ML090400748
| ML090400748 | |
| Person / Time | |
|---|---|
| Site: | Monticello, Salem, Indian Point, Point Beach, Surry, Turkey Point, Ginna, Robinson  |
| Issue date: | 02/05/2009 |
| From: | Weinstein M SCIENTECH |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| Download: ML090400748 (6) | |
Text
143 West Street, New Milford, CT 06776 Phone: 860.210.3000 - Fax: 860.210.3021 www. scientech. com OSCIENTECH February 5, 2009 Attn: Document Control Desk U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001
Subject:
Update to Part 21 Report Notification, Texas Instruments Operational Amplifier, P/N: OPA 2111
References:
Scientech Part 21 Notification #21-03-01, Event #39516, Date 1/16/03.
Scientech Part 21 Follow-up Report #21-03-01-02, Event #39516, Date 2/15/03.
Dear Sir or Madam:
The purpose of this letter is to provide an update report to the original Scientech Part 21 Notifications referenced above. At the time of the original notifications regarding failures of Burr Brown OPA2111 KP chips manufactured between January 2000 and October 2002 as described in the references, Scientech conservatively prohibited purchase of similarly manufactured chips in three other amplifiers including the 3556BG amplifiers.
In October 2008, per the attached Technical Bulletin Volume 4, Failures of Texas Instruments Amplifiers, Scientech requested and received approval from the utilities listed in Attachment 2 to use 3656BG chips that were manufactured between January 2000 and October 2002 based on an extremely low failure rate. Although the data showed no need for additional precautions, to be conservative, Scientech performed a two-week bum-in of all modules that use 3656BG chips built in that time period.
Should you have any questions regarding this matter, please contact Robert Queenan, Operations Manager, Scientech/NUS Instruments, at (208) 524-9311.
Sincerely, Michael B. Weinstein Director of Quality Operations Scientech (860) 514-9203 B
Page 1 Flow Control Nuclear
Update to Part 21 Report Notification, Texas Instruments Operational Amplifier, P/N: OPA 211 Attachments:
- 1. NUS Instruments Technical Bulletin Volume 4, Failures of Texas Instruments Amplifiers, Dated October 2008.
- 2. List of Utilities Approving use of 3656BG Chips.
Cc:
Scott Robuck Robert Queenan LaWanda Wold Rebecca Baumgartner QA File Page 2
Update to Part 21 Report Notification, Texas Instruments Operational Amplifier, P/N: OPA 211 NUS Instruments Technical Bulletin Volume 4, Failures of Texas Instruments Amplifiers, Dated October 2008
NUS Instruments LHI cZD A\\
Volume 4 Page I of 2 IssueDate.
October 2008.
From:
Scientech, a Curtiss Wright Flow Control company Technical Point Of
Contact:
Bob Queenan 200 S.'Woodruff Avenue Idaho Falls, ID 83404-(208) 524-9311 Subject-Failures-of Texas-Instrument Amplifiers-
Background
Texas Instruments (TI) provides Burr Brown amplifiers. There were three failures of Scientech modules in early 2003 that were traced back to a Burr Brown OPA21"IT KP chip. T-.I-identified a-manufacturingproblem-that affect-ed.the reliability of some of their OPA21 I 1KP chips built between January 2000 and October 2002, and determined that a two-week bum-in of the chips would identify any problems.
Scientech uses OPA2111 KPs in many NUSI modules, including safety related and non-safety related applications. We issued a notification pursuant to 1 OCFR21 informing all affected clients of the problem, and recommending that if any modules built with these components-had not been-in-operation for at least two-weeks, that-the modules-undergo-a-two week bum-in to identify any OPA21 1 KP op amps with theproblem. At this time, Scientech committed to test any chips in house for the problem, and to prohibit purchase of more chips from the problem time frame.
Texas Instruments used the same manufacturing process for three other chips in use at Scientech/NUSI - the OPA404AG, the OPAlIBM, and the 3656BGamplifiers.
Although-there had-been--no-failures of these -chips, Scientech~extended-:the-ýcorrective action to these three types as well,. testing chips in house and prohibiting. further purchase of suspect chips.
Current Issue Recenty,. Texas Instruments, informed Scienterh that. they were putting the. manufacture of 3656BG chips on hold. They were experiencing a low yield due to contaminants getting into a coating used on the chip. Although the problem was detected during their normal quality control inspections, they were unwilling-to continue the line until-the yield was sufficient to-be cost effective.
Since Scientech-uses this product in many -safety-and non-safety related-products, -we-immediately attempted.to.procurea-sourceof3656BGs.from distributors.- We.quickly exhausted the limited supply of new chips. However, we found a supply of unused chips manufactured between January 2000 and October 2002.
Scientech/NUSI maintains a repair facility, and suspected that the 3656BG had a very low failure rate. We examined the repair records for any modules built in 2000, 2001, or 2002. There wereover300modules built duringthat time-frame, and many used
.multipie3656BG-chips., Ofall._those modules, tbhere.has-been.only one.failure-recorded due to a 3656BG chip - a module built in 2001 that failed in 2004. This extremely low failure rate demonstrates that the 3656BG chips manufactured between
NUS instruments cc ml Volume 4 Page 2 of 2 January 2000 and October 2002 do not have the reliability problem that the OPA2111 chips manufactured in this same time frame evidenced.
In addition, Scientech asked Texas Instruments for failure data on the four types of chips. They supplied the following:
0 S
0 0
96,482 OPA21 1 KP shipped - 3 confirmed failures 561,352 OPA I IBM shipped - 6 confirmed failures 219,760 OPA404AG shipped - no failures reported 29,496 3656BG shipped - no failures reported Scientech also explored the failure modes likely in the 3656BG amplifier. Only time and over voltage applied to the integrated circuit power rails while driving the output to the rail, have been identified as failure mode inducers. Buffering of the internal op amps within the 3656BG makes failure of the internal op amps much less likely. The isolation function of the 3656BG is in no way compromised by the process problem and any of its effects. Failures are not temperature related and the AC power rails are buffered, no common cause event could be identified which might induce simultaneous failures in modules of redundant channels.
In summary, Scientech has found a very low failure rate of 3656BGs in modules in the field, Texas Instruments has had no reported 3656BG failures, and there is no failure mode that would lead to simultaneous failures in multiple safety channels.
Therefore, Scientech was overly conservative in prohibiting the use of the 3656BG chips built between January 2000 and October 2002. The failure history of these chips demonstrates that the problem that affected the reliability of the OPA211 KP did not affect the reliability of the 3656BG. Therefore, Scientech is revising its eommitment-and will no longer prohibit the purchase of 3656BG chips built in that time period.
Although the data shows no need for additional precautions, for conservatism Scientech will perform a two-week bum in of all modules that use 3656BG chips manufactured between January 2000 and October 2002, and will obtain client approval before shipping any modules with such chips in them.
Note: Technical Bulletins are for information only and are not controlled under the NUSI QA program.
Any actions to be taken by NUSt in response to an issue shah be tracked. through implementabtion.
OA-controlled documents independently of the Technical Bulletin.
Update to Part 21 Report Notification, Texas Instruments Operational Amplifier, P/N: OPA 211 Utilities Approving the use of 3656BG Chips Constellation Energy R.E. Ginna Nuclear Power Plant 1503 Lake Road Ontario, NY 14519 Dominion Virginia Power Surry Nuclear Plant 5570 Hog Island Road Surry, VA 23882 Entergy Nuclear Operations Inc.
Indian Point Energy Center Town of Cortlandt Buchannan, NY 10511-0249 Florida Power & Light Point Beach 659 Nuclear Road Two Rivers, WI 54241 Florida Power & Light Turkey Point 9760 SW 344 Street Florida City, FL 33035 Progress Energy H.B. Robinson P.O. Box 1551 Raleigh, NC 27602 Public Service Electric & Gas Salem P.O. Box 236, MC N47 Hancocks Bridge NJ 08038 Xcel Energy Monticello Nuclear Generating Plant 2807 W. County Road 75 Monticello, MN 55382