ML18038A465
| ML18038A465 | |
| Person / Time | |
|---|---|
| Site: | Nine Mile Point  |
| Issue date: | 05/08/1984 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML18038A464 | List: |
| References | |
| NUDOCS 8905120245 | |
| Download: ML18038A465 (22) | |
Text
'~a ai t
UNITED STATES I'JUCLEAR REGULATORY COM SION WASHINGTON. O. c, 20555 EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGL'LATION MAIN STEAM ISOLATION VALVE LIMIT SWITCH TESTING NIAGARA MOHAWK POWER CORPORATION NINE MILE POINT NUCLEAR STATION, UNIT NO. I OOCKET NO. 50-2?0 INTRODUCTION The Ni ne Ni le Pointi Uni t 1
(NNP-1) Techni ca l Speci I' cat i ons (Table 4.6.2a(6) ) require that an "Instrument Channel Test" be performed quarterly to verify operabi Lity of the main steam isolation valve (NSIV) position instrument channels used to initiate a reactor scram on NSIV closure.
To cause a reac" tor scram't Least one NSIV (either inboard or outboard) in both main steam Lines must close to the 10X closed position as sensed by Limit switches mounted at the NSIVs.
The Limit switch contacts in the reactor protection system (RPS)
Logic circuits open to de-energize the associated channel trip re-Lays (11K16 and 11K17) causing a scram.
During a
review of Licensee surveillance
- testing, the Resident Inspector noticed that the Licensee was simulating NSIV clo" sure by pulling fuses in series with the Limit switch contacts.
The contacts are closed when the NSIVs are greater than 90K opens maintaining the channel trip re Lays in an energ di tion.
Performance of the instrument channel test i
ized con-n this manner (i.e., pulling fuses) does not verify operability of the Limit switches relied on to initiate a reactor scram under actua l conditions The NRC regionaL office (Region
- 1) be-Lieves that current 8905I20245 890504 PDR ADOCK 05000220 P 'DC testing of the NS IV position instrument
'l
>r
channels may be deficient in this regard.
Furthermore, Region 1 believ'es that the technical specification definitions of.
"Instrument Channel" and "Instrument Channel Test" indicate that the testing should include provisions to verify operabil" i ty of the sensors (i.e.~
the NSIV Limit switches).
A review of other operating BWRs in Region 1 indicates that operabi l-ity of the Limit switches is verifie tially closing the NSIVs until a
cha d during testing by par-nnel trip occurs.
The Licensee contends that it was never intended to verify oper" abi lity of the Limit switches as evidenced by the Lack of a
Technical Specification requirement for a "Sensor Check".
EVALUATION The requirements concerning testability of nuclear power plant protection systems are set forth in 10 CFR 50, Appendix Ap General Design Criterion (GDC) 21 "Protection System Reliabil-ity and Testability" and IEEE Standard 279(Criteria for Pro-tection Systems for Nuclear Power Generating Stations)
Section 4.9 "Capability for Sensor Checks".
GDC 21 states, in part/I that the protection system shaLL be designed for high func-tional reliability and inservice testability commensurate with the safety functions to be performed.
GDC 21 further states that the protection system shaLL be designed to permit peri-odic testing of its functioning when the reactor is in opera" tion including a capability to test channels independently to
e
determine failures and Losses of redundancy that may have occur-red.
Section 4.9 of IEGER Std'79 states that means shall be provided for checking~
with a high degree of confidence~
the operational availability of each
( rotection) s stem in ut sen-sor during reactor operation.
Therefore, verification of sen-sor (in this case, NSIV Limit switch) operability during reac-tor operation is required.
The NIP-1 Technical Specifications provide the following defini" tions fo~ "Instrument Channel" and "Instrument Channel Test":
Instrument Channel An instrument channel means an arrangement of a sensor and auxiLiary equipment required to generate and transmit to a trip system a single trip signal related. to the plant parameter monitored by that instrument channel.
Instrument Channel Test Instrument channel test means injection of a simulated signal into the channel to verify its proper response including~ where applicable~
alarm and/or trip initia" ting action.
These definitions do not distinguish between analog channels (here the input to the protection system is typically a
4 to 20 mA variable signal from a process sensor/transmitter) and digital/bistable channeLs (here the protection system input is a fixed signal that is either presenb or absent depending
I I
C
on the state of the sensor; e.g.~ limit switch contacts).
Furthermore~
the point of injection of the simulated signal into the instrument channel during testing is not specified.
Typically~ for analog channels~
the test signal is injected as close to the sensor as practicable>
thus testing the por-tion of the instrument channeL downstream of the sensor.
Operability of the sensor is then verified by cross compar-ison of its output (i.e.~ control room indication/readout) with other instrument channels monitoring the same parameter.
This cross comparison of sensor readouts is known as a chan-nel check or sensor check.
A sensor whose output varies from the outputs of the remaining sensors is suspected of being f ai led.
- Thus, the instrument channeL test (also referred to
~ as a channel functional test) in conjunction with the channeL check provide complete end-to-end testing of the instrument channeL.
The General Electric Standard Technical Specifica-tions (GESTS; NUREG-0123) define a channel functional test for analog channels and a channel check as follows:
Channel Functional Test Analog channeLs - the injection of a simulated signal into the channel as close to the sensor as practicable to verify operability including alarm and/or trip functions and channel failure trips.
I I
VI
Channel Check A channeL check shaLL be the qualitative assessment of channeL behavior during operation by observation.
This determination shall include~
where possible~
comparison of the channel indication and/or status with other indi-cations and/or status derived from independent instru-ment channels measur ing the same parameter.
The NSIV limit switches provide binary (also referred to as bistable or digital) outputs to the RPS (i.e.,
the Limit switch contacts are in one of two possible states corresponding to the two possible NSIV positions; open or closed)
~
The Limit s~itches do not monitor a "Live" parameter (i.e.~ the process is not being continuously received/monitored at the sensor input)~ or provide continuous indication in the control room.
The exact positions of the HSIVs are unknown.
A channel check is not applicable to sensors providing binary/bistable outputs.
These sensors are often referred to as blind sensors.
There-
- fore, the Licensee's argument that it was not intended to re-quire testing of the limit switches as indicated by the Lack of a requirement to perform a sensor check is not valid.
The GESTS define a channel functional test for bistable chan-nels as follows:
Channel Functional Test Bistable channels - the injection of a simulated signal into the sensor to verify operability in-cluding alarm and/or trip functions.
I 0
For bistable
- channels, such as the NSIV Limit switches, the simulated test signal must be injected into the sensor.
This is done to verify operability of the sensor itself as well as the portion of the instrument channel downstream of the sen" sor.
The ideal method of performing a channel functional test for the NSIV position channels is by partial closure of the NSIV past the 10K closed scram Limit swi tch setpoint.
This method of testing duplicates sensor operation under actual conditions
~
Performance of the NSIV position instrument channel test by pulling fuses does not satisfy the requirements of GDC 21 or IEEE Std. 279 regarding verification of'ensor operabi lity during reactor operation.
However~
due to the vagueness of the NNP-1 TechnicaL Specifications+ it cannot be concluded that testing of the Limit s~itch contacts is required and therefore~
that testing was not being conducted by the Li" I
censee in accordance with the Technical Specifications.
The Technical Specifications and associated test procedures should be revised to include provisions for periodic testing of the Limit s~itch contacts.
The inboard NSIVs at HNP-1 are motor operated.
The closure times for these valves are six seconds.
The Licensee contends
I I
that testing the Limit switches by partial closure of the NSIVs during operation could result in an unwanted reactor scram.
The Licensee believes that given the fast closure times of the NSIVs, i t ~ould be di fficult for the control room operators to manually reverse the direction of valve travel following partial closure before the valve closed to the point where a reactor scram occur" red on high steam Line f Low in the adjacent steam Line.
The NSIV motor operator s
do not have the capability for slow opera-tion.
The inboard NSIVs (as well as the outboard NSIVs) at other BWRs are air operated (air pressure is required to open the valves'nd vented to aLLow'pring force to close the valves) and have the capabi Lity for slow closure (the air pres-sure is slowly vented) to allow for testing of the NS/V Limit switches.
In addition~ other BWRs have four steam lines as opposed to two at NIP-1.
Therefore~
the steam f Low rejected f rom a
steam Line due to partiaL NSIV closure during testing must be accommodated by the single remaining steam Line.
This makes the NIP-1 design more susceptible to high steam Line flow reactor scrams during testing.
The outboard NSIVs at NNP-1 are air operated+
but do not have the slow test capabil-ity.
Staff guidance regarding testing of protect%on system circuits during reactor operation is provided in Regulatory Guide 1.22
0 I
a
a 8
a (RG 1 '2) "Periodic Testing of Electric Power and Protection Systems".
Position D4 of RG 1.22 allows exceptions to the requirements for testing of protection system equipment during reactor operation if 1) there is no practicable system design that would permit operation of the actuated equipment without adversely affecting the safety or operability of the plant; 2)
The probability that the protection system will fail to initiate the operation of the actuated equipment is~
and can be maintainedi acceptably Low without testing the actuated equipment during reactor operation) and 3) the actuated equip-ment can be routinely tested when the reactor is shutdown.
According to the Licensee~
in order to provide test-capabil-ity for the NSIV Limit switches at power without disrupting plant operation would require valve control circuit modi fica-tions providing the capability for sLow closure.
The staff believes that although such modi fications would probably not cause an unnecessary burden to the Licensee~
that the bene" fit gained from the modifications would not contribute signi-ficantlyy to the safety of the plant.
This assessment is based on the following: 1)
The NNP-1 Chapter 15 transient analyses takes credit for a reactor scram initiateg by the NSIV Limit switches to help terminate a
HSIV closure transient.
- However, if the Limit switches should faiL~ two other independent and
I
diverse scram functions (reactor high pressure and high neutron f Lux) are available to terminate the transient.
Peri-odic testing of the instrument channels monitoring the diverse parameters is performed during reactor operation as required by the NNP"1 Technical Specifications
~
- 2) Although not presently required by the NIP-1 Technical Spec ifications~
the Licensee has commi tted to calibrate the NSIV Limit switches, to ensure that a
scram occurs at the 10X closed setpoint, during each refueling outage.
The NNP-1 Technical Specifications wiLL be revised to include provisions for calibration at refueling in-tervals.
The staff's position~
however~
is that verification of NSIV Limit switch op'erabi lity at a frequency of once per refueling outage is not sufficient to satisfy the requirements of GDC 21 and IEEE Standard 279.
Operating experience has sho~n that reactors are typically shutdown at Least several times between refue ling outages.
Therefore, the staff recommends that the Licensee include in their technical specifications a require-ment to verify operability of the NSIV L'imit switches prior to startup follo~ing each plant shutdown by actual valve closure past the 10K actuation setpoint~
unless the test has been performed within the previous 92 days'esting in this manner wiLL prevent the operational problems encountered when per" forming the test at power.
In addition~ testing prior to startup by actuaL valve closure wi LL eliminate the need for pulling fuses during testing which is discouraged by the
J I
. r
10 staf f.
I t is reco>nized that pul Ling fuses in the fail safe RPS instrument channel circuits results in a channel trip.
The staff is concerned~
however~
regarding the potential for the practice of pulling fuses to be carried over to energize-to actuate safety related circuits during testing~
which is cLearly unacceptable.
The Licensee has indicated that with the exception of NSIV pos< tion~ fuses are not pulled during RPS instrument channe L testing.
CONCLUSION The Lack of a Technical Specification requirement to perform a sensor check does not relieve the Licensee from periodicaLly testing the operability of the NSIV Limit switch contacts used to initiate a reactor scram.
However~
due to the operational problems encountered when testing the NSIV position instrument channels (including verification of Limit switch operability) at power~ the staff has determined that strict adherence to a
fixed test frequency is not necessary.
The staff recommends that the Licensee incorporate a requirement in their technical specifications to test the NSIV position instrument channels prior to star tup following plant shutdowns by actual closure of the NSIVs~ unless the test has been performed within the previous 92 days.
In addition~ it is recommended that the Technical Specifications be revised to distinguish between analog and bistable channel functional tests consistent with the GUSTS.
I t
\\
\\
May 4, 1989 THRU:
Robert A. Capra, Director Project Directorate I-l Division of Reactor Projects I/II Office of Nuclear Reactor Regulation Docket No. 50-220 MEMORANDUM FOR:
John A. Skoczlas, Jr., Chief Document Control Branch Office of Administration and Resources Management DISTRIBUTION oc et e
PDI-I Rdg CVogan MSlos son RBenedict RCapra FROM:
SUBJECT:
Marylee M. Slosson, Project Manager Project Directorate I-I Division of Reactor Projects I/II Office of Nuclear Reactor Regulation MAY 8, 1984 MEMORANDUM FOR R.
STAROSTECKI FROM D. EISENHUT,
SUBJECT:
NINE MILE POINT 1 - EVALUATION OF TECHNICAL SPECIFICATION RE(UIREMENTS FOR MAIN STEAM ISOLATION VALVE LIMIT SWITCH TESTING - ACCESSION NUMBER 8405180131 The subject document
{copy enclosed),
an internal NRC memorandum, has been referenced by the Nine Mile Point 1 licensee in a licensing action, and it should be made available to the general public.
The licensee already has a copy of it, and we have determined that it need not be withheld.
Please provide appropriate distribution to the PDR/LPDR.
Enclosure:
As stated
)MEMO SKOCZLAS" ORIGINAL SIGNED BY Marylee M. Slosson, Project Manager Project Directorate I-I Division of Reactor Projects I/II Office of Nuclear Reactor Regulation W&
NAME :CVogan WWWWWWW DATE: 0/9/89 W
W WW
~
- RBenedict:vr:MSlosso W
- -"/: /
89
~
~
- RCapra
~ 8/~/89 OFFICIAL RECORD COPY
t