ML20247F183

Design Calculation for OCNGS ECCS Strainer Modification

ML20247F183
Person / Time
Site:	Oyster Creek
Issue date:	05/05/1998
From:	GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20247F180	List: ML20247F176 ML20247F183 ML20247F189
References
NUDOCS 9805190189
Download: ML20247F183 (35)

Text

-- - - - - - - - - _ - - _ - _ - - - - _ _ _ - - - - _ _ - _ _ ___-

e i a

(

Attachment II Design Calculation for the '

l Oyster Creek Nuclear Generating Station  ;

i l

ECCS Strainer Modification  !

Table of Contents l

1. Introduction 2 l

2. References 2

3. DrywellInsulation Debris Sources 3 3.1. Pipe Break Imcations 3 3.2. Break Jet Zone Of Influence 5 3.3. Other Drywell Debris Sources 7 3.3.1. Transient Debris Sources 7 l 3.3.2. Fixed Debris Sources 8 )

3.3.2.1. Drywell Coating 9 3.3.2.2. Paint Chips 9 3.3.3. Latent Debris Sources 9 3.3.4. Additional Operational Debris 9 3.3.5. Summary 9

4. Drywell Debris Transport 10

5. Suppression Pool Debris 11

6. Suppression Pool Transport 14

7. Core Spray and Containment Spray NPSH Evaluation 14 7.1. System Description 15 7.1.1. Containment Spray 16 1 7.1.2. Core Spray 17 I 7.2. Changes to the Current ECCS NPSH evaluations 19 7.2.1. Containment Response to a LOCA 19 7.2.1.1. Large Break LOCA 19 7.2.1.2. Small Break LOCA 21

8. EOP Actions That Impact Suction Strainer Performance 22 8.1. Primary Containment Pressure Control 22  !

8.2. Drywell Temperature Control 23 8.3. Torus Temperature Control 23 j 8.4. NPSH Limit Curves 24

9. Suction Strainer IIead Loss 25 9.1. Containment Overpressure 26 9.2. Flow Management 29 ,

10. Analytical Results 30 l

11. Strainer Design 35 1 1

9805190189 980505 PDR ADOCK 05000219 .

O PDR l I

k _ _ ._ - . . . _ _ _ . _ . _ _ _ . . . . _ . _ . _ _ _ ___ ______

i i 4

a 1940-98-20124 Attachment 11 Page 2 1.0 Introduction The purpose of this document is to justify the use of containment overpressure as an analytical input to the calculation of the NPSH available to the core spray pumps. Overpressure is i defmed to mean atmospheric pressure greater than 14.7 psia. The current evaluation of the core spray NPSH does not take credit for conutinment pressure in excess of 14.7 psia. For example, if the containment pressure is predicted to be 20 psia, the NPSH calculation would assume 14.7 psia. During the next refueling outage, scheduled for the fall of 1998, new suction strainers will be installed within the suppression pool to address concerns associated with debris blockage of the strainers. It has been postulated that the jet forces from a LOCA j could damage insulation and transport the debris from the drywell to the suppression pool.

Once within the suppression pool, the insulation fiber is drawn onto the suction strainers by the operation of the core spray and containment spray pumps. The insulation will form a filter on the strainers and begin to capture small particulate such as iron oxide (sludge), dirt, and paint chips. The accumulated affect of this debris would produce a significant pressure drop across i the existing suction strainers. With the strainers blocked, the pumps drawing suction from J.:

suppression pool wil! cavitate.

l To ensure that the core and containment spray systems function as designed, the suction strainers will be modified to accommodate the postulated debris loads. The design of the new strainers will include consideration of containment pressure. The pressure will be conservatively established to ensure that margin is available within the design values. The use of containment pressure is a change to the current plant design basis. This document will outline the operation of those systems that take suction flow from the suppression pool during a design basis accident with emphasis given to the EOP's. The use of overpressure will be shown to be a reasonable physical assumption and consistent with the plant's operational procedures.

2.0 References

1. GE Nuclear NEDO-32686 Class 1 November 1996 DRF A74-00004 ' Utility Resolution Guidance for ECCS Suction Strainer Biockage' prepared by Boiling Water Reactor Owners' Group

2. RG 1.82 Rev. 2, "%ter Sources for Long Term Recirculation C% ling Following a Loss of Coolant Accident"

3. Blockage Computer Code 4 NRC Bulletin 96-03, " Potential Plugging of Emergency Core Cooling Suction Strainers by Debris in Boiling Water Reactors" l- 5. NRC Bulletin 95-02, ' Unexpected Clogging of a Residual Heat Removal Pump Strainer While Operating in Suppression Pool Cooling Mode"

6. NRC Bulletin 93-02, " Debris Plugging of Emerget.cy Core Cooling Suction Strainers"

7. Supplement I to NRC Bulletin 93-02, " Debris Plugging of Emergency Core Cooling Suction Strainers" t---- - _ _ _ _ _ _ _ _ - - - - - - - - _ . _ _ _ _ _ _ - . - - - - - - - - - _ _ ___

tL ,

4 i

1940-98-20124 Attachment 11 Page 3

8. Generic letter 85-22, " Potential for Loss of Post-LOCA Recirculation Capability due to

' Insulation Debris Blockage"  ;

j 9. FSAR Section 3.6, Appendix 3.6 B " Analysis of Pipe Breaks Inside Containment" l

10. GPUN Calculation . C-1302-241-5450-074 Rev. 0 ' Core Spray NPSH Assessment'

11. GPUN Calculation C-1302-241-5450-080 Rev.1 ' Calculation cf Torus Temperature as  ;

NPSH Input'

{

12. GPUN Calculation C-1302-241-e610-081 Rev. 0 ' Suction Strainer Debris Generation and Transport'

13. GPUN Calculation C-1302-240-5450-006 Rev. 0 'OC Simulator model of Suction Strainer Blockage' l 14. SE-000187-001 Rev.1 " Evaluation of Blistered Torus Coating"

15. NEDC-3131462P " Oyster Creek Nuclear Generating Station Safer /Corecool/

GESTR-LOCA Loss-of-Coolant Accident Analysis' l 3.0' Drywell Insulation Debris Sources The drywell represents a significant source of debris following a loss of coolant accident (LOCA). For Oyster Creek, the primary debris type is found in the unjacketed NUKON fiber insulation blankets, which are wrapped around the process piping as well as the reactor vessel.

In addition to drywell debris, this evaluation accounts for other debris sources such as dirt, dust, and coating debris.

3.1 Pipe Break Locations The first step when evaluating debris generation is to identify the break locations. From

, these locations, the LOCA zones ofinfluence are determined and the identified targets  ;

! evaluated for debris generation. Each location will produce a different amount of i debris and the strainer design must accommodate any break within the licensing basis I of the plant. Oyster Creek has identified within the FSAR (reference 9) a set of breaks defined using Regulatory Guide 1.46 methods.

1 The break locations used in LOCA debris-generation evaluations are most properly l identified as pipe locations with the greatest potential for failure under cyclic-loading -

J conditions. This methodology is detailed in the FSAR and is used for all pipe break, pipe whip, and jet impingement calculations at the Oyster Creek Nuclear Generating Station.

I 4

r

. 1940-98-20124 Attachment II t

Page 4 Break locations are selected for use in determining appropriate measures for the protection of structures, systems, and components important to safety that are within the reactor containment. Regulatory Guide 1.46 provides an NRC accepted approach for identification of break locations which need to be addressed in order to resolve concerns about pipe whip. Regulatory Guide 1.46 does not suggest that the break location are based upon the effect, but rather that the selection of break locations be l based upon the likelihood of occurrence.

The approach used in Regulatory Guide 1.46 is also applicable to the current issue of debris generation and suction strainer clogging. The intent of this debris generation assessment is to quantify the impacts of breaks on systems and components that are important to safety. In the case of the suction strainer plugging scenario, this represents the impact on suction strainers used by systems that provide emergency core cooling. The plan to resolve the suction strainer clogging concern establishes goals that are consistent with those outlined in Regulatory Guide 1.46. Therefore, for the evaluation of suction strainer clogging, it is appropriate to evaluate breaks at high stress and fatigue locations as described in the Branch Technical Position MEB 3-1.

This approach is endorsed within the URG (reference 1) and provides a symmetric distribution of break points along virtually all systems included within the containment.

Consequently, the jet zones of influence for an assumed break target much of the i drywell. This approach encompasses 115 major break locations within the drywell.

The breaks include the following systems:

1. Main Steam Lines (8 break locations)

2. Core Spray Lines (24 break locations)

3. Feedwater Lines (22 break locations)

4. Recirculation Lines'(20 break locations)

5. Shutdown Cooling Lines (8 break locations)

6. Isolation Condenser Lines (16 break locations)

7. Cleanup Lines (17 break locations)

The variety of break sizes and locations is sufficient to ensure acceptable ECCS performance under the most severe conditions.

The break locations also include breaks within the bio-shield wall. Thus, the vessel insulation is included in the overall evaluation of the suction strainer. A summary of the break locations is provided in Reference 9.

l 1

___.__________________________.__________d

1 f 4

• 1940-98-20124 p

Attachment II Page 5 I

3.2 Break Jet Zone Of Influence -

The break jet zone of influence is a function of the pipe size (ID), the type of pipe restraint and the insulation type. For this evaluation the break specific pipe diameters are used along with the assumption that the piping is unrestrained. The lack of pipe restraints is appropriate for Oyster Creek and produces a greater amount of debris than restrained pipe failures. In addition, all breaks are assumed to produce jets from botti sides of the break. This will double the zone of influence for those breaks such as the  ;

main steam lines and core spray lines that connect with the reactor vessel from only one '

side.

l The insulation type is unjacketed NUKON which is identified within the owners group .

evaluations. In addition, trace amounts of other insulation types are included within the drywell. The table below provides a summary of the different types of insulation and their quantities. These trace amounts of insulation will not be handled within the zone of influence section of this evaluation. The fiberglass and asbestos will be treated as NUKON. The rubber will not be considered due to its small volume. Reflective i Metallic Insulation (RMI) existing on the upper portion of the reactor vessel was also considered in the new strainer design.

I Insulation Type Volume of Pipe Insulation Within the Drywell

! Unjacketed NUKON 1480 ft' L Fiberglass- 26 ft' i Asbestos 8ft' i Rubber 96 ft'

.l The simplified BWROG approach (reference 1) is used to estimate the break zone of influence using the assumptions listed above. The approach conservatively identifies a sphere that extends from the break point having an outer radius defined by a dynamic pressure surface sufficient to destroy the insulation type being evaluated.

This approach is referred to as method 3 within the URG (reference 1).

l-L

1 f 4

1940-98-20124 Attachment 11 Page 6 The zone of influence is calculated for each size pipe included in the evaluation.

The table below summarizes the resulting zone of destruction volume (reference 12) and the volume size relative to the drywell free space volume of 180,000 ft'.

Pipe Size Zone of Destruction Relative Volume Volume (Drywell 180,000ft')

26" OD 42,571 ft' 23.7 %

24" OD 37,664 ft' 20.9 %

18" OD 15,889.5 ft' 8.83 %

14" OD 7,476 ft' 4.2%

10" OD 2,724.5 ft' 1.5%

6" OD 58a.5ft' .33%

4" OD 174.4 ft' .097 %

3" OD 73.5 ft' .041 %

2" OD 21.8 ft' .012 %

l As can be seen from the table, the large pipe sizes (e.g. recirculation loops, main steam) will introduce the largest zone of influence into the containment. Any insulation within the sphere is assumed to become potentially transportable debris. A summary of the amoum of debris for each break location is provided in Reference 12. This includes any reactor vessel insulatlan for those breaks lecated at nozzles (i.e. wiJiin the bio-shield).

i L ' An assessment of the large break LOCA indicates that the entire volume of reactor vessel fiber is within the break zone of a recirculation loop failure at the vessel nozzle (Reference 12). This debris is included in the debris loading assessment.

l c _ ___________-____-______________a

O o 1940-98-20124 Attachment II l Page 7 3.3 Other Drywell Debris Sources This section addresses sources of drywell debris other than that associated with pipe insulation. The URG (reference 1) classifies three types of other drywell debris.

Transient debris - Transient debris is non-permanent plant material brought into the - l l drywell during an outage. This debris is controlled by FME and housekeeping l programs.

Fixed debris - Fixed debris is material that is part of the permanent plant. This fixed material becomes a debris source only after exposure to the effects of a LOCA. Such debris will include paint, concrete dust, etc.

• Latent debris - Latent debris is debris that would not be present until later in the LOCA event following prolonged exposure to a LOCA environment. For example, unqualified coatings may fail and flake offlater into the event after prolonged exposure to harsh environmental conditions.

. 3.3.1 Transient Debris Sources As is described in the URG (reference 1), transient debris is material left in the '

drywell. FME and housekeeping at Oyster Creek is controlled in accordance with the following station procedures:

. Procedure 119, " Housekeeping" e Procedure 119.3, ." Foreign Material Exclusion" During extended outage periods, the drywell is open for general access with no formal I

! tool and equipment accountability program in place. To ensure that no debris is left in j the drywell and none is introduced into the wetwell via the vent downcomers the i following measures are taken:

• After the initial entry and prior to allowing general access, special covers are placed over the vent downcomers to prevent objects from entering or being dropped into the wetwell. The covers are subsequently removed prior to drywell close-out. The removal and installation of the covers is controlled in accordance with Procedure 233, "Drywell Access and Control".

• Several days prior to closeout of the Drywell, plant personnel commence a series of inspections that include a thorough search for debris generated during the outage period. All horizontal (i.e., floors, cable trays, etc.) and equipment surfaces are inspected for the presence of trash. All trash is properly removed. The final closeout inspection is conducted by a licensed (SRO) operations supervisor and documented in accordance with Procedure 233.

. 4

.- 1940-98-20124 Attachment II Page 8 l During a refueling outage, the drywell is cleaned (debris removed) periodically to prevent excessive buildup of loose debris.

The BWROG position is that with an effective FME and housekeeping program, the amount of dust and dirt as well as ablated concrete should not exceed 150 lb. Based upon the above discussion, it is the position of GPU Nuclear that Oyster Creek has l adequate programs in place to allow the use of this value in the strainer design. )

I l

3.3.2 Fixed Debris Sources As part of the strainer design, fixed debris sources in the drywell other than from the pipe insulation previously described, are considered. A review of the potential sources of fixed debris reveals the following:

1. The Oyster Creek Containment does not contain the following:

e fabric equipment covers

)

e cloth bags used to hold equipment left in the drywell e fire hoses e ropes e ventilation system filters (Note: filters are used on the Drywell Recirculation fans only during refueling outages and are removed prior to plant partup) e cloth

• thermal insulation (other san piping insulation which is addressed separately)  !

e tape e plastic (e.g., plastic sheeting or laminates covering signs) l

• rust from unpainted steel surfaces - the Oyster Creek containment is operated as an inert environment with no more than 4% O2 when the plant is on line. ,

Consequently, there is little or no rust on the un-coated surfaces inside l containment. However, as a conservative estimate 50 lbs of rust is assumed in the design calculation.  !

' 2. Equipment labeling consists of labels made of ceramic layer on a steel plate. They  !

are attached with stainless steel braided cable. The labels are designed to withstand  !

LOCA conditions. Materials and attachment methods were intentionally selected to withstand LOCA conditions. As no significant amount of debris from labels is expected, O lbs is assumed in the design.

I

3. Paper based products maybe contained in the drywell. At any time, based on plant material conditions, equipment outage tags, (danger or information) or temporary modification tags may be hung. The number of tags, is minimized. Normally, l there are no equipment control tags in the drywell during plant operation. A conservative estimate would be a tot .: of 10 paper tags being in the drywell. The tags are approximately 3 in wide and 5 in long and are made of heavy paper.

l 1940-98-20124 Attachment 11 f Page 9 No temporary signs are allowed to remain in the drywell on closeout. As these tags would weigh well below I lb, O lbs is assumed in the design.

3.3.2.1 Drywell Coating The drywell is coated with Inorganic Zinc Primer (Zinc Silicate Complex) which is qualified for a LOCA environment. This type of coating is identified in the Bechtel report referenced in the URG (reference 1). The Bechtel report conservatively estimates the mass of coating released following a LOCA jet impingement to be 47 lbs.

Since the coating is qualified, there will be no additional flaking due to the harsh environmental conditions following the reactor vessel blowdown. GPU Nuclear is participating in the BWROG coating committee and the issues are monitored as part of the strainer design effort. There is not any reason to alter the assumed coating debris.

3.3.2.2 Paint Chips There is little or no paint in the drywell other than the coatings, which are addressed separately. However, to account for paint located on valve motor operators as well as recirculation pump and motor bodies a conservative value will be assumed for paint chips.

3.3.3 Latent Debris Sources This category includes unqualified coatings and adhesive backed labels. All coating used within the Oyster Creek drywell are LOCA qualified as such there is no source term for unqualified coatings to be included as part of the strainer design load. GPU Nuclear is participating in the BWROG coating committee and the issues are monitored as part of the strainer design effort. There is no justification to alter this assumption.

3.3.4 Additional Operational Debris The BWROG URG recommends (does not require) that the utility consider the inclusion of unexpected debris such as rags which may be left behind following an outage. The quality and evolving nature (lessons learned assessments) of the Oyster Creek FME programs are sufficient to ensure that additional operational debris will not ,

contribute to the strainer debris load.

3.3.5 Summary DW Coating Inorganic Zinc l 47 lbs ll Paint Chips (DW equipmer t) l 40 lbs l Paint Chips (wetwell coating) l 10 lbs l Dust / Dirt / Concrete l 150 lbs l Rust l 50 lbs l Iron Oxide, Wetwell (See Section 5.0) l 300 lbs l Fiber (Nukon) l 140 ft' l

f f 1940-98-20124 Attachment 11 s Page 10 4.0 Drywell Debris Transport i I

For the fibrous insulation material, a transport factor will be applied when establishing a strainer debris load. All other drywell debris types are assumed to be 100% transportable.

The BWROG testing has demonstrated that the fibrous debris generated by the jet forces will have a variety of sizes from large intact blankets to small fine fiber material. The small fine fiber material will completely transport while the larger material will be less likely to transport to the wetwell. ' This is especially true if there is a grating between the insulation and the suppression system vents. The transport factors obtained from the testing includes the effect of core spray and containment spray flow on the fiber material which was not immediately transported during the reactor vessel blow-down portion of the accident. The primary transport mechanism following the vessel discharge is the erosion of debris caught in the flow exiting the vessel break location.

1 It is the position of the BWROG that the conservative nature of the erosion factor sufficiently covers any containment spray effect. Specifically, the erosion of the debris occurs when the '

fluid pours out of the reactor vessel from core spray injection. The entire population of large debris is assumed to be subjected to the erosion effect. However, it is 'very unlikely that all of the debris is located in the small area where the fluid spills out of the reactor vessel. The j violent nature of the accident will likely scatter the debris throughout the containment, and as  ;

such sending most of it away from the spill over location for the core spray flow exiting the  !

vessel.

The transport factors (reference 1) to be applied to the insulation in the Oyster Creek

, evaluation are as follows. 1 Above the lowest grating (23ft elevation) - 28%

Below the lowest grating (23ft elevation) - 78%

When evaluating the individual break locations the zone of influence was assessed to determine if the affected insulation was above or below the grating. For simplicity, a conservative approach was used to assess the insulation location. If the zone ofinfluence extended

! significantly below the lowest grating elevation the entire insulation volume is assumed to be located below the lowest grating and is subjected to the higher transport factor. The

- recirculation loops are the one exception to this rule where a conservative averaging approach (reference 12) is used to establish a transport factor for those break locations where the ZOI extends below the grating. The highest debris transported for each piping system considered is

1. provided on the next page in tabular form.

. . I 1940-98-20124 Attachment II Page 11 l

i I S3ste m tireak 10 Transport l Tran por1ed i Vactor V ser I Recirculation l PSA004 l 28 % 140ft' l Feedwater l FW2021 l 28 % l 94ft' l lSDC l NU3047 l 28 % l 25ft' l l Isolation Condenser l NE5%3 l 28 % l 84ft' l l Cleanup l ND1073 l 78 % l 23ft' l l Main Steam South l MSS 084 ,j 28 % l 93ft' l

[ Core Spray l NZ3096 l 28 % l 82ft' l 5.0 Suppression Pool Debris There are several sources of debris within the suppression pool. Included in these are:

1. Iron oxide sludge

2. Suppression Pool Coating

3. Rust

4. Outage Debris

1. The amount of iron oxide sludge has been measured for Oyster Creek and documented in reference 12. The results presented in reference 12 indicate that the sludge generation rate is less than 50 lbs/yr. However, the strainer design will conservatively assume 75 lbs/yr.

The total sludge to be assumed in the design will be 300 lbs that assume an initially clean pool, which is cleaned every 2 cycles.

- 2. The FSAR was reviewed for historic data. In 1983, the wetwell interior surfaces, the interior of the Vent System up to the drywell, and all external surfaces of the Vent System were grit blasted to SSPC-10 or SSPC-5 at 1 1/2 - 3 mils profile. Welding repaired pitted surfaces of immersed wetwell shell. Grinding blended rough areas of wetwell shell.

Mobil 46-X-16 Epoxy Filler was applied to selected-pitted areas of the wetwell immersed shell portion prior to coating. Welding repaired surfaces in the Vent System thinned by corrosion. The immersed bottom half of the wetwell shell, the interior of the downcomer and the entire interior surfaces of the Vent Sys2m were given 3 coats of Mobil 78 Hi-Build Epoxy (DFT-16 mils). The vapor phase upper half of the wetwell shell, exterior of the Vent Header and vent lines portions inside the wetwell were given two coats of Mobil 78 Hi-Build epoxy (DFT-10 mils). Following coating application, the entire wetwell interior was heat cured at 108*F for 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />. Demineralized water was put back in the wetwell.

i 1940-98-20124 Attachment 11 l Page 12

3. In addition to the FSAR discussion, the licensing basis includes a Safety Evaluation of wetwell coating blisters (reference 14). The 12R visual inspection of the vapor phase and immersed Mobil 78 coated surface of the Oyster Creek torus found it to be in excellent condition. A few random blisters and areas of minor mechanical damage

were found as would be expected for a coating of this type after four years of l immersion service. Areas along the wetwell invert which have a combination of Mobil l'

78 over the top of Mobil 46-x-16 patching compound, however, have blistered at the 46-x-16 and tours shell bonding surface. These blisters have been classified as medium to medium dense with blister sizes of number 6 to number 2 (1/8" to 1/2" diameter with a few random blisters about 1" diameter) per ASTM D714. This Mobil 46-x-16 i

had been applied uniformly along the bottom surface in 10 bays and randomly in the l other ten bays.(also along the bottom surface) from approximately 5 o' clock to 7

o' clock positions.

Although this blistered condition was unexpected and unde:irable, no evidence of l corrosion damage associated with this condition was observed nor was there any

evidence to suggest that the coating is not adequately adhering to the steel shell and still providing corrosion protection.

l Although the above discussion focuses on the corrosion issues associated with the

- wetwell, the safety evaluation also addresses the concern of suction strainer blockage.

The discussions and conclusions are repeated here for completeness.

l The presence of blisters on the wetwell coating requires an evaluation to determine whether or not these blisters will spall during a design basis LOCA, be transported to l the suction strainers, and cause sufficient blockage to prevent the core spray and containment spray systems from performing their intended function. If a Design Basis Accident occurs, the suppression system experiences three phases of blow down. They are Pool Swell, Condensation Oscillation, and Chugging. Pool Swell is the bulk rise of

- wetwell water above the bottom of the downcomer followed by its falling back.

Condensation Oscillation is the steady condensation of steam with large steam bubbles exiting the downcomer and Chugging is condensation within the downcomer. The Pool Swell is the most violent and turbulent phase.

l To establish that the coating was adherent and not susceptible to spalling, a series of tests were devised. The type and magnitude ofloading in the vicinity of the blisters l under design basis LOCA conditions were calculated. Suction pressures of about l

2.2 psig and compression pressure of about 43 psig in the vicinity of the coating during a design basis LOCA condition were calculated.

L_________________-______-_________________-____

o 1940-98-20124 l

Attachment II l

i Page 13 A test was performed which showed, no size change, no undercutting ofligaments, no

, ~ delamination or spalling of any blister. Only one blister showed very minor leakage, but no size change, crackling, or fragmentation was observed. The compression load portion of the test revealed that the blisters were in fact elastic. When a heavy load r

was applied the compressed blister, it cracked but did not spall or grow.

In order to establish that the coating adhesion was adequate to withstand design basis

! - LOCA loads, an in-situ adhesion test method was developed. The coating ligaments in between the blisters were tested in different bays using an Elcometer adhesion test.

device. The results showed that the underwater coating ligament adhesion strength was between 37 - 135 psig. Since the design basis LOCA vacuum load was defined to be about 2.2 psig, it was concluded that the test result has demonstrated adequate adhesion strength of the coating ligaments.

A third test using a hand held paint scraper was used to peel the coating in between the l

blisters. The purpose was to evaluate the adhesion strength of the coating in and i around the blistered regions. This testing was conducted in several bays at locations j that had high blister density and representative sizes. The individual areas peel tested were limited to 1 square foot. These peel tests showed that the coating adhered well to the surface. I In addition to the above tests it was also noted that between January 1984 and ,

December 1986, there were eleven occurrences of EMRV actuation's as part of Plant  !

Surveillance Tests which resulted in steam blow down through the " wye" quenchers in  :

Bays 7,9,13 and 15 for a minimum of 5 seconds per occurrence. The visual 1 inspection of the affected bays during the 12R outage revealed no spalling of the coating and no spalling of blisters in these bays.  ;

Based upon review of the above results and observations, it can be stated that the blisters are not expected to spall during normal plant operation, transients, and design basis accidents.

l The Safety Evaluation went on to provide an evaluation of the impact of coating failure on the strainer performance. This evaluation is a defense in depth argument that will not be used in the sizing of the suction strainers. It is reasonable to assume that the coating within the vent system and the wetwell contribute 0 lbs of debris to the suctior strainer load. However, despite this it will be assumed that 50% of the identified biisters fail contributing 10 lbs of paint chips to the suction strainer debris load.

, 1940-98-20124 Attachment II Page 14

4. The Oyster Creek Primary containment is inert during operations with an atmosphere that contains less than 4% O2 . As such, there is little or no rust present within the suppression pool on any surfaces that are not coated. Therefore, wi'hin the suppression  ;

system, O lbs of rust is expected to contribute to the strainer debris loading from any l surfaces that are not coated. l

5. The wetwell and vent systems are maintained clean. As described previously, the l

wetwell vent system is maintained clean. During drywell entries, the vents are covered 1 to prevent debris entering into the downcomers. Wetwell access is controlled in accordance with Procedure 233.1. The wetwell, unless drained and opened for general l access is controlled as a zone 2 area. Tool accountability is maintained with all tools, l l equipment, and material logged in and out to ensure all is removed prior to closeout.

l This process is controlled in accordance with Procedure 119.3 and the respective work-controlling document (e.g. job order).

6.0 Suppression Pool'IYansport The Oyster Creek analysis took credit for suppression pool transport of fibrous debris. The NRC blockage code (Reference 3) was used to access fiber debris loading time. The l corresponding head loss was determined using the vendor correlations. Fibrous debris that is transported to or resides within the suppression pool was assumed to load onto the suction j strainer at a rate based upon the maximum system flow rates with a minimum suppression pool volume. Credit was not taken for any debris settling within the suppression pool. This ensured that a bounding evaluation of the suction strainer debris loading was provided.

7.0 Core Spray And Containment Spray NPSH Evaluation The purpose of this section is to demonstrate the need for contJnment overpressure as an l

analytical input to the calculation of the NPSH available to the core spray pumps.

Overpressure is defined to mean containment pressure greater than 14.7 psia (one atmosphere).

The current evaluation of the core spray NPSH does not take credit for containment pressure in excess of one atmosphere. For example, if the containment pressure is predicted to be 20 psia the NPSH calculation would assume 14.7 psia. During the next refueling outage, scheduled for the fall of 1998, new suction strainers will be installed within the suppression pool. The suction strainer modification is intended to address concerns associated with debris blockage of the strainers located within the suppression pool. It has been postulated that the jet forces from a LOCA could damage insulation and transport the debris from the drywell to the suppression pool. Once within the suppression pool, the insulation fiber is drawei onto the suction strainers by the operation of the core spray and containment spray pumps. The insulation will form a filter on the strainets and begin to capture small particulate such as iron oxide (sludge), dirt, paint chips, etc. The accumulated affect of this debris would proouce a significant pressure drop across the existing suction strainers.

L________________________________--___-___

. 1940-98-20124 Attachment 11 L Page 15 l

With the strainers clogged, the pumps drawing suction from the suppression pool could cavitate.

i l

To ensure that the core spray and containment spray systems function as designed, the suction l strainers will be modified to accommodate the new, postulated debris loads. The design of the new strainers includes consideration of containment pressure. The pressure is conservatively l established to ensure that margin is available within the design values. The use of containment '

pressure is a change to the current plant design basis. This document outlines the operation of those systems that take suction from the suppression pool during a design basis accident with emphasis given to the EOP's. The use of overpressure will be shown to be a reasonable 7 physical assumption and consistent with the plant operating procedures.

l 7.1 System Description l Oyster Creek is a BWR 2 with a Mark I containment. Three suction strainers supply a l common ring header (depicted below) that service four containment spray and four core spray pumps. i Ring Header NZolB NZOIC NZ01D NZOIA Y

CS 1 1 4 X68B

> X68A X69 4 y CS 1-3&4 CS 12 X68A, X68B, X69 - Current Suction Strainers NZ01 A/B/C/D. -

Core Spray Pumps CSI-1/2/3/4 -

Containment Spray Pumps l

l

< .I

4 1940-98-20124 Attachment Il Page 16 l

l 7.1.1 Containment Spray The containment spray pumps are divided betwe,en two systems. Refer to figure below l for a schematic diagram of the system configurative. Each system is initially aligned to cool the suppression pool. The control room operators in accordance with the j Emergency Operating Procedures (EOPs) manually start the pumps. The control room operators are trained to manually trip the containment spray pumps when containment pressure indicates 1.0 psig. The pumps are then placed in wetwell cooling mode.

When aligned to spray the drywell the pumps automatically trip on low pressure j (currently set at 0.6 psig). This trip is not safety grade so it only represents a backup i to the operator action. The significance of the low pressure pump trip will be I emphasized in operator training on this modification.

I NC To Drywell Atmosphere

> NC To Wetwell Atmospherv X >

l y To Suppression Pool NO i

From Suppression Pool Heat Exchanger  !

Pump M

y i l

From Suppression Pool Heat Exchanger p,,p Wqzy p i 5 s U

l Containment Spray System Arrangement f

L L------_ _ _ . _ _ _ _ - . - _ _ _ _ _ - . - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ - _ _ _ _ _ _ _ _ _ _ - - - _ _ - - - - - - - - - - - - - - - . _ - - . _ _ _ _ - - . _ _ _ _ _ _ _ _ _ _ _ _ - . _ - - - - _ _ _ _

, ,- l

. 1940-98-20124 -

Attachment II Page 17 7.1.2 Core Spray i

The Core Spray System contains two completely independent systems each containing two sets of pumps. One main pump from each system along with one booster pump from either system can supply 100% rated flow. Either low-low reactor water level or high drywell pressure signal actuates both systems. The single failure in the Core  ;

Spray System (loss of an Emergency Diesel Generator) cannot impair the capability of the system to perform the required safety function.

For the purpose of this evaluation, the single failure assumed is selected to minimize the NPSH conditions with maximized flows. To accomplish this, the failure of the core spray pump NZ01A is assumed. This failure will cause the start of the pump NZOIC with a booster. This pump has the least available NPSH and is appropriately included in this evaluation.

The two ECCS subsystems operate in various combinations to maintain peak cladding temperatures below 2200"F, and within the limits of 10 CFR 50.46, for any size break LOCA even if a single failure occurs in the ECCS.

The ECCS was provided to meet definite design criteria with respect to the design basis LOCAs. The criteria to which the ECCS was originally designed included the following conditions:

For all the design basis LOCAs, sufficient cooling capacity must be provided so that a conservative design evaluation will indicate that, as a result of the abnormal temperature transients which occur due to the loss of coolant, no cladding melting would occur. By preventing cladding melting, the core would be held in a definable geometry and subsequent core cooling can be predicted with confidence.

The above criterion of no cladding melting was satisfied for the entire spectrum l of reactor primary system break sizes from small leaks up to and including the complete double ended severance of one of the recirculation lines. Detailed evaluations were required for various break sizes, as the entire behavior of the reactor system is dependent on the size of the break assumed.

i The two ECCS subsystems, which together perform the emergency core cooling l

function must have sufficient redundancy of active components to be able to accomplish their emergency core cooling function even under the condition of a single ' lled component. This criterion is satisfied in that the Core Spray System was provided with a 100% redundant pumping and valving system.

Core Spray System configuration for the LOCA mitigation requires one main l pump and one booster pump from one subsystem and one additional main pump from the other subsystem.

. 1940-98-20124 l Attachment II l Page 18 i The ECCS was designed to satisfy the criterion of no loss of function upon i

Loss-of-Offsite-Power (LOOP). This figure illustrates the system configuration.

NC X > To RPV NC X

> To Wetwell NC From Suppression Pool Booster Pump Main Pump A

v v From Suppression Pool l

Main Pump BoosterPump p A V

V Core Spray System Arrangement I

l

~

1940-98-20124 Attachment Il Page 19 7.2 Change to the current ECCS NPSH evaluations The change to the NPSH evaluations is strictly analytical. The difference from the current analysis is that the design evaluation includes containment overpressure. The containment conditions have been established to ensure that the analysis is conservative.

The operational response (EOP) to NPSH concerns will be conducted in a manner much like that employed currently at Oyster Creek. The operators will monitor containment conditions (pressure, temperature, and water level) and pump flows to assess the NPSH of the operational ECCS pumps.

7.2.1 Containment Response to a LOCA 7.2.1.1 Large Break LOCA l

The containment response to a large break loss of coolant accident is depicted on the next page. The drywell pressure rapidly rises above 34.7 psia. The wetwell pressure follows closely behind until the blowdown subsides and the containment sprays reduce the drywell pressure below that of the wetwell. Once the drywell pressure drops 0.5 psid below the wetwell, vacuum breakers open returning the previously purged non condensable gases to the drywell atmosphere.

The suppression pool temperature rapidly rises during the reactor vessel blowdown.

Following the depressurization of the reactor vessel, the core spray system automatically initiates adding additional energy to the suppression pool as it cools the core and overflows into the suppression pool from the drywell. An example of the pool temperature response is depicted on page 21.

The containment spray system is manually initiated in response to the accident. For the analysis presented in this study the containment spray systems are assumed to be started early in the accident (i.e., within a few seconds of the accident). This will reduce the containment overpressure early in the event as well as maximize the flow through the common suction strainers. Maximizing the flow through the strainers will produce the largest pressu: e drop across the strainers and reduce the strainer debris loading time (minimum debris loading time is approximately 30 minutes).

l 1

l -

1940-98-20124 l Attachment II j Page 20 l Once the drywell pressure is reduced to 1.25 psig, the containment spray system is ,

) assumed to have been tripped either by operator manual actions (operator actions are

)

l presently in the EOPs) or automatically by pressure switches. These switches are l routinely surveilled and provide an excellent defense in depth for the operator actions.

Note that the assumed actuation limit is greater thart that currently in the procedures. ]

The appropriate setpoint and procedural changes will be made in support of the i modification.

I The pressure profiles given below show the pressure response following the trip of the j containment spray pumps. As can be seen from the figure, the containment pressure '

rises again as the spray cooling terminates on low pressure. As will be discussed later in this document (section 9), no credit is taken for the re-pressurization of the wetwell above 1.25 psig (15.95 psia) in the suction strainer design.

l The containment spray system that is tripped on low drywell pressure is realigned to {

cool the suppression pool. This maintains the high flows through the suction strainers. I With the containment spray flows maximized, the cooling of the suppression pool is also maximized. This is demonstrated in an additional case presented later in the analytical results section of this document.

1 Containment Pressure Response f

l 45 00 ,

I 40 00 #

35 00 '

/ \ \

/

_ \

/ . .. -

- A- g _ _ _ . ,

= = WW Press (pm.)

~

gw - -

,,,, /

10 00 5 00 0 00 1.00 10 00 100 00 1000 00 10000 00 Time (sec)

{

I

l l

l

• 1 suppe.'*8oa Pool Temp 1940-98-20124 v.. ,, Attachment II m Page 21 "5 "

uo m I

e I i lll

,es l y

'25 " . _ _ p i

mm /

/ j l

,_ / \

l mm f

m. /j aw /

j/

\ ,w w

)

.0 x 1.00 10.00 100 00 ,000 00 10000 00 Time (sec)

I 7.3.2 Small Break LOCA l

l The containment response to a small break LOCA is similar to that of the large break.

The primary differences are the magnitude of the containment pressure rise and the amount of debris. The large break will produce a greater pressure rise and the most insulation debris. The suppression pool temperature response will also be somewhat different for the small break, however, the actuation of the ADS will tend to equalize l

the overall response. Certainly, the large break will heat the pool more rapidly i providing an early challenge to the ECCS NPSH requirement.

l For a small break, the containment pressure will rise above the high drywell pressure l setpoint (3.5 psig). It is conservatively assumed that for all break sizes the operator actuates containment sprays, this reduces overpressure and increases strainer head loss.

When level in the reactor reaches Lo-Lo-Lo level, ADS actuates reducing reactor ,

pressure by releasing coolant into the suppression pool. This will cause the core spray )

pump flow rate to rise reducing the NPSH available and increasing the NPSH required.

This brings the small break LOCA into line with the large break LOCA in terms of flows through the common suction strainers.

Based upon the discussion provided above it is concluded that the large break will sufficiently bound all breaks sizes within the design basis accident spectrum.

Therefore, the remaining analytical results provided are based solely on the large guillotine failure of a recirculation loop.

1

_.__.____._.___.___.s

, 1940-98-20124 Attachment 11 Page 22 8.0 EOP Actions That Impact Suction Strainer Performance l

In evaluating the use of containment pressure in the design of the suction strainers, the Emergency Operating Procedures (EOP) were given direct attention. GPU Nuclear recognizes that the EOPs must work with the design basis of the overall plant as well as the systems required for safe operations. However, credit is not taken for EOP actions such as aligning core spray to take suction from external sources. Such actions are considered to be defense in

' depth and are not considered appropriate for establishing a suction strainer design basis.

The EOP actions that are credited are associated with maintaining containment over pressure (removing the containment spray system from drywell cooling mode) and NPSH to the pumps.

Concerning the low containment pressure action, several points should be kept in mind. First, the manual action is backed up by an automatic trip of the pumps. Second, the containment spray system is manually started by a conscious operator decision. The operator is trained to trip the containment spray pumps on low pressure and as such, attention is given to the s containment pressure response to the sprays. It is not expected that the operator would walk away from the controls after starting the pumps. As can be seen from the graphs provided earlier, following the blowdown, the containment spray system will reduce containment pressure quite rapidly. Within several minutes, the operator will have a good idea as to whether a pump trip will be required. These issues will be re-emphasized to the operators during the training associated with this modification.

A review of the EOPs was conducted to ensure that this was properly considered in regards to this modification. The principal EOPs that will influence the overpressure assumption and the l overall NPSH assessment are those associated with Primary Containment. These include l

Primary Containment Pressure Control, Drywell Temperature Control, Suppression Pool Temperature Control, and the NPSH limit curves, all of which are currently included within

' the plant specific EOPs.

8.1 ' Primary Containment Pressure Control When the wetwell pressure exceeds 12.0 psig the operators are instructed to initiate drywell sprays. This action is expected after all but the smallest LOCAs. For the large break LOCA, the EOP step is reached almost immediately while for the smaller LOCAs, the step may not be reached for several minutes. This action will reduce the containment pressure hence the available NPSH to the operating ECCS pumps. In addition, initiation of drywell sprays will increase the flow and pressure drop across the common suction strainers and accelerate the strainer debris loading.

- Despite this, the action is an appropriate response for the loss of coolant accident. The action is included in the evaluation of the suction strainer head loss and NPSH evaluation. No procedural change will be made to accommodate the overpressure assumption used in establishing the design basis of the suction strainer.

1 4

_ _ . _ _ . _ _ _ _ . _ _ _ _ . _ _ . _ _ . . _ _ _ _--- U

.. 1940-98-20124 Attachment II l Page 23 i

l Existing plant pressure switches trip the containment spray pumps when drywell J l pressure drops below 0.6 .1 psig. The EOP has the operator confirm that drywell l sprays have tripped if drywell or wetwell pressure drops below 1.0 psig. This particular step will be modified to ensure that the drywell sprays are terminated at a l pressure greater than 1.25 psig in the wetwell. This will include a change to the l automatic trip setpoint as well as the EOP. However, the purpose of the EOP will l remain unchanged as the pumps are tripped to prevent the actuttion of reactor building to wetwell vacuum breakers and the subsequent deinerting of the primary containment.

l l l l Tripping the spray pumps will reduce the suction strainer flow and maintains the minimum assumed containment overpressure. No credit is taken for the resulting reduction in strainer flow since the operators will likely place the tripped system in suppression pool cooling mode in response to elevated suppression pool temperatures.

No procedural change will be made to accommodate the overpressure assumption used in establishing the design basis of the suction strainer other than that of the action level to trip the drywell spray pumps.

Several EOP limit curves have the operator emergency depressurize the reactor such as l the Pressure Suppression, and the Torus Load Limit /SRV tail pipe limit. These actions will deposit energy into the suppression pool reducing the available NPSH to the ECCS pumps. These actions are appropriate and will not be excluded from the analysis. .

l Within the design basis accidents, these actions are associated with small break loss of

{

coolant accidents.

8.2 - Drywell Temperature Control Before the bulk drywell temperature reaches 281*F the EOPs have the operator l Emergency Depressurize the Reactor and Spray the drywell (if the Spray Limit curve is

! not exceeded). These actions reduce the NPSH available to the core spray pumps by  !

L raising the suppression pool temperature and reducing the containment overpressure.

However, these actions are appropriate and will be maintained as currently defined.

Therefore, the suction strainer design will accommodate these EOP procedural actions.

As was the case previously, the operators will confirm the drywell spray pump trip prior to wetwell pressure dropping below 1.25 psig.

! 8.3 Torus Temperature Control l

The EOPs provide guidance to the operator to control suppression pool temperature.

When the bulk suppression pool temperature exceeds 95* F the operator is instructed to initiate a containment spray pump in each system. This action will increase the flow through the suction strainers and reduce the NPSH available to the operating core spray

f 1940-98-20124 Attachment II Page 24 l

pumps. In the long term (several hours) this action will reduce the peak suppression pool temperature optimizing the NPSH available for the given set of flow conditions.

In addition, the operation of these pumps will accelerate the debris accumulation on the suction strainers. This will also reduce the NPSH available in the early part of the accident. However, the EOP actions are deemed to be appropriate and will not be modified. The suction strainer design will accommodate the action in a conservative manner.

l l 8.4 NPSH Limit Curves i The EOPs presently contain NPSH limit curves for the core spray pumps. These

! curves will be modified to account for the new strainer design. However, the curves will fundamentally remain the same. The curves guide the operators in the i

management of containment overpressure and NPSH available to the ECCS pumps.

l The NPSH limit curves have the operator evaluate:

1. Containment Pressure l 2. Suppression Pool Water Level

3. Suppression Pool Temperature

4. Core Spray Configuration and Flow Rate The figure provided immediately below is used by the operator to assess the containment conditions. From this figure, the operator is able to determine which NPSH curve (A - G) should be evaluated on the figure on the next page.

OC Station (Static Head)

( 5 G

a I,

F

,e. re

m. > >

E i

2 D

l C

i B

A 0

100 125 150 175 200 Wetwell WaterLevel(in)

• i 1940-98-20124 l Attachment 11 Page 25 j l

NPSH Limit Core Spray C Pump l Te (

220 N

% N N. .t . _ x

'*e -=**==A N, .,

u.:  ;

v- R E 1.

140 v N. .3 -

130 120 2500 3000 3500 4000 4500 5000 Flow (gpm)

Note that each core spray pump has its own limit curve.

l 9.0 Suction Strainer Head Loss The acceptable head loss across the new suction strainers was calculated for a dynamic condition. The limiting core spray pump NZ01C was assumed to be in operation at its run out flow rate, a Loss of Coolant Accident was assumed, and the containment and core spray systems associated with its mitigation were initiated. The impact that these systems have upon l the containment response was assessed. The resulting containment pressure, level, and temperature response as a function of time was calculated and used to assess the NPSH l available for the core spray systems. The benefit of the full containment pressure is credited i for the initial part of the transient. With the initial part of the transient defined to be that time prior to the trip of the drywell spray pumps. A wetwell pressure of 1.25 psig is assumed for the remainder of the evaluation. The overpressure was conservatively calculated to minimize the benefit of the containment overpressure. The suppression pool response was also calculated in a conservative manner to ensure that the result is bounding and conservative.

s' 3

. 1940-98-20124 Attachment II Page 26 i-l As described previously, it is assumed that the operator trips the containment spray pumps in f j drywell spray mode when containment (wetwell or drywell) pressure reaches 1.25 psig. In the  !

I' analysis, this occurs at approximately 3.75 minutes after the accident. From the period beginning ten minutes after the start of the accident, credit was taken for operator action to reduce core spray flows as needed to address any NPSH concerns associated with the core l l spray pump operation. In the analysis, the "C" booster pump is tripped at 10 minutes, and the "B" booster pump is tripped at one hour. ' After one hour, containment over pressure is no j longer required. It should be noted that core spray pump NPSH limits bound those of the containment spray pumps.

l 9.1 Containment Overpressure The following assumptions minimize the benefit of containment overpressure and maximize the head loss across the strainer.

i

1. Containment Spray was assumed to start early in the event (Time 0) l e Maximized strainer flows - strainer pressure drop L e Minimized containment pressure

! e Minimal impact on suppression pool temperature early in the event e Peak suppression pool temperature was not changed by pump start time (Controlled by decay heat, pool water mass, and heat exchanger capacity energy balance) e Containment spray system placed in drywell spray mode after tripping off at l containment pressure of 1.25 psig.

2. Containment Heat structures credited as heat sink in the analysis

3. Initial containment conditions selected to provide bounding NPSH evaluation e Minimum Suppression Pool Level This assumption minimized the heat capacity of the pool providing the maximum temperature response as well as the minimum static head available to ,

the suction of the pumps. In addition, minimizing the suppression pool l

maximized the wetwell vapor space. With a large vapor space, the minimum containment pressure response was established. This was a direct result of the larger volume available for the drywell non-condensable gases that were transported to the wetwell during the reactor vessel blowdown.

e Wetwellpressure 0 osie This assumption minimized the containment pressure response by minimizing the number of non-condensable gases present within the containment.  !

• Wetwell Humidity 100%  ;

With an initial wetwell humidity of 100%, the non-condensable gases are at their minimum value for the given pressure and temperature conditions.

However, this is also a reasonable assumption since the vapor space is in direct contact with a large suppression pool.

I l

l 1940-98-20124 Attachment II Page 27 e Wetwell Temperature 95'F This assumption provided two effects on the NPSH evaluation. This value was l applied to the atmosphere as well as the pool. It represented an operational l maximum and as such served to minimize the number of non-condensable gases l initially present within the wetwell. For the suppression pool, this temperature j represented a maximum initial value, would tend to minimize the pool mass for 1

a defined volume, and maximized the pool temperature response.

• Drvv> ell Pressure 0.0 osip This assumption minimized the containment pressure response by minimizing the number of non-condensable gases present within the containment.

! e Drywell Humidity 100%

With an initial drywell humidity of 100%, the non-condensable gases were at their minimum value for the given pressure and temperature conditions.

• Drvwell Temperature 150* F 1

! As with the wetwell atmosphere, this value represented an operational maximum

that minimized the number of non condensable gases within the containment.  !

Minimizing the number of non-condensable gases served to minimize the l

l containment pressure response to the loss of coolant accident. i

! l

4. Degraded Emergency Service Water Heat Removal Capabilities

• Maximum emergency service water temperature assumed (95* F) l l This assumption minimized the heat removal capacity of the system and produced a higher long term pool temperature response.

! e Fouled Heat Exchanner This assumption minimized the heat removal capacity of the system and produced a higher long term pool temperature response.

e Minimum ESW Flow rates

, This assumptica minimized the heat removal capacity of the system and produced a more severe suppression pool temperature response.

l 5. ANS 1979 decay heat with a two-sigma uncertainty was assumed.

l i

l f

c__ _ _- _ ___ - - _ __ - __ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ - _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ - _ - _ _ _ - _ _ - _ _ _ -_

1940-98-20124 Attachment II ;

l Page 28 l The figure provided below makes a comparison of the drywell response to a large break l LOCA with nominal containment initial conditions and also with those assumed in establishing the suction strainer design. The figure on the next page compares De drywell response used to evaluate NPSH with that of the peak drywell pressure evaluation. A,. can be seen from these figures the drywell pressure response is conservatively established for use with the NPSH calculations, g l

i Graph of pressure comparing nominal and minimum pressure results 50 00

1. ,2 4 ft

• Mvn> "" , r 4 ' f.gg lb a' ~'

• l l 'l , ' '

W; y *

+, . ~

'+ p' ^ % i s

u p ,' ~ ,,,

- j gg <

; y ,

l 40 00

< ,' i s ~< f i preneure < <

l (M) - h >rt s er

• 3500 ,, .' , ,

g - +, ^

OW Press Nominal Corukbons

> -s

1. g#' 1 ,

'5 a , i ,

= = EW Press Boutubn0 Conchbons l - v 4 ' , ., 4 l s es > ,' kn > ' '

i

}! , t >

's

> ' 4 >A &

3

,A>$

, , MC l -^t> r 3

'>s  ; 5

[ t'

)

[h

_ t% , - ,>

y"' ^ , '

>8, ' 'p t

f. ' ' , y '

h$ p f x , j 4 , ,2 , + r . , (<f b'k,',,res Y, '. ; ,

! $m som imm som m imm m l

Time (sec)

I 1

i i

1

~

. 1940-98-20124 Attachment II Page 29 I Graph of presaure comparing peak and minimum pressure results e0 00

'k' ' f',N , i,l ',; q >

'l;l, , , , ,

~& :'='~ ,t' n, n .;. ,17 's a 'A 2.' ,1 -

f j k! k h  ? .' , c '; [ u .

x s , n . ,,: ,

~ > < >

~ '

m

, , . . ' .e ' ;'  ; '

e  :~ wv:

< y

. y , , ,, , ,

(pe4a) 3 '

..> - ,s  : e g p q w- ,v' qs >' .a >

C 00 za - , p* . Y,,::: ; ,

- w er w m e m

}s J', -== Peak M Proesure Analysm

$5 m ,'g r o ,

,' ,m

, s g m

a o .

3 2 2

,3 , s>

J 'pt ,r 'y[. :s

) 4s ['> % y <

5'(,

s , a ' '

e ,~ , <~ A . . ,

f~;^,

M -

v

'l t

y ,,

3,

_o

>+,v>,

; 'a '}'[} y p ' '

. '3

,~'

sl f >, L

, ' .c ~n.. ~ , , [<S, s

e,

>\

• • L . . .

ll .

'q g

15 00

' :;'r; 'j l N

1 00 10 00 10000 Time (sec)

It should be noted that for the minimum pressure case the containment pressure recovered following the trip of the drywell spray pump. This results from the containment structures acting as heat sources. However, as stated above this was not credited in the evaluation of the NPSH available to the core spray pumps. Instead, the wetwell pressure was held at 1.25 psig (15.95 psia).

9.2 Flow Management Using the EOPs, the operators will evaluate the NPSH available to the core spray

pumps. If a pump is operating at or above its NPSH limit, the operator will trip the booster pump as required to reduce the pump flow. This has the benefit of increasing the available NPSH while reducing the required NPSH.

l i

This action has been evaluated and it has shown that the 10 CFR 50 Appendix K flows will be maintained for the large and small break LOCA. The analytical evaluations conducted to support the suction strainer design included an operator action to start the containment spray system, and then trip it at 3.75 minutes.

t

1940-98-20124 Attachment 11 Page 30 f

l The analysis assumes the core spray booster pump associated with NZ01C is tripped at

{

10 minutes to ensure that NPSH was maintained. The booster pump associated with j NZ01B is tripped at approximately one hour. At this point, containment overpressure i is no longer required.

I Of the pumps involved, only one of the core spray booster pumps (NZOIC-Pump) required tripping this early. The other pumps can run for at least an hour prior to any potential operator action.

The tripping s the booster pump associated with NZOIC does not violate the pump combinations associated with the Appendix K evaluation. The Appendix K evaluation assumes two main pumps and one booster for core cooling. In addition, it should be ~

noted that the Appendix K evaluation also assumed degraded flow conditions for the pumps (4100 gpm main and booster pump). If these flows were to be assumed there would not have been a / eed to trip any of the operational booster pumps. For analysis purposes in this document, the core spray pump is assumed to operate at its runeut flow of 5000 gpm when a booster pump is running.

10.0 Analytical Results Two cases are presented one to maximize strainer flows and the other to maximize suppression j pool temperature. The first case presented represen'.s an evaluati on of the maximum strainer '

flows. The table below provides the flows assumed in the first (maximum flow) evaluation.

.'Timg 4 ,

1 ttPCS 1LPCS '1 CSg -) iCSS

,or ' kwat(Bqosty) . (with Booster) l 5000 gpm l 5000 gpm 3650 gpm 4200 gpm l l 0 to 10 min

' 10 min to 60 min ' 5000 gpm l 4260 gpm l 4200 gpm 4200 gpm l 60 rnin on 4260 gpm l 4260 gom l 4200 gpm 4200 gpm l

1940-98-20124 Attachment II Page 31 Two graphs are provided, the first shows the pool temperature response.

,, Suppression Pool Temp

("F) 145 00 k:bM hb hb -

) !! N S  % *

@M 5idH @hf E$N5 i I N  %.k[i. **; %6 WC y

s n $$.$ e,...:

J  ! * '

,,,oo ___

] , ..

d .. u o ..we w e?,!3$wh jj h i

F m:

0 U MhIhilkhk $hhh 130 00 1%$f$ l&

_ $bb ?h) pJ 'g Nm 2 a '

! pp w e m!v *ev ,

o_ m ,

'a lj- 2d$  ; l . $d$i ,f

$4R i@jg

^

q j NP g U

h, l [ .j!jhfhi ihk 4 fhh@ m s?

y en gy > syf[

,g, #  : u:2m.s  :

w a n m ab <

@ mie u s sm? # ire .o  % .u.;

1 ; UO NW 14 m

% W$igr ) igg & # W sgW  %  %. pw gg_yg ,

110 00 y,f hh' gg s h.k<

i g:] fj fhkh[

$ i i

I.

p gp wp w ,

5

g j sj jq >

'h: . h.>

105 00 . . .

c ,

Ep d .

..j 11$$$ " '4 Mi $s

@ 3 ij$lhy#..

9pp%g g '

g -

x

p i

g t

3:  ;

p y A g .; p ., gg Wp

. et, I Q, $ 3 ffh i l_i@ A, m

2 EU >

1 00 10 00 100 00 1000.00 10000 00 T6me (sec)

The second graph provides the head less remaining between the available and required NPSH (labeled as NPSH C and NPSH B) as well as the overpressure assumed in the analysis. The first two curves actually represent the tolerable strainer head loss (i.e., provided the strainer head loss does not exceed the curve, adequate NPSH is assured). The overpressure assumed is based upon the containment response to a large break LOCA with the ECCS and containment spray flows provided in the previous table.

The NPSH labeled curves clearly show the benefit of overpressure. As the pressure inside containment drops the tolerable strainer head loss also drops. Note that at 600 seconds when the 'C' pump booster is tripped the tolerable head loss rises for the 'C' pump. A similar response is seen with the B pump at 60 minutes.

Since the strainers are sized for 4 feet of head when fully loaded, the time for the debris to load is an integral part of the evaluation. As can be seen from the figure, the 'C' pump would i not be able to tolerate 4 feet of strainer head loss prior to the booster pump trip at 10 minutes.

However, the strainer debris loading time is in excess c,f 30 minutes; therefore, there is not a concern with the 'C' pump since the booster pump can be tripped prior to reaching full debris loads on the strainer.

I k

, 1940-98-20124 Attachment Il Page 32 Over Pressure Contribution To Head 40 00

/%

35.00 /- -

B booster 30 W y'

't c booster pump

\

25 00 k si tnpped j

y .

E I g \  ; NPSH Pump C

$ 2000 g g --- NPSH Pump B I  %

-+--Over Pressure Contribution To Head 15 00 i I h

10 00

\ k r

--~

1 3' _f i

i 5 00 ---

1 0 00 Ili~llll l I I lilll 10 100 1.000 10,000 T1rne (sec) l The minimum NPSH available to the C core spray pump for this scenario was 2.55 ft while l i the B core spray pump would have 4.29 ft available. Note that without credit for containment pressure, the 'C' pump would cavitate. This occurs when the curve labeled 'NPSH Pump C' overlapped with the curve labeled ' Overpressure Contribution To Head'. Once the booster pump is tripped the need for overpressure is no longer critical. Therefore, for the high flow l condition containment pressure in excess of 14.7 psia is only required for the first hour of the accident.

l

, 1940-98-20124 Attachment II Page 33 The second case demonstrates the response with minimum suppression pool cooling. The pool temperature is most severe for this case. However, the minimum strainer flows overcome the NPSH issue.

Timing -

1 LPCS 1 LPCS 1 CSS -

1 CSS (with Booster) (with Booster)

O to 10 min l 5000 gpm l 5000 gpm l 0 gpm l0gpm l 10 min to 60 minl 5000 gpm l 4260 gpm l 3200 gpm l0gpm l l 60 min on ] 4260 gpm l 4260 gpm l 3200 gpm l0gpm l

,, Suppression Pool Tomo (*F) rF) 165 00 v

155.00 /

, .5 . /

7 135 00  ! d /

125 00 f

,,5 . /

.. J 7 ,

95 00 l

1 00 10 00 100 00 1000.00 t0000 00 100000 00 Time (sec)

1 e .

1940-98-20124 Attachment Il i Page 34 l l

1 Comparison of NPSH with Over Pressure 45.00 40 00 - -

\ I 35 00 30 00 g ,

's g C txxater pump ~~

I inpped I

, K {

g 2500

\ ' -~ +

B booster -*-NPSH Pump C gi

[ pumptapped

-NPSH Pump B

-.,P_,._,.e s

I f i V / \

'5 "

l d

/ i j 10 m

[ u

• mk;g%

~

5" i u o 00 III I I I IIIIII 10 00 100.00 1000 00 10000 00 100000 00 Time (sec)

The minimum NPSH available to the C core spray pump for this scenario was 6.9 ft while the  ;

B core spray pump would have 6.92 ft available. The analysis that maximizes pool

{

temperature, as well as the high flow analysis, only required overpressure for the first hour of j the event. Note that the need for overpressure is less critical for conditions where the pool 1 temperature is maximized. The comparison of these two results illustrates the benefit of flow management strategies.

The evaluations presented in this document have demonstrated the need for containment overpressure during the first hour of the event. After that time, reducing flow eliminates this i I

need. The containment pressuie response was conservatively calculated. However, the design does depend upon the low-pressure trip of the containment spray systcm when in drywell spray mode to ensure wetwell pressure does not drop below the 1.25 psig as requested by this l submittal. This manual trip is backed up by non-safety related pressure sensors following a j manual start of the system, fraining will be provided to th: operators to ensure they fully 1 understand the significance of this action in terms of the suction strainer design. Therefore, the requested overpressure is conservative, reasonable and represents the minimum change I required to ensure this modification is a realistic response to the NRC Bulletin 96-03 concerns.

~

! .- 1940-98-20124

- Attachment II Page 35 11.0 Strainer Description The strainers to be installed at the Oyster Creek Nuclear Generating Station are designed by General Electric. ' The selected design was the Stacked Disk ' Array. The strainers will be sized to ensure that the pressure drop across each will not exceed 4 feet. The following design

. values were assigned:

l Total Flow 8 18,400 gpm

[- Pool Temperature 8- 125 F Fiber Load 8 0 to 242 ft'

i. DW Coating Inorganic Zinc 47 lbs Paint Chips (DW Equipment) 40 lbs L Paint Chips (Torus Coating) 10 lbs l Dust / Dirt / Concrete 150 lbs Rust 50 lbs Iron Oxide, Torus 300 lbs 1

The strainer dimensions to be installed are:

L l Number 3 p Height 57 in L

Diameter

• 51 in Weight 3300 lbs

. Surface Area 475 ft' Maximum Head Loss 4 ft l-

• The flow value corresponds to two core spray systems at 5000 gpm each and two containment spray systems at 4200 gpm. These are maximum system flow rates.

-2 The pool temperature is selected to be below the expected temperature as this will give a more conservative result in terms of pressure drop across the debris laden strainer.

' The fiber debris used in the design is 102 ft' greater than that calculated for the worst  ;

design basis break location. The strainer is designed for the full range of fiber debris values to demonstrate the capabilities in the thin film regions of the debris loading.

• Note that these dimensions exceed the opening area of the wetwell. As a result, the strainers ,

I will be segmented and reassembled in the wetwell. The performance characteristics of the strainers are not affected by the segmentation.

_ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ - _ - _ _ _ _ _ _ - _ _ _ - - - _ _ - - _ _ _ _ - _