IR 05000267/1985013

Insp Rept 50-267/85-13 on 850415-19,0717-19 & 1022-25.No Violations or Deviations Noted.Major Areas Inspected: Corrosion Phenomenon in Pcrv post-tensioning Tendon Wires & Extent of Pcrv Tendon Degradation

ML20151T532
Person / Time
Site:	Fort Saint Vrain
Issue date:	01/28/1986
From:	Ireland R, Stewart R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20151T517	List: IR 05000267/1985013 ML20151T515
References
50-267-85-13, NUDOCS 8602100296
Download: ML20151T532 (9)

Text

p

" '

.

'

L, APPENDIX U.S. NUCLEAR REGULATORY COMMISSION-REGION IV NRC' Inspection' Report: 50-267/85-13 License: DPR-34.

i Docket: 50-267

~ Licensee: ,

Public Service Company of Colorado (PSC)

P. O. Box 840 Denver, Colorado 80201

' Facility Name: Fort St..Vrain Nuclear station (FSV)

Inspection At: Platteville, Colorado Inspection Conducted: April 15-19; July 17-1 and October 22-25, 1985

-

/

/ /2

'

Inspector: - _ w /ma / .

R.' C. Stetia'rt Reactor Inspector, Engineering Date /

Section, Reactor Safety Branch l

A M & n /s R. E. Ireland, Chief, Engineering Section,

//2.F/'Y4 Ditte Reactor Safety Branch Inspection Summary Inspection Conducted April 15-19;' July 17-19 and October 22-25, 1985 (Report 50-267/85-13)

Areas Inspected: Special, unannounced. inspections involving the review of the corrosion phenomenon occurring in the prestressed concrete reactor vessel (PCRV), post-tensioning tendon wires; extent of PCRV Tendon Degradation; Proposed Monitoring and Surveillance Program; and Corrective Measures being proposed and/or initiated. The inspection' involved 68 inspector-hours by one

' '

NRC inspector.

Results: Within the four areas inspected, no violations or deviations were identified.

0602100296 060203 gDR ADOCK 05000267 PDR

p

..

DETAILS 1. Persons Contacted

• J. W. Gahm, Manager, Nuclear Power Operations

'

• C, H. Fuiler,' Station Manager

,

• L. W.-Singleton, Manager, Quality Assurance

• R. L. Craun, Site Engineering Manager

• T. Prenger, QA Services Manager

• F. J. Borst, Support Services Manager

• M. J. Ferris, QA Operations Manager

• G. L. Redmond, MQC Supervisor

• F. J. Novachek, Technical / Administrative Services . Manager

• D. L. Frye, Sr. Licensing Specialist

• W. A. Craine, Maintenance-

.D. W. Warembourg, Manager, Nuclear Engineering T. M. McIntire, Supervisor, Nuclear Site Engineering M. Fisher, QA Engineer.

L. Parmley, QC Inspector B. McCloskey, Stone & Webster, Mechanical Engineering J. Jackson, QA Services A. Reed, Technical Services T. V. Erlewine, Mechanical Engineer The NRC inspector also interviewed other licensee staff personnel during the inspection periods.

• Denotes those present at the final exit interview on October 25, 1985.

2. Background a. PCRV Function The PCRV encloses the reactor core, primary coolant system and part of the secondary coolant system. The structure is a polygonal thick walled, hollow prism with thick flat heads. The vessel structure is designed so that for all load combinations the overall structural response of the PCRV will be essentially elastic and that the vessel is in compressive condition for all internal pressures from zero up to reference pressure 845 pounds per square inch gauge (psig).

(Reference pressure is equal to 1.2 x Peak working pressure).

b. Tendons The FSV PCRV uses the Brikenmaier Brandestini Ross Vogt (BBRV) post-tensioning system of high-strength steel tendons housed in carbon

.

...

steel tubes. The unbonded prestressing system for the PCRV consists of 448 multiwire tendons that provide the forces to resist the inter-nal pressure of primary coolant. The tendon system consists of 90 longitudinal, 310 circumferential, 24 top crosshead, and 24 bottom crosshead tendons.

A single tendon consists of 169 (152 for circumferential tendons around the barrel section) 0.?50 inch diameter Type BA low-relaxation carbon steel wires meeting requirements of ASTM-421, each of which is anchored-by a cold-detormed end called a buttonhead. The buttonheads anchor each wire against a washer fabricated from AISI 4142H steel.

The washer and remaining components of the anchor assembly transmit-the tendon force to the PCRY. The tendons are ungrouted and are not in contact with the PCRV since they are inserted into previously placed~ carbon steel tubes. The mechanical properties of the prestressing tendons used in the design are as follows:

Minimum Guaranteed Ultimate Tensile Strength 240,000 psi

Tendon load immediately after ,

anchoring 168,000 psi

Minimum yield strength at one percent elongation 180,000 psi

Total relaxation losses at 30-year design life at 120*F 8 percent c. Corrosion Protection The corrosion protection of-the tendons consisted of coating each wire with Meta-Bond 39 (a calcium zinc phosphate coating) plus Rustarest 452 which seals the phosphate coating. The tendon wires were also coated during tendon fabrication and installation, with No-Ox-ID CM casing filler, a product of the Dearborn Chemical Division, W. R. Grace and Company, Chicago, Illinois. No-Ox-Id CM is a chemically neutral grease-like thixotropic compound of paraffin-based refined mineral oil and microcrystalline petroleum-derived wax (petroleum) containing additives of lanolin and sodium petroleum sulphonate.

d. Margin of Safety The " Fort St. Vrain Generating Station, Updated Final Safety Analysis Report," Section 5 contains a summary description of the Design and Construction of the FSV PCRV. Appendix E provides the principal backup information on which the design of the PCRV was based.

i

.

..

Considerable redundancy in the prestressing system has been provided in the design of the PCRV. Section E.14.2, Appendix E delineates the quantitative analyses performed to establish the approximate number of tendons which could fail during operation of the reactor before a compromising condition would exist. The analyses reflect that the-1oss of a certain number of adjacent tendons for each type of tendon and the loss of net compression across a section under operating conditions (i.e., 9 of 12, circumferential tendons, in any 5' cross section could fail at 845 psig without' reducing the safety factor below 2.0) would not cause a hazard with respect to the reduction in the vessel's safety factor. In addition, the analyses showed that it is possible to operate with any one tendon detensioned and still comply fully with the design criteria for the PCRV. Twenty-seven load cells on selected tendons monitor tendon loads as a means to assure uniform loading throughout the structural elements and are continuously monitored with low point alarms in the control room, e. Initial Proof Test The structural design verification, wherein the PCRV was subjected to the maximum initial proof te demonstrated in August 1971.{t Thepressure IPTP level(IPTP)

is 1.15ofx 970 psig, was the reference pressure of 845 psig or 1.41 x the normal working pressure of 688 psig, of the PCRY. The IPTP established that the structural response of the PCRV to pressure changes up to 970 psig is essentially linear and a condition of net compression was maintained at'all cross sections in the vessel at 970 psig. In addition, comparison of measured and calculated load and displacement data shows acceptable agreement indicating predictability of vessel structural behavior up to a pressure of 970 psig.

f. PCRV Tendon Degradation During March and April 1984, several tendon' anchor assemblies were inspected as required by Technical Specifications SR 5.2.2.c., and the following conditions were observed:

evidence of broken tendon wires indicated by wire buttonheads raised off their washers

evidence of corrosion at the ends of failed tendon wires

evidence of corrosion within anchor assemblies 1 Gulf General Atomic Report, Gulf-GA-A10839, December 1971

-

.

evidence of moisture within anchor assembly covers

evidence of incomplete coverage of corrosion inhibiting compound As a result of the above findings, the 1984 surveillance program was substantially expanded.to involve visual inspection, measurement of liftoff force and detensioning of tendons (March 1984 through February 1985). In addition, an extensive investigation wa!.

initiated in order to determine the corrosion mechanism.

Based on the results of the investigation the licensee has concluded that the corrosion of tendon wires was caused by microbiological attack on the tendon No-0x-Id CM organic grease.2 The microbial

.

l activity caused the formation of formic and acetic acids. The acids, in conjunction with moisture in the tendon tube, vaporized and recondensed on the cooler portions of the tendon ends (predominately ,

8"-10" from the buttonhead end and in areas relatively absent of ,

grease). Based upon past surveillance information, the corrosion !

process has been taking place since 1971 in a gradual and progressive manner.

The number and location of tendons with broken wires gre identified in the licensee's 1984 and 1985 surveillance reports. The number of tendons with known broken wires, included 12 longitudinal tendons with 1 to 22 broken wires, 6 circumferential tendons with 1 to 15 broken wires, 9 bottom crosshead tendons with 1 to 28 broken wires, and 2 top crosshead tendons with I broken wire in each tendon. In some cases, the total number of broken wires included wires broken during lift-off tests, or during retensioning.

The results of 74 longitudinal lift-off tests indicated that tendons with identified broken wires generally had a slightly smaller lift-off value than intact tendons. Lift-off values on 30 circumferential tendons showed little change in values. Some of the 15 bottom crosshead tendon lift-off tests showed definite reduction in lift-off values for tendons with multiple wire breaks, however, all lift-off test values exceeded the minimum design lower limits.

2 Thurgood, Roberts and Epstein, " Evaluation of the Causes of Corrosion in the Fort St. Vrain Post-Tensioning Tendon Wires " GA Technologists, PSC submittal 1/24/85.

3 Lab Report No. 52, Examination of Failed Tendon Wires from Fort St. Vrain Unit 1. PSC Submittal P-04543-4, 1/24/85 and tendon surveillance,"

March 1985.

4 " Liftoff Tests," Attachment 1 to " Examination of Failed Wires From Fort St. Vrain Unit 1, dated 12/31/84.

'

.

l i

3. Assessment Reports / Restart Irsues As a result of the corrosion phenomenon occurring in the tendon assemblies I and the unknown exgent of tendon degradation, the subject was added to the NRC restart issues as a matter requiring an evaluation of the licensee's assessment prior to plant restart.

The licensee was required to determine the mechanism causing the corrosion (see f above) and to initiate corrective measures to eliminate further tendon degradation and to coatinue a tendon surveillance program that will determine the extent and rate of current degradation in the PCRV tendon system.

DuringameetingheldFebruary 20-22, 1985,6 and through PSC letter, ,

P-85071, dated March 5, 1985, the licensee proposed a tendon surveillance program establishing a six-month inspection / testing frequency for a period of three years, or until such time the effective corrosion control is established.

In response to the information provided by the licensee (see f. above) the NRC staff evaluated the integrity of the PCRV with de,raded tendons in a safety evaluation report (SER) dated May 16, 1984. The staff findings were that the reactor vessel was capable of withstandins the operating pressures with the degraded tendons as determined at that time. In addition,aTechnicalEvaluationReport(TER)condugtedbyLosAlamos National Laboratory and incorporated in an NRC SER, concluded that the license's commitments concerning tendon surveillance were acceptable with the proviso that:

1. The licensee incorporate the modified tendon surveillance program !

into the Technical Specifications.

2. The licensee provide the NRC with the results of the tendon .

surveillance program every 6 months after plant restart. l l

5 " Preliminary Report Related to the Restart and Continued Operation of Fort St. Vrain Nuclear Generating Station," Docket No. 50-267, PSCO, October 1984.

6 Letter, R. P Denise, NRC, to 0. R. Lee, PSCO, dated April 8, 1985. l

'

7 PSCO Letter, P-85071, " Revised Tendon Surveillance Program," dated March 5, 1985.

8 Safety Evaluation for Fort St. Vrain PCRV Tendon Corrosion Investigation and Proposed Remedy - May 21, 1985.

l

~

.

7 l l

3. Thelicenseeadgressthematterofpotentialforanchoragestress washer failure WithreferencT0to Item 3 above, the licensee responded to this matter on June 7, 1985. The licensee's inspections / investigation concluded that i the conditions / failures identified in Information Notice 85-10, Supple-ment 1, were not applicable to the PSC tendon system.

4. Inspection Findings During the period April 15-19, 1985, the NRC inspector reviewed the prior history of tenden corrosion findings 2, 3,/; lift-off tests 4/; QA/QC Procedures and Records; and Review of the Status of the licensee's attempt to establish a nitrogen purge on tendon tubes 8] as a means to halt any further tendon wire corrosion.

The NRC inspector verified that the corrosion of tendon wires is random in nature and that there is no immediate concern of several tendon failures .

that would result in localized PCRV concrete tension. In addition, the NRC inspector conducted a c designated by the licensee,grelation December review of inaccessible 14, 1985, tendons with as-built conditions.

There were no descrepancies noted.

The inspector examined tendon QC records that documented lift-offs /inspec-tions conducted on 37 tendon assemblies (Circumferential tendons) from the period tiarch 5,1985 through April 15, 1985. No additional failed wires were found during that period, and only isolated rust was observed by the licensee QC inspector. 'a NRC inspector also witnessed ongoing lif t-off/

inspection activities in order to verify QA/QC procedure compliance as prescribed by PSCO, "PCRV Prestressing Tendon Liftoff Procedure," MP 11-6, Issue 4, dated January 26, 1985. During the lift-off tests (circumferen-tial), the inspector also accompanied the licensee's QC inspector in visually exan.ining tendon wires. Tendon wires were found to be moderately greased and absent of any evidence of rust. The licensee plans to continue lift-off/ inspections of all remaining accessible tendon assemblies until an overall base line can be established prior to the start of the 3-year surveillance cycle (paragraph 3. above).

In reviewing the status of the licensee's attempt to established a nitrogen purge on a tendon tube, the inspector was informed that nitrogen 9 NRC Information Notice, 85-10, Supplement 1, " Post Tentioned Containment Tendon Anchor Head Failure," date March 8, 1985.

10 PSCO Lettwr, P-85193, " Fort St. Vrain Additional Tendon Inspection," dated June 7, 1985.

11 PSCO Letter, P-84523, dated December 14, 1984.

- __

.

-

purge losses around the tendon anchor assemblies were too great to be a viable means of purging tendon tubes. The licensee has contacted a local sealant supplier in order to attempt a means of sealing around anchor assemblies.

During the period July 17-19, 1985, the NRC inspector observed additional lift-off/ inspections. It was noted that approximately 41 tendon, circum- ,

ferential tendons (both ends) were examined since February, with isolated

'

rust found and all lift-offs exceeding minimum design limits. The licensee estimated that lift-offs / inspections were complete on vertical, top and bottom crosshead tendon assemblies. Lower PCRV levels of circumferential tendons were yet to be completed.

During this inspection, the NRC inspector reviewed the proposed modified tendon surveillance program (see footnote 7) which is to be incorporated

, in the UFSAR. Specifically, lift-off/ visual inspections would involve two populations of tendon groups: a population of tendons that have not been previously identified as being corroded, and a control population with know corrosion. Acceptance criteria would include a mandatory engineering evaluation on any tendons that have 20% of total wires broken on 169 wire assemblies, and/or 15% of total wires broken on 152 wire assemblies (circumferential barrel tendons).

Although 12 control tendons have tentatively been selected, the program has not been fully implemented pending further baseline tendon lift-off/

inspection data.

Sealing of an anchor assembly was completed on one tendon in an effort to reduce nitrogen purge losses. Although the effort reduced the losses by 60% at 10 psi, this method of preventing further corrosion does not appear to be an efficient method. In the meantime, the licensee has continued developmental work at GA Technologies, San Diego, CA., on other methods of corrosion arrest, including removing existing grease, oil flushing, new grease applications, etc.

During the period October 22-25. 1985, the NRC inspector observed the on going lift-off/ inspection of the remaining circumferential tendons and L

that the results were well within acceptable limits (paragraph 3 above).

With the majority of li't-offs. complete, all appear to be above the lower design limits. During the period February 26 through October 25, 1985, 21 additional tendon wires were found broken in 16 tendon assemblies. The

~ worse cases being; 3 broken wires in a circumferential tendon (152 wires assembly); and 3 broken wires in a bottom crosshead tendon (169 wire

'

assembly). The licensee continues to make preparations for the

,

!

-

.- - -- - _ _ -

, __

. S .-

t

.

replacement of two tendon assemblies for further examination. This effort is principally intended to determine the extent of corrosion if any, in the middle portion of the wire assemblies. The two assemblies to be replaced have not been selected at this time.

.As stipulated by the licensee's commitment to-the restart issues (paragraph 3 above), the licensee plans to submit the results of the tendon surveillance activities since plant restart July 21, 1985, in mid-January 1986.

No violations or deviations were identified during the above inspection periods.

5. Exit Interview An exit interview was held October 25, 1985, with Mr. J. W. Gahm, Manager, Nuclear Production, and other members of the PSC staff as denoted in

. paragraph 1 of this report. The NRC senior resident. inspector also attended this meeting. At the meeting, the scope of the inspection and findings were summarized.