ML19308B927
| ML19308B927 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/08/1978 |
| From: | Jennifer Davis NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| To: | |
| Shared Package | |
| ML19308B928 | List: |
| References | |
| TASK-TF, TASK-TMR SECY-78-495, NUDOCS 8001170636 | |
| Download: ML19308B927 (13) | |
Text
I b4a( fd SECY-78-495 September 8, 1978 l
! O COMMISSIONER ACTION i
For:
The Commissioners From:
John G. Davis, Acting Director Office of Inspection and Enforcement Executive Director for Operations NO, Thru:
I
Subject:
LETTER TO O!!B CONCERNING THE HUCLEAR REGULATORY COMillSSION'S LICENSEE CONTRACTOR AND VENDOR j
IllSPECTION PROGRAM
Purpose:
To obtain Commission approval of a letter to OMB which transmits an IE staff study and an independent contractor study of the Licensee Contractor and Vendor Inspection Program (LCVIP).
Discussion:
The LCVIP is the program for direct NRC audit inspection of selected design and manufacturing activities that occur away -- offsite -- from the i
site on which a reactor is constructed. LCVIP complements and is founded on the same direct inspection philosophy as the m're extensive onsite NRC inspection effort. The LCEIP determines that nuclear steam system suppliers, architect engineers, and selected fabricators of components important to safety have implemented quality assurance programs capable of producing a service or product sufficient for the safety needs of the user. The LCVIP -- as the other inspection activities of the NRC --
includes both a preventive element and a reactive element. The preventive element consists of rou-tinely planned and scheduled inspections; the reactive element consists of inspections and investigations conducted in response to events, incidents, defect reports or allegations.
Cur -
rently,)the LCVIP is centralized in Region IV (Dallas.
800117
The Commissioners /~N b
The Office of Management and Budget has questioned the need for the LCVIP during the last two NRC budget reviews.
In staff level contacts with Of1B, NRC agreed to provide a study of the LCVIP as mutually determined by NRC and OMB.
Mr. Joseph Kearney, Ot1B, confirmed this agreement with his letter of March 27, 1978 to lir. Gossick.
Assignment of a new Ot1B staff analyst delayed the study some-what.
It would be advantageous if the letter and report could be received by OMB in time for consid-eration of the FY 80 budget.
The study has now been completed by the Division of Reactor Construction Inspection, IE; the effort includes an independent study conducted by TRW, Incorporated under contract to IE.
Both the in-house and contractor studies are provided as attach-ments to the proposed letter to OMB.
Both studies confirm the need for a continuing LCVIP. This assessment is based primarily on the need -- and philosophy -- of NRC for independence in arriving at judgments.
The studies also conclude that LCVIP has identified and caused correction of some specific problems. The studies identified areas for improveiiieni, that IE will evaluate for incorporation into the program.
The TRW study considers seven alternatives for conducting LCVIP.
TRW distinguishes between
" certification" audits (approving the vendors QA program as described in manuals, procedures, etc.)
and "impleme:itation" audits (determining that the QA program is implemented as described).
TRW evaluates several aiternatives for performance of these separate audits.
The IE program, as now carried out, does not separate into " certification" and " implementation" audits.
IE does not (and never intended) to review a vendor's complete QA program and " certify" that it meets Appendix B.
Rather, in its routine inspections, IE samples the vendors QA program by examining the effectiveness of its impicmentation.
This is done by inspecting a sample of the work product (hardware),
observing work undenvay, and reviewing records that lp b
The Commissioners -
l l
i attest to quality.
In this way, IE determines, based on the sample, whether the implemented QA program satisfies Appendix B.
Thus, LCVIP is developed and implemented according to the same philosophy as all IE routine inspection effort -
qualitative extrapolation from a small sample.
The IE staff study considers five alternative --
elimination of LCVIP and four alternative ways of conducting LCVIP.
IE does not consider elimination as a viable alternative if the flRC is to make independent judgments. TRW does not treat elimina-tion as an alternative.
The four viable alterna-tives are:
I 1.
Host Concept (Alternative 2).
Under this l
decentralized concept, regional-based IE r
inspectors would accompany utility representa-tives to audit the utility's performance in auditing its contractor.
The agency used this technique at one time, but discarded it.
(This is similar to TRW Alternative I.)
2.
Reactive effort only (Alternative 3).
This alternative would eliminate any centralized preventive (routinely scheduled) inspections but would provide for centralized reactive inspection or investigation of events, inci-dents, defects, or allegations at licensee contractors and vendors.
(This is similar to TRW Alternative II.)
3.
Current LCVIP (Alternative 4).
This is the current minimum program that provides for both preventive and reactive inspection from a centralized office and continuation of the trial with ASME of third party inspection.
(This is similar to TRW Alternative III.)
4..
Augmented LCVIP (Alternative 5).
This alternative involves a continuation of'the current program for preventive inspection of vendors and contractors, an increase in resources for the reactive program, and addi-tional resources to expedite efforts toward third party agreements with groups in addition to ASME.
(TRW Alternatives IV, V, VI, VII are modifications or variations of this Alternative.}
r r
n m
i
1 The Commissioners. ')
y Recommendation:
The study concludes that the current LCVIP should s,
continue.
Recognizing budget constraints, IE recommends continuation of the current program (29 positions) through FY 80.
By the close of FY 80, the study of third party inspection will be completed.
At that time, IE believes the program should be re-evaluated with augmentation as an alternative.
IE recommends that the Commission approve the letter to OMB.
Concurrences:
This paper has the. concurrence of NRR, MPA and CON.
oh G. Davis Ac ing Director Office of Inspection and Enforcement
Enclosure:
^
Proposed letter to OMB with attachments NOTE:
Commissioners' comments should be provided directly to the Office of the Secretary by c.o.b. Thursday, September 21, 1978 Commission Staff Office comments, if any, should be submitted to the Commissioners NLT September 15, 1978, with an information copy to the Office of the Secretary.
If the paper is of such a nature that it requires additional time for analytical review and comment, the Commissioners and the Secretariat should be apprised of when comments may be expected.
DISTRIBUTION:
Commissioners Commission Staff Offices Exec. Dir. for Opers.
Regional Offices Secretariat
[ \\
q.
e e
f* **%q 8
UNITED STATES
[
NUCLEAR REGULATORY CoMMiss!Of[
g r,,
t
., w g
WASHING TON. D. C. 20555 e,,, *#
O
~
Mr. Joseph P. Kearney Branch Chief Energy and Technology Branch Office of Management and Budget 8002 New Executive Office Building Washington, D.C.
20503
Dear Mr. Kearney:
NRC has conducted a thorough study of the Vendor Inspection Program in response to your letter of March 27, 1978.
This study concludes that I
the Vendor Inspection Program is essential to the public safety mission l
of this agency and should continue.
The Office of Inspection and Enforcement staff has prepared a paper, provided as Attachment IV, that sumarizes the study.
A detailed evaluation of the vendor program was conducted by TRW, Incorporated, an independent contractor.
The results of the TRW study are documented in a formal report (NUREG/CR-0217) that is provided as an attachment to the in-house study.
Recent discussions with your staff have helped us identify specific concerns about the program:
1.
Does the direct NRC inspection of vendors and contractors duplicate inspections that are performed by others?
2.
If the direct NRC inspections are duplicative, why is there a need for direct inspection?
3.
What benefits have resulted from the NRC vendor inspection program?
We have prepared short attachments (I, II, and III) specifically addressing each of these issues.
We are confident that our responses to these three questions, supported by the in-house oaoer and contractor charts in Appendix E of Attachment IV). study, demonstrate the value o Our studies, as well as a recent GA0 audit of the vendor program, reinforce our belief in the value of the program.
Each of these evaluations has identified areas of improvements that will make the vendor program even more effective.
b 7
m
+j l
i I
Mr. Joseph P. Kearney 2-O The comments and criticisms of your staff have likewise been helpful in
\\
identifying areas of potential improvement.
We look forward to any l
future discussions you wish to have about the program.
Sincerely, l
j i
1.ee V. Gossick Executive Director for Operations Attachments:
I - III Responses to Questions IV.
IE Study V.
TRW Study
- 1 SECY NOTE:
only five copies, one for each Commissioner, provided e
9 0
f D
e Y-I' Q.,
Attachment I DOES THE DIRECT NRC INSPECTION OF VENDORS AND CONTRACTORS
(~]
DUPLICATE INSPECTIONS PERFORMED BY OTHERS?
V The direct inspection of vendors and contractors by NRC has some appearance of duplicating inspections done by others, because the NRC program examines some of the same areas.
However, the purpose and perspective of NRC vendor inspections are distinctly different from those performed by licensees.
A licensee inspects vendors for two basic reasons.
First, the licensee assures that the vendor is performing in accordance with his specific contract. Second, the licensee must assure that the vendor has a quality assurance program that meets the NRC requirements as the licensee understands them.
NRC inspections of vendors cover some of the same areas as those conducted by licensees.
However, NRC's goal is to assure that the vendor has a quality assurance program that meets NRC requirements and is thereby capable of producing products that will meet safety standards and criteria. The NRC inspections are not influenced by the demands of satisfying a particular contract requirement, nor are they oriented toward " acceptance" of a particular product.
Rather, NRC inspections focus on the vendor's ability to produce safe products.
Furthermore, NRC inspections are generic across the nuclear industry, whereas each licensee inspects against specific provisions of his contract.
The common areas of inspection between NRC and licensees apply only to the NRC regularly scheduled " preventive" inspections of vendors.
A major element of the NRC vendor inspection program has no licensee counterpart. This is " reactive," or unscheduled, inspectio.n taken in response to problems identified at licensees' sites or at vendor shops.
Feedback from the NRC onsite inspection program, evaluation of licensee events reports, incidents, allegations, defect reports (under 10 CFR 21) all identify matters that must be investigated for generic as well as plant-specific impact. These essential investigations are performed by the NRC vendor inspection program.
About 300 ASME system inspectors currently inspect vendors that manufacture nuclear quality pressure retaining piping and vessels as required by most state codes and insurance indemnity considerations.
The NRC has long recognized that this ASME program, with appropriate modification and NRC audit, may serve to reduce, and perhaps replace, the NRC direct inspection of certain vendors.
For several years, NRC has been working toward such a modified and NRC audited program.
It is undergoing trial at the present time.
In addition, NRC is exploring the possibility that a similar program,can be initiated with IEEE for inspection of non-mechanical components.
2-O V
in summary, the preventive inspections of vendors by HRC examine many of the same activities as do inspections by others, but they are performed for a different purpose and from a different perspective.
!!RC reactive vendor inspections and investigations have no counterparts in the licensee inspections of vendors.
l
~
l 1
o e
9 r
Attachment II IF THE DIRECT NRC INSPECTIONS ARE DUPLICATIVE, WHY IS THERE A NEED FOR DIRECT INSPECTION?
As discusseo in Attachment I, NRC inspections of nuclear vendors do not duplicate those conducted by licensees.
direct inspections of vendors should be explained.Nonetheless, the value of There are two principal reasons why HRC independently inspects vendors, l.
Direct inspection of selected vendors is a continuation of HRC's practice of independently verifying information and activities concerning licensees.
2.
Experience has shown that techniques not employing direct inspection of vendors provide an unsatisfactory basis for the discharge of NRC responsibilities.
t Independent review and verification is a cornerstone of NRC's regulatory program.
The agency cannot depend, as the basis for its judgment, on information concerning licensee activities without independently examining, verifying, and evaluating it.
The principle of independent and inspection activities. verification is evident in NRC's research, standard s If it is to provide assurance of public health and safety, NRC must have its own independent sources of informa-tion upon which its judgments are based.
NRC must routinely audit licensee activities, including those provided by vendors, to assure their adequacy.
Accepting the adequacy of a vendor's quality assurance to issuing a license without an independent review of a licen application.
NRC recognizes that independent verification is a key element in its inspection.
NRC audit inspections cover both onsite and offsite activities.
By sampling the activities of licensees, contractors, and vendors, NRC can provide significant insights into the performance of the quality assurance activities of each.
The opportunity for more direct verification of licensee activities is a driving force behind NRC's recent decisioit to assign resident inspectors to nuclear power plants.
The vendor inspection programs. program has the greatest leverage of all NRC With less than five percent of NRC's inspection resources, the vendor program covers a.ctivitids that account for about one half of the total capital cost of a nuclear plant.
NRC's direct independent inspection of vendors evolved from earlier experience.
of a very few vendors.In the early 1960s, the agency performed direct inspections AEC relied upon the start-up testing of a reactor to identify problems.
During the late 1960s, problems Q
~
U hi f
2-identified during late construction of a major reactorl/ ed to l
reexamination of the earlier practice and the adoption of new standards.
This case emphasized that quality during the offsite design and fabri-cation impacted strongly on the plant quality.
Initially, NRC inspec-tors accompanied utility inspectors who were evaluating vendor's quality assurance programs; this approach proved difficult and unsatisfactory.
The agency shifted its vendor inspection practices in 1974 by initiating a two-year trial program for direct vendor inspection.
At about the same time, serious negotiations began with ASME to establish an acceptable "Th'rd Party" inspection effort.
If the trial program now underway with ASME proves successful, NRC's ability to independently verify the activities of vendors will be increased substantially.
Experience has shown that other techniques -- relying on preoperational testing; auditing I
of the licensee's vendor inspections -- are unsatisfactory in providing l
a basis for discharge of HRC responsibility.
,l/ liarginal quality assurance and quality control during the design, i
fabrication and erection of the Oyster Creek facility caused lengthy delays in completion and substantially increased costs.
O-
Attachment III WHAT BENEFITS HAVE RESULTED FR0ft THE ttRC VENDOR INSPECTION PROGRAM 7 NRC believes that significant improvements in the safety of nuclear power plants are attributable to the Vendor Inspection Program.
While we cannot demonstrate in absolute terms the safety benefits of this or any other regulatory program, we believe that the benefits of the program -- as well as the consequences of not having it -- can be inferred from a review of selecteri cases.
1.
. Sensitization of reactor pressure vessel safe-ends.
A transition piece called the safe-end is attached to the reactor press'are vessel nozzles by the manufacturer.
I become more prone to stress corrosion cracking if subjected toThe material l
improper heat treatment (sensitization).
Improper heat treatment by the manufacturer was observed by the AEC (HRC) vendor inspector.
As a consequence of this observation, the safe-ends of approxi-mately 15 reactor vessels were replaced or repaired prior to placing the vessels in service.
The safety significance of this finding was the avoidance of potential stress corrosion cracking in a critical location (high radiation area of the primary pressure boundary). Stress corrosion cracking in stainless steel piping due to sensitization is the type of piping integrity problem that has occurred more frequently than any other reactor piping system problem. The avoidance of cracking in the reactor nozzle area has saved the industry millions of dollars and hundreds of man-rems of exposure in repair costs.1/
2.
Major flaws in pressure nozzle welds.
During a two-man NRC inspection of a reactor pressure vessel manufacturer, both inspectors identified independently the fact that the manufacturer had been having difficulty in meeting weld acceptance standards on the main piping nozzles.
Investigation disclosed the problem to be generic to a number of vessels.
Subsequent to identification of the problem, one vessel which had i
been shipped to the facility site was found to have unacceptable defects.
Extensive repairs were required c.a this vessel at the field site as well as to others.still in the shop.
if Noteworthy is a current outage occurring at the Duane Arnold plant.
A reactor vessel nozzle problem'due to another cause is resulting in a lengthy unscheduled outage and a direct cost of approximately 500 man rems in radiation exposure to accomplish the repair.
O
_4 w
The safety significance of this finding relates to the critical O
importance of the main reactor vessel nozzles to maintaining core cooling capability, both during normal operation and during any accident condition.
The optimum quality of workmanship must be obtained in the manufacture of pressure vessels since there is no redundancy associated with this critical component.
The manu-facturer would not have taken corrective action had not the NRC (AEC) identified the problem.
Furthermore, individual owners were unable to identify the problem as generic since their access to the manufacturer's records was limited to their individual vessels.
Access to all contractual records of a manufacturer is a unique advantage of URC direct inspection in the identification of generic problems.
3.
Inadequate packaging of fuel assemblies During a routine inspection of a inajor fuel manufacturer, the NRC i
inspector observed steel chips inside the plastic envelope used to protect the fuel bundles. It was determined that the chips resulted from drilling modification of the shipping containers.
The safety significance of this finding is that the metal chips could become wedged between the fuel pins and other parts of the fuel bundle leading to premature failure of the fuel from fretting and/or blockage of the coolant path through the bundle.
Either result would require early replacement of the fuel, a multi-million dollt.r expense.
Noteworthy is that licensees currently inspecting fuel fabricators failed to identify the problem.
Also noteworthy is that licensees did not inspect or audit fuel fabricators until about 1973 when the AEC/NRC initiated inspections of fuel fabricators because of multiple fuel problems occurring at that time.
Subsequent to initiation of the NRC's inspections, a marked reduction in the incidence of fuel problems has occurred.
4.
Upgrading of Vendor Quality Assurance Programs (see charts in Appe A significant direct result of NRC vendor inspections is the upgrading effect that such inspections have had upon vendor QA organizations.
By focusing senior managements' attention on the importance of QA in their organizations, the NRC Vendor Inspection Program caused a major reorganization to be made at one major AE, motivated improved performance of the QA organization at one NSSS, I
and influenced the addition of personnel to the QA departments of i
several other vendors.
By providing stronger management support to j
the internal QA organizations more effective implementation of QA w
L._
3-()'
programs is obtained.
Evidence of improved effectiveness has been demonstrated by a declining incidence of identified deficiencies since initiation of direct NRC inspection.
Further verification of this benefit was obtained by TRW in statements from vendor per-sonnel during private interviews.
TRW observed that licensees and their agents tend to concentrate their inspection and audit attention in the same areas as the NRC.
The implication of this observation is that the. industry may relax in their audit programs should the NRC discontinue its direct inspection.
This action may have an adverse effect upon the health and safety of the public.
t e
1 a
]
s t
l l