IR 05000324/1985035

Insp Repts 50-324/85-35 & 50-325/85-35 on 851002-04 & 1113-15.No Violation or Deviation Noted.Major Areas Inspected:Performance of Unit 2 MSIV Solenoid Valves, Including Failures of Seat/Seal Matl

ML20138Q489
Person / Time
Site:	Brunswick
Issue date:	12/05/1985
From:	Jape F, Schnebli G NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
To:
Shared Package
ML20138Q471	List: IR 05000324/1985035 ML20138Q469
References
50-324-85-35, 50-325-85-35, NUDOCS 8512270255
Download: ML20138Q489 (10)

Text

(-

.

. UNITED STATES

[AMEIoq'o NUCLEAR REGULATORY COMMISSION

[" n R EGION 11 101 MARIETTA STREET, $ j

• t ATL ANTA, G EoRGI A 30323

'+ ,. . . . . ,o Report Nos.: 50-325/85-35 and 50-324/85-35 Licensee: Carolina Power and Light Company P. O. Box 1551 Raleigh, NC 27602 Docket Nos.: 50-325 and 50-324 License Nos.: DPR-71 and DPR-62 Facility Name: Brunswick 1 and 2 Inspection Conducted: October 2-4 and November 13-15, 1985

, 8, /2/f/8f

'

Inspector:

G. A. Schnebli

~

Date Signed Accompanying Personnel: *F. Jape, Region II, *V. Thomas, IE-HQ, and

• V. Hodge, IE-HQ l

• Attended meeting at Harris Energy Center on November 13, 198 Approved by: Am -; kusep "

/MJ T F. Jape, Section Chief

~

gf Date Signed Engineering Branch Division of Reactor Safety SUMMARY Scope: This special announced inspection entailed 31 inspector-hours on site and 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> at Harris Energy Center in the area of followup on problems associated with several Unit 2 Main Steam Isolation Valves (MSIV) failure to clos Results: No violations or deviations were identifie .

8512270255 Q [200g24 PDR ADOCK PDR G

__ _

.

.

REPORT DETAILS

~ Persons Contacte'd Licensee Employees

• J.-Allen, Technical Aide - Regulatory Compliance R. J. Bloch, Project Engineer - Metallurgy Unit W. Bracey, Specialist - Electrical D. W. Brinkley,-Engineer - Metallurgy Unit W. D. Brown,-Project Specialist - Analytical Chemistry

• J. W. Chase, Assistant to General Manager R. A. Coutant, Engineer - Metallurgy Unit

• C. R. Dietz, Plant General. Manager

• D. Hickman, Project Engineer - Nuclear Systems

• B. E. Hinkley, Manager - Technical Support

• D. E. Novotny, Senior Specialist - Regulatory Compliance

8. Parks, Engineering Supervisor B. Phillips, Consultant Engineer B. Poulk, Senior Specialist - Regulatory Compliance

• A. L..Richards, Principal Engineer - Quality Assurance / Quality Control QA/QC W. Schade, Senior Engineer - Electrical Systems .

D. Sullivan, Principal Engineer - Metallurgy Unit G. Thompson, Project Engineer - Electrical Systems Other licensee employees contacted included engineers, technicians, mechanics, and office personne NRC Resident Inspectors W. Ruland, Senior Resident Inspector L. Garner, Resident Inspector

• Attended exit. interview 2. -Exit Interview The inspection scope'and findings sere summarized on November 15, 1985, with

'those persons indicated in paragraph 1 above. The inspector described the areas inspected and ' discussed in detail the inspection finding No dissenting comments were received from the license Inspector followup item, 324/85-35-01, Followup on problems associated with MSIV dual solenoid valves including EP seat material and open coils, paragraph 5 The licensee did not identify as proprietary any of the materials provided to or reviewed by the inspector during this inspectio .

.

3. Licensee Action on Previous Enforcement Matters This subject was not addressed in the inspectio . Unresolved Items Unresolved items were not identified during the inspectio .5. Followup on the Failure of Several Unit 2 MSIV Solenoid Valves (617ttl, 92700,93702) Failure of Ethylene Propylene Seat / Seal Materia).

This reactive inspection was initiated on October 2,1985, te ft.iiewp on the failure of three MSIVs to fast close. The failurrs were identified on September 27, 1985, during the performuch of Periodic Test (PT) 25.1, Nuclear Steam Supply System Main Ster.m and Feedwater

-

Isolation Valve Operability -Tes This test was reauired prior to returning the Unit to service, which had been placed in cold shutdown on September 26, 1985, as a precautionary measure against the then approaching hurricane Glori In performance of the test, MSIVs 2-M21-F028A, F022C and F028C failed to fast clos Initial troubleshooting of the failure isoiated the probable cause-to the double solenoid valves on the MSIV operator's (Automatic Switch Company (ASCO) Model No. 8323A36E). Licensee engineers performed a detailed visual inspection of the three solenoid valves associated with the failure, which showed the following results:

-

F028C - The valve disc on the de solenoid portion of the valve appeared to be very brittle and had a piece missing from its center. This piece was found wedged in the valve exhaust port, thus blocking it off. The valve 0-ring on that side was also very brittle and came out in small pieces. The valve disc and 0 ring in the ac end of the valve was not degraded to the same extent encountered on the de solenoid side of the valve. The 0-ring on the ac end came out in one piece; however, its lower half was stuck to the valve body. The valve disc on the at end had minor indentations where it had contacted the supply port but it was i r. tact . This disc is normally not in contact with the supply port sea The disc on the dc end of the valve is normally pushed against the exhaust port seat. The valve body appeared discolored and there was a fine layer of dust on its internal surface F022C - The valve disc on the de end of the valve was found adhered to the exhaust port. When the disc was freed, it was found to have a severe indentation where it had contacted the exhaust port and a raised nipple existed in the center of this region. The valve disc was removed in one piece. The valve r

~

.

.

.

0-rings were -intact and the disc on the ac end of the valve and the valve body appeared to be similar to that encountered on the F028C-valve.

..

^

-

F028A - The valve disc on the dc end of the valve also had a piece missing from-its center as encountered with F028C; however, this piece could not be locate The condition of the valve 0-rings, the ac solenoid valve disc, and the valve body were similar to those found on the F028C valve solenoi ~

Based.on the results of the visual inspection, the licensee determined that.the failures of the actuator solenoids were caused by:

-

F028C - Valve failed due to the exhaust port being blocke F022C - Valve failed because the solenoid valve disc was adhered

'

-

to the valve seat and the disc could not mov F028A - Valve failed either because the solenoid valve disc was stuck to .its seat or because disc material had broken off and plugged the exhaust port and was subsequently dislodged during troubleshooting of the valv The licensee took valve F023C to the valve vendor (ASCO) for an ev 1uation. The licensee stated that ASCO concluded that the elastometer i .ne valve seats and seals, ethylene propylene (EP); had degraded due to contamination. This conclusion was based upon evidence that the elastometer only degraded where it was in contact with the brass pieces of the valve. The valve disc on the ac solenoid portion of the valve, which is normally open and therefore not in contact with the supply port, was not nearly as degraded as the dc solenoid valve disc. The ,

valve 0-ring on the ac end was stuck to the brass body; however, the portion that contacted stainless steel was removed in one piece. The vendor did not determine the contaminant; however, they believed it was not introduced during manufacturer or assembly of the solenoid valve assembl Based on vendors conclusion that the failure mechanism was caused by a contaminant, the licensee reviewed past records of samples taken from the air system which indicated the maximum hydrocarbon content in the

'

air te be approximately 7 parts per million (PPM). Additionally, air samples were obtained from the air accumulators in the vicinity of the solenoid valves and showed no appreciable hydrocarbon content. Labora-l tory analysis of the three subject valves showed a significant amount of hydrocarbons in the valve body of the F028C solenoid valve. The precise hydrocarbon has not yet been identified.

l This particular model solenoid valve was installed in Unit 1 in June 1983 and in Unit 2 in August 1984 for the Environmental Qualifi-

, cation Program (EQ). The Unit 1 solenoids were subsequently replaced

'

during tne 1985 outage when the MSIV actuators were replaced. The new solenoids were identical to the old ones, with the exception that the

. _ _ _ . - _ . __ _ __ _ _ __._ -

_

..

.

new solenoids contained Viton seats and seal materials in lieu of E The information currently available from the vendor shows the following for the specific materials:

-

Ethylene Propylene - Rated for service to 300 F, EP is resistant to high levels of radiation and is therefore the material of

. choice for EQ applications. However, EP absorbs hydrocarbons, and like a sponge, softens and swells up. This exposure also lowers l

~ its' meltir.g ~ point. It is undesirable in applications where

-

exposure to hydrocarbons is likely. Overheating of EP causes it to turn brittle. General Electric (GE) is utilizing EP in its EQ

. program Viton - This material is rated for use to 400 F and is impervious to hydrocarbon Its major disadvantage is that it is less resistant to radiation than EP by a factor of ten. GE and ASCO recommend Viton for applications that are not oil-fre A licensee review of the Brunswick maintenance history of these and similar ASCO solenoid valves produced only one occurrence which could have:been attributed to the.same cause. This occurred in August 1984, on Unit I and a cause was not determined. Prior to this event, there were no recorded events involving the newar type' valves on Unit 2 which indicated a failure of the EP materia Based upon a review of the available literature and industry experience including NOTEPAD, NPRDS, vendor, other utilities, and preliminary

' failure analysis results, the licensee replaced all of the Unit 2 dual solenoid valves with valves having Viton seals. Their justification was as follows: The seat material in the failed valves (EP) swells and changes its characteristics when exposed to hydrocarbons. One such change is a possible lowering of its melting poin . Hydrocarbons were found in the valve bod . The failed seat material (EP) showed evidence of being heated to the melting poin . The failed disk seat material had become lodged in the exhaust ports of the failed solenoid valve . Based upon items 1 through 4, the failure mechanism appeared to be a ' combination of hydrocarbon contamination, temperature, and -

internal valve geometr . Viton,.another seat material is impervious to hydrocarbons and has a higher temperature rating than EP. However, Viton is more susceptible to radiation effect Viton has a recommended

. radiation-related changeout of 3.3 years, whereas EP has approxi-

--

.

.

mately a 4.2 year changeout based on the valve environment at Brunswic /. Viton is recommended for use in systems that are not oil-fre In addition to replacing tle subject valves the licensee has performed the following: A sample of other ASCO solenoid valves in the plant were disas-sembled and inspected. No deterioration of the seals was identi-fie . An exhaustive failure analysis program is being conducted at the Harris Energy and Environmental Center (HE&EC) Metallurgical Laborator See paragraph 5.c for details of the inspection performed at HE&E . The licensee committed that during the first scheduled outage for either unit of sufficient length, following six months of opera-tion, a sampling of the MSIV double solenoid valves will be replaced. The removed solenoids will be evaluated as part of the continuing failure analysi In addition, all of these valves will be replaced at approximately three year intervals due to the radiation susceptibility of Vito *

b. Failure of Solenoid DC Coils Following replacement of the ASCO valves with Viton seal material, the licensee has experierced several failures of the dc coils on Unit On October 5, 1985, the de coils associated with F022D and F0288 failed during the performance of post maintenance tes;ing required after replacement of the valves. Troubleshooting indicated an open circuit in the de coils. The coils were replaced and the valves subsequently retested satisfactoril On October 15, 1985, an unplanned closure of F022A occurred while Unit 2 was operating at 99% power, which in turn caused a reactor scram. At the time of this event the licensee was performing Periodic Test PT-01.1.12P, Reactor Protection System and Primary Containment Isolation System. Closure of F022A occurred when the ac solenoid coil on the actuation air supply solenoid valve for F022A was de-energized in accordance with the test procedure. At the time, it was not known that the corresponding dc coil on the solenoid valve had failed at a prior undeterminable tim Consequently, wnen the ac coil was de-energized, closure of F022A resulted. The failed coil was replaced and subsequ-ently retested satisfactoril Initial investigation of the de coil failure was unable to detect a failure mechanis _

_7_

..

Due to the number of failed coils the licensee initiated a temporary surveillance program on October 16, 1985, to monitor operability of the-solenoid coils. The program involved a temporary modification to g install a resistor in the supply line to each coil. Continuity of the coil circuitry can then.be verified by measuring a voltage drop across the resistor. The licensee is presently using this system to check the-circuitry on a daily basis and prior to performing a test which involves use of these coils. Use of this system is accomplished directly from cabinets in the control roo On October 20, 1985, F028C showed an open dc coil during performance of the daily resistor chec Since the unit was operating, trouble-shooting on the circuitry could not be performed; therefore, the C main steam line was isolated to prevent an unplanned closure of the valv Troubleshooting will be accomplished when the unit is shutdown to deterrnine if the problem is in the coil or in another portion of the ci rcuitr Two of the failed coils (F022D and F022A) and several coils from ready spares were sent to HE&EC for ft.ilure analysi See paragraph below for preliminary results of this analysi Failure analysis information will be provided by the licensee as it becomes availabl Failure Analysis Conducted at Harris Energy and Environmental Center Metallurgical Laborator As previously stated in paragraphs 5.a. and b, HE&EC commenced a failure analysis of the solenoid valves in the areas of the failed EP material and the open de coils. The analysis is presently in progres The initial results obtained by'the licensee, were presented to the NRC at a meeting on November 13, 1985, the meeting was attended by personnel from Region II and IE Headquarter The meeting lasted the entire day and was considered to be beneficial and informative to all parties concerne . The following facts were discussed concerning the failure of the EP seat / seal material:

-

In additicn to the three solenoid valves (F028C, 22C and 28A), the remaining Unit 2 valves and all Unit 1 valves, which were replaced during the 1985 outage, were disassembled and inspected. Most of the valves indicated some signs of seat degradation, although not as severe as the initial three that faile The outboard valves seats and seals were degraded more than the inboard valve .

-

.

This observation led the licensee to believe one of the contributing factors was heat related, due to the fact the temperature in the MSIV pit is usually higher than drywell temperatur Drywell ambient temperature normally is less than 135 F, a T/S limit, and the MSIV pit has exceeded 165 F on several occasions due to an inadverter.t loss of ventila-tio Another aspect of temperature, determined by the licensee, was self-heating of the valve due to current flow in the coil Four thermocouples were placed inside a valve, then the valve was heated in an oven to an ambient temperature of 160 F to 180 F corresponding to the maximum temperature expected in the MSIV pit. With a coil voltage of 125V de the temperatures inside the valve ranged from 257 F-to 290 F and with the voltage increased to 140 V de the temperatures ranged from 265 F to 290 F. The test voltage of 140 V de for the upper limit was chosen as this would be the highest voltage that the coils would be exposed to because 140V dc is the voltage obtained during a battery charge (equalizer charge).

-

Testing of new samples of EP material showed that the material fails at 675 F when heated on a hotplat The failure mode was nearly identical to the failures observed in the solenoid valves. During testing the elastomer in the EP material is driven off, leaving only the carbon black filler mater.al. This was also demonstrated using thermogramametric analysis, in which a sample is heated while weighing. The EP material showed a rapid mass loss at 225 to 325 Laboratory analysis of the valves showed a significant amount of hydrocarbons in the valve body of the F028C valve. The

- precise hydrocarbon or amount could not be determined. The possible sources of hydrocarbons identified by the licensee included contaminates in the air system, contaminated during

,

manufacture, and the use of pipe dope during installatio As stated previously, air system samples taken near the valves did not identify hydrocarbon Test samples of EP material were soaked in oil and analyzed. Test results did not indicate failure until the oil soaked EP was also exposed to heat and in this situation the material degradation only occurred at the point of contact with the heat source. The

- results of this test. led the licensee to believe that heat was the major contributor to the failure mechanis Samples of Viton material were analyzed in parallel with EP and in all respects the Viton proved superior.

f I

p

.

'

[ _

? ..

,#

I I

8 j'

L b -

The licensee contacted the supplier of the EP material (Minnesota Rubber) and they stated that their data indicated EP material should be replaced after 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> at 250 F or 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> at 300 F. These figures differ from the informa-tion currently provided by the vendor and will be looked into L further as part of the followup to this inspection (see l~ paragraph 5.d. below).

t

-

In addition to the valves inspected that came from the

!

Brunswick plant, Grand Gulf provided one solenoid valve of i the same type for inspection by HE&EC. A visual examination,

!' under a microscope, showed that the EP seat material exhibited a noticeable dimple in the area where the seat contacts the exhaust por However, the EP seat / seals were still resilient and showed no visible signs of degradation. It should be noted that this valve was only in service approxi-mately eight months and the ambient temperatures (<135 F) in the vicinity of these valves at Grand gulf is somewhat cooler than at Brunswic (2) The following preliminary findings were also discussed concerning the failure of several de coils:

-

As previously stated, two coils that failed and five coils from ready spares were sent to HE&EC for failure analysis.

!

-

One coil from the lot of spares indicated an open circuit, tie other four were satisfactory. Visual examination of this failed coil showed exterior damage to the coil; it appeared that the coil had been dropped or damaged by an external forc The failures of the remaining two coils appeared to be separation of the very fine coil wire at the point it connects to the larger pigtail lead. This connecting point is a soldered connection which is then taped and lacquere Under magnification the separation appeared to be corrosion induced. Further analysis of the coils at this junction indicated the presence of chlorides in four of the coil The licensee is continuing its efforts to determine the source of the chloride The effects of voltage on the coils were also discussed; however, this aspect was dismissed as a significant contri-butor because the coils are rated for a voltage of 90V to 140Vdc as stated in paragraph 5.a. A maximum voltage of 140Vdc is expected during a battery charg In addition, the voltage at the solenoid valves would be somewhat lower due to line losses in the long run of wire from the battery to the valve.

I

g_~

~

. .

} 9 A

-

Failure analysis of the coils is continuing .and additional

-

information will be provided by the licensee as it becomes available, Based on the licensee's preliminary conclusions and failure analysis completed : to date, the licensee's action in replacing the subject valves containing EP material with valves utilizing Viton appears to

'be the. proper course of action. At present, the'NRC is preparing an IE Notice discussing this event. Further actions will depend upon the final results obtained from~ HE&EC and discussions within the NRC. -This area will continue to be followed by the NRC and is identified as Inspector Followup Item 324/85-35-0 _

-

t

. ",

L.