IR 05000267/1989019

Insp Rept 50-267/89-19 on 890821-26 & 29-30.No Violations or Deviations Noted.Major Areas Inspected:Region 19 CRD Assembly Recovery Activities,Followup on Region 19 CRD Clevis Bolt Failures & Steam Generator Ring Header Problems

ML20248H780
Person / Time
Site:	Fort Saint Vrain
Issue date:	10/02/1989
From:	Barnes I, Stewart R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20248H778	List: IR 05000267/1989019 ML20248H775
References
50-267-89-19, NUDOCS 8910120072
Download: ML20248H780 (12)

Text

'

. . . ,

, ,

.

,

APPENDIX U.S. NUCLEAR REGULATORY COMMISSION

REGION IV

.

NRC Inspection Report: 50-267/89-19 Opeiating License: DPR-34 Docket: 50-267 Licensee: Public Service Company of Colorado (PSC) j P.O. Box.840 Denver,' Colorado 80201-0840 Facility Name: Fort St. Vrain. Nuclear Generating Station (FSV)

Inspection'At: FSV, Platteville, Colorado Inspection Conducted: August 21-26 and 29-30, 1989 4 k

Inspector: / M' /o/2/G7 R. C. Stewart, Reactor Inspector, Materials Date

{

anct Quality Programs Section, Division of j Reactor Safety j Approved: V w, m 4/e9 1. Barnes, Chief, Materials ano Quality Date Programs Section, Division of Reactor Safety j

Inspection Summary l

Inspection Conducted August 21-26 and 29-30, 1989 (Report 50-267/89-19)

Areas Ins)ected: Nonroutine, announced inspection of' Region 19 control rod drive (CRJ) assembly recovery activities, followup on Region 19 CRD clevis bolt failures, and review of steam generator ring header cracking problem Results: Licensee recovery activities for the Region 19 CRD assembly were <

found to be well planned and executed, with detailed attention given to both !

assuming the viability of the recovery methodology and ALARA considerations, i Preliminary review of the Region 19 CRD clevis bolt failures identified the absence of quantitative installation tightening criteria for these bolts. The absence of such criteria, coupled with imposed tensile stresses resulting from -l differing thermal coefficients of expansion for the clevis and clevis bolt materials, are currently considered to be the most probable cause of the clevis 1 bolt failures. The licensee is continuing to investigate the root cause of the j failures, including a hot cell metallographic examination of one of the i fractured bolt heads. Review of the steam generator main steam ring header 8910120072 891003 PDR ADOOK 050002e7 O FDC ,l ( -

p 61 ,

, . .

.

-2 cracking problems confirmed that the cracking was characteristic of elevated temperature tertiary stage (i.e., final accelerating phase to failure) cree Examination of typical stress rupture data for annealed Incoloy 800 material indicated that creep failure should not have been a credible event for the estimated operating life at a temperature of 1000*F. Additional review was not performed to establish whether local steam ring header inlet temperatures could have exceeded 1000 F. The results of review of ring header design calculations will be documented in NRC Inspection Report 50-267/89-2 No violations or deviations were identified in this inspectio i i

l

--_- -___ - _ - - I

~

,

.

..' . <

t

-.- n< . ,

'

u ,

-3-g y*<

. T-DETAILS

< Persons Contacted

,.

4.PSC ,

• C.Fuiler, Manager,NuclearProduction

< Evans, Operations Manager L. Hutchins,-Health Physics Supervisor ,

D. Johnson, Systems Engineering Supervisor ,

t D. Alps,* Security Supervisor M. Ferris, Qual.ity Assurance (QA) Operations Manager R. Craun, Manager, Nuclear Site Engineering-T. Mcintyre Superintendent Materials Management

• • P. Harrington, NED Supervisor, Materials Engineering

., **M. Deniston, Superintendent of Operationt

. **B. Ring,;QA Engineer - -_

>

• • R.' Hellner, QA Materials Engineering Supervisor-

• • D.rScott, QA-Services Manager

'

• • D. Frye, Senior Nuclear Licensing Analyst

.

-

i

• • B.LBrunsdon, Systems Engineer ,

P. Tomlinson,-QA Division Manager

'

NRC

• • R. Farrell," Senior Resident Inspector P. Michaud,' Resident Inspector '

• Denotes the senior licensee representative present at the informal exit i meeting held on August 26, 198 '

• • Denotes those persons that attended the exit meeting on August 30. 198 'The NR,C inspectors also contacted other licensee personnel during the

,

performance of the inspectio _

2 '. . ' Observation of Region 119 CRD Assembly Recovery Activities- (92701)

During the performance on August 17, 1989, of weekly scram surveillance testing of CRD assemblies, the Region 19 control rods failed to insert as required. The licensee declared the Region 19 CRD assembly inoperable and commenced plant shutdown.- After shutdown, additional testing was

. performed of the Region 19 CRD assembly which finally resulted in no rod motion being attainable. LAdditional details of the results of troubleshooting are documented in NRC Inspection Report 50-267/89-16.

<

f rt

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

'

hg '

,

jN~- ,

.

.

-

,

2.1 Attempted Retraction of the Region 19 Control Rods Using the Overhead Crane f

. As a result of the Region 19 control rods being 120 inches below the fully retracted position, the licensee was unable, to use the auxiliary transfer cask (ATC) for removal and, transfer of the CRD to the hot service facility (HSF). The ATC is sized for insertion of CRDs with. control rods in the fully retracted position. Dn August 22, 1989, the licensee attempted to raise the Region 19 control rod pair to the' fully retracted position using the' overhead crane, in order to, allow ATC us . . Disassembly Activities and Preparation for Re;noval

-

Details of.the CRD. assembly are depicted in Attachment To obtain access to the CRD' cables for connection to,the overhead crane required initially the unbolting of the CRD penetration primary seal

, and raising of the CRD assembly approximately 3 feet above the reactor. isolation valve. Prior to start of this activity, the licensee developed and issued a detailed procedure which included hold points for Quality Control (QC), Health Physics (HP), and Engineering sign-offs. In additioncto review of.this procedure, FHPWP-154 dated August 22, 1988, the inspector made the following observations:

The Region 19 penetration was cordoned off with positive HP

"

control of personnel access under a specific work permit for that are *

Access to the reactor fuel deck was also limited to specific personnel operating under a work permit.-

The cognizant lead supervisor gave a briefing to all involved personnel prior to commencement of work regarding significant work objectives and safety criteri A check stand, with an inventory log, was maintained for all tools and material entering or leaving the Region 19' penetration are The calibration status of test equipment and tools was verified to be within required due dates, including the dynamometer crane gage and electronic load cel !

Core reactivity was monitored and recorded throughout the CRD movement activity, with no change observe . Final Work Activities After the CRD assembly was raised 3 feet above the reactor isolation valve, an inspection was performed of the CRD cable drum assembly and

____________-____-_A

,- . - - - -

_ - - - - - -- - _- - -

e y ,

e

-

n ;

-5-s

>

the orifice drive mechanism. This. inspection did not reveal any

'

malfunctions or evidence of cable binding. Both control rod cables were then securely clamped, cut, and 25 feet cable extensions swaged to the existing cables (using a 40-ton swaging tool) to facilitate

_

'

use.of the overhead crane for the cable pulling activity. Maximum

-

load cell readings were pre-established at 1600 foot. pounds for the cable pulling' activity. On initiation of the' cable pulls one of the control rod pair raised easily with an applied force of 200 foot pounds. The companion rod bound, however, after a cable movement of approximately 6 inches, with the force gradually increasing to 1500 foot pounds. At this point, the licensee' ceased the retraction effort, secured the CRD assembly, and prepared for recovery using a specially fabricated shielded cas .2 Final Recovery of the Region 19 CRD and Transfer to the HSF 2. Special Transfer Cask and Rigging In anticipation that control rod' retraction activities could be unsuccessful, the licensee fabricated a special lead shielded cas . A sketch of this cask is shown in Attachment 2. The cask was fabricated from 20 feet long standard 24 inch and 30 inch welded i steel piping, thus resulting in a 23 inch ID and an annulus of '

2 1/2 inches. The annulus was filled witn lead shot through two pipe nipples welded into the top annulus plate. The bottom annulus plate was extended beyond the side of the cask in order to accommodate a remotely operated slide gate (present for retention of loose contamination and loose parts). The inspector reviewed the fabrication details of the cask and attachment welds made on the ]

overhead crane trolley. All weld designs indicated a safety factor

'

of 2 or bette As noted in Attachment-2, the cask was supported from the overhead crane trolley by two nylon strap slings, a steel cable sling, and a ten ton chainfall on each side of the cask. The chainfall and sling )

combination was " proof load" tested on August 25, 1989, prior to the J Region 19 CRD assembly recovery activity by lifting a known weight of -

ten ton . Verification of Equipment Operability and Applicable Dase Lates Prior to the recovery attempt of the Region 19 CRD, the Region 15 CRD (which had a similar in-core service history to the Region 19 CRD) was removed from the core utilizing the ATC and placed in the HSF for dose rate surveys. The surveys indicated a maximum dose rate of approximately 2200 Rem / hour at 6 inches from the surface of the )

absorber string. A simulated recovery exercise was then conducted {

using a spare CRD assembly. This exercise included establishment of i alignment markings, use of remote TV monitors, and establishment and I maintenance of a 3 inch vertical clearance below the buttom of the l I

l

- - - - - - - - _

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

!. -

. .

L  ;

<

!

'

I --6-V cask to allow remote TV monitoring of the CRD alignment and recovery operations. The exercise was successfully performed. ALARA and recovery team briefings were then conducted prior to the actual recovery of the Regicn 19 CRD assembl . _ Final Recovery of the Region 19 CRD Assembly

. During the evening of August 25, 1989, the Region 19 CRD assembly was successfully removed and transferred to the HSF without inciden ,

i The inspector observed that all activities were conducted in a safe  ;

manner, with the precautions described in the " Safety Evaluation for Control Rod Pull Using Reactor Building Crane," dated August 25, 1989, adhered t No violations or deviations were identified in this area of the inspection. Additional information on this subject is documented in NRC l Inspection Report 50-267/89-16, 3 .' . Followup on Region 19 CRD Clevis Bolt Failures (92701) l After successful transfer of the Region 19 CRD rod pair to the HSF, an inspection was performed to determine the cause of ane of the absorber strings becoming immobilized. This inspection revealed that the head of ];

the self-retaining bolt (which connected one of the control rod cable assembly ball ends to the absorbar string clevis) had broken off and {

dropped into the guide tube, resulting in jansning of the absorber strin I The inspector was informed by licensee personnel that the actual fracture {

location was at the square' groove in the shank of the bolt which contained j the bolt retaining element. Inspection of the self-retaining bolt in the  ;

clevis of the other absorbs string identified that the head was also {

missing from this bol ;

i 3.1 Procurement of Clevis Bolts J i

The inspector ascertained that the failed bolts were installed during '

June 1985, in conjunction with cable replacement and control rod drive mechanism refurbishment activities. Drawing D'-1201-179, Revision C, was revised per Change Notice (CN) - 1933H to alter the selected bolt material f rom AMS 5737 (Type A286 precipitation hardened, high strength stainless i steel) to AMS 5667 (Inconel X-750) with a minimum precipitation hardened )

ultimate tensile strength of 165,000 psi. The choge permitted the i continued use of the stainless steel bolts if Inconel X-750 bolts were not available. Work records established that Inconel X-750 bolts were installed in 1985 in the Region 19 CRD rod pai Review of the vendor certified material test report (CMTR) applicable to the Inconel X-750 bolts showed reported ultimate tensile strength values j which were in excess of the 165,000 psi required minimum; i.e.,

209,060 psi, 210,553 psi, and 213.540 psi. The location of fracture for the tensile tests was reported as "urdercut" which was interpreted by the  !

I l

l

_

)

_ - _ - _ _ _ _ _ _ _ _ _ _-_ _ _ ,

. ,

,

.. . .

-

-7-inspector to mean the bolts failed at the retaining element groove in the bolt shank. No ductility or yield stress values were reported for the tests. The inspector additionally noted, as shown below, that the recorded heat treatment information was not in accordance.with the requirements of AMS 566 AMS 5667 Requirements Heat Treatment Performed Equalization - 1625 F 1 ?5 for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> 1800*F for 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> Precipitation Hardening - 1300*F 1 25 for 1350*F for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, furnace 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> i 1 cooled 100 F/ hour to-1150 F, held 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> at 1150*F Included in the data package for the bolts was GA Technologies Nonconformance Report (NCR) 10119. This NCR identified that GA Technologies originally invoked AMS 5670 as the applicable Inconel X-750 specification, but received AMS 5667 from the vendor. This was dispositioned "use as is" by GA Technologies without apparent recognition that the vendor had used an inappropriate heat treatment for AMS 566 Licensee personnel informed the inspector that the maximum service temperature the clevis bolts were exposed to was approximately 750 If this estimate was correct, the incorrect heat treatment would not be expected to have been a contributory factor to the subsequent bolt failures. No specific data was seen by the inspector which would confirm the accuracy .of. this information. The inspector also questioned the value as a result of recalling from a previous inspection that the maximum design temperature for the control rod cable had been identified by PSC to be approximately 1200 .2 Clevis Bolt Installation Criteria The inspector reviewed Fuel Handling Procedure Work Package FHPWP-100-41 to ascertain the criteria that were applicable in 1985 for installation eT clevis bolts. This review identified that torque values were not specified for this activity, the instructions simply requiring the bolts to be tightened with an installation tool until shouldering occurred against the clevis. This approach conceptually permits application of any level of tensile stress below that which would result in either stripping of the bolt from the retaining nut or bolt fracture. Actual stresses I applied to the clevis bolts cannot thus be determined with any degree of certaint .3 _E,ffects of Differing Material Thermal Coefficients of Expansion One potential contributor to bolt failure that was under active i consideration by PSC personnel during the inspection related to the  !

effects of differing thermal coefficients of expansion of materials used j l

i

_____-____--_-__-__-__-____________A

'

. ,

• .

• - .

,

F-8-in the clevis assembly. The clevis was manufactured from Type 321-austenitic stainless steel, with the ball end of the control rod cable and the clevis bolt being, respectively, Inconel 625 and Inconel X-750. The thermal coefficient of expansion of the austenitic stainless steel is approximately 25 percent higher than the Inconel alloys. The greater thermal growth of the clevis prongs during heating would.thus be expected to impose additional tensile stresses on the clevis bolt, with the level of applied stress being proportional to the maximum temperature attaine .4 Root Cause Analysis Licensee personnel informed the inspector that it was their intent to establish the root cause for the bolt failures. As part of this effort, the head of the bolt which caused jamming of one of the absorber strings was retrieved and sent to GA International Services Corp. for failure anlysis. Licensee Event Report (LER)89-015 committed to provide the results of this analysis in a supplemental repor i No. violations or deviations were identified in this area of the ,

inspectio . Review of Steam Generator Ring Header Cracking (92701)

During the reactor shutdown resulting from the Region 19 CRD problem, a maintenance followup was made regarding an observed water drip below a steam generator module; The leak, which was originally attributed to be a valve packing leak, was found by the mechanics to actually originate from the Incoloy 800.B-1-4 main steam ring header. After removal of insulation, a through wall crack approximately 5 inches in length was discovered which was located adjacent to 1 of the 18 steam entry nozzles  !

(i.e., Nozzle N-118) welded in the header. A liquid penetrant examination {

was then performed of the 12 steam generator module ring headers. This l examination revealed linear crack like indications were present in header '

base material in the vicinity of 36 of the 216 steam inlet nozzles present .

in the 12 ring headers. The linear indications were confined to 7 of the j 12 ring headers, with the majority of indications being associated with l nozzles from the same 180' segment of each header; i.e., 29 out of the 36 i nozzles exhibiting adjacent linear penetrant indication .1 Results of Metallurgical Examination of Boat Samples Removed From Ring Headers B-1-1 and B-1-4  :

A boat sample was cut from the B-1-4 ring header to remove the through wall crack for metallographic examination. An additional two boat samples were cut from ring header B-1-1 in areas adjacent to Nozzle N-117 where )

penetrant examination had detected widespread linear indication Sections were cut from the boat samples and then examined using conventional metallography and a scanning electron microscope. The examinations revealed that the through wall crack was intergranular in

- ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ -

y1 y .

-

.

_

-

e

' w \

n p ,

,

pg 4 - 3f ,

-9-

'

, ,

n n .

"

nature with oxides present on the crack surfaces. Intergranular cracks wereipresent.throughout the sample and examination at higher-magnifications revealed both microvoids distributed along grain boundaries -d and microcracks formed by coalescence of voids. These metallographic features are characteristic of elevated temperature tertiary stage (i.e.,

final accelerating phase to failure) creep cracking. Similar observations were made with respect to sections cut from the two boat samples that were removed from the B-1-1' heade .2 Ring' Header Material and Forming Practice The inspector reviewed the vendor (Semitomo) CMTR that was applicable to the single heat of Incoloy 800-(ASME Material Specification SB-407, Grade 1) 5.563 inch OD, 0.75 inch wall, pipe material used to fabricate the ring headers. A copy of the material specification revision that was in effect at the time of the 1968 procurement was not available. The examinations performed and reported properties were, however, consistent with the requirements'ofilater versions of SB-407. A check analysis was also perfomed by PSC, of a ring header sample which confirmed the composition of the material was in conformance with SB-407 requirement .The inspector reviewed the forming requirements contained in Stearns - )

Roger Specification SR-PP-105, " Cold Bending Specification for SB-407

'

Gr. 1 Pipe." This specification required packing of the pipe with dry sulfur free sand, heating of the pipe in a slightly reducing sulfur-free atmosphere to 950-1000*F, followed by bending and cooling to room temperature in still air. This specification, while containing good practices for processing'of Nickel based alloys, provided no information l regarding the forming methods or equipment used. The ring headers have a l diameter of 51 inches, as measured from the center line of the pipe. The )

inspector calculated the fiber strain created by the forming process, using an ASME Section III Code equation, with the value obtained being 10.9 percent. This value appeared to be of a magnitude that would normally result in a fabricator performing an annealing heat treatment subsequent to the forming operation. No information was present in the fabricator data package to indicate that this heat treatment was perfomed. Technical data was not available to the inspector during the inspection that would allow assessment of the effects of the forming percent strain and preforming temperature used on the elevated temperature creep properties of the Incoloy 800 ring header materia .3' Review of Creep Properties of Incoloy 800 The inspector ascertained that nominal header steam temperature and pressure durf r.g operations were, respectively,1000'F and 2400 psi Estimated operating hours at temperature were indicated to be 38,000 by licensee personnel. Review of published typical stress rupture data for annealed Incoloy 800 showed that to cbtain a 40,000 hour0 days <br />0 hours <br />0 weeks <br />0 months <br /> rupture life at 1000'F would require an applied stress of approximately 33,000 psi. In that the desigt calculations indicated an allowable stress of 13,600 psi j i

i

---____ ____

i

.

!

.

, .- .

!

-10-in the ring header, it appeared to the inspector that creep failure should not have been a credible event at 1000*F operating conditions. As a I result, and factoring in the information obtained relative to header cracking locations, the inspector questioned licensee personnel regarding their knowledge of the uniformity of steam temperature in individual steam ,

generator module tubes. Steam temperature was measured only at the ring i header outlets and it thus appeared possible to the inspector that i individual steam inlet temperatures to the ring headers could exceed (

1000*F with a mean outlet temperature of nominally 1000 F after mixing being maintained. Licensee personnel indicated that they did not believe significant temperature variations could occur in individual steam generator module tube '

As a result of the PSC announcement on August 29, 1989 (at the beginning of this part of the inspection) that a restart of the plant would'not be made, the inspector did not review subjects such as feedwater flow balance, controls for regulating feedwater flow to individual steam generator tubes, and ring header outlet temperature ranges maintained

during reactor operatio .4 Review of Design Calculations j i

The inspector requested the resident inspector to perform a review of the main steam ring header design calculations in order to establish whether the initial design was contributory to the early header failur Preliminary review by the resident inspector indicated little er no margin was factored into the design analysis. For example, the design i temperature of 1050 F is extremely close to the nominal operating temperature of 1000'F and no allowance was apparently made for temperature measurement inaccuracies, flow distribution variations, or accident and transient conditions which could exceed 1050 F. In addition, the stress values calculated by using 1050 F were very close to the maximum allowable stress values. Further review of these calculations will be documented in NRC Inspection Report 50-267/89-2 No violations or deviations were identified in this area of the inspectio . Exit Interviews An informal exit interview was held with Mr. C. Fuller and other members of the plant staff on August 26, 1989, regarding the Pegion 19 CRD assembly recovery activities. An exit interview was held on August 30, 1989, with the personnel denoted in paragraph 1 of this report regarding the Region 19 CRD clevis bolt failures and the steam generator ring header cracking problems. At these meetings, the scope of the inspection and the findings were summarized. The licensee did not identify, as proprietary, any of the information provided to or reviewed by the inspector I

>

___ _ _ _

..

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

____,

?

.

...

,

ATTACHMEtlT 1

.j

-

,

a e-- ,

.;

a

- h,-c t . . . )

..t

Control Rod Drive # . e - )  :

b Orlfice Drive .

f 3 Primary St

,

Mechanism i 'y l M Lead

'

.

-

.

"

l

.

! Granulated Baron @ [ ,[ { > , > n i

,

' l

, .a ,

0.rltice Drive G%

Shaft

"

3 *

,

{ ll , ,

$ M SIDE. VIEW i

..

, '

Control Rod Orive Cable y, 4 p- ,

,

I

-; .

l ;

. .

I i l 1 -,2 I ,

pi Reserve Shi '

FRONT VIEW Hopper 26'4'

.-

.* l ,' ,

'

l, t l .- a; ,

. .

i .

.

I t Thermal l

y Insulation

-N[ M. Rupture Disc

.

"

s _N i

j 1 - j r l I & -

i ._

-

.

i q

da ?.

,

f *'. Control

'

?

. '

i i

y Ao d_, Reserve Shut j q q)

?

%, Guide M ,

Guide'\ Tube l ru f- ~

[

n T,=' j= 6

4 orifice vaive % ,

.pg-OF

.

_

' e {L:.

- - Control ~

,

e ,%

-

.M.Roda ,

,

. -

i,  !

I

.

-

\. . i

_

_ _ _ _ _ _ - _ _ _ - _ _ -

a ;. h: ATTACHiiENT'2

. -e .. l

~

. .

• ~ enhMif Z Anh H

'

r': ' !'_^.JP:. aul.9Gp L u 4 5

y -~ ~ aa

,

,

I i

b,p Wb pefG

'

, I MOVE () TO

'

I, L#N l

• i'

/

k

, i,

. i .

\ .h I l

'

l I

i ,

.

'

I

",

yes \ 7- -- /W r/M 5///M CM M

'

l; f%C$4 APhf.QX W6/W" /0 nws

'

, $

'

~

,

r

~-- Mn 2 %"tx .sim- \

. '

,

I

1, .

.

l

!

! , i

., 1 J.Od M /4/(d M '

l

~w.s.e m e _ .

/U96VC RW AMEFt .-

uJ a.uemm a m-owns dAC aFWWan m gay I /7 7DV Noc _ _ - _ _ _ _ _ - _ _ _ _ .