ML20059M810
| ML20059M810 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 08/17/1993 |
| From: | Groeger J ALTRAN CORP. |
| To: | |
| References | |
| OLA-2-I-MFP-021, OLA-2-I-MFP-21, NUDOCS 9311190343 | |
| Download: ML20059M810 (7) | |
Text
)
do - 2 75/3 2 3 - OGA-2
_L-/M/2P-4 /
@pp @ekT a,(
ly, MY-13-1993 11:34 R.TRAN F1ATERIR.S Eb5 g l 95 buls.
P.61 Y
j%iuGEL, Arp 1t-
'p
'T,$p ALTRnN MATERIALS 93 0: 2E P6 :16 ENGINEERING INCORPORATED
,~
FACSIMILE MESSAGE 200 High St.
Telephone: 617/ 330-1080 or 1130 Boston, MA 02110 Facsimile: 617/ 330-1083 or 1055 i
l Date: May 13,1993 l
Time: 111:15 l
~!
4 l
l number of pages to follow:
3 TO:
Fax #: (415),972-9423 l
Firm:
Pacific Gas & Electric j
l i
Attn: Weffs Fargo l
A From:
. foe Groeger l
I I
~
)Vi*
\\
Cp "4 g J1 m;ua ucuaccRr c3M'AISSION
[f a no.nif e-a l g,.AY Ne$r
- 51- !15-tw om,sEMAnat ec ar am3L
. <. n mitrit9 /
h ?.' ;!: V d _ _.
_ ___. t 5 - n.
0% f_dW hh2bn 4-144 %
- r :-
w-..____
. pus b
eep s t L L 9311190343 930817 000002360 PDR 'ADOCK 05000275 O
PDR PAGE.0el My to 'on en?
j-MAY-14-los3 19:59 AL.TRAN MATER 1R 5 ENG P.G2 s.
REVIEW DRAFT i
May 6,1993 Mr. Wells D. Fargo Pacific Gas & Electric Company facsimile transmission
Dear Wells:
In accordance with our recent discussions, we have prepared this letter as a summary of our analyses and root cause evaluation of the 4 kV and 12 kV cable failures at the Diablo Canyon Power Plant. As you know, there are a few specimens remaining to be obtained to fill in the details, but the major conclusions have been reached.
& 4 31 %
The 4 kV cable that failed in (recent failure date) was dissected at your TES d
Center and subjected to visual examination. It was then cent to Okonite for evaluation. - Later, we obtained the specimen and subjected it to a range of visual, microscopic, chemical, and electrical tests. The cable showed no signs of mechanical damage as might have been evidenced by deformation of the copper shield tape, scuffs marks, cuts, elongation of the Jacket, or deformation of the conductor. The fault site was inspected with a stereomicroscope and no anomalies in the construction were noted at this site. No contaminants or foreign objects were found.
within the radial fault ' site. The insulation of this cable was prepared into thin sections for examination with a microscope. This was found to be void free, well mixed, and the interfaces between the insulation and shields were found to be in good condition. Some sections of the insujation were stained to reycil water trees.
d None could be found in any of the sections examined near the fault site and at_a l
distance of approximately 4 in. and 12 in. sway from the fault site. The dielectric loss characteristics of the insulation were evaluated at a number of frequencies, ranging -
from 12 H2 to 10 kHz. No anomalies were noted in the response spectra.
The chemical agents found to have attacked the jacket of the 12 kV cable were absent in the sections of 4 kV cable that were subjected to chemical analysis.' The physical properties of the Jackets of the 4 kV and 12 kV cables were found to be e
.000002861 j
7 I
- I#
- U
'E*
l l
l' P.03 nAV-14-1993 20:00 R.TRAN MATERIR.S ENG l
1 The 4 kV jackets exhibited significantly higher tensile l
significantly different.
strength than even the 12 kV specimens from the warehouse. This is consistent with a higher degree of crosslinking of the 4 kV jacket. The crosslinking density was measured using a solvent extraction technique. This confirmed that the 4 kV cable i
jacket had a greater crosslinking density when compared with the 12 kV samp The from the rect and from the switch gear end of the affected CWP I-2 cable.
higher crosslinking density would render the jacket more resistant to chemica l
degradation.
The reason for the failure of the 4 kV cables remains an open questioni A short-term high voltage test program has been designed to further explore the
" weakest links"in the 4 kV specimens presently at Okonite. Further monitoring of In the 4 kV cables will be conducted in the Cable Aging Management program.
addition,in-situ field testing of the installed 4 kV ASW caules has been proposed to l
test, for example, possible switching-induced surges on these circuits.
The analysis of the 12 kV cable failures has been completed and the root cause for the failures has been identified. The electrical failures resulted from a c induced loss of the copper shield tape. This allowed a floating shield condition to develop, with high electric stress concentration resulting at the end of the ground section of the shield. During the fault condition, high local shield temperatures resulted which further degreded the remaining jacket. The shield temperature excursion was verified by the grain growth profile in the shield tape.
Comparative chemical analyses of the jacket in th. badly damaged, less damaged, non-damaged, and warehouse specimens was conducted. Chemical test were also conducted on specimens of drain sump water, water from the affected pull boxes and ducts, chemical waste sumps, groun'd water, sea water, and water-removed from undemeath the jacket of the March 12 failed CWP cable. Tests indicated that the cables had been exposed to concentrated sea water in the duct.
)l This would have resulted from rain water washdown of salt spray from the buildings and other surface areas. The higher-than-expected sulfate level in the water samples was found to be consistent with the local' sea water in all cases.
Solvent extractions from the cable jacket indicated the presence of two fatty-acid ur' esters and two organic chloride compounds. The chemical structure and molecular weight characteristics of the fatty acid compounds are fully identified, but
)
1 060002862 2
PAGE.e03'
.. h3 9@: 48
f1AY-14-1993 20:01 R.lRAN MATERIALS ENG P.04 g
9 insufficient material was available to allow a full analysis of the chemical properties of the chlorides. 'The nature of these materials, though, is consistent with y,!
ingredients found in a cleaning agent or other surface-active material. Efforts to further identify the actual product from which they came are ongoing. The presence of these compounds is significant. As a class of materials, they would soften and eventually degrade a neoprene cable jacket. It is this attack that allowed the jacket to be breached, subsequently allowing the sea water intrusion which led to the electrical failure that was experienec' r.; e !
The initial chemical analyses of the jacket from the October 12 kV CWP cable
>l failure failed to detect the presence of any unusual chemical agent that was unique to the damaged area of the jacket. This cable, though, had been badly contaminated through use of a lubricant which had been pumped into the duct to aid removal of this cable. This material contained a wide range of components that effectively i
masked the chemical analyses. Following identification of the fatty ester / chloride material in the jacket of the March CWP failure, the jacket from the October was again analyzed, using a more refined extraction procedure. This verified the presence of the sa.ae fatty acid / chloride compounds.
o Other than where the 12 kV cables failed due to sever jacket damage, the cables weTe found to be in good condition. The insulation was free of voids, j
contaminants, interfacial defects, or trees. The insulation compound was found to be well mixed and the metallic components were free of corrosion. The jacket tear strength was found to be low in the good areas of the cable but this was j...
approximately the some as that tested on the non-aged 12 kV cable from the warehouse.
a
'W Cable jacketing materials were considered in detail from the basis of the degraded neoprene and the Hypalon used for the most recent cable replacement applications. Neoprene is a chlorinated polyethylene polymer which has enjoyed a long history of application to class 1E cables.
It is generally not used for underground power cable applications though, due, in part, to its long-term instability in water immersion applications. This material is well documented as suffering a significant degree of swelling when exposed to moisture for long periods.
Swelling increases its susceptibility to chemical attack and increases its rnoisture permeabiljty rate. Hypalon, which is chlorosulfonated polyethylene,is quite similar 000002863 3
MCY 14 '93 1E: 49 PAGE.ced
P.oS t
T1AY-24-1993 20:o1 ALTRAN MATERIALS ENG dc.ja L n in its response to chemical and moisture exposure, but offers some toIf I
improvement in mechanical properties. It, too, is prone to swelling following lon term moisture exposure. Though water has not been specifically implicated as a cause for the 4. kV cable failures, the use of neoprene or Hypalon jacketing materials ct is not what we would suggest. Instead, the cables should be protected with a l
~
moisture barrier with an overall linear low density polyethylene (LLDPE) Jacket. If this cannot be accommodated due to installation and/or engineering constraints, an Careful monitoring of the LLDPE jacket should be used in place of the Hypalon.
installed, Hypalon jacketed cables fluough the next outage is suggeste'd.
The potential link between spills from existing, known chemicals and the C
cable Jacket damage was explored at length and no connection was found. The FA
, attacking agent is chemically inconsistent with any of the known materials stored g
near the duct bank. Further, the chemical agent was not found in ar, of the pu))
..p box water samples nor in any of the sumps. The concentration of the ester / chloride material in the damaged jacket is so low at this time that it would be unlikely to cause further degradation. This would suggest that its introduction to the cable may have been a single event and that it was not recent.
The sulfide-induced jacket degradation hypothesis suggested by Okonite has been considered and found to be inconsistent with all of the circumstances, chemical analyses, and site conditions. A letter from us reviewing their hypothesis and comparing it with our tests has been prepared and submitted on April 24 for your cmideration.
i Our analyses of the volds resulting from the high voltage breakdown tests conducted at CTL indicates that they are not artifacts of the insulation sectioning operation. Similar voids, however, were not founc in any of the field-failed or non-failed cable sections that we examined. In our experience, the voids found in the high voltage breakdown test specimens could be formed during the fault conditions.
For some EPR formulations, when significant moisture is present, this is a normal, This issue will be briefly explored in the short-term high expected occurrence.
voltage tests that we will be conducting or supervising. Since the possible presence of voids is a concern for the condition of the remaining in-plant cables,_ their significance must be addressed.
.u 000002664 4
PAGE.005 MAY 14 '93 16: 49
MAY-14-1993 22:22 R. TRAM r1ATERIR S EM3 p, %
9 Summary and closing remarks, etc. etc.
Final report date, follow-up I
supplement on high voltage testing.
Sincerely, Joseph H. Groeger VP (please list all PG&E recipients) cc
/
t i
4 s
000002865 5
TOTR. P.06 MAY 14 '93 16:50 PAGE.006 l
L
I, l
To:
Wells D Fargo@NECSEZ@SFNPG Cc:
k RAM 7@ Chem @TES 6
.m:
Subject:
Date:
Thursday, March 18, 1993 18:00:09 PST Attach:
Certify:
N Forwarded by:
- Wells, So We had a very short but productive meeting with Joe today, here at TES.
far, our chemical results are in the same ball park as his, looking at preliminary tests.
I also found out today from Richard Potter (DCPP Environmaental Coordinator) that on 6/21/92, there was an accidental mixing of con. sulfuric acid and resulted in an exothermic reaction causing some con. sodium hydroxide that details are in an A/R for operatians. Maybe Chuck Short spilleage. The exact can help Vern track this down on Friday ?
with Joe that the After looking at the samples of cable I am in agreement I will FAX was chemically attacked. We are still working on the source.
jacket down to Vern tomorrow for your meeting. We are also the latest test info.
dring some other lab tests, and we will keep you appraised.
I will be home most of this weekend. Remember what happened last Friday!!
Phones (510) 828-7881, home (510) 866'-5303, work and voice mail 8-251-5303, Co.
-Rich McCurdy (XiOOU2844 i