IR 05000528/1984048
| ML17298B905 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 01/31/1985 |
| From: | Crews J, Hernandez G, Miller L, Narbut P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML17298B904 | List: |
| References | |
| 50-528-84-48, 50-529-84-41, 50-530-84-31, NUDOCS 8502200248 | |
| Download: ML17298B905 (20) | |
Text
U.S.
NUCLEAR REGULATORY COMMISSION
REGION V
Report Nos. 50-528/84-48, 50-529/84-41 and 50-530/84-31 Docket Nos. 50-528, 50-529 and 50-530 License Nos.
CPPR-141, 142 and 143 Licensee:
Arizona Public Service Company P.
O. Box 21666 Phoenix, Arizona 85036 facility Name:
Palo Verde Nuclear Generating Station - Units 1, 2 and
Inspection at:
Palo Verde Construction Site, Wintersburg, Arizona Inspection conducted:
September 5 - December 14, 1984 Inspectors; 4h ~
p. p.
arbut, c
r spector G.
n nd Other Accompanying Personnel:
J. L.
r Construction Resident r
eactor Engineer Approved By:
L. M'r, Jr.,
'e Reactor Projects ection
'/>oi55 Date Signed J r Date Signed J%'
P te igned il Date Signed Summary:
Ins ection on Se tember 5 - December
1984 (Re ort Nos. 50-528/84-48, 50-529/84-41 and 50-530/84-31 of the licensee's and the architect engineer's (Bechtel) activities in the areas of as-built configuration control for pipe supports and design control administration and implementation.
The inspection involved approximately 224 manhours of inspection onsite and at the Bechtel offices in Norwalk, California and approximately 160 manhours of direct inspection of calculations at the EGRG offices in San Ramon, California by two NRC consultants and three NRC inspectors.
Results:
No violations or deviations were identified.
8502200248 850201 PDR ADOCK 05000528 Q
DETAILS 1.
Persons Contacted The below listed key management/supervisory personnel were among those contacted:
a.
Arizona Public Service Com an (APS)
A. C. Rogers, Nuclear Engineering Manager L. Souza, Assistant Corporate QA Manager R. Papworth, Operations Engineering Manager b.
Bechtel Power Cor oration (Bechtel)
R. Elias, Plant Design Chief Engineer W. Bingham, Project Engineering Manager A. Iangmo, Manager of Engineering R. Steins, Project Engineer R. Keidel, Manager Materials and equality Services R. Welcher, Project QA Manager D. Freeland, Plant Design Engineering Group Supervisor In addition numerous design engineers, field engineers, (}A and gC personnel were contacted.
2.
Dates of In section and Personnel Involved a
~
Because of the extended nature of this inspection, a summary of activities is provided for clarity.
Examination b~
Date 9/5-7/84 Location Bechtel, Norwalk A~ctivit Procedure familiarization.
Design drawings acquisition.
LLNL NRC 9/24-27/84 PVNGS Field inspection and walkdown 10/1-3/84 Bechtel, Norwalk Sample calculations acquisition.
Preliminary review.
10/5/84 PVNGS Exit interview
. regarding walkdown results.
Findings for
,. action identifie li
LLNL 10/8-25/84 EGSG, San Ramon In office detailed calculation review by consultants.
LLNL NRC 11/8/84 Bechtel, Norwalk Meeting to present unresolved questions from calculation review, and elicit responses.
LLNL NRC ll/15-16/84 Bechtel, Norwalk Detailed review of substantiating data for responses offered in November 8 meeting.
LLNI NRC ll/30/84 Bechtel San Francisco Presentation of APS actions.
LLNL NRC 12/7/84 NRC Walnut Creek Presentation of APS actions.
b.
NRC Personnel J.
Crews, Senior Reactor Engineer P.
P. Narbut, Project Inspector, PVNGS G. Hernandez, Senior Resident Inspector, Construction, PVNGS c.
Consultant Personnel"-
Clyde H. Morton - Mechanical Consultant Mechanical Engineer with 12 years experience in the design and construction of nuclear power plants.
Currently an Engineering Specialist in Applied Mechanics Section of EGSG Energy Measurements, Inc.
Charles B. Simkins'
Mechanical Consultant Registered Professional Engineer (Mechanical) with 13 years experience in the field of piping design, stress analysis, and mechanical design.
Currently Senior Engineer for Special Measurements Section of EGSG Energy Measurements, Inc.
-The consultants were obtained Livermore National Laboratory employees under a subcontract participated in meetings, but under an NRC contract with Lawrence (LLNL).
They are, in fact, EGSG with LLNL.
Mr. W.
Wade of EGRG also not the actual inspection effort.
3.
Pur ose of Ins ection The purpose of this inspection was to:
Perform a field inspection of pipe supports to assess the accuracy of as-built documentation of the pipe supports.
This was a followup
0,
verification inspection to the September, 1983 CAT inspection findings regarding pipe supports.
Inspect pipe supports and piping final field installations utilizing consultant personnel with piping design experience.
The inspection was to independently assess how well design specification control and implementation provided a well-engineered final product from a designer's point of view.
This was a verification followup to a
portion of IE Bulletin 79-14.
Inspect the piping and pipe support calculations performed by the architect engineer (Bechtel).
The inspection was to verify field changes were properly analyzed when required, or alternately, properly judged to be acceptable without calculations when engineering judgements were made.
Verify, by sampling, that all field changes affecting design were properly controlled administratively.
Specifically, that they were reviewed by the designer and properly reconciled with the design calculations.
Verify that field change options (which do not require designer review by the licensee's procedures)
were being properly implemented by field personnel, i.e., that field personnel do not work outside the design specifications.
4.
Sco e and Methodolo of Ins ection The following was used:
Procedures:
The applicable procedures for design control, design reconciliation, and field implementation were reviewed for familiarity and general acceptability.
Judgements of procedure adequacy were made by assessing the products of those procedures, the hardware and design calculations.
Hardware Walkdown:
A safety-related system, the Unit 1 Auxiliary Feedwater System, was selected as a representative sample of systems important to safety.
All piping and pipe supports (in both Train A and B) from the Auxiliary Feedwater Pump discharges to the Steam Generators which they serve were examined.
There were 46 supports involved.
All applicable drawings and amending documents such as Nonconformance Reports listed on the Design Document Register were used to inspect the final installed pipe supports in the field.
This provided an assessment of the accuracy of the Design Document Register.
Alternately, this was to confirm that the as-built documentation accurately reflected the as-built condition.
Additionally, the hardware walkdown was used to judge, from an independent designer's standpoint, whether there were any apparent design oversights such as questionably long unsupported pipe spans, insufficient two way and three way supports, or inadequate piping load isolation of critical areas (such as pump nozzles).
Finally, the field walkdown used the change documents which were authorized to be issued by the field (without designer approval).
These field approved documents
l.
(which were not permitted to make design changes)
were assessed to determine if unauthorized design changes had been made.
Desi n Calculation Review:
The design calculations for 14 of the pipe supports were examined.
The computer methodology for pipe stress was not comprehensively included since this had been reviewed by NRC at the time of PSAR review.
The calculations for piping and pipe supports were reviewed on a sampling basis for reasonable input assumptions and reasonable numerical results.
Hand calculations were reviewed for reasonable methodology and results.
Where designer approval of changes was based on "judgements" rather than calculations, the "judgements" were independently assessed by the NRC consultants to determine if the judgements were reasonable.
Second, the design calculation review was used to assess, by sampling, whether design changes were in fact reviewed by the designers.
Third, the calculation review was used to determine if the designer had analyzed the proper configuration of the support when pipe loadings were revised due to changes in engineering assumptions, or field changes to piping.
Fourth, the calculation review results were used to measure the adequacy of the architect-engineer's and licensee's QA programs for design activity overview.
5.
Si nificant Results of Ins ection a.
General The details of the examination performed by the NRC consultants are attached to this report (Enclosure 2).
The issues and findings remaining to be resolved as identified in the consultant report have been restated in NRC report format in the body of this report.
b.
~Summar In regard to the objectives of the inspection discussed in paragraph 2,
and the results of the inspection described in the attached consultant report, it is concluded that:
1)
As-built Documentation vs As-built Hardware The walkdown results show thatthe as-built. documentation listed on the Design Document Register generally accurately reflected the as-built hardware.
Although a number of deficiencies in pipe supports were identified which had not been recorded on nonconformances by the licensee, these were generally "more-of-the-same" types of deficiencies identified during the'September, 1983 regional CAT inspection.
The deficiencies included undersize and
underlength welds and dimensional discrepancies (see attachment, section 4).
These types of deficiencies were expected by the NRC based on the extent of corrective action taken subsequent to the CAT inspection.
The licensee had inspected a sample of 2199 pipe supports and identified 1269 similar deficiencies.
Since all of these deficiencies were evaluated as technically acceptable-as-is, the licensee did not perform inspections of each and every pipe support.
Therefore, the consultant's findings were not surprising nor disturbing.
The nonconformances written by the licensee as a result of the consultant's findings have also been technically resolved as acceptable-as-is.
Therefore it can be concluded that as-built documentation was, overall, satisfactory.
En ineerin Evaluation of Final Installation Xn general, the engineering of the installed piping and supports was satisfactory from a designer's standpoint.
Three potential problems in the adequacy of design specifications transmitted to the field were identified which require further examination and resolution.
L~inesto s
The first of the potential problems is a question regarding
"linestops" (see attachment, paragraph 5g, issue 813).
A
"linestop" is a pipe support located near a critical component such as a pump nozzle.
The line stop is installed to take up any pipe loads resulting from thermal growth, or a seismic event, and prevent those loads from being t'ransmitted to the critical component (e.g.
a pump nozzle).
The consultants noted that field construction tolerances allow 1/8 inch of "play" in construction of the line stop, whereas the calculations of loads imposed on the nozzle assume a zero inch "play".
This means that piping as constructed can move 1/8 inch and apply larger loads to pump nozzles than is apparently calculated for.
However, Bechtel engineering management strongly indicated that the apparent contradiction between the design methodology and
, the field tolerances was a widely accepted and valid engineering practice.
After a series of meetings (including NRR personnel versed in design methodology)
the licensee submitted letter ANPP 31473-EEVB/ACR on December 14, 1984, which addressed the "line stop" question using the term "close proximity supports".
The letter stated that the basic design (for line stops) at Palo Verde was conservative but that complex, time consuming calculations would be necessary to demonstrate this to NRC's satisfaction.
Therefore, the licensee committed to shim all
"proximate" supports (as defined to the staff at the Diablo Canyon Project) adjacent to rotating and reciprocating equipment to 1/16 inch gap on Unit 1 prior to initial criticality.
ANPP reserved the right to perform the extensive calculations, or other alternate design approaches for Units 2 and 3, in lieu of shimmin The licensee's actions to shim "proximate" supports in Unit 1 prior to initial criticality is a followup item (Followup Item 50-528/84-48-01).
Zero Clearance Free-to-Slide The second potential problem regarding design specification adequacy deals with a specification instruction to the field personnel which allowed "zero clearance free to slide" (see Enclosure 2, paragraph 5.g.,
issue 15).
This applies to pipe supports which are not "line stops" but are designed to hold a
pipe in place vertically and horizontally but allow the pipe to move (slide) axially for reasons of thermal growth.
In one case, the consultants found such a support with "zero clearance" around the pipe
~
Their question centers on how the field (and the designer)
can be assured that such an installation is, in fact, "free to slide" as assumed by the designer.
If the pipe is "bound" in the support (vs. free to slide), then unintended stresses can be transmitted to the support, and reactively to the piping.
The licensee, in their letter of December 14, 1984, reported the results of their investigation of the potential problem.
They reported that 6200 safety related supports are potentially affected.
Of these, 4350 involve hot piping.
Any such restrictions of hot piping would have been identified during the "piping verification program," which predicted and verified by measurement the movement of piping due to thermal growth.
Of the 1850 supports on cold pipe, the licensee selected a 5/
sample (100 supports) for field examination.
All piping was found free to slide.
Additionally, the regional CAT inspection in September, 1984 had examined similar supports, and found them free to slide.
Based on the results of the licensee's sampling and the regional CAT inspection, this issue is considered resolved.
Vertical Liftoff During the review of calculations the consultants noted that the computer modeling used for pipe analysis assumes the pipe is restrained from moving upwards or downwards by a pipe support, regardless of whether the pipe support actually had an upward movement restraint (see Enclosure 2, paragraph 5.g.,
issue 14).
This was usually not a concern since upward forces (such as seismic and thermal)
were generally overcome by downward forces (such as dead weight).
However, if upward forces exceed downward forces, the piping could liftoff the support and experience a different vibrational mode, which could affect the stresses on the pipe.
The licensee initiated a study, the status of which was reported in their December 14, 1984, letter.
They committed to analyze uplift forces for 100/ of the affected pipe support If any are found (where uplift exceeds downward forces),
necessary corrections will be made prior to initial criticality.
The study results, as of December 10, 1984, showed about 15/ of the calculations had been preliminarily reviewed, with no safety concerns identified.
The licensee commitment to complete the liftoffstudy and make any necessary hardware changes prior to initial criticality in Unit 1 is a followup item (Followup Item 50-528/84-48-02).
Other Items Other relatively minor observations regarding design specifications were made, including questions on weld symbol usage and the appropriateness of using a non-design document to change a design document.
The licensee committed to corrective actions as described in paragraph 6 of this report.
Desi n Reconciliation of Field Chan es In general, the review of the designer's reconciliation of field changes was found to be adequate.
Where field changes were reconciled by calculation, the calculations were found to be satisfactory.
Where the field changes were reconciled by
"judgement",
the judgements were found to be sound.
Regarding field changes accepted by engineer
"judgements",
the designer used a system whereby there was documented evidence that each field change or nonconformance was evaluated.
However, the rationale for acceptance, when based on
"judgement",
was found to be lacking (see Enclosure 2, section 5, finding 1).
Although there was not a requirement for the rationale to be documented, the consultant considered it to be a good practice.
The licensee's senior engineering representative stated that the licensee had noted the lack of documented rationale in 1982 and had taken corrective action which will preclude recurrence in Units 2 and 3.
The licensee's actions will be verified in the course of future inspections.
It can be concluded, however, that design reconcilation of field changes was, properly performed and met regulatory requirements.
Desi n Administrative Controls In general, the designer's administrative control of design was found to be proper and accurate.
In one case, however, revised piping loads were erroneously applied to the wrong drawing revision of a pipe support (see attachment, paragraph 5f, issue 12).,
In essence, the pipe support evaluated was not the support. in the field.
The individual who made this error was not part of the normal work force that performs this task.
The licensee committed to investigate the extent and significance of such errors.
In the
l'
I j
i
case of the error found by the consultants, the administrative error did not affect the technical acceptability of the installed pipe support.
The preliminary results of this investigation by the licensee were reported in their December 14, 1984 letter.
Approximately 30% of the calculations were reviewed to determine if similar errors were made.
This included 1175 supports and 2,138 modification documents.
Four additional, similar errors were identified for a total of five errors (0.2/).
All were reanalyzed and found acceptable-as-is.
To prevent recurrence the designer has implemented a
"checklist" procedure.
All calculations will be re-reviewed using this checklist by mid February.
This issue is considered resolved based on the licensee's actions, taken and planned, and the fact that all errors identified have been acceptable without hardware changes.
Field Chan es Which Are Not Reviewed b
the Desi ner During the pipe support walkdown, the consultants reviewed field change options whi'ch were not r'equired to be approved by the designer, These included modification change notices (MCN's) and field revisions to the support drawings.
Both are formal mechanisms designed to allow field personnel to exercise options already permitted in design specifications without further designer approval.
In general these mechanisms were found to be properly used.
No problems were 'found with the use of MCN's.
(Note:
The consultant's report (see Enclosure 2, paragraph 5e) refers to the MCN issue as "unresolved" in reference to the existing NRC item regarding field change documents, Followup Item 50-528/84-39-01.
The subject of MCN's had been extensively evaluated in NRC reports as discussed in Report 50-528/84-39.
There is a remaining action for the licensee to perform an audit to assure the corrective actions taken were effective.
Therefore, this issue MCN's was considered
"unresolved" by the consultants.)
One problem was found with a field drawing revision which should have been submitted for designer approval (see attachment, paragraph 5f, issue 8).
Although the change was subsequently shown to be technically satisfactory, the change should have had prior designer approval.
Based on this single finding, further sampling to determine the scope and significance of changes which bypassed the designer was committed to by the licensee.
The licensee reported the results of their review in the December 14, 1984 letter.
An approximate 10/ sample (1106 supports)
was conducted to determine whether the designer had been similarly bypassed.
Three additional cases were identified wherein a field change had been approved by the designer for a single unit but had
been approved by the field for all three units (an administrative error).
In all three cases the changes (if they had been implemented)
would have been acceptable.
This issue is considered resolved based on the licensee's actions and the lack of safety significance of the results.
The following items represent other relatively minor items identified as a result of the inspection:
During the walkdown the consultants noted certain weld symbols they considered could have been more clearly indicated on support drawings (see attachment, paragraph 5.f(3), issue j/11).
None of the symbols were improperly interpreted by the field personnel except one.
Support Dwg 13-AF-015-H003 shows a 3/16 inch fillet between items C (angle iron) and B (channel iron).
The designer's intent was to have the horizontal side of the angle iron welded to the cut end of the channel.
The field welded the vertical side of the angle iron to the bottom of the channel.
Subsequent analysis showed the installation to be satisfactory.
Additionally, this was simply another example of similar problems identified in the extensive sampling done after the CAT inspection which were found not to have technical significance (that is, the involved supports were acceptable as is).
However, the welding Quality Control engineer accompanying the consultants indicated at that time that either weld location was acceptable.
At the December 7,
1984 meeting the licensee's Assistant Corporate QA Manager stated that all welding inspectors had been interviewed as to their understanding of weld symbology.
Additionally, a test had been administered and none of the inspectors scored less than 90/.
This item will be followed up further to verify the results of the examinations and interviews (Followup Item 50-528/84-48-03).
b.
Flare Bevel Welds During the field walkdown the consultants noted that the welding of flare bevel welds was done differently than at most other sites, in that the welds are not filled out to the full radius of the tube steel (see attachment,,
paragraph 5.g. issue 816).
This question was an existing NRC unresolved item (reference 50-530/82-09-02).
The licensee had previously committed to provide substantiating information that shows that the effective weld throat assumed by the designer was in fact being achieved by welders in the fiel Ik
At the December 7,
1984 meeting the A/E representatives provided Technical Report No. 8412-02 EV, BLN No. 0484-06 dated December, 1984.
The report shows that the effective throat assumed by the designer was achieved in actual welds using Palo Verde weld procedures and tube steel.
Therefore this issue is considered resolved and Unresolved Item 50-530/82-09-02 is closed.
c ~
S rin Can Deflected b
Fire Protection Insulation The consultants found a pipe support spring can deflected from vertical by the installation of fire protection insulation which had occurred after the spring can installation had been accepted.
At the December 7 meeting the license presented a
TMX from Corner and Lada, the spring can manufacturer, which confirmed that the deflection angle (6~) was acceptable as is.
The licensee chose to
,correct the installation in any event, since it exceeded the Palo Verde installation specification maximum of 5~.
The licensee stated the fire protection insulation installation specification (MM301)
had been revised to specifically check for insulation interference.
This issue is considered resolved based on the licensee's actions.
d.
Flare Bevel Welds With Added Fillet, Ca s
The consultants questioned the licensee's use of fillet caps on flare bevel welds since 1)
a later version of the AWS D.l.l Commentary than is in effect at Palo Verde indicates that credit should not be taken for added fillet caps on groove welds and 2) the addition of a fillet cap does not substantially increase the overall weld throat size.
On December 13, 1984, the licensee provided calculations which show that the addition of a fillet cap to a flare bevel weld is a technically sound method to increase the strength of the welded connection.
This is because the strength of the base metal (in contact with weld metal) versus the strength of the weld throat controls the weld strength in the flare bevel weld configuration.
This issue is considered technically resolved.
7.
Exit Interview
\\
Due to the extent of this inspection and the number of meetings held to exchange information no single exit interview was conducted.
The meetings attended and the participants are detailed in paragraph 2 of this report and in section 2 of the attached consultant's repor ~r b
P i
1