ML20005G878

From kanterella
Jump to navigation Jump to search
Insp Rept 999000519/89-01 on 891023-26.No Noncompliances Noted.Major Areas Inspected:Followup to NRC Concern Re Piping Sys at DC Cook & Prairie Island Exceeding SAR Design Limits Due to Copes-Vulcan Valve Weight Problem
ML20005G878
Person / Time
Issue date: 12/27/1989
From: Cwalina G, Petrosino J
Office of Nuclear Reactor Regulation
To:
Shared Package
ML20005G876 List:
References
REF-QA-99900519 NUDOCS 9001230234
Download: ML20005G878 (11)


Text

.

ORGANIZATION:

Bechtel Corporation i

i I

Gaithersburg, Maryland j

i REPORT INSPECTION INSPECTION 1

NO.: 99900519/89-01 DATE: October 23-26, 1989 ON-SITE HOUR $:

56 2

CORRESPONDENCE ADDRESS:

Mr. R. H. Stone, Senior Vice President Gaithersburg Regional Office I

Bechtel Corporation 15740 Shady Grove Road Gaithersburg, Maryland 20877-1454 ORGANIZATIONAL CONTACT: Mr. Sidney A. Bernsen, QA Manager 4

TELEPHONE NUMBER:

(301)258-3776 NUCLEAR INDUSTRY ACTIYlTY: Provides design and consulting services associated with basic components to various nuclear utilities.

i h

ASSIGNED INSPECTOR:

) <-

V-Af 78 Jospph4. P(tttsino, Reactive Inspection Section e

jfo.1 (Rp-1)

OTHERINSPECTOR(S):

T. L. Tinkel, Sonalysts Incorporated

?!M APPROVED BY:

/2 G. C. cwalina, Acting chief RI5-1, Vendor Inspection Branch E i

?

l INS"ECTION BASES AND SCOPE:

A.

BASES: Appendix B to 10 CFR Part 50 and 10 CFR Part 21.

B.

SCOPE: This inspection was conducted to follow-up an NRC concern regarding piping systems at D.C. Cook and Prairie Island that were found to exceed their Safety Analysis Report design limits due in part to a Copes-Vulcan valve weight and center of gravity problem which Bechtel-Gaithersburg was aware of in 1979.

l PLANT SITE APPLICABILITY: All sites.

9001230234 900117 PDR QA999 EECBECHT 99900519 FDC m

__-___-,w.

w

.. me..

.._...,.....w,.,

_,.--m-

--.,..,c,,

w

ORGANIZATION:

BECHTEL CORFORATION

],

GAITHERSBURG, MARYLAND i

l REPORT INSPECTION I

NO : 99900519/89-01 RESULTS:

PAGE 2 of 11 i

A.

VIOLATIONS:

None B.

NONCONFORMANCES:

j l

None C.

UNRESOLVED ITEMS:

)

1.

Processing of Generic Deficiencies AreviewofBechtelEngineeringDepartmentProcedure(EDP)-4.74,

" Generic Deficiency Processing," revision 5, dated November 16, 1987 was performed. The scope and purpose of EDP-4.74 is:

to define the method for identification, review, evaluation, and disposition of generic design and e of Problem Investigation Requests (quipment deficiencies by means PIRs). The scope of such review is limited only to incoming problem documents."

Generally, the procedure covers the method for engineering review and evaluation of generic problems detected on nuclear power projects. PIRs are also used to obtain problem resolution for computer program errors. The procedure states that incoming documents can be generated by any of the Bechtel nuclear projects and can be from sources external to Bechtel (e.g. problem notifi-cationlettersfromsuppliers). The Bechtel QA Management stated to the NRC Inspector that EDP-4.74 is not one of the procedures they have adopted to implement 10 CFR Part 21.

The NRC inspectors reviewed several of the PIR issues (discussed below and in Section E.2), and are concerned that deviations that are potentially reportable pursuant to Part 21 are being processed by Bechtel using this procedure and not being evaluated in accordance with the intent of 10 CFR Part 21.

Additionally, EDP-4.74 indicates that it controls computer codes and other problems found during Bechtel quality-related work ac-tivities for nuclear power plants.

Since Bechtel management does not consider EDP-4.74 to be one of its procedures to implement Part 21, the processing of " generic deficiencies" under EDP-4.74, may circumvent the provisions of 10 CFR Part 21. The following are examples of two " generic deficiencies" that were processed under EDP-4.74 and the Bechtel Project responses associated with each example:

,-r r

- - + - - - - -

e

., =

,.-w.

---,,.....-m,,-.--v-=


,r-w--m------mm.m.

w.--=------==

wen---Cwr-

i ORGANIZATION: BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 3 of 11 a.

Bechtel was aware of a problem which questioned the integrity of all four CVI Corporation containment cooling units installed at Vogtle nuclear power plant Units 1 and 2 2

due to high stress on the nozzles. This issue is identified on PIR G87-07-MQ, dated February 11, 1987. Bechtel internally distributed the problem description to the cognizant Bechtel project engineers in accordance with EDP-4.74 The responses received from the project engineers appear below.

(1) Point Beach Project

" Inactive Job. No action taken."

(2) Zion, Quad Cities, Dresden "Bechtel does not have any design responsibility for containment cooling units on any current Ceco nuclear plants. Therefore, no action will be taken."

(3)

Callaway

"...several p[roblem investigation reports (PIRs)issuedtoSNUPPS Callaway and Wolf Creek] proj-ect have not been processed yet because Union Electric has previously objected to using job hours, unless the task has been specifically assigned by them. We

[Bechtel]haverequestedUnionElectric'sapproval...to review these and other PIRs for applicability to the Callaway plant. Until UE's authorization to proceed is received, we will continue to hold the existing and all l

future PIRs..."

l (4) Palo Verde, South Texas Project, San Onofre - Bechtel's PIR response was incomplete for these three projects.

The inspectors were not able to determine what action Bechtel had taken.

b.

Bechtel was aware of a problem which indicated that the qualified life of some Skinner solenoid valves was based on the solenoid being used in a nonenergized state. This issue is identified on PIR G87-06-EQ, dated January 28, 1987.

However, Model V5H65600 Skinner solenoid valves are used at Palo Verde Units 1 and 2 in an energized state for its MSIVs.

Therefore, the qualified life was indeterminate.

When informed of the issue, Bechtel project engineers provided the following responses:

4 t

ORGANIZATION: BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 39900519/89-01 RESULTS:

PAGE 4 of 11 (1) Point Beach

" Point Beach is an inactive job, and Bechtel has not performed any EQ activities. No action ta ken. "

(2) Dresden and Quad Cities- "The subject PIR...is not applicable...because CECO does not use these valves in EQ applications. No further action...is required."

(3) Grand Gulf- "There are no Skinner solenoid valves used in

... applications requiring environmental qualification.

Therefore, no further action is required."

Furthermore, there were numerous other technical issues and computer code errors that were documented in the Bechtel "PIR Log," which may be examples of deviations as defined in Part 21.

This subject was not fully reviewed by the inspectors and therefore, this et will be classified as UNRESOLVED ISSUE No. 89-01-01.

D.

STATUS OF PREVIOUS P-

TION FINDINGS

l-Not reviewed durb inspection.

E.

INSPECTION FINDlw.

.ND OTHER COMMENTS:

1.

Copes Vulcan Weight and Center-of-Gravity issue l

The NRC concern was identified in part during a 1988 Northern States Power Company (Prairie Island) engineering review regard-ing the applicability of the seismic issues discussed in NRC Bulletin 79-14. The same problem was subsequently identified during an NRC inspection at the D.C. Cook Units 1 and 2.

During the Prairie Island engineering review of its seismic piping system analyses design basis information for one specific Copes-Vulcan (CV)valvecouldnotbefound. As a result, the engineeringconsultant,FluorDaniel(FD),requestedthevalve design information from CV. Prior to FD receiving the valve design information from CV, the original design basis documents were found by FD, who compared the information to the original seismic analyses input and found the data to be consistent. How-ever, after receiving and reviewing the requested CV design information, FD noted that it contained different, nonconservative valve assembly weights and center-of-gravity (CG) 1 1

i

ORGANIZATION:

BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 5 of 11 l

data. As an example, the original small bore CV valve assembly information showed a specific valve assembly weight to be 170 l

pounds and its CG to be just above 5 inches from the datum point.

Conversely, the new CV valve assembly information indicated the t

same valve to weigh 240 pounds and have a CG of 19 inches from the datum point.

The new CV supplied weights and CG data were then inserted into the original stress calculation at Prairie Island. As a result of the new analyses, several Prairie Island piping systems stressesforseismicapp(FSAR) requirements.

lication were found to exceed their final safety analysis report In one instance stresses increased from approximately 32,000 psi to approximately 161,000 psi. A review of this issue by NRC staff determined that as a result of licensee action due to the issuance of Bulletin 79-14, the Farley nuclear plant (NP) notified Bechtel of 42 questionable CV valve drawings and Bechtel contacted Westinghouse about the drawings.

i During this inspection, the inspector found no examples where Bechtel failed to execute its 10 CFR Part 21 responsibility. At this time, no additional review is planned in this area.

2.

Review of Generic Deficiency Processing Methods (Unresolved Issue) a.

Bechtel Engineering Department Procedure (EDP) 4.74,

" Generic Deficiency Processing," Rev 5,. dated November 16, 1987, was reviewed. This procedure states in part:

(1)

"1.0 PURPOSE AND SCOPE.

To define the method for identification, review, evaluation, and oisposition of generic design and equipment deficiencies by means of Problem Investigaticn Requests (PIRs). The scope of such review is limited only to incoming problem documents received by the Gaithersburg officer. 2.0 GENERAL. This procedure covers the method for engineering review and eval-uation of generic problems detected on nuclear power projects and replaces Problem Alerts formerly used for this purpose. PIRs are also used to obtain problem ~ resolution for computer program errors."

(2) Paragraph 4.0statesin part: "In those cases where a reported problem does not appear to be uniquely limited to a specific project, the PIR Coordinator consults with a cognizant engineering discipline Y

ORGANIZATION: BECHTEL CORPORATION GAITHERSBURG, MARYLAND i

REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 6 of 11 Chief Engineer's representative to confirm whether any

)

active project is, or could be, affected.

If confirmed, the cognizant discipline Chief Engineer designates which i

project (s)istoreceivethePIR. A PIR is then issued to-that project (s) by the cognizant discipline Chief Engineer.

A PIR may also be issued by the Manager of Quality Assurance, BEPC, for action by Engineering."

(3) Paragraph l

5.0 states in part:

"5.0 ENGINEERING RESPONSIBILITIES.

{

5.1 DISCIPLINE CHIEF ENGINEER 5.

Engineering discipline-Chief Engineers, or tneir designees, are responsible for the following:... D.

Approving or rejecting and recycling PIR responses." (4) Paragraph 5.2 states in part:

"5.2 i

PROJECT ENGINEERING. The Project Engineering 1

Manager / Project Engineer or action addressee, is responsible for the following:

D.

Applying the provisions of EDP-4.66 if the evaluation of the generic deficiency identifies a condition reportable to the NRC under 10 CFR 21 or 10 CFR 50.55(e)." (5). Paragraph 6.5 states:

"6.5 ENGINEERING l

APPROVAL, CLOSE0VT AND FOLLOWUP. The cognizant discipline Chief Engineer will approve the action taken by signing and dating Block 7 of the PIR form and returning the form to the PIR Coordinator for the closeout.

If the PIR response is not acceptable, then the cognizant discipline Chief Engineer will, in writing, request additional action or clarification, as appropriate, until an acceptable response

)

is received and approved. Delinquent responses will be followed up as necessary by the discipline Chief Engineer."

b.

The Bechtel PIR Logs for 1986-1989 were reviewed. A number of individual PIR Reports were selected for review. The re-view included the following PIRs:

q PIR No.

PIR Date Subject G-88-14-MQ 4/28/88 Gas Accumulation in Charging Pump i

Piping G-88-10-CQ 3/23/88 Three-bolt C1 amps.Used as Axial Restraint in Lieu of Riser Clamps G-88-06-MQ 2/26/88 Rockwell Motor Operated Valves with Double Key Design

ORGANIZATION:

BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 7 of 11 G-88-04-EQ 2/3/88 Agastat Relays-Mistacked Relays and Mating Sockets G-87-07-MQ 2/11/87 Containment Cooling Units Overstress Concern G-87-06-EQ 1/28/87 Inadequately Qualified Skinner Solenoid Valves c.

The following is a sumary of information found during'the review of PIR G-87-06-EQ dated January 28, 1987 titled,

" Inadequately Qualified Skinner Solenoid Valves", and mentioned in Section C.1 above:

.1 (1) The problem was originally identified by the Bechtel Project for Palo Verde via Bechtel Deficiency EvaluationReport(DER)No.86-11. The DER identified that a number of Skinner Class 1E solenoid valves were used in a system installation for which they were not qualified. Specifically, the Skinner solenoid valves in question were subjected to qualification testing with their respective coils deenergized. These solenoid valves were used for operating the Main Steam m

l Isolation Valves (MSIV) and Main Feed Iso?ation Valves (MFIV). However, in these applications, the solenoid valves were normally energized.

Reanalysis of the Skinner solenoids for a normally energized condition using the original environmental test conditions (140*F, ambient temperature) indicated a design life of less than 4 months.

(2) The existing Skinner solenoids were then reanalyzed for-the energized condition using a less severe in situ ambient temperature and found to have a qualified life of at least 5 years. This qualified life was about as long as that originally determined for the deenergized valves using the higher 140*f environmental temperature.

(3) The safety assessment performed by Bechtel for this deviation at Palo Verde concluded that the condition would not affect the safety function of the valves op-erated by the solenoids (in this case MSIV and MFIV).

The basis for this conclusion was that the circuit de-sign for the solenoid coil is fail-safe (the primar valve would close in the event of solenoid failure)y

\\

q

=.

ORGANIZATION: BECHTEL CORPORATION GAITHERSBURG, MARYLAND l

REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 8 of 11 l

Thus, coil failure, if it did occur, would not prevent the primary valves from failing in their fail-safe (closed) position.

(4) The reportability assessment performed for this deviation at Palo Verde concluded the deviation was not safety significant (based on the Bechtel safety assessment) and, therefore, the condition was not reportable.

(5) PIR G-87-06-EQ requested other Bechtel projects to determine if "a similar deficiency exists on your projects" and to act upon any findings in accordance with project procedures.

(6) The project responses submitted for PIR G-87-06-EQ were signed off by the Bechtel Chief Engineers as being acceptable. The project responses varied from project to project.

include:-(a)QuotesfromBechtelselectedresponses i

"Farley does not use Skinner valves in applications requiring environmental qualification."

(b) " Point Beach is an inactive job and Bechtel has not performed any EQ activities."

(c) " Skinner valves are not used in MSIV (Duane Arnold)," (d) " Skinner valves not used for IE applications (Davis-Besse)," and (e)"DonotusethesevalvesinEQapplications (Dresden)."

d.

Bechtel's overall handling of PIR G-87-06-EQ was reviewed and discussed with the Bechtel QA Manager, Deputy QA Manager, and two senior electrical. engineers. The following points were made by the NRC inspectors:

(1) The scope of the postulated failure modes for the safety assessment analysis appears to have been very i

narrow. Specifically, the analysis apparently assumed that the additional heat input from constant solenoid energization will only result in loss of the solenoid magnetic field.

The analysis appears to ignore the possibility that overheating can cause nonmetallics to break down resulting in swelling or binding that could keep the spring from returning the solenoid valve to the loss of electrical power fail position. Had i

Bechtel considered a broader range of failure modes, the deviation may well have been determined to be safety significant, and this may have required the deviation to be reported.

l

ORGANIZATION:

BECHTEL CORPORATION GAITHERSBURG, MARYLAND l

l REPORT INSPECTION l

NO.: 99900519/89-01 RESULTS:

PAGE 9 of 11 l

The Bechtel Representatives reiterated that they considered the original safety assessment to be satisfactory.

(2) The nonspecific requested action statement in the PIR (i.e., determine if a similar deficiency exists) makes it difficult to understand exactly what question was being asked by the PIR and what question was being answered by the various projects.

For example, (a) Was the scope of the PIR question directed at all IE solenoid valves or just Skinner solenoid valves? The responses from Duane Arnold, Grand Gulf, Davis-Besse. Dresden, Callaway, and Pilgrim suggest that these particular projects only addressed Skinner solenoid valves. The Bechtel Senior Electrical Engineer stated that Bechtel Gaithersburg was-primarily interested in Skinner solenoids because other solenoid valves that they were aware of had qualification test reports for the energized condition.

l (b) Was the scope of the PIR question directed to all l

1E Skinner solenoids or only 1E Skinner solenoids used in applications requiring environmental-qualification (EQ)? Were non-EQ safety applications considered? The responses from projects such as Farley, Point Beach, Grand Gulf, Davis-Besse, and Dresden suggest that the scope of the answer may have been limited to applications requiringEQ(10CFR50.49).

The Bechtel Senior Electrical Engineer indicated he was primarily concerned with EQ applications.

His rationale was

.that non-EQ valves were outside containment and were accessible to permit replacement following failure. EQ applications, on the other hand, were not readily accessible and, therefore, these valves needed to be. identified for replacement prior to failure.

(c) Was the scope of the PIR question directed to all IE solenoid valve applications or just IE solenoid valve applications for MSIV and MFIV? The responses from projects such as Turkey Point, Duane Arnold, Callaway, and Pilgrim suggested that the scope of the answer may have been limited to MSIV and/or MFIV applications.

1 i

ORGANIZATION: BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 10 of 11 i

(d) Does Bechtel ensure that all projects for which l

Bechtel has or had respon.iibility are considered l

t for generic applicability of safety problems? The

{

response provided for Point Beach stated, in part, i

that Point Beach was an inactive Bechtel project.

This response suggests that Bechtel may not have evaluated deviations for plants where Bechtel no I

longer has an active contract with the licensee.

i In this regard, the Bechtel QA Management representative stated that Bechtel is not I

responsible for evaluating deviations with hardware or services of Bechtel's vendors for inactive contracts. The NRC Inspectors stated to the Bechtel representatives that their positions regarding evaluating deviations is contrary to the intent and requirements of 10 CFR Part 21.

i l

(3) Two project responses, one for Callaway MFIVs and one i

for Pilgrim MSIVs and MFIVs, indicated that the Palo Verde deviation was directly applicable to their i

i project.

In both cases, these Bechtel projects used the Palo Verde resolution to justify the technical acceptability of their installations in these plants and to conclude the deviation was not reportable.

It was noted with interest by the NRC Inspector that the Palo Verde installation was apparently only found to be acceptable after the qualification reanalysis was performed using an in situ environmental temperature j

that was lower than the original 140"F qualification i

test temperature.

Neither of the projects adopting the Palo Verde resolution provided any confirmation that the lower in situ temperatures used for Palo Verde were actually applicable for their particular installations.

e.

PIR G-87-07-MQ dated 2/11/87 titled " Containment Cooling Units - Overstress Concern," was rev,iewed. The Bechtel project response for Point Beach stated, " Inactive job - no action taken." This response was discussed with the Deputy QA Manager.

He reiterated his position that Bechtel is not responsible for evaluating conditions on. inactive projects for hardware or service deviations caused by Bechtel vendors. The NRC Inspector reiterated his understanding that Bechtel was responsible for the evaluation under 10 CFR 21 if the vendor could not perform one, regardless of whether the project was active or inactive.

As an additional point related to this matter, the NRC Inspector notes that Bechtel

c.

1 ORGANIZATION:

BECHTEL CORPORATION GAITHERSBURG, MARYLAND REPORT INSPECTION NO.: 99900519/89-01 RESULTS:

PAGE 11 of 11 l

EDP 4.74', Rev. 5, paragraph 4.0, refers to evaluating l

reported problems on any active project, not any project, as required by 10 CFR Part 21.

1 The Bechtel processing of generic deficiencies under EDP 4.74 was identified as a matter requiring further NRC review and evaluation and was classified as an Unresolved Issue No.

'l 89-01-01.

F.

PERSONS CONTACTED:

Name Organization S. Bernsen, QA Manager Bechtel i

D. Kansal, Deputy QA Manager Bechtel P. Dadlani, Senior QA Engineer Bechtel

4. Gandhi, Farley Project Manager Bechtel M. Dugan, Senior Electrical Engineer Bechtel F. Gardner, Senior Electrical Engineer Bechtel R. McCormick, Farley Project QA Manager Bechtel J. Love, Farley Project Bechtel k

l e

a i

.