ML18092B473
| ML18092B473 | |
| Person / Time | |
|---|---|
| Site: | Salem  |
| Issue date: | 03/24/1987 |
| From: | Corbin McNeil Public Service Enterprise Group |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML18092B474 | List: |
| References | |
| NLR-N87041, NUDOCS 8703310420 | |
| Download: ML18092B473 (6) | |
Text
1-:-
Public Service Electric and Gas Company Corbin A. McNeill, Jr.
Vice President -
Public Service Electric and Gas Company P.O. Box236, Han cocks Bridge, NJ 08038 609 339-4800 Nuclear
- MAR 2 4 1987 NLR-N87041 U. S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 Gentlemen:
AUGUST 26, 19~6 REACTOR TRIP AND FALSE BLACKOUT SIGNAL AT SALEM UNIT NO. 2; STATUS OF CORRECTIVE ACTIONS SALEM GENERATING STATION UNIT NOS. 1 AND 2 DOCKET NOS. 50-272 AND 50-311 This letter forwards the sixth report on the status of corrective actions take~~in response to the August 26, 1986 reactor trip and false blackout signal which occurred at Salem Unit No. 2. The submittal has been delayed in order to: 1) fully address concerns expressed by our Offsite Safety Review Group relative to the 10CFR50.59 analysis prepared for the proposed short term modifications, and 2) to assess the impact of the outage of the Hope Creek -
Keeney line on the minimum expected 5 0 0 kV system voltage. Each of these items has been addressed in the enclosed Safety Evaluation (Attachment 4).
In our previous reports we identified the completion of a
detailed transient analysis and the preparation of* a long term action plan as the outstanding actions to be completed by Public Service Electric and Gas Company (PSE&G). These activities have been completed and the results were discussed with Region I representatives on February 24, 1987 in a meeting at the Salem Generating Station.
r *
.e MAR 2 41987 Document Control Desk 2
The detailed transient analysis (Attachment 2) conducted by Power Technologies, Inc. (PTI) demonstrates that adequate voltage for operation of safety related equipment can be assured, provided that Salem electrical distribution system loads remain at their presently reduced levels.
This load reduction is being accomplished by providing electric power to six Circulating Water pumps (3/Unit) from the Hope Creek 13 kV system. It is expected that this arrangement will continue until an overall plan for system improvement is formulated. Additionally, PSE&G conducted a
Validation Study of the PTI computer model used for the transient analysis.
Attachments 1 & 2 to this letter includes the PTI computer simulation data for the validation test. is the PSE&G evaluation of the acceptability of the computer model for predicting system transient response.
A Long Term corrective action program has been established and consists of the following elements; 1) determination of the root cause of the electrical distribution system
- anomaly,
- 2) identification of corrective actions to prevent recurrence, and
- 3) identification of electrical distribution system improvement options. Items one and two are complete. A Task Force has been assembled to perform the necessary work associated with item 3. A brief discussion of each of these items is presented in the following paragraphs.
Root Cause After substantial analysis and evaluation, it has been determined that the present design of the degraded grid protection scheme is deficient. The use of a 95% setpoint for the reset of the degraded grid protection relays, and the time coordination between the transfer and bus protection relays, provides insufficient margin to prevent the flip-£ lop of the vital buss es between their preferred power sources. on August 26, 1986 immediately following the reactor trip and subsequent safety injection, the 4 kV vital bus voltage recovered to
~95%.
Due to the inherent uncertainties associated with the relay calibration process, this recovery value was insufficient to assure positive reset of all relays. The lack of positive reset allowed the vital buses to flip-flop until the timing sequence completed. At that point the Safeguards Equipment Controller logic initiated an automatic start of the emergency diesel genera tors and transferred all vital bus loads to their respective emergency source.
Attachment 4
includes a
.e UAR 2 4 1qg7 Document Control Desk 3
discussion of the selection of the relay setpoints and their contribution to the root cause of the event.
Corrective Actions In order to preclude recurrence, PSE&G proposes to redesign the present degraded grid protection scheme. The present design utilizes 2 relays per vital bus (91% to trip, 95% to reset, 10.5 second time delay) to initiate individual vital bus transfers, and an additional relay (91% to trip, 95% to reset, 13 second time delay) to provide single bus input to the SEC logic. This design: 1) initiates a bus transfer if voltage drops below 91%
for 10. 5 seconds and does not recover to above 95%, 2) separates the bus from the offsite source if voltage does not recover above 95% within 13. 5 seconds and, 3) automatically transfers all vital busses to the emergency source if a sustained degraded voltage occurs on 2-out-of-3 vital busses.
Primary protection against catastrophic failures (i.e. the loss of a Station Power Transformer or a genuine loss of offsite power event) is also provided. This system is similar in design, with the exception of relay setpoints ( 70% to trip, Inverse Time relays) and function to the degraded grid protection design.
The new design will eliminate the potential for flip-flop by removing the 91% bus transfer logic. The 2 transfer relays per vital bus would be rewired into the individual bus SEC to provide a
2-out-of-3 relay per bus logic for SEC actuation.
Upon satisfying the 2-out-of-3 relay per bus logic, only the affected bus would be separated from its pref erred source and loaded onto its emergency source. A degraded voltage on 2 busses would result in all three vital busses being loaded onto their emergency sources, however; this condition would be initiated by the 70% undervoltage logic. The 70% undervoltage logic would be completed due to the time interval (~10 seconds) between the stripping of the bus from the preferred source and the time that it is reenergized from the emergency source. This modification provides additional conservatism in the design by improving the redundancy of the SEC logic inputs. Additionally, the change would be consistent with the NRC's originally proposed Technical Specification for degraded grid protection.
Subsequent to the February 24 meeting, that the modifications described above change to the Salem Unit No.
1 it was would and 2
determined require a
Technical
. e MAR 2 4 1981' Document Control Desk 4
Specifications.
The required changes are identified in.
In consideration of the time requirements imposed by the Shelly process and the desirability of proceeding with the proposed modifications, a
two step installation plan has been prepared. The initial step will proceed in accordance with the requirements of 10 CFR 50.59 and would delete the transfer scheme (not a
Technical Specification item) and rewire the transfer relays ( 2 per vital bus) to the individual bus SEC. This will provide a 2-out-of-3 relay per bus logic which would operate in parallel with the existing 2-out-of-3 bus logic. The relays would also be recalibrated during this step to reset at a value of 92.6%.
Upon completion of this relay modification, the AUXILIARY POWER TRANSFORMERS will be returned to service to power the non-vital group bus loads with the exception of the circulating water pumps which will continue to be powered from the Hope Creek 13 kV system.
The single bus protection relay (Technical Specification item) which presently provides inputs to the 2-out-of-3 bus protection logic would remain in its existing configuration until such time as the required Technical Specification Changes have been approved. At that point, step two of the installation process would consist of reconfiguring the SEC inputs to provide only the desired 2-out-of-3 relay per bus logic protection. This 2 step process was discussed with Region I representatives in a telephone conference on March 3 I 1987 o An additional concern relative to the Technical Specifications was raised by our Offsite Safety Review Group during its review of the 10CFR50.59 Safety Evaluation. Upon review of the minimum allowable vital bus voltage specification, it was determined that the value specified
(~90% of bus voltage) was non-conservative in that it failed to account for the voltage drop due to cable losses for the worst case load. An extensive review of the basis for this specification identified additional inconsistencies in several supporting documents. It was determined that no immediate safety concern existed as the relay setpoint of 91% provides sufficient margin to account for the calculated voltage drop. In order to correct the situation, a new setpoint calculation was performed and all relays are being recalibrated. The new setpoints will be administratively controlled until a formal change can be processed. A more detailed discussion of this
- matter, including the appropriate setpoint calculations, is provided in Attachment
- 4. to this letter includes a computer simulation supporting the acceptability of the proposed modification.
. e MAR 2 4 1987 Document Control Desk 5 is the PSE&G Safety Evaluation and Attachment 5 provides a
complete description of the proposed modification.
Plant Improvements A Task Force has been formed to conduct a detailed review of the Salem electrical distribution system design basis and identify potential improvements which could result in increased safety and reliability, and add flexibility to the present system. Attachment 6 identifies the overall goals and objectives of the Task Force and defines the scope of review to be performed. The present schedule for completion of all review work and proposal of system improvements is July 1, 1987.
At the conclusion of the meeting a question was raised regarding the existing level of separation in the "blackout" protection inputs to the SEC. It was concluded that, because this system exists in its original configuration, no immediate concern relative to acceptability was presented.
However, PSE&G agrees that it would be appropriate to review the design in this area as part of any long term improvement program.
With respect to future status reports, we believe that our outstanding commitments are satisfied with this submittal and, unless specifically requested, this will be the last regular report on this issue.
We plan to initiate the configuration changes which will enable the auxiliary power transformers to power the non-vital group bus loads within one week.
We will contact you at that time to ensure you concur with our approach.
In the interim, should you have any questions on the information presented herein, please feel free to contact us.
Attachments Sincerely, C. A. McNeill, Vice President
. e Document Control Desk 6
c Mr. William F. Kane Director, Division of Reactor Projects Mr. Donald c. Fischer (w/attachment)
Licensing Project Manager Mr. Thomas J. Kenny (w/attachment)
Senior Resident Inspector Mr. Om Chopra (w/attachment)
Electrical Instrumentation & Control Systems Branch MAR 2 4 tsar