IR 05000211/2007017
| ML17083B636 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon, 05000211  |
| Issue date: | 07/19/1985 |
| From: | Dodds R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML17083B635 | List: |
| References | |
| 50-275-85-27, 50-323-85-24, IEB-79-14, NUDOCS 8508080588 | |
| Download: ML17083B636 (50) | |
Text
U. S.
NUCLEAR REGULATORY CO>iMISSION
REGION V
Report Nos.
Docket Nos.
50-275/85-27 and 50-323/85-24 50-275 and 50-323 License Nos.
',Pacific Gas and Electric Company 77 Beale Street, Room 1451 San Francisco, California 94106 Facility Name:
Diablo Canyon Units 1 and
Inspection at:
Diablo Canyon Site, San Luis Obispo County, California Inspection conducted:
February 11 - July 17, 1985 Inspectors:
LLNL Contract Inspectors Approved By:
R. T.
Do ds, Chief Reactor Projects Section
7i ate Signed Summary:
Ins ection from Februar 11 throu h Jul
1985 (Re ort Nos. 50-275/85-27 and 50-323 85-2 Areas Ins ected:
Followup of various allegations pertaining to plant construction.
Results:
No violations or deviations were identified.
8508080588 8~0719 1~
PDR ',ADOCK 05000275 a
poR
~
1)
I
~>)
,
Persons Contacted f}
DETAIES, J'.
Tressler, Project Engineer D. Rockwell, Unit 2 Construction Superintendent
'.
Bratton, PG&E Allegation Coordinator D. Dunham, PG&E QC Engineer T. Pierce, QC Engineer S. Pascal, PTGC Electrical Department Engineer W. Cadwell, PTGC Electrical Inspector R. Ortega, PTGC Electrical Department Engineer C. Warne, PTGC Instrument and Control Department Supervisor J. Gisclon, PG&E Assistant Plant Manager K. Doss, PG&E Onsite Safety Review Group Member
.
H. Karner, Pullman, QA/QC Site Manager J. Guyler, Pullman, Internal Auditor C. Neary, Pullman, QC Supervisor C. Gunderson, PG&E Allegation Coordinator C. Dougherty, PTGC, Pipe Support Group, Engineer D. McGrew, PPP, Engineering, Engineer G. Thomas, PTGC, Pipe Support Group, Supervisor G. Vincent, QC Hotline G. Horrison, PTGC Electrical Department Engineer R. Cook, PG&E Allegation and Rate-Case Supervisor M. Iepke, PG&E Onsite Engineering Supervisor G. Klemm, PTGC Mechanical Department Piping Engineer R. Seewald, PTGC Mechanical Engineer G. Johnson, PG&E, PTGC Special Projects Engineer R. HcBay, BCI, PTGC QA/QC Records Supervisor D. Delfino, Pullman Power Products, Mechanical Engineer, NOTE:
The allegation characterization statements contained in this report are either a paraphrasing of the staff's understanding of the allegers concern or statements taken from,the allegation source document.
The characterization statements do not represent a staff assessment, conclusion or position.
2.
Task:
Alle ation or Concern Nos.
200 and 201 ATS No:
RV-84-A-030 a.
Characterization The PG&E nondestructive examination (NDE) weld inspection reports are inconsistent with Foley inspection rep'orts and/or the NDE reports were changed by Foley without the approval of the PG&E inspection team.
b.
Im lied Si nificance to Desi n Construction or 0 eration The allegation implies that the NDE inspection results maybe compromised due to record falsification which may allow uninspected or defective welds to exist in a safety-related structur L
c ~
Assessment of Safet Si nificance The staff reviewed the original NDE reports related to this allegation referenced by PG&E resp'onse DCL-84-186.
Changes in the weld NDE identification numbers were noted by a single line cross-out with the initials of the person performing the change.
No changes were made that totally obliterated the original.
The changes were for identification only because each,NDE report identifies the weld joint by connection number, detail drawing number and reference drawing, eliminating the possibility that the wrong joint. was inspected and/or repaired.
For example, a full penetration V-groove b'utt joint: is identified by
'Foley by two separate weld numbers.
One number is for the face and one is for the root.
In certain instances on vertical or horizontal weldments these weld numbers were inadvertently reversed due to the inspectors orientation while inspecting the weld. If the inspector was on the outside looking into the'uel handling building he may call the face of the weld No.
1 and the root No. 2. If he was oriented from the "inside looking out" these numbers were reversed.
This situation was rare due to Foley's gCP-6 requirement that the inspectors participate in a continuing training program at, least once every three months.
This training, emphasized orientation of prints and weld inspection sheets.
d.
Conclusion and Staff Position The allegation that PG&E inspection reports are inconsistent with Foley's inspection reports is substantiated in rare instances, but the inconsistencies are of weld identification number only and present no adverse'afety significance.
The staff, upon review of the NDE REPORTS found no evidence of unauthorized alterations of the reports.
e.
Action Re uired None.
3.
Task: Alle ation or Concern No.
0219 ATS No:
RV-84-A-045 a
~
Characterization Construction work on the PAM-2 panel in Unit 2 was transferred from HP Foley to PG&E. It is alleged that this action was contrary to procedures and was done to bypass equality Control inspections.
b.
Im lied Si nificance to Desi n Construction and 0 eration Bypass of procedures in this case could result in inoperability of this portion of the Post-Accident kionitoring System when called upon to operate and could indicate the possibility of a more general failure to fully implement the Diablo Canyon gC progra "h kkk
'ak 4
=
h r
kk k )h kl
'kh h
c ~
Assessment of Safet Si nificance The process for PGSE modifications to equipment during construction was reviewed.
Modifications of this nature are controlled by QC procedures DCP-10 and DCP-19.
Both procedures are applicable to work on safety-related equipment and require QC review of the work plans and QC inspection and sign off of modifications to safety-related equipment.
Documentation of activities conducted under these procedures is maintained via form E-72 for DCP-10 and E-53 for DCP-19.
d.
In order to determine if these procedures were correctly applied to the PAM-2 panel in Unit 2, the E-72 and E-53 logs were examined to identify modifications made by PGGE.
This examination looked for log entries identified as being associated with PAM or with Design Change Notices that affected PAM-.,2, as determined from the revision description for the PAM-2 wiring diagrams 435796 Rev.
2 and 435797 Rev.
2.
This process, identified a number of E-53's and E-72's that may have been related to the Unit"AM-2 panel.,
Each of these candidate records were, reviewed; E-72 number 642 and E-53 number 2082 were identified as the only PG&E conducted modifications on the Unit 2 PAM-2 panel during construction.
In both cases, the work was identified as Class,l,and the,<required QC approvals were obtained.
I Position I
Modifications to the Unit,: 2 PAM-2 panel that were done during construction by the PGSZ Proj'ec't Team',General Construction and the Dry Run Test Group were,'performed
'in accordance with approved and applicable PGRE procedures.
'ocumentation generated by these modifications provide evidence that the QC inspections and reviews required by the procedures were conducted.
e.
Action Re uired None.
4.
Task:
Alle ation or Concern No.
0378 ATS No: RV-84-A-062 a
~
Characterization:
A QC inspector alleges that he was never provided with support to perform his duties, nor was he provided with proper training.
b.
Assessment of Safet Si nificance Staff reviewed the affidavit to determine when the alleger was employed by the H. P. Foley Company as a
QC inspector.
The staff subsequently examined copies of QC inspector training records encompassing the indicated time perio I I
I I
I
's i
it II
II I
I I
pl
II
i I
I I A I I'1 li
r 44~
ig>>
ir II I 4"
I 4,
Mr I i,
I
11
0, It II I
The staff review of the.docume'nted training records shows that H. P. Foley Company conducted over 750, quality related training sessions from January 1981 to December 1983.
The number of people attending each session ranged fiom one'o twenty.
A review of specific training records indicate that the individual involved received classroom training in at least twelve different QA/QC Procedures, as documented on H. P. Foley "Quality Indoctrination and Training Records" dated March 25, 1981, April 1, 1981, February 16, 1983 and March 8, 'l983.
The training records are all signed by the alleger.
In addition to classroom training, the alleger also received over 90 days of on-the-job training.
c.
Conclusions and Staff Position The allegation that.
a QC inspector was not provided with adequate training cannot be substantiated.
Staff review of H. P. Foley training records provide evidence that extensive QA/QC training was indeed conducted.
The portion of the allegation that implies the inspector was "not provided with support" lacks specificity, and cannot be substantiated.
The allegation has no safety significance.
d.
Actions Re uired None.
5.
Task:
Alle ation or Concern No.
0624
~
~
ATS No:
RV-84-A-052 a
~
Characterization PG&E officials failed to correct a false statement to the NRC that the Company was ignorant of inaccuracies in the plant manual for operators.
b, Assessment of Safet Si nificance Staff review of the allegers affidavit indicates that, an inaccuracy in Volume 16 of the Plant Manual is the source of the allegation.
The necessary correction involved a hi-lo level alarm on the reactor coolant pump lube oil system which was incorrectly identified as a
lo-level alarm only.
Volume 16 of the Plant Manual is discussed on three separate occasions in the affidavit, specifically:
1)
Page 13: "...I had repeatedly notified PG&E management (including a memo to Mr. Thornberry, the Plant Manager, which is enclosed as exhibit 8) of an error in Plant Manual Volume 16...."
2)
Page 14: "...PG&E claimed ignorance of any errors in Plant Manual, Volume 16, even though I had written memo's on the subject to both John Gisclon and Mr. Thornberry, the Plant
~
'
i(
g
~
$
[
Jk'k
h
'f I*
I t
3)
Manager, and given copies of these memos to the NRC the previous'August."
Page 31: "...I had complained to the NRC that an error which I discovered in the Plant Manual, Volume 16, 'Annunciator Response'as still uncorrected'eight months after I had reported it.
NOTE: this problem was ultimately not corrected until January of 1983, over 20 months after I identified the problem...."
I GAP letter dated February 2,
1984, page 49, paragraph 170 states that "at a 1982 meeting, PG&E failed to correct a false statement to the NRC staff that the company was ignorant of inaccuracies in its Plant Manual for Operators."
PG&E's response to this allegation dated March 29, 1984, substantiates that the subject of inaccuracies in the Plant Manual were not discussed at the exit interview on January 7, 1984, and that Mr. Gisclon and Mr. Thornberry did not have the opportunity to correct this impression.
The PG&E response further stated, "...in fact, however, corrective action was already underway."
The "Corrective Action that was already underway" consisted of:
Response to the alleger's memo dated June 9, 1981 acknowledging the problem and indicating that it would be corrected.
Copies of the proposed changes were sent to the alleger by Rob Fisher on January 11, 1982.
"An on-the-spot change" to Volume 16 of the Plant Manual was issued on January 5, 1983, at the request of NRC, Permanent revisions to Volume 16 of the Plant Manual, including correction of the annunciator response problem were incorporated in Revision 3, dated July 15, 1983.
Conclusions and Staff Position The allegation that inaccuracies in the Plant Manual were not discussed at a 1982 meeting can be substantiated.
However, PG&E has taken the appropriate action to correct the Manual (Ref: Rob Fisher letter dated January 11, 1982).
The manual was corrected in September 1983 prior to loading fuel in Unit 1 in November of 1983.
The staff concludes that this allegation has no safety significance.
Non,I t
I'.
~
~
P 6 ',
~
~
~
6.
Task:
Alle ation or Concern No.
0?89 ATS No:
RV-84-A-064 a.
Characterization f
Welded extension bars at the 119 level in Unit 2 were inspected and bought off, but did not meet specifications or drawing requirements.
b.
Im lied Si nificance to Desi n Construction or 0 eration Welded extension for decking supports not meeting specifications or drawings may present a safety hazard due to failure upon loading.
C.
Assessment of Safet Si nificance d.
The staff on another allegation (reference allegation 995, ATS:
RV-84-A-064) conducted an investigation to establish its validity and as a consequence, examined inspection records for the 119 foot level.
Several inspection discrepancies were found.
One of the three corrective actions taken by Guy F, Atkinson, Co.
(GFA) was to reinspect all related field work at the 119 level.
Of the 900 items to be reinspected, 42 were found inaccessible due to status of the construction work.
The remainder were either found acceptable or minor repairs were made.
The inaccessible items were accepted without reinspection by PGRE following an evaluation by the PGSiE engineering staff.
Staff Position Based on the significance of identified discrepancies, the staff concludes that there is reasonable assurance that the welded extensions meet the specifications and drawings.
e.
Action Re uired None.
7.
Task:
Alle ation or Concern No.
0833 ATS No:
RV-84-A-064 a
~
Characterization A welder disregarded the ISO drawing, field installation instructions, process sheet instructions, weld rod requisition instructions and then incorporated his own welding procedure.
b.
Lm lied Si nificance to Desi n Construction or 0 eration The failure of a welder to weld in accordance with the qualified weld procedure could result in a defective pipe weld and the inability of a safety-related pipe system to perform its intended functio Il IO
Assessment of Safet Si nificance The staff reviewed the Pullman Power Products (PPP) internal audit number 85 for 1981, which was the basis for the allegation.
Audit number 85 was performed on
CFR 50 Appendix B, Criterion XVI, Corrective Action.~ Audit action request No. 4-4 was the specific finding in the audit report.
sr The original documentation wa's retrieved and reviewed from the records vault 'and all procedures or specifications in effect at the time of the audit were obtained.
The field work was shown on PPP isometric drawing No. 23-21 Revision "P".
Field welds 477 and 478 attach a stainless steel valve PCV 658 to a 4-inch, schedule
carbon steel pipe.
These,welds were'ade on February 10 and ll, 1981.
The first of the three-part audit Cinding stated that the welder disregarded the GTAW/SHAM weld procedure (WP 149)
and continued to use GTAW for the weld-out (WP 150).
This statement is incorrect.
It misrepresent's, the requirements of weld procedure 149, which permits 1 to 3 passes of GTAW after the root pass.
Xn this case,'
consumable insert was used to weld the 0.237 inch wall thickness pipe to the valve.
In a weld of this configuration, no more than three GTAW cover passes would be required to complete the weld.
Hence, the complete GTAW weld-out per procedure WP 149 is identical to MP 150.
The weld procedure change to the process sheet was made
. by the field engineer on February 10, 1981, which means the change was prior to or at the time of the welding.
There was no attempt made by the welder to use "his own" procedure.
In this case, the design of the pipe is not affected by use of either WP 149 or WP 150.
The weld procedure specifications, procedure qualification records, and welder qualification records for both procedures were found acceptable.
The second part of the audit finding was that according to the weld rod requisition sheets, welding occurred on February ll, 1981, while the process sheet noted welding was completed on February 10, 1981.
ESD-264 requires the PPP inspector to Jill all the blocks on the field process sheets and not'he individual performing the operation (the preferred method).
The method used required the welder to notify the inspector when welding was complete and to enter it on the process sheet.
ESD-202 required the weld foreman to fillout the weld rod requisition sheets the dsy ~rior" to the weld being made.
A detailed review of the field process and weld rod requisition sheets for both PW 477 and 478 showed that the inspector, not the welder, filled in the completion date as February 10, 1981, a day earlier than the filler requisition date.
The process sheet indicated that no welding was performed after the final inspection by the PPP inspector.
All welds were subsequently visually inspected and radiographed before acceptance.
The third part of the finding states that the field engineer did not initial, date and explain changes made on the process sheets.
After
I'
I'l I I
r gltr I
I I
r h
VL
I j
I f
. lr
~ %
'I',>
I I
I'
I r'I r
~4 I
r
)
I}
r iI l~,
t t. It t IT
'
r II I
I I
reviewing the FW 477 and 478 process sheets, it was found that the field engineer did sign and date the weld procedure change on both process sheets.
The field engineer did not initial and date where the same change occurred redundantly on the same form.
There was no written explanation of the change but it was clearly evident the change was from WP 149 to WP 150.
ESD-264 paragraph 7.2 only requires explanation of changes and not the reason for the change.
Again, from the foregoing evaluation, there was no need to make this change to the process sheet.
The staff reviewed several subsequent audits for the follow-up actions to this audit.
1982 Internal Audit Number 31 dated July 6, 1982, on
CFR 50 Appendix 8, Criterion XVI, Corrective Action and 1982 Internal Audit Number 35 dated December 1,
1982 on Criterion IX, Special Processes were reviewed to see if similar problems noted in 1981 Audit Number 85 were still detected.
Audit.'31 was an
"unscheduled" audit. It was never completed and was voided by the internal auditor on approximately July 6, 1982.
The checklist for Audit 35 has an item No.
4 which was <o verify that there were no deviations from the required weld procedures.
However, the audit results did not indicate whether'he, status of work performed was acceptable or not.
Subsequent audits'examined, such as Audit 118 dated May 27, 1983 on Criterion X, Inspection and Audit 131 dated April 24-30, 1984 on, Criterion IX, Special Processes, were complete, accurate and had a checklist 6f'"reported result's.
For example, the response to Audit Action Request No. 5 of Audit 118 showed the corrective action and steps taken 'to prevent. reoccurrence were appropriate and consistent with the,~finding.
Conclusions and Staff Position The allegation that a welder disregarded the weld procedure and process sheet cannot be substantiated.
The staff has concluded that Audit Action Request No. 4-4 of PPP Internal Audit Number
appeared to misrepresent the situation as evidenced by the original documentation.
The weld procedure was correctly implemented.
It appears that the inspector errored in his entry of the weld completion date, not the welder.
The changes made to the process sheet were not necessary but, even having been made, they still met the intent of the requirements.
There is no apparent safety significance to this allegation.
Actions Re uired Non ~.
I K
I I
II
\\ I F
k F
R
~ C (
'- I C,
$ -g~
I (F Fl
l 8.
Task:
Alle ation or Concern No.
0995
'TS No:
RV-84-A-064 a
~
Characterization Inspection work performed under Atkinson's control in. 1978 in the northwest corner roof area of turbine building was performed by a questionable inspector.
b.
Im lied Si nificance to Desi n
Construction or 0 eration An unqualified inspector could permit unsatisfactory construction work to go unrepaired resulting in structural failure.
c.
Assessment of Safet Si nificance Upon disclosure of the identity of the alleged unqualified inspector, the staff proceeded to investigate the individual's qualifications and the extent of his participation in the inspection of the construction of the Turbine Building roof.
In accordance with the Guy F. Atkinson Company (GFA) Diablo Canyon Project Specification No. 5442, Quality Assurance
- Personnel Q.8-9, the inspector was indoctrinated, trained and certified (effective August 2, 1978)
as a Structural Steel Welding and Erection Inspector by GFA's Quality Assurance Hanager consistent with GFA's Quality Control Procedure 13.
The supporting documentation (references 2,
3, and 4) was reviewed and found acceptable.
Contrary to the allegation, the inspector did not perform any work of record in the turbine building roof area.
Inspection records for the roof structure generated by GFA during the period of the inspectors employment at Diablo Canyon were examined and his signature did not appear.
Although the foregoing clearly resolves the specific allegation, further investigation was conducted to establish the validity of the allegation with respect to the work actually performed by the inspector.
As a consequence, inspection records for the 119 foot level (location derived from reference 1) were examined for the period of the inspector's employment.
Several inspection discrepancies were identified that had also been found and documented by NCR 293 dated April 17, 1979, during an audit of inspection records performed by GFA.
The audit was conducted in accordance with GFA - QCP-15 (reference 5).
Virtually all of the discrepancies described by the NCR were attributed to the inspector in question.
The corrective actions taken by GFA were to:
1)
conduct a thorough review of all of this inspector's documentation, 2) reinspect all related field work and, 3) not rehire the individual who had already terminated, to return to college.
The inspector was found to have been involved with 900 items and 750 (83@ of these were reinspected and found acceptable (reference 7).
fh II i(
'i f
E
"V
Of the remainder 108 (12%) were found to have minor deficiencies that were quickly repaired and 42 (5Q were inaccessible due to the advanced state of construction at the time of the GFA audit.
Half of the inaccessible items had been ultrasonically tested and accepted by another'nspector.
and,'herefore,,also accepted by PGSE based upon the availability of test records.
The remaining 21 items were accepted by.the PG&E engineering staff on the basis that "welds were either not critical to resist loads, or margins between weld size shown on design drawings and weld size required to resist design loads were more than adequate considering limited deficiencies that might exist."
d.
Conclusions and Staff Position The allegation as stated could not be substantiated to the extent that the inspector met all of the requirements established for the job and that no evidence could be found that he had performed inspection work in the northwest roof area.
However, further investigation did validate significant deficiencies in the inspector's performance in another location which had been identified by the licensee's gA program and properly dispositioned, including the evaluation of inaccessible welds.
e
~
Actions Re uired None.
~
~
9.
Task:
Alle ation or Concern No.
1062 ATS No:
RV-84-A-064 a
~
Characterization Missized or improperly installed beam clamps have been used to fasten raceway support struts to structural beams.
b.
Im lied Si nificance to Desi n Construction and 0 eration Inadequately supported safety-related raceway could result in degradation or failure of critical power, control or instrumentation circuits during events (e.g.,
that cause abnormal loading of the raceway supports.
C.
Assessment of Safet Si nificance The sketches provided as part of the allegation indicate that the Unistrut P2785 and P2786 style beam clamps are the items of concern.
This is the predominate beam clamp style in use at Diablo Canyon.
This type of clamp is a angled piece of plate steel that captures the foot of a beam between it and the framing channel.
The plate is held in place and tightened against the beam by a U bolt which is installed around the framing channel and through the plate (see
~
~
V I
k'
t
ll j"
lt I
Sketch).
Vendor catalogs and PG&E drawing 50030 Revision 30 require that these clamps always be used in pairs.
I Review of the Unistrut, Power, Strut and Super Strut catalogs shows that Unistrut produces two clamp size's 'in this style, Super Strut three sizes and Power Strut two sizes. 'n all cases, the only difference between the different size clamps of each vendor is the U-bolt length.
If,a clamp with too short of a U-bolt is selected, the clamp will not fit in the application because U-bolt threads will not extend sufficiently beyond the clamping plate to allow installation of the U-bolt nuts. If a clamp with too long of a U-bolt is selected, then it is possible that the U-bolt nuts would reach the end of the threads before the clamp was completely tight.
The selection of the wrong size clamp will not, in and of itself, result in a loose attachment of strut to beam.
Loose clamps will only occur if the U-bolt nuts are not properly tightened.
The results of previous NRC raceway support inspections were reviewed and additional inspection of beam clamp installation was conducted.
As a part of the NRC inspection of the seismic reverification program, 300 raceway supports were inspected including 24 supports whose design allowed for the use of beam clamps.
No loose beam clamps were identified during these inspections.
In response to this allegation an additional 200 beam clamps on raceway supports in both Unit 1 and Unit 2 were inspected for tightness.
During this inspection one loose beam clamp was identified.
In this case, the proper size U-bolt was employed, however, the nuts were simply not sufficiently tightened.
PGSZ issued Minor Variation Report No.
3281 on April 2, 1985, to tighten the clamp.
d.
Conclusions and Staff Position The review of vendor information, establishes that the size of beam clamp used does not directly affect the tightness of the clamp. It is further concluded from the results of raceway support inspections that loose beam clamps has not been a generic problem at Diablo Canyon.
e.
Action Re uired None.
10.
Task:
Alle ation or Concern No.
1270 ATS No:
RV-84-A-071 a
~
Characterization PGSE response that an individual was not aware of orders to stop working on hangers is misleading.
Another individual gave the stop
I if k
V
D
work order, but the individual who claims to be unaware of the order should have known about it.
Im lied Si nificance to Desi n Construction and 0 eration The allegation is no't sufficiently detailed to allow an assessment of its significance to design, construction and operation.
Generally, stop, work orders, are"issued to hold construction while problems with design, materials or procedures are resolved so the amount of rework-'equired is minimized.
In this regard, failure to be aware of ox'o comply, with a" stop work order would be more likely to have economic significance than safety significance.
Assessment of Safet Si nificance The portion of the allegation provided to the investigation team did not specifically identify the subject or time period of the order to stop work that is in question.
It is assumed that the oxder in question was a. fo'xma1 stop work order on pipe support installation and that the order could have occurred between Mar'ch 1983 and April 1984.
The general subject of stop work'orders on pipe supports was
'iscussed with a pipe support supervisor and a group engineer of PTGC.
They stated that one significant stop work order issued during this interval involved supports that were being evaluated by engineering as part of the seismic reverification pxogram.
Pullman Power Products (PPP)
was instructed to stop construction on these supports by a memo from G. A. Thomas to H.
W. Karner dated August 8, 1983..
The intent of this order was to stop work on supports that might be subject to a design change as a result of the reverification program. It was 'stated that PPP issued stop work orders on the supports in question and hung hold tags on each support.
Failure to stop work on any support in this case would have no safety significance, as any design change to a support would be implemented whether or not construction on the support was continued.
The PPP Pipe Support stop work orders were reviewed for the interval in question to audit 'Pullman's performance in response to stopping work on the supports discussed above and to determine if other irregularities exist that might be the source of the allegation.
A sample of the supports identified in the August 8, 1983 memo was selected and it was verified that for each sample case, a
corresponding PPP stop work order was issued.
During the review of the stop work orders it was noted that one PTGC memo dated September 15, 1983, was referenced by stop work orders thxough March 1984.
This apparent discrepancy in dates was investigated and it was determined that the September 15, 1983 memo instructed PPP of a policy change that could also be the source of the allegation as discussed below.
Pullman Power Products procedures require that construction work on a support be stopped when a new revision of the drawing for that
I
)I
ly ll I"
y p
I y
I v
K il
"I I
)
)I
I I K
I)
'l,
)I IK X
I
~,yy y
l~
,t I
'
I w
yI P
) I
- '
I
'1 e
ll I y I
I
K>>
1 t-.
K I J
'j I
I
II I
Il II I
y
)
I I
't I
I y5 4
)K I
~ 'I
.
1
support is received.
Mhen a new revision is received a stop work order is issued, the old construction package is pulled back from the Iield, a new construction package is issued and the construction hold is removed.
As a result of problems encountered in construction, PTGC instituted a pipe support preinspection program that evaluated new designs for constructibility before they were turned over to PPP.
During the time new drawing revisions were under review by the preinspect group, PPP was not formally aware of the new revisions thus construction work on the affected supports continued even though new drawing revisions were onsite.
This apparent violation of PPP procedures was noted by preinspect group inspectors.
However, since PPP had'ot received the new revision, their actions were in accordance with written procedures.
Initially, no action was taken on this problem because PGSZ judged that a minimal amount of rework would be required as a result of construction continuing while the >ew design revisions were under consideration by the pr'einsp'ect group.
After an instance was identified where significant rework"',would have been required had construction continued during preinspection, PPP was instructed to stop work on any support when. the.PTGC preinspect group received a
new revision of:.drawings, for that support.,
This policy change was transmitted by the September 15," 1983 'memo from G. A. Thomas to H.
M. Karner.
Review of the PPP stop work orders from the time of the memo forward indicated that PPP consistently implemented this policy.
Even if this were not the case, failure on the part of PPP to stop work on a support would have no safety significance since the support would eventually have,to be reworked to comply with the latest drawing revision.
d.
Conclusions and Staff Position No irregularities were noted in the PPP pipe support stop work orders during the interval of March 1983 through April 1984.
The two issues that resulted in the majority of stop work orders during this interval were investigated and appear to have been handled properly by PGSE and PPP.
Therefore, an allegation that stop work orders during this were not properly implemented or that key individuals were not aware of stop work orders cannot, be supported.
Furthermore, failure to take proper action on the stop work orders reviewed would have had economic impact, but no safety impact.
e.
Action Re uired None.
11.
Task:
Allegation or Concern No.
1381 ATS No:
RV-84-A-073 a
~
Characterization A nut could not be properly installed on an embedded stud that attaches a main feedwater pipe support to the containment wall.
The
'h (
I *
I I
I I'
I h
I
,>>'l
~
>>
~
~
b.
k nut was cut in half and tack welded, to'the base plate.
This problem is described in Diablo Canyon Problem Report 1-2335-P.
Im lied Si nificance to Desi n Construction and 0 eration An improperly supported feedwater pipe could result in a loss of feedwater during a transient that'-'places abnormal loads on the support.
The pipe support in question is upstream of the main feedwater check valves so that problems with this support would not impair the ability to provide auxiliary feedwater.
Nevertheless, the support in question is classified as safety-related by PGSZ.
c ~
Assessment of Safet Si nificance Diablo Canyon Problem Report DP-1-2335-P contains a detailed description of the condition that is the source of this allegation.
The embedded mounting studs for support 1048/8SL were too short to allow complete thread engagement for the base plate nuts.
As a field fix, the nuts were cut in half longitudinally, installed on the studs and tack welded to the base plate.
Engineering approval of this configuration was provided on February 14, 1979, in response to the problem report.
Approval was based upon a determination that enough threads were engaged to allow the full bolt strength to be developed.
The unusual configuration of the base plate attachment was again noted during final gC inspection of the support.
The discrepancy report generated as a result of this inspection (PPP DR 5734)
questions the suitability of the nut tack welds since neither the nut material or the welding procedure used could be established.
This item was resolved by memo from J.
A. Longworth to J. Arnold and R. Tinkle dated January 20, 1984, which recommended that the nut welds be accepted as is, since the welds serve only to prevent the nuts from backing off and are not assumed to add any structural strength to the support.
The partial engagement of the base plate nuts is documented on the as-built drawing for the support and the as-built configuration was accepted by PGSE engineering.
Some confusion about PGM's ultimate acceptance of the nut tack welds may have resulted from the fact that DR 5734 also identified a number of undersized structural welds on the support.
Pipe Support Design Tolerance Clarification Form No. TC-1-14461 indicated that some of these welds were unacceptable as is and should be reworked.
The rework required by TC-1-14461 was completed and the resulting support configuration was incorporated into the as-built drawings.
TC-1-14461 does not address the nut tack welds.
d.
Conclusions and Staff Position The condition described in the allegation exists in the field.
This condition was documented as a discrepancy on two separate occasions and was properly dispositioned on both occasion t
'
I fl I
P ll(
ll
e.
Action Re uired None.
12.
Task:
Alle ation or Concern No.
1264 ATS No:
RV-84-A-071 a
~
Characterization b.
PGSZ response to GAP allegation No.
259 on, walkdown by personnel who marked-up drawings used by equality Control is misleading.
The people were not qualified.
Assessment of Safet Si nificance The implied safety significance of this allegation is that the quality of safety-related systems may be compromised if they were analyzed and constructed using improper drawings.
C.
Assessment.
of Safet Si nificance The source of this allegation was GAP submittal No.
120 dated June 21, 1984, attachment 10, page 12, which states that PGSE's response to GAP allegation No.
259 is misleading regarding qualification of personnel performing walkdowns.
The alleger is concerned that unqualified personnel worked on the drawings and that "those drawings were used for gC inspection work throughout the plant and for engineers to make decisions on the acceptability of design changes."
It is not apparent that the alleger understands the use of drawings that are marked-up as a result of walkdowns.
Marked drawings that result from walkdowns are used for updating as-built drawings which are later analyzed and used for gC inspection.
The NRC staff examined the concern regarding qualification of walkdown personnel by: reviewing PGSE instructions and procedures for walkdown inspections; reviewing a sample of ISO drawings used for the walkdowns; and verification of training of individuals who signed the drawings used for walkdowns.
PGSZ instruction entitled "Field Reverification Procedure for Piping Isometrics" was prepared in response to NRC Bulletin 79-14 on March 10, 1982, and was revised on April 12, 1982.
On May 6, 1982, PGSZ engineering assumed control of the procedure, and re-issued it as instruction numbers I-17 for large bore pipe and I-18 for small bore pipe.
The NRC staff randomly selected seven isometric drawings that were marked-up during walk-downs, and recorded the names of individuals who signed the drawing lk N
Training records were later reviewed, and, in all cases, personnel who performed the walk-downs and signed the drawings had been previously trained in the walkdown,procedure.
Review of PG&E's 79-14 Training 'File indicates that over 50 people have been trained in the walkdown procedures.
No violations were'identified.
d.
Staff Position The staff concludes that personnel who performed walk-downs were adequately trained and qualified to do the work.
None.
13.
Task:
Alle ation or Concern No.
1397 ATS No:
RV-84-A-073'
~
Characterization On diesel generator fuel oil transfer system pipe support (support number 20/85R),
a nut is bearing on a weld and half of a shim plate.
This condition.is not shown on the as-built drawings, but is shown on pipe support tolerance clarification form TC-1-11497.
b.
Im lied Si nificance to Desi n Construction and 0 eration A improperly supported pipe in the diesel fuel oil transfer system could result in eventual failure of one diesel generator following conditions (such as a seismic event) that imposed abnormal loads on.
the piping system.
c ~
A'ssessment of Safet Si nificance A copy of the drawings for support 20/85R was obtained and the support was inspected against the drawing.
The inspection confirmed that the condition described by the allegation exists.
Specifically:
Iine 1-2-S-2599-2" rests on a shim plate that is welded to one of the support cross members.
A U-bolt is installed around the line and through holes that were drilled partially through the shim plate and partially through the weld material.
U-bolt nuts are installed on the top side of'he ship plate to provide for the design spacing between the pipe and the U-bolt.
The U-bolt is secured from underneath by a nut and a lock nut on each side.
The top nuts rest partially o'n the shim plate and partially on the weld.
These nuts are the source of the allegation.
Contrary to the allegation, the correct version of the hanger drawing for this support details this configuration (Drawing 049318 sheet llOB Section D-D in conjunction with section E-E on sheet
- l
'N f
C N
llOH).
These details were added by either revision 3 or revision 4 of the drawing.
Xn both cases, the Field Correction Transmittal form documents that the as-built configuration has either been factored into the support calculations or that the calculations have been reviewed to verify that the condition does not affect the adequacy of the support.
d.
Conclusions and Staff Position The condition described by the allegation exists in the field, however, PGSE has"documented, assessed and accepted the configuration described.
More than enough weld material remains after drilling holes for the U-bolt to insure secure attachment of the shim plate to the support.
The incomplete seating provided for~the top U-bolt nuts is of no consequence since the nuts are only used to support the U-bolt in order to provide the designed gap between the U-bolt and the pipe.
Therefore, the staff. agrees that the condition observed does not adversely affect the adequacy of the support.
e.
None.
14.
Task:
Alle ation or Concern No.
1486 ATS No:
RV-84-A-113 a
~
Characterization An individual questions the acceptability of capped off lines coming from the reactor vessel support pads.
b.
Im lied Si nificance to Desi n Construction and 0 eration C.
Capped lines could be indicative of incomplete construction work.
Assessment of Safet Si nificance The piping and Mechanical Drawing (Drawing 50152 Change 3) of the reactor vessel suppo'rt block was examined to identify pipes coming from the support pads.
Twenty four pipes were identified.
Inspection of this piping in the Units 2 containment, confirmed that lines 2K1696, 2K3003, 2K391, 2K1702, 2K2947, 2K400, 2K1700, 2K2254, 2K395, 2K1698, 2K3008 and 2K393 are capped.
The piping and instrument diagram for the Component Cooling Water System (Drawing 108014 sheet 7) shows these lines as Cooling Water lines to spare Reactor Vessel support coolers and that they are capped per desig ti I*
.I
/
It, I
l'/'l.'
f II.,
'.
/
'I II w
t l
'l.q II E
I I
,.)
/
I
'l II ~ e E fl
/
I t
I 'l
!
E Ill lt Il IE I
l I
- N E
I,/
E l
I ll Itk I
l'
d.
Conclusions and Staff Position The capped lines coming from the reactor vessel support pads are for spare equipment and are capped per design.
e.
Action Re uired None.
15 'ask:
Alle ation or Concern Nos. 1566'nd 1567 ATS No:
RV-84-A"115 a
~
Characterization b.
C.
Welders worked without shielding gas flow meters and incompatible meters.
I'i Im lied Si nificance 'to Desi n";Construction or 0 eration Welds made without shielding 'gas flow'rate information or inaccurate information could possibly re'suit in oxidized weldments, with lower mechanical and physical properties of the., base material, thus not meeting the original design criteria.
Assessment of Safet Si nificance
'uring an audit of welder. performance by Pullman Power Products equality Assurance on June 4, 1984, it was discovered that a welder was using a flowmeter that was not identified as an acceptable flowmeter in weld procedures ESD-213 and ESD-219.
Further investigation by Pullman revealed that improper and uncalibrated pressure regulators were also in use.
These deficiencies were documented by Pullman Power Products on Deficient Condition Notice (DCN) Number 96-058, dated June 7, 1984.
Pullman Power then prepared Deficiency Reports (DR's)
8657 and 5946, and submitted them to PGSE for disposition.
PGSE's investigation resulted in a revision to PPP Procedures ESD-213 and ESD-219.
The revisions included a description of and instructions for the use of additional types of flowmeters, including those found during the audit.
Supplementary instructions were issued to all welders and QC personnel were indoctrinated in proper flowmeter and regulator use.
PGSE investigated regulator calibration requirements with the manufacturer, and reviewed the ASME code,Section IX.
The ASME code addresses the shielding gas flow rate as a "non-essential" variable.
Because of this, changes in gas-flow were deleted from the PPP ESD's, and the existing welds were accepte I
4 I
w
'I
~
~
d.
)
Conclusions and Staff Position i
lt Jl The allegation that welders'ere using improper flowmeters and uncalibrated regulators is substantiated.
However, the staff further concludes that PGRE acted properly in the corrective action taken to resolve the situation of the use of uncalibrated weld shielding gas pressure regulators and incorrectly matched flowmeters to regulators by revising PPP's welding procedures ESD-213 and ESD-219 to reflect conformance to ASME code Section IK which considers the shielding gas flow rate for manual gas tungsten arc welding as a non-essential variable.
All welds performed required NDE assuring weld soundness.
This allegation, therefore, has no safety significance.
e.
Actions Re uired None.
16.
Task:
Alle ation or Concern No.
416 ATS No:
RV-84-A-062 a
~
Characterization Inconsistent weld acceptance standards were used by day and night shift inspectors inspecting structural steel welds for the Guy Z. Atkinson Company.
b.
Im lied Si nificance to Desi n Construction and 0 eration Inconsistent standards could have resulted in unacceptable welds being accepted.
This could cause structural steel welds to fail under abnormal loads if weld strength is less than assumed by the design.
Co Assessment of Safet Si nificance Exhibit 1 of Attachment 10 to the GAP letter indicates that one example of the night shift requiring rework on welds accepted by day shift inspectors is excess reinforcement on stiffener plates 3 and
on column 14.
Review of weld inspection records indicated that column 14-A was inspected on the date in question by the inspector identified in Exhibit 1.
Stiffener plates 1 through 16 on column 14-A and stiffener plates 7, 8, 15 and 16 on column 12.1-A were inspected as part of the investigation into this allegation.
No evidence of excess reinforcement was noted.
Thus, the example cited does not support the allegation.
To address the question of inconsistent inspection standards, two thousand two hundred and fifty (2250) structural steel weld inspection records were reviewed for 1855 welds reflecting the work of. 22 different weld inspectors to determine if any groups of inspectors had significantly different, weld reject rates than other groups.
The records reviewed were from two groups.
One group was
J IC J
E(
i I
t e
S 1,
randomly selected from the structural steel welding records for the turbine building.
The second group was selected to include the records for the stiffener plate welding referred to in exhibit 1 of the allegation.
The records encompassed the time period June 1978 through August 1979," thus spanning several months before and several months after the date of exhibit'.
Of the 2250 inspection records reviewed, 126 (5Q were rejects.
Five inspectors accounted for all of the rejects.
The bases for these rejections were reviewed to determine if they were attributable to different inspection standards.
J V
J Of the 126 rejects found in the sample, 111 resulted from reinspection of 395 welds that had been originally accepted but which subsequently cracked at the ends as documented by NCR's 243 and 313.
The resolution of these NCR's required grinding out the weld cracks, visually inspecting the welds and conducting liquid penetrant testing to insure no cracked weld material remained to serve as a starting point for further crack growth.
All welds that were reinspected were visually acceptable.'he rejects resulted from the liquid penetrant testing.
Since the circumstances surrounding these rejections are significantly different from the other weld inspections reviewed, the high reject rate in these cases clearly result from circumstances other than acceptance criteria.
The reject rates for the two inspectors who conducted the reinspection were roughly the same.
The remaining 1855 weld inspections reviewed resulted in 15'ejects (<lQ by three inspectors.
Seven of the rejections were by a single inspector who conducted 906 of the inspections in the sample.
This represents a rejection rate of less than one percent, which is not clearly inconsistent with the other inspectors rejecting no welds out of a much smaller number of inspections.
The eight remaining rejects were by two inspectors who inspected a total of 52 welds.
It could not be determined if both of these inspectors were on the night shift.
The eight rejects occurred on three joints, therefore, it appears that the relatively large percentage of rejects for these two inspectors resulted from the same welder making more than one reject weld on a joint.
To verify that different acceptance criteria did not result in welds being more readily accepted by one group than another, the productivity rate for two of the inspectors that inspected a large percentage of the welds in the sample, were examined to verify that no more than a credible number of welds per day had been inspected by these individuals. It was determined that one inspector had accepted a very large number of welds in one day.
Review of these welds, however, indicated that all of the welds were grouped together so that it was possible to inspect the number of welds indicated.
Conclusions and Staff Position The specific column referred to in the allegation as an example of unacceptable work being accepted by day shift showed no indication
J f
li
,( N
of excess weld reinforcement as. alleged.
Review of a large sample of weld inspection records did show differing weld rejection rates by different inspectors, but in all cases the higher rejection rates noted could be attributed to factors other than different acceptance standards.
No evidence was noted that different standards resulted in inspection of more than a credible number of welds per day by an inspector.
Therefore, it was not clear that there was any group of inspectors (day or night shift) that had higher rejection rates due to differing acceptance criteria.
Thus, the allegation of differing acceptance criteria between day and night shifts cannot be substantiated.
e.
Action Re uired None.
17.
Task:
Alle ation or Concern No.
1543 ATS No:
RV-84-A-114 a
~
Characterization An individual feels that PG&E made false statements that led to the resolution of previous allegations regarding welder qualification test surveillance.
Specifically, the existence of welder qualification records referenced by a J.
P.
Watson affidavit presented before the Atomic Safety and Licensing Appeal Board is questioned.
b.
Im lied Si nificance to Desi n Construction and 0 eration Lack of adequate welder qualification test records would impair the traceability of weld quality for any welds made by welders lacking adequate qualification records, invalidating the acceptance of these welds.
As a result some of the accepted welds might fail under abnormal loading conditions because they lack the structural strength assumed by the design.
Assessment of Safet Si nificance The allegation stems from apparently conflicting testimony regarding welder qualification records that was given by J.
P.
Watson on two different occasions.
In the ASLB affidavit, it was indicated that records of fitup, root pass, filler pass, final visual and bend test had been reviewed for a group of welders.
In subsequent testimony during a Department of Labor hearing Mr, Watson indicated that there were no permanent records of examinations conducted at qualification weld hold points.
The process used by Pullman Power Products (FPP) to conduct and document welder qualification tests was reviewed to determine what documentation of qualification tests was generated and maintained.
Welder performance qualification is controlled by PPP procedure ESD 216.
This procedure requires the QC Welding Supervisor to
I f
i
<
'pl
22, supervise welder performance qualification testing and document the results of the guided bend tests or radiography used to determine the acceptability of the qualification weld.
These requirements and the documentation generated'nd'aintained.
by PG&E fulfillthe welder qualification requirements of "ASHE Code Section IX Article III, gW 301.4.
Wl Discussions with 'a'PP mechanicalngineer indicate that it was also standard practice fo',PPP
'(}C'Weelding, Supervisor's to inspect the qualification weld p'rogressi~on at selected hold points, usually fit up, root pass and final visual inspection.
The results of these inspections were maintaine'd,in-informal notes by 'the Welding Supervisor, but were not.maintained as"permanent formal Diablo Canyon gC documents.
d.
Conclusions and Staff Position The apparent.conflict in testimony seems to result from Hr. Watson referring on one hand to records consisting of informal notes and on the other hand to formal permanent records.
Therefore, no real inconsistency between the statement that (informal) records of gC inspection for fitup, root pass, filler pass, final visual and bend test were reviewed and the statement that (formal) permanent records of these inspections do not exist.
Therefore, the allegation that PG&E made false statements regarding welder qualification test documentation cannot be substantiated.
Furthermore, the welder qualification records maintained by PG&E fully comply with the ASME requirements for such records.
Therefore, failure to maintain records of inspections conducted at hold points during the qualification testing has no safety significance.
The specific questions raised in the alleger's affidavit are addressed as follows:
Are inspection records of welder qualification testing required at nuclear power facilities?
Yes, however, the ASME code only requires that the essential variables, test results and background data be maintained as records of welder qualification.
The welder qualification records maintained by PG&E satisfy these requirements.
Does the welder qualification test have hold points that must be inspected before the test can be allowed to proceed'?
No, although inspection at hold points was standard practice by the PPP gC Welding Supervisors, it is not required by either the ASME Code or by the written PPP procedures.
If a gC inspector performed an inspection of a mandatory hold point, would that inspector be required to document the results of such an inspection?
I I'
V
,k>>
I I II
r I
$
lh I
I
!
kkk 1'
lt (I It
),kl'l I
6, rk I,
rk
~
k I
>>
,I I
I tt I
V
1 I
k
)Vl II
~
No, as discussed above neither the inspection hold points nor documentation of the results of hold point inspections is mandatory.
e.
None.
18.
Task Alle ation or Concern No.
169 ATS No:
RV-84-A-022 a
~
Characterization:
Pullman failed to conduct support welds as required by procedure and QA wrongfully approved the technique.
b.
Im lied Si nificance to Desi n Construction or 0 eration Welds made without the proper procedure could result in improperly made welds in critical areas.
c ~
Assessment of Safet Si nificance The alleger stated that two W14x90 wide flange beams were welded together on support nos.
2/45R and 2/49R by use of an unqualified welding technique.
It was stated that Pullman welding procedure specification (WPS) 7/8 was used to join steel shapes without the required backing bar in place of which back-gouging was performed and the Pullman QA/QC Manager wrongly approved the technique utilized.
The staff reviewed packages from support nos.
2/45R and 2/49R and Pullman's WPS 7/8 to evaluate welds in question.
The joint type is a groove butt in which a backing bar is placed to ensure full penetration when welding.
Pullman's General Field Support Process Sheets for these supports indicate a hold point after back-gouging to verify sound weld metal was attained from the back side before welding the back side.
Both techniques are approved by the AWS and ASME codes.
Both techniques achieve the same purpose:
to obtain full penetration at the root.
The AWS Code has both of these particular weld joint types as prequalified in their Structural Welding Code Dl.l, but the omission or inclusion of a backing bar is an essential variable in which a WPS has to be requalified if the weld joint was not prequalified.
ASME Section IX addresses backing bars as nonessential variables requiring no requalification of a WPS.
Pullman's WPS 7/8 closely resembles ASME Section IX but does not state that Section IX is to be used as a supporting document when a
interpretation is required.
Therefore, technically the alleger was correct in stating that the support welds made without a backing bar was performed without a properly qualified procedure because back
P P
1 h
I, I
II
gouging was not addressed on VPS;.7/8.
On March 12, 1984, MPS 7/8 was revised to allow for back grinding'and back welding.
Pullman's gA/gC Manager has no "recolle'ction in approving the use of backgouging as an alternative to the use of a backing bar for welds in question.
d.
Staff Position The staff concludes that no structural deficiencies resulted from the change of welding technique, because full penetration of the root was achieved in both cases.
None.
19.
Exit Interview Mr. M. Tressler, was informed of the results of these inspections by telephone on July 17, 198 t f