012 8814 8221 50

5 486 50

40 9412 11050 9020 4816 6410 64

8 608 708 62

As with the first test. the second test indicatedthat some of the high readings are probably dueto cross coupling, as in the case of slots 140, 141,146, 147, and 148; and slots 167 and 173. Thesecoils are all physically connected in series. Thehigh frequency discharges detected by the coro-

na probe are coupled into adjacent coils throughthe copper coil jumpers. Although there are nocoils with high readings above 150, there are 15coils with moderate readings between 50 and150 that should be monitored closely to forewarnof an increased rate of deterioration.

Corona probe tests were conducted on unit 1 onApril 27, 1982; results are given in table 3.

Table 3.-Results of corona probe tests on unit 1

Slot No.Coil

(In top of slot) Readi ng

626991959798

101183193194195197198199200203204205206210211214216

C3-1 5C3-1483-4C3-483-283-1C3-3C1-1181-5A1-4A1-5Cl-681-381-4A1-3C1-4C1-581-281-1C1-2C1-1A4-1 6C4-1 7

5074767854

140505060

1005263

730240110130120270100

56806082

High readings due to coupling appeared in thefollowing slot sequences: 62-69, 91-97-98, 95-101, 194-195-200, 193-198-199-205-206,197-204-203, and 210-211. This unit appearedto have several coils with badly deteriorated insu-lation, and minor deterioration in the remainderof the coils with moderate readings between 50and 150.

3

Corona probe tests were conducted on unit 2 onMay 4, 1982; results are given in table 4.

Table 4.-Results of corona probe tests on unit 2

Coil(In top of slot)

C4-1 6C4-1 5B4-1 2B4-1 3A4-1 2A4-11C4-1 3C4-1 2C4-11A2-3A2-2C2-5B2-1A1-17A1-11C1-12C1-11B1-9

Slot No.

789

101112142021

146147150152153174182183186

Reading

625662665461907260856255608078585053

There were no coils with high readings in unit 2.

After collecting all of the corona probe data, thefollowing conclusions were reached on the con-dition of the insulation systems:

1. Unit 1 had the highest corona probe readingof 730 and may be in danger of failure. Coils

removed from unit 3 had readings of 280 andwere in a very deteriorated and dangerouscondition.

2. The insulation in unit 3 was generally in amore deteriorated condition than the insula-tion in units 2 or 4, but was not in danger ofimminent failure.

All four units had a problem with the stationaryturbine wear rings. The rings on units 3 and 4were replaced when the units were rewedged in1979.

Because of the condition of the insulation sys-tems, the reliability of the units was questionableand a decision was made to rewind the genera-tors. It was decided to rewind unit 1 first becauseit had severely deteriorated coils and also neededa new stationary turbine wear ring. Unit 2 wasselected as the second one to be rewoundbecause it also needed a new stationary turbinewear ring. This is predicated on the condition ofthe unit 3 insulation. If, for some reason the unit3 insulation starts deteriorating at a faster rate,unit 3 may have to be rewound before unit 2. Ifnot unit 3 would be rewound third and unit 4 last.

BIBLIOGRAPHY[1] Milano, B. and R.C. Arbour, A PortableRamped D-C Test Set for Electrical InsulationTesting, Bureau of Reclamation Report No. GR-80-2, E & R Center, Denver, Colo. June 1980.

4

cD'" Ct!

~.r:C

LIIIai>~()

:>

>en

.x2c0Q

.0'" .QCt!

Q()-0"0O

JC

LEC

t!a:I~:JQ

1

~...

'""In5

""~i:=II......

EE

uu

.0I

4u

+-<7'>

.,.

...x

:1,

0>o.--(\J

-0

-;;-"""

>-rocr><>-~

'"

'" N0N00

15

10

<:t.

(J')

5

Yellowtail Unit 3

C0 9-17-80R Dyex = I kV/cmy,. Ij.LA/cmT ,. 25.C

00 5 10 15 20 25

kV

Figure 2.-Ramped d-c absorption test V-I curve, C phase.

15

10

J <:l

5

Yellowtoil Unit '3

A' 9-17-80R Dyex ~ I kV/cm

Y'"' IIA-A/cmT

'"'

25.C

00 5 10 15 20 25

kV

Figure 3.-Ramped d-c absorption test V-I curve, A phase, (flashover).

15

10

(X)

ca:

:l

5

Yellowtail Unit 3A0 9-17-80 Retest IR Dyex - I kV/cmY - I JLA/cmT - 25.C

00 :I 10 15 20 25

kV

Figure 4.-Ramped d-c absorption testV-I curve, A phase, retest (flashover).

I~

10

CDoC(

::l

~

00

Adjusted ramprate

~ 10 15

kV

Figure 5.-Ramped d-c absorption test V-I curve, A phase (retest 2).

Yellowtail Unit 3AI 9-22-80 Retest 2R Dyex - I kV/cm

Y - 'J-LA/cmT - 25'C

20 25

0

15

10

«::!...

5

00

I5

I15

I10

kV

Figure 5.-Ramped d-c absorption test V-I curve, A phase (retest 3).

Yellowtail Unit 3A0 9-22-80 Retest 3R Dyex = I kV/emY = '/-LA/emT = 25.C

I20

I2<;

15

10

-'-'C(

:t.

5

12-10- 80

Yellowtail Unit 3AIR Dyex - I kV/cmy,. IJ.LA/CmT - 25.C

00 5 10 15 20 25

kV

Figure 7.-Ramped d-c absorption test V-I curve, A phase after repairs.

Figure B.-Dissected coil prior to removal of all groundwall and turn insulation.PBO1-D-79795.

.

Figure 9.-Closeup view of dissected coil prior to removal of all groundwall andturn insulation PBO1-D-79796.

Igure 10.-View showing discoloration of binder material. PBO1-D-79797.

12

Figure 11.-Groundwall and turn insulation removed from coil. PBO1-D-7979B.

Figure 12.-Portion of unused spare coil showing normal colorof binder. PBO1-D- 79799.

13

Mission of the Bureau of Reclamation

The Bureau of Reclamation of the US. Department of the Interior is responsible for the development and conservation of the Nation's water resources in the Western United States.

The Bureau's original purpose "to provide for the reclamation of arid and semiarid lands in the West" today covers a wide range of interre- lated functions. These include providing municipaland industrial water supplies; hydroelectric power generation; irrigation water for agricul- ture; water quality improvement; flood control; river navigation; river regulation and control; fish and wildlife enhancement; outdoor recrea- tion; and research on water-related design, construction, materials, atmospheric management, and wind and solar power.

Bureau programs most frequently are the result of close cooperation with the U.S. Congress, other Federal agencies, States, local govern- men ts, academic institutions, water-user organizations, and other concerned groups.

A free pamphlet is available from the Bureau entitled "Publications for Sale." It describes some of the technical publications currently available, their cost, and how to order them. The pamphlet can be obtained upon request from the Bureau of Reclamation, Attn D-922, P 0 Box 25007, Denver Federal Center, Denver CO 80225-0007.