AECM-85-0233, Forwards Weir Wall Overflow Probability Analysis. Because of Very Low Probability of Occurrence & Leak of Safety Significance for Worst Case Postulated Overflow (Ref AECM-85/0046),corrective Actions Not Warranted

Forwards Weir Wall Overflow Probability Analysis. Because of Very Low Probability of Occurrence & Leak of Safety Significance for Worst Case Postulated Overflow (Ref AECM-85/0046),corrective Actions Not Warranted

ML20138B129
Person / Time
Site:	Grand Gulf
Issue date:	10/04/1985
From:	Dale L MISSISSIPPI POWER & LIGHT CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
AECM-85-0233, AECM-85-233, TAC-55759, TAC-57139, TAC-57140, NUDOCS 8510150103
Download: ML20138B129 (5)

Text

MISSISSIPPI POWER & LIGHT COMPANY EmethMM

] Helping Build Mississippi P. O. B O X 164 0, J A C K S O N, MI S SIS SIP PI 39215-1640 October 4, 1985 NUCLEAR UCENSING & SAFETY DEPARTMENT U. S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, D. C. 20555 Attention: Mr. Harold R. Denton, Director

SUBJECT:

Grand Gulf Nuclear Station Units 1 and 2 Docket Nos. 50-416 and 50-417 License No. NPF-29 File: 0260/0272/L-860.0 Reference 1: AECM-85/0046 Weir Wall Overflow AECM-85/0233 Mississippi Power & Light (MP&L) previously addressed in Reference 1 the issue of weir wall overflow following a postulated inadvertent upper containment pool dump. In that letter it was stated that if all worat case conditions are combir.ed weir wall overflow would occur. MP&L believes that this issue has minimal safety significance as discussed in Reference 1; however, MP&L committed to perform a further evaluation of this issue and to advise the NRC of actions which could be taken to resolve this issue.

The evaluations which were performed by MP&L included a probability analysis of the event, a review of equipment which could be affected by overflow, and an evaluation of options for preventing overflow. The results of these evaluations indicate that corrective actions are not warranted based on the l very low probability for occurrence of weir wall overflow and the lack of safety significance for the worst case postulated overflow as discussed below.

Details of the probability analysis which MP&L performed are described in the attachment to this letter. The range of probabilities calculated was from 3.0E-7 to 4.5E-6 events per year. These probabilit'les are governed primarily by operator actions. Based on the results of the analysis, MP&L believes that the occurrence of weir wall overflow is highly unlikely.

In addition, if overflow were to occur, MP&L has determined through drawing reviews, calculations and walkdowns that the overflow water contacting safety-related equipment would not present a safety concern. This conclusion was made based on the following facts:

1) Safety-related electrical cables contained by raceways passing through the affected area have been qualified in accordance with IEEE-383 and IEEE-323 and have been qualified for submergence.

8510150103 8D1009 PDR ADOCK 05000416 PDR i

/

P J l l J13AECM85072601 - 1 Member Middle South Utilities System

AECM-85/0233 Page 2

2) Class IE terminal boxes are not' located in the affected area.

3) . Safety-related piping containing hot process fluids that would be submerged by the overflow water consists of portions of the 2 inch reactor water cleanup drain lines and portions of the 3/4 inch reactor recirculation drain lines. Stress calculations performed on these lines show that they would maintain their structural integrity following the overflow.

4) Other affected lines which would be submerged or wetted ccanist of small, uninsulated piping near ambient temperature. Flooding will have negligible effects on these lines due to the relatively small temperature difference between the lines and overflow water.

Various options for preventing overflow were evaluated including raising the height of the weir wall, further lowering of the setpoints of existing vacuum breakers, changing to narrow range transmitters, and lowering of the upper pool and/or suppression pool water levels. The results of this evaluation indicate that these optians are either not cost effective or have potential negative safety implications:

-Raising the height of the weir wall would involve extensive work during a lengthy shutdown even assuming only a few interfering components would require modification.

. Lowering the existing setpoints is not feasible because the vacuum breakers would be open a significant amount of time.

Changing to narrow range transmitters would require purchase and installation of new transmitters and controls to supplement the existing transmitters and possible control room modifications.

Lowering the upper pool level is not practical because it would preclude the use of the skimmer system (pool cleanup).

The suppression pool level is limited to a narrow operating range.

Lowering the upper operating limit without changing the lower

. operating limit would result la an overly restrictive narrow operating range. Lowering the lower operating limit would result in a minimum pool inventory below the required Technical Specification limit, thus resulting in a reduced safety margin.

J13AECM85072601 - 2 T

. w AECM-85/0233 Page 3 MP&L believes that such: corrective actions are not warranted based on the very low probability for occurrence of weir wall-overflow and the lack of safety

~ significance for the worst case. postulated overflow.

Yours truly, L. F. Dale-Director ARR/GWS/SHH:vog Attachment-cc: Mr. : J. ' B. ~ Richar'd Mr. O.~D.'Kingsley. Jr.

~

Mr. R.' B. McGehee Mr. N. S. Reynolds

- Mr..H. L. Thomas-

Mr. R. C.. Butcher Mr. James M. Taylor, Director

- Office of Inspection & Enforcement ,

U. - S. Nuclear Regulatory Commission

- Washington, D. C. '20555 L " Dr. J. Nelson Grace, Regional Administrator U. S. Nuclear Regulatory Commission Region II-101 Marietta St., N. W., Suite 2900 Atlanta, Georgia 30323-e Y

f 4

O w

w

~

~-J13AECM85072601 -

3 ,

a :. . , . . . . . , -

Attschment to AECM-85/0233 Page 1 of 2-t WEIR WALL OVERFLOW PROBABILITY ANALYSIS MP&L has performed a conservative analysis to estimate the probability of weir wall overflow. In order for overflow to occur, the following conditions must simultaneously exist:

1. Containment pressure must exceed drywell pressure by at least 0.16 psi, and

2. The suppression pool and upper containment pool water levels must be near their upper operating ranges, and 3.. - An inadvertent upper pool dump must occur.

The most likely scenario which could result in weir wall overflow is a sudden loss of drywell esoling, followed by venting of the drywell and, then, a restoration of drywell cooling. As a result of the loss of drywell cooling, =

the drywell temperature and pressure will increase. At a pressure of

-approximately 2 psi, the high drywell pressure sensors will generate a LOCA l signal. The operator is- expected to vent the drywell in an attempt to preclude reaching the LOCA setpoint; but the LOCA signal is still assumed to be generated because the venting may not limit the pressure rise quickly enough.

After the drywell is. vented and the LOCA signal has been generated, drywell cooling is assumed to be reestablished. Since some of the drywell air volume was vented, the pressure drops as the remaining air is cooled, creating the required 0.16 psid between the drywell.and containment. With the LOCA signal present, the suppression ~ pool makeup system'(SPMU) will automatically initiate

~

in 30 minutes or can be initiated manually using two control room handswitches.

Thus, if th'e operator fails to reset the LOCA signal within 30 minutes or inadvertently. manually; initiates SPMU, conditions w'ould exist that. allow the suppression pool to overriow the weir wall.

The analysis that was performed is very conservative in nature. The following major assumptions were made:

1. The operator vents the drywell upon loss of drywell . cooling to preclude reaching the LOCA setpoint. Note that if the operator fails to vent or does not attempt to vent before the LOCA signal is

' generated (which precludes venting), the drywell negative pressure cannot develop'and weir-wal ' overflow does not occur.

2. Any suppression pool level instrumentation. failure generates a LLow-Low level signal.

3. . The pressure differential required for weir wall overflow is always present if the operator vents the drywell-(even though this

• Ldifferential may actually only be present for a fraction of the time that any negative. pressure exists).

4 '. .The suppression pool and upper containment pool levels are always near the upper limit within the range required to experience weir wall-overflow.

J13AECM85072601 - 1 m__ma '. O _

1 Attachmint to AFCM-85/0233 Page 2 of 2

5. The frequency of loss of off-site power is 0.15 per reactor-year

-(conservative in comparison to a recent study suggesting a national average-value of 0.09 per reactor-year).

6. The LOCA signal is generated even if the operator initiates drywell venting.

7. Drywell inleakage is neglected.

The dominant branch of the event tree is governed primarily by operator actions. For a major loss of drywell cooling, the LOCA setpoint will be reached very quickly (tpproximately 1.5 minutes). The realistic probability that the operator can vent the drywell in this period of time is low. In this case, the probability of weir wall overflow is 3.0E-7 events per year.

If it is assumed that the operator performs all functions correctly (i.e.

vents within 1.5 minutes following loss of off-site-power-and resets the LOCA signal), then the probability of weir wall overflow is 4.5E-6 events per

, year. This case is dominated by a spurious upper pool dump resulting from an equipment failure.

MP&L believes that because this analysis is based on a series of conservative assumptions and dominant events, the figures 3.0E-7 to 4.5E-6 represent a bound on the realistic range of event probability. The occurrence of weir wall overflow is, therefore, very unlikely.

J13AECM85072601 - 2

_.u