ML20206U937
| ML20206U937 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 09/25/1986 |
| From: | Nixon D TENNESSEE VALLEY AUTHORITY |
| To: | |
| Shared Package | |
| ML20206U867 | List: |
| References | |
| C011306-SQN, C11306-SQN, NUDOCS 8610080457 | |
| Download: ML20206U937 (11) | |
Text
TVA EMPLOYEE CONCERNS
~
REPORT NUMBER: C011306-SQN SPECIAL PROGRAM REPORT TYPE: Sequoyah Nuclear Plant Element REVISION NUMBER:
1 PAGE 1 OF 9 TITLE: Testing of Anchors REASON FOR REVISION: To incorporate TAS and SRP comments.
PREPARATION PREPARED BY:
f*S.5~hh Albv/
r
/
SIGNATUlfEs DATE
/
~/
REVIEWS PEER:
ph g f Of'2f'AS SIGNATURE DATE TAS:
SIGNATURE DATE CONCURRENCES
/If 7"~7 7 ~8d CEG-H
/
f/ //
~
SIGNATURE DATE SIGNATURE DATE APPROVED BY:
1 ECSP MANAGER DATE MANAGER OF NUCLEAR POWER DATE CONCURRENCE (FINAL REPORT ONLY) 1313T 8610090457 860925 PDR ADOCK 05000327 P
PDR u
~
REPORT NUMBER: C011306-SQN TVA EMPLOYEE CONCERNS SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 2 0F 9 I.
Introduction The concerns in this element, IN-85-285-002, IN-86-115-001, IN-86-190-003, and XX-85-023-001, address the testing of anchors.
II.
Summary of Perceived Problems A.
Shell-Self-Drilling (SSD) type anchors were improperly tested or the pull test was bypassed.
B.
SSD type anchors are tested by sampling methods instead of on an individual basis.
C.
SSD type anchors are being overtorqued.
III.
Evaluation Methodology A.
Reviewed the Sequoyah Nuclear Plant (SQN) Generic Concerns Task Force (GCTF) Report on Incorrect Installation and Inspection of Anchors for overall content, adequacy, and findings with respect to the subject concerns.
B.
Reviewed the Watts Bar Nuclear Plant (WBN) Employee Concerns Task Group (ECTG) Element Report on Testing of Anchors (C011306) for content and generic applicability to SQN.
C.
Reviewed Nuclear Safety Review Staff (NSRS) Investigation Report I-85-439-WBN to determine the methodology used in addressing the sampling program issue, the findings and the adequacy of the report with respect to the answering of the subject concern.
D.
Reviewed the Division of Nuclear Engineering's (DNE) response to NSRS Investigation Report i-95-657-WBN with respect to anchor lot designations / definitions and justification for sampling methodology.
E.
Reviewed SQN Modifications and Additions Instruction (M&AI) 10 (Testing of Expansion Anchors Set in Hardened Concrete), M&AI 11 (Fabrication, Installation and Documentation of Seismic Supports and Supports Attached to Seismic Category I Structures), and Gonoral Construction Specification C-32 to determine requirements for anchor testing and identify any criteria for designation of anchor lots.
Also, reviewed other applicable procedures and documentation as required to obtain information relevant to the subject concerns.
TVA EMPLOYEE CONCERNS REPORT NUMBER: C011306-SQN SpECIAL PROGRAM REVISION NUMBER:
1 PAGE 3 OF 9 F.
Interviewed knowledgeable personnel in responsible units to obtain additional input applicable to the subject of this element.
IV.
Summary of Findings A.
The SQN-GCTF Report addressed the subject of this element as well as the parameters of the anchor installation process. With respect to anchor testing, the report referenced sampling programs done to satisfy NRC IE Bulletin 79-02 as well as SSD reinspection and replacement exercises initiated as a result of employee concern XX-85-023-001.
Although not specifically referenced in the GCTF report, Nonconformance Report (NCR) 720 and Nuclear Regulatory Commission (NRC) inspection item 328/78-01-14, along with Employee Concern XX-85-023-001, were the basis for the initiation of a Special Maintenance Instruction (SMI-2-317-24R2) to review anchor bolt / baseplate installations for the SQN unit 2 shield building wall.
The report concluded that existing historical information from pull test data, 79-02 inspections, and the SMI-2-317-24R2 anchor survey, produced results that supported the adequacy of SQN's anchor installation / inspection program. The status of the open issues addressed by the SQN report will be addressed in section G.1. of this report.
B.
The WBN-ECTG Element Report for Testing of Anchors did not identify any specific issues generic to SQN.
Specific instances of conditions adverse to quality (CAQ) at WBN were documented and corrected according to site NCRs. The report identified no CAQ with respect to the use of sampling programs for proof loading of SSDs.
C.
A detailed review of the NSRS Investigation Report I-85-439-WBN was made to evaluate the issue of using sampling techniques to test SSDs.
It was discovered that all aspects of the NSRS report were applicable to SQN, except where specific W8N procedures were referenced. The conclusion of the report indicated that
" determination of adequacy of the anchors based on sampling is an acceptable technique endorsed by industry standards, TVA procedures and the NRC in IE Bulletin 79-02."
This conclusion was based on a j
detailed review of American National Standards Institute (ANSI) and American Society of Testing and Materials (AGTM) Gtandards, TVA Design Standards, and TVA General Construction Specifications.
D.
A detailed review of the DNE response to recommendation 03 of NSRS Investigation Report I-85-657-WBN was made to evaluate the applicability of defined anchor installation lots at SQN. TVA General Construction Specification G-32 provides the procedural criteria for all TVA nuclear plants for the number of anchors
TVA EMPLOYEE CONCERNS REPORT NUMBER: C011306-SQN SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 4 0F 9 IV.
Summary of Findings (cont'd) selected and the testing frequency required to prove acceptable anchor installations. However, G-32 was not intended to provide evidence of anchor acceptability based on individual lot test results. The results of proof load tests are evaluated monthly by DNE, where they are categorized by anchor type and size, not by lot.
The defect rate is then determined for each group of anchors, and the results provide the means for evaluating whether the proof load rate is acceptable.
A failure rate of greater than 5 percent for any group of anchors requires additional action (s) and a failure rate close to 5 percent requires further evaluation to determine if trends exist. This methodology recognized the high rate of proof test failures in the SQN unit 2 annulus area and eventually led to identifying the understrength surface concrete problem addressed by NCR 72D.
It should also be noted that the statistical sampling plan employed was a recommended method in NRC IE Bulletin 79-02.
E.
A review was made of TVA General Construction Specification G-32, RIO and M&AI 10, R10 with the following results:
1.
The criteria which addresses the designation of anchor lots in M&AI 10 is taken directly from G-32 and provides adequate definition and description of anchor installation lots.
This review also determined that no specific criteria existed to require proof testing (pull testing) of SSDs before baseplate installation (although G-32 implied this method) nor was there specific criteria for shimming baseplates when "through the plate" testing was performed.
See section F.4 of this report for additional information on this subject.
F.
The responsible engineer in SQN/ Modifications Unit was interviewed to obtain information on mechanisms employed to address NSRS recommendation Q-85-023-001-01.
The results of this interview are as follows:
1.
NCR-72D was originally issued by DNE to document continual failure of anchor pull tests in the SQN Unit 2 annulus area.
The corrective action for this NCR was to replace all 1/2" diameter and larger self-drilling (SSD) anchors in the affected area with equivalent size wedge bolts. Excluded from this replacement were 1/4" and 3/8" diameter SSDs installed in specific conduit supports. The QTC investigation of concern XX-85-023-001 revealed these corrective actions had not been fully implemented. In response to NSRS recommendation Q-85-023-001-01, initiated as a result of the aforementioned QTC investigation, SQN agreed that the corrective action of NCR-720 had not been fully implemented as detailed in a letter from H. L. Abercrombie to R. K. Siberling dated April 24, 1986.
TVA EMPLOYEE CONCERNS kEPORT NUMBER: C011306-SQN SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 5 0F 9 IV.
Summary of Findings (cont'd) 2.
Further discussion revealed that the SQN Modifications Unit had initiated SMI-2-317-24 on January 15, 1986 to conduct a field survey of anchor bolt / baseplate installations in the unit 2 shield building wall. This SMI included a sampling program for all anchorages on the vertical interior and exterior wall at all elevations using the following selection process:
a.
System 67 (ERCW) in the annulus only b.
Systems 30 and 65 (HVAC ductwork) in the annulus only c.
System 26 (fire suppression) in the annulus only d.
Electrical systems conduit 21/2 inch diameter and greater (including some junction boxes and cable trays on the interior / exterior wall) e.
Unistrut attached to wall surfaces that supports pipe or tubing on the interior / exterior wall f.
Cantilever type supports for tubing and conduit two-inch diameter and smaller g.
The thirty supports previously identified in the QTC investigation ' report on the subject concern (XX-85-023-001).
i The results of this exercise are as indicated in the letter from C. R. Brimer to H. L. Abercrombie concerning the SQN-GCTF recommendations, and reflect an overall acceptance rate of greater than 95 percent. The actual number of supports inspected was 1281 (1193 was referenced in the aforementioned letter). The total number found unacceptable was 43, all of which have been or will be corrected by specific Maintenance Requests (MRs).
Initially, 1/4 inch and 3/8 inch diameter SSDs had been excluded from the inspection under certain conditions (MEMO SWp 781013 005). However, since that time, conduit loads have been increased so the referenced memo is no longer applicable in almost every case, except for aluminum conduit.
Workplan 11963 was written to address this issue and SQN Modifications has initiated a test program to qualify the existing anchorages to greater loads. If this program is unsuccessful, further action will be required. In addition, the inspection revealed no SSD anchors to be installed in any
)
of the mechanical systems in the inspected areas - all had been replaced as required by the SQN Final Report for NCR 720.
TVA EMPLOYEE CONCERNS REPORT NUMBER: C011306-SQN a
SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 6 0F 9 IV.
Summary of Findings (cont'd) 3.
The responsible engineer in SQN Project Services Section (Site Services) who was also involved in the unit 2 annulus inspection was interviewed. He provided the following information:
a.
The strategy behind choosing the System 67, 30, 65 and 26 supports for inspection was that these supports encompassed a very large percentage of the installations in the annulus area.
b.
M&AI 10 RIO (Testing of Expansion Anchors Set in Hardened Concrete), section 3.3 states, "no 1/2 inch diameter or larger SSD type anchors are to be installed in the SQN unit 2 shield building wall."
M&AI 11 R12 also contains this statement as well detailing specific wedge bolt anchors for installation in the affected area. These statements provide procedural means of assuring future anchor installations will not violate the criteria of the NCR 72 D Final Report.
4.
The implications of concern IN-85-285-002 were discussed with the responsible DNE-Civil Engineering Branch Central Staff engineer with respect to SQN.
He stated that the inspection criteria of 79-02 was sufficient to prevent the anchor shell from contacting the baseplate during "through the plate" proof testing. He also stated that when anchor reinspection required proof testing as a result of a NCR, the reinspection process would require plate shimming if "through the plate" proof i
loading was performed. However, neither G-32 nor M&AI 10 specifically state that proof testing is to be done before I
baseplate installation, nor is there criteria included which requires shimming of the baseplate when through the plate proof testing is performed.
G-32 only implies that the proof test will be done before base plate installation.
It was stated, however, that proof testing before base plate installation was standard practice at all plants and shimming of the plate for "through the plate' testing was also a standard, recognized practice.
At this point, the positive results of SQN's 79-02 inspection program were discussed as well as the 79-02 inspoction procedures for both units which required verification that the anchor shell was not contacting the base plate. The final comments were that a revision to G-32 could be made to include specific statements addressing proof test before base plate installation and shimming when through the plate testing was performed but he did not feel as though a revision was a necessity.
I 1
L
i TVA EMPLOYEE CONCERNS
~
REPORT NUMBER: C011306-SQN SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 7 0F 9 IV.
Summary of Findings (cont'd) 5.
G-32 and M&AI 10 were further discussed with the SQN Modifications responsible engineer. He agreed with the comments made in section F.3 above by the DNE-CEB engineer that pull testing before base plate installation was a standard, recognized practice even though not a specific requirement. He also stated that during reinspections the base plate was removed in many cases instead of using shims which would allow "through the plate" proof testing.
G.
Conclusions 1.
This evaluation is in agreement with, and verified the SQN-GCTF report on the subject of this element.
SQN's 79-02 reinspection program, the reinspection program initiated according to SMI-2-317-24, and the reinspections performed as a result of the QTC investigation of concern XX-85-023-001 served as a mechanism to further prove the overall adequacy of SQN's concrete anchor installations. The recommendations in the SQN-GCTF report on Incorrect Installation and Inspection of Anchors address 2 other issues:
Base plate flexibility - this issue is addressed fully in a.
the WBN-ECTG report for SQN on Design of Plates, b.
Overtorquing - before outlining the details that relate to this issue, it should be pointed out that overtorquing is not the proper terminology to be used. The concerned individual used the word overtorquing in describing the concern on expansion anchors. However, since expansion shell anchor bolts are " tightened" with a wrench of some type, and not " torqued", the appropriato terminology is "overtightened." The terminology overtorquing and overtightening are not interchangeable in this application.
This issue was initially raised at WBN by employee concern IN-86-115-001.
NSRS Investigation Report I-85-659-WBN substantiated the concern and recommended a program to verify anchor integrity based on bolt tightening practices. The WON-ECTG Element Report on Installation of Anchors further verified the NSRS report and indicated that PIR-WBNCEB8644 had been initiated to evaluate the deficiency.
TVA EMPLOYEE CONCERNS REPORT NUMBER: C011306-SQN SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 8 0F 9 IV.
Summary of Findings (cont'd)
Interviews with responsible personnel in SQN Site Services, Modifications, Compliance, Site DNE and CEB Central Staff revealed no evidence to indicate the anchor overtightening issue had ever been addressed by SQN.
It is conceiveable that sufficient data exists from G-32, 79-02 and other reinspection programs to adequately address this issue.
However, SQN should initiate an exercise to evaluate possible anchor overtightening using methodology that will sufficiently answer this potential generic problem.
2.
This evaluation did not attempt to verify whether pull tests were bypassed, as stated in concern XX-85-023-001, because SMI-2-317-24 R2 and Workplan 11693 are sufficient mechanism to evaluate all anchorages installed in the unit 2 annulus area where understrength surface concrete was identified as a result of NCR-72D.
In addition, specific procedural mechanisms have been implemented to control future installations in this area.
The SQN Modifications Unit is performing the aforementioned evaluations.
3.
This evaluation is in agreement with and verified the WBN-ECTG Element Report findings as well as the NSRS investigation report I-85-439-WBN findings with respect to the use of sampling programs being acceptable.
Agreement with the DNE response to NSRS Investigation Report I-85-657-WBN-03 was also determined according to this evaluation. The sampling program issue has been adequately addressed in all cases and no deficient areas were identified.
G-32 is adequate with respect to lot definition and testing frequency of concrete anchors.
4.
This evaluation addressed specifically the issue of anchor shells contacting base plates during pull tests (concern IN-85-285-002) and determined:
No specific criteria in G-32 or M&AI 10 address this issue a.
but 79-02 reinspection programs were governed by mechanisms to verify no contact existed both before and upon completion of the proof test.
Since that time sufficient sample, review and reinspection programs have been performed to reveal generic failures with respect to this issue. None have been identified. Therefore, this evaluation does not recommend additional reinspections for this condition.
t TV.1 EMPLOYEE CONCERNS REPORT NUMBER: C011306-SQN SPECIAL PROGRAM REVISION NUMBER:
1 PAGE 9 0F 9 IV.
Summary of Findings (cont'd) b.
This evaluation did identify that DNE-CEB should review existing G-32 criteria to determine if specific requirements should be implemented which would:
(1) require pull tests to be performed before base plate installation (2) require shimming of the base plate when reinspections cause through the plate proof tests to be performed.
SQN Modifications unit should review M&AI 10 to determine c.
if specific requirements should be implemented to address recommendations 1 and 2 above.
V.
Root Cause This evaluation did not identify specific conditions adverse to quality, therefore, no root cause is assigned.
c l
VI.
Corrective Actions Further review to be performed by DNE-CEB and the responsible line organization as required to determine applicable corrective action (s).
VII.
Generic Applicability The WBN-ECTG Element Report (C011306) on the subject of this element did not identify generic applicability.
However, this evaluation concludes that generic implications exist for all TVA nuclear plants on the following issues:
A.
adequacy of inspections and criteria with respect to anchor shells contacting base plates during proof tests 8.
potential overtightening of SSD type anchors e
.1>
CONSTRUCTION CATEGORY Enclesura 2
~.
.=
ECTG WRITER'S GUIDE ELEMENT REPORT FORMAT 1.0 Issue I.
Introduction Characterization II.
Sumary of Perceived Problem 2.0 Sumnry IV.
Summary of Findings Conclusions (Portion) 3.0 Evaluators Cover Sheet Information 4.0 Evaluation III.
Evaluation Process Methodology
5.0 Findings
IV.
Summary of Findings l
6.0 Root Cause V.
Root Cause (Collective Significance)**
- Overall Root Cause will
.A not be considered at this report level.
7.0 Attachments /
VIII. Attachments List of Concerns i
i i
Other items included VI.
Corrective Actions in Construction Element (To be provided by Reports line orgainzations)
VII.
Generic Applicability 1905T
Page 1 of 1 SEQUOYAH EMPLOYEE CONSTRUCTION CATGORY CONCERN ELEMENT REPORT NUMBER ISSUE Construction 11103 IN-85-288-001 Methods Used During i
IN-86-116-001 Installation XX-85-070-007 l
Construction 11306 IN-85-285-002 Testing of Anchors IN-86-115-001*
IN-86-190-003 XX-85-023-001
- Data base change to add this concern to SQN list has been submitted.
(
1960T f
.