IR 05000312/1985019

Insp Rept 50-312/85-19 on 850623-0809.Major Areas Inspected: Evaluation of Causes of Hot Leg B High Point Vent Leak, Including Areas of Engineering & Design Control,Const,Qa Insps & QA Oversight

ML20135B527
Person / Time
Site:	Rancho Seco
Issue date:	08/22/1985
From:	Crews J, Dangelo A, Eckhardt J, Miller L, Perez G, Wong H NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:
Shared Package
ML20135B452	List: IR 05000312/1985019 ML20135B449
References
50-312-85-19, IEB-79-14, NUDOCS 8509110091
Download: ML20135B527 (13)

Text

{{#Wiki_filter:. . . - - - - - - _ . ._ . .- _- . - - - . . , '

 ' -
.

, , U. S. NUCIIAR REGULATORY COMMISSION , ! , ' REGION V s

         '
        '

Report N /85-19 ,

Docket N ,

       +

License N DPR-54

        . 7-  ,

f 4 , Licensee: Sacramento Municipal Utility District' P. 0. Box 15830 - . Sacramento, California 95813' . Facility Name: Rancho Seco Unit ,

         ,
          ,
           .

Inspection at: Herald, Califortiia (Rancho Seco Site)

,

i Inspection conducted: June 23 - ugust 9, 1985 ' Inspectors: '

           '27'k3 J.lf! ckhard   eporResidentInspector    Date Signed
   /[
   ~
       '(    h'2 L[

i G. PT rez etInsp7ctor Date Signed

 .

r %, . w + nac J. Iy Cri s , Sen' eactor Engineer- Date Signed

   // //lo, )dide )n   b t Inspector, San Onofre F-22 4f Date Signed

.; A . J . @ '/ h 2 Z '[[

   '

l ~

    [   (

j H. J. yo g, Of fice ' spection , Date Signed and ceme - Approved By: A L2 Y ! L. F.lyfller Jr. , yC ' ef -Date Signed

Reactot Projects, Eection 2 Summary: Inspection between June 23 - August 9,'1985 (Reporc 50-312/85-19) i . . ! ' Areas' Inspected: This special inspection evaluated the causes of the B hot leg high point. vent leak, and included the areas of' engineering and design ! control,~. construction, quality control inspections, and quality assurance !- oversight. This inspection involved 298 hours by five inspector Results: Enforcement actions relative to the. issues discussed in this report t will be the subject of separate correspondenc i e G c

           ,

________.._1_._.__.___.-.____ - - _ . _ . . . . . _ _ . - . - . _ - _ m __

 - - - . . - - - - .. ___ --. . - . -_ - - . .-. . - - .

t

'
.         i
.

I !

1 1-DETAILS

Persons Contacted i Licensee Personnel

R. Rodriguez, Assistant General Manager, Nuclear

,  *P. Oubre', Manager, P clear Operations
 *G. Coward, Plant Superintendent

' L. Keilman, Manager, Nuclear Engineering L. Schwieger, Manager, Quality Assurance 1*J. Jewett, Site QA Supervisor

 *J. McColligan, Nuclear Project Engineer     !

'.

 *J. Field, Technical Support Superintendent     ,
 *D. Abbott, Supervising Mechanical Engineer
 *S. Crunk, Regulatory Compliance; Engineer J. Barker, QA Engineer  .

1 -

[ 'Bechtel Personnel
 *T. Kahn, Group Supervisor,; Mechanical
     ,

In addition to the above personnel, the inspectdrs^ contacted other'

supervisors, operators, enginesrs, and inspectors during this inspection,

       '

i * Attended the exit meeting on Augus 9, 198 ,

       ,   ,

j . Description of the Event initiated a plant cooldown, and declared an unusual event due to a RCS t leakage rate of greater than 10 gallons per minute. .At 0735 hours a i licensee inspection team entered the reactor building in an attempt to

. isolate the leak. The leak was verified to be on a one inch line in the

[ _"B" high point vent. system and not isolable. The plant.cooldown and

,  depressurization occurred in'an orderly manner. The plant reached a cold

! shutdown condition at 1838 hours on June 23, 1985, and secured from the unusual event. By the time the RCS was depressurized, approximately j 16,000 gallons of the coolant leaked into the reactor building. There , '

!. was no observable increase in the reactor building surface or airborne
 ' contamination due to the coolant. leakage.

' The: source of the leak was found to be a crack located on a segment of

 . piping that connects the "B" OTSG nozzle to the high point vent system and the nitrogen supply system for the RCS. The crack was a 120 degree through wall opening azimuthally about the pipe wall. The pipe was a one' inch diameter schedule 160 line. The licensee's investigation
         .
   .
<          f

1 * v w --,f- --- ----m <*r- r ---4 r y - -- -'--

[

'
.    -

    &
    .

subsequently identified that the "B" high point vent system had deficiencies in the as-built pipe support configuratio . Background Information -

    ,

The original construction of the, Rancho Seco Nuclear Power Plant. included two vent lines connected to the top portion of.the RCS hot leg pipin ~ The vent line on the "B" side included a.one inch diameter quality-Class 1, seismic Category 1 nuclear piping (Class 11N1) . traveling from

    ~

the hot leg steam generator nozzle to the,first isolation valve (RCS-017). The section of the piping,from'the, first isolation valve to the second isolation valve (RCS-Oll) was classified:as quality Class 2, seismic Category 2 nuclear piping (Class'22N2). The remaining portion of the nitrogen supply piping was classified' quality Class 3, seismic Category 3 (Class 33).

The licensee dispositioned a nonconformance report (NCR S-046) in October 1974, which called for the removal of a' portion of the Class' 33 piping in the nitrogen supply system. A removable spool piece was to replace that portion of piping and be removed during plant operation to prevent the potential contamination of the nitrogen supply syste The inspectors were not able to locate any modification package for this change. Therefore, it appears that the licensee did not perform the proper stress calculations to assess the effect of the modification on the Class 11N1 portion of the syste In response to a requirement of the NUREG-0737." Clarification of TMI Action Item Requirements", the licensee initiated a design for reactor coolant system vents. The purpose of the vents was to remove noncondensable gases from the RCS which might inhibit core cooling during natural circulation. A contract was issued to Bechtel Power Corporation to design three vent systems; two were for the "A" and "B" sides of the RCS hot legs and the other for a vent system to connect to the pressurize Pipe stress calculation No. 316 and the accompanying isometric sketch dated September 26, 1981, (Rev. I to Bechtel Job No. 12334) for the "B" side high point vent and nitrogen supply system identified the following changes to the existing piping: " Spring hangers at points 295 and 345 replaced with rigid supports."

This was to modify an existing guide hanger assembly described in drawing M-485 sheet 1-32 " Point 247 connected to 248." This change was also identified in Note 1 on the isometric as..." Field to weld rigid member between two flanges." This was to be placed in the nitrogen supply piping between points 247 and 248 to support the piping with the spool piece removed during plant operatio " Points 437 and 438 connected together by 1\"x1\"xk" angle." This change was also identified in Note 2 on the isometric as..." Field to

. - . - .. - - - . _ . .  . . -__ - .
.
*

install 1\"xl\"xk" L, 12" long data point 437 and 438 with two U-bolts as shown." "U-bolts at points 326 and 290 to be tight on pipe." This was for an existing spring hanger assembly described in drawing M-485 sheet 1-33 " Rigid Z restraint added at 290 and at 295." This was to add

horizontal restraints to the assemblies identified in items a and b abov Furthermore, a Bechtel project engineer sent a letter to the licensee on October 7, 1981, discussing the changes that were necessary to modify the existing "B" vent piping as noted in a through d abov The licensee installed the solenoid operated high point vent modification required by NUREG-0737 during the 1983 refueling outage. The work was

,  described and controlled by engineering change notice ECN-A-2934. The 1  existing vent and nitrogen supply piping support additions and j  modifications described in a through d above were included in changes to

drawings M-485 sheets 1-329 and 1-330 and in the stress calculation isometric sketch and were part of the drawing package of ECN-A-2934.

7 However, these support additions and modifications were not included or described in the actual ECN and were not installed. The quality control

,

inspection conducted after completion of ECN-A-2934 did not identify that i the additions and modifications were not installe . Licensee Investigation i As a result of the June 23, 1985, pipe crack in the RCS, the licensee's safety review committees, the Plant Review Committee (PSRC) and the i

 'Hanagement Safety Review Committee (MSRC), developed an extensive

, investigatio In scope, the investigation was to determine possible cause(s) of the crack, the potential generic implications, and the , corrective actions required with respect to the licensee's finding In respon e to the identification of the configuration difference between the "B" high point vent line and the design, the licensee initiated a . detailed walkdown of all safety-related piping modifications made since j the as-built verification required by NRC Bulletin 79-14, " Seismic Analyses for As-Built Safety-related Piping System" (IEB 79-14) (the

 " Post 79-14 Walkdown").
      ' Post 79-14 Walkdown The post 79-14 valkdown included all safety-related piping modifications installed since the as-built verification required by IE Bulletin 79-14. To. ensure thorough walkdowns, a detailed inspection procedure was' written and the walkdown crews were trained. The walkdown packages included both the stress' analysis calculations and construction drawing ~

This effort included 349 safety-related piping supports installed since the original IEB 79-14 effort and resulted in nonconformances

     ~
    ,

    . .  - ._ - - - . .. . .- .
 '
 .

' identified in 225 supports. Of these nonconformances, 148 vere determined by the licensee to be of a minor nature so as not to affect the function of the support. They were determined to be acceptable as is and only drawing changes were made to reflect the as-built conditio The remaining 77 nonconforming supports were evaluated by the licensee. Of those, 75 were reworked to meet the original design condition although they were determined to not effect system operabilit The remaining two supports were determined to exceed the code allowable limits for their related systems for a seismic design basis event. These supports were between the two solenoid valves on the A and B RCS emergency vent piping and involved missing frames needed for both horizontal and vertical restraint during a seismic event. The licensee has subsequently installed these two supports j to meet the original desig i Supplemental 79-14 Walkdown

       '

l Based on reviews of the scope of the original 79-14 walkdown, the licensee determined that their original walkdown should be expanded , to include all safety-related systems, and not just those emergency systems as originally performe To. perform the balance of.79-14 ' valkdowns (referred to as the supplemental 79-14 walkdown), the licensee developed a walkdown inspection procedure. The walkdown teams consisted of SMUD and contract engineers with expertise in the pipe design area who.were trained to the walkdown procedur The supplemental 79-14 walkdown included visual inspection of 401 lines consisting of over 1800 components (supports and valves).

Approximately 25 percent of the components were identified as having nonconformances. These were documented on 107 nonconformance ' reports. Many of the nonconformances were determined to be

acceptable and only drawing changes were made to reflect the

'

as-built conditio Approximately 255 supports were determined to exceed the code allowable limits for supports for a seismic design basis event and require fixes to meet the design requirements for supports. The

,

problems noted were of the same types as identified in the ' post 79-14 valkdowns and included items such as weld items, out of

tolerance support locations, out of tolerance gaps, configuration l

discrepancies, et It is noted that the licensee did not identify , any piping that exceeded code allevable limits even though the supports exceeded code allowable limit The licensee also performed an investigation of the pipe's failure ' mode, which included an elaborate metallurgical study of the pipe segment with the crack, and an analytical comparison of the calculated stresses needed to produce the crack versos the stresses calculated from the as found condition. The licensee also performed

-__ . - - _ _ _ _ _ _ _ _ - _ _ _ - - _ _ _ - _ _ _ . _ _ - - _ - -

.

,

a plant _ transient analysis on the possible consequences of a total severance of the one inch diameter pip The quality assurance organization performed three special audits: an investigation of the cause.of the "B" high point vent line break; an audit of the adequacy of the post 79-14 walkdowns; and an assessment of the supplemental 79-14 walkdown . 5. NRC Evaluation

   /

The inspectors reviewed the licensee's evalsation' of the failure mode for the vent system piping. The licensee's evaluation was, based upon an investigation conducted by General Electric Co., Nuclear Energy Engineer's Division (GE), for their determination of the cause of the crackin ,

  '

The GE investigation consisted of opt'ical and electron scanning microscopy, metallography, hardness measurements, elemental analysis, and gamma scanning. The results from the General Electric Co.'s tests concluded that the through wall circumferential transgranular crack was most likely caused by high cyclic fatigue. The metallugical evaluation appeared to be complete and comprehensiv The inspectors also requested the licensee consider a mechanical vibration calculation of the as-found condition of the "B" high peint vent system to verify that the configuration did have the potential to

     "

generate the stresses needed for the pipe failure. The inspectors reviewed the results of the calculation, performed by the licensee, and agreed with the determination that the piping system was sufficiently unsupported to create the cyclic load required for the pipe crac Based on the recent piping failure of the RCS vent line, the licensee's actions for IE Bulletin 79-14 were reviewed. This IE Bulletin (issued on July 2, 1979 with subsequent revisions and supplements) requested verification that the seismic analysis reflect the actual, as-installed configuration of safety-related piping systems. The Bulletin had been previously reviewed in Inspection Report Nos. 50-312/80-04, -09, -24, and -3 The areas reviewed as followup to the licensee's activities for IE Bulletin 79-14 were: coverage of systems by the licensee, followup to selected I.ER's, review of the re-analyses of piping systems based on IEB 79-14 walkdowns, review of QA audits concerning the bulleti In addition, the licensee's QA surveillance of the engineering walkdowns, dispositioning of nonconformances, and pipe failure mode analyses were reviewe Guidance was provided in the IEB 79-14 as to which safety-related piping systems were to be considered under the bulleti Safety-related piping was to include Seismic Category 1 systems as defined in Regulatory Guide 1.29 or the FSAR, and also Seismic Category 1 piping regardless of size if it had been dynamically analyzed by compute . ._ . _ _ _ __ _ ._ - . _ __

 .
*

6 - , The S?fUD engineer who was cognizant of the previous lEB 79-14 1 actions stated that the philosophy that was used in the selection of systems was that only those portions of systems were analyzed which were engineered safety features systems (i.e. , high pressure injection, reactor building spray, core flood tanks and piping, decay heat removal, etc.). In addition, it was assumed that no piping under 2\ inches in diameter (small bore) had been dynamically analyzed by computer. Upon further review (including the licensee's submittals to the NRC dated July 30, September 5, October 31, 1979 and August I and October 10, 1980) by the NRC and SMUD, it was determined that the intent of the IEB 79-14 had not been met, in that there existed piping aystems which were Seismic Category I systems as defined in the licensee's FSAR Appendix 5B which were not d part of the previous scope of bulletin activities. Examples of such systems were: primary coolant system components, including vent and drain piping inside containment.;' main steam and feedwater piping up to the stop valves; atmospheric dump and main steam safety valves and associated piping from the main steam headers; and portions of the spent fuel cooling system.' Also, it was; determined that some small bore piping had been computer analyzed and had not been a part

;  of the IEB 79-14 review. These systems include <the RCS vent line
     -

piping and pressurizer vent line pipin The portion of system piping which significantly contributed to the cause of the recent pipe leak was the. nitrogen supply system. The original design of this system was a straight run of pipe without the use of a removal spool piece. The' design had been dynamically , analyzed by computer. A nonconformance report was initiated in 1974 j to preclude the potential for contaminating the nitrogen system by

;  designing a removal spool piece into the system to physically l  separate the nitrogen system and the RCS during normal operation l  The addition of the removal spool piece required a change in the i

piping geometry in the vicinity of the spool piece additio However, there were no engineering documents to show that the original analysis was revised and properly reviewed. This condition was postulated to have existed since 1974. Since the actual

as-built condition (spool piece and piping geometry change) did not

!  match the seismic analysis, if this line had been subjected to  '

- IEB 79-14 inspection as intended, there was reasonable assurance j

'

that the necessity of a replacement spool piece during operations would have been detected and corrected, possibly preventing the overstress condition which ultimately resulted in the piping i failure, During the previous IE Bulletin 79-14 activities, the licensee had described two instances in Reportable Occurrences 79-14 and 79-15 (transmitted to the NRC on November 9 and November 16, 1979 respectively) in which the as-built configuration did not match the , seismic analysis. Analyses had indicated that in the as-built  !

'

conditions, allowable design loads and stresses would be exceeded ' l under seismic conditions. Additional pipe supports and the i modification of existing supports were required. The associated Work Requests (39375, 40643, 43370, 43047) and support drawings were

. -_-- -      __ . _ - .

c:

.   .
*

7 , ,

    , .

reviewed and found to reflect the proper installation, modification and inspection of the affected support i

    <

c. The previous reanalyses of.pipi'ng systemsf were reviewed to determine whether the Bechtel scope of responsibilities was properly implemented. From the SMUD submittal to the NRC (Region,V) dated September 5, 1979, three stress calculations which required" reanalysis were reviewed (stres's calculation No. 8,l76, and 106).

In addition, the computer input sheets for these three stres calculations were reviewed. The reanalyses were found; except in one instance, to properly reflect-the as-built ~ conditions determined in piping system walkdowns. In the one instance, it appeared that the stress isometric sketch for calculation No. 8/106 correctly located data point 185, but the computer! input' sheet was missing data point 185. Therefore, a short span'(l'-10") of 2\ inc diameter piping was not correctly modelle ' - In addition,'in stress calculation' No. 8/106, an 11 foot . span of 3 inch diameter pipe was modelled without a midpoint data point (between data points 240 and 245). Since masses and stresses are calculated at the data points only, the omission of a data point could under-estimate piping stresses and support loads. In Bechtel's preliminary estimation, the midpoint data point could have been legitimately omitted from the analysis as the span's natural frequency was approximately 30-33 H After the inspectors identified this concern, the stress calculation was redone using the proper coding of data point 185 and an intermediate data point in the 11 foot span of piping. The results of the re-analysis showed that pipe stresses remained low (significantly below stress allowables), and pipe support loads were within design allowabl It was shown by reanalysis that the discrepancies identified in stress calculation 8/106 were of a minor nature in that stresses remained within design allowables. The inspectors concluded that, although some discrepancies were identified in the review of reanalyses done for IEB 79-14 walkdowns, no additional review was necessary based on the minimal effect on stresses and load d. Two SMUD audits of the Bechtel activities for IEB 79-14 were reviewed. Audit No. 0-284 and 0-312, dated March 6 and July 22, 1980 respectively, were found to adequately verify the Bulletin responsibilities of Bechtel (i.e., preparation of stress problem package for all designated lines, evaluation of walkdown inspection results, reanalysis of lines if required, and proper disposition of discrepancies). Since it was not Bechtel's responsibility to , designate the lines or systems to be covered under the Bulletin, ' these audits would not be expected to have identified the fact. that ' there was an inadequate scope of review of systems as previously described.

' e. The quality assurance organization conducted a surveillance of the engineering walkdown teams by independently verifying the as built-condition of seven systems. The seven systems were selected by QA at random. The surveillance included inspection of the as built

_ -. - -- -- _- . . ._

*
.

g drawings for pipe routing and correct pipe support location, , reconciliation of the piping stress isometrics with support hanger drawings, and the evaluation of the engineering walkdown teams' findings by the QA organization. Seven lines were walked down as follows: SYSTEM CALCULATION N LOCATION Feedwater System CAL #49b Auxiliary Building

Reactor Coolant System CAL #98 Containment Building Seal Water Injection System CAL #45 Containment Building Letdown System CAL #113 Auxiliary Building

.i Main Steam System CAL #201 Turbine Generator Bldg Borated Water System CAL #75 Tank Farm / Auxiliary ! Bldg Reactor Coolant Drain System CAL #120 Containment Building l ! j The quality assurance walkdown was accomplished with assistance from stress analysts from the Stone & Webster Corporation. The seven packages reviewed were originally walked down by six engineering walkdown teams; therefore, the sample reviewed by the quality assurance organization was comprised of twelve members of the engineering walkdown tea , l The QA surveillance identified a discrepancy with the engineering walkdown teams' reviews. This discrepancy concerned a support on the 10" main steam line to the turbine generator reheater. The j specific problem involved a support which was restraining the line ' in a direction which was not assumed in the original stress analysis. The discovery of this restraint in the improper direction was not identified by the engineering walkdown team because the hanger drawing which was used for construction and for the walkdowns contained several details of standard hangers used at the sit However, only one of the standard details which appear on the hanger drawing was to be installed. The other typical support - configurations shown on the drawing were not to be use Apparently, during construction also,.the correct hanger to be used was not clear, and, therefore, an incorrect configuration was fabricated and installe , This discovery by the quality assurance audit was made by checking the hanger drawing against the piping stress isometric. The piping

stress isometric did show the directions in which the pipe was to be restrained as assumed by the stress analysis. The problem which surfaced appears to have been caused by the use of a standard , support hanger drawing by the walkdown team which contained several standard support configurations but only one of the. configurations was to be used. The support configuration to.be used was labelled with a title and therefore should have been the only configuration used. Neither the seismic stress isometric ~ nor the -labelled hanger drawing detail was taken to the field by the field walkdown team, t As corrective action for this deficiency, identified by quality assurance, the engineering organization began a re-review of all

    .
   %

t

-_- . -.- - - . - .. -_ . . . . - - . _ . - - ~ . - - _
'
.

packages which were issued to the engineering walkdown teams. This review was to identify any similar support drawings which show standard details, are not labelled, and therefore, should not have been fabricated and installed. If one of these support drawings ' should be discovered, licensee personnel stated that a reinspection will be performed to verify that only the correct support was installe The inspector concluded that the QA surveillance appeared rigorous and it identified a significant finding which was being followed u , The licensee generated a procedure for controlling the review and dispositioning of noncenformance reports generated by the engineering walkdown team. This procedure (flERPT-M0002) described the steps which the stress analyst / pipe support designer took to review findings identified by the engineering walkdown tea , Responsibilities for the individuals participating in the review were defined and established responsibilities of the supervising enrineer, the initiating engineer who performed necessary calculations to resolve any nonconformances, and the checking enginaer who would verify the assumptions and completeness of the initiating engineer's wor The p. cedure also included the piping evaluation criteria, such as the analysis methodology, modeling techniques, and computer program * to be used (the standard Bechtel pipe stress analysis code HE101 Version K 2). Also included were the Piping Overlap Method to be used (should it be necessary because of extremely large piping), and acceptance criteria such as allowable piping stress, nozzle loads, valve accelerations, and pipe support load criteria. Load combination tables for the, analysis were also contained within the procedur The inspector concluded-that the procedure provided the necessary guidance and information to the analyst to perform the evaluation in a controlled fashio . V The inspector interviewed'three engineers involved in the NCR review process. The individuals interviewed were performing in process evaluations of discrepancies identified bycthe engineering walkdown team on the following three systems: ' Main Steam to the Main Feedwater Pump, the Reactor Coolant Drain System, and the Feedwater ; System. Topics discussed included the specific work the engineer ras evaluating, such as redesign of a pipe support or reanalysis of a piping system to determine maximum postulated stresses within the piping system. Discussions were held with a pipe support design engineer, two stress analysts, and four team group leaders. During these discussions the in process work was~ discussed such as applicable piping codes, modeling techniques, applicable welding codes, nozzle allowable loads for large equipment such as pumps and heat exchangers, and geometry data input to the piping stress analysis computer cod !

      !

f

      !

i !

.. _   . . . = ..- . ._  -- .
   *
.

, The inspector determined that all engineers were familiar with the applicable codes used and analytical techniques employed for this , reanalysis. The stress analysis being performed for the main steamline was adequate and in agreement with the licensee's i procedure and design criteria documents and industry practice for the portion reviewe < The Rancho Seco plant was originally designed under the rules of ! ANSI-B31.7. During the review phase, the licensee elected to perform their review under the. rules of ASME Section III. The c licensee has elected to use a more current code in order to expedite ' a more efficient review pro' cess,'in that procurement of pipe support parts under the rules of ASME were readily'available. In addition, the engineers performing the' review were more familiar with the

rules of ASME (current day piping code) than the rules of ANSI-B3 '

       (a pipi'ng code no longer used _ in the design of nuclear power i        plants). This decision by the" licensee would also increase the conservatism of the analysis slightly, due to the increased requirements within the ASME Code on stress intensification factors used for various pipe fittings such as tees.

j The licensee indicated that some of their reanalyses may require use of Code Case N411. The licensee has requested that the Office of

;        Nuclear Reactor Regulation approve the use of this Code Case.

The inspector concluded that performing these analysis under the rules of ASME Section III was adequate for the IE Bulletin 79-14 analyses which were performe g. To substantiate high cycle fatigue failure of the "B" hot leg event, the licensee performed a hand calculation to assess the fatigue life a of the "B" hot leg vent. This calculation was performed to l' substantiate the findings that the "B" hot leg vent mode of failure

             '

was high cycle fatigue due to a improperly supported small bore lin The licensee elected to use a procedure for calculating the i vibrational fatigue life of the piping system as described in , i ANSI /ASME OM3-1982, titled " Requirements for Preoperational and l Initial Startup Vibration Testing of Nuclear Power Plant Piping Systems". Using this document as guidance, the licensee then established a procedure for determining the permissible number of cycles before failure would be obtained. This procedure required assumptions to be made on the predominate frequency of the piping system. The frequency is needed in order to determine the fatigue life of the piping syste The predominant frequency was calculated by the licensee from the dynamic analysis performed on the piping system. Therefore, the value generated by the computer program was a natural frequency of

;

the piping system at rest rather than the predominant frequency of system in operation. Currently, no data exist for the Rancho Seco , facility which would allow the determination of the predominant frequency of the system with all reactor coolant pumps runnin *

             \

_ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ . _ . . _ . _ _ _ _ . . . . . _ _ _ _ _ _ _ . _ . _ _ . _ _ . . _ _ .

*

.

Therefore, the licensee has utilized a best estimate approach of using the natural frequency of the piping syste The inspector concluded that although this method for determining fatigue life of the "B" Hot Leg Vent was not the most rigorous method available, it was probably the only method available with the limited amount of data currently availabl The licensee's analysis showed that acceleration levels for the "B" hot leg vent greater than 0.633g could produce stresses at the nozzle connection which were beyond.the code stress limit for steady state vibration. The calculated acceleration vould produce a deflection of the piping system of approximately 0.59 inches. The , estimated deflection of the piping system in the as found condition ' after the leak had been discovered was approximately 0.9 inche Therefore, the licensee concluded that the maximum deflection the

,, piping system could tolerate from an assumed acceleration was exceeded with the as'found piping configuratio The inspector concluded that the piping system,. as found, might not have tolerated the current configuration without ultimate failure of the pipty system. However, this calculation method.was a,very simplified approach, and would onl/ yield preliminary result This method supported previous conclusions by the licensee that failure of the hot leg vent was caused by,high cycle fatigu . Conclusion The inspection clearly indicated that a quality assurance system failure
   ~

had occurred in the "B" high point vent's design, construction, and inspection. Necessary design information was not properly. included in the design package even though this information was detailed on the pipe stress isometric as well as the October 7, 1981 letter (mentioned above) that were transmitted to the licensee. Lack of control over the construction work that installed the modification was evidenced by the omission of several pipe supports. The quality control inspection performed was not detailed enough to identify the conflict between the as-built configuration and the design. Finally, the quality assurance organization's program of review and surveillance did not identify any of the deficiencies that plagued the "B" high point vent modificatio The nitrogen supply and vent header systems connecting to a one inch reactor coolant system vent located on the high point of the "B" hot leg were not identified as systems to be covered under the licensee's QA program required by 10 CFR 50, Appendix "B". These systems, therefore, were not designed, constructed and inspected pursuant.to QA Class I requirement On June 23, 1985, a crack was identified in the above mentioned reactor coolant vent line that resulted in a 17 gpm primary coolant leak that was non-isolable. The cause of this condition was determined to be fatigue from high cycle vibration, induced by the nitrogen and vent header systems because of design and construction errors in 1974, 1981, and 1983

.

. ..* g that resulted in missing pipe supports, improper pipe supports and a missing pipe section as described below:

* missing cross-brace between lines'20574-1"-HE and 20555-\"-CA
*

missing rigid flanged spool piece thatlshould have been Installed in the nitrogen supply line during operation 'of :the plant

   ' '
     '
*

missing east-west stops on support numbe l'20'67-3 S 5 ,

     ,
    '
     ^
* missing east-west stops on suppor't tiumb'er IS$0567-2 l '
* no replacement of vertical spring support with req'uired stops on
    '  '

support number IS20567-2 The above deficiencies in the sys[ ems were identified 1n a stres ~ analysis performed by the licensee's contractor in 1981 in~ preparation for addition of the high point vent modifications which were made to the system in 1983. The required piping supports were not installed, even though the need for these piping supports was transmitted to the licensee by letter from their design contractor on October 7, 198 In addition, the design drawings for construction of the reactor coolant system 3/'+ inch high point vent lines on both the "A" and "B" sides specified seismic frames located between the two sets of solenoid valves, but these supports were found missing in June 1985. Numerous additional examples of similar deficiencies in safety-related systems were identified and documented in the licensee report to NRC's Region V office dated August 6, 198 Similar concerns had been identified by previous inspections. The licensee demonstrated an inadequate control of the design and installation of auxiliary feedwater flow transmitters in Inspection Report 84-1 Inadequate construction and ineffective quality contro' inspections associated with the Nuclear Service Electric Building were discussed in Inspection Report 85-0 The inspectors concluded that the previous and present findings had showed the same pattern of performance in the engineering, construction, quality control and quality assurance area Enforcement action related to the issues discussed in this report will be the subject of separate correspondenc . Exit Meeting The inspector met with licensee management personnel denoted in paragraph I at the conclusion of the inspection on August 9,1985. The scope, observations, and findings of the inspection were discussed. }}