ML20155E248
| ML20155E248 | |
| Person / Time | |
|---|---|
| Issue date: | 10/22/1998 |
| From: | Spessard R NRC (Affiliation Not Assigned) |
| To: | Amer A AFFILIATION NOT ASSIGNED |
| References | |
| REF-QA-99901292 NUDOCS 9811040253 | |
| Download: ML20155E248 (8) | |
Text
_
q CQq%
3 h
b UNITED STATES f
j NUCLEAR REGULATORY COMMISSION
~
R WASHINGTON, 6Ic355SidOI I ~ "
- '+,
o
- g ocG6eEif,5168 2: l!O PUBLIC CO:U. Eni a:i Mr. Ahmad E. Amer, Chairman Amor Industrial Technologies, Inc.
Wilmington, DE 19801
Dear Mr. Amer:
This letter confirms the upcoming public meeting between the Nuclear Regulatory Commission (NRC) and Amer Industrial Technologies, Inc. (AIT) to be held at the NRC headquarters office in Rockville, MD at 10:30 a.m. on November 9,1998, as discussed during an October 22,1998, telephone conversation between you and Ms. Suzanne C. Black of this office.
The purpose of the meeting is to discuss the remaining issues that the NRC has concerning your responses to the Notice Of Nonconformance contained in the NRC's letter to you dated, March 21,1996, and your subsequent correspondence relating to the mechanical testing and chemical analyses of the tubes used for the manufacture of the heat exchangers on AIT Job No. 442. The attachment to this letter provides a detailed description of a recent staff review of the testing and analyses documentation you provided for these tubes. Please be prepared to discuss the results of the staff's review dunng the meeting.
Your cooperation in providing the support needed to complete our review of these issues is appreciated. Should you have any questions concerning this letter, please contact Ms.
Suzanne Black at (301)415-1017.
Sincerely, l hJY W R. Lee Spessard, Director Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation Docket No. 99901292
Enclosure:
As stated j
/
9811040253 981022 PDR OA999 E*******
o
>J 99901292 PDR
~J
/ f 3 l/f AU$
t
' -' ? r1 7n' qui 722)
REVIEW OF AMER INDUSTRIAL TECHNOLOGY INC.
The following are the results of the staff's review of correspondence to and from Amer Industrial Technologies, Inc. (AIT) and the NRC regarding the stainless steel tubes used to construct l
three replacement charging pump gear oil heat exchangers for the South Carolina Electric &
Gas Company's (SCE&G's) V. C. Summer Nuclear Plant. The heat exchangers were constructed in accordance with SCE&G PO No. Q650395, dated May 19,1994, and were l
processed as AIT Job No. 442. The tubes were required to meet the following requirements:
Material: SA-249, " Specification for Welded Austenitic Steel Boiler, Superheater, Heat i
Exchanger, and Condenser Tubes," and Code: ASME Section ill-1971 Edition, Class 3.
l.
Mechanical and Chemical Test Data VS Allowable Tube Length Testing Performed by Testing Laboratorv Inc.
j AIT's February 5,1996, letter to the NRC contains documentation from Testing Laboratory Inc.
(TLI) dated March 23,1995, that indicates that a minimum of 24 inches of tubing (from each tube) would be required to perform the required mechanical testing on the 35 tubes to be subjected to such testing. AIT issued PO 20824 to TLI for the mechanical testing of the tubes.
AIT assigned a unique heat number for each of the 35 tubes to be tested for traceability purposes. The traceability numbers assigned for the tubes were M-2627-1 through M-2627-35.
Further, the March 23,1995, TLI letter indicates that TLI had previously received 35 samples of 5-1/2 inches long, but it was not discussed why TLI had these 35 test pieces.
Review of TLI Test Report dated 4/17/95 and supplemented on 4/18/95 indicates that 35 samples were subjected to tensile testing and that a flange and flatting tests were performed on 70 pieces (2 from each heat trace), and that 35 pieces from each tube were subjected to a reverse-bend and hardness tests. The TLI test report refers to AIT PO 20824.
According to the TLI letter dated March 23,1995, and as a result of reviewing the TLI test report dated 4/17&18/95, the following lengths of tubing would have been required to perform the specified mechanical tests on each tube (note TLl's 4/17 & 18/95 test report indicates that these tests were performed for each of the 35 tubes):
Tensile test:
10 inches long,1 piece Flattening test:
5 inches, 2 pieces, each 2-1/2 inches long, one piece from each end of tube Flange test 5 inches,2 pieces, each 2-1/2 inches long, one piece from each end of tube Reverse Bend test:
4 inches,1 piece
.].
1 1 -
l The totallength of tube sample / coupon from each of the 20 ft long tubes required for l
performing the mechanical tests (performed as documented by TLI) was: 10+5+5+4=24 inches l
(2 feet).
On September 23,1998, the staff called TLI located in Dublin, PA at (800) 627-3966 and discussed the above with Sherri L. Lengyel (QA Coordinator), the individual that approved / authorized the test reports for the mechanical tests performed for AIT. Ms. Lengyel i
informed the staff that the records had been sent to archives and it would take her some time to review Lem. On September 23,1998, the TLI QA Manager, Arnold Horoff, called and provided the staff the following information:
I For each of the 35 heat traces (M-2627-1 through M-2627-35) AIT provided the following lengths of tube samples for testing. Each test sample was identified as coming l
from one of the tubes with this heat number:
35 pieces, each 5-1/2 inches long (one flattening test, one flange test, and one hardness was performed on each piece) Note: each 5-1/2 inch long piece provided was cut into pieces 2-1/2 inches,2-1/2 inches and 1/2 inch long with each the 2-1/2 inches pieces being subjected to a flange and flattening test) 35 pieces, each 10 inches long (one tensile test was performed on each piece) 35 pieces, each 2-1/2 inches long (one flattening test was performed on each piece) 35 pieces, each 2-1/2 inches long (one flange test was performed on each piece) 35 pieces, each 4 inches long (one reverse bend test was performed on each piece) l Testina Performed by Ramball Test Lab. Inc.
When AIT received the 36 tubes from Marmon/ Keystone, it initially sent samples from one tube to Ramball Test Lab, Inc. (Ramball). Ramball performed mechanical testing and chemical analyses on the samples received. Subsequent to Ramball performing these mechanical tests and chemical analyses, an ASME survey team identified that all of the tubes received from Marmon/ Keystone should have been subjected to mechanical tests and chemical analysis. The i
following are the results of the staff's review of the activities that AIT requested Ramball to i
j perform on the samples taken from the tubes received from Marmon/ Keystone.
AIT's February 5,1996, letter contained test report data from Ramball which indicates that a full section pipe ter.sile test was performed with a gauge length of 2.0 inches on the one tube initially received from AIT. AIT incorrectly stated in its February 5,1996, letter that the sample length was 2 inches. The gauge length was 2 inches with several additional inches required to l
test the sample. The gauge length is the minimum length of the sample that is subjected to l
testing and the test specification requires certain diameters of !cngth beyond the gauge length plus the amount of tube to be placed in the test fixture. Ramball also performed chemical analyses and flattening, bend, and flange tests on the samples cut from one of the 36 tubes that AIT received from Marmon/ Keystone.,,
On September 23,1998, the staff called Ramball and talked to Fred Fetterolf, General Manager, concerning the chemical analyses and mechanical tests that Ramball performed fo-AIT on July 22,1994. Ramballinformed the staff that a minimum length of 1 inch of tubing would have been required to perform the chemical analyses on the sample tested on July 22, 1994. Ramball subsequently performed chemical analyses on samples from the remaining 35 tubes having assigned heat trace numbers M-2627-1 through M-2627-35. During this telephone conversation, Ramball also confirmed that the 2 inches gauge length would have probably required a test specimen (tube length) of approximately 8-10 inches.
Conclusions based on NRC Review of Documentation and Discussion with Ramball and TLI Based on the above information, AIT provided TLI a total of 5-1/2 + 10 + 2-1/2 + 2-1/2 + 4 which equals = 241/2 inches of tube.
Adding the 1 inch of tubing that AIT provided to Ramball[it is assumed that this is also the minimum amount required for all chemical analyses), a total of 25-1/2 inches of tubing from each of the 35 tubes would have been required to be provided to the test laboratories.
Available Tube Lenoth Provided by the Tube Sucolier and Manufacturer On 4/30/98, Jim DeLer of Marmon/ Keystone (the supplier of the 36 tubes to AIT) provided the shipping document (packing slip) for the tubes in question to the staff via FAX. The following information was on the packing slip:
36 pieces of.500 OD X.0.49, T316, tube Random lengths of tubirg was provided with each length being a minimum of 20 feet +
Tube weight was 0.2360 per foot Actual weight of 170 pounds and actuallength of 720 feet shipped On September 25,1998, the staff called Jim Deler of Marmon/ Keystone Corporation and discussed the 36 tubes that had been provided to AIT. Mr. DeLer clarified that the stainless steel tubes had been special ordered by Marmon/ Keystone in 20 ft lengths. He further confirmed that the hand written entries, for actual weight and actual length, on the packing slip were tne actual measurements taken by personnel packing the tubes and that each tube was menured and that the total measured length of tube was hand written on the packing slip. He furtner stated, again, that the stainless steel tubes were 1/2 inch diameter and supplied by Gibson Tube and that the tube length was 20 feet.
On October 13 and 14,1998, the staff called the manufacturer of the tubes, Gibson Tubo (732) 469-8990, and discussed the length of the tubes supplied to Marmon/ Keystone with Anthony Graviano (Manager of Quality Assurance) who was the individual that signed the certification supplied to Marmon/ Keystone. Mr. Graviano stated that Gibson Tube had never been accredited by ASME and that Gibson Tube was a commercial manufacturer which does not have a nuclear QA program. He confirmed that the tubes supplied to Marmon/ Keystone in
.1994 (which were then shipped to AIT) were not subject to Part 21 or Appendix B type QA controls and were sold as commercial grade tubes. On October 14,1998, Mr. Graviano called l-the staff and provided the following information concerning the tubes supplied Marmon/ Keystone PO No. 75S31820 (Gibson Shop Order No. 46707):
l l
Mr. Graviano was unable to locate the records for the specific order sent to Marmon/ Keystone. His commercial program does not require that he maintain such records, but he was able to find other records of similar tubes supplied to Marmon/ Keystone during that same year.
The records that could be found for similar tubes supplied to Marmon/ Keystone during the period when the tubes in question were ordered were supplied in 20 foot straight lengths.
Gibson Tube's procedures require that when an order is for a tube 20 feet long, the tube length tolerance is 20 feet (+ 1.0 inch, -0 inch).
Mr. Graviano stated that he was reasonably certain that the length of tubes supplied to Marmon/ Keystone were 20 feet long (+1.0 inch, -0 inch).
Further, AIT signed off and stamped the Marmon/ Keystone packing slip for the 36 stainless steel tubes that contained the above information (e.g., total linear feet shipped) as " RECEIVED
- SUBJECT TO INSPECTION - A.I.T."' on May 11,1994. It appears that the AIT review of the packing slip and inspection of the 36 tubes, as annotated by the stamped and sign off on the Marmon/ Keystone packing slip, indicates satisfactory acceptance of the shipment as required by the AIT quality assurance program. Because the documented AIT sign off of the Marmon/ Keystone packing slip indicated acceptance of the information contained on the shipping slip, it is assumed that the information on the packing slip regarding the weight and length of each tube supplied was correct.
Based on the above, the length of tubes provided by Marmon/ Keystone to AIT were 20 feet long + a fraction of an inch up to 1 inch.
Tube Lenoth Fabrication According to AIT Fabrication Drawing No. N-442-1, Revision 1, dated 5/26/94,56 pieces of tube, each 48 inches long were required for each of the three heat exchanger fabricated. A total of (56 X 3 X 48 inches)/12 = 672 feet of tubing would be required to fabricate the three heat exchangers.
Discrecancy in Tube Lenoths According to AIT's PO and Marmon/ Keystone shipping documents, a total of 720 feet (+ one inch maximum according to Gibson Tube; 36 tubes X 1 inch = 36 inches] 720 + 3 = 723 feet of tubing was shipped. Therefore 723 - 672 = 51 feet of tubing was available for testing.
According to the test labs, a total of 35 (the additional test sets] x (25-1/2 inches) + 1 (the initial test set] X (10 + 1= 11 inches) = 75 feet and 3 inches of tubing would be required to perform I
the testing. The staff was not able to determine the length of the tube samples that AIT provided to Ramball for performing the one set of flattening, bend, and flange tests. Therefore, 4
A the 75 feet and 3 inches of tubing required to be sent to the test laboratories is conservative and does not include the lengths of tube that AIT provided to Ramball for the one set of flattening, bend, and flange tests. Therefore, there was a shortage of approximately 24 feet of tubing. [24 ft being conservative - giving AIT benefit of doubt]
Further, looking at the fabrication of the heat exchangers, if 24 to 251/2 inches were cut from each tube for testing, that would only leave 18 [ conservative] feet of tubing to ute to fabricate the heat exchanger from which only four lengths of 4 feet could be cut (from each 18 feet of tubing). Therefore, a total of 36 X 4 = 144 pieces of heat exchanger tube could be cut from the leftover tube and a total of 168 pieces each 4 feet long would be required to fabricate the three heat exchangers. Therefore, (168-144=24), there would be a shortage of 24,4 ft long tubes, and all three heat exchangers could not have been fabricated, if out of the 720 to 723 total feet of heat exchanger tubing (36 pieces, each 20 ft + 1 inch max.
long), each 20 feet long piece was cut in 4 foot lengths, %re could possibly be a total of no more than 48 feet of tubing left over for testing; therefore, only 12 tubes could be tested. To fabricate the heat exchangers, all of the tubing in 24 of the 20 foot lengths (24 X 5=120,4 foot lengths) and all but 4 feet in each of the remaining 12,20 foot long pieces (12 X 4 = 48,4 foot lengths) would need to be used. Therefore, there would be 12,4 foot lengths, of tubing leftover to cut the test samples from and that is only 48 feet. Note: The total of 120 + 48 is the number of 4 foot long pieces of tube required to fabricate the three heat exchangers.
Summary: There is an unexplained discrepancy between the amount of tubing supplied to AIT and the amount required to fabricate the heat exchangers and to perform the claimed testing. It appears that either some test material came from another lot or heat of material, or some heat exchanger tubes came from another lot or heat of material.
2.
Compliance to the 1971 Edition of the ASME Section lil Code in AIT's August 14,1998, and September 1,1998, letters to the NRC, Mr. Amer has implied that the tests being questioned by the NRC are beyond the requirements set farth in the 1971 Edition of ASME Ill and Appendix B. The tests that the staff reviewed were the chemical analyses and mechanical tests which are required to be performed by the ASME Section lli Code,1971 Edition.
AIT is not in compliance with the 1971 Edition of the ASME Code (1971 Section til Code) in that AIT did not purchase the 36 tubes from a Matenals Manufacturer as defined in ND-2110,
" Pressure Retaining Materials," of the 1971 Section 111 Code. Further, neither the manufacturer nor the supplier of the 36 tubes provided the Materials Manufacturer's Certification required by ND-2130," Certification by Materials Manufacturer," of the 1971 Edition of Section Ill.
Throughout the AIT correspondence to the NRC (e.g., AIT letters dated March 7,1996 and October 11,1996, reference is made to NA-3711(b) of the 1971 Section 111 Code and the fact that this section of the Code permits AIT to test only one piece from one of the 36 tubes and thus, AIT meets the 1971 Edition of Section Ill. The 1971 Edition of Section ill contains no such section. The 1971 Section ill Code had no provisions for material upgrading and required v-.
~
Q d
that the component Manufacturer (AIT) purchase material from a Materials Manufacturer that had a QA program that me' the applicable provisions of the 1971 Section lli Code.
AIT purchased the tubes from non-qualified ASME source and according to the NRC inspection report, a non AIT qualified supplier. Further, it appears that the AIT's PO for the tubes did not invoke the supplier's QA program and there were no supplier's certifications indicating that the tubes met the requirements for Class 3 of the 1971 Section 111 Code. AIT states that it did audit Marmon/ Keystone and approved it for maintaining traceability. The below note provides additional discussion on AIT's qualification of Marmon/ Keystone.
In order to meet the 1971 ASME Section lll Code requirements, AIT would be required by the 1971 Section lli Code to confirm that the tubes purchased met applicable Section lli and Section 11 (Material Specifications) Code requirements. Because AIT had not established material traceability *(see Note) from the manufacturer to the point of installation as required by NA-8122, " Quality Assurance Program Requirements for Class 3 Construction," ND-4120,
" Certification of Materials and Fabrication by Component Manufacturer And/Or Installer," and ND-2600, " Material Manufacturers' Quality Assurance Programs," AIT had to perform the required chemical and mechanical tests to meet the Code requirements for materials.
a Therefore, the chemical and mechanical tests that the NRC is questioning would be required to be performed by AIT as specified in Section XVil," Supplying and Upgrading of Materiai," of the AIT Quality Assurance Manual, Revision 8. Section XVll of AIT's QA Manual requires that when upgrading material, each piece to be upgraded must be subjected to the chemical and i
mechar'70al tests required by the applicable material specification and any additional test requireo oy the applicable Edition of the Section lli Code.
determine if tney maintained traceability of material and implementation of the QA program (See AIT letter to NRC dated February 6,1996). The tubes were manufactured by Gibson Tubes and supplied to Marmon/ Keystone. It is unclear as to whether Gibson Tube was on Marmon/ Keystone's approved supplier list, but based on 4
j the staff's contact with Gibson Tube on October 13,1998, it is clear that Gibson tube was never accredited by ASME and only has a non-nuclear commercial type QA Program. Further, Gibson has no Part 21 reporting responsibility since it only provides a commercial grade product. This appea,s to be a moot point because AIT's PO did not invoke any quality requirements on Marmon/ Keystone. The AIT audit / survey of, and PO to, Marmon/ Keystone is not discussed in the inspection report and the AIT audit report of Marmon/ Keystone is not included in any of the documentation contained in the files that the staff reviewed. It appears that both the NRC inspection team as well as the ASME survey teams considered the AIT audit / survey of, and PO to, Marmon/ Keystone to be an unacceptable basis for testing only one of the 36 tubes, and was not consistent with the provisions contained in AIT's QA Manual. Because the tubes were manufactured and supplied by Gibson Tube as commercial grade products and as such Marmon/ Keystone performed no tests to confirm that the tubes met the requirements of the 1971 Edition of Section Ill, AIT's ASME accredited QA Program requires that such tube material be upgraded.
_= -
In summary, if AIT had purchased the 36 tubes as required by the 1971 Section lli Code from a Materials Manufacturer (approved source) and had received the required Section lli material certification and associated documentation and records, the additional chemical and mechanical testing would have not been required. However, since AIT purchased the 36 tubes apparently from a non-qualified source (AIT contends that it did survey Marmon/ Keystone and this was provided to the NRC inspectors in a February 6,1997, letter - a year after the conduct on the inspection] and with no QA program requirements invoked, and the tube manufacturer, Gibson Tube, supplied the tubes as commercial grade tubes, AIT was required by its QA Manual and the Code to perform the chemical and mechanical tests. Further, there was no certification provided by the manufacturer or supplier of the tubes that indicated that the tubes met the requirements of ASME Section Ill.
3.
Mixed Heat / Lot Discussion AIT contends that the differences in the Carbon and Sulfur contents of the tube sections tested (e.g., Carbon varied from 0.012 to 0.032 percent and Sulfur varied from 0.015 to 0.009 percent) vs Mill Certification (Carbon of 0.016 and Sulfur of 0.004 percent) could have been caused by using an analytical method at the testing lab different from the Mill methods. The staff's review indicates that there is a possibility that the material sent out for testing the set of 35 tubes may not have come from the same heat as documented on Marmon/ Keystone's certification.
The staff's review of tolerances provided in ASTM E-1086-85, " Optical Emission Vacuum Spectrometric Analysis of Stainless Steel By the Point-To-Point Excitation Technique," the variances from the original mill certification for the elements Carbon and Sulfur are conservatively (+,-) 25 percent. Therefore, based on E-1086-85 and the staff's engineering judgement, for the reported Carbon content of 0.016 percent and reported Sulfur content of 0.004 percent identified on the original material certification provided by Gibson Tube to Marmon/ Keystone, an acceptable range for these elements would be: Carbon (0.012 to 0.20 percent) and Sulfur (0.002 to 0.009 percent).
It should be noted that the results for the one sample that AIT sent out initially (to Ramball) when it first received the 36 tubes, indicated that the piece of tube tested contained: Carbon (0.020 percent) and Sulfur (0.006 percent) These results are within the (+,-) 25 percent tolerance.
However, subsequent to constructing the heat exchangers, AIT sent 35 pieces of tube out for testing after AIT was informed by the ASME Survey team that testing was required for each tube used to construct the heat exchangers. The following are some of the subsequent test results for these 35 samples::
M-2627-4, Carbon (0.019 percent), Sulfur (0.012 percent)
M-2627-7, Carbon (0.016 percent), Sulfur (0.014 percent)
M-2627-10, Carbon (0.023 percent), Sulfur (0.012 percent)
M-2627-30, Carbon (0.023 percent), Sulfur (0.015 percent)
Note: Of the 35 test samples,9 had Sulfur of 0.012 percent, and 15 had Sulfur of 0.011 percen'.
r.-
- 1 a
The above results are not within the (+,-) 25 percent tolerance for Sulfur.
In summary, while not conclusive, there is a variability in the claimed test results that exceeds the expected variations that are inherent in the spectrometric and wet chemistry analysis methods.
i 1
6 i-8-
4 y
7,
.