ML070440175
| ML070440175 | |
| Person / Time | |
|---|---|
| Site: | Oconee, Mcguire, Catawba, McGuire  |
| Issue date: | 01/31/2007 |
| From: | Gordon Peterson Duke Energy Carolinas, Duke Power Co |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| GL-06-002 | |
| Download: ML070440175 (9) | |
Text
Duke GARY R. PETERSON D uke Vice President Energy McGuire Nuclear Station Duke Energy Corporation MGO1 VP / 12700 Hagers Ferry Rd.
Huntersville, NC 28078 704 875 5333 January 31, 2007 704 875 4809 fax grpeters@duke -energy. corn U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555-0001
Subject:
Duke Power Company LLC d/b/a Duke Energy Carolinas, LLC (Duke)
Oconee Nuclear Station, Units 1, 2, and 3 Docket Nos. 50-269, 50-270, 50-287 McGuire Nuclear Station, Units 1 and 2 Docket Nos. 50-369 and 50-370 Catawba Nuclear Station, Units 1 and 2 Docket Nos. 50-413, 50-414 Response to NRC Request for Additional Information Letter dated December 5, 2006, Generic Letter 2006-02, Grid Reliability and the Impact on Plant Risk and the Operability of Offsite Power.
Pursuant to 10 CFR 50.54(f), the enclosed information is being provided regarding the NRC Request for Additional Information (RAI) letter dated December 5, 2006 "Generic Letter 2006-02, Grid Reliability and the Impact on Plant Risk and the Operability of Offsite Power." As directed in the request, Duke Energy Corporation is responding to Questions numbered 1, 3 and 5 in accordance with NRC letter dated December 13, 2006 extending the RAI due date to January 31, 2007.
Please note that the Duke responses to the RAIs for Generic Letter 2006-02 included in this letter provide information about analyses, procedures, and activities concerning grid reliability that have been provided by the Power Delivery Department of Duke in response to the request of the Duke nuclear power plants. The Power Delivery Department operates under requirements and procedures that are different than those of the Duke nuclear power plants.
This letter and attachments do not contain any new NRC commitments.
Inquiries on this matter should be directed to B.K. Marrow at (704) 875-5878 or K.L.
Ashe at (704) 875-5715.
www. duke-energy. corn
U.S. Nuclear Regulatory Commission Page 2 Date January 31, 2007 7 inerel, f4zzm Gary R. Peterson
Attachment:
RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02
U.S. Nuclear Regulatory Commission Page 3 Date January 31, 2007 xc (with attachment):
W.D. Travers U.S. Nuclear Regulatory Commission Regional Administrator, Region II Atlanta Federal Center 61 Forsyth St., SW, Suite 23T85 Atlanta, GA 30303 D. W. Rich Senior Resident Inspector (ONS)
U.S. Nuclear Regulatory Commission Oconee Nuclear Station J.B. Brady Senior Resident Inspector (MNS)
U.S. Nuclear Regulatory Commission McGuire Nuclear Station A. T. Sabisch Senior Resident Inspector (CNS)
U.S. Nuclear Regulatory Commission Catawba Nuclear Station L.N. Olshan NRC Senior Project Manager (ONS)
U.S. Nuclear Regulatory Commission One White Flint North, Mail Stop 0-8 G9A 11555 Rockville Pike Rockville, MD 20852-2738 J.F. Stang NRC Project Manager (MNS and CNS)
U.S. Nuclear Regulatory Commission One White Flint North, Mail Stop 8 H4A 11555 Rockville Pike Rockville, MD 20852-2738
U.S. Nuclear Regulatory Commission Page 4 Date January 31, 2007 Gary R. Peterson affirms that he is the person who subscribed his name to the foregoing statement, and that all the matters and facts set forth herein are true and correct to the best of his knowledge.
Gary R. Peterson, Vice President Subscribed and sworn to Nofr fi My commission expires:
Date
/-3/- 67 4C SEAL
ATTACHMENT RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02
RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02 Offsite Power Operability RAI Question 1: Switchyard Minimum Voltage In response to question 1(g) you did not identify specific minimum switchyard voltage limits (kV) that you supplied to the local transmission entity. Please, provide the following information:
What is the specific minimum acceptable switchyard voltage included in your protocol agreement with your grid operator (GO) and what was the basis for this value?
How is this value related to your technical specification degraded voltage relay setpoints?
Duke Response The established Service Level Agreement and related Department Directives require that the Nuclear Power Plant (NPP) periodically provide the GO with the calculated minimum acceptable degraded grid voltage values for each NPP. This update is currently provided at an annual frequency as required by these documents. The bounding values for normal offsite power alignments that are provided to the GO as of the date of this response are as follows:
McGuire: 224.14 kV (Unit 1) & 503.37 kV (Unit 2)
Catawba: 221.85 kV (Units 1 & 2)
Oconee: 230.206 kV (Units 1, 2, & 3)
These required minimum Switchyard voltage values are subject to change based upon system design changes.
Catawba & McGuire The Degraded Voltage relaying is set up to monitor the 4 kV Emergency Bus voltage.
The minimum safety equipment voltage is determined at the 4 kV Emergency Bus voltage level. This required bus voltage is translated from the 4 kV voltage level to the applicable corresponding Switchyard voltage level (230 kV or 525 kV). This is the value that is provided to the GO and is documented in approved calculations.
These minimum Switchyard voltage values are determined based upon preventing the actuation of the Degraded Voltage relaying scheme on the 4 kV Emergency Bus when the Unit generator is tripped off-line and LOCA loads are sequenced on.
The minimum Switchyard voltage values provided to the GO ensure that the voltage will remain above the maximum reset value for the Degraded Voltage relays at the 4 kV I
RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02 Emergency Bus level, and above the Technical Specification Degraded Voltage relay Allowable Value and Nominal Trip Setpoint.
Oconee The Degraded Grid Voltage relaying monitors voltage at the 230 kV Switchyard Yellow Bus. This value provided to the GO is above the Degraded Grid Voltage relay maximum reset setpoints. The value is supported by voltage drop calculations for bounding design basis LOCA loading events.
The minimum 230 kV Switchyard voltage value provided to the GO is above the Technical Specification (TS) Degraded Voltage relay Allowable Value and the relay maximum reset value.
RAI Question 3: Verification of RTCA Predicted Post-Trip Voltage Your response to question 2(g) indicates that you have not verified by procedure the voltages predicted by the online grid analysis tool (software program) with actual real plant trip voltage values. It is important that the programs used for predicting post-trip voltage be verified to be reasonably accurate and conservative.
What is the range of accuracy for your GO's contingency analysis program?
Why are you confident that the post-trip voltages calculated by the GO's contingency analysis program (that you are using to determine operability of the offsite power system) are reasonably accurate and conservative?
What is your standard of acceptance?
Duke Response The range of accuracy for the GO's contingency analysis program has not been determined quantitatively. The contingency analysis program only archives result data when a contingency causes a limit violation. The contingency analysis program would have to have a contingency limit violation at the exact time of a unit trip in order to have the contingency analysis program's predicted post unit trip voltage so it then could be compared to actual unit post trip voltage readings. The accuracy of the contingency analysis application is dependent upon a variety of items that can change each time the application solves. To reduce the probability of inadequate nuclear switchyard voltage (and a contingency analysis program limit violation), the GO intentionally operates the grid around the nuclear unit switchyards with voltage margin above GO's calculated pre-contingency nuclear switchyard voltage.
2
RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02 Duke has not compared the actual voltage changes to those that would have been computed by our network applications following a unit trip. This is because Duke does not archive the voluminous amounts of pre and post trip information. In addition, archived information may not provide exact results since it may not be the exact information the contingency analysis program used just before and after the unit trip because it depends on when the information is captured.
Duke keeps up with the latest revisions of the vendor's network analysis software on its Energy Management System. The vendor that supplies Duke's network application software follows a formal development and testing process to certify that the revisions have been properly tested. This QA process has been registered against the ISO 9001:2000 standard.
The GO's contingency analysis program is an industry standard commercial grade tool that is used by many of the large entities within NERC. Duke's real-time contingency analysis program is used daily and configured to run automatically every ten minutes, at a minimum. It also runs automatically after generator trips and other defined system events, and can also be run manually by the system operators. Therefore there is continual feedback on the validity of the program results. The GO uses this tool to analyze the entire grid within Duke and a significant portion of the grid external to Duke and not just for the nuclear power plants. The detailed model used by Duke's network applications contains thousands of analog and status point readings that are brought in from hundreds of stations. The model is continually updated to reflect changes in the grid internal and external to Duke (e.g., new lines, stations, etc.) and is tested prior to being put into production. Based upon the reasonableness of the values provided to the GO thru the daily use of this program for the whole grid system, Duke has confidence that the values provided for the specific post-trip voltages are reasonably accurate. The contingency analysis program typically works adequately and operational comparisons show no discrepancies to observed conditions. There are very few instances where the contingency analysis program has generated information that involves invalid output (i.e. model does not produce a valid solution or numerous violations are generated or system anomalies during data collection). When suspect information is generated, these items are investigated for problem identification and correction. Periodically the GO will verify the reasonability of the predicted voltages and flows versus actual voltage and flows following removal of equipment or lines for maintenance. These actions help serve to validate the reasonable accuracy of the tool and provide assurance of accurate results for predicted post trip voltages of the nuclear plants.
The network model used by Duke's network applications, and the alarm limits themselves, have been developed using conservative parameters for the equipment.
The GO intentionally operates the grid around the nuclear unit switchyards with voltage margin above GO's calculated pre-contingency nuclear switchyard voltage. Based upon this preceding information Duke has confidence that the values provided for the specific post-trip voltages are conservative.
3
RESPONSE TO QUESTIONS 1, 3 and 5 of the REQUEST FOR ADDITIONAL INFORMATION RELATED TO GL 2006-02 Duke's standard of acceptance is qualitative in nature based upon the GO's stated confidence and successful daily use of the contingency analysis program. The contingency analysis program is currently a state of the art tool that is the best available tool for use in performing this type of analysis.
There are currently no specific NERC standards for real-time contingency analysis and other such applications. However, the NERC Real-time Tools Best Practices Task Force (RTBPTF) that was created by the NERC Operating Reliability Subcommittee (ORS) as a result of the August 14, 2003 black out is expected to make recommendations for standards in this area when its final report is released in 2007.
Duke is an active participant in this task force.
Maintenance Rule RAI Question 5: Seasonal Variation in Grid Stress (Reliability and Loss-of-offsite Power (LOOP) Probability)
Certain regions during certain times of the year (seasonal variations) experience higher grid stress as is indicated in Electric Power Research Institute (EPRI) Report 1011759, Table 4-7, Grid LOOP Adjustment Factor, and NRC NUREG/CR-6890.
Do you adjust the base LOOP frequency in your probabilistic risk assessment (PRA) and Maintenance Rule evaluations for various seasons? If you do not consider seasonal variations in base LOOP frequency in your PRA and Maintenance Rule evaluations, explain why it is acceptable not to do so.
Duke Response Duke does not adjust LOOP frequency for various seasons. Maintenance Rule a(4) risk assessments performed by Duke account for actual and predicted Grid conditions, actual and predicted severe weather; and planned/emergent maintenance activities.
The seasons (including the summer months) are not homogeneous. Each season has periods of weather that vary in temperature and associated grid stress. Duke manages unit and system outages such that the maximum available margin is maintained during periods when higher system loads are expected. This minimizes the potential for grid stress during peak load seasons. Therefore, Duke believes it is acceptable to use the average LOOP frequency and assess known risk conditions versus adjusting the LOOP frequency seasonally.
4