TXX-3442, Final Deficiency Rept Re Unauthorized Field Mods to Hilti- Kwik Concrete Anchor Bolts During Installation,Initially Reported on 801001.Bolts Removed or Cut Off & Driven Into Concrete Structures
| ML20039A158 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 12/04/1981 |
| From: | Gary R TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC) |
| To: | Madsen G NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| References | |
| 10CFR-050.55E, 10CFR-50.55E, TXX-3442, NUDOCS 8112160285 | |
| Download: ML20039A158 (6) | |
Text
,s TEXAS UTILITIES GENERATING COMinNY 2001 HRYAN TOWER = DALLAM,TF:X AH 75201
- h. C December 4, 1981
'/
n.a.oxnv TXX-3442
, y -r lr R-c'
- q 4
N.I bh Mr. G. L. Madsen, Chief 7
Reactor Projects Branch U. S. Nuclear Regulatory Commission 4
Office of Inspection a Enforcement
'w / i D i,j, 611 Ryan Plaza Dr., Suite 1000 Docket Nos. 50-445 Arlington, Texas 76012 50-446 COMANCHE PEAK STEAM ELECTRIC STATION HILTI-KWIK CONCRETE ANCHOR BOLTS FILE NO:
10110
Dear Mr. Madsen:
In accordance with 10 CFR 50.55(e), we are submitting this final report concerning identified deficiencies as a result of unauthorized field modifications of Hilti-Kwik concrete anchor bolts during the installation process. This matter was verbally reported to your Mr. R. G. Taylor on October 1, 1980.
Interim reports have been forwarded by our letters TXX-3223, TXX-3266 and TXX-3303 dated October 21, 1980; January 12, 1981; and April 15, 1981, respectively.
The 6ttached report includes information previously submitted and defines the corrective actions that were developed as part of our overall evaluation.
Records supporting our study and the conclusions reached are available at the CPSES site for your inspector's review.
If we can provide any additional information, please advise.
Very truly yours, R.
. Gary RJG:kh Attachments cc: NRC Region IV - (0 + 1 copy)
Director, Inspection & Enforcement - (15 copies) c/o Distribution Services Branch, DDC, ADM.
U. S. Nuclear Regulatory Commission Washington, D. C.
20555
/kcr
$hkk o 00
/ /
S
F December 4,1981 TXX-3442 ATTACHMENT A HILTI-KWIK CONCRETE ANCHOR BOLTS
SUMMARY
Description of Deficiency As a result of an allegation to Construction management, Ultrasonic Testing (UT) of a number of Hilti-Kwik and Super-Kwik concrete expansion anchor bolts was accomplished to verify that the installed lengths were consistent with a code letter designation on the end of the bolt and in compliance with design requirements. Two of approximately 50 expansion anchor bolts examined were indicated by UT as being shorter than required.
Both anchor bolts were associated with a single conduit support in the Unit 1 Cable Spread Room.
One expansion anchor bolt, which was removed by jacking, was found to have been field modified in an unauthorized manner to shorten the bolt by cutting off the mandrel and grinding a new mandrel into the shorter shaft. The second expansion anchor bolt was checked for torque prior to removal by jacking, and the required torque value could not be obtained. This second expansion anchor bolt had been nodified in a manner similar to the first.
More extensive UT examinations, as described subsequently, also indicated a condition in HVAC support installation wherein the second or lower mandrel of three Hilti Super-Kwik expansion anchor bolts (out of a sample of 818 bolts UT examined) had been removed. Torquing or load data was not obtained for these three bolts.
Analysis of Safety Implications After evaluating the previously described Hilti-Kwik expansion anchor bolts, it has been determined that due to the unauthorized and uncontrolled modification of these items, sufficient data is not available to determine their structural capabilities. Therefore, it has been concluded that this deficiency could have adversely affected safe plant operation due to the inability of the supports to sustain the design loading conditions.
Corrective Actions The following actions have been taken to correct this deficiency and to evaluate the overall integrity of installed Hilti-Kwik concrete expansion anchors.
December 4,1981 TXX-3442 1.
It has been emphasized to craft personnel by Construction management that any unauthorized modification or alteration to permanent plant components will result in disciplinary action.
2.
Field modified Hilti-Kwik bolts have either been removed from or cut off and driven into the concrete structures (and patched per existing CPSES procedures).
Supports involved have been reworked as required to comply with design requirements..
3.
To evaluate the exter,t of the identified deficiencies and to correct problems, an extensive program of ultrasonic testing was planned and implemented by the site Ouality Engineering Group.
The initial phase of the program consisted of a statistically sound sampling program of Hilti-Kwik expansion anchor bolts installed by functional craft disciplines and subcontractors (mechanical, electrical, instrumentation, fire protection and HVAC). This initial phase indicated adequate confidence that the mechanical, instrumentation and fire protection Hilti-Kwik bolt installations had not been modified in an unauthorized manner. The second phase of the program included further examination of HVAC and electrical installations.
Details and results of the UT examinations are provided in Attachment B.
Based on all of the examinations performed, we are confident that the support installations involved are now in full compliance with design and installation requirements and will perform their intended function.
r
3 December 4,1981 TXX-3442 ATTACHMENT B HILTI-KWIK CONCRETE ANCHOR BOLTS DETAILS AND RESULTS OF UT EXAMINATIONS Procedure Development and Implementation Ultrasonic Examination of Hilti bolts was developed to determine the length and type of Hilti bolt after they were installed.
Actual Hilti bolts were used as reference blocks to produce an acoustical equivalent to the installed bolts.
One bolt of each size and type (Kwik or Super-Kwik) was drawn frum existing stock and collectively these bolts were used to develop the testing procedure used in the field.
The UT instrument used is capable of generating frequencies of at least 1 Mz through 5 mz and provided linear presentation (within + 5 percent) for at least 75 percent of the screen height (sweep line to tiip of screen).
Glycerin was used as a couplant and the transducer was in the frequency range of 1 MHz through 10 MHz.
To ensure that the ultrasonic presentation seen by the inspector on the CRT screen during inspections would be interpreted correctly,.a predetemined screen pattern for each bolt type und length (as developed from the reference block graph) was produced and provided to the inspector in figure form.
All testing was performed under close supervision of the Quality Engineering Group.
All inspectors perfoming UT of Hilti bolts are senior in status, having obtained at least their Level II certification in accordance with the criteria established in SNT-TC-1A. - All inspectors have a minimum of 7 years proficiency as a Level II Technician in UT NDE.
Additionally, all inspectors were given additional instniction in testing of Hilti bolts by ultrasonic means by the CPSES Site Level III NDE Supervisor.
Initial Sampling Program Using the Estimation Theory *III, a number was derived, such that, if the observed data corresponded to the predicted results we could be 99.5 percent confident that 99 percent of the bolts have the correct embedment and configuration. The required number of bolts to be inspected was 782 with 1 or fewer bolts failing inspection.
The equation is shown as follows:
- (1) Prcbability and Statistics for Engineers and Scientists, R. E. Walpole and R. H. Myers, Macmillan Company, N.Y., 01972, pages 201-205.
December 4, 1981 TXX-3442 29% = 2.810 (normal random variable) or =
.005 (1-u = confidence coefficient) p =.01 (estimate of proportion unacceptable)
G =
.99 (estimate of proportion acceptable)
.01 (maximum error)
=
e n = (Z-v2Y a 2 (2.810)2 (.99) (.01) 782
~
=
2
~
(.01)2 c
Test results were analyzed using the standard test of hypothesis where the null hypothesis (H ) is that the proportion defective (p) is equal to.01.
o The alternative hypothesis is that p is greater than.01.
The tests were conducted at the 99.5% confidence level.
Since the sample size (n) is large and p is very close to zero, the Poisson approximation of the binomial was used, as follows:
(x;.01n)
, where the critical region is t.005.
x=0 For four of the five groups (Mechanical, Electrical, Instrumentation and HVAC) a minimum of 782 Hilti Bolts were tested for correct length. The Fire Protection group had fewer than 782 bolts installed at the time of the sampling program; therefore all bolts in that group were tested.
The results of the examinations are as follows:
Fire Protection - All bolts UT'd (161).
All bolts acceptable.
Instrumentation - 783 bolts UT'd.
All bolts acceptable.
Mechanical - 816 bolts UT'd.
All bolts acceptable.
Three bolts found to be unacceptable.
Electrical - 796 bolts UT'd.
All bolts acceptable.
The results of the sampling nrogram indicated additional testing was required in the HVAC.
Additional testing was also undertaken in the areas of the plant worked by the crew involved in the original allegation.
Note that the sampling program results indicate that unauthorized modification in electrical installations was isolated.
I
December 4, 1981 TXX-3442 Expanded Examination in HVAC and Electrical Installations Based on the UT examinations discussed previously, expanded examinations were initiated in the HVAC and Electrical Hilti bolt groups.
Except as noted herein, all installed HVAC and electrical conduit support Hilti expansion anchor bolts installed in areas worked by the involved crew were examined.
In isolated instances, where personnel safety would have been compromised or where subsequent installation of material rendered installed Hilti expansion anchor bolts inaccessible (as determined by the Quality Engineer), ultrasonic examinations were not performed. The majority of the inaccessible installations involved ceiling mounted supports with the predominant mounting device being the Richmond insert.
One-quarter inch (1/4") Hilti bolts were also excluded from the UT examination for the following reasons:
1.
Hilti Super-Kwik bolts are not available in 1/4" diameter.
2.
All Hilti-Kwik bolts, by design, only have one mandrel.
The findings in the electrical arena show that a Hilti-Kwik bolt with a modified mandrel cannot achieve the required torque value.
3.
1/4" Hilti-Kwik bolts have not been utilized in HVAC installations.
4.
Comparisons of embedment requirements for the 1/4" Hilti-Kwik bolt versus the amount of concrete coverage over rebar indicate that the probability of rebar interference is minimal, and thus there should have been no reason to field modify the Hilti-Kwik-bolts.
A total of 4870 additional Hilti bolts were examined in the HVAC group; 1 bolt was found to be nonconforming due to mandrel modifications.
In the Electrical group 3641 additional Hilti bolts were examined; 7 bolts were identified as nonconforming due to mandrel modifications.
In summary, a total of 8511 additional Hilti bolts were examined, representing all bolts installed with the exception of the inaccessible and 1/4-inch bolts described above.
m.
- m. m 2