ML19259D204

Responds to 790730 Request for Info Re Degraded Voltage Protection Circuitry.Submits Criteria Re Voltage Selection, Voltage Drop Analysis & Load Injection Scheme

ML19259D204
Person / Time
Site:	Prairie Island
Issue date:	10/12/1979
From:	Mayer L NORTHERN STATES POWER CO.
To:	Office of Nuclear Reactor Regulation
References
NUDOCS 7910170359
Download: ML19259D204 (4)

Text

-.

o NSIB NORTHERN STATES POWER COMPANY M I N N E A f.0 L.l S. M I N N E S OTA 55401 October 12, 19 79 Director of Nuclear Reactor Regulation U S Nuclear Regulatory Commission Washington, DC 20555 PRAIRIE ISLAND NUCLEAR GENERATING PLANT Docket Nos. 50-282 License Nos. DPR-42 50-306 DPR-60 Degraded Voltage Protection Information The following information concerning the degraded voltage protection circuitry installed recently on both units at Prairie Island is provided in response to a letter dated July 3r, 1979, f rom A Schwencer, Ch f, Operating Reactors Branch

1. 1, Division of Opetating Reactors, USNRC:

1. In selecting voltage and time delay setpo:.nts for the low voltage protection circuitry, we have chosen values which minimize the probability of spurious transfer during all modes of plant operation, but which still initiate transfer when voltage levels on the worst case bus (480 V) fall to unacceptable levels based on a conservative analysis.

The undervoltage relay time delay is limited to < 8 seconds to assure completion of transfer of the bus to an available source within 10 seconds. This is consistent with the maximum dalayed availability assumed for the diesel generators when receiving a Safety Injection start signal. The 4160V Safeguard Bus undervoltage relay setpoint is limited to > 88% to assure adequate 480V Bus voltage to pickup motor starter con-tactors. These settings are slightly different than the preliminary values reported earlier.

A voltage drop analysis has been done with the worst postulated case 480V Bus loads. The maximum postulated 480V loads would be applied af ter a Saf ety Injection / Turbine Trip. Loss of voltage would occur transferring the 4160V Safeguards Bus to the alternate offsite source. During the transfer all of the motor starter contactors would drop out and it is assumed all of the motors would s top. Af ter the transfer is completed the 4160V Bus voltage drops to the 88% setpoint limit and it is assumed to remain there for the duratiu of the event. All notor starters that l ._ g 7910170 4 r

NORTHERN STATES POWER COMPANY Director of Nuclear Reactor Regulation Page 2 October 12, 1979 receive an automatic start signal are assumed to pick-up simultaneously starting their motor loads. With all motor loads drawing locked motor current, the 480V Bus voltage would drop to 68%.

More than half of the starting load consists of motors that start unloaded or have starting torque several times that required by the load and are capable of starting at voltabes on the order of 50% of nominal. With these motors up to full speed and the remaining motors still in locked rotor the voltage would increase to 76%. Many loads are capable of starting at 70% voltage, however, only the f ans are assumed to reach full speed. With these two groups of motors up to speed and the remainder in locked rotor, the voltage would be at 80%. At this voltage the remaining motors start allowing the voltage to reach 84%.

With all of the manual non-safety related loads restarted, after being automatically shed on the transfer, the 480V Bus voltage would be 83% of nominal. The total postulated 480 volt load is 1.02 but, 0.44 MVAR.

This loading includes no motor operated valves since most of their repositioning is done automatically before some of the larger loads are sequenced on. The Turbine Generator Airside Seal Oil Back-up Pum,r (20 HP or 0.017 MW, 0.008 MVAR) is not included since it is not expected to run. The Turbine Generator Turning Gear (40 HP or 0.034 MW, 0.016 MVAR) is also not included since it starts approximately one hour af ter the turbine trip. With these postulated conditions the starting times will be longer, perhaps two to three times normal. The effect of this longer duration on the motor overloads would be of fset by the reduced starting currents. The 83% Bus voltage would actually correspond to 86% of the motor nominal voltage of 460 V. Since the motors are designed to operate continuously at 90% voltage and deliver full rated horsepower at 75% voltage, operation at 86% voltage would be accepte b le. Due to slightly higher winding temperatures, a fully loaded motor would lose no more than two days life expectancy for every day of reduced voltage operation.

The 480V motor ararter contactors are being naintained and tested to pickup at or b elow 400V (83%) . Most actually pickup well below 80%

voltage. Even in the postulated case, the voltage becomes marginal only for the non-safety related loads that are tanually started. If a con-tactor does not pickup due to inadequate voltage a delay of about a half ninute may cause the control fuse to blow. Delays of this duration will not be experienced.

j . <. -,_

.J

NORTHERN STATES POWER COMPANY Director of Nuclear Reactor Regulation Page 3 October 12, 1979 This postulated case conservatively assumed many load to be on or running:

1. All heaters on full including; Pressurizer, Diesel Techet Coolant and Boric Acid.

2. All Prelube and Aux Lube Oil Pumps.

3. Compressors: Diesel Start Air, Station Air, and Uaste Gas.

4. Sump Pump.

5. At least two of the three Charging Pumps running at full output.

6. Boric Acid and Fuel Oil Transfer Pumps.

7. All f ans including Spent Fuel Pool Special and Containment In-Service Purge

8. Control Room Water Chiller normally running at reduced load.

All of these loads would actually either not run at all, run for a short time and not run again, or cycle on and off as re-quired. For all credible events, the equivalent 480V Bus load would actually range from less than half of the postulated load to one or two hundred horsepower less than the postulated load.

Many motors would actually continue to spin through the bus transfer without stopping. Many starters would not pickup immediately but would sequence their load on as the voltage improved. Realistically the under voltage setpoini and time delay are more than adequate to assure proper electrical equipment operation.

2. Both the voltage restoration and load rejection schemes require all source breakers, including the diesel generator breaker, to be tripped before transferring to another source or initiating an under voltage load rejection. We have completed a design change in both units which blocks the undervoltage trip signal to the diesel generator source breakers with the Safety injection signal, thereby preventing load shedding of the emergency buses when they are being supplied by the diesels.

This complies with Position 2 of Enclosure (1) to Mr D K Davis' June 3, 1977 letter.

j. 7

' .a

NORTHERN STATES POWER COMPANY Director of Nuclear Reactor Regulation Page 4 October 12, 1979

3. The load rejection scheme has been described in our September 20, 1976 letter. In the case of the diesel generator the voltage restoration scheme and load rejection / sequencing scheme is ccmpletely and automatically electrically reset wnen bus voltage is rer.tored.

This reset must occur before the logic allows loads to be sequenced.

At the time of reset, all of the relay coils in the load rejection are deenergized. Pneumatic time delay drop out relays sequence the start permissives to the loads. Once the voltage restoration and load reject schemes are reset they have:

a) no memory of previous source searches.

b) no way to distinguish which source breakers are closed.

c) no way to know whether or not all of the load sequenciag time delay relays have dropped out.

With the schemes reset, an undervoltage event will trip all sources breakers except for the Safety Injection block ref erred to in iten (2) above. At this point, with all source breakers tripped, the circuit has no way to determine the previous state of any of the source breakers. The logic will search for a source beginning with the offsite sources and ending with load rejection and load sequencing onto the diesel generator.

Therefore, as stated in our letter dated May 4,1978, all of the logic involved in coping with interruption of the on-site sources is verified to be operable Curing the simulated loss of offsite power in conjunction with a safety infection actuation signal. We continue P.o believe that a special periodic test of the design for bus de-energization, load rejection, voltage rest; ration, and load sequencing for interruption of the on-site source is act a reasonable or well founded requirement.

Please contact us if you have any questions related to the infornation we have pro-vided.

i i L 0 Mayer, PE Manager of Nuclear Support Services j < , -,,

i LOM/DMM/ak --

cc: J G Keppler G Charnoff