ML20126C606
| ML20126C606 | |
| Person / Time | |
|---|---|
| Site: | Maine Yankee |
| Issue date: | 06/07/1985 |
| From: | Whittler G Maine Yankee |
| To: | John Miller Office of Nuclear Reactor Regulation |
| References | |
| 5708L-GDW, GDW-85-166, MN-85-111, NUDOCS 8506140476 | |
| Download: ML20126C606 (2) | |
Text
o
/
o MAME HARHEE ATOMICPOWERCOMPARUe avausrajain"s%
(207) 623-3521 g
June 7, 1985 MN-85-lll GDW-85-166 Director of Nuclear Reactor Regulation United States Nuclear Regulatory Commission Washington, D. C.
20555 Attention:
Mr. James R. Miller, Chief Operating Reactors Branch No. 3 Division of Licensing
References:
(a) License No. O'R-36 (Docket No. 50-309)
(b) MYAPCo Letter to USNRC dated January 14,1985 (MN-85-10)
(c) USNRC Letter to MYAPCo dated March 12, 1985
Subject:
Testing Frequency for Emergency Feedwater Auto-Initiation Logic Gentlemen:
In Reference (b), Maine Yankee submitted proposed technical specifications to require testing cf the emergency feedwater automatic initiation logic every refueling.
The proposed test frequency is the same as the test frequency for similar circuitry in the safety injection system.
In Reference (c), the NRC stated that testing of the emergency feedwater logic should be required every 62 days.
The Maine Yarkee emergency feedwater system automatic equipment is not designed to be conveniently tested while the plant is at power. Performance of this test with the plant at power would reduce system reliability by removing a complete train from service, increase the likelihood of an inadvertent plant trip through personnel error, and increase the risk of damage to equipment.
In our opinion, testing of the automatic start logic of the emergency feedwater system with the plant at power would require a modification to the design of the system to specifically provide for such testing. We have determined that such a modification would be a BACKFIT as described by part 50.109 of the Commission's regulations. Part 109 states that the NRC may require backfitting of plant design only after it has determined that the proposed backfit would provide substantial additional protection to public health and safety.
The NRC has not demonstrated, in accordance with 10 CFR 50.109, that this backfit would result in substantial additional protection. Our reasoning follows.
More frequent testing of the auto start logic will not necessarily improve the reliability of emergency feedwater system.
The auto start logic has been designed to be, and similarly designed logic has proven to be, extremely reliable. More frequent testing might instead increase the likelihood of failure.
8506140476 850607 g[
DR ADOCK 050 9
O 5708L-GDW
MAINE YANKEE ATOMIC POWER COMPANY
.a e
United States Nuclear Regulatory Commission Page Two Attention:
Mr. James R.. Miller, Chief MN-85-lll Additionally, the nuclear industry has objective evidence that testing safety equipment involves a significant risk that the equipment will not be returned to service following.the test because of operator error.
Thus, more frequent testing could increase the likelihood of this type of personnel error, thereby decreasing the reliability of the equipment.
Even if the more frequent testing of the auto start logic were to increase the reliability of the auto start of the emergency feedwater system, there would not be substantial additional protection to public health and safety.
Under even worst case hypothetical design base accident scenarios requiring emergency feedwater, the emergency feedwater is not required for approximately thirteen minutes following the onset of the accident.
This provides ample time for the operators to manually start emergency feedwater flow in the highly unlikely event that the redundant auto start feature of both emergency feedwater pumps failed to operate on demand.
In recognition of this fact, the plant was originally designed with manual start emergency feedwater pumps.
On the other hand, a safety injection pump must start almost immediately after the onset of a postulated loss of coolant accident in order to 1
successfully mitigate the accident. The EC accepts testing of the auto start logic for the safety injection pumps each refueling outage.
In summary, the burden of proof to demonstrate that backfitting of the emergency feedwater system to allow logic testing with the plant at power is necessary, rests with the NRC. The NRC has not demonstrated that the backfit is necessary or will result in substantial additional protection to public health and safety.
Therefore, we have concluded that backfitting of the emergency feedwater logic is not warranted and therefore, the automatic i
actuation logic should be tested each refueling as we have proposed.
We would be pleased to meet with you if you desire to discuss this matter further.
Very truly yours, i
MAINE YAWEE ATOMIC POWER COWANY A b NV f
G. D. Whittler, Manager Nuclear Engineering & Licensing i
GDW/bjp cc: Dr. Thomas E. Murley Mr. Cornelius F. Holden s
l
[
5708L-GDW
-