ML19330B664

Final Deficiency Rept for Significant Const Deficiency 15 Re Procedural & Performance Deficiencies in Ultrasonic Testing of Structural Welds.Caused by Mfg Procedural & Ebasco ASME Insp Deficiencies.Mfg & Insp Responsibilities Established

ML19330B664
Person / Time
Site:	Waterford
Issue date:	07/31/1980
From:	Aswell D LOUISIANA POWER & LIGHT CO.
To:
References
LPL-14739, NUDOCS 8008050348
Download: ML19330B664 (9)

Text

,

I OUISi ANA u2 csuacNee srassr .

P OW E R & LIG H T! a scx scc 8

• NEW OALEANS LCuts:ANA 70174 + (504) 368-2345 YlCOLE 3CU%

UT;L.T E5 5YSTEM 2

• Jul7 31, 1980 w, %.% 'wu." ,

LPL 14739 0-3-A35.07.15 Mr. K. V. Seyfrit, Director, Region IV U. S. Nuclear Regulatory Coc=ission Office of Inspection & Inforcecent 611 Ryan Plaza Drive, Suite 1000 Arlington, Texas 76012

Subject:

Waterford SES Unit No. 3 Docket No. 50-382 Final Report of Significant Construction Deficiency No. 15 " Procedural and Perfor:ance Deficiencies in Ultrasonic Testing of Structural Welds Performed by Industrial Engineering '4orks"

Reference:

LPSL Letter LPL 14395 dated June 27, 1980

Dear Mr. Seyfrit:

In accordance with the requirements of 10CFR50.55(e), we are hersby providing two copies of the Final D.eport of Significant. Construction Deficiency No. 15, " Procedural and Perfor=ance Deficiencies in Ultra-sonic Testing of Structural Welds Performed by Industrial Engineering

~4orks".

Very truly yours, D. L. Aswell DLA/LL3/grf Enclosure ,

I cc: 1) Director Office of Inspection and Enforcement U. S. Nuclear Regulatory Co==ission Washington, D.C. 20555 O/7 I i

(with 15 copies of report) $ l

2) Direcror, Office of Management Information and Progra: Control //

U. S. Nuclear Regulatory Co= mission Washington, D.C. 20555 s (with 1 copy of report) 80080503[/g

I l

1 LOUISIANA POWER & LIGHT COMPANY WATERFORD SES UNIT No. 3  ;

Final Report of Significant Construe:1on Deficiency No. 15 Frocedural and Performance Deficiencies in Ultrasonic Testing of Structural Welds Performed By Industrial Engineering Works 1

Reviewed by

, A , ,

Crnich - Site Manager 7h O/ [O

' ' Date /

Reviewed by # -w 8# D R. J7 MidIiiser - Proj ect Superintendent Date Reviewed by ,

/M [- 88 " ((

J. H' art - Project LicenAng Engineer Date Reviewed by O i R. A. Hartnett - Q. A. Site Supervisor Date July 28, 1980

FINAL REPORT OF SIGNIFICANT CONSTRUCTION DEFICIENCY NO 15

. PROCEDURAL AND PERFORMANCE DEFICIENCIES IN ULTRASONIC. TESTING OF STRUCTURAL WELDS PERFO.tfED 3Y INDUSTRIAL ENGINEERING WORKS INTRODUCTION This report is sub=itted pursuant to 10CFR50.55(e). It describes procedural and performance deficiencies in Ultrasonic Testing of structural welds by Industrial Engineering Works-(IEW) for Louisiana Power & Light Company, Water-ford Steam Electric Station Unit No. 3. The problem was identified on November 26, 1979, by Ebasco's Corporate ASNT Certified Level III Examiner during a routine surveillance visit of IEW's facility. In addition, this report in-cludes all corrective =easures .taken by Ebasco and IEW to correct deficiencies and preclude recurrence.

DESCRIPTION The extent of the deficiencies which are identified on Nonconfor=ance Reports in accordance with Parts 1 and 3 of Ebasco's Nuclear Quality Assurance Program Manual (ETR-1001) is as contained herein.

I. PERFORMANCE DEFICIENCIES Insufficient Ultrasonic Examination a .' Affects Purchase Order NY 403573 and NY 403611.

In accordance with the ASME Section III Subsection NF Boiler and Pressure Vessel Code, Ultrasonic Examination is required of the-entire weld. Furthermore, Paragraph T535.2D.1 of the .

2 code specifies'that "The search unit shall be placed on the contact surface with the beam aimed at about 90 degrees to the veld and manipulated laterally and longitudinally so the

, Ultrasonic beam passes through all of the weld metal in two different approaches of the beam to the reflector."

b. Affects Purchase Orders NY 403593, NY 403573, and NY.403611.

In accordance with AWS Dl.1,100* volumetric inspection is required when specified by the design engineer.

The IEW examination was performed in one direction only using a single search unit from the web surface (s) only. This resulted in not achiev-ing 100 volumetric coverage. The attached sketch of a typical "T" well representing the worst condition illustrates the area where the beam did not pass through the 'reld in rwo directions. The portion of the weld ; that did not receive complete examination was approximately 20".

w

II'. IMPROPER CALIBRATION Potentially all , Purchase Orders with IEW affected.

The Ebasco Corpora:e ASNT certified Level III Examiner during his November 26, 1979, surveillance visit at IEW's facili:y, noted that IEW was not calibrating its equip =ent in strict accordance with Para.

6.18 of AWS Dl.1 edition or 1:s1own procedure UT-120, Revision 1.

The Ebasco Level III Exaniner required IEW~to calibra:e equip =en:

.as specified by its procedure in order for him to verify compliance with AWS Dl.1. The ' equip =ent was calibrated by IEW with Ebasco's Level III Exa=iner present. It was deter =ined : hat when equipment calibration was perfor=ed as specified by its procedure, UT-120, Revision 1, IEW complied with or exceeded applicable AWS Dl.1 re-quirements, :hereby eli=inating :he following:

(a) Insufficien: Ul: asonic Tes:ing Sensitivity (b) Equipment not meeting performance requirements (c) Incorrect range and location calibration III. ?RCCEDURAL ' DEFICIENCIES .

A) Affects NY Purchase Orders 003573 and 403611.

Incorrec: procedure requirements in shop drawings

- The IEW Procedure csed for UT exa=ination was in accordance with ASMI See: ion III,. Subsection NF requirements in lieu of Ebasco's specification requirement of AWS Dl.1

3) Affects NY ?urchase Orders 003593 and 403611.

Incorrec: requirements for detecting and evaluation of fusion line indications Note: Item III.3 should have been discussed as a Performance De-ficiency and was incorrectly identified as a Procedural De-ficiency _in the Interin Report dated January 2,1980.

SAFETY IMPLICATION The below listed Ebasco Purchase Orders require Ultrasonic examination on Seismic I full penetration welds to ensure that the subject velds comply with the quality standard of the ASME or the AWS Code, when specified by the Design Engineer. The welds requiring Ul:rasonic examination are associated with the following' structures:

ASME Welds a) Reactor Coolant Pu=p Suppor:s NY_ Purchase Order NY 403573

-b) Reactor Coolant Pipe Stops NY Purchase Order NY 403573 '

c) Pressuricer Support NY Purchase Order.NY~403611 l i

Note: These welds were' designed in accordance with AISC requirements 1 but welded and examined _in accordance wi:h the requirements of l the ASME Code See:1on III Subsection NF.

~

l l

1

AWS Welds a) ' Pipe. Whip Restraints NY Purchase Order NY 403593 b) . Safety Injection Tank Suppor:s NY Purchase Order NY 103611 c) Fra=ing over steam' genera:or NY Purchase Order NY 403573 If corrective action had not been taken, potential weld defec:s may have existed in these Seismic I structures and may not have been detected. Such

_ potential weld defects may under certain condi: ions lead to higher than allow-able s tress levels on Lehese structures and if lef t uncorrected result in f ail-ure of the weld and possible subsequen: failure of the affected component when subjected to the dynamic events postulated as the basis for design.

. CORRECTIVE ACTION Due to the ~ findings of Ibasco's ASNT Certified Level !!! Examiner during his November 26, 1979, surveillance visit at IEW's f acility, the following steps were i= mediately taken by Ebasco's Quality Assurance Engineering Group to en-i sure that construction progress did no: preclude problem evaluation and any required correc:1ve action. All equipment on site and a: the supplier's fa-cility was placed on restricted hold in accordance wich_Section QA-III-6, Revision 1, Nonconfor=ances, and QA-III-10, Revision 1, Iden:1fication and

' Control of Items, of Ebasco's Nuclear Quali:7 Assurance ?rogram Manual. In addition, the Ebasco Project Manager, Project Superintendent, and Projec: En-gineer were. no:ifed resulting in :he establish =en of an Ebasco project team with the Project Quality Assurance Engineer acting as team coordinator. This team solicited IEW's input and in conjunction with IEW, established IEW and Ebasco discipline responsibili:ies.and determined' priori:1es for. Engineering evaluation, reinspection and possible rework of welds in :he field based on Engineering s tatus, safety i= pac: and construction schedules. The priority list and rates: progra= were based on construe: ion schedule requirements to the extent practicable, taking into consideration the final position of the structural members and the resultant access for re-examination and possible rework in place. In order to syste=atically accomplish the above, the follow-ing responsibilities and corrective action i=plementation requiremen:s were established by IEW and Ebasco.

IEW a - Initiation and disposition of nonconformance reports.

b - Preparation of NDE procedures for ul::asonic testing in the field to assure uniformity of inspection techniques. One procedure was prepared by IEW in accordance with the requiraments of ASME Section III, Subsection NF.

This procedure was deter =ined to provide acceptable examination of compo-nents which _ hai previously been partially examined in the IEW shop using either ASME or AWS criteria.

c - Qualification for ASNT certification of ultrasonic technicians (levels 1,=2, and 3) as applicable per ASME.

d - Correction of all performance deficiencies.

e - U1:rasonic examination or re-examination either at its facility or in the

-field.

J 4 y W D W

. I l

. )

f - Rework of all welds deter =ined to be unsa:isf actory. 1 g - Sub=1ttal of a, b, c and d to Ebasco for review and coc=ent.

E3ASCO The project tea = consis:ing cf =e=bers fron Engineering, Quality Assurance Engineering, Ma:erials Application, Construction and Licensing was responsible

'for the following:

a - Identifying field and shop' welds of concern (Engineering) .

b - Verifica: ion of codes used (Engineering, QA & Materials Application).

c - Weld accessibility to support UT Re-exa=ination and possible rework (Engineering, QA & Construction) .

d - A co=plete review of all IEW Shop Drawings, including identifying all af fected welds and indicating classification of each weld (Fngineering) .

e - A s:ress investiga:ica by Engineering was conducted to determine whether stress levels actually encountered in operation would require the pre-viously specified full penetra:icn welds.

- A tabulated presentation of all concerned welds for each purchase order drawing and each jcin: to indicate the classification, design criteria, and code standards (All).

3 - Preparation of his:orical data fro = the PSAR stage, :hrough design and

, specification develop =ent, and in:o the present status of the FSAR to in-clude standards involved, design classifica: ion, design cri:eria and code standards (Engineering and Licensing) .

h - Review of IEW UT procedures to AWS and ASME-NF require =ents (QA).

1 - Develop =en and =aintenance of Priority Lists (Construction, QA) .

Ebasco Engineering conducted an engineering analysis based on design criteria to deter =ine which welds actually required full penetration welding in ac-cordance with AISC, AWS 31.1 and Ebasco Standard Practices for Structural De-sign. Based on this engineering study it was deter =ined that certain welds specified as full penetration actually could have been designed as partial penetrations and as such would not have required ultrasonic examination.

3ased on this engineering evaluation, all full penetration welds which could have been designed as partial penetration welds were identified and reinspec-tion by Ultrasonics was not performed. The engineering evaluation perfor=ed was as follows:

There are a total nu=ber of one thousand two hundred four:een (1,214) weld joints associated with the reactor coolan: pu=p supports and reac:or coolan: pipe stops. One thousand sixty-six (1,066) welds were reclassified as partial penetration leaving the re=aining one hundred forty-eight (148) weld =ents as full pene: ration. The pressurizer support structure had one hundred sixty-six (166) weld joints in which all welds were reclassified as partial penetration welds. Data suppor:ing this analysis is available in Ebasco's Engineering files.

The result of' this effort was the determination tha: 148 full penetration weld =ents required re-examination by ultrasonics at the Waterford III Site.

Parallel.with and aided by the engineering analysis, on January 3, 1980,

. IEW =obilized on site and co==enced ultrasonic tes ting. Prior to start- '

ing; UT Examina: ion.Ebasco reviewed IEW NDE procedures and personnel quali-fications.

Ultrasonic examination at the site was conducted in order to re-inspect ite=s previously :ested at the IEW facili:y to ensure 100 percent volu=e:ric exa=ination of the required welds. To minimize impact on construction scheduling, initial UT was performed without removal of A=ercoat 71 and Amercoat 90 paint coatings. Since the surf ace to be scanned was coated, it was therefore necessary to develop a technique (documented by Addendum A :o procedure UT-124) for ul::asonic attenuation correction required to co=pensate for the surf ace coating.

During the week of February 25, 1980, a team of Quality Assurance Engineers from Ebasco's New York Office revisited the Waterford III Cons truction Site. The purpose of this visit was to analyze test.

results being obtained from IEW's ultrasonic examination of struc-

ural s: eel 1:e=s.

To maximize the effectiveness of this investigation, attention was

' directed in the following =ajor areas:

A. Equipment Suitability

3. Personnel Qualifications C. Procedure and Technique A. The ultrasonic equipment being used by IEW consisted of both Krautkra=er and 3ranson instruments and a variety of transducers.

Equipment certifica: ion records and linearity checks were main-tained by IEW. In addition, IEW had calibration blocks repre-senting the range of wall thickness for the welds involved.

3. Additional certification records for IEW NDE personnel were re-viewed. This review disclosed one =inor deficiency which was immediately resolved.

C. The procedure used by IEW for on-site exa=ination was UT-124, Revision 1, dated 12-8-79 supplemented by UT-124S for defect sizing. The procedure and supple =ent had been reviewed by Ebasco prior to use and it was observed that IEW was implementing the procedure properly.

D. The variables in coating condition and :hickness were such that required confidence in the results could not be obtained and the procedure was determined to be ineffective and discarded. Based on this decision, all items in the field requiring reinspection by UT examination were stripped of all coatings by =echanical means and cleaned prior to examination. Testing validation of'this con-clusion is discussed in Paragraph F. Further, Ebasco deter =ined tha:

the additional nitrasonic. testing would be performed on only those welds classified as full penetration welds to satisfy stress levels.

E. The required UT re-examinations were made with test sensitivity established in accordance with procedure UT-124, Revision 1. Both straight and angle bea= exa=inations were perfor=ed with satis-factory resul:s.

l i

l

4 F. Welds- as specified by the Design Engineer representing various join: configura ions no: previously inspected 1000 volu=etrically were re-exa=ined. This re-exa=ination resulted in no rejectable indica:icns based on signal response. However, three (3) indica-1ons were interpre:ed as " lack of fusion" and were rejected.

These rejec:able indica:icns were recorded and disposicioned as accep:able on Nonconfor:ance Repor: W3-1802-007 Add A.

On March 20, 1980, the sa=e team revisited the Waterford III Site to review the results of all IEW field activities related to UT reinspection and to assist si:e personnel in the close-out of IEW's site activities.

G. During ultrasonic re-exa=ination of eight_(S) stops known as '_'D" S: ops ,

it was de:er=ined by IER.tha: the four;(1) welHs at :he top' of ^each of these eigh: "D" Stops had been =ade as par:ial penetration velds in lieu of full penetra: ion welds. This problem was docu=ented on Nonconfor=ance Report W3-lS02-Oli dated March 19, 1980. Exploratory ultrasonic testing was perfor:ed on these welds to establish the depth of weld penetration for the welds. An engineering evaluation of :his infor:a: ion concluded that based on :he stress levels in :hese welds, partial penetration welds were acceptable. At the request of a Nuclear Regulatory Co==ission Inspec:ce during the week of April 7, 1960, these "D" Stops were sand blasted and reinspected. During this reinspection, it was noted that s=all areas of non-fusion or incomplete weld existed a: each end of the seven-inch welds at the top of the "D" Stop. This proble= was then docu=ented on a Nonconfor=ance Report W3-1992. The "D" Stops were returned to the =anufac:urer's facility for repair and rework. As a result of this proble=, IEW reviewed their inspection procedures with IEW inspectors and established additional weld inspec-tion criteria to preclude recurrence of this type of proble=. Ibasco i Vendor Mechanical / Welding' Inspectors were provided' copies.cf the J l entire documentation package relating to the "D" Stop proble=s as a training aid .and guidance for their future inspec: ions :o preclude re-currence of this type of proble=. The "D" Stops are planned for return to the Waterford III Site on August 4, 1980.

H. The items listed in the DESCRIPTION paragraph of the Interi= Report dated January 2, 1980, were further evaluated during the progress of this effort. I:e=s listed in paragraphs II and III.3 were deter =ined to be isolated occurrences not contributing to the overall proble=.

CONCLUSION Since all AWS and certain NT welds were reclassified by Engineering as partial pene: rations based on low stress levels , re-exa=ination by UT was not required.

This determination was further reinforced by the fact that the UT inspection (approxi=ately- 80% of weld in lieu of 1000) ini:ially performed did not uncover discontinui:ies .which exceeded code acceptable li=its.

As a result of :he analysis and testing progra= described herein, it has been deter =ined that the original proble= definition should have been li=ited to proper'i=ple=entation of procedures to assure that 100 percen: volumetric

- examination of all required welds was acco=plished. As stated above, all IEW

- welds requiring: full penetration'to satisfy stress levels were re-exanined at the Waterford III Site and found acceptable.

TTPICAL "T" WELD JOINT ULTRASONIC EXAMINATION COVERAGE l

/

s -

~

Weld & HAZ Examined with Ultrasonic Beam Passing

\NN through in two directicas E Weld & HAZ Exa=ined with Ultrasonic 3eam Passing through in one direction HAZ - Heat Affected Area