IR 05000346/1979006
| ML19259C516 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 04/20/1979 |
| From: | Norelius C, Streeter J, Vohler J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML19259C517 | List: |
| References | |
| 50-346-79-06, 50-346-79-6, NUDOCS 7906230117 | |
| Download: ML19259C516 (26) | |
Text
.
.
<
.
U.
S. NUCLEAR REGULATORY COMMISSION OFFICE OF INSPECTION AND ENFORCEMENT
REGION III
Report No.
50-346/79-06 Docket No.
50-346 License No.
NPR-3 Licensee: Toledo Edison Company Edison Plaza 300 Madison Avenue Toledo, OH 43652 Facility Name :
Davis-Besse Nuclear Power Station, Unit 1 Investigation At:
Davis-Besse Site, Oak Harbor, OH Babcock & Wilcox Office, Lynchburg, VA Investigation Conducted: January 6, 14, 29, 1979 l -i ] ]c k..
Inves t i ga to r :hlJ.
4 [t8 /7 7 E. Foster
. c.:
- o..,,., e y f, y/
- ,' /
In sp ec to r :
,,,J.
E. Kohler l-d N-Il 6 ta ks
,
Reviewed By:
C. E. Norelius V !' d !?'I As e + 1,n t to the Director
.= %d j) A$ 2)I
!
a.
F. Streeter, Chief
Nuclear Support Section 1 Investigation Summary Investigation on January 6, 14 and 29, 1979 (Report No. 50-346/79-06)
Areas investigatec: Inspector concerns regarding loss of pressurizer level indication and violation of technical specification table 3.3-4 involving setpoint settings to 90% undervoltage relays.
The investigation involved 72 inspector-hours onsite by two NRC personnel.
Results:
No items of noncompliance were identified.
2283 298 7 90623 0 ll]?'
.
.
INTRODUCTION The Davis-Besse Unit 1 Nuclear Power Plant, licensed to the Toledo Edison Company, is an operating plant located near Oak Harbor, Ohio.
The facility has been in operation for approximately two years, and utilizes a Pressurized Water Reactor designed by the Babcock 6 Wilcox Company (B&W).
Bechtel Power Corporation was the Architect-Engineering Firm for the plant.
REASON FOR INVESTIGATION Subsequent to his review of first cycle power operations at the Davis-Besse Nuclear Power Plant, an NRC Region III (RIII) inspector indicated concerns relative to operation of the facility. The inspector indicated that he felt two areas of concern warranted an NRC investigation, as they appeared to reflect a lack of timely analysis of a possible safety p robl em and possible unsafe operation of the f acility. The first
,
area of concern was related to a loss of pressurizer level indication (LOPLI) which occurred during a loss of offsite power transient. The second area of concern was that the plant had operated with its 90:
undervoltage relay setpoints below those specified in the Plant Technical Specifications, which could have resulted in delayed emergency system response.
Discussions with the inspector developed specific questions within each broad area of concern.
An investigation was initiated into these concerns.
SUMMARY OF FACTS on De cembe r 19, 1978, the RIII inspector advised the Chief, Reactor Operations and Nuclear Support Branch RIII, of concerns related to LOPLI and pressurizer voiding.
On February 29, 1979, discussions with the inspector indicated that he was also concerned that the licensee had delayed implementation of an undervoltage relay setpoint change, possibly delaying Emergency Core Cooling System (ECCS) response had it been needed prior to the change.
Within the area of LOPLI, the inspector indicated that he questioned whether a timely analysis of the phenomenon had been performea, and if it was a generic occurrence at Babcock 6 Wilcox (B6W) reactor facilities.
He also questioned whether feedwater design was unique, and whether 2283 299
- 2-
.
.
LOPLI indicated a violation of General Design Criteria 13 (instrumen-tation and control).
With the area of the undervoltage relay setpoint, the inspector questioned whether the setpoint delay was deliberate, and if ECCS would have been delayed prior to the setpoint change. He also questioned the relationship of the setpoint implementation delay to a Safety Features Actuation System (SFAS) test.
RIII personnel visited the Davis-Besse site on two occasions, and reviewe'
records related to the undervoltage relay setpoint change. Correspondence between the licensee and NRR related to the relay was also reviewed.
A visit was mal' to the B5W office in Lynchburg, Virginia where discussions were held with E&W and representatives of several utilities operating B6W plants.
Documents related to analysis of LOPLI were also reviewed during these visits.
It was found that LOPLI has occurred at several B&W plants, with an analysis of the phenomenon being perforned by B&W in 1975. A letter indicated that B&W had advised Davis-Besse of the possibility of LOPLI prior to operation.
B&W personnel advised that their analysis of LOPLI indicated core coverage would be maintained during analyzed reactor transients, ECCS would initiate when pressure dropped to the ECCS setpoint, and f unction properly.
Discussions with Toledo Edison (TECo) representatives, and documents reviewed indicated that TECo discussed LOPLI with B&W shortly after LOPLI was exper-ienced at the Davis-Besse facility, and had been pursuing measures to limit LOPLI wi th the vendor.
Discussions with B&W personnel indicated the feedwater system for the Davis-Besse facility differs from that of other E6Y plants.
The auxiliary feedwater pump system for Davis-Besse consists of two 100% pumps, each feeding a steam generator (a connecting pipe and valve is designed to open only should one pump fail). At other B&W plants, B&W personnel advised that the auxiliary feedwater system consists of three pumps of approxi-mately 802. capacity, which feed a common header to the steam generators.
B&W personnel stated that the LOPLI was an operational problem, not a safety problem, since the pressurizer cannot fully empty without a drop in the main reactor coolant system pressure. They advised that such a pressure drop would cause the High Pressure Injection system (HPI) to function, maintaining core coverage.
A recent submittal to the Of fice of Nuclear Reactor Regulation (NRR) by Toledo Edison, (December 22, 1978)
contains their analysis, and indicates that if the pressurizer should void, small voids would form and flow through the RCS system without impairing cooling flow.
Discussion with NRR on March 13, 1979, indicated there are no open items relative to LOPLI.
2283 300-3-
.
.
With regard to the unde rvoltage relay setpoint, it was found that the licensee had requested a change to the facility Technical Specifications to revise the setpoints of the 90% undervoltage relays as final resolu-tion to NRR's questions on undervoltage protection. The NRC granted this change, but the relays were not reset for approximately 7 months, apparently due to oversight and failure to control Facility Change Requests on the part of the licensee.
A review of the chronology of events related to the relay setpoint change revealed that the relays had been set to the figure required by the licensee's analysis of undervoltage conditions. However, this setpoint did not in-clude margin for setpoint drift, a consideration which was included in the Technical Specification change request. The result of this condition was that theoretically, for one sequence of events, ECCS might have been de-layed one-half second.
The licensee had reported to the Commission the discovery that the relays had not been reset, found during a review of an SFAS test.
The relays were correctly set shortly after this discovery.
CONCLUSIONS 1.
Toledo Edison performed a proper review of LOPLI following its occurrence at the Davis-Besse plant.
2.
B&W personnel advised that LOPLI had been analyzed, and was not a safety problem.
3.
NRR has reviewed the B&W analysis and concurred with its conclusions.
4.
The licensee did not implement a revised setpoint for the 90'4 under-voltage relays for approximately 7 months, after receiving a Technical Specification change they had requested.
5.
No evidence could be developed to indicate the licensee knowingly delayed implementation of the revised relay setpoint, and the licensee reported the discovery that the relays had not been reset.
6.
A review of the events surrounding the undervoltage relay indicates that the relays were set in accordance with the licensee's analysis of undervoltage conditions, but did not include margin for setpoint drift.
2283 301-4-
.
7.
Delay of implementation of the relay setpoint indicated lack of control of facility changes.
8.
No items of noncompliance are cited within this report.
2283 302
- 5-
.
.
DETAILS 1.
Personnel Contacted Toledo Edison Company T. Murray, Station Superintendent W.
Green, Assistan t to Station Superintendent
- C.
Domeck, Nuclear Project Manager, DB-1
- F.
Miller, Plant Nuclear System Engineer
- E.
Novak, Superintendent Power Engineering and Const ruction
- S.
Jane, System Engineer
- C. Calcameggio L. Stalter, Technical Engineer T. Beele r, Assistant Engineer J. Lingeaf elter, Nuclear Engineer Babcock & Wilcox E.
R. Cane R. C. Luken S. H. Klein F. R.
Faist B. M.
Dunn J. T. Willse L. R. Cartin Metropolitan Edison Company J.
F. Hilbish, Superintendent, Licensing Sacramento Municipal Utility District R. A.
Dietrich, Senior Nuclear Engineer Arkansas Power & Light M. O. White J. T. Enos 2283 303
"" "*"' " '"
D. G. Anderson, Inspector
- 6-
.
.
2.
Scope This investigation focused on two areas of concern indicated by an RlII inspector; loss of pressurizer level indication (LOPLI), and noncompliance with a Technical Specification regarding undervoltage relay setpoints. Emphasis was placed on the timeliness of the licensee's actions (evaluation of LOPLI, implementation of facility changes) rather than on technical aspects of the events. The in-vestigators did not make technical evaluations of the infor;aation supplied by TECo or B&W as this had been accomplished by NRR.
3.
Loss of Pressurizer Level Indication A.
Conce rn :
Is LOPLI a generic occurrence at B&W facilities?
Findings Loss of pressurizer level indication during certain transients has occurred at the following B&W plants: Arkansas Nuclear 1 Three Mile Island 2, Rancho-Seco, and Davis-Besse 1.
De t ails RIII investigators met with representatives of B&W at the Lynchburg, Virginia facility on February 14, 1979 to discuss LOPLI at B&W Plants.
Present at this meeting were representa-tives of four similar B&W facilities as follows:
Metropolitan Edison Company (Three Mile Island 2), Sacramento Municipal Utility District (Rancho-Seco), Arkansas Power and Light Company (Arkansas Nuclear 1), and Toledo Edison Company (Davis-Besse 1).
Each utility made a brief presentation regarding LOPLI. Each indicated they had experienced LOPLI periodically early in plant life following certain reactor transients. The LOPLI at these facilities was attributed to less than ideal main steam safety settings and feedwater flowrate.
Documents provided indicated that B&W informed Davis-Besse 1 in 1976, during final construction, that main steam safety relief valve blowdown settings should be modified to reduce occurrence of LOPLI.
However, modification was not pe r forme d until af ter power operation, early in 1978 (Exhibit 3).
2283 304
.
- 7-
.
Discussion indicated that, in the case of Davis-Besse 1, LOPLI was found to be attributable to main steam safety relief settings and the following compounding conditions:
1.
Overcapacity of auxiliary feedwater pumps.
2.
Makeup pumps tripping on loss of of fsite power.
3.
320-inch pressurizer level indication span.
The investigators were advised that changes had been init iat e d to modify the auxiliary feedwater system which addresses items 1 and 2 above. However, at the time of this investi-gation, these facility changes had not been implemented; however, Davis-Besse personnel advised that they were operating the auxiliary feedwater system with administrative controls to reduce the probability of LOPLI.
B.
Conce rn :
Is there a possibility that voiding of the pressur-izer during a transient could occur which would adversely affect emergency core cooling system (ECCS) injection?
Findings The inspectors reviewed memoranda f rom Toledo Edison Company (TECo), the Nuclear Regulatory Commission, Arkansas Power and Light, and Babcock & Wilcox.
The conclusion drawn in these memoranda is that LOPLI was not a safety question.
B&W's analysis concluded that if the pressurizer should void, high pressure ECCS injection would automatically be initiated maintaining core coverage. The vendor concluded that ECCS injection would not be affected.
De ta il s RIII personnel had previously requested TECo to determine what the actual level was in the pressurizer during the LOPLI which occurred at Davis-Besse on November 29, 1977. The licensee was also requested to determine whether an unreviewed safety question existed.
Two analyses were provided RIlI: with makeup flow, and without makeup flow.
These analyses, provided in September and October 1978, indicated that the pressurizer did not empty during the November 29, 1977 transient.
Furt he rmore,
the conclusion drawn by Toledo Edison was that if the pres-surizer were to empty, high pressure injection would actuate, because the reactor coolant system (RCS) woul( have depre-ssurized to the 1600 pounds per square inch h.gh-pressure safety injection 2283 305
_s_
.
setpoiat. Therefore, core coverage would be maintained.
The licensee concluded, with NRR concurrence during December 1978, that LOPLI did not represent an unreviewed safety question, because HPI would be available to maintain core coverage should RCS depressurize to 1600 pounds per sq. in, and ECCS injection would not be affected. As of March 13, 1979, NRK advised RIII that they had no more questions regarding LOPLI at Davis-Besse Unit 1.
C.
Concern: Were the design requirements for the Davis-Besse Unit 1 auxiliary feedwater system unique, resulting in a unique design of two 100% steam driven auxiliary feedwater pumps?
Findings The Davis-Besse Unit 1 auxiliary feedwater system was apparently not designed against any unique NRC criteria. A decision was made to supply two pumps as opposed to three pumps at other B&W facilities.
During plant review, this design was found to be acceptable.
De ta ils The licensee review of the November 29, 1977, transient resulted in a determination that one of the causes of LOPLI was an exces-sive auxiliary feed pump flow rate.
The auxiliary feedwater system had been programmed upon actuation to maintain 120-inch levels in the two steam generators automatically after certain reactor trips.
The investigators discussed with utility representatives at the B&W meeting of February 14, 1979, design requirements for the auxiliary feedwater system. These discussions indicated that the feedwater system at Davis-Besse Unit 1, designed to meet the limiting case specified by B&W (small break), was different than other plants represented. They advised that most B&W plants have three auxiliary feed pumps with approximately 80% capacity each. Davis-Besse Unit 1 has two 100% steam driven pumps with each pump feeding one steam generator. The decision to design two large capacity pumps as opposed to three lesser capacity pumps was apparently not based on any unique B&W or NRC requirements, but was an engineering decision by the architect-engineer.
2283 306
- 9-
.
The present design of two 1007 capacity pumps was found acceptable by NRR in the Davis-Besse Unit 1 FSAR issued Decembe r 1977, with one exception.
The auxiliary feed pump control valves are to be modified f rom AC power to DC power on one t rain during the first ref ueling outage, in order to meet diversity of power requirements as specified by NRR.
Through discussion it was found that the licensee has investi-gated several different modifications to the auxiliary feed-water system in order to bet ter cont rol the flow rate.
Possible modi ficat ions involved:
1.
Orificing.
2.
Installing a flow control valve.
3.
!bdification of the auxiliary feed pump level controller by installation of a dual setpoint (35 inches /120 inches)
TECo personnel advised the decision had been to enginear a dual setpoint controller which would automatically control at the lower level of 35 inches unless the steam and feedwater rupture control system activated and was followed by a safety f eat ures ac t ua t ion.
Then, the level of the steam generator would be automatically controlled at 120 inches.
In the interim, unt il this f acility change is implemented, administ rative controls have been developed to control steam generator levels.
NRE has reviewed and approved the licensee's correc t ive ac t ion involving a dual setpoint.
In addition, NRR has reviewed and approved the interim administrative controls on manual operation of the auxiliary feed pumps and has found that there is reasonable assurance that operators have enough time to respond adequately to a transient.
D.
Concern: Was the licensee's analysis of the LOPLI experienced on November 29, 1977 performed in a timely fashion? Was LOPLI considered as a possible unreviewed safety question?
Findings Toledo Edison's review of loss of pressurizer level was timely, based on the following facts:
1.
Loss of pressurizer level indication was previously analyzed at Arkansas Nuclear 1 (ANO) in 1975 and it was concluded to not represent an unreviewed safety question by B&W and ANO (Exhibit 5).
2.
Toledo Edison was in contact with B6W soon af ter the event and determined that a setpoint change to the auxiliary feed pump level controller would enhance auxiliary feedwater system performance (Exhibit 4).
2283 ;07 10 _
.
-
Details During the meeting with B6W and other utility representa-tives, LOPLI was discussed.
Arkansas Nuclear 1 provided documentation of the analysis performed by B6W on the LOPLI which occurred during the 1975 startup of ANO 1 (Exhibit 5).
This analysis stated that core coverage would be maintained because high pressure injection actuation would occur well before pressurizer voiding.
In addition, the representative from Metropolitan Edison described a transient which occurred on the Three-Mile Island Unit 2 station caused by excessive feedwater flow and main steam safety valve lifting.
This event caused rapid cooldown of the primary system and pressurizer level indication was lost briefly.
However, as the RCS pressure dropped to the 1600 pound per square inch setpoint for high pressure injection, the high pressure inj ection system actuated and pressurizer level indication was restored.
This event substantiated the 1975 ANO 1 analysis, and appears applicable to the Davis-Besse Unit 1 station.
Statements and documents indicate Toledo Edison was in contact with B6W after the November 29, 1977 transient to discuss LOPLI.
Since B6W had concluded in 1975 that LOPLI did not represent an unreviewed safety question, they indicated that no new analysis of LOPLI was performed.
In February 1978, B6W and Toledo Edison were exploring ways of enhancing pressure level indication and had decided on a dual level setpoint (Exhibit 4).
The investigators found the following:
The safety-related analysis regarding possible voiding of the pressurizer during certain transients was first reviewed by B&W in 1975 for Arkansas Nuclear 1 and the conclusion was that no unreviewed safety question existed; that the safety-related analysis regarding possible voiding of the pressurizer performed at ANO 1 in 1975 was applicable to Davis-Besse Unit 1 (as stated in vendor meeting of February 1979); that the licensee began investigating corrective action regarding auxiliary feedwater system as early as February 1978.
The Toledo Edison review of the November 29, 1977 tran-sient which resulted in LOPLI was not untimely based on the following considerations:
1.
No new information was provided by Toledo Edison regarding the safety significance of worst case voiding of the pressurizer during certain transients.
2.
Exhibit 9 shows that the licensee began corrective action discussions with B6W soon after the event.
2283 308
- 11 -
.
.
3.
NRR concluded that no unreviewed safety problem existed at Davis-Besse Unit 1.
E.
Concern:
Did the loss of pressurizer level indication at Davis-Besse Unit I represent a violation of General Design Criteria (GDC) 13?
Findings Pressurizer level indication loss at Davis-Besse Unit I was similar to the event which occurred at Arkansas Nuclear 1 in 1975. Arkansas Nuclear I concluded that CDC 13 was not violated because RCS pressure indication was never lost, and there was a correlation between pressurizer level indication and RCS pressure.
Details During the meeting held at B&W, General Design Criteria 13 and loss of pressure level indication were discussed.
Representatives from Arkansas Nuclear 1 ststed that although pressurizer level indication had been lost during certain loss of offsite power transients, reactor coolant pressure indication was never lost.
ANO stated that the correlation involved to relate pressurizer level to RCS pressure is sufficient te meet GDC 13 requirements.
This argument appears applicable to Davis-Besse Unit i The investigators were advised by B&W personnel ths. the following additional measures had been taken by B6U to improve the performance of 9tessurizer level indication:
1.
Repositioning of the pressurizer level taps w11ch presently spans 320 inches to incorporate a larger span (new plants).
2.
Recommendations to licensees regarding main steam safety relief blowdown settings.
3.
Recommendations to licensees regarding auxil..ary feedpump flow.
F.
Small Break Analysis Documents reviewed indicated that during the licensee's review of LOPLI, it was discovered general operating procedures and emergency procedures governing manual operation of the auxiliary feedwater system violated certain B&W assumptions used in the FSAR analysis of the 2283 309
- 12 -
.
.
small break. Davis-Besse personnel advised they had corrected these procedures.
The item was licensee identified and reported in LER 50-346/78-117.
4.
90% Undervoltage Relav 5etpoint Concerns A.
Concern: Technical Specification Table 3.3-4 was amended effective November 29, 1977, and changed the setpoint associated with the 90% undervoltage relay; however, the facility change was not performed until June 15, 1978.
The plant was in noncompliance with Technical Specification Table 3.3-4 for a period of 7 months.
Was implementation of the setpoint change deliberately delayed?
Finding The licensee operated in noncompliance with the noted Technical Specification for 7 months.
The licensee reported this to the NRC when discovered.
No evidence was developed to show that the setpoint change delay was deliberate.
Details The investigators found that the 90% undervoltage relay and its associated setpoints had been the subject of much correspondence over a one year period spanning October 1976 to November 1977 between Toledo Ediscn and NRR.
Review of the correspondence indicates that the reley in question was installed in the plant in response to an NRR generic letter on degraded grid voltage (Exhibit 6).
Al NRR had not completed its review of Toledo Edison's 90% relay submittal at the time the safety evaluation report was issued, (December 1976), the plant was licensed with an offsite power grid stability requirement.
At the time of cn IE verification inspection, Technical Specification Table 3.3-4 called for a 10 +
1.5 second delay setting for the undervoltage relays.
The setpoint safety analysis for the 90% undervoltage relay called fer a 9 second setpoint.
Site records indicated the 9 second setpoint was installed at that time and was in compliance with the existing Technical Specifications.
(See Exhibit 9).
In November 1977, NRR finished its detailed review of the Toledo Edison 90% undervoltage relay protection system and issued Amendment 7 to the facility Technical Speci-fications which accomplished the following:
2283 '10
- 13 -
'.
.
a.
Approved the licensee's analysis for a 901 under-voltage relay with a 9 second maximum time delay for starting diesel generators and tripping in-coming 4.16 KV source breakers.
b.
Referenced an IE memo dated October 31, 1977, (Exhibit 6) which verified that the 90% relay had been installed.
Dropped License Condition 2.C.3.q regarding the c.
offsite network during power operation.
d.
Changed Toledo Edison's Technical Specification Table 3.3-4 per their request, dated October 27, 1977, (Exhibit 7) from starting the diesel generator at 10 seconds i 1.5 seconds to 7 seconds
+ 1.5 seconds.
Statements received and documents reviewed indicated that Toledo Edison made the request to change Tech-nical Specification Table 3.3-4 for the following reasons:
Their NRC approved 90% undervoltage relay setpoint a.
analysis was based upon starting the diesels in 9 seconds (maximum). Although 9 reconds was within the existing Technical Specification limit of 10 + 1.5 seconds, the existing Technical Specifi-cation would have permitted a setpoint of up to 11.5 seconds which would violate their 9 second under-voltage relay setpoint analysis.
To comply with the 9 second maximum analysis, Toledo Edison would need a 7.5 + 1.5 second setroint.
b.
Toledo Edison was required to comply with Regulatory Guide 1.105 " instrument spans and setpoints."
Regulutory Guide 1.105 requires an extra margin to be added to a setpoint for drift allowance.
Com-pliance with Regulatory Guide 1.105 then lowered the required 90% undervoltage relay setpoint from 7.5 + 1.5 seconds to 7.0 + 1.5 seconds.
Since 7.0 + 1.5 seconds was below the bounding condi-tion of 10 + 1.5 seconds existing in the Technical Specifications, 7.0 1 1.5 seconds became the new Technical Specification requirement for Table 3.3-4 and was granted by the NRC effective November 29, 1977.
Documents reviewed indicated that when Amendment 7 to the Technical Specifications became effective, Toledo Edison had not completed the detailed engineering 2283 311
- 14 -
.
.
required to change the setpoint from the then existing 10 1 1.5 second specification to a new 7.0 i 1.5 second specification. During January 1978, Toledo Edison contracted with the Bechtel Corporation to provide the set-point package for the revised delay setting.
This package was completed by Bechtel on February 19, 1978, and was forwarded to Toledo Edison.
Following Toledo Edison's schedule, the earliest they could have been able to implement the change would have been after February 1978.
This change was designated Facility Change Request No.77-430 and was assigned a priority of "7" defined as "the work is required at the earliest convenience", a low priority.
The following events took place which affected the plant during this time:
a.
Coal strike of 1978. The plant management advised during this period they wished to maintain a 70% power level without tripping.
b.
Toledo Edison's (corporate) apparent judge-ment that installation of the facility change would require the plant to go off the line, as well as perform an installation test which would be equivalent to the major 18-month integrated safety features actuation test, due in June 1978.
c.
An unplanned outage from April 1978 to July 1978, when the plant was in cold shutdown for 3 months for removal of all burnable poison rod assemblics from the core.
While Toledo Edison was in cold shutdown, the quality assurance department generated a memo dated April 1978 which identified various facility changes that were still open.
FCR 421, involving development of the undervoltage relay setpoint table for submission to NRR, was one of those identified (Exhibit 8).
-
In June, 1978 Toledo Edison performed the integrated safety features actuation system (SFAS) test.
This test f ailed initially (IE Notice of Violation, Sep-tember 1, 1978).
Toledo Edison was required to per-form a review of their safety features actuation system in order to determine if any other deficiencies in 2283 312
- 15 -
the SFAS system existed.
TECo advised that during this review on June 12, 1978, they realized that FCR 77-430 had not been implemented, and reported they had been in noncompliance with Technical Specifica-tions (LER 50-346/78-061).
Toledo Edison changed the priority of FCR 77-430 from "7" to
"4", which meant the facility change should be impicmented as soon as possible.
Corrective action implementing FCn 77-430 was completed on June 15, 1978.
The successful 18 month integrated safety features actuation test was performed on June 21, 1978.
Investigators concluded that Technical Specifica-tion Table 3.3-4 was violated for a period of 7 months from November 29, 1977 to June 1978.
The event is described in LER 50-346/78-061, was licensee identified, and corrective action has been taken to implement the FCR.
This item is classified as a licensee identified item.
The review of this event identified a failure to control and implement a facility change request.
Similar failures have been the subject of discussions with the licensee and were referenced in a management meeting of August 1978, as well as inspection report 50-346/79-05.
B.
Concern:
The licensee's delay in implementing the set-point package associated with the 90% undervoltage relay was related to the unsuccessful 18 month SFAS test first performed on June 2, 1978.
Findines A review of the SFAS test information apparently led to the licensee's discovery that the undervoltage relay set-point had not been changed as previously noted.
No other relationship was indicated.
Details Toledo Edison corporate personnel initially advised IE investigators that the 90% undervoltage relay with 7.0 1 1.5 second time delay was not implemented immediately because the integrated SFAS test would have to be performed as a verification of installation.
TECo personnel stated that the SFAS test would require the plant to be shutdown at what was thought to be an inopportune time (during a coal strike).
This information regarding testing was later found to be incorrect.
Documents, procedures, and statements indicated that the 90% undervoltage relay is not functionally tested during the SFAS test and has no bearing on the success or failure of that test.
The 2283 313
_ 16 -
relay setpoint can be changed during power operation.
Procedures indicate that since June 1978, the licensee has included the 90% undervoltage relay setpoints in their surveillance procedures.
C.
Concern:
The licensee did not meet the FSAR assumption of 30 second injection of ECCS following a loss of coolant accident (LOCA).
This prompted the licensee to change the setpoint from 10 seconds i 1.5 seconds to 7 seconds i 1.5 seconds.
Findinas The 90% undervoltage relay setpoint is not considered in the FSAR and was added to the plant as a result of an NRR letter.
The delay setpoint was changed to 9 seconds to assure 30 second ECCS injection, then changes to 7 seconds + 1.5 seconds to account for setpoint drift.
Details The FSAR considers the design basis accident in whic?
there is a total loss of all offsite power coincider with a loss of coolant accident.
The loss of offsite power trips an installed 59% undervoltage relay, starts the diesel generators, and begins sequencing all necessary equipment.
ECCS is assumed to begin injection approximately 30 seconds after the 59% relay trips.
TEco's response to the NRR letter resulted in an instal-lation of an additional relay set at 90%.
This relay would trip after "X" seconds when voltage on the incoming 4.16 KV source buses had fallen between 59% and less than or equal to 90%.
The plant went into power operation with a 90; relay installed and functioning, set at a 10 second time delay period.
The time delay was later changed to 9 seconds to support the Toledo Edison 90%
undervoltage relay safety analysis.
During plant operation, before NRR completed their detailed review of the Toledo Edison submittal, the plant operated with an additional requirement as follows: power operation would only be permitted during times when of fsite 4.16 KV grid voltage was between 98.3% and 102.3% of rated voltage. After Amendment 7 was issued, the 4.16 KV offsite power stability requirement was dropped.
NRR concluded that installed undervoltage protection was sufficient to isolate safety related equipment from a degraded offsite power condition.
As stated above, the licensee did not implement Amendment 7 for a period of appreximately seven months.
The result of this oversight is as follows:
a.
The 90% undervoltage relay was calibrated at 9 seconds and verified by RIII to be 9 seconds to
- 17 -
2283 314
.
Lapport the Toledo Edison safety analysis.
b.
A half-second drif t was not included in the settina, and nonconservative drift could have theoretically pushed the action setpoint to 9.5 seconds thereby delaying starting of the diesel generators by a half second.
5.
Toledo Edison Manacement Deficiencies The review of the delay in undervoltage setpoint implementation indicates that weaknesses exist in areas of nuclear license administration, facility change tracking, and engineering for facility changes.
A.
Encineering for Facility Changes Review of events related to the 90% undervoltage relay setpoint revealed two instances where engineering for a facility change had not been completed when the facility change was required to be implemented as follows:
1.
7.0 +
1.5 second setpoint for the 90% undervoltage relay.
2.
Pushbutton defeat of 90% undervoltage relay.
(Covered in IE inspection report 50-346/79-30).
During discussions with TECo staff the investigators were advised that engineering to support facility change re-quests involving Technical Specifications changes had not been performed until after NRC approval was granted.
Detailed engineering supporting the 90% undervoltage re-lay setpoint was not performed until after NRC granted approval.
The practice of delaying performance of engineering for Technical Specification-related facility changes indicates inadequate facility change administration. Reviews, administration and performance of engineering to support Technical Specification modifications were apparently not performed on a timely basis.
B.
Nuclear License Administration Toledo Edison did not identify to the NRC that an imple-mentation date would be required for Amendment 7, thus the amendment became effective the day of issuance even though Toledo Edison had not performed the necessary revision to the 90% undervoltage relay setpoints to be in compliance with the new license requirements.
This resulted in noncompliance with the Facility Technical Specifications as amended.
2283 al5
- 18 -
C.
Review of Facility Changes Review of the 90% undervoltage ralay event indicated that noncompliance with Technical Specifications occurred because of inadequate adminstrative procedures for carrying out facility changes.
This noncompliance was the result of failure to implement the FCR related to Amendment 7 in a timely fashion.
Attachment:
Exhibits 1 through 9 2283 316
- 19 -