ML12334A647

Official Exhibit - NYS000145-00-BD01 - W. Shu, S.A. Boggs, Effect of Cable Restoration Fluid on Inhibiting Water Tree Initiation, IEEE Transactions on Power Delivery, Vol. 26, No. 1 (January 2011) (Shu)

ML12334A647
Person / Time
Site:	Indian Point
Issue date:	01/01/2011
From:	Boggs S, Shu W Institute of Electrical & Electronics Engineers
To:	Atomic Safety and Licensing Board Panel
SECY RAS
References
RAS 21545, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML12334A647 (4)

Text

United States Nuclear Regulatory Commission Official Hearing Exhibit Entergy Nuclear Operations, Inc.

In the Matter of:

(Indian Point Nuclear Generating Units 2 and 3)

ASLBP #: 07-858-03-LR-BD01 NYS000145

~

.~\

I!!

..

<

o ~

.

0

~

0

~

r7; Docket #: 05000247 l 05000286 Exhibit #: NYS000145-00-BD01 Admitted: 10/15/2012 Identified: 10/15/2012 Withdrawn:

Submitted: December 15, 2011

~1-1) +o~ Rejected: Stricken:

--*** .. Other:

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. I, JANUARY 201 I 97 Effect of Cable Restoration Fluid on Inhibiting Water Tree Initiation Wen (Essay) Shu, Student Member, IEEE, and Steven A Boggs, Fellow, IEEE Abstract-Silicone fluid has been applied in cable rejuvenation 2550V

.=,,:: (5 kV/mm) for decades to "cure" existing water trees. However, whether it in-hibits future water tree initiation is not known. An electrical aging test has been carried out at 5 kV/mm (127 V/mil) based on a sample which includes semicon electrodes on both sides of the insulation.

Samples were tested untreated, treated prior to the test, and treated at the midpoint of the test (3500 h). The results suggest that treat-ment reduces water tree initiation by a factor of 25 for bow-tie and 100 for vented water trees for the samples treated prior to aging. In the case of samples treated at the mid-point of the test, both bow-tie and vented water tree density are reduced by an order of magni-tude relative to the density present at the time of treatment.

Fig. I. Experimental setup for electrical aging test. The sandwiched Index Terms-Cable rejuvenation, water tree. semicon-PE-semicon sample is employed for simulation of stress exposure in cable. Standard conductor semicon and 0.1 M NaCI are used in aging as sources of ions.

I. INTRODUCTION ATER TREEING or electrochemical degradation, first water in oxidized hydrophilic regions (water trees) to form a hy-W reported in 1969 [1]-[5], is an electro-chemo-mechan-ical phenomenon which has caused premature failure of XLPE drophobic oligomer (short polymer) which fills the water treed region [8]. While the effect of silicone rejuvenation in "curing" and HMWPE cable. Water treeing takes place in dielectrics existing water trees and extending the life of the cable is well es-which can be electro-oxidized from a highly hydrophobic state tablished, the effect of silicone fluid on initiation of new water to substantially more hydrophilic, which causes water to con- trees, especially for vented water trees, has not been reported.

dense into the hydrophilic region, resulting in a self-propagating A well controlled laboratory test was carried out to study the "water tree". Electrochemical degradation is accepted as the degree to which formation of new water trees, including both primary cause of unreliability of medium voltage XLPE cable bow-tie and vented trees, is inhibited by rejuvenation.

[3]. As a result of the high failure rate of HMWPE and XLPE The only related research of which we are aware was carried cable during the 1970s and 1980s, a great deal of effort was put out by Hydro Quebec [10], which removed aged cable from the into reducing the effects of water trees, including dry curing field and carried out further aging in the laboratory. Their work of cable, semicons with substantially reduced ionic content, showed that the rejuvenated cable had a much lower density of and eventually, introduction of tree retardant compounds which bow-tie water trees as well as a lower than usual water con-reduce the propensity toward water treeing, usually by making centration in the cable insulation. Very few vented trees were the dielectric more hydrophilic. Cable rejuvenation technology present in either rejuvenated or non-rejuvenated cable samples.

was introduced around 1987 to "cure" existing water trees in II. EXPERIMENTAL PROTOCOL HMWPE and XLPE cable, thereby extending substantially the life of the cable [6], [7]. The effect of silicon fluid on inhibiting water tree initiation Cable rejuvenation technology has been employed success- and growth was investigated using a sample configuration ( Fig.

fully in the field for over 20 years to cure existing water trees in 1) with semicons on both sides of the XLPE insulation which URD cable [8]. To date, over 25 million meters of URD cable was developed at the University of Connecticut in the 1980s have been treated and remain in service [9]. As the silicone re- [11], [12]. The test protocol, based on the samples of Fig. 1, juvenation fluid diffuses into the cable insulation, it reacts with was as follows (Table I).

Group 1 (15 samples) was aged in an untreated state to 3500 h, at which time 5 samples were removed for water tree counting, Manuscript received April 28, 201 0; revised June 21, 201 0; accepted July 16, 5 samples were aged to 7000 h in an untreated state after which 2010. Date of publication August 26,2010; date of current version December they were subjected to water tree analysis, and 5 samples were 27,2010. This work was supported by Utilx Corp. Paper no. TPWRD-00313-removed to form Group 2.

2010.

The authors are with the Electrical Insulation Research Center, Univer- Group 2 (5 samples) was taken from Group 1 after 3500 h of sity of Connecticut, Storrs, CT 06269 USA (e-mail: wen.shu@ieee.org; aging in an untreated state and subjected to silicone rejuvena-steven.boggs@ieee.org).

tion, after which they were aged for a further 3500 h.

Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Group 3 (10 samples) was subjected to silicone rejuvenation Digital Object Identifier 10.1109/TPWRD.2010.2061241 before aging, aged to 3500 h after which 5 samples were re-0885-8977/$26.00 © 2010 IEEE OAGI0001273_00001

98 IEEE TRANSACTIONS ON POWER DELIVERY. VOL. 26. NO.1. JANUARY 2011 TABLE I 1.0,--------------------.,

LONG TERM AGING TEST PLAN Aging Time Material 3500 hrs 7000 hI's Untreated Treated moved for water tree analysis and the remaining 5 samples were 2 3 4 5 6 7 8 aged to 7000 h after which they were subjected to water tree Exposure Time {Days}

analysis. Fig. 2. Phenyl Methyl Dimethoxy Silane (PMDMS) content in the The plaque samples of Fig. 1 were aged at room temperatnre semicon-PE-semicon plaque as a function of exposure time in 100% rel-and 5 kV/mm across the XLPE insulation. Silicone rejuvena- ative humidity at 50 0 C. Four days exposure of the fluid-saturated samples is sufficient for nearly complete curing.

tion was applied by soaking the samples in rejuvenation fluid to satnration after which the samples were subjected to moist air until the fluid in them had oligomerized, as determined using gas chromatography/mass spectroscopy (GC/MS). The curing time in moist air was 4 days.

A. Sample Preparation Cable grade polyethylene (Dow 4201) and conductor semicon are used in the test. The XLPE dielectric is moulded with two depressions for the semicon "buttons", which are pre- Fig. 3. Sample plaque is cut in half through the center, and 10 slices of 250 pared separately. The two semicon buttons are then cross linked JlIll (10 mil) thickness are prepared from each plaque, stained, and mounted on to the XLPE in the final step. All components are inspected at glass slides using Canada balsam for water tree count and size measurement.

every stage of manufacture to assure freedom from cavities, etc. The insulation thickness between the center of semicons is 0.51 mm. The completed test sample is glued to a polyethylene insulation for about 33 years [8]. In the present case, the rejuve-tube using a silicone RTV which does not evolve acetic acid nation fluid within the sample was oligomerized through expo-during curing. The chamber is filled with 0.1 M NaCl which sure to 100% humidity air at 50°C. GC/MS was used to mon-has a conductivity of '" 1 S/m. The semicon on the bottom side itor curing as a function of exposure time. The GC/MS analysis of the sample is grounded while the other side is exposed to the shows that 4 days exposure to 100% humid air after fluid satn-sodium chloride solution which is connected to the high voltage ration is sufficient for full curing, as shown in Fig. 2.

using a platinum wire. The test setnp is shown in Fig. 1. After removing samples from aging, the plaque samples are microtomed to 250 {Lm (10 mil) thickness, dyed with Methylene B. Sample Treatment Blue (MB), and mounted on microscope slides (Fig. 3) for water tree count and size measurement. At least 10 slices are analyzed Silane treatment in this context involves both satnration of the from each plaque.

test sample with fluid and curing of the fluid within the sample using humid air. "Curing" causes the fluid within the sample to oligomerize, which fixes it within the sample. If the fluid is III. RESULTS not "cured" prior to aging, the fluid will elute into the sodium Water tree analysis shows very large numbers of bow-tie trees chloride solution. The rejuvenation fluid used in this research in the untreated plaque samples, as a result of which the number was provided by Utilx (2-2614) [l3]. The sample plaques are of bow-tie trees has been recorded without size measurement.

first immersed into silane fluid at 50°C until saturation. The Vented water trees are far fewer, and the size of each vented fluid saturation level is 7.8 wt-% for XLPE and 64 wt-% in water tree has been measured. Bow-tie water tree number den-the semicon at 50°C. As detailed in [8], rejuvenation fluid dif- sity is presented as number of trees per unit volume of insu-fuses into aged insulation and reacts with water therein to form lation, while vented water tree number density is presented as non-fugitive, hydrophobic oligomers which replace the water. number per unit semicon-insulation surface area from which it Since oligomers are larger molecules relative to the monomer, can initiate.

they have much lower diffusivity through polyethylene. For ex- Results from the untreated test set at 3500 hand 7000 h ample, the diffusivity of the tetramer (four units of the monomer (based on 50 slices analyzed for each) are consistent. An average reacted together) through PE is one tenth that of monomer at 50 number density of 49 mm- 3 bow-tie trees and 7 mm- 2 vented DC and therefore the tetramer is retained in 15 kV XLPE cable water trees is observed after 3500 h aging, while 99 mm- 3 OAGI0001273_00002

SHU AND BOGGS: EFFECT OF CABLE RESTORATION FLUID 99 o 03 /~/_----, 14 -,-----------------------------------------------------------------------------------------------------------------

. 1 // 1 1 ___ Untreated E 1 - 3500 hours0.0405 days <br />0.972 hours <br />0.00579 weeks <br />0.00133 months <br /> ~ 1: 12 ~ "....,', .... Treated (13)

i. 'j 7000 hours0.081 days <br />1.944 hours <br />0.0116 weeks <br />0.00266 months <br /> .!

E 1 ~ Treated at 3500 hours0.0405 days <br />0.972 hours <br />0.00579 weeks <br />0.00133 months <br />

~R 0.02-1i;

>.

......

iii ';

~j'",:1 >.

-+-,.

'iii

,...

101 -:

<

e l i oQ3 8 -j1

(!)  ; <

o 1 ci ~

z 6~ /'

f

.0 0.01.1 ~

- .1,

,//'

1l e

'j

-j

~ 4~, ///

2c < /

.L--'------~-----.lL~",-

<ll 2 -;~ /p'/'

>  :

0. 0.00 o 25 50

... "'-,-r, 75 100 125 240 270 1 :/

/

/"

0<&<'**************************************************,c*****

Maximum Tree Length (~lm) o 3500 Time (hours)

Fig. 4. Probability density for maximum vented water tree length in untreated sample plaques. The typical maximum vented water tree length per slice under Fig. 6. Plot of vented water tree number density as a function of aging time for examination is 50 to 75 Jim at both 3500 hand 7000 h aging time. the various sample classes. The numbers in parentheses provide the numerical value of the datum. Vented water tree number density is presented as number of vented water tree per unit semicon-insulation interface area (mm-O).

100 .,----------------------------------------------------------------------------------------------------- --------

<'") .1 - Untreated (99'1 10 - --'---C--'--y--r--y--C--'---*--T-*---c--.---c--,--.------.---C-T--*------*-----/ ....(.---r*----y-*-T-:

'EI -....:,.-... Treated

..s 80 -[ ~ Treated at 3500 hours0.0405 days <br />0.972 hours <br />0.00579 weeks <br />0.00133 months <br />

~E 10~ rr-.~-:_-.- ;-'--r-.-'--.--.-~:.::~:.:.:.r~~~~::t~~~-;;~~-*rl*~*:I-.-,-.-.-.-r-,-~ j

• s ~ . .". .:\ ._.;.:... Treat6d (3500hr5! ~ i

.{ 6J o :1 a

'q; l\

\\

.a

.~

.:3

.,

1 00 ~

j 1

.\

t.:-;.,,,>;,,)..

'"\,

\:,

~ Untreated (7000 hrs)

T rea~ed C1000 hrs}

Treated af 350D hrs(70DO hrs)

~

31 i

1 1

~

c ci  : (49) (!) 6 \ ~ 10-' l t l o

~ 40 ~ /'/'/ -"'.-"'--,,- ci

\ ~ ~ ~~~.~~\ j

~, / ..... z ,I:' \ ~.10-2j ':'. \. 41 t;;; j /'/ "--.. , -  :, \ C j .... '<~::....."":',;;;~__. ~ ~i

'"f

>

20 -[

-1 //

/ "-,,-

"-,,-

\\, \~ 10- 3 L.".".,.,,: -~:':-,-:-:~.'-J 1 S i

~ /~

/' .....

'~, [SA}.

2- \. '\

.:::....\

0 25 50 75 Tree Length (flm) 100 125 200 250~

1 o k< . . _____ .................WNNN'_._._._._._._*::Lr~'.'.:::::::::'::':::::::::::::'~:':'::::::::::~_:8:::.):

o 3500 Time (hours) 7000 o

o

.~Ar'Y2!'*~.t:-:-,"t' 25 50

75~ 1$$ t 100

.", /~"Y*i*,;t 125 200 250 Fig. 5. Plot of bow-tie water tree number density as a function of aging time for Tree Length (~lm) the various sample classes. The numbers in parentheses provide the numerical value of the datum. Bow-tie water tree number density is presented as volume Fig. 7. Vented water tree length distribution for all five sets of sample plaques number density (mm- 3 ). after 3500 and 7000 h aging at 5 kV/mm and exposure to 0.1 M NaCI solution.

The main plot has a linear ordinate while the subplot presents the same data on a logarithmic ordinate. Treated samples, whether treated prior to aging or at the midpoint of aging, contain very few vented water trees.

bow-tie and l3 mm- 2 vented is observed after 7000 h aging.

The total number of vented water trees counted in the untreated samples, 66 at 3500 hand 125 at 7000 h over an effective surface bow-tie and 0.11 mm- 2 vented at 7000 h of aging. The data area of 9.65 mm 2 , suggests that the insulation has been aged to a indicate that treatment prior to aging reduces the initiation rate sufficient degree that the effects of the rejuvenation fluid can be of vented water trees by a factor of > 100 and decreases bow-tie evaluated. The maximum vented tree length in each sample in- tree initiation by > 25 times compared to untreated samples.

spected at both 3500 and 7000 h is typically in the range of 50 to This result provides credible evidence that fluid treatment 75 {.lm or about 10% of the total insulation thickness of 508 {.lm, impedes water tree initiation significantly.

as shown in Fig. 4. The data suggest that the vented water tree The maximum vented water tree length among the few such initiation rate is approximately a linear function of time for both trees observed in treated or 3500 h treated samples is in the bow-tie and vented trees (Fig. 5). No electrical trees were found range of 50 to 87.5 Itm, while over 200 I"m long vented water in any of the cable dielectric samples examined. trees are observed in untreated samples, relative to a total insu-Figs. 5 and 6 plot the measured data for water tree density in lation thickness of 508 flm (Fig. 7). By reducing the diffusion untreated, treated, and samples treated after 3500 h of aging. rate of water into the insulation, water tree growth is apparently Without treatment, the water tree density increases linearly retarded.

with time, and large numbers of water trees occur. The treated The samples treated after 3500 h of aging contained, on samples have almost no bow-tie or vented water trees, with average, 49 mm- 3 bow-tie and 7 mm- 2 vented water trees.

an average number density of 0.54 mm- 3 bow-tie trees and After silicone rejuvenation and aging for an additional 3500 h, 0.04 mm- 2 vented watertrees after 3500h aging and 3.8 mm- 3 the samples contained only 5.4 mm- 3 bow-tie and 0.7 mm- 2 OAGI0001273_00003

100 IEEE TRANSACTIONS ON POWER DELIVERY. VOL. 26. NO.1. JANUARY 2011 vented trees, as shown in Figs. 5 and 6. Thus, the rejuvenation ACKNOWLEDGMENT not only "cured" the existing water trees, it retarded the growth The authors thank the CableCURE Division of Utilx for sup-of additional water trees so that the water tree density after plying the fluid used in this research as well as providing exper-treatment and an additional 3500 h of aging was reduced by an tise in the relevant chemistry.

order of magnitude relative to the density present at the time of treatment. At 7000 h, the density of water trees differs little between the samples treated prior to aging and those treated REFERENCES after 3500 h of aging. [1] T. Miyashita, "Deterioration of water-immersed polyethylene coated wire by treeing," in Pmc. IEEE-NEMS Electrical Insulation Con}:,

IV. DISCUSSION 1969, pp. 131-135.

[2] L. A. Dissado, Electrical Degradation and Breakdown in Polymers.

The effect of rejuvenation in inhibiting water tree initiation London, U.K.: Peter Peregrinus, 1992, p. 19.

and growth is expected for the following reasons: [3] S. A. Boggs, J. Densley, and J. Kuang, "Mechanism for impulse con-

1) Silane-based rejuvenation fluid reacts with water as it version of water trees to electrical trees in XLPE," IEEE Trans. Power diffuses rapidly into the cable insulation. As a result, it Del., vol. 13, no. 2, pp. 310-315, Apr. 1998.

[4] C. T. Meyer and A. Chamel, "Water and ion absorption by polyethylene "consumes" water in the insulation and within degraded, in relation to water treeing," IEEE Trans. Elect. Insul., vol. EI-I5, no.

hydrophilic regions created by water trees and fills the free 5, pp. 389-393, Oct. 1980.

volume within the insulation with hydrophobic oligomers [5] R. Ross, "Water treeing theories-Current status, views and aims," in Pmc. Int. Symp. Electrical Insulating Materials, 1998, pp. 535-540.

which remain in the insulation, deny water access to [6] W. R. Stagi, "Cable injection technology," in Pmc. IEEEIPower Eng.

electro-oxidized regions, and reduce the moisture diffu- Soc. Transmission Distribution Con}: Expo. Latin America, Caracas, sion rate through the insulation. Reference [10] reported Venezuela, 2006, pp. 1-4.

[7] G. J. Bertini and W. J. Chatterton, "Dielectric enhancement tech-that only 2 ppm (by weight) water content was detected in nology," IEEE Elect. Insul. Mag., vol. 10, no. 2, pp. 17-22, Mat*./Apr.

field-aged silicone treated cable insulation. This suggests 1994.

that the silicone fluid cures pre-existing water trees and [8] W. J. Chatterton and J. Dionne, "Chemical treatment of URD cables,"

impedes water tree initiation and growth by filling the in Pmc. IEEE Electrical Insulation Con}:, 2009, pp. 500-503.

[9] W. J. Chatterton and J. Steele, "A chemical and electrical analysis of free volume within the insulation to reduce the moisture aged CableCURE rejuvenated cables," in Minutes of IEEE Insulated diffusion rate. Conductor Committee 122nd Meet., pp. 45-58.

2) The silane-based fluid fills micro-voids from which [10] S. Pelissou, J. Cote, R. Savage, and S. St-Antoine, "Influence of cor-roded conductors on the performance of medium-voltage extmded ca-bow-tie water trees grow. Although dry curing has re- bles," presented at the Jicable'03 Int. Conf. Insulated Power Cables, placed steam curing, which decreases the density of Versailles, France, 2003.

micro-voids in XLPE by two orders of magnitude [14], [11] M. S. Mashikian, J. H. Groeger, S. Dale, and E. Ildstad, "Role of semiconducting compounds in the premature aging of XLPE cable the formation of micro-voids during cable manufacture is insulation," in Pmc. IEEE Int. Symp. Electrical Insulation, 1988, pp.

still inevitable as a result of byproducts of cross linking, 314-320.

creation of gas during cross linking, migration of antioxi- [12] S. A. Boggs andM. S. Mashikian, "Role of semiconducting compounds dants, etc. [2]. The long term growth of bow-tie water trees in water treeing of XLPE cable insulation," IEEE Elect. Insul. Mag.,

vol. 10, no. 1, pp. 23-27, Jan./Feb. 1994.

is diffusion limited, as once the local supply of moisture [13] CableCure 2-2614, Material Safety Data Sheet Utilx. Kent, is consumed, the only source of further moisture is by WA, Aug. 11, 2008. [Online]. Available: http://www.utilx.com/

diffusion. Thus, filling of micro-voids with a hydrophobic pdfs\MSDS_2-2614_08_11_08.pdf

[14] M. T. Shaw and S. H. Shaw, "Water treeing in solid dielectrics," IEEE oligomer impedes bow-tie water tree initiation, while Trans. Elect. Insul., vol. EI-19, no. 5, pp. 419-452, Oct. 1984.

reduction of the moisture diffusion rate reduces the growth rate of any existing bow-tie trees.

3) The moisture diffusion rate in the cable insulation may be Wen (Essay) Shu (S'09) received the B.S. and M.S.

degrees in electrical engineering from the School of reduced to a degree that water tree formation and growth Electrical Engineering at Southwest Jiaotong Univer-is not possible. In part, this could be the result of reducing sity, Sichuan, China, in 2003 and 2006, respectively, the moisture diffusion rate through the semicons, as they and is currently pursuing the Ph.D. degree in mate-are also treated. If the moisture diffusion rate of a semicon rial science from the Electrical Insulation Reseat'Ch Center, University of Connecticut, Storrs.

is reduced to that typical of a highly polar polymer, vented Her reseat'Ch interests focus on water tree phe-water trees will not grow. Unfortunately measuring the nomena in power cable and pattial dischat'ge water diffusivity through a treated material is complicated measurement and analysis.

by reactions between the fluid products and moisture. Fu-ture work will continue attempts to measure water perme-ability in treated cable materials. Steven A. Boggs (M'79-SM'92-F'93) was graduated with a B.A. degree from Reed College, Portland, OR, and the Ph.D. and MBA degrees from the Univer-V. CONCLUSION sity of Toronto, Toronto, ON, Canada, in 1972 and 1987, respectively.

The effectiveness of silicone rejuvenation for life extension He spent 12 yeat*s with the Reseat'Ch Division of Ontat'io Hydro working in the at'eas of soil thermal properties, pattial dischat'ge measurements, high field-of HMWPE and XLPE cables is well established. The present induced degradation in solid dielectrics, and SF 6-insulated systems. From 1987 contribution has improved understanding of the mechanisms by to 1993, he was Director of Reseat'Ch and Engineering at Underground Systems, which life extension is achieved and established that fluid treat- Inc. He is presently Director of the Electrical Insulation Reseat'Ch Center and the Reseat'Ch Professor of Material Science, Electrical Engineering, and Physics ment not only cures existing water trees but impedes future for- at the University of Connecticut, Storrs, and Adjunct Professor of Electrical mation of water trees. Engineering at the University of Toronto.

OAGI0001273_00004