ML20210T752

Responds to NRC 860612 Request for Addl Info Re Scram Discharge Vol Vent & Drain Valve Closure Times

ML20210T752
Person / Time
Site:	Hatch
Issue date:	09/23/1986
From:	Gucwa L GEORGIA POWER CO.
To:	Muller D Office of Nuclear Reactor Regulation
References
10844, 51-1219, SL-1219, TAC-53327, TAC-53328, NUDOCS 8610090184
Download: ML20210T752 (4)

Text

.. . - ._

Georgia Fewer Corrpany 333 Piedrnont Annue Attznta, Georgia 30308 Telephone 404 5264526 Masting Address:

Post Office Box 4545 Atlanta, Georgia 30302 Georgia Power L.T.Gucwe Ito sothen ektre 3:yavn Manager Nuclear Safety and ucensing SL-1219

%. "u",'i-September 23, 1986 Director of Nuclear Reactor Regulation Attention: Mr. D. Muller, Project Director BWR Project Directorate No. 2 .

Division of Boiling Water Reactor Licensing U. S. Nuclear Regulatory Commission Washington, D. C. 20555 NRC DOCKETS 50-321, 50-366 OPERATING LICENSES DPR-57, NPF-5 EDWIN I. HATCH NUCLEAR PLANT UNITS 1 AND 2 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING SDV VENT AND DRAIN VALVE CLOSURE TIMES Gentlemen:

By letter of June 12, 1986, Georgia Power Company (GPC) was requested to

' provide additional information in regard to proposed Technical Specifications for Scram Discharge Volume (SDV) vent and drain valve closure times. The NRC questions and GPC responses are hereby provided:

NRC Question 1:

Your letter of June 14, 1984 indicates that extensive modifications I

would be required to meet a 30 second closure time for the Scram Discharge Volume (SDV) vent and drain valves. Your letter of September 13, 1985 seems to indicate that the necessary plant modifications have l been made. If they have been made, then we do not understand why you can't agree with the 30 second closure limit. If they have not been made, quantify your estimate of " extensive modifications": For example,

(

approximate amount of additional piping, extent of modification labor I (in man-hours or dollars), length of time for completing the modifications.

t GPC Response 1:

l As explained in GPC's letter of June 14, 1984, two types of modifications were considered in order to meet the 30 second closure time requirement. The " minor" modification involved replacing the solenoids which actuate the SDV vent and drain valves with solenoids

! having a larger exhaust port, and, therefore, quicker response time.

I The " major" modification involved assignment of new, independent i

8610090184 860923 \

PDR ADOCK 05000321 P PDR O

Georgia Power d Dir!ctor of Nuclear Reactor Regulation Attention: Mr. D. Muller, Project Director BWR Project Directorate No. 2 September 23, 1986 rage Iwo solenoids for each vent and drain valve, and installation of associated air piping. The " minor" modification was performed on both units, and did result in closure times of less than 30 seconds for the vent and drain valves for both units. However, there was very little margin to the 30 second time limit. (The limiting valve closed in 29.95 seconds.) GPC is proposing a closure time value of 45 seconds for two reasons: 1) The 30 second value does not provide enough margin to the actual closure times (following the " minor" modification), and 2)

Analyses have been performed justifying longer closure times. GPC has no present plans to proceed with the " major" modification.

NRC Question 2:

Your analysis of valve closure times seems to be predicated on the assumption that the water level in the SDY during normal operation prior to a scram is zero, i.e. , the volume does not contain water. However, considerations of drainage piping flow resistance due to friction, as well as a postulated single failure (such as a closed drain valve),

could lead to a non-zero water level within the SDY. In view of the above, indicate the maximum potential water level that could accumulate within the SDV prior to a scram.

GPC Response 2:

During normal operation the SDV vent and drain valves are open and the

. SDV would be empty. If for any unspecified reason the drain valve was

! closed, the maximum water level that could accumulate within the SDY l prior to a scram with rods out is that level which would result in an SDV high high level scram signal. This value is 71 gallons for Unit 1 and 57 gallons for Unit 2. The SDV high high level scram is designed to l

scram the reactor while enough free volume, with appropriate margin, exists in the SDV to accommodate the volume of water which will be discharged from the upper side of the CRD pistons through the scram outl et valves during scram. This ensures that all rods are fully inserted prior to SDV fillup.

Following a scram and prior to scram reset, the scram outlet valves will remain open and the SDV vent and drain valves will be closed.

l i Under these conditions, the SDV will fill with water. When the scram is reset, the scram outlet valves close, stopping inflow into the SDY.

Al so, the vent and drain valves open, allowing the SDV to drain.

Postulated additional scram signals immediately following scram reset provide the limiting case since the SDV would be nearly full at the time of the subsequent scram signal. A shorter closure time value actually 1084y

\ mm

. , s l

Georgia Power d Director of Nuclear Reactor Regulation Attention: Mr. D. Muller, Project Director BWR Project Directorate No. 2 September 23, 1986 Page ihree exacerbates this condition, since it allows less time for the SDV to drain from the initial scram prior to re-closure of the vent and drain valves. The important consideration is that this condition could only occur with all rods in (from the initial scram). Consideration of single failure is not appropriate since SDV overfill is not associated with a design basis accident. The pertinent point is that, with certain assumptions, it can be demonstrated that the SDV will overfill, but this condition can only occur after all rods are in and the SDV has performed  :

its function. The proposed 45 second closure time limit is designed to

! preclude SDV overfill for normal scrams. No closure time limit will preclude SDV overfill for all postulated conditions including multiple scram signals and single failures. The consequences of SDV overfill in these remote situations are water in the vent l ines, increased discharges to radwaste, and potential radiological concerns in the reactor building. These consequences are undesirable from an operational standpoint but do not represent design basis considerations.

NRC Question 3:

Your letter of November 18, 1985 provides stall flow measurements made over two full cycles of operation. This information was used as a basis for establishing a maximum post-LOCA CRD leakage. Provide justification that the stall flow value and hence your estimates of maximum post-LOCA i

CRD seal leakage will not be exceeded. Include a discussion of the error bounds of the data that relates the stall flow to post-LOCA flow.

r GPC Response 3:

Previous GPC letters have addressed stall flow criteria for CRD rebuild. These criteria provide a high degree of confidence that excessive post-scram leakage by the CRD seals will not occur. CRD seals i tend to degrade gradually with time. As long as the maintenance criteria are preserved, there is no reason to believe that CRD seal.

leakage will increase significantly over the values provided in our letter of November 18, 1985. It is not possible to guarantee that the stall flows provided could not be exceeded by some small degree in future cycles. However, calculations using conservative assumptions show that adequate SDV margin is preserved to accommodate some increase in average seal leakage for a 45 second vent and drain valve closure time. No information is available regarding error bounds for the data that correlate stall flow to post-scram flow. The correlation was

. excerpted from General Electric Topical Report NED0-24342.

1084y 700775

_____m. _ _ _ _ - _ - _ _ _ _ . . - _ _ _ _ - . . . _ _ _ . _ . - . . . __

d I

Georgia Power A Director of Nuclear Reactor Regulation Attention: Mr. D. Muller, Project Director BWR Project Directorate No. 2 -

September 23, 1986 r Page Four NRC Question 4:

j The General Electric report " Relaxation of Scram Discharge Volume Vent l and Drain Valve Closure Times" dated December 1984 and your letter of i December 22, 1983 list the Unit 1 SDV capacity as 482 gallons. Your letter of January 6,1986 lists SDV capacities of 549 gallons for Unit 1 and 636 gallons for Unit 2. Explain these differences and provide '

detailed information (e.g. calculations) that will permit verification of the "as built" volume of the SDV for each unit.

GPC Response 4:

The FSAR specifies 482 gallons as the Unit 1 SDV capacity. This is a The SDY conservative,ignatedlicensing capacity des basisisvalue in the FSAR based used on by General" Electric.

a specific volume per drive" criterion which provides the m)aimum SDV volume, with appropriate j margin, to ensure the scram function, which is a requirement for design

, basis events. For consideration of non-design basis events such as SDV overfill, it is appropriate to use the actual, as-built volume rather <

than the FSAR volume. The volumes of 549 gallons for Unit 1 and 636

gallons for Unit 2 are based on as-built calculations performed by our Architect / Engineer. As always, these calculations are available for NRC inspection at the offices of our Architect / Engineer.

If you desire further information, please contact this office.

i l Sincerely, il d#ft 2 L. T. Gucwa

! REB /1c f

c: Georgia Power Company U. S. Nuclear Regulatory Commission Mr. J. P. O'Reilly Dr. J. N. Grace, Regional Administrator Mr. J. T. Beckham, Jr. Mr. P. Holmes-Ray, Sr. Resident Mr. H. C. Nix, Jr. Inspector - Hatch

GO-NORMS i

i i

1084y 700775

, , _ _ . - - . . , _ _ . - _ . - . - . . - _ . . _ - _ , ~ _ - . - _ . . , . . . . . - _ - ,,_, - -- _ ,_ _ --_ _ .- - ... _ _ _ _.. _ _ _ - - ,