ML20052B117

Forwards Request for Addl Info Re Deficiencies in Structural Steel Bolting Mfg by Morrison-Knudson

ML20052B117
Person / Time
Site:	Satsop
Issue date:	04/21/1982
From:	Bishop T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:	Leddick R WASHINGTON PUBLIC POWER SUPPLY SYSTEM
References
10CFR-050.55E, 10CFR-50.55E, NUDOCS 8204300048
Download: ML20052B117 (4)

Text

._.

y jg.,

4 3

4 g

• g

.q-APR ?,11982 s

xg Docket No. 50-508 V,\\p 50-509 9

APR2 91982"' T V

c a n c m R T';"

11 Washington Public Power Supply System b

A P. O. Box 1223 4

fe Elma, Washington 98541 m

Attention: Mr. R. S. Leddick WNP-3/5 Program Director

References:

1) Letter, C03-81-503, D. E. Dobson to R. H. Fauskenberry, Potential 50.55(e) Morrison-Knudsen Structural Steel Problem, dated February 19, 1981

2) Letter, G03-81-1027, R. S. Leddick to R. H. Faulkenberry, Potential 50.55(e) Morrison-Knudsen Structural Steel Problem, dated May 7,1981

3) Letter, G03-81-2370, R. S. Leddick to B. H. Faulkenberry Potential 50.55(e) Morrison-Knudsen Structural Steel Problem, dated August 14, 1981

4) letter, G03-81-2756, R. S. Leddick to B. H. Faulkenberry Potential 50.55(e) Morrison-Knudsen Structural Steel Problem, dated November 17, 1981

5) Letter G03-82-101, R. S. Leddick to B. H. Faulkenberry, Potential 50.55(e) Morrison-Knudsen Structural Steel Problem, dated January 29, 1982 Gentlemen:

Thank you for the above referenced reports submitted pursuant to 10 CFR 50.55(e) which describe deficiencies in the structural steel bolting at Unit Nos. 3 and 5.

We have reviewed your submittals, particularly the final reports listed as reference Nos. 4 and 5, and have found that the following additional information will be required to enable us to evaluate your resolution of the deficiencies.

8 2 04 3 0 00 T8 TOl S /

([,,

omu>..... RV/j k, sua m 4 $ HAIST D0DDS BISHOP om> '"'"Elj)}W Nf

, "Alh(l5El NRC FORM 318 410 80l N RC M O 2 4 0 OFFICIAL RECOFsD COPY

' " "* 32"2 '

. i' Washington -Public Power Supply System 2i N

~

~

A.

Reference No. 4 - Final Report of Potential 10 CFR 50.55(e)

Structural Steel Connections - Unit No. 5 (1) Visual inspection and torque testing prooran Since your conclusions of acceptability of the structures and the supporting statistical analysis is based upon the projected occurrence or nonoccurrence of loose bolts, it is important that we understand how the reinspections were conducted. We are aware that at the beginning of the reinspection effort,100% torque testing was performed on all accessible bolts,.but this was later reduced to-torque testing of 20% of the bolts of each connection.

In addition, page E4-14 adds the caveat that, "1 bolt out of 20 may be loose, but only 1 bolt was able to be tested. This report will only indicate 1 out of 20 bolts were loose." Please explain the effect of this method on your study of connection integrity (discussed on page E4-10 and E4-11), i.e. on what basis you assume that 992 (1020-28) connections have 100% design strength if less than 100% of the bolts in those connections were torque inspected.

We understand, also, that incorrect job inspection torque values were used during the torque reinspections.

Please discuss the effect of these values on your analysis.

You have not provided a definition of " inaccessible" bolts. Our inspections have disclosed that many of the pipe chase bolts, although having the bolt head embedded in concrete, are accessible for partial visual inspection and torque inspection.

Please indicate why these bolts were classified as " inaccessible" and provide your definition of this term.

(2) Statistical Method Synopsis Please explain the basis for your assumption that the percentage of design strength can be estimated based on a simple ratio of bad versus good bolts in any one connection. This assumption is valid for simple shear connections but may not be valid for moment and other types of connections.

Probabilities and confidence levels have been determined using a binomial distribution. curve which is valid only if the items checked are selected at random.

Please indicate the randonness of your sampling techniques.

On Page E4-6 you have indicated that 9940 bolts were tested with 4

61 bolts failing to pass the test.

Please indicate if all 9940 bolts were torque tested.

In addition, indicate if the 61 bolts failing to pass the test include the deficient bolts documented on NRC 5504 and, if not, the basis for their omission.

.-r--r

Washington Public power Supply System APE 21 7582 (3) Non-Randon Distribution of Loose Bolts The raw data listing on page E4-10 indicates that loose bolts tend to be distributed among relatively few connections (28 connections out of 1023 connections).

For this reason we have calculated the binomial distributions for connections with zero design strength and for connections with less than, or equal to, 50 percent design strength using the raw data of page E4-10. These calculations indicate a probability of 16 percent that 5 connections will have zero design strength and a probability of 9 percent that 11 connections will have less than or equal to 50 percent design strength.

Please address these results or justify your design strength probabilities on page E4-ll wherein you have combined 992 connections with 100% assumed design strength with the 28 connections exhibiting one or more loose bolts thereby arriving at a random loose bolt distribution and correspondingly low probabilities of low design strength. Also, justify your conclusions on page E4-3 in light of the above questions.

B.

Reference No. 5 - Final Report of potential 10 CFR 50.55(e)

Structural Steel Connections - Unit No. 3 (1) General Backup data showing the distributions of deficient bolts within connections and methods of statistical analysis were not provided with this report as had been provided with the Unit 5 report. Accordingly, if you intend to utilize this data, please provide sufficient infomation to pemit us to evaluate the deficiency and your corrective actions.

(2)

Description of the Deficiency and Items of Concern Results of reinspections are provided but are not definitive, e.g. " fifteen (15) connections with loose bolts, two (2) connections with missing bolts". These results do not pemit us to perfom an evaluation of the deficiency and evaluate the adequacy of your statistical analysis.

In this regard, please provide the same type of infomation as was supplied for the Unit 5 report. Also, please indicate if the unsatisfactory bolt projections were cases of excessive projection and, if so, whether or not the nut was run up on the bolt shoulder giving an erroneous torque indication.

In addition, indicate the reinspection procedure used (visual and or torque); the percent which were torque inspected (twenty percent or 100 percent); and the actual job inspection torque applied.

..s

^

I 002

_. Wash (4gton Public Power Supply System We understand that torque inspections were discontinued shortly after being; initiated on Unit 3.

Please discuss the rationale for discontinuance of these torque inspections in light of the deficiencies reported on page 2 of your report.

(3) Corrective Actions Taken Please review your data base, statistical analysis techniques, and conclusions in view of our comments on the Unit 5 report and '

provide the same supplemental information for Unit 3, as applicabl e.

Please provide the above requested information within thirty (30) days of the date of this letter.

Sincerely, T. W. Bishop, Acting Chief Reactor Construction Projects Branch 1

bcc:

DitB/ Document Control Desk (RIDS)

Distributed by RV:

State of WA (Lewis)

Resident Inspector (2)

Engelken (ltr) l 1

[

~.. -. _...,.

r

.