B12497, Provides Current Status of Plans to Address Degraded Grid Issue at Facility.Proposed Scheme to Address Issue Involves Use of Complex Panels Containing Electronic Relays & Bistables

Provides Current Status of Plans to Address Degraded Grid Issue at Facility.Proposed Scheme to Address Issue Involves Use of Complex Panels Containing Electronic Relays & Bistables

ML20214P388
Person / Time
Site:	Millstone
Issue date:	05/22/1987
From:	Mroczka E NORTHEAST NUCLEAR ENERGY CO., NORTHEAST UTILITIES
To:	NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM)
References
B12497, NUDOCS 8706030352
Download: ML20214P388 (8)

Text

_ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ ,

s NORTDMAST UTHJTIES cenerai Omces . seiden street. Beriin, connecticut 9 E.IS$nEs N(

m. .an. .o- . cw '

P.O. BOX 270 HARTFORD. CONNECTICUT 06141-0270 k k J $)$ ((E Y.'2.[, (203) 665-5000 May 22,1987 Docket No. 50-245 B12497 Re: Degraded Grid Protection U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D.C. 20555 Gentlemen:

Millstone Nuclear Power Stati , Unit No.1 Degraded Grid Protection for Class IE Power Systems By letters dated November 14, 1985(l) and January 13, 1986,(2) Northeast Nuclear Energy Company (NNECO) identified the need to delay implementation of plant modifications related to degraded grid protection for Class IE power systems from the Cycle 10 outage to the Cycle 11 outage scheduled for summer, 1987.(3)This delay was approved by the NRC in a letter dated December 12, 1985. The purpose of this letter is to provide the NRC Staff with the current status of our plans to address the degraded grid issue at Millstone Unit No.1.

Background

NNECO has strived to incorporate a loss of voltage and degraded voltage detection and actuation Position PSB-1 (BTP PSB-1) sc}epe 41 on meeting the intent three different occasions,of NRCeach Branch Technical of these timed to occur during refueling outages. On all occasions, we have had to defer implementation of modifications due to the uncovery of additional safety concerns.

(1) 3. F. Opeka letter to 3. A. Zwolinski, " Degraded Grid Protection for Class IE Power Systems," dated November 14,1985.

(2) 3. F. Opeka letter to C.1. Grimes, " Integrated Safety Assessment Program -

Supplement to ISAP Topic No.1.25," dated January 13,1986.

(3) C. I. Grimes letter to 3. F. Opeka, " Implementation of Degraded Grid Protection for Class IE Power Systems," dated December 12,1985.

(4) NRC Branch Technical Position PSB-1, " Adequacy of Station Electric Distribution System Voltages." r N

s i' ER61883E$78Eh5 P

U'.S. Nuclear Regulatory Commission B12497/Page 2 May 22,1987 The proposed scheme to address this issue involves the use of complex panels containing electronic relays and bistables. Our experience in the early 1980s, wherein the incorporation of an ATWS-related electronics-based scheme resulted in several spurious and unnecessary trips of the. unit, led NNECO to make the decision to " burn-in" these new voltage scheme-related panels prior to placing them into auto-actuation service. With the NRC's concurrence, Millstone Unit No. I operated through Cycle 10 with the new scheme in a monitor and alarm-only mode. The units proved to be highly reliable, thus, it was our intent to place the scheme into full service prior to startup for Cycle 11 operation.

Shortly before the Cycle 10 refueling outage, during which the scheme was to be placed into operation, NNECO, as part of the Integrated Safety Assessment Program, performed a probabilistic-based study of the loss of voltage and degraded voltage detection system and identified that the incorporation of the scheme as designed would result in a significant (240%) increase in frequency of station blackout. The. study found that the original scheme, wherein intradivision crosstalk was allowed, was considerably more reliable than the new proposed scheme, primarily due to the doubling of functions performed by each safety division backing its redundant counterpart (diesel and gas turbine auto start, for example). In the new scheme, the BTP PSB-1 criterion for the auto-reinstatement of load shed feature led NNECO to remove the intradivisional crosstalk, resulting in a situation wherein single device failures would cause a loss of their associated power division. While this fully met the intent of the single failure criterion, the probabilistic-based study identified a significant increase in core melt frequency as a result of that design. As such, NNECO requested a one fuel cycle deferral in implementation schedule to improve the design such that BTP PSB-1 specifications could be met without a reduction in plant safety. The NRC approved the request and the design was subsequently revised to provide a doubling of critical devices within each division. The new design was subjected to a rerun of the probabilistic-based study, the result being a design with reliability essentially equal to that of the original scheme. (Itis noteworthy that NNECO has on many occasions demonstrated the value of PRA-hased studies. Once again, the avalbbility of these new techniques prevented a previously unidentifiable reduction in unit safety.)

More recently, due to the complexity of the new scheme and the rather significant change in plant response to voltage-related problems, a decision was made to .model the new scheme on the Millstone Unit No. I plant-specific simulator. Once again, the design features necessitated by the auto-reinstatement of load shed capability requirement proved to be troublesome resulting in operator confusion due to the significantly different unit response to a loss of power. Concerns arose as to the operators' ability to assess the developing scenario in time to prevent such events as serious level transients, since the reactor is not directly scrammed by the new scheme, but rather scrams some time later due to a process . variable exceeding its setpoint. In the final analysis, the Millstone Unit No. 1 Plant Operations Review Committee concluded, with Northeast Utilities Service Company's Generation Engineering and Millstone Unit No. I's Plant Engineering concurrence, that the new scheme should not be placed into service without, as a minimum, the addition of a turbine generator or reactor trip initiated directly as a result of low or degraded voltage.

U.S. Nuclear Regulatory Commission B12497/Page 3 May 22,1987 Current Sta+us We are presently reviewing and redesigning the loss of voltage and degraded voltage scheme to address these issues, and have determined that insufficient time remains for completion of these activities for incorporation during the upcoming Cycle 11 refueling outage. NNECO thus finds it necessary to defer the implementation schedule to allow appropriate design changes to be made.

In summary, we have determined that the Millstone Unit No. I electrical bus arrangement, due to its lack of symmetry, is not readily compatible to a backfit which requires the auto-reinstatement of load shed feature. The original loss of voltage scheme was designed to create a single response, that being to put the unit into a full loss of offsite power condition, not a partial loss like most other events. When we attempted to design a partial loss capability (i.e., one safety division can suffer loss of normal power (LNP) without the other and vice-versa),

problems as described above arose. NNECO is rp, at this poin , fully confident that the auto-reinstatement of load shed featu" can be safely incorporated. We intend to pursue one more round of design change considerations to definitively determine our position on this subject. The Staff should be appraised that we may find it necessary to deviate from some of the BTP PSB-1 criteria. NNECO believes the lack of the auto-reinstatement feature is acceptable from a single failure standpoint and will provide appropriate documentation should it become necessary.

For your information, Attachment 1 provides a description of the existing undervoltage protection scheme at Millstone Unit No.1. Attachment 2 is a comparison of the loss of voltage and degraded voltage mheme presently in place (since 1976) with the criteria of BTP PSB-1. Wh;1e all criteria are not met, NNECO is confident that the intent of BTP PSB-1 is essentially met, and that Millstone Unit No. I can operate safely until new designs are incorporated.

Per discussions between the NRC and members of our staff on March 20,1987, we are submitting this information to enable Staff review of our plans. We are also willing to meet with the Staff to discuss this issue.

If you have any questions, please feel free to contact my staf f.

Very truly yours, NORTHEAST NUCLEAR ENERGY COMPANY h ff A ~

E. JAiroczka f' Serftor Vice President cc: W. T. Russell, Region I Administrator

3. 3. Shea, NRC Project Manager, Millstone Unit No.1 .

T. Rebelowski, Resident Inspector, Millstone Unit No. I

Docket No. 50-245 B12497 Attachment 1 Description of Millstone Unit No. I's Existing Undervoltage Protection Scheme May,1987

Attachment i B12497/Page 1 Existing Design The current undervoltage protection scheme senses voltage at the 345 kV level on the high voltage side of the RSST. The loss-of-voltage relays are General Electric (GE) ModelIAV55s which have an inverse time characteristic. Degraded voltage relays are solid-state Devar transmitters. A 120-VAC sensing source is provided by transformer high voltage bushing potential devices. The Devar relay contact outputs operate into Agastat timers for both alarm and tripping functions. Both sets of voltage relays are designed to protect the station electrical distribution system from transients on the offsite supply system. The loss-of-voltage relays protect against a complete collapse of the switchyard voltage, while the degraded voltage relays alarm a degraded voltage condition.

The loss-of-voltage relays will trip the offsite system if the RSST is in use; the degraded voltage relaying is interlocked with an accident signal to trip the offsite supply system if degraded voltage occurs coincident with an accident.

The existing logic consists of two redundant circuits that trip the offsite supply, load shed the buses, start the diesel and gas turbine generators, and input the sequencing logic.

The loss-of-voltage setpoint on the 2 GE lAV voltage relays is set at 32V, which corresponds to 246 kV in the switchyard. The degraded voltage setpoint on the Devar voltage relays is set at il5V, which corresponds to 345 kV in the switchyard. The GE lAV relays have an associated inverse time delay characteristic. The Devar relays operate into Agastat timers which provide a 10 second time delay to alarm, and I second delay to trip if degraded voltage occurs coincident with ECCS. The loss-of-voltage (71 percent or less) logic is one-out-of-two taken once. The degraded voltage (90 percent of 4 kV buses) logic is one-out-of-two taken twice interlocked with the ECCS signal. The logic described above is in effect only when the plant is aligned to the RSST.

Should loss of either onsite power source occur subsequent to it providing power to the emergency buses, reload shed is achieved by operator action. Tripping of all the required circuit breakers can be accomplished using control switches on the main control room panel.

At the safety-related 4kV buses, degraded voltage is sensed and provided as an alarm to the control room operator.

Docket No. 50-245 B12497 Attachment 2 Conformance Review to BTP PSB-1 May,1987

Attachment 2 B12497/Page1

~

By letter dated June 2, 1977,(1) the NRC Staff transmitted a Safety Evaluation and Statement of Staff Positions on the susceptibility of safety-related electrical equipment with regard to: 1) sustained degraded voltage conditions at the offsite power sources, and 2) interaction between the offsite and onsite power systems. At that time Millstone Unit No. I was requested to: 1) propose plant modifications as necessary to meet the Staff Positions, or 2) provide a detailed analysis which shows your facility design has equivalent capabilities and protective features. Additionally, it was requested that certain technical specifications be incorporated into the operating license.

Conformance Review The following is a conformance review to BTP PSB-1 of Millstone Unit No. I's ability to cope with: 1) degraded voltage cond;lons, and 2) automatic reload-shed capability.

Position 1 -Second Level of Undervoltage Protecticn with a Time Delay a) NNECO complies with the position that the voltage and time setpoints should be determined from an analysis of the voltage requirements of safety-related loads at all onsite system distribution levels. The basis of the setpoints have been determined by analysis and test.

b) The voltage protection scheme includes coincident logic to preclude spurious trips of the offsite source. The degraded voltage trip (coincident with ECCS) is one-out-of-two taken twice with voltage sensed on the high voltage side of the RSST. The degraded voltage alarm signal for the voltage sensed on the 4 kV buses is two-out-of-three on any one of three 4 kV buses.

c) The selected time delays: 1) do not exceed the maximum time delays assumed in the FSAR accident analyses, 2) minimize the effect of short duration disturbances from reducing the availability of the offsite power source and 3) do not allow the duration of a degraded voltage condition to result in failure of safety systems and components.

d) The degraded voltage monitors at the 345kV level automatically initiate l disconnection of the offsite power sources only when an ECCS signal i exis ts. NRC goncurrence with this position was obtained in a letter dated j June 23,1982.t2) e) The degraded voltage monitors at the 345kV level on the high voltage side of the RSST which provide trip and alarm functions do not meet IEEE 279-l l (1) G. Lear letter to D. C. Switzer,

Subject:

Millstone Unit No.1 Emergency l Power System, dated June 2,1977.

(2) D. M. Crutchfield letter to W. G. Counsil, " Millstone Unit No.1 - Degraded Grid Protection for Class IE Power Systems," dated June 23,1982.

t

Attachment 2 B12497/Page 2

-1971; however, the degraded voltage sensors located on the 4 kV buses for the purposes of alarming a degraded voltage condition do meet IEEE 279-1971 and procedures for operator response to degraded voltage will be implemented during the Cycle 11 outage. Specifically, Section 4.8 of the standard states, "to the extent feasible and practical, protection system inputs shall be derived from signals that are direct measures of the desired variables." It is NNECO's intent to relocate the degraded voltage tripping logic to the 4 kV buses, f) Technical specifications for the second-level voltage protection monitors were submitted July 16, 1980.(3) Upon completion of the redesign of the degraded voltage scheme, new technical specifications will be submitted once the voltage relaying that provides tripping is relocated.

Position 2 -Interacticn of the Onsite Power Sources with Load Shed Feature The Staff's position is that: 1) once the onsite power sources are supplying the emergency buses, the load-shed feature should be bypassed, and 2) automatic reinstatement of the load-shed feature shall be provided should an emergency generator trip. NNECO concurs with the Staff's bypass position and the existing design does bypass the load-shed feature once the onsite supplies are powering the buses. However, while the Staff does not specify the reason for autornatic reinstatement of the load-shed feature, it is NNECO's position that this capability should not be mandated for all designs. For Millstone Unit No.1, NNECO prefers to credit a manual load shed in the event that one of the two emergency generators trip. In this regard, there are a minimal number of breakers that need to be tripped prior to re-establishing an onsite source, and the control switches for the breakers are located on the main boards. Operators ice fully trained on which breakers need to be tripped. As stated in the cover letter, NNECO will attempt to incorporate an auto-reinstatement feature.

(3) W. G. Counsil to D. M. Crutchfield, " Proposed Technical Specification Changes for Station Electric Distribution System Voltage," dated July 16, 1980.