ML20210Q999

Safety Evaluation Concluding That Since 25th Tendon Surveillance on Few Yrs Away,Adequacy of Remaining Prestressing Force Will Be Critical to Verify

ML20210Q999
Person / Time
Site:	Three Mile Island
Issue date:	08/28/1997
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20210Q969	List: ML20210Q963 ML20210Q999
References
NUDOCS 9709020244
Download: ML20210Q999 (4)

Text

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

~ _ . _ - -

l  !

l'  !

L  ;

4 a I

SAFETY EVALUATION '

i. REVIEW OF TWENTIETH YEAR TENDON SURVEILLANCE  !

THREE MILE-ISLAND UNIT 1 l

l DOCKET NO. 50-289 >

(

1.0 INTRODUCTION

i ,

The staff has reviewed the Three Mlle Island Unit 1 (TMI-1) toth Year Tendon 4

Surveillance Report submitted on April-7,1995 and supplemented by additional information-as requested by the staff and provided by the licensee, the GPU ,

i Nuclear Corporation, on November 29, 1995. The surveillance was performed in L

accordance with TMI-l Technical Specification (TS) 4.4.2 which requires the '

licensee to perform periodic tendon surveillance on 11 randomly pre-selected i tendons. The surveillance consists of sheathing filler material inspection and testing, tendon w ire inspection and tensile testingtendon grease leakagelift-off check, force tendon measurement, detensioning and retensioning, and grease r,efilling and resealing. The group l

! of Il tendons inspected consists of 3 dome, 5 horizontal and 3 vertical. .This 1- is in com i 1.35 Rev.pliance with the tendon number requirement of Regulatory Guide (RG

3. In addition to the 11 tendons ins

areas were inspected for concrete crack growth.pected, 7 dom I. 2.0' EVALUATION i '

The staff has reviewed the results of the 20th year surveillance of the

! various elements of the post-tensioning system as contained in the report and j in tha licensee's response and clarification to staff's request for additional

," information. The staff found the licensee's tendon surveillance procedure to  ;

be generally in conformance with the criteria established in the relevant TS

- and the guidance contained in RG 1.35 Rev. 3. However, after an independent check of licensn's analysis of the tendon lift-off forces, the staff has-found-that the hoop tendon may go below the minimum required force starting in j the 25th year. This is in contrast to the licensu's conclusion of its '

i analysis that the hoop tendon will not go below the minimum required for the L rest of the 40 year plant-life. Therefore, the staff has reservation about the licensee's a discussed below.pproach and method of determining the tendon force levels as 2.1 Tendon Lift-Off Force

.The licensee has used the average values'_of the tendon lift-off forces instead- '

of individual tendon lift-off forces for each surveillance to study the trend of-the tendon forces for each group of tendons. In addition, the average *

.value for each survelliance is the average of the normalized lift-off forces. -

The use of such a procedure may result in-a trend which is not the trend of '

the actual lift-off forces. Instead of using the average of the normalized tendon forces the individual tendon forces as obtained from the lift-off of the-same group, of tendons should be used in the trend regression analysis.

The bases for such a procedurr tre: (1) the normalization factors (NF) are ENCLOSURE  :

Fa' ifE! 74%"=,

.P PDR 4

t-determined by the licensee on the basis of a method presented in a paper published in Nuclear Engineerin staff has some reservation licensee for the same thegroup, normalizedo(2)g tendon forces f tendons are not the same as the result of the andasDesign, shown by the October 1 and NFs (3) thebeing regression detemined analysis may on-the basis of subgroups be considered as a form of normalizat of a group of tendons, io process, and the regression line represents the means of the probability distributions of the tendon forces for each of the points of time. Therefore, it is inappropriate to take the avera surveillance for the trend analysis. le of the lift-off forces of eachThe licensee w analysis of the tendon force trend wi hout any averaging and by using the individual lift-off forces. The resu' ts of the re-analysis indicate that the hoop tendon force may go below the milimum required-force at about 26 years after structural integrity test SIT whichindicatesnotendongrouph(avin) its in forces contrast go to below the the original analysis minimum required force for the life of the p1 nt. Based on the data provided by the licensee the staff performed an independent linear regression analysis, by usingalogarithmicscalefortime,andfoundthatthehooptendonforcemay go below tae minimum required force at about 25 years after 51T. The result of the staff's analysis confirms that of the licensee. Since the next tendon surveillance is the 25th year one, it becomes very critical to see if this will be truly the case and to determine what corrective action is to be taken for such an eventuality.

At the staff's request, the licensee provided the force-elongation data as well as a plot of the data for the retensioned tendons. From the plots, as expected..there is a linear relationship for the dome and vertical tendons.

However, for the hoop tendon there is a kink in the slope of the line. The licensee attributed it initially to the compression of the shims with a finite thickness,-the validity of'which was questioned by the staff. It ap inaccurate measurement, most likely in elongation may be the cause, because pears that other data points can-be fitted on the same. slope,line. The force-elongation data on retensioning the detensioned tendon can be used as a check of the overall performance of tendon lift-off tests to see how carefully the surveillance was conducted. From the data provided. it appears that the licensee has conducted the tendon surveillance with reasonable care, despite

. the one inaccurate measurement.

2.2 Grease Volds The grease voids are determined from the difference between the grease removed and the grease added for tendons detensioned and retensioned. The difference should not exceed the 4-gallon limit. In cases where the 4-gallon-limit'is exceeded, an engineering evaluation of the discrepancy is requir6d.- According.

to the licensee. the 4-ga11on limit is derived from RG 1.35, Rev.- 3 and is .

based on 5 percent of net tendon sheathing volume.- It appears that the licensee has calculated the not tendon sheathing volume on the-basis of the average of the lengths.cf all tendons in the containment. The licensee used >

the average value of gallons of grease used for all the surveillances since 5 the-fifth year surveillance for.each group of tendons.as the criterion. For ,

each surveillance, the average of the additional grease used to fill the sheathings of the tendons inspected is used. The use of such averages does

'--7 . . _

,www+-+.-win,-a-m.,a--vm,-m,,n-,.,,.nw,,--,,*e,-,,-,,-vre-.-v.ww,, e-nwy,w__ _ -v .w -w w .w -w - o w .v.-r,. ,,.-w- .-www.r,mw.e.,emew., rom---.. .-e ---m s.-

l 3-

! not meet the intent of the RG 1.35 Rev. 3. The criterion of additional grease i

should be based on the 5 percent of the not volume of the sheathing of each

, individual tendon. For vertical and hoop tendons the limit for each sheathing expressed in gallons thould '>e the same,since each of the vertical and hoop tendons has respectively the same length. However, since the tendon lengths vary for the done tendons from the upper part to the lower I

done, the limits in gallons will vary with the lengths of tendons. part of the For cases when the added grease volume is less than that removed, it should not be treated ss presently by the licensee as a reduction of the additional grease for the 3endon grou). For such a condition, the licensee should provide an explanation as to-wiy less grease needs to be added. In spite of the deficie:icies in the licensee's procedure in analyzing the data. there appears to be little loss of grease discovered in the 20th year surveillance.

However, for future surveillances the licensee's method of analyzing the data should be modified to meet the intent of RG.1.35. Rev. 3 as discussed above.

The grease voids should be considered for each individual tendons inspected, if the licensee's intent is to assess the average grease voids for each group of tendons, a regression analysis of the leakage data for that group of tendons appears to be more appropriate. However, the basic concern of grease voids is that, with the presence of the voids, the tendon wires may not be protected against corrosion. Unless the wires withdrawn for inspection and testing show signs of corrosion, the need for such an analysis appears to be unnecessary. Even in such cases it will be more assuring to have additional tendons detensioned and wires withdrawn and inspected than to perform the analysis. Another concern is that missing grease may indicate grease leakage into concrete which may reduce the concrete strength. To reach such a condition, the leakage must be widespread and most likely there will be grease stains on the containment surface.

2.3 Anchorage Assembly and Tendon Wire The licensee inspected and measured the anchorhead threads to assure proper thread engagement during the lift-off, detensioning and retensioning. The anchorage components consisting of anchorhead, busing, shims, buttonheads and bearing plate were inspected for corrosion and cracks. Only minor surface corrosion and slight pitting were found on some of the components. There was no. evidence of cracking on any of the components.- With the exception of one wire in one tendon, no wires were found to be protruding or broken. The results of wire inspection and testing indicated that-the ultimate strength exceeded 240 ksi and the elongation at failure exceeded 4 percent in all sam >1es tested. Based on.the-information provided, we can conclude that the anciorage components and the tendon wires are in good condition.

In addition to the anchorage assembly and tendon wire, the concrete around the bearing plate was also inspected. No cracks greater than 0.010 inch were

.found for-the tendons inspected except around the two ends of a done tendon-wher(cracks as long as 4 inches and .020 inch wide were found. This is one

' of the- seven done teedons chosen to study the crack growth of the concrete around their ends. The licensee evaluated the crack growth and found it to be' acceptatie..

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

4

3.0 CONCLUSION

Based on the review prestress in the hoo,)as described above, the staff concludes that the loss of tendon is more than predicted and it may reduce the hoep tendon force below tie minimum required force at about 25 years after the Sil.

Otherwise the overall condition of the containment tendon system is found to be satisfactory. Since the 25th year tendon surveillance is only a few years away, it will be critical to verify the adequacy of the remaining prestressing force. If it is found to be inadequate, corrective actions should be planned and taken by the licensee. The licensee should inform the staff of such finding and the planned action. The licensee should modify its procedure in assessing the grease voids for future grease inspections as discussed in this Safety Evaluation.

Principal Contributor: Chen P. Tan '

Date: August 28, 1997 i

4 i

1

- - - - - , - -m._ < vee a _

& - * --r--- .y. .-. , c.-.--e-we, -e.- m,.+