ML20214R439

Rev 1 to Employee Concerns Task Group Operations, Mechanical Equipment Reliability/Design,Kerotest Valve Leakage & Corrosion

ML20214R439
Person / Time
Site:	05000000, Sequoyah
Issue date:	09/19/1986
From:	Lagergren W, Mccomb S, Scott R TENNESSEE VALLEY AUTHORITY
To:
Shared Package
ML082421095	List: ML20214R399 ML20214R413 ML20214R423 ML20214R439
References
1601T, 301.01-SQN, 301.01-SQN-R01, 301.01-SQN-R1, NUDOCS 8609290175
Download: ML20214R439 (10)

Text

_

i i

I TENNESSEE VALLEY AUTHORITY SEQUOYAH NUCLEAR PLANT l

EMPLOYEE CONCERNS TASK GROUP OPERATIONS CEG j

}

i Subcategory: Mechanical Equipment Reliability / Design (301)

Element: Kerotest Valve Leakage And Corrosion Report Number: 301.01-SQN Revision 1 EX-85-085-003 IN-85-594-001 IN-86-285-001 XX-85-090-001 XX-85-090-002 f

f-/f-86 Evaluator:

Scott McComb Reviewed by: [

//

9-/,

((

~

ber Date OPS C,E6,M Approved by: S'd_bl 4, 7[/f[thd

[]W.'R.Lagergren

' Ddte t

8609290175 860923 PDR ADOCK 05000327

)

P PDR 1601T l

l

I

}

Revision 1 l

I.

Kerotest Valves Leakage and Corrosion j

f This report evaluates the use of Kerotest valves in CSSC systems at the

{

i Sequoyah Nuclear Plant. The five employee concerns which address this problem have identified Sequoyah (SQN), Watts Bar (WBN), and Bellefonte Nuclear Plants (BLN). The concerns IN-86-285-001 IN-85-594-001, i

XX-85-090-001, and EX-85-085-003 are considered generically applicable to SQN. The NSRS reported on the " Generic Implications of Watts Bar Kerotest Valve Problems on SQN" (I-85-861-SQN, Revision 1) when they investigated concern XX-85-090-002.

In all cases, these valves were reported to have leakage, corrosion, and seating problems.

1 l

}

II.

Specific Evaluation Methodology A.

Concerns / Perceived Problems i

The five employee concerns evaluated in this report are as follows:

]

IN-86-285-001: " Watts Bar unit 1 & 2.

Globe valves (Kerotest) were received from vendo: in a corroded condition due to vendor's hydro of valve and inadequate i

drying. These valves leaked after installation. A i

generic NCR was written to correct this problem but the full implementation of the NCR disposition is I

questionabin.

Examples of the systems with these valves are:

CVCS, Safety Injection, RHR & Reactor j

Coolant. Construction Dept. Concern. CI has no j

further information."

IN-85-594-001:

"3/4" Kerotest valve (possibly globe) 30 valves inspected with a 901 reject rate - bearings were missing / busted / frozen. These valves were installed throughout the site (units 1 & 2) and all may not j]l have been identified as evidence of an NCR, hold tags, or further investigations was not known.

(Names / details known to QTC)"

4 XX-85-090-002: "Sequoyah: Unit 1 and 2.

Per CI TVA used globe valves (Kerotest) extensively in both plants, Watts Bar and Bellefonte and had leakage & corrosion t

l problems. CI questions the usage of these valves i

at Sequoyah - the sister plant - for leakage &

j corrosion problems. The systems to be checked as examples are CVCS, Safety Injection. RHR & Reactor 4

t l

Collant, etc. CI has no further information. Nuc

]

Power Concern."

t I

1 I

J l

j Page 1 of 9 i

Revision 1 XX-85-090-001:

"Bellefonte all units. The globe valves (kerotest) need to be checked for corrosion & leakage due to vendors hydro and inadequate drying. Examples of these systems are:

CVCS, Safety Injection, RHR, Reactor Coolant.

CI stated this problem has existed for six years. Construction Dept.

Concern. CI has no further information."

EX-85-085-003:

"Kerotest valves are extremely poor. They seldom seat properly.

Construction Dept. Concern. CI has no additional information.

B.

Methodology The evaluation of these concerns, the associated NSRS evaluation, and line response concentrate on five aspects.

1.

The validity of the concern.

2.

The adequacy of the NSRS cvaluation report and the recommendations contained therein.

3.

Verification of the line response and corrective action taken by line management.

4.

The adequacy of corrective action to resolve the concern and prevent recurrence.

Appropriate regulatory requirements and SQN specific procedures were reviewed to ensure prog.ammatic compliance for safety valve availability and operability.

Interviews were conducted to verify any actions required to close this issue at SQN.

Revision 1 to this report to incorporate Senior Review Panel l

comments was performed by an evaluator who was not part of the lR1 original evaluation.

1 III.

Findings A review of the referenced documents and interviews with cognizant personnel revealed the following:

1.

NSRS Evaluation Findings-lR1 The NSRS report has generated the following findings as a lR1 result of its investigation into this problem.

l Page 2 of 9

Revision 1 III. Findings (continued) a.

A problem with Kerotest valves at WBN was identified and l

documented in a Division of Construction Nonconformance l

Report (NCR) - 2501R (Reference 2).

l l

b.

In the final report on this problem (Reference 5) the lR1

" Description of the Deficiency" identified several hundred 3/4,

l l, and 2-inch valves with leakage and corrosion problems. The l

" Safety Implications" section of the report states:

l "While some of the subject Kerotest valves are installed in essential safety-related systems: CVCS, SIS, RHR, UHI, RCS, and CSS, operation of the valves is not required for the safe shutdown of the plant during a loss of coolant accident. As a result TVA could identify no valves that perform a safety function. However, to document this, a failure effect analysis was performed. The analysis identified no detrimental effect on plant safety as a result of the failure of any of these valves."

The report also stated in the " Corrective Action" Section:

"TVA does not consider the valves to perform a safety function; however, we do believe that the corrosion identified with the valves could result in a maintenance problem during the life of the plant. Therefore, TVA has instituted a maintenance program to dismantle, inspect, and replace parts as required for those valves installed at WBN."

In addition, the report noted the " generic applicability of the corrosion problem" and stated:

" Verbal discussions with personnel at SQN indicated that during normal disassembly for maintenance, they have identified no Kerotest valves with what they consider excessive corrosion."

c.

The response to NCR 2501R does not address the safety l

consideration of a one percent fuel failure as stated in the l

safety analysis. If the valve stems leaked, and the valves I

are in areas where they might inhibit personnel entry, this I

in turn might impair safety functions.

I lR1 d.

The letter (Reference 6) which transmitted the above report to l

NRC in accordance with the requirements of 10/CFR 21 did not l:

mention whether the WBN procurement specification was deficient l

by not requiring valve packing removal and bonnet drying after l

hydrostatic testing.

l Page 3 of 9 i

l Revision 1 e.

A memorandum entitled, "Sequoyah Nuclear Plant Units l

1 and 2 - Deficient Kerotest Y-Type Globe Valves - Report l

No. 4 (Reference 7)," responded to the commitments made l

in NCR-2501 to follow up on generic implications of the l

Kerotest valve failures. This memorandum stated:

l l

"We have reviewed the Kerotest valve installation at i

Sequoyah Nuclear Plant and find that these valves' safety l

function would not be compromised by problems developing l

from a water saturated stem packing. Westinghouse l

specified packing replacement after hydrostatic testing l

on their procurements."

l l

f.

The memorandum also stated that most of the valves at l

SQN were procured by Westinghouse and implied that there l

was no problem at SQN. The Westinghouse specification l

applicable to these valves at SQN is G-678824 Revision 1.

l Westinghouse also had an applicable cleaning and l

cleanliness specification which did not specify packing i

removal. The Westinghouse procured valves were dedicated l

i to the Nuclear Steam Supply System (NSSS).

l lR1 i

g.

Most of the remaining small valves not initially l

)

purchased for SQN by Westinghouse were procured by NAVC0 l

as part of the principal piping contract 71C37-92615. TVA l

l specification 9923 (no revision) was used on this contract.

l This specification did not require that valve bonnets be l

dried out or packing removed after hydrostatic tests. The l

6 remainder was purchased directly from Yerotest on TVA l

contract 79K3-824147. TVA General Specification 3640 was l

made applicable to this contract.

Paragraph 7.10 of this l

specification requires removal of the packing after l

hydrotest and that valves shall be shipped without the l

packing installed.

l l

h.

A follow-on purchase of small Kerotest valves for SQN was l

j made under TVA contract 84P73-836406. The specification l

applicable to this contract was MEB-SS-10.19, Revision 0.

l This specification also requires removal of the packing l

i after hydrotest and that valves shall be shipped without I

the packing installed.

l i

4 i

Page 4 of 9

Revision 1 1.

Most of the Kerotest valves used at WBN were also l

procured through the Westinghouse NSSS contract. Many l

WBN valves were also purchased through Dravo on the I

i principal piping contract. Dravo, in turn, issued l

purchase order (PO) E2897/98-KN-1 which included TVA l

Design Specification WBNP-DS-19835-1521-CX, Revision 2, I

as a technical requirement. This specification does not i

require removal of the packing after hydrotest. TVA also l

made a direct purchase of Kerotest valves for WBN by l

contract 79KA2-824589-4. TVA General Specification 3640 l

was also applicable to this contract with requirements as l

noted previously.

l l

j. A search of TVA's Equipment Identification System for l

Kerotest valves installed at SQN found 1,528, two inches I

or less, Westinghouse purchased valves. A second search I

for non-Westinghouse valves found over 500 Kerotest globe l

valves installed in SQN CSSC systems.

These valves are l

used extensively in the ERCW system and as instrument l

isolation valves, root valves, sample connections, vent l

valves, and drain valves for other systems.

l lR1 k.

A space Kerotest Y-type 1-inch globe valve was found in l

the ECN warehouse at SQN and disassembled by Mechanical l

Maintenance (MM). A clear liquid that looked like water I

was in the bonnet area. The bearing on the stem was l

corroded. Rust was visible on parts inside the bonnet.

I The bearing would rotate, but it was not smooth.

l Photographs were taken of the valve parts and will be l

l retained in the NSRS flies.

l l

1.

Three former TVA field inspectors agroed that packings l

were generally removed at the various suppliers because l

it was good practice and not because it was stated in the I

specifications. The asbestos / graphite packing generally l

used during the time period when these valves were shipped I

would retain water; and, hence, could cause corrosion l

and/or seizure of the valve stem.

Source inspection of I

smaller valves was sometimes waived making their condition l

uncertain.

l 2.

NSRS Recommendations a.

1-85-861-SQN-01, Kerotest Valve Inspection Perform an inspection of representative Kerotest valves at SQN.

If corrosion is found in the stem area, or stem leakage is found, an engineering evaluation should be performed to determine the reportability and proper resolution of this problem.

If corrosion and leakage are not found, an evaluation l

should be performed to determine why the difference exists between WBN and SQN valves.

Page 5 of 9

i a

Revision 1 l

OE should evaluate the' methods used to determine generic i

applicability and not rely solely on verbal information such as j

that received from SQN.

j Justifie4 tion The condition of the Kerotest valves is uncertain at SQN, I

l similar to WBN, and potentially reportable. NCR-2501R I

l documented similar problems at WBN.

1,528 Kerotest valves I

were purchased by Westinghouse for NSSS. 500 Kerotest l

valves were purchased by others for CSSC systems.

lR1 l

l The Westinghouse and TVA specifications are inconsistent l

l regarding packing removal and valve drying after l

l hydrotesting.

l b.

I-85-861-SQN-02, NRC Reportability i

i SQN should provide objective evidence that the statements 3

regarding the Westinghouse testing procedure contained requirements for drying out bonnets and/or replacement of j

bonnet packing after hydrotesting were actually in the j

applicable Westinghouse specif1 cation.

1 j

WBN should notify the NRC in accordance with 10 CFR 50, Part 21 of the root cause of the WBN valve failures. This is 4

2 evidently a lack of uniform enforcement of specification requirements to be taken after hydrostatic testing, either by contract or by inspection.

Justification A spare Kerotest Y-Type Globe Valve from Sequoyah Nuclear Plant I

was disassembled and corrosion and rust were found inside the lR1 bonnet.

l c.

I-85-861-SQN-03, Analyze One Percent Fuel Failure Accident The impact of Kerotest valve stem leakage coupled with the one i

percent fuel failure accident stated in the safety analysis should be investigated, l

Justification The response to NCR-2501R does not address the safety I

consideration of a one percent fuel failure as stated in the safety analysis.

I i

i Page 6 of 9 l

r E

/

Revision 1 3.

Line Response to NSRS Recommendations (Reference 9)

SON Response to I-85-861-SON-01 Based on the search of the maintenance requests (MR) filed at SQN since the plant went into operation, only one Korotest Y-type globe valve has experienced failure due to leakage or corrosion problems. Approximately 1,500 Kerotest globe valves (two-inch and under) are listed as installed at SQN.

In addition, the search conducted by the NFRDS (a nationwide data base for operating nuclear plants) reveals that only four Kerotest valves out of over 1,600 in thale data base experienced failure due to corrosion over the past 10 years.

In view of the above information, an inspection of Kerotest l

valves at SQN is not warranted, and this review satisfies the lR1 2

requirement for an evaluation of generic applicability at SQN.

l SQN Response to I-35-861-SON-02 As referenced in the NSRS report (section B.3), a memorandum from R. W. Cantrell to J. A. Raulston dated September 10, 1981, stated that " Westinghouse provided the majority of the valves installed in safety systems of SQN. Westinghouse specified packing replacement after hydrostatic testing on their procurements." In addition, section B.6 of the NSR$ report states, "... purchased directly from Kerotest on TVA contract 79K3-824147. TVA General Specification 3640 was made i

applicable to this contract.

Paragraph 7.10 of r.his specifiestion requires removal of the packing after hydrotest and that valves shall be shipped without packing installed."

Section B.7 of the NSRS report also states that a follow-up purchase of small velves from Kerotest required removal of the packing after hydrotest, and that valves shall be shipped without the packing installed. Based upon the above information and the information contained in the response to Recommendation I-85-861-SQN-01 on the essentially non-existent failure rate of Kerotest valves at SQN, a search to provide objective evidence concerning the Westinghouse testing procedure is unwarranted and would provide no useful information.

I Page 7 of 9

Revision 1 SON Response to I-85-861-SON-03 The safety analysis report indicates evidence of seat leakage, not stem leakage.

In order to have stem leakage, both the i

packless metal diaphragm and the stem backup packing would have i

to fail. The failure analysis which was previously performed and documented identified no detrimental effect on plant safety as a result of the failure of any of these valves.

I Based on the documented results from the failure analysis l

and the backup packing arrangement of the Kerotest valves, l

it would be both repetitious and unwarranted to further l

investigate this issue.

l I

j 4.

Conclusions l

i l

}

Based on the findings, the NSRS has substantiated that there l

is a problem with Kerotest Y-type globe valves corroding if the 1

j packing is not removed and valve body dried after hydrotesting.

l The findings also indicate that due to the inconsistencies of l

the purchase contracts for the Kerotest' valves at Sequoyah j

Nuclear Plant it cannot be assured that the packing was removed l

3 and valve body dried prior to shipping. However, only one out l

j of 1500 Kerotest globe valves installed at Sequoyah has

]

experienced failure due to corrosion indicates that the problem l

1 l

that Watts Bar has had with Kerotest valves is not generic to lR1 Sequoyah. Inerefore, this evaluation is in agreement with the l

N3RS report I-85-861-SQN with the following addition. Due to I

the inconsistencies of Kerotest purchase specifications and l

l purchase orders, a review of the current specifications should l

l be initiated to ensure they are consistent and the requirement I

is included for ensuring the valves are dry and packing is l

j' removed after hydrotesting for future purchases. The l

=

recommendations from these concerns do not affect the safety l

of plant and therefore should not have an impact on plant l

startup.

l 1

j IV.

Root Cause l

g The root cause of TVA's problem with Y-type Kerotest root valves is I

that Kerotest did not routinely change out the back-up packing after l

i hydrotesting and TVA had incor.sistent purchase specification for l

ensuring the that the valves were dried out properly upon receipt.

I 1

I V.

Generic Applicability l

2 l

1 Kerotest Y-type valves were used at Watts Bar, Sequoyah, and l

Bellefonte Nuclear Plants. The problem of the packing not being l

removed after hydrotesting was found to exist at Watts Bar but not l

Sequoyah. Bellefonte and Browns Ferry have not as yet been evaluated l

for potential Kerotest valve problems.

l 1

i j

l Page 8 of 9

)

o.'

O Revision 1 VI.

References 1.

MI-15.3, Revision 0. December 1984, " Maintenance, Inspection and Repair of Kerotest Manual "Y" Globe Valves" 2.

NCR 2501 R, Revision 1, dated October 22, 1980 (801 027B0 425) 3.

NSRS Investigation Report I-85-861-SQN, Revision 1, April 9, 1986, " Generic Implications of Watts Bar Kerotest Valve Problems on Sequoyah" 4.

Memorandum from John A. Raulston, to L. M. Mills, "WBN Units 1 and 2, Deficient Kerotest Y-Type Valves - Report No. 4 (final) -

NCR-2501 R, dated April 27, 1981 (810 43060 153) 5.

Letter from L. M. Mills, to James P. O'Reilly, NRC, " Watts Bar Nuclear Plant, Units 1 and 2 - Deficient Kerotest Y-Type Globe Valves - NCR 250lR - Final Report," dated April 24, 1981 (A27 810424 014) 6.

Memorandum from R. W. Cantrell, to J. A. Raulston, "Sequoyah Nuclear Plant Units 1 and 2 - Deficient Kerotest Y-Type Globe Valver - Report No. 4 (final)," dated September 10, 1981 (810916 F0144) 7.

10CFR21, " Reporting of Defects and Noncompliance Dated January 1, 1985 8.

Memorandum from H. L. Abercrombie, to R. P. Denise, dated May 19, 1986 - Subject - Response To the Nuclear Safety Review Staff (NSRS)

Investigation Report No. I-85-861-SQN, R1 " Generic Implications of Watts Bar Kerotest Valve Problems on SQN" VII.

Immediate or Long Range Corrective Action A review should be made on the current specifications regarding l

Kerotest valves to ensure valves are dried upon completion of lR1 hydrostatic testing and valve packing is removed.

l Page 9 of 9-

-.