ML20039B357
| ML20039B357 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 12/11/1981 |
| From: | Carey J DUQUESNE LIGHT CO. |
| To: | Varga S Office of Nuclear Reactor Regulation |
| References | |
| IAL-80-18, NUDOCS 8112220571 | |
| Download: ML20039B357 (13) | |
Text
r i
I
%-?* ?b me s.m
> 7 W, "G
Nf*
(412) 4564or I
us s,.m ne~.
['s""*
December 1 N
&\\
F-[- _
\\/,
Director of Nuclear Reactor Regulation
\\9
' 'f United States Nuclear Regulatory Commission Attn:
Mr. Steven A. Varga, Chief
,: \\h
~,
Operating Reactors Branch No. 1
{
D{g2 2 n31% Q>
Division of Licensing 10 k '. 3, (._
Washington, DC 20555
\\(
a r;
Reference:
Beaver Valley Power Station, Unit No. 1 y
Docket No. 50-334, License No. DPR-66 N.
i v,
Followup Report - Immediate Action Letter 80 GravM24nd7 Gentlemen:
The enclosed report was prepared to document the Graver Tank inspections and repairs that were initiated in response to the subject letter.
We have completed these inspections and repaired the penetrations and seams that were considered deficient in accordance with Engineering and API 650 require-ments.
This problem was originallly reported to you with the issuance of LER 80-025/0IT dated April 30, 1980, and LER 80-037/0IT dated Jm.e 20, 1980. A subsequent investigation by the Office of Inspection and Enforcement was documented by Report Number 50-334/80-17 and forwarded to us September 2,1981.
This information is being forwarded to close out this issue, should you have any further questions in this regard, please contact my office.
Very truly yours, J. J. Carey Vice President, Nuclear Enclosure cc:
Mr. D. A. Beckman, Resident Inspector OO U. S. Nuclear Regulatory Commission 5
Beaver Valley Power Station Shippingport, PA 15077 U. S. Nuclear Regulatory Commission c/o Document Management tiranch Washington, DC 20555 9112220571 011211 PDR ADOCK 05000334 S
Paga 1
)
I I.
INTROOUCTION This report prevides the details, corrective action and final conclusions resulting from investigations perforne! subsequent to the issuance of LIR-80-025/OlT dated April 30, 1980 and LER-80-037/OlT dated June 20, 1980.
II.
FI5LD FARRICATED S'1tRAGE TANKS These tanks were fabricated and erected to the requirements of Engineering Specification BVS-183, Purchase Order BVC-198, which invoked API-650. The prime QC/QA functions were the responsibility of Graver. Stone i Webster, as the agent for the Duquesne Light Company, was responsible for surveillance during the erection / fabrication of the tanks.
A.
Refueline Water Storace Tank - OS-TK-1 Subsequent to the initial finding of lack of fusion in one tank nozzle /shell penetration, other noz:le/shell penetration welds were examined and found not to be in accordance with the approved Engineering and Graver drawings. With the exception of two (2) roof penetrations, all other penetrations (total 16) were excavated and rewelded to meet the requirements of a full penetration weld as required by the Engineers. Since the as-found weld condition did not conform to the Engineering requirements, several calculations were performed.
These calculations confirmed that, prior to repair, the 12" gnetratior, examined would have been' adequate under normal operating conditions plus seismic loadings i= posed by the connecting piping. Calculatiens were also performed for specified roof penetrations and, although not in accordance with the drawings, these penetrations are capable of withstanding the design conditions without repair.
A copy of these calculations has been submitted to the USNR0 Resident Inspector.
During the weld repair of the tank nozzle /shell penetrations, a linear indication was cbserved in a vertical seam weld made by Graver. This indication, lack of fusion, was remeved by grinding, rewelded and inspected for acceptance.
The Graver seam. weld radiographs were reviewed and reinterpreted by a DLC Level III examiner. Based on this review, three (3) radiographs previously accepted by Graver, were rejected.
Additional radiography was performed to verify several previously taken Graver radiographs and cther rande locations were also radiographed.
These additional radiographs were acceptable; however, the weld areas shown on the three (3) Graver radiographs that were rejected by DLC could not be re-radiographed as they were below the concrete i
line and, therefore, not accessible for radiography. These three (3) areas were repaired from inside the tank by surface grinding and were determined to be acceptable by ultrasonic inspection.
i All necessary weld repairs have been satisfacterily completed en QS-TK-1 and the tank has been returned to service.
i I
/
Paga 2 B.
Demineralired Water Storace Tank - W -TK-lO As a result of the as-found condition of the nozzs/shell penetrations welds on QS-TK-1, a drawing review was performed to ascertain if the same weld joint configuration, full penetration, was specified for the other Category I tanks erected by Graver.
This review showed that the full penetration weld requirements and details 'were typical for all Category I tanks listed in the Engineering specification and shown on the Graver drawings.
Information radiography performed on a segment of three (3) randomly selected penetrations in WT-TK-10 confirmed that lack of fusion existed and that repairs were necessary.
Engineering instructions were issued and, with the exception of the two (2) roof penetrations, all remaining ten (10) penetrations were ground and rewelded to full penetration welds as required by the Engineers. An analysis of the safety implications involving the roof penetrations was performed which confirmed that no repairs were necessary. A copy of this analysis has been submitted to the USNRC Resident Inspector.
Concurrent with the weld repairs of the tank nozzle /shell penetrations,the Graver radiographs for the tank seas welds were also reviewed and. re-interpreted by a DLC Level III examiner.
This review disclosed that three (3) radiographs, originally accepted by Graver, contained code rejectable defects and were rejected by the DLC examiner. Areas previously radjographed by Graver were re-radiographed in addition to other randomly selected areas.
This radiography confirmed the DLC re-interpretation of the Graver radiographs and also showed that additional code rejectable '.r.lications were present in two horizontal seams.
A second set of additional radiographs was taken to determine the extent of the unacceptable condition in the two horizontal seams.
These additional radiographs also contained rejectable indications and it was apparent that both entire seams were unacceptable. Both horizontal seam welds, made by Graver, were removed, repaired and meet the code requirements.
All weld repairs on WT-TK-10 have been completed and examined by the non-destructive techniques required by API-650, 5th Edition, i
1973 and the tank has been returned to service.
1
(
C.
Review of OS-TK-1 and WT-TK-10 Documents As a result of the as-found conditions of the no::le/ tank shell penetrations, the following Engineering and Graver documents were reviewed:
Erection Control Sheets (Travelers)
Drawings Weld rod issue slips Engineering specification Graver Q.A. Manual l
l l
peg 2 3 Thi; review disclosed that the as-found condition was not always in agreement with the Engineering requirements.
The Travelers were signed indicating acceptance of inspection points; hewever, the as-found condition on nozzle /shell penetrations welds leads to the conclusion that certain inspection points had not been properly verified. Travelers were signed off en one day and had the appearance of being signed af ter the work had been perfor ed rather than while the werk was in process. As such, it appeared that the signed Graver Travelers were not representative of work actually done er inspections actually performed.
D.
Primarv Gra'de Water Sterace Tanks - BR-TK-6A and 6B Re-interpretation of the Graver radiographs, by CLC, resulted in rejecting one (1) radiograph of a sea = weld, in each tank, which had previously been accepted by Graver. Engineering instructions were issued to re-radiograph the rejected weld area in sich tank and to take one (1) infor ation radiograph of the tank noz:le/shall penetrations in each tank.
1.
ER-TK-6A The results of th-s re-radiography showed that the seam weld had apparently been ground and the defect had been re=cved.
The re-radiograph was acceptable; therefore, no repairs were required for any seam weld on BR-TK-6A.
Of the ten (10) penetrations in the tank, nine (9) informatien radiographs were taken.
The vent penetration in the outer roof was inaccessible because of the floating roof inside the tank. The radiographs showed that five (5) penetratiens were acceptable and that four (4) require investigatien and repair.
These four (4) no::les have been repaired in acccrdance with instructions issued by the Engineers, and the tank meets the code and engineering recuirements.'
2.
BR-TK-6B Engineering instructions were issued for perfor=ing infor=ation radiography en all nc==le/shell penetraticas and for re-radiographing a previously accepted Graver radiograph which was subsequently rejected by OLC.
Four penetration welds (reference penetrations 41, N2, N5, N6) and one sean weld had rejectable defects and were repaired.
E.
Boron Recovery Test Tanks 1.
BR-TK-2A Six penetration welds (reference pene.trations NIA, NIB, N2, N3, C1, C3) and three seam welds were found to have rejectable defects and required repairs to be made. These repairs have been made and the tank was returned to service.
L
Pr.gs 4 2.
BR-TK-23 Five penetration welds (reference penetrations NIA, NIB, N3, C1, C2) and one seam well were found to have rejectable defects. These repairs have been made and the tank returned to service'.
17.
Ccolant Recovery Tanks 1.
BR-TK-4A Five penetration and one seam weld had rejectable defects and were repaired.
2.
BR-TK-4B One seam weld was required to be repaired on this tank.
III. REACTOR NAINMDIT LINER A.
General The shop fabrication and field erection of the Reac*er Centain=ent steel plate liner was performed in acco: dance with Engineering Specificatica BVS-136, Purchase Order BVC-65.
All shcp and field welding was specified to be randers radiographed in accordance with Paragraph UW-52 of Section VIII of the ASME Code for Pressure vessels. The ASME III Code was used as a guide fe:
the selection of =aterials and fabricatien as described in the FSAR.
As such, the Liner is not ASME Code sta= ped.
The prime QC/QA respensibilities were Graver tank. Stene & Webster, as the agent fer the Duquesne Light C :::pany, was responsible for surv.tillance and auditing during erection.
A review was performed cf the available Graver Tank C0=pany ccnstructicn and inspectico records. The purpose of this review was to determine if these records c=ntained i::egularities similar to those previously encountered and described during review of the Graver records for the Refueling Wate: Storage Tank (QS-TK-1) and the Demineralized Water Storage Tank (W:'-TK-10) a-described in Mr. Grier's letter to Mr. Dunn dated June 26, 1980.
The review of the contair:sent liner records was conducted by a
l Page 9 team of personnel. representing DLC-QA, the DLC On-Site Engineering Group and S&W.
The foncwing records / documents were reviewed:
Engineering Specificatica for Shop Fabrication and Field Erection of Reactor Contain=ent Steel Plate Liner, BVS-136, Purchase Order No. BVC-65.
ASME Boiler and Pressure vessel Code, Pressure vessels,
Division I Section VIII dated 1968 through and including Su:me: Addenda 1970.
Graver Q.A. Manual Grave: Erection Drawings Graver
- Travelers (Erection Centrol Sheets)
Grave: ICE Procedures and Records Graver Spct weld X-ray MAPS Grave: Spot radiographs (Total 447)
Graver Radiographic Interpretation and Techniques Sheets.
Construction photographs B.
Anomalies Noted In Grave Records The types of i :egularities detected in the tank erection record review were ret de tected in the record review of the reacter containment liner; however, other an==alies were found.
These anomalies are categorized and described in the following pa:agraphs.
1.
Review of the Graver Radiocrachs All available Graver Spot radiographs (447) were re-interpre ted, by two (2)
Level III examine:s to the requirements cf ASME Sec tion VIII, Paragraph UW-52.
The results of this review have been conselidated into three (3) categories:
a.
Accarent Duclication of Radiocrachs (1)
Three (3) sets (a set consists of 2 radiographs to the same weld area) of radiographs, appear to be the same with the exception cf the veld identificatien number.
Identifying characteristics such as the veld surface, weld contours, marks in the plate materials, slag and porosity in the weld appeared to be identical en both radiographs.
This left little doubt that the radiographs were duplicated.
It appears that cer tain radiographs were missed
P 9a 6 during erection.
utplicate films were made of subsequent weld jointe and substituted for the radiograph which had not been tak en.
The substitui. tion (or duplication) was apparently achieved by taking a second radiograph on the same date at the same location. The equipment set-up and film location were unchanged, only the weld identification number was changed.
It did not appear to the reviewers that the film substitutien was done to subvert the process control, but was an attempt, by the Graver radiographic crew, to icg a sufficient number of radiographs to meet the spot radiography regairements. A review was made of the Travelers to identify the dates the welds in the affected area were made and the dates on which the radiographs were taken.
In addition, construction photographs were' reviewed to determine the erection segaence.
Reducing this data leads to the postulation that the radiographer determined he had missed a radiograph at a lower elevation which apparently could not be taken because of rework regaired to remove the leak chase channel which had been instaHed and a radiograph at the higher elevation was substituted for the icwer elevation. The apparent duplicate radiographs are:
H7P3T1 and H8P8 H8P4 and H9PS R9V10R and R10V12R (2)
An apparent duplication of radiographs, identified Item @ on Attachment 1,
was also noted en as radiograph HilP1, but the scenario of its cause is significantly diff erent than described above. The original radiograph, HilP1 was taken -on June 3, 1971 and rejected by Graver.
The radiograph of the weld repair, HilPlR was taken on June 23, 1971 and was accepted by G raver.
In addition, as required by code, two (2) Tracer radiographs were taken cr. June 23, 1971 and were rejected by Graver. Subsegaent record review identified that the rejected weld seam was cut-out and rewelded.
The repaired area was qualified by radiograph HHPlR1 taken on July 6,
1971 and accepted by G raver.
Rad iographs H11PlR1 and HHPlR are duplicates.
The duplication indicates that the repair excluded the seam area originaHy repaired (H HPlR) and the repa ir was qualified by a radiograph (EllPlR1) of a known acceptable area.
l f
l
i i
Pigs 7 These radiographs (total of 20 films) were of; :ially tranr:aitted and given to the NRC Region I InvestigatOfr Mr. R. K. Chistopher on July 3, 1980.
See Attachment 1 for the location of these radiographs.
b.
Repair of Weld Seams Two (2) sets of radiographs contained film of weld areas wbtch did not appear to be representative of a repaired weld.
The radiographs of repaired welds were compared to the original radiographs. The general appearance of the repair radiographs was similar to the original radiograph which led the reviewers to question whether a repair had actually been made since a weld which has been repaired normally exhibits a weld seam that is wider than the original weld and has a different appearance.
Subsequent investigaticn of records available showed that the weld seam had been cut cu t and rewelded.
Consequently, the weld repair radiograph represents a new weld which should be similar in width to the seam weld observed on the original radiograph.
Disagreement with Graver Radiocraphic Inter =retations c.
Of the 447 radiographs reviewed and re-interpreted by DLC, the examiners identified only four (4) cases of disagreement with the original film interpretation made by Graver.
This disagreement is not surpricing when consideration is given to the number of radiographs involved, the qualifications and experience of the DLC reviewing team.
The four cases of disagreement, radiographs RlV8, KRlP4, DV19T1 and E4P6, are considered minor in nature and have been evaluated as acceptable by the engineers and the evaluatica is available.
d.
Indeoendent Review r.n independent review of the preceeding three (3) anomalies was performed and confirmed by a Level III examiner frem the Hartferd Steam Boiler Inspection and Insurance Non-Destructive Test Engineering Division.
2.
Correlation of Graver Radiocraphs and Associated Records a.
An investiga tion was conducted to relate the 447 available radiographs and the Graver Records of Radiographs with the Graver Spo X-Ray Maps. The areas examined were the cylindrical shell, rings 1 through 12, and the 6 me assemblies.
The list of reviewed radiographs was cccpared with the Record cf Radiographs and the Spot X-Ray Maps.
This resulted in the
Paga 8 identification of:
445 radiegraphs per. the Record of Radiographs having been tak en and interpreted; 447 radiographs were reviewed by the DLC review team: 372 radiographs reviewed were traceable to both the Record of Radiographs 'and. the Spct X-Ray Maps.
the Record of Radiographs indicate six Additionally, (6) weld areas which were rejected and subsequently repaired; however, no records exist which indicate that the required tracer radiographs were taken, b.
It was observed that there were numerous changes in the radiograph numbering systems used by Graver en the radiographs, on the Record of Radiographs and on the Spot X-Ray Maps.
These differences made 100%
correlation of an information impossible, e.g: Vertical welds for the do=e were identified with the same numbers for two diffe-ing areas, the only difference being the welder's symOc,1 noted on the Spot X-Ray Map.
To determine which radiograph pertained to a specific locatien, it was necessary to obtain the welder's symbol from each radiograph and then correlate this radiograph to the Spot X-Ray Map. The results of this correlation are surmarired as fonews:
95% of the radiographs matched the Spot X-Ray Map.
83% of the radiographs were traceable to both the Spot X-Ray Maps and the Record of Radiographs.
Ten out of the eleven radiographs for Ring 1 of the cylindrical sheH are missing.
This is noted cn a drawing contained in a Graver documentation package.
Errors were roted in identification between radiographic film and locations noted on the Spot X-Ray Maps.
3.
Construction and Insoection Record Ccucarison a.
Data Cemoarison A comparison of similar data, i.e. dates, procedures, personnel, of Graver construction and inspectica records was made.
An attempt was also made to correlate I
informatico fren this compa;ison with the results of the 447 radiographs re-interpreted as previously described.
A total of twenty-two (22) areas were reviewed: Ten (104 from the dome portion of the liner and twelve (12) from the cylindrical shell portion.
b.
Document Cencarison The EIection Centrol Sheets (Travelers) and Record of Radiographs covered large segments of work (generally an entire ring). Individual records covered days or weeks
<y.
u.3
Page 9 of an activity. It was not possible to pinpoint exactly when an activity occurred. Based upon the results of the 22 areas reviewed, the dates recorded for activities were generauy censistent.
The dates on repair radiographs were exceptions since the Travelers recorded the date of the original radiograph rather than the date of the repair radiograph.
c.
Observations The Graver " Record of Radiographs" identified the following:
(1)
R. Welch, of Graver, was the machine operator and processor for the entire shell.
(2)
C. W. Funkhouser, of Graver, was. the interpret =
and :eviewer fc: the entire shell.
(3)
A. Hollid, of Stone & Webster, signed Record of.
Rads > graph forms for the cylindrical shell. There were no S&W signatures fer the dome portions of the liner.
(4)
The Authorized Nuclear Inspector up through Ring 8 was J. Gall of Factory Mutual.
(5)
The Authorized Nuclear Inspector f:cm Ring 9 through the reminder of the shell liner and deme was an inspector, using the symbel A1, f:cm Lu=berman's Insurance Cmpany.
1 (6)
On two (2) berizental dome seams, the machine operator was L.
Bruder of Graver.
For the remaining dome seams the machine cperator and processor was B. Welch of Graver.
(7)
Interpretation and review of the deme radiographs changed with different joints; however, the interpretation and review was performed by either B. Welch c: C. W. Funkhouser of Graver.
(8)
The results of the data suggest that, except for a change in the ANI organi:ations around the time of the apparent radiographic duplication,
the personnel involved in the erection of the cylindical shell region were the same.
i
Paga 10 d.
Code Aeolicability The review of Graver cbcu=ents was initially inconclusive relative to the acceptance criteria applied by Graver for interpre tation of radiographs.
The Engineering Specificatien required random radiography of seam welds to AS!'E Sectica VIII, Paragraph W-52.
The review of Graver records identified the following:
(1)
The record copy of the Graver " Quality Assurance and Control Manual for Shop Fabrication and Field Erection of Reactor Contain. ment Steel Plate Liner" (copy assigned to C.
W.
Funkhouser of Graver) identified Radiographic Examination Procedure No.
RT-2, Rev.
1, dated 9/12/72, in the index.
Another cbcument in the Manual dated 1/22/70 (which appears to be a change authori:atien to Graver drawing B30799-4), states that "All shop and field welding shall be random radiographed in accordance with paragraph W-52 (Spot Examination of welded joints) of Section VIII of the ASME Code for Pressure vessels and Graver's Radiographic Examination Procedure No.
RT-1."
The procedure contained in the Manual is " Exam! nation Precedure No. 1 Radiographic Examination for ASME Section III, Class A & B Construction."
Ecwever, the procedure in the Manual contains two (2) different -
interpretation sheets.
One sheet is " Record of Radiographs ASME III, Class A & B Vessels" and certifies that the radiographs meet N624 of ASME III, Appendix IX of Section III and Radiographic Examination Procedure No.
1.
A second sheet is
" Record of Radiographs for use with Procedure RT-2" and certifies that the requirements of Paragraph W-51 of the 1968 Edition o.f the ASME Boiler and Press ure vessel Code,Section VIII, Division I.
(2)
A review of the erection control sheets showed that the words "ASME VIII Rr-2" was manually changed on erection control sheets for vertical seam welds to read "ASME III, RT-1."
The radiographic interpretation sheets are consistent with the manual changes. The following data has been extracted frem the Travelers i
.. ~. - - -
o Paga 11 Ring Vertical Hori:cntal Note Location Welds Welds 1
VIII,RT2 VIII, RT2 2-12 III, RTl VIII, RT2 Cl III, RTl III, ICl III, RTl Roof Subassemblies Connect Roof VIII, RT2 VIII, RT2 Subassemblies III, RTl Install Top Plate Manua11y chang".d f:cm "ASMI VIII, RT-2" to "ASME III, RT-1" for vertical seam wel'ds.
The majority of the " Record of Radiograph" forms certify. that radiographs meet N624 of ASME Section III, Appendix IX of Sec tion III and Radiog raphic Examination Procedure No.
1.
A total of four (4)
Record of Radiograph ferns certify to ASME VIII, but identify paragraph W-51 which deals with 100%
radiography rather than W-52 for spe: radiography.
(3)
The Taality of the Graver radiographs was judged satisfactery.
(4)
Revision 4 of the Enginees ing Specificatien BVS-136 regaires radiography to be performed and evaluated in accordance with paragraph W-52 of ASME VIII, Division 1.
This was achieved by Graver using their RT-2 p:ccedure, gaalified to W-51 of ASME VIII, Division 1 for the hori: ental seams and their RT-1 procedure, gaalified to ASME III, Division 1.
These two procedures resulted in radiographic film that meets or exceeds the regairements of W-52 of ASME VIII, Division 1.
The Record of Radiographs for these p:ocedures contain statements of certification to ASME III and/or W-51 which are incorrect and by the specificaticn, invalid. The DLC re-intepretations confirmed that the Graver evaluation and acceptance was in accordance with W-52 of ASME VIII, Division 1 and satisfied the Engineering Specification.
C.
Evaluations Performed To Resolve Liner Record Anomalies
- 1. The applicable code per the Engineering Specification at the time of liner e:ection, ASME Section VIII, Paragraph W-52 states in part that " spot radiography in accordance with these rules will not insure a fabrication product of predetermined gaality thoughout.
It must be realized that an accepted vessel under these spot radiography rules may still contain defects which might be discicsed on further
Pige 12 examination." Spot radiography is used as a means to prove the welding process, not the-quality of the vessel. The code also acknowledges and accepts weld defects as stated in W-52 Paragragh (d) evaluation and tests, subparagraph (a) which states in part that "The defective velding disclosed by the first of the three radiographs may be removed and the area repaired by welding, or it caybe allowed to remain in the weld joint, at the discretion of the inspector."
- 2. Regardless of the anc=alles of E'te applicable code as relating to the Graver Record of Radiographs, all 447 Graver radiographs were re-interpreted to W-52 cf section VIII by two (2) Lovel III examiners, and with the exception of four (4) radiographs, described in paragraph
- III, B 1 c,
the remainder of the radiographs were determined to be acceptable and meet the code requirements.
- 3. The integrity of the seas welds has also been preven by ncn-destructive examination en.100% cf the weld surfaces during erection.
- 4. Halogen leak rate testing on 100% of the sea: welds was perfor=ed during erection after leak chase channels were installed.
- 5. The integrity of the Reactor Containment Lina: has also been de=enstrated by perfer ance el an overpresurization conducted pric:
to initial cperatien and has. subsequently been periodically re-verified by the performance cf leak rate tests in aceordance with 100FR50, Appendix J.
t L