ML20236G401
| ML20236G401 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 10/27/1987 |
| From: | Gridley R TENNESSEE VALLEY AUTHORITY |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| NUDOCS 8711020523 | |
| Download: ML20236G401 (12) | |
Text
,
T:. '
TENNESSEE VALLEY' AUTHORITY CH ATTANOOGA, TENNESSEE 374of SN 157B Lookout Place:
.08T271987 E
U.S. - Nuclear Regulatory Comission-
- j-ATTN
- .;' Document Control' Desk' Washington,.D.C.
20555 s
U' l Gentlemen:
lIn-the Matter'of.
.)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 n SEQUOYAH NUCLEAR PLANT: (SQN) -- RESPONSE TO FINDINGS. IDENTIFIED DURING THE
' FINAL NRC INSPECTION OF THE. DESIGN BASELINE AND VERIFICATION PROGRAM (DBVP)
On July 24, 1987, an NRC-inspection _ team from the Office of Nuclear Reactor
' Regulation (NRR) -'Special Inspection Branch concluded a,two-week' inspection of the SQN DBVP.
The team assessed the Engineering' Assurance (EA) oversight of'the DBVP'and examined corrective' action for EA and NRC' observations /open it' ems. -In addition,-the team reviewed resolution of.punchlistl items;-the EA.
report (EA-OR-001) submitted to NRC as an enclosure'to a letter dated May 15,.
-1987; and the Unit'2 Phase I (prerestart) DBVP Report.
In' advance of receipt offNRC. Report Nos. 50-327/87-31 and 50-328/87-31, which' idocuments. NRC's findings during. the subject ' inspection, enclosed is TVA's g
. response to the! observations. identified during the exit meeting. Upon
-issuance'of the inspection report, this response will'be supplemented or 1
modified as required.
Very truly yours, TENNESSEE VALLEY AUTHORITY' a
R. Gridley, rector i
Nuclear Licensing and m-Regulatory Affairs 1
, Enclosures l
4
- cc:.-see page 2
.j
%[
P
.l
'O h
t3 f
An Equal Opportunity Employer 1
. U.S. Nuclear Regulatory Commission 00'I271987" cc (Enclosures):
Mr. G. G. Zech, Assistant Director for Inspection Programs Office of Special Projects U.S. Nuclear Regulatory Commission 101 Marietta Street, NW, Suite 2900 Atlanta, Georgia 30323 Mr. J. A. Zwolinski, Assistant Director for Projects Division of TVA Projects Office of Special Projects U.S. Nuclear Regulatory Commission 4350 East-West Highway EWW 322 Bethesda, Maryland 20814 i
Sequoyah Resident Inspector Sequoyah Nuclear Plant 2600 Igou Ferry Road l
Soddy Daisy, Tennessee 37379 l
l
^'
ENCLOSURE 1 Following are the TVA descriptions of the observations identified by.the NRC inspection team on July 24, 1987, at the conclusion of the final inspection of the Sequoyah Nuclear Plant (SQN) Design Baseline and Verification Program (DBVP).
Each observation description is followed by a TVA response.
It should be noted that these descriptions of the observations and subsequently, the responses, may change when NRC Report Nos. 50-327/87-31 and 50-328/87-31 is issued.
Most of the corrective actions identified in the TVA responses are associated with Significant Condition Reports (SCRs), Problem Identification Reports (PIRs), or Condition Adverse to Quality Reports (CAQRs).
These will be
(
dispositioned in accordance with TVA's Condition Adverse to Quality (CAQ)
]
process.
e Observation 3.16 - Potential Generic Condition Evaluation (PGCE) l No PGCE was requested for PIR SQNMEB86127, revision 0, and the team disagrees j
with this decision; however, an informal memorandum from a DBVP engineer to j
the Civil Engineering Branch (CEB) was sent May 27, 1987, requesting that a I
PGCE be performed.
TVA Response 1
PIR SQNMEB86127, revision 1, was issued on July 24, 1987, requiring a PGCE to be performed.
The PGCE request has been sent to TVA's various nuclear plants, and the final evaluations are still ongoing.
Observation 3.17 - Calculations for Solenoid Valve Support Variances The team reviewed workplan 9503, which installed solenoid valves in accordance with Engineering Change Notice (ECN) 5457. There are at least two examples of variances without supporting calculations (variances 54-33-A43 and 54-33-A45).
Sketches are provided without appropriate calculations.
l TVA Response No calculations could be found for these variances.
Calculations were either not easily retrievable the variances were approved without issued calculations.
To corrase this, a calculation package will be generated that I
will qualify instrument line support variances 54-33-A43 and 54-33-A45 to current or unit 2 restart criteria.
This calculation package will be generated before restart of unit 2.
It should be noted that instrument lines as now installed represent no safety problems to the plant.
The following is from section 3.3 of enclosure C of a letter transmitted to NRC on April 8,1987, concerning the Alternate Analysis (AA) Program Phase II (postrostart).
Various concerns have been identified regarding documentation for the installation and inspection of instrument lines and supports.
These concerns do not represent a safety problem and will be addressed after restart as part of the docketed AA Phase II program.
. Section 4.0 of enclosure C referenced above commits to the following postrestart activities to ensure that all instrument lines and instrument line supports are in compliance with design basis requirements. Walkdowns will be performed for all safety-related instrument lines and supports.
These walkdowns will develop the "as-constructed" configuration for instrument lines and supports that do not pass screening criteria developed for the walkdowns.
Using the sketches generated by the walkdowns, evaluations will be performed for instrument lines and supports that do not pass the walkdown screening criteria. Modifications will be performed as necessary to ensure compliance with design basis requirements.
In addition, a sampling program will be conducted to establish confidence that variances to standard support designs have not resulted in unacceptable support configurations.
With regard to action to prevent recurrence, section 5.0 of enclosure C referenced above states that existing procedures will be revised and new procedures will be developed to address all instrument and support issues and prevent recurrence of similar problems in future installations.
Procedure revisions and now procedures will include:
(a) Use of variances on instrument line supports will be eliminated.
All future chang 0s will be made using issued drawings.
(b) An " Engineering Requirements Specification" for instrument line installation and inspection will be issued for SQN.
Existing installation and inspection procedures will be revised, and new procedures will be issued to implement this specification.
(c) The instrumentation drawings and typical support drawings will be revised to upgrade and clarify design requirements.
By initiating the above actions, variances 54-33-A43 and 54-33-A45 and all other instrument line supports will be adequate for restart of unit 2.
Phase II of the AA Program (including modifications) will be completed by the end of the cycle 4 refueling outage for each unit.
Observation 4.8 - Radiation Monitoring System Punchlist item 4426, written on System 90 (Radiation Monitoring System), is associated with SCR SQNNEB8615. The SCR was initiated by the DBVP because there are components for reactor coolant pressure boundary leakage detection that are supplied with nonessential air.
The restart category for the punchlist item is postrestart, justified by the TVA SQN Safety Analysis Group under Quality Information Release (QIR) NEB 86241. The conclusion of this QIR was that the safety function of this equipment would be performed upon the loss of the nonessential air, and thus the postrestart category is allowable.
The team was given copies of the punchlist item's Sequoyah Engineering Procedure (SQEP)-45, attachments 1 and 2 documents, as well as the QIR.
The team's concern is that NRC Regulatory Guide (RG) 1.45 may be violated by the SCR.
b
. TVA Response
. TVA's response-to this observation is captured in the Engineering Report (ER),
revision 1, prepared for SCR SQUNEB8615. The most salient points of the ER
.are highlighted below:
1.
There is only one safety concern'that is raised in this issue; q
namely, the ability to isolate potential containment leak paths as a consequence of a Safe Shutdown Earthquake (SSE) event.
The sample line isolation valves, FCV-90-107_ thru -111 and FCV-90-113 thru -117, for radiation monitors RE-90-106 and RE-90-112, serve as containment-isolation valves. These valves are presently served by nonessential control air, which is not qualified to survive a seismic event.
In the event of loss of control air, these sample line isolation valves will fail closed and thereby satisfy their safety function to isolate potential containment leak paths.
Monitors RE-90-106 and RE-90-112 will cease to be functional in this event.
2.
Monitors RE-90-106 and RE-90-112 do not provide any primary safety function.. Their primary service is to satisfy the RG 1.45 requirement for detecting and monitoring radioactive airborne particulate activity as a consequence of a leak from the primary coolant pressure boundary.
While satisfaction of RG 1.45 recommendations with these monitors cannot, at present, be ensured during or after a seismic event (even something less severe than an SSE event), there is no safety concern issue.
3.
'SQN's compliance with the recommendations of RG 1.45 is stated in Final Safety Analysis Report (FSAR) Section 5.2.7, as.follows:
The leakage detection systems comply with applicable parts of NRC
(
General Design Criterion 30 and Regulatory Guide 1.45.
These systems provide a means of detection, to the extent practical, leakage from the reactor coolant pressure boundary.
In NRC's Safety Evaluation Report for SQN (NUREG-0011, issued in March 1979), the staff recognized that the airborne particulate radioactivity monitoring system had not been specifically designed ~to remain functional when subjected to an SSE occurrence.
It was concluded, however, that the degree of SQN compliance to RG 1.45 constitutes an acceptable basis for satisfying the requirements of General Design Criterion 30.
TVA therefore considers that no further action is required for SQN to demonstrate compliance with applicable l
parts of General Design Criterion 30 and RG 1.45.
The corrective action stated in part B (item 14a) of SCR SQNNEB8615 is to provide the sample line isolation valves with control air service that is fully qualified to Seismic category I requirements, to revise SQN-DC-V-32.0, and to update' appropriate drawings.
This SCR, however, has not been fully dispositioned; and the corrective action determinations may change before the SCR is closed.
. Observation 6.21'- Post Accident Monitoring (PAM) System-
~
l SCR SQWNEB8722 identified a condition that the present PAM instrumentation for channel 1 does not comply with Design Input Memorandum (DIM) SQN-DC-V-19.0-1 for electrical separation or with SQN FSAR section 7.5.2.
The SCR corrective action specified adding isolation devices for isolation of nonqualified circuits or performing an evaluation to determine the feasibility i
'of disconnecting the nonqualified circuits from the PAM channels. However, a revision to this SCR is presently in process that changes the corrective action to a postrestart category and revises the PAM design criteria and FSAR (although not the electrical separation criteria) to delete the requirement that one PAM channel be. fully separated from nonqualified circuits. The NRC team notes that this SCR revision has not been approved and issued, but does not agree with such a revision and considers the original corrective action to be appropriate.
The team is concerned that there may not be adequate controls of changes to punchlist item corrective action after a DBVP system engineer has concurred with the restart category and corrective action recommendation documented on the SQEP-45 attachment 2 form.
TVA Response As part of the Electrical Engineering Branch (EEB) disposition'of SCR SQNNEB8722, an ER was prepared that reviewed all PAM 1 and 2 channels required for Class IV events, in accordance with FSAR Table 7.5.1-2.
The following summary presents the findings and conclusions from the ER and related. item disposition (SQEP-45, attachment 2 form).
. Findings 1.
PAM 1 channels do interface with nondivisional loops at the computer racks. This is consistent with other indication loops and the original design philosophy of the plant for pulling computer points off of instrument loops.
2.
PAM 2 channels follow the same scheme as the PAM 1 channels but are physically separate and tie to different computer terminal strips than the PAM 1 channels.
3.
Both the PAM 1 and PAM 2 channels are isolated with qualified current-to-current isolators in the Reactor Protection System racks.
Investigation revealed no compromise of the divisional to nondivisional interfaces for these subject loops.
4 PAM 1 and PAM 2 are physically separate from sensor to display.
NOTE:
Indicators are not currently seismically qualified as specified in the DIM-SQN-DC-V-19.0-1 exception.
. i CONCLUSIONS Accept as is.
Existing PAM loops exhibit separation and a level of availability consistent with their respective displays and the original design banis. No qualified to nonqualified isolation problems are evident.
NEAR-TERM CORRECTIVE ACTIONS (Postrestart)
Clarify physical and electrical separation requirements in DIM-SQN-DC-V-19.9-1 and section 7.5 of the SQN FSAR.
LONG-TERM CORRECTIVE ACTIONS (Postrestart)
Upgrade PAM loops consistent with the NRC-accepted commitment in TVA's RG 1.97, revision 2 plan (fuel cycle 4 outage for unit 2), as stated in an NRC letter to TVA dated February 12, 1987.
The DBVP issued Program Directive No. DBVP-D-87-008 on August 5, 1987, to address the above concerns regarding changes to the punchlist. This directive requires that punchlist changes be categorized as one of the following:
(A) administrative changes, (B) implementation status changes, and (C) technical changes.
Technical changes are those changes to punchlist corrective actions and/or restart categories other than those changes resulting from the normal progression of implementation action. Technical changes are required by this Directive to be approved by the system engineer (SE) and by the Discipline Evaluation Supervisor (DES) in full accordance with SQEP-45, including documentation on attachment 2 forms.
Directive DBVP-D-87-008 was submitted to NRC as an enclosure to a letter dated
-August 20, 1987.
In addition, as a requirement of Program Directive DBVP-D-87-002, each SE prepared a final system closcout statement. verifying that all system punchlist items (both open and closed) have his concurrence. He also discussed in this closeout statement any restart categorization differences from the System Evaluation Report (SYSTER) statement.
System Closure Statements have been completed and issued for all the DBVP Phase I systems; however, it should be noted that they have been and are being revised as current information dictates.
Observation 6.22 - Auxiliary control Air (ACA) System An earlier observation from NRC Inspection and Enforcement (IE) Report Numbers 50-327/86-45 and 50-328/86-45 (Observation 6.6 - Auxiliary Control Air System Design Criteria) had been closed, in which NRC noted that a commitment had been made by a TVA response to NUREC 0737 that physical separation of air headers inside containment to prevent adverse interaction was required.
Based upon TVA's confirmation that the design was satisfactory, the item was closed in IE Report Numbers 50-327/86-55 and 50-328/86-55.
Since that time, SCR SQNMEB86121 has reopened the issue of inadequate separation of ACA headers.
The SCR initially identified 24 interactions between ACA headers and process l
piping (critical cracks). CEB calculation CEB-PR/NZ-065, revision 1, evaluated these interactions. Separation on five of the interactions was l
l l
l l
\\
L l
o_ - _ --- -
u
' determined to be acceptable based upon a closer look at the separation-criteria (CEB Report CEB 79-12).
One interaction was a small break Loss of Coolant Accident (LOCA) (critical crack) near the pressurizer spray valve.
'This interaction has been resolved by installing an isolation plug on the ACA header upstream of the potential interaction under ECN 7102. The ACA header was not serving equipment downstream of the interaction.
The 18 remaining interactions were determined to be non-LOCA breaks (critical cracks). Mechanical. Engineering Branch (MEB) calculation SQN-32-D053
)
(" worst-case" ACA system pressure transient) determined that ACA pressure of 70 psig emuld bs rostored within five minutes following an ACA header break inside containment because of containment isolation valvo closure at 50 psig.
Nuclear Engineering Branch (NEB) QIR SQP87386 evaluated the loss of ACA and determined that~ACA pressure is not required for the first five minutes following a non-LOCA break. The team does not agree with the results of the NEB QIR. One specific concern involved the suspect qualification and j
instrument accuracy (CAQR SQP8713) for the pressure indicating controllers
]
(PIC) on the ACA containment isolation valves (CIV).
Failure of the PICS i
could prevent isolation of ACA following an ACA pipe break, thereby increasing the amount to time before ACA is restored. Another concern involved the heating, ventilating, and air conditioning (HVAC) logic for i
switchover/ rollover capability (nonsymmetrical logic) following loss of ACA pressure (i.e., if ACA pressure to Train A is Icst because of interaction and Train B is lost to assumed single failure, will HVAC logic provide for switching back to restored Train A after 5 minutes.) The 480-V board rooms and 6.9-kV shutdown board room HVAC were cited as examples where the switchover feature may bo inadequate.
j i
~The NRC inspection team also identified a new concern involving potential interactions between ACA tubing and process piping inside and outside containment.
A specific concern was identified for.the tubing supplying air j
to the pressurizer spray valves. A subsequent walkdown determined that a a
critical crack of Reactor Coolant System piping (small break LOCA) could interact with ACA tubing.
It was noted that an evaluation was being performed i
by NEB to address the loss of ACA because of a small break LOCA.
The Division of Nuclear Engineering (DNE) agreed to perform a walkdown inside and outside l
containment and to evaluate any additional interactions discovered under SCR SQNMEB86121.
i IyA Response Revision 2 to SCR SQNMEB86121 addresses the concerns identified by NRC as part l
of this observation.
Following is a brief description of each concern identified by NRC followed by the document that addresses this concern.
1.
NRC questioned taking credit for the PICS for the ACA containment isolation valves in MEB calculation SQN-32-D053 (worst-case ACA pressure transient).
Documents:
1.
EEB QIR SQP87469 2.
MEB Calculation SQN-32-D053, revision 2 (Page 1-1, 1-2, and 1-3 of 83)
s l
j L
_7_
1 L2. - NRC questioned liVAC logic for switchover/ rollover capability
.(nonsymmetrical logic).
Document:
QIR NEB 87229 R3 (Page 2 of 3) l 3.
Generally, it is NRC's position that ACA pressure and its safety-related air users would be required immediately following a small break LOCA or small non-LOCA break (critical cracks) inside containment.
[
Documents:
1.
NEB QIR.SQP87427 L
2.
QIR NEB 87229, revision 3 I
4.
NRC identified a new concern involving potential interactions.between ACA tubing and process piping inside and outside containment. A walkdown was performed, and additional interactions were identified and addressed.
Document:
CEB Calculation PR/NZ-065, revision 2, Part B - Pages 22-46.
CONCLUSION
.}
The CEB calculation referenced above addressed all of the interactions for the
.ACA headers and tubing. The corrective action response to SCR SQNMEB86121, revision 2, includes the CEB calculation as'part of the SCR response.
ECN 7102 installed an isolation plug to eliminate ACA header interactions with the i
pressurizer spray valves. The two NEB QIRs referenced above document the fact that ACA pressure is not required for 10 minutes following a small break LOCA or'amall break non-LOCA inside containment. A walkdown was performed inside
)
and outside containment, and additional interactions were addressed in the CEB l
calculation. The CEB calculation evaluated these interactions and eliminated all but three interactions through critical crack exclusion.
ECN 7272 has I
been issued to provido pipe rupture shielders on the Auxiliary Feedwater System to' eliminate the remaining ACA tubing interactions outside containment.
l Observation 7.5 _ Punchlist Accuracy A biased sample of five punchlist items associated with implemented modifications was reviewed, and all were found to be unsatisfactory as follows:
Punchlist Number Comment l
0386 Related to a CAQR that was actually outside DBVP scope but reported as implemented.
9111, 8514 Reported the item as implemented when actually it was not.
9670, 7442 No attachment 2 documents could be located to show corrective action status.
In addition, the DBVP should address the following issues identified as.a result-of inconsistencies found in the punchlist:
(1)
Use of the punchlist when there is a time lag between input of data from the SQEP-45 attachment 2
)
form and reflection of it in the punchlist; (2) incorrect input of data; and (3) lack of verification of the data entered into the punchlist.
N
. l TVA' Response-
. Punchlist[itemnumber0386-Thecontrollingprocedur'efortheDBVPPunchlist,
- SQEP-45 jallows'only'three categorizations for punchlist action items:
Open" --Initial identification of action;1 tem.
Closed" - Cderective action has been defined and documented, restart categorization.has been made, concurrence from the SE and the DES has been obtained, and.ccreective. action has been scheduled in PROJECT /2.
Implementation Status " Complete" - Corrective action has beenLimplemented.
When an item.s determined.to be outside the scope of the DBVP or is a i
duplicate of another existing item,-there is no further DBVP action required, and the item is statused as implementation complete. Although a more descriptive status categorization could have been used if procedurally.
allowed. the status is not. technically incorrect since it relates to-implementation'of DBVP-related action items only. Additional action outside
. the scopeLof DBVP is typically tracked and implemented in-accordance with'the TVA CAQ. process or'other existing mechanisms, which'are referenced on the SQEP-45 attachment 2 forms.
l With regard to the other punchlist items,.the implementation complete status I
- has now been deleted on the listing, except for punchlist item 9111 for which I
- DBVP corrective action was complete on August 19 1987.
While every effort
- has been made to input punchlist' data accurately and completely. errors have occurred occasionally.
As a result,'the DBVP issued Program Directive
- DBVP-D-87-008,~" Control and Processing of Changes to SQEP-45 Punchlist," as
~
stated previously in the TVA response to' Observation 6.21'.
Section III of the
' directive clarified the administrative requirements for verifying' accuracy of the punchlist data before distribution.
The control process was strengthened
- l
' by requiring that all, changes to the punchlist be submitted to the punchlist coordinator'on an SQEP-45' attachment 2 form. The punchlist coordinator.
performs an'administrativeLreview to ensure that the. requirements of SQEP-45 H
and the directive have been met. After data entry and before distribution of-i the punchlist, the punchlist coordinator verifles the correctness of the data entry.
TVA considers that these measures as well as others, which involvo SE review of punchlist items, have greatly improved the consistency and correctness of the punchlist data base. Applicable DBVP directives include:
DIRECTIVE NO.
TITLE DBVP-D-87-002 DBVP System Engineer Concurrence with Closeout of Punchlist Items DBVP-D-87-006 Statusing Punchlist Items DBVP-D-87-007 Closure of Punchlist Items It should be noted, however, that the computerized punchlist data base was
- developed only as a management tool to track action items and to provide sorting' capabilities.
The progression of each action item from open to closed to implemented is evidenced with the SQEP-45 attachments 1 and 2 forms, which 4
--_____.m.m_..
._.-___--_____.__-._______.-...___.m_-
l i
h l-j l'l' are maintained in-a controlled flie. These forms provide documentary evidence of other pertinent factors as well, such as restart categorization / justification and SE/ DES concurrence.
Implementation of corrective action is typically effected through already existing, procedurally controlled mechanisms, such as f
' CAQRs (including SCRs), ECNs, drawing deviations, and work requests.
These mechanisms provide documentary evidence of corrective action implementation and, for the most part, are tracked-by computerized systems such as PROJECT /2 and Tracking and Reporting of Open Items (TROI).
The time lag between input of data and reflection'of it in the punchlist was caused by the sheer number of changes being submitted.
During the time of the NRC' inspection, individual punchlist changes numbered as many as 2,000-2,500 changes per week, and a time' lag was an inherent by-product of this volume of data changes.
At present, the volume has' decreased substantially to an average of 300 changes per week, and the punchlict 6ata base is updated typically within 1 to 2 days.
l 1
-m____.______
r-
- o--
'j f:t e J
(
ENCLOSURE 2' 1
t LIST OF NEW COMMITMENTS FOR SEQUOYAH NUCLEAR PLANT UNIT 2 1.
A calculation'packago'will be generated before restart of unit'2 that will:
1 qualify instrument'line support variances.'54-33-A43 and'54-33-A45 (as part-of. ECN 5457) to current or unit 2 restart criteria, j
1 J
.j i
.< j
- I I
I i
l k
i-i I
I
p ?I 6 00T 2 8 1987 lom, Deputy Director, Division of Industrial 1 Nuclear Safety, NMSS
,, Director, Division of Radiation Safety and l,RegionIII
'IAL PROGRAM FOR MATERIALS PERFORMANCE INDICATORS R
lREGIONIII i
l l trial program as discussed in my June 11, 1987 e presented in Attachment 1 to this memorandum.
a success and recommend its use by the other nce indicators should also be incorporated in the ince licensing reviewers through their review of ensees, and site visits also have an opportunity degraded performance.
In fact, the reviewers on e of potential problems prior to the inspection ly in the process of developing a list of ch may be used by license reviewers.
We are also joint inspection and licensing reviews for ilizing the performance indicator concept, ease contact Dr. Bruce Mallett or Mr. Roy Caniano uk f. /hc
,-Qack A. Hind, Director 2 Division Radiation Safety and Safeguards 8080526 871020 2 LIC30 PDR 7
l
\\
t gea