ML20059B209
| ML20059B209 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 10/19/1993 |
| From: | Colburn T Office of Nuclear Reactor Regulation |
| To: | Gipson D DETROIT EDISON CO. |
| Shared Package | |
| ML20059B212 | List: |
| References | |
| TAC-M83095, NUDOCS 9310280089 | |
| Download: ML20059B209 (3) | |
Text
.s Docket No. 50-341 October 19, 1993 o
Mr. Douglas R. Gipson
' i Senior Vice President Nuclear Generation Detroit Edison Company 6400 North Dixie Highway Newport, Michigan 48166
Dear Mr. Gipson:
SUBJECT:
FERMI PROPOSED TECHNICAL SPECIFICATION CHANGE RELATING TO RESPONSE TIME TESTING REQUIREMENTS (TAC NO. M83095)
By letter dated March 27, 1992, you requested a technical specification (TS) change to eliminate selected response time testing requirements at Fermi Unit 2.
This request was based upon the BWR Owners' Group Licensing Topical Report, " System Analyses for Elimination of Selected Response Time Testing Requirements," General Electric Report NEDC-32013P, dated March 1992. The review of the subject topical report has been completed, and the staff has determined that there was an insufficient bases to justify the elimination of the selected response time testing requirements. Therefore, the topical report was not approved. A copy of the safety evaluation (SE) for the topical report is enclosed. This SE was transmitted to the BWR Owners' Group on August 19, 1993, i
Because the topical report was not approved, we are unable to approve your March 27, 1992, amendment application to delete selected response time testing in the Fermi 2 TS.
Therefore, this portion of your request is denied.
Your request to add a response time testing requirement for main steam line flow-high for the MSIVs will be handled by separate correspondence. A copy of the Notice of Denial of Amendment to be published in the Federal Reaister is also enclosed for your information.
Should the staff and the BWR Owners' Group reach resolution on the unresolved items identified in the enclosed SE related to NEDC-32013P, the staff would document that resolution is a subsequent SE of a revised topical report. At that time, you may wish to resubmit your application.
If you have any questions, please contact me at (301) 504-1341.
Sincerely, "hf G "oNuk,Sr.ProjectManager
$ boob 0 0 1
P Project Directorate III-l gwq Division of Reactor Projects - III/IV/V orggqL bbgy nyp EsLib a s.
aI Office of Nuclear Reactor Regulation t ht h.w i.
Enclosures:
DJSTRIBUTION 1.
Safety Evaluation _
JZwolinski DHagan Docket < File, 2.
Notice of Denial NRC & LPDRs WDean ACRS (10)
PD31 Rdg CJamerson BJorgensen, RIII cc w/ enclosures:
JRoe TColburn OC/LFDCB See next page OGC JWermiel, 8/H/l OPA offlCE LA:PD31 N PM:PD31 k OGC /6/6 (A)D:PDAlb
- E CJamerson TColburn:jkd
[ b 1cinn WDean$ k i
'} / /93 9 /22/93 7 /@/93
/ 0/_fh93 WE
/hNo
/fis)'No bl/No
( es/;No I
COPY s
0FFICIAL ret 0'RD COPY fitENAME: G:\\WPDOCMFERMI\\F E83095~O R h
. I
kC1 Rf C em O
f-W[ E UNITED STATES
- g R.
NUCLEAR REGULATORY COMMISSION g vf,/
WASHINGTON, D.C. 20555-0001 October 19, 1993 Docket No. 50-341 Mr. Douglas R. Gipson Senior Vice President Nuclear Generation Detroit Edison Company 6400 North Dixie Highway Newport, Michigan 48166
Dear Mr. Gipson:
SUBJECT:
FERMI PROPOSED TECHNICAL SPECIFICATION CHANGE RELATING TO RESPONSE TIME TESTING REQUIREMENTS (TAC NO. M83095)
By letter dated March 27, 1992, you requested a technical specification (TS) change to eliminate selected response time testing requirements at Fermi Unit 2.
This request was based upon the BWR Owners' Group Licensing Topical Report, " System Analyses for Elimination of Selected Response Time Testing Requirements," General Electric Report NEDC-32013P, dated March 1992. The review of the subject topical report has been completed, and the staff has determined that there was an insufficient bases to justify the elimination of the selected response time testing requirements. Therefore, the topical report was not approved. A copy of the safety evaluation (SE) for the topical report is enclosed. This SE was transmitted to the BWR Owners' Group on August 19, 1993.
Because the topical report was not approved, we are unable to approve your March 27, 1992, amendment application to delete selected response time testing in the Fermi 2 TS. Therefore, this portion of your request is denied. Your request to add a response time testing requirement for main steam line flow-high for the MSIVs will be handled by separate correspondence.
A copy of the Notice of Denial of Amendment to be published in the Federal Reaister is also enclosed for your information.
Should the staff and the BWR Owners' Group reach resolution on the unresolved items identified in the enclosed SE related to NEDC-32013P, the staff would document that resolution is a subsequent SE of a revised topical report. At that time, you may wish to resubmit your application.
If you have any questions, please contact me at (301) 504-1341.
Sincerely, k-Timothy G. Colburn, Sr. Project Manager Project Directorate 111-1 Division of Reactor Projects - III/IV/V Office of Nuclear Reactor Regulation
Enclosures:
1.
Safety Evaluation 2.
Notice of Denial cc w/ enclosures:
See next page
Mr. Douglas R. Gipson Fermi-2 Detroit Edison Company cc:
John Flynn, Esquire Senior Attorney Detroit Edison Company 2000 Second Avenue Detroit, Michigan 48226 Nuclear Facilities and Environmental Monitorin5 Se: tion Office Division of Radiological Health Department of Public Health 3423 N. Logan Street P. O. Box 30195 Lansing, Michigan 48909 Mr. Wayne Kropp U.S. Nuclear Regulatory Commission Resident Inspector Office 6450 W. Dixie Highway Newport, Michigan 48166 Monroe County Office of Civil Preparedness 963 South Raisinville Monroe, Michigan 48161 Regional Administrator, Region III U.S. Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, Illinois 60137 Mr. William E. Miller Director - Nuclear Licensing Detroit Edison Company Fermi-2 6400 North Dixie Highway Newport, Michigan 48166 Juh 1993
August 19, 1993 ENCLOSURE 1 Mr. L. A. England BWR Owners Group Chairman PO Box 220 St. Francisville, LA 70775
Dear Mr. England:
SUBJECT:
BWR OWNERS' GROUP LICENSING TOPICAL REPORT " SYSTEM ANALYSES FOR ELIMINATION OF SELECTED RESPONSE TIME TESTING REQUIREMENTS,"
MARCH 1992.
(GENERAL ELECTRIC REPORT NEDC-32013P) (TAC M83089)
We have completed our review of the subject topical report submitted by the BWR Owners Group by letter dated March 27, 1992. The enclosure provides our Safety Evaluation Report (SER).
The topical report attempted to show that response time testing (RTT) time requirements could be eliminated by showir.g that changes in response beyond acceptable limits would be detected by other means. We found that the proof of this assumption was neither comprehensive nor compelling. The re i
did not refer to better methods of testing, or list any substitute tests. port We find that there is an insufficient basis in this report to justify the elimination of the selected RTT requirements and, therefore the staff does notapproveGeneralElectricNuclearEnergyReportNEDC-320l3P. A structured engineering approach to RTT in accordance with Section 4 of the SER could form the basis for the requested relief.
Should you have any questions, or wish further clarification, please feel free to call me at 504-1003, or Paul Loeser at (301) 504-2825.
Original signed by:
Bruce A. Boger, Director Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulatio;.
Enclosure:
DISTRIBUTION As stated Central File HICB R/F PDR
.' Y ffbb b
b J. Wermiel C. Thomas /B. Boger k
W P
C PDR
i l
EVALUATION OF GE TOPICAL REPORT ON THE ELIMINATION OF SELECTED RESPONSE TIME TESTING REQUIREMENTS
1.0 INTRODUCTION AND BACKGROUND
i The periodic testing of reactor trip systems is established la Section 50.55a, j
" Codes and Standards," of 10 CFR Part 50, " Domestic Licensing of Production l
and Utilization Facilities." The NRC in 1975 implemented a program making response time testing (RTT) a requirement of the technical specifications (TS).
IEEE Std. 338-1975, " Criteria for the Periodic Testing of Class IE Power and Protection Systems," and its later version IEEE Std. 338-1977,
" Criteria for the Periodic Testing of Nuclear Power Generating Station Safety Systems," (Ref.1) provided generic guidance in the form of response time verification tests. The NRC staff in Regulatory Guide 1.118, " Periodic Testing of Electric Power and Protection Systems," Revision 2, June 1978 (Ref.
- 2) endorsed IEEE Std. 338-1977, and described a method acceptable to the staff for complying with the requirements of the periodic testing of the reactor protection system and other systems important to safety.
The Instrument Society of America (ISA) embarked on a parallel path and in 1986 issued ISA-S67.06-1984, " Response Time Testing of Nuclear Safety-Related Instrument Channels in Nuclear Power Plants," dated August 29, 1986, (Ref. 3).
This document was more definitive than the IEEE standard inasmuch as'the ISA document was devoted to RTT only and not to periodic testing in general.
ISA Standard S67.06-1984 has not been endorsed by the NRC staff. However, the document relates quite well to the response time procedures and test methods currently employed by the nuclear industry.
i
, In May 1991 the Electric Power Research Institute (EPRI) issued Report NP-7243, " Investigation of Response Time Testing Requirements" (Ref. 4). The stated reason for this report was the nuclear industry's efforts to increase i
plant availability and to minimize personnel radiation exposure. The purpose of this report was to justify the elimination of the RTT requirements for the pressure sensors used in the instrumentation loops.
l The General Electric Company in March 1992 issued Report NEDC-32013P, " System -
Analyses for Elimination of Selected Response Time Testing Requirements" (Ref. 5). This report took the justification for the elimination of RTT one step further than did the EPRI report and included the instruments and components of the instrumentation loops.
2.0 DISCUSSION The two most important tests to determine the electrical quality of an 4
installed instrumentation or control channel are response time tests and steady state tests. The response time test of an instrumentation or control channel is time dependent and is defined as:
An output expressed as a function of time, resulting from the i
application of a specified input under specified operating conditions.
J (ANSI /ISA-S51.1-1979, Process Instrumentation Terminology).
j 1
g 9
The definition of an instrument channel response time is:
The time interval from the time the monitored variable exceeds its trip setpoint until the time a protective action is initiated.
(ISA-S67.06-1984, RTT of Nuclear Safety-Related Instrument Channels in Nuclear Power Plants).
Steady state tests are independent of time and use a constant input or a slowly changing input to verify the circuit continuity.
The steady state analog tests on a channel indicate that the channel is present and that most if not all of its components are operating and connected; in other words, the-static condition of the channel. The analog tests do not and cannot indicate the dynamic qualities of an instrument channel, such as its ability to respond j
to a stimulus as a function of time, or how well the channel will behave under the stress of time. A properly engineered response time test will evaluate these qualities.
The GE Topical Report, NEDC-32013P, (Ref. 5) and the EPRI Report, NP-7243, (Ref. 4) were reviewed and evaluated to determine if_ there was an adequate basis to support the elimination of RTT of electrical circuits important to plant safety.
The justification to eliminate the RTT is based upon the concept that the action required and specified by the Technical Specification (TS) is required a comparatively long time after the monitored variable exceeds its trip
4 i i setpoint, and no failures have been detected by RTT. Several examples of this are given. The first example is for Isolation Actuation Instrumentation and the Emergency Core Cooling System (ECCS).
In these cases, the required response is after a number of seconds, i.e.,10 to 13 seconds for Isolation Actuation and 27 to 64 seconds for ECCS actuation. The GE topical report states that since the overall instrument loop will respond in significantly less time, on the order of microseconds or milliseconds, any degradation less than total failure will be in small increases of that response time as compared to the protective function system response time, and therefore, have no significant affect on the required protective function.
In the second example, that of the reactor protection system (RPS) or main steam isolation -
valve (MSIV) actuation, the TS limits are much smaller, on the order of 1 second.
In this case, the entire instrument loop is judged to have a significant enough impact to effect the time response required by the TS, and therefore, only the sensor should be eliminated from the testing requirements.
Again, it is assumed that any degradation less than total failure of the sensor will not affect the overall response time of that protective function ~
L as required by Technical Specifications.
1 IEEE Std 338-1977 (IEEE Standard Criteria for the Periodic Testing of Nuclear Power Generating Station Safety Systems) states in Paragraph 3 of Section 6.3.4:
"RTT of all safety related equipment per se, is not required if, in lieu of RTT, the response time of the safety equipment is verified by
t functional testing, calibration checks or other tests, or both. This is acceptable if it can be demonstrated that changes in response time beyond acceptable limits are accompanied by changes in performance l
characteristics which are detectable during routine periodic tests."
IEEE Std 338-1977 has been endorsed by Regulatory Guide 1.118, Revision 2.
The EPRI report includes both an industry survey and a Failure Mode and l
Effects Analysis (FMEA).
In the FMEA, a number of sensors are analyzed.
These instruments include all sensors which are identified in the GE topical l
report to be exempted from RTT. The FMEA concludes that with the exception of fill-oil loss in Rosemount sensors, there is no credible failure which can result in a response time failure which is not accompanied by a failure in 1
calibration.
The new Maintenance Rule, 10 CFR 50.65, requires that each licensee " monitor the performance or condition of structures, systems, or components, against licensee-established goals, in a manner sufficient to provide reasonable assurance that such structures, systems, and components, as defined in Paragraph (b), are capable of fulfilling their intended functions".
NUMARC 93-01, Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants, dated May 1993, states in Paragraph 10.2.1.3, Performance Trending:
Performance should be trended against established criteria so that adverse trends can be identified. When adverse trends are identified,
i i E
appropriate corrective action should be promptly initiated. The utility's historical data, when combined with industry operating experience, operating logs and records, and station performance monitoring data, can be useful in analyzing trends and failures in equipment performance and making adjustments to the preventive maintenance program.
3.0 EVALUATION l
l The GE report, NEDC-32013P and the EPRI report, NP-7243 both state:
The majority of the allowable instrumentation response times are much longer than instrument circuits require for signal processing from sensor input to final output signal.
The GE report states that changes in circuit response time beyond acceptable.
j limits will affect the normal performance characteristics of the circuit and that these changes will be detected during other types of surveillance testing techniques. The elimination of RTT is being recommended for those channels where the channel response time (e.g., 300 milliseconds) is a small fraction of the TS allowable system response time (30 seconds for example). Although on the surface this appears logical, there appears to be an inherent conflict in this situation. The response times used as the limiting factors during most response time testing are those required by TS. This does not measure l
i s
i i
the designed response time of the instrument, as required to obtain a true indication of the health of the instrument.
If a measure of the channel response time does not adequately measure the response of the instrument, it would be preferable to improve the test rather than eliminate it. Since this
[
was not the recommendation, the argument here appears to be that there is no need to know the response time of the instrument.
l The GE report listed the following as benefits that would be derived by the elimination of RTT:
Improvements in plant safety and increased availability.
Reductions in plant personnel radiation exposure.
Reductions in outage manpower requirements.
According to the GE report, the improvements in plant safety and the increase in system availability would come from the reduction of the time the system is out of service during the testing, and reducing the high potential for plant personnel error during the performance of the response time test including a reduction of the number of half trips. The reductions in plant personnel radiation exposure and the reductions in outage manpower' requirements would be achieved by not having to do the tests, although the performance of the TS required periodic calibration and functional tests may reduce these benefits to some extent.
i J
t 0 However, in discussing the benefits to be derived by the elimination of RTT, the GE report did not address the original purpose of the tests or the benefits originally expected to be gained from the tests.
The GE report and the referenced EPRI report provided valuable insight as to the nature of the RTT activities being performed by the licensees.
Basically, the selected input test signal was so restricted, 5 to 10 percent of instrument setpoint, that only a small segment of the total monitored range was tested and, the documentation of the test and of the failed equipment was so incomplete that no quality data as to the merits of RTT could be obtained.
I The documentation of the failed equipment was entered into a database which contained over 4200 test measurements from over 2100 instruments. An EPRI analysis of the database showed that:
No apparent response time failures were contained in the data.
Some response time measurements were zero or negative, due to slow response time of the reference sensor.
l Response times from the same model of sensor varied from plant to i
plant.
Significant variations occurred between hydraulic and electronic response time test measurements on the same sensor.
-I Significant variations occurred in hydraulic test results, either using a step or ramp input.
l
4-
. Sensor failure trends could not be determined from the response time test data due to limited history of RTT results in the database.
Some contributing factors that may have influenced data accuracy and measurement repeatability:
Changes in the reference transmitter.
Varying distances between the hydraulic tester and the transmitter Transmitter non-linearity.
=
Temperature differences.
Differences in testing techniques among personnel.
Differences in personnel training in time domain testing.
The EPRI analysis of the database went on to describe in greater detail the deficiencies of the database with some of the comments repeated. Where the database showed a failed process pressure transmitter, it was not possible to ~
determine from the information contained in the database whether or not the response time had degraded in the failed transmitter. Response time tests or other types of tests were usually not performed on failed instruments unless root cause analysis was required.
In addition, most response time test input data did not depict the test method used.
Large variations in response time test data on the same instrument were found from one plant to another. The variations in the response time test measurement data reduce the usefulness of the database for predicting response time failure trends. As a result, very
t i I lit *'e if any failure trending information could be concluded from the database. However, the EPRI report states: "No response time failures were found in over 4200 measurements contained in the database" and the GE report concurred in this conclusion.
As shown by the above discussion, the data contained in the database are inconclusive with respect to the merits or usefulness of RTT. The results of i
the response time test activities as described in both reports apparently could not be correlated either intraplant or interplant. Therefore, any conclusion drawn in the two reports that is based on the database and is being used to justify relaxation of RTT is invalid.
The GE topical report quoted the RTT relief paragraph from Regulatory Guide 1.118 as follows:
RTT of all safety-related equipment per se, is not required if, in lieu of RTT, the response time of the safety equipment is verified by functional testing, calibration checks or other tests, or both. This is acceptable if it can be demonstrated that changes in response time beyond acceptable limits are accompanied by changes in performance characteristics which are de W, 4 3 during routine periodic tests.
The GE report intended to demonstrate u.c. other means of testing would verify response time.
6 d
r.
Report NEDC-32013P contains analyses of the failure modes of the components that make up a trip channel.
In the analyses of the failure modes, NEDC-32013P tried to demonstrate that the components that make up a trip channel had no failure mechanisms which would affect the response time of a trip channel except for the log radiation monitor and the variable damping potentiometer. The report concluded that any degradation in response time of the trip circuit components would be negligible when compared to the overall system response time requirements and that any loss of steady state circuit performance would be easily identified by other TS required surveillance i
tests. However, the failure mode and effects analysis " 'd a number of items q
where it was stated that there were "no known means" of such a failure j
occurring. An example of this would be a change in viscosity of fill-oil.
It should be noted that until it occurred, fill-oil-loss in Rosemount transmitters was also not a known failure mechanism. Throughout the analysis, only individual part failures were considered, not interactions between parts, such as friction or binding between moving and static parts. The analysis does not appear to be all-inc' and, therefore, is not a valid argument for elimination of RTT.
In addition, any reliance on calibration as a means of detecting response time failures is suspect. There has not been any prcposal to include RTT into the currently conducted calibration, nor is a method proposed for sensitizing i
technicians performing calibration to response time problems.
In addition, j
there have been a number of proposals and topicals which have received NRC approval for extension of calibration intervals, as well as proposals to.
I
L f replace calibrations of some ins;truments with cross channel comparison tests.
None of these proposals or topicals has taken response time requirements into account, nor has the GE Report NEDC-32013P taken into account the possibility of decreased calibration frequency or eliminated calibrations.
The new Maintenance Rule,10 CFR 50.65, and the NURARC interpretation of that rule now imply that drift or calibration trending of safety related instrumentation is a requirement. The GE topical report did not address a program of this type.
Drift trending may be considered an alternate method of verifying correct performance of an instrument channel, although it does not verify the response time of the safety equipment. A combination of trending -
and adequate calibration can form a basis for relief from RTT.
)
i i
4.0 CONCLUSION
AND RECOMMENDATIONS The discussion on RTT in General Electric Topical Report NEDC-32013P and the referenced EPRI Report NP-7243 states that RTT, as now conducted by the licensees, is not a very useful activity and that the tests do not accomplish their intended purpose of detecting actual or impending safety system failures. However, these reports did not define the purpose of the tests, why the tests are being performed, what is learned from the tests, or how the information is being used. The report did not refer to better or alternate methods of testing as a substitute for current RTT. No mention of the requirements of the Maintenance Rule,10 CFR 50.65, was made, nor was the
e t
1 method of compliance as applied to safety related instrumentation shown.
It is, therefore, concluded that RTT continues to be considered a necessary activity that must be performed on electrical circuits important to plant safety in accordance with existing Technical Specifications. Poor test results which are the result of an improperly designed test are not a proper justification for the elimination of important system tests.
1 l
)
Based upon the staff review of the GE report and referenced EPRI report, the staff does not approve General Electric Nuclear Energy Report NEDC-32013P.
Any resubmission of the GE topical report requesting elimination of RTT should address the following topics:
1.
The purpose and goals of response time testing.
2.
What information is gained from the tests?
3.
How is the information being used?
4.
If the tests are eliminated, what tests would replace them and still provide the above information?
1 5.
How would the information of Item 4 be used?
6.
What would be the frequency of the replacement tests?
t i
7.
How will application of 10 CFR 50.65 affect safety related instrumentation maintenance and calibration methodology?
8.
An all inclusive FMEA, including not only individual part failures
~
but also interactions between parts, such as friction or binding between moving and static parts.
l I
.u-f
, It should be noted that even with a resubmittal covering the eight items listed above, special justification should be provided for elimination of P
testing for channel response times of less than 30 seconds and for sensors used in automatic actuation of RPS and ESF systems.
Prir:cipal Contributor: Paul Loesser s
i l
l l
I 1
I l
9 e
g
5.0 REFERENCES
1.
IEEE Std. 388-1977, " Criteria for the Periodic Testing of Nuclear Power Ge.,erating Station Safety Systems."
i 2.
Reg Guide 1.118, Rev 2, 1978, " Periodic Testing of Electric Pcwer and Protection Systems," Revision 2, June 1978.
i 3.
ISA-567.06-1984, " Response Time Testing of Nuclear Safety-Related Instrument Channels in Nuclear Power Plants," dated i
August 29, 1986.
4.
Electric Power Research Institute Report NP-72433, " Investigation of Response Time Testing Requirements," Final Report May 1991.
.)
5.
General Electric Nuclear Energy Report NEDC-32013P, " System Analyses for Elimination of Selected Response Time Testing Requirements," Herc'.1992.
6.
Nuclear Management and Resources Council 93-01, Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants, dated May 1993.
,