Text

Response to Request for Additional Information Re NRC Generic Letter 2004-02
	ML092380647
Person / Time
Site:	Farley
Issue date:	07/27/2009
From:	Ajluni M; Southern Nuclear Operating Co
To:	; Document Control Desk, Office of Nuclear Reactor Regulation
References
	GL-04-002, NL-09-0982
	Download: ML092380647 (30)
	v • d • e

Southern Nuclear Operating Company, Inc.

Post Office Box 1295 Birmingham, Alabama 35201-1295 Tel 205.992.5000 July 27, 2009 Energy to Serve Your World'"

Docket Nos.: 50-348 NL-09-0982 50-364 U. S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, D. C. 20555-0001 Joseph M. Farley Nuclear Plant Response to Request for Additional Information Regarding NRC Generic Letter 2004-02 Ladies and Gentlemen:

In letters dated February 28, 2008 and April 29, 2008, Southern Nuclear Operating Company (SNC) submitted supplemental responses to Generic Letter (GL) 2004-02 "Potential Impact of Debris Blockage on Emergency Recirculation During Design Basis Accidents at Pressurized-Water Reactors," dated September 13, 2004.

On March 9, 2009, a Nuclear Regulatory Commission (NRC) Request for Additional Information (RAI) was received regarding the supplemental responses.

By SNC letter dated June 3, 2009, SNC stated that a reply to the RAIs would be submitted by July 31, 2009. The SNC response to the requested information is provided in the Enclosure 2.

A listing of significant conservatism information, that supplements the overall compliance information included in the supplemental response to Generic Letter (GL) 2004-02 dated February 28, 2008, is provided in Enclosure 1.

This letter contains an NRC commitment, described in Enclosure 3, to revise procedure(s) to address latent debris.

A // (P

U. S. Nuclear Regulatory Commission NL-09-0982 Page 2 Mr. J. J. AjIuni states he is Nuclear Licensing Manager of Southern Nuclear Operating Company, is authorized to execute this oath on behalf of Southern Nuclear Operating Company and to the best of his knowledge and belief, the facts set forth in this letter are true.

If you have any questions, please advise.

Sincerely, M. J. Ajluni Manager, Nuclear Licensing to and subscribed before me thisMZ._.__ day of ,J..J )2009.

Notary Public My commission expires: 7--/ /'

MJA/JLS/phr

Enclosures:

1 - Significant Conservatisms in FNP Emergency Core Cooling System Sump Design, Testing, and Analysis 2 - Response to Request for Additional Information Regarding Generic Letter 2004-02 3 - Commitment Table cc: Southern Nuclear Operating Company Mr. J. T. Gasser, Executive Vice President Mr. J. R. Johnson, Vice President - Farley Ms. P. M. Marino, Vice President - Engineering RTYPE: CFA04.054 U. S. Nuclear Re-gulatory Commission Mr. L. A. Reyes, Regional Administrator Mr. R. E. Martin, NRR Project Manager - Farley Mr. E. L. Crowe, Senior Resident Inspector - Farley

Joseph M. Farley Nuclear Plant - Units 1 and 2 Enclosure 1 Significant Conservatisms in FNP Emergency Core Cooling System Sump Design, Testing, and Analysis

Enclosure 1 Significant Conservatisms in FNP Emergency Core Cooling System Sump Design, Testing, and Analysis This information supplements the overall compliance information included in the supplemental responses to Generic Letter (GL) 2004-02 dated February 28, 2008 and April 29, 2008.

Detailed analyses of debris generation and transport ensure that a. bounding quantity and a limiting mix of debris are assumed at the containment sump screen following a design basis accident (DBA). Using the results of the analyses, conservative evaluations were performed to determine worst-case screen head loss. Other conservatisms were applied to ensure that net positive suction head (NPSH) margins were conservatively calculated and conservative testing was done to demonstrate that vortexing and air ingestion would not occur.

The following is a list of significant conservatisms in Joseph M. Farley Nuclear Plant (FNP) emergency core cooling system (ECCS) sump design, testing and analysis. It is provided to demonstrate that a conservative holistic approach for the resolution of General Safety Issue (GSI) -191 is in effect at FNP.

1. Debris interceptors are installed in both FNP Units 1 and 2 containments. No credit is taken in either the analysis or testing for debris captured by these interceptors. The interceptors are located in the debris flow path between the large-break loss-of-coolant accident (LBLOCA) zone of influence and the ECCS sump screens in the secondary shield wall access points. While the amount of debris intercepted by these interceptors is not quantified, they provide defense-in-depth.

2. No credit was taken for near field debris settling. The test arrangement for FNP was highly stirred using multiple mechanical mixers along with test facility flow to lift the debris and chemical surrogates to the extent practicable so that the maximum amount practicable deposited upon the screens. As the ECCS sump has many quiescent areas, it is reasonable to expect that significant settling of coating debris would occur following an LOCA scenario and much less debris would transport and lift upon the screens than tested.

3. FNP has separate ECCS sump screens for each residual heat removal (RHR) and containment spray (CS) pump. There are a total of four screens in each unit. Screen testing was done with the assumption that only one train of RHR and CS operate (2 of 4 screens) thus doubling the amount of debris loading to each screen as compared to all four pumps operating. Assuming only one of the four pumps failed to operate would reduce the amount of debris deposited on each screen to approximately 2/3 of the tested values.

4. To generate the total debris loading for the screens, the debris quantity for the limiting break location that generated the most coatings debris is combined with the debris quantity from the one location that generates the most insulation debris. In reality, these are two separate break locations that cannot occur simultaneously. Thus, the tested debris loading for the screens is maximized.

5. FNP assumed that all failures of acceptable coating in the zone of influence (ZOI) were as chips. Since FNP is a very low fiber plant, this is more conservative than the assumption that the coating failed as particulates. FNP specific testing demonstrated El-1

Enclosure 1 Significant Conservatisms in FNP Emergency Core Cooling System Sump Design, Testing, and Analysis that chips increase head loss more than particulates for the FNP (very low fiber) debris loading.

6. Nonqualified containment coatings are all assumed to fail. Electric Power Research Institute (EPRI) report "Design Basis Accident Testing of Pressurized Water Reactor Unqualified Original Equipment Manufacturer Coatings" for OEM coating failures documented testing on various types of unqualified coatings, alkyds, epoxies and inorganic zinc. A 100% failure of all unqualified coatings is conservative, since the EPRI report has indicated that only about 20% of unqualified OEM coatings actually detached as a result of autoclave DBA testing.

7. The head loss associated with the Reflective Metal Insulation (RMI) transported to the sump was treated as separate from the head loss associated with the other debris. This is considered conservative, as a mixed debris bed containing RMI would be expected have a lower head loss.

8. All debris is assumed to be present at the sump screens immediately upon initiation of RHR recirculation. No credit was taken for time to transport while the sump continues to fill due to continued addition of water to the sump resulting from containment spray operation.

9. For testing purposes, twice the inventoried quantity of unqualified labels was assumed to detach and transport to the sump screens. In reality, many of the labels are tightly adhered and many are protected from direct containment spray and likely would remain in place. In the event of detachment, many of these labels would not be transported to the sump screens due to torturous paths between the labels and the screens.

10. Two hundred pounds of latent debris was assumed for testing purposes while the surveyed value was 125 pounds. In addition, the debris was assumed to be 15% fiber per NEI guidance, although the source of fiber in the FNP containment is very limited as FNP is primarily a RMI insulation plant. Very limited amounts of fibrous insulation are installed on the steam generator instrument lines and around the reactor vessel nozzle penetrations. All latent fiber in containment and all other fiber within the break ZOI are assumed to transport to the sump screens.

11. Measured tested screen head losses were increased by 43% to account for uncertainties. These conservatively increased head loss values were used to calculate NPSH margins.

12. FNP does not credit containment pressure above pre-accident pressure for net positive suction head available (NPSHa) calculations. In reality, post LOCA pressures in containment would provide significant NPSH margin above calculated values. Analysis shows that this would add a minimum of 16 feet of NPSHa immediately upon initiation of ECCS recirculation and would increase during the event.

13. A very detailed and conservative calculation is used to determine minimum ECCS sump level. The containment sump level calculations were performed using "stacked" El -2

Enclosure 1 Significant Conservatisms in FNP Emergency Core Cooling System Sump Design, Testing, and Analysis conservatisms. For example, maximum reduction in Refueling Water Storage Tank (RWST) mass due to level instrument uncertainty was assumed even thought this would involve opposing instrument uncertainties; positive on the high end of the instrument range and negative on the low end of the range. In addition, minimum allowable initial water volumes were assumed for both the ECCS accumulators and the RWST. Also, the switch over to recirculation is assumed to occur instantaneously at the RWST low level set points. Operator action time is required for the operator to manually perform the swap over from injection to recirculation mode, during this time, additional inventory is added to the ECCS sump. A realistic value for containment sump level would be at least 6 inches higher than used for NPSH calculations.

14. Testing for FNP's screens was conducted at lower than expected minimum sump levels.

In addition, maximum ECCS pump flows were used for these tests. These tests clearly demonstrated that that FNP screens are not susceptible to air ingestion under worst case LBLOCA or small-break loss-of-coolant accident (SBLOCA) conditions.

15. No credit for'leak-before-break was taken in the FNP sump analysis scenario.

El -3

Joseph M. Farley Nuclear Plant - Units 1 and 2 Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 On March 9, 2009, a Nuclear Regulatory Commission (NRC) Request for Additional Information (RAI) was received regarding the supplemental responses, dated February 28, 2008 and April 29, 2008, to Generic Letter (GL) 2004-02. The Southern Nuclear Operating Company (SNC) response to the requested information is provided as follows.

Head Loss and Vortexinq NRC RAI Question 1:

Provide verification that bore holes were not present in the debris bed during testing. Provide the method used to determine this was the case (e.g., were flows sweeps at the end of testing used for such a determination?).

SNC Response:

Flow sweeps were not used for Joseph M. Farley Nuclear Plant (FNP) testing as these tests were conducted prior to that guidance being issued.

For testing done for high temperature conditions without chemical effects, the surface inspections after the completion of head loss testing did not indicate borehole formation. No evidence of boreholes or other surface anomalies was found. Head loss plots show fairly smooth curves, without jittering, and no boreholes or other surface anomalies credit was taken for the testing. In addition, since the maximum measured debris bed head losses under these conditions was 2.7 inches, differential pressures needed to create boreholes were not present.

The bounding calculated net positive suction head (NPSH) margin condition for FNP is early in the recirculation phase when temperatures are at 212 OF. For NPSH calculations, FNP does not take credit for containment pressure greater than normal pre-event atmospheric pressure. This results in the assumption that the sump water is at saturation conditions. With water temperatures above 212 OF, reduced water viscosity reduces head losses. At 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months following a large-break loss-of-coolant accident (LBLOCA), containment sump temperature will be below 212 °F and will continue to decline throughout the event. With water temperatures below 212 OF, assuming only normal atmospheric pressure in containment, the increasing amount of subcooling of the sump fluid creates substantial available NPSH above that calculated for saturation conditions. Chemical precipitates are not present until sump temperatures are substantially below 200 OF. Impact of these precipitates, even when applied to highly conservative conditions of 200 OF, will not challenge NPSH margins, as is shown in Figure 1. There is reasonable assurance that, over the 30-day LOCA mission time, NPSH margin exists.

E2-1

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 35 30 25 0

20 -

15-10 5-0-

Sump Temperature Figure 1: Sump Subcooling Impact on NPSH Margin (Chemical Effects Head Loss Extended to 200 OF)

For testing done for lower temperatures with chemical precipitates present, post testing inspection of the screens indicated that boreholes likely were present. As discussed above, the FNP NPSH margins at the temperature at which chemical precipitates occur are much greater than the measured non-temperature scaled head loss. The minimum margin of net positive suction head available (NPSHa) above net positive suction head required (NPSHr) for the residual heat removal (RHR) and containment spray (CS) pumps at 140 OF was calculated to be 19.8 feet. The non-temperature scaled debris bed head loss was less than 5 feet. As this is not a limiting case for FNP, using unscaled values presents no challenge to NPSH margins and has no impact on limiting values of margin. The applicable test curve (5M-CS-U2B-40H-CE) is supplied in response to RAI Question 8.

In conclusion, although boreholes were likely present for chemical effects testing, the magnitude of the head loss even without viscosity adjustments does not challenge NPSH margins over the LOCA mission time of 30 days.

(For related information - see the SNC response to NRC RAI Question 18.)

NRC RAI Question 2:

State whether the test results were extrapolated to different flow velocities. If extrapolation of flow was used explain why it is conservative or prototypical.

SNC Response:

The flow rates for the test were scaled per equations which yield the same perforated plate flow velocity for the module test as in the plant installed screen. The calculated circumscribed flow approach velocity for the module test was slightly higher than the plant screen for conservatism.

E2-2

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Scaled full flow velocities equivalent to 4500 gpm and 3400 gpm were used for RHR and CS respectively. Therefore, no extrapolation was used.

NRC RAI Question 3:

Provide verification that the fibrous size distribution used during testing was prototypical or conservative compared to the size distribution predicted by the transport evaluation. (All fiber used in strainer head loss testing should have been prepared as fines)

SNC Response:

SNC did not explicitly quantify the size of the fibrous debris as the testing was done prior to the inspector guidance issuance. However, testing was completed following the Waterford audit and lessons from that audit were applied. Measures were taken to ensure that fiber debris sizing was adequately small, incrementally added, and stirred to allow for representative transport and even deposition upon the screens. The fiber addition procedure and schedule are provided in the response to RAI Question 4. Fiber preparation was done as follows:

Tempmat Blanket was purchased from State Insulation Corporation and cut into squares of approximately 3 inches. Transco Thermal Wrap, purchased from Transco and used to simulate latent fiber, was shredded by the manufacturer following procedures. Both insulation types were shredded 5 more times by Continuum Dynamics Incorporated in a leaf shredder (following procedures) to produce smaller shreds and more individual fibers, see Figure 2. The fibers were shredded and handled to achieve composition consistent with number 2 and 3 sizing per the guidance of NUREG/CR-6808, Table 3-2. Sufficiently small incremental additions, along with a highly stirred test arrangement, adequately ensured representative deposition upon the screen surfaces. The material was visually inspected.

Photos of the fiber distribution for both the Transco and the Tempmat are shown in Figure 2.

Figure 2: Fibers suspended in water over 1/2 inch x 1/2 inch grid showing that fiber is shred as individual fibers and small shreds. Transco is on the left and Tempmat is on the right.

E2-3

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 NRC RAI Question 4:

Provide details of the debris addition procedures used. Include a description of fibrous concentration during debris addition and the method of adding fibrous debris to the test tank.

Provide verification that the debris introduction processes did not result in non-prototypical settling or agglomeration of debris.

SNC Response:

For the testing that did not include chemical precipitates, the following fiber addition procedure was followed:

Water was removed from the tank and added to the particulate debris (coating and silicon carbide) to form slurries. The agitators were turned on and then each particulate debris type in slurry form was added to the tank. When the particulate debris was well mixed in the tank the pump flow was re-initialized. The flow re-initialization time was recorded. Head loss and flow rate data were recorded manually at intervals of approximately 5 minutes. Water temperature was recorded manually at intervals of no more than 30 minutes. After the pumps were started with all of the particulate debris in the tank, wet fiber was added to the tank based on the schedule for each test shown in Table 1. The flow re-initialization time was Time 0 for fiber addition.

E2-4

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 IM-RHR-401H 2M-CS-U2B40H, 3h1-CSU2B-30H 04-CS-UB-.30V TempMatltrans:co TemprAaelrrransco .e. Mat can TenTh*-lTransce Interva Time(n lb Timelmin fib lib) mtmin lb memn 0 Test Start 0 Test Slar 0 Test Star' 0 Tes. Start 1 18 0.07S 0,M18 0 0,073 0.15 20 0.075 0.190 10 0.075 0.,00 2 22 0.073 0.185 9 0.073 0.185 25 0.075 0.190 15 0,075 0.1W0 3 28 0.073 0,185 12 0.073 0.185 30 0.075 0.190 20 0.075 0.190 4 30 0.073 D.O8 15 0.073 0.185 35 0.075 0.190 25 0.075 0.190 5 34 0.073 ORBS5 le 0.073 0.1 P5 4-0 0.075 0.190 30 0.075 0.190 a 38 0.073 0DA85 21 0.073 A.185 45 0.075 0.190 35 0-075 0.190 7 42 0.073 O.S. 24 0.073 0.185 50 0.075 0.190 40 0.075 0.190 6 4a 0.073 0.165 27 0.073 0,165 65 0.075 0.100 45 0.075 0.10-2 52 0.073 0.,. 30 0.073 0. 1 5 60 0.075 0.190 50 0.075 Q.P0 1.3 C 0.07* 06 s 42 0.073 a..185 N.1 0.075 0.1 '0L 570 0.07 0.L0 12 70 0.073 O. B55 44 0.073 0. 185 85 0.075 0. 190 75 0.075 D.AY0P 15 74 0.073 O.vS& 48 0.073 . 185 0C 0.075 0.190 80 0.075 0.1P0 18 78 0.073 DRAM S5 0.073 0.185 85 0.075 0.190 17 82 0.073 0.165 54 0.073 0.185 o0 0.075 0.190 18 88 0.073 0.185 57 0.073 D. 185 19 90 0,073 0QA85 60 0.073 0.1S5 20 83 0.073 0.185 21 -0 0.073 0.185 22 69 0.073 0. 85 23 72 0.073 0.185 24 75 0.073 0. 1P-5 25 78 0.073 0.18 25 _1 0.073 0,165 '............

_ "

27 84 .073 0.185 29 - - 90 0.073 0.18M Table 1: Fiber Addition Schedule E2-5

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 For the chemical effects (CE) test the following procedure and fiber addition schedule was used:

The particulate debris (silicon carbide and coating chips) and fibrous debris were wetted with water removed from the pool. The agitators were turned on. The particulate debris was added to the pool. After the particulate debris was in suspension the pump was turned on.

Wet fiber was added to the pool in accordance with the schedule shown in Table 2.

In all tests, the particulates were added before the fiber and fiber was added in small increments. Thin bed formation would have been noted, if it had occurred, using this method.

Figures 3 through 7 provide additional information.

5M-CS-U2B-40H-CE Time mi TempMat (Ib Transco (Ib 0 Test Start 15 0.171 0.434 30 0.171 0.434 45 0.171 0.434 60 0.171 0.434 75 0.171 0.434 90 0.171 0.434 Table 2: Fiber Addition Schedule NRC RAI Question 5:

[not used]

E2-6

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 NRC RAI Question 6:

Provide the amount of each type of debris added to each test.

SNC Response:

For the tests that did not include chemical precipitates, the following were used:

Debris Test Tempmat Transo Coatiny Silicon Test Tepaso Chips Carbide (Ibs) (Ibs) (Ibs) (Ibs) 4M-CS-U1B-30V 1.28 3.24 13.95 21.56 1M-RHR-40H 1.38 3.51 15.10 23.35 2M-CS-U2B-40H 2.12 5.37 23.13 35.78 3M-CS-U2B-30H 1.13 2.86 12.30 19.01 10 mil thickness Table 3: Debris Matrix For the test including chemical precipitates, the following was used:

Tempmat Transco Coating Silicon Temma (IChipsI Carbide (Ibs) (Ibs) (Ibs) (Ibs) 1.03 2.61 11.23 17.37

'10 mil thickness Table 4: Fibrous and Particulate Debris Test Amount (Ib)

Chemical Surrogate Sodium 'Aluminum 0.64 Silicate Aluminum Oxyhydroxide 85.83 Calcium Phosphate 0.07 Table 5: Chemical Effects Debris E2-7

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 NRC RAI Question 7:

Provide the area of the test strainer for each test.

SNC Response:

Four test articles were used for the FNP modular testing.

Net Perforated Test Flow Rate Test Article ID Ar_..a (.ft2)

A e *ft2Test Number q m 40 x 40 horizontal 179.6 1M-RHR-40H 1017 40 x 40 horizontal 179.6 2M-CS-U2B-40H 1594 30 x 30 horizontal 95.4 3M-CS-U2B-30H 809 30 x 30 vertical 97.3 4M-CS-U1B-30V 925 33 x 48 for CE test 87.2 5M-CS-U2B-40H 666

Net perforated area = gross perforated area minus blockage from internal structures Table 6: Test Articles NRC RAI Question 8:

Provide the test termination criteria and the methodology by which the final head loss values were extrapolated to the emergency core cooling system mission time or some predicted steady state value. Provide enough test data that the extrapolation results can be verified.

SNC Response:

The test termination criteria are as follows:

Each test could be completed by meeting a stabilization (steady state) criterion of less than or equal to a 1% increase in head loss over a 30 minute period. If the head loss was varying up and down then an average head loss was used to determine the termination criterion.

Figures 3 through 7 provide head loss test data for the five FNP tests. Debris addition points and other items of interest are noted on the figures. Figures 8 through 9 provide extrapolations for the five tests.

E2-8

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 9

8 7

6 5

0 4

-J 3

2 1

0 50 100 150 200 250 Time (min)

Figure 3: Head Loss Test 1M-RHR-40H I I.

2M-CS-U2B-40H Test termination criteria met begin to lower water level i

0 50 100 150 200 250 300 Time (min)

Figure 4: Head Loss Test 2M-CS-U2B-40H 3M-CS-U2B-30H Test ternination criteria met begin to low er w ater level.

C0

.-01'

/

7 6 *" " uer udlditiou S 15inammients - . . .. / _

5 4

0 3

-J M2 ]* ' -

C oating D ebris AirIngestion

, .- __Added, agitat o Air Ingestion 1

0 running 0 50 100 150 200 250 300 350 Time (min)

Figure 5: Head Loss Test 3M-CS-U2B-30H E2-9

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 0 50 100 150 200 250 300 350 400 Time (min)

Figure 6: Head Loss Test 4M-CS-UI B-30V 5M-CS-U21B-40H-CE 70 HL Stable, Test Terninated 40 HL SHL Stable Add Chem Batch #3 40o HL Stable. Adid Chain Batch #1 If.Stable, Add Chem Batch #4 l'O -

30 0 it-f Stable Add Chem Batch #2 10-10-200 400 600 800 1000 1200 1400 1600 1800 Coati ng rnd fibrous debris added Time (min)

Figure 7: Head Loss Test 5M-CS-U2B-40H-CE Five modular screen tests were done for FNP. The last 60 minutes of test data before test termination are extrapolated, as shown in Figures 8 and 9. Three of the tests had extrapolated values with negative slopes. Two had slightly positive slopes. The equation of the most positive slope is as follows: Head Loss (inches of H2 0) = 3E-04 x where x is in minutes.

Extrapolation of this over a 30-day LOCA mission time yields an increase of 1.1 feet of additional head loss. This does not challenge NPSH margin at any time during the 30-day LOCA mission time. This is shown in Figure 10. See the response to RAI 18 for a discussion of CE headloss.

E2-10

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Headloss Test 1M-RHR-40H Headloss Test 2M-CS-U2B-40H 60 minute data extrapolated 60 minute data extrapolated 7.8 7.6 18.8 7.4 18.6 7.72 18.4 0 18.2

6.8 18 6.6 17.8 6.4 6.2 9J 17.4 5.8 17,2 17 Minutes Minutes Headloss Test 3M-CS-U2B-30H Headloss Test 4M-CS-U1 B-30V 60 minute data extrapolated 60 minutes data extrapolated HL=3E04x+3.682 ]

42 4,

a(3.4.

o~2 3.2

3. 0 Minutes Minutes Figure 8: Non-Chemical Effects Test Extrapolations Chenical Effects Headloss Test 5M-CS-U2B-40H-CE 60 ninute test data extrapolated 57 56 55 6~ 54 53 10 52 0= 51 50 (0 Cf) to (0 to co to U1J 0 rD 0 Lo to Minutes Figure 9: Chemical Effects Extrapolations E2-1 1

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 35 30 I

25 20 15 U)Cu10 5

0 0 0.5 1.15 2.2 30 t

Sump temp 200 OF Days t

Sump temp 125 OF Figure 10: NPSH Margin Extrapolated Over Mission Time Without Chemical Effects NRC RAI Question 9:

Provide the quantitative margin to flashing through the strainer and describe the methodology used to determine this margin.

SNC Response:

The minimum pressure margin was determined to be 8.3 psi shortly after the initiation of recirculation, increasing monotonically to approximately 17 psi through the end of the analysis at 2 x 106 seconds. The margin to flashing through the screen was taken as the minimum difference between the GOTHIC calculated containment vapor pressure and the saturation pressure at the calculated sump temperature at any time for the design basis containment response.

NRC RAI Question 10:

Provide the minimum strainer submergence at the onset of recirculation considering both the small and large-break loss of coolant cases. Provide a vortexing and air entrainment evaluation for the case at the minimum submergence. In general, the small-break loss of coolant accident provides the limiting submergence and this occurs at the onset of low head safety injection recirculation while the containment spray pumps are still supplied by the refueling water storage tank. Limiting pool level is determined by the minimum reactor coolant system contribution to E2-12

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 the pool and the refueling water storage tank volume injected at the time that the low level alarm occurs (alerting the operator to align the residual heat removal pumps to recirculation)

SNC Response:

The limiting case for screen submergence upon initiation of recirculation for both large and small break scenarios is for RHR. CS screens are well covered for both small and large break scenarios.

For the LBLOCA the calculated minimum emergency core cooling system (ECCS) sump level is 56 inches above the containment floor at initiation of RHR recirculation. The RHR screens are 44.75 inches in height. Therefore, expected screen coverage is approximately 11 inches. The CS screens have greater coverage than RHR screens due to increased sump inventory upon initiation of CS recirculation. CS screens are fully covered under all recirculation scenarios.

All scaled full flow screen testing was conducted with 3.5 +/- 0.5 inches of water level above the screen with no vortexing or air ingestion indicated for either CS or RHR tests.

As CS approach velocities are greater than those of RHR, CS screen testing was used to determine the point at which air ingestion occurs. Following test completion criteria, water level was lowered for two of the full flow CS tests until air ingestion occurred. One test was scaled for approximately 173% of full RHR flow while the second was scaled for approximately 165% of full RHR flow. In the higher flow case, air ingestion was observed at 39 7/8 inches above the tank floor (approximately 5 inches below screen top). Minimum calculated LBLOCA level at initiation of recirculation is 56 inches. Minimum long term LBLOCA level is 54 inches.

Therefore it is clearly demonstrated that air ingestion following the LBLOCA event will not occur.

The second air ingestion test was run with scaled flow equivalent to 165% of maximum RHR flow of 4500 gpm. Maximum small-break loss-of-coolant accident (SBLOCA) flow is 675 gpm per screen. As post SBLOCA reactor coolant system (RCS) pressure is above the shut-off head of the RHR pumps, maximum screen flow is dictated by the high head safety injection system (HHSI) flow. The maximum capacity of these pumps is 675 gpm. Therefore, this air ingestion test was run at a scaled flow rate approximately 11 times the maximumSBLOCA RHR flow. For this test air ingestion occurred at a water level of 31 3/8 inches, which for this arrangement, was approximately 9 inches below the top of the screen. Minimum calculated SBLOCA sump level is 39.3 inches at the initiation of RHR recirculation.

SBLOCA minimum level was calculated using highly conservative assumptions. It is assumed that no water is contributed from the RCS and minimum water level occurs at the onset of low head safety injection recirculation while the CS pumps are still supplied by the refueling water storage tank. Minimum refueling water storage tank volume inventory is assumed to be injected at the time that the low level alarm occurs, thus alerting the operator to align the RHR pumps to recirculation. As CS flow will continue to fill the ECCS sump from the RWST, RHR screen uncoverage time would be limited to minutes under SBLOCA conditions. Under LBLOCA conditions the RHR screens are fully covered at all times.

Therefore, it is clearly demonstrated that, under SBLOCA conditions, air ingestion will not occur.

E2-13

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 NRC RAI Question 11:

Provide an explanation of the difference between the head loss shown in Figure 3.f.13-1 [SNC letter dated February 28, 2008], typical head loss plot for test 2M-CS-U2B-40H, and the head loss provided in Figure 3.f.10-1 [SNC letter dated February 28, 2008].

SNC Response:

The head losses shown in Figure 3.f.10-1 is the head loss across debris beds. It is the difference between measured clean screen head loss and loaded screen head loss and is a compilation of test results.

The head loss shown in Figure 3.f.13-1 is the head loss measured across a single test article which includes debris bed, screen, and internal losses.

NRC RAI Question 12:

Provide an explanation for the behavior of the head loss shown in Figure 3.f.13-1 [SNC letter dated February 28, 2008], typical head loss plot for test 2M-CS-U2B-40H, especially at the end after the test termination criteria have been met, water level is being lowered, and head loss increases significantly.

SNC Response:

The response near the end of the head loss plot, while water level is being lowered, is not prototypical for a LBLOCA condition at FNP. During the early portion of test, flow is initiated and adjusted for performing the test and to determine clean screen losses. The pump is then shut down and preparations are made to add debris. The pump is restarted and debris is then added in accordance with the testing program.

The behavior at the end of the test is not prototypical of a LBLOCA event at FNP as the water level was reduced until air ingestion occurred. This was done to determine water level at which air ingestion would occur.

The tank was filled with water to 3.5 + 0.5 inches above the test article. After the test met the termination criteria, the pump and data acquisition system were kept running while the tank slowly drained. Air entrainment became obvious at a water height (measured from the tank floor) of 39 7/8 inches. This resulted in the higher head losses seen at the end of the test at 5 inches below the top of the screen. Minimum LBLOCA sump level at FNP is 54 inches above the floor.

E2-14

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Coatings NRC RAI Question 13:

The supplemental response [SNC letter dated February 28, 2008] states on page 53 that the transport fraction for qualified coatings is 100%. However, on page 29, Table 3e-1 indicates that for the 4D zone of influence (ZOI) failed coatings modeled as chips, a transport fraction less than one is assumed. Please clarify this apparent contradiction and describe the methodology and technical basis for the reduction in debris transport if a transport fraction less than one is assumed for the failed coating chips. If test data was used as part of the basis for the transport reduction, please compare the properties of the failed coatings at Farley Nuclear Plant (FNP) to the coating chips that were used for the transport testing.

SNC response:

A reduction in qualified coating modeled as chips was made as indicated in the Table 3e-1.

The previously reported 100% transport fraction for qualified coatings was in error.

The transport of coatings as chips is analyzed by dividing the containment into sections and deciding what fraction of the debris in that section transports, based on Computational Fluid Dynamics (CFD) determined pool velocities and coating chip transport data. The coating chip transport data was taken from NUREG/CR-6916. The CFD data was used to determine in what percentage of each containment section the pool velocity is greater than the chip transport velocity. The percentage of coating debris that was generated in each section of containment that transports as chips to the screens was set equal to the percentage of that section of containment that had a pool velocity greater than the chip transport velocity.

NUREG/CR-6916 guidance was used for coatings considered as chips transport. Coatings type E2 was used for coating transport analysis. A comparison of FNP coating with the NUREG/CR-6916 is provided for information in Tables 7 and 8. FNP coatings have both higher density and thickness than the E2 coatings, therefore, it was chosen as conservatively representative in the transport analysis.

E2-15

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Coating Primer Topcoat 3 Thickness (mil) 3 3 name density (g/cm ) name density (g/cm ) Unit 1 / Unit 2 Steel Ameron, Dimetcoat 6 3.15 Amercoat 66 or 90 2.60 or 2.58 14.3 / 15 Concrete Ameron, NU-KLAD 10OAA 1.95 Amercoat 66 or 90 2.60 or 2.58 18.7 / 11 Table 7: FNP Coating Characteristics Coating Primer Topcoat Thickness (mil) Density (g/cm 3)

ALK Ameron, Amercoat 5450 (low density alkyd) 2.2 1.00 ZE Ameron, Dimetcoat 6 Ameron, Amercoat 90 7.1 2.58 E2 Carboline, Carboguard 890N 8.6 (two coats) 1.78 E6 Carboline, Carboguard 890N 23.0 (six coats) 1.77 E3C Keeler Long, KL4129 Keeler Long, KLD1 25.0 1.85 and KL6548S I I I Table 8: NRC Tested Coatings - NUREG/CR-6916 NRC RAI Question 14:

Please provide the amounts of qualified and unqualified coatings assumed to fail as chips and the amounts of qualified and unqualified coatings assumed to fail as particulate that were used in the strainer qualification tests. If chips were used in testing, then please justify treating qualified and unqualified coating debris as chips, given that page 37 of the supplemental response indicates that a thin bed is expected to be formed during strainer operation. From the NRC review guidance and safety evaluation, ifthere is a thin bed present, all coating debris should be treated as particulate that would transport to the sump, unless proper justification and/or data are provided. Proper justification for the treatment of coatings as chips could be provided by verifying that testing has shown that the coatings will fail in a manner that their properties match the surrogates used in testing or justifying that a filtering bed would not actually cover the plant strainer.

SNC Response:

Under a single train failure scenario, FNP debris loading could approach a thickness value that indicates thin bed formation could result. However, FNP screens are not strictly flat plates in that there is a wire screen mesh mounting on them that is designed to prevent thin bed formation. In addition, the plates are largely vertically oriented, which may have an impact on the formation of a thin bed. Although fiber thickness that under some conditions may indicate a thin bed is present, assuming a single train of both RHR and CS fails to operate, the behavior of the FNP screen tests indicate that given FNP specific conditions the high head losses associated with thin bed formation did not occur. A series of FNP specific tests were conducted using varying amounts of epoxy coating flakes and 10 pm silicon carbide particles to simulate a variety of plant LOCA debris-generation scenarios. The test results clearly indicated that higher head losses were obtained with coating chips rather than with particulates.

E2-16

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Test results are given below:

Transco Silicon Coating Inorganic Temp Q HL Tempmat Test (emp Tibm) Carbide Chips Zinc Tem p (in.

(Ibm) (Ibm) (Ibm) (°F) (gpm) H20) 1-OF- 59 239 19.2 100SC-2X100S_________-2X______

0.24 0.89 123.1 0 47.9 63 59 198 29 18.3 19.2 1 A-50F-68 0.89 18.95 52.1 479 68 24 234 0.

20.8 1SC-2X 0.24 OSC-2X 70 198 18.6 2-5O0F-60 47.9 60 29 239 3.

33.0 50F- 0.24 0.89 71.1 52.1 50SC-2X 65 196 41.5 3-100OF-OSC- 0.24 0.89 18.92 104.2 47.9 62 103 OSC-2X 186 Table 9: Test Matrix and Result Summary Based upon the results of the tests and considering that the likely screen fiber loading would be lower than the tested values as the likelihood a single train of both CS and RHR failing is very low, the more conservative approach is to assume that the coatings fail primarily as chips. If both trains of RHR and CS were to operate, the fiber loading is clearly below "thin bed quantities."

Debris Source Term NRC RAI Question 15:

How will your containment cleanliness and foreign material exclusion programs assure that latent debris in containment will be-controlled and monitored to be maintained below the amounts and characterization assumed in the emergency core cooling system strainer design?

Will latent debris sampling become an ongoing program?

SNC Response:

An enhanced containment cleaning program will be performed on a three-outage basis. This enhanced cleaning program will focus on removal of latent debris. -Latent debris in containment consists of dirt, hair and other particles or fiber type debris that generally is carried into containment on personnel / equipment during outage maintenance periods. The amount of latent debris inside FNP containment has been sampled and the results were provided as input in the containment sump screen design calculations. The values used in these calculations included a substantial margin. This enhanced cleaning program, performed on a three-outage bases, provides reasonable assurance that latent debris in containment will remain below the conservative values used in the containment sump screen design calculations.

E2-17

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Chemical Effects NRC RAI Question 16:

The value given for the mass of aluminum oxyhydroxide in your April 29, 2008 letter (page 5 of

11) is 988 Ibs, but the mass stated in your February 28, 2008 letter (page 88 of 101) is 729 lbs.

Please explain the discrepancy in these values and identify which amount of aluminum oxyhydroxide was used as a basis for the chemical effects head loss testing. If less than the predicted amount of chemical precipitate was used in head loss testing, discuss how this affects the interpretation of the test results.

SNC Response:

Subsequent to the submission of the February 28, 2008 letter, a review of the chemical quantities was conducted assuming longer containment spray run times. This resulted in higher aluminum oxyhydroxide values. The higher of the two values was used for testing.

NRC RAI Question 17:

The WCAP-16530-NP chemical spreadsheet predictions show most of the Farley plant-specific precipitate is aluminum oxyhydroxide. Your April 29, 2008 letter provides the one-hour precipitate settlement data for calcium phosphate and sodium aluminum silicate. Please provide the one-hour precipitate settlement data for the aluminum oxyhydroxide precipitate used in head loss testing.

SNC Response:

Testing of the aluminum oxyhydroxide turbidity indicated 98% of the solution in the graduated cylinder was turbid at one-hour. This indicates that less than 90% of the solution settled. This meets the WCAP acceptance criteria.

NRC RAI Question 18:

Based on bench testing performed at Alion Science and Technology, Farley assumes aluminum-based precipitates will not form at temperatures above 140 OF. Please provide the experimental data that supports that assumption.

SNC Response:

FNP specific bench top experiments performed by ALION Science & Technology identified a visible precipitate occurring on or about day 17 while temperatures were reduced from 200 OF to 140 OF. At day 17, the FNP sump temperature is calculated to be approximately 130 OF. FNP post LOCA sump temperature is reduced to below 200 OF within about 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months following a LOCA event.

E2-18

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 Additionally, other utility test data, such as Three Mile Island, indicate flow loop testing done with similar sump pH as at FNP, demonstrated that no appreciable head loss increase was observed due to chemical effects until temperature was below 140 OF and also demonstrated that head loss due to chemical effects is time dependent. This test demonstrated that increased CE head loss were small at 140 OF and then showed an increase with both time and temperature reduction.

With recognition that NPSHa increases dramatically due to normal atmospheric pressure as sump temperatures drop below 212 OF, and based upon the test information, applying full chemical precipitate head loss at 140 OF for FNP is appropriate.

To demonstrate that CE head loss does not challenge FNP NPSH margins even if precipitates are assumed to come out of solution at much higher temperatures, Figures 11 and 12 illustrate the impact of CE head loss at FNP. The head loss used is conservatively not adjusted for reduced viscosity at higher temperatures and is also conservatively increased by 43% over the measured value to address testing uncertainties.

35 30 25 20o 5

0 Sump Temperature Figure 11: Sump Subcooling Impact on NPSH Margin (Chemical Effects Head Loss Extended to 200 'F)

E2-19

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 35 30 25 20 L

I 15 10 5-0 -I 0.5 1.15 2.2 30 t

Sump temp 200 'F Days t Sump temp 125 'F Figure 12: Sump Subcooling Impact on NPSH Margin (plotted vs. time)

The post LOCA sump temperature is reduced to below 200 OF within about 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months following a LOCA event.

Licensing Basis NRC RAI Question 19:

Please provide a general description and schedule of the changes to the plant licensing basis that were made as a result of the evaluations and plant modifications made to resolve GSI-191 for FNP.

SNC Response:

In letter dated February 28, 2008, SNC stated in response to NRC issue 3.p that the FNP licensing basis was changed in accordance with the requirements of 10 CFR 50.71. FNP Final Safety Analysis Report, Appendix 6D describes the new Emergency Core Cooling System screens installed to address GL 2004-02. A description of the new Unit 1 and 2 screens, including size, assembly details, and figures was added. A summary approach used to size the new screens using the guidance of NEI 04-07 and the containment walk down used to confirm installed installation is included. Pipe break characterization, debris generation, latent debris accumulation and debris transport to the containment sump is described. Residual Heat Removal Pump and Containment Spray Pump head loss as a result of debris accumulation, including the vortexing analysis is included. The sump structural analysis, including a description of the passive screen, is included. The upstream effects of debris accumulation, downstream effects associated with any debris bypass, and chemical effects testing are also included. Tables for debris generation ZOI, LOCA generated insulation debris inside ZOI, E2-20

Enclosure 2 Response to Request for Additional Information Regarding Generic Letter 2004-02 debris generated from coatings based on ZOI = 4D, latent and foreign material debris used in the analysis, and summary of debris generated and transported to the screen modules is included.

E2-21

Joseph M. Farley Nuclear Plant - Units 1 and 2 Enclosure 3 Commitment Table

Enclosure 3 Commitment Table Type Scheduled Commitment One-Time Continuing Completion Date Action Compliance Revise the appropriate procedure(s) to implement an enhanced containment cleaning X November 15, 2009 program to address control of latent debris.

E3-1