ML20035E842

Provides Estimate of ECCS Design Margins in Response to Unresolved Item 92-03-01 Noted in Safety Insp Rept 50-155/92-03 on 920127-0429 of Engineering & Technical Support Programs

ML20035E842
Person / Time
Site:	Big Rock Point File:Consumers Energy icon.png
Issue date:	04/15/1993
From:	Donnelly P CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9304190310
Download: ML20035E842 (12)

Text

~

~

1 Consumers Power P irick = o.aa.nr i

Plant Manager PCFWERING AUUNGAN'S PROGRESS Beg Rock Point Nuclear Plant.10269 US-31 North, Charlevoem. MI 49720 t

April 15, 1993 Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 DOCKET 50-155 - LICENSE DPR BIG ROCK POINT PLANT -

EMERGENCY CORE COOLING SYSTEM (ECCS) DESIGN MARGIN - UNRESOLVED ITEM 92-03-01 During the period January 27th through April 29th 1992 a routine announced safety inspection of the engineering and technical support programs, including modifications and design changes implemented during the most recent refueling outage, was conducted by the Nuclear Regulatory Commission (NRC)

Region III staff at the Consumers Power Company (CPCo) Big Rock Point (BRP) plant. As a result of the review the Region III inspector expressed concerns with the methods used to verify the Emergency Core Cooling System (ECCS) design requirements. The methods include using the FLOWNET computer code to predict emergency core cooling flow rates. The inspector concluded that FLOWNET consistently over predicts system flow rates and since FLOWNET predicted ECCS flow rates are only slightly higher than the required flow rates, and that littic or no margin exists in the ECCS performance. This concern was classified as an unresolved item.and forwarded to the Office of Nuclear Reactor Regulation (NRR) for resolution.

(155/92003-01).

On September 30, a conference call was conducted between NRR-and the Big Rock Point staff to discuss the concerns. This conference call resulted in an agreement by CPCo to summarize the inherent.ECCS design margins contained in the ECCS/ Core Spray design analyses and associated submittals. The following provides a brief history of the BRP ECCS design evolution, a description of the inherent margins in the ECCS design, a discussion of the Palisades Service Water System FLOWNET Model Validation / Benchmark Report, a comparison of-the RETRAN and FLOWNET computer code results, and a review of FLOWNET ECCS i

sensitivity analyses.

1 This letter provides our estimate of the ECCS design margins contained in the referenced analyses and submittals.

j

( -l i

nn4

.4 9304190310 930415

~

DR ADOCK O 5

A CA6 ENERGYCOMPANY

{

t i

NUCLEAR REGULATORY COMMISSION 2

BIG ROCK POINT PLANT ECCS DESIGN MARGIN t

April 15, 1993 l

l

SUMMARY

Considerable margin exists in the BRP LOCA analysis design flow requirements I

of 292 and 296 gpm. The 292 and 296 design rated flows were satisfactorily demonstrated in the CPCo ECCS test program. The design flows were further conservatively enveloped by the development of a maximum bundle power i

technical specification for each bundle. The creation of the maximum bundle power specification which included many additional conservatisms justifies the operability of the Ring Spray System (RSS) and the redundant Nozzle Spray i

System (NSS).

l Some examples of conservatisms include the following:

- A minimum of 20% margin exists in the minimum flows used in the Tech Spec maximum bundle power calculations when compared to the nozzle and sparger spray limiting test data respectively.

- Over 45% margin exists when using the limiting 0.375 ft2 LOCA analysis time-to-rated spray decay heat value.

l

- Over 5% margin exists in the experimentally derived ring spray l

nozzle loss coefficient and over 45% margin exists in the experimentally developed nozzle spray loss coefficient.

l

- Margin exists in the form of the administrative limits imposed on the maximum allowed bundle power that is, the 1% margin and the i

numerical uncertainty applied to the 4.0 MWt specification.

l

- The BRP design is the only GE BWR ECCS with design spray test data.

l Because of these significant conservatisms, the successful bench mark of F10WNET to the RETRAN code, a better understanding of the Palisades i

FLOWNET bench mark test data, and the experimental verification of the ECC.,

spray geometry, it can be concluded that considerable margin exists in the l

BRP ECCS design.

Therefore, using FLOWNET calculated results to establish the minimum acceptable fire pump head curves for the monthly surveillance test is acceptable. This approach ensures that the Big Rock ECCS satisfies its licensing basis.

INTRODUCTION The Big Rock Point ECCS system is composed of two core spray systems:

the Ring Spray System (RSS) and the redundant Nozzle Spray System (NSS).

In the Commission Memorandum and Order dated May 25, 1976 [1], CPCo was granted an exemption until the refueling outage scheduled for the spring of 1977 from the single failure criterion in 10.CFR 50.46 and Appendix K as applied to a loss of Coolant Accident (LOCA) followed by a failure in the RSS. This exemption was necessary since test data indicated that the NSS may not perform adequately in a steam environment. As a condition of the Order,

t

' NUCLEAR REGULATORY CONNISSION 3

BIG ROCK POINT PLANT i

ECCS DESIGN MARGIN April 15, 1993 s

I r

the Commission required CPCo to provide test data showing that the existing l

NSS provided adequate spray distribution during LOCA conditions, or to modify the system to provide the required spray flow.

j l

CPCo conducted a test program to measure the NSS bundle spray oistribution in-a steam environment. Due to spray inadequacies, CPCo developed, tested and installed a new NSS which was composed of 12 small nozzles. The staff j

reviewed the test and design information of the new NSS and found the system to be acceptable [2].

An inherent assumption in the Commission's granting CPCo the one-cycle exemption (Cycle 14) was the adequate performance of the RSS. The staff was i

sent data (References [3] and [4]) which indicated that the ring spargers may also undergo performance degradation due to steam effects. Therefore, BRP was requested to substantiate the performance adequacy of the RSS prior to the Cycle 15 startup. This was not possible as Big Rock had no data that defined the RSS spray pattern in either an air or steam environment, and there was insufficient time to conduct tests to determine the spray pattern.

j CPCo requested a one-cycle exemption from the single failure criterion of 10 i

CFR 50.46 with respect to any '0CA followed by a failure of the redundant NSS. The exemption and supporting calculations were presented to the staff in Reference [5].

l The staff's evaluation of the requested one-cycle exemption for Cycle 15 [2]

was granted based on sevtral factors, and required the following [6]:

" Prior to Cycle 16 startup, CPCo must provide an evaluation of the RSS demonstrating acceptable performance at the anticipated LOCA environments, or modify the RSS such that acceptable performance is achieved; and" "If a new core spray sparger design is developed, the hydraulic characteristics of the ECCS must be evaluated to ensure adequate i

performance of both spray systems considering the most limiting single failure."

CPCo elected to design, fabricate and test a new ring sparger rather than to try to substantiate the adequacy of the existing sparger.

Early in the design effort, CPCo developed a conservative set of Minimum Acceptable Bundle Spray flows (MABS). These specified MABS were used by both GE and NUS in the design of the sparger for the ring spray system. The sparger design was'to deliver spray flow to every bundle at or above the MABS at all LOCA usage conditions. The MABS were defined to encompass the NSS design requirements as well.

ECCS INHERENT DESIGN MARGIN The following discussion focuses on the ECCS design margins. The spray flows used in the MABS specification, the time-to-rated spray (tas), the decay heat, the limiting LOCA analyses, and the Tech Spec maximum bundle power calculations are presented.

NUCLEAR REGULATORY COMMISSION 4

BIG ROCK POINT PLANT ECCS DESIGN MARGIN April 15, 1993 MABS l

As noted above the scope of the sparger design required that the ECCS be able to deliver spray flow to every bundle at or above the MABS at all LOCA usage conditions. The recorded spray flows to each bundle generally i

exceeded the MABS. However, two bundles received flows slightly below the MABS during certain sparger test runs. Consequently CPCo reviewed the test data for nozzle spray system as (evaluated by the staff in Reference [2]),

and noted that eleven bundles received spray flows slightly below the MABS and below the spray flows recorded from the ring spray system tests.

Of the thirteen bundles found to receive less flow than the MABS specifi-cation, the largest recorded deviation was found to be only 0.29 gam and only 1 channel recorded a value this high. That is, the most limiting test data failed to deliver the required minimum acceptable bundle spray flow by 0.29 gpm.

As a result of the ring and nozzle spray test results, CPCo developed a maximum bundle power technical specification for each bundle such that the performance adequacy of both the RSS and NSS designs could conservatively be justified.

L The allowable bundle power at each core location was derived based on the minimum of the lowest redundant nozzle flow for bottom-only and top / bottom combined steam entry (Reference CPCo letter 1977); the design flow rates for the sparger as described in Appendix I to Referenco [8]; the measured flows from the 75 psig confirmatory tests of the modified sparger,- Figure 26 of Appendix V to Reference [8]; and the CPCo defined MABS values.

The selected nozzle flow test data were the minimum flows that were recorded for bottom only and top and bottom combined steam entry mode configurations.

The top only steam entry mode was dismissed as the BRP design precludes any situation in which steam entry could occur in the top of the bundle only, during a LOCA. As was the case for the nozzle test data the most limiting sparger test data at 75 psig were used in comparison to the MABS. The pressure criterion is based on the BRP ECCS analysis which does not take credit for any spray cooling for reactor pressures greater than 75 psig.

Table 1 lists the average bundle flow for the nozzle and sparger flow tests and the MABS average values as well. Table 2 records the specific values for the 13 channels that registered test results slightly less than the MABS specification and demonstrates how the minimum values were selected for the Tech Specs.

Using the limiting test data a factor of 1.7 to 4.0 flow margin exists in the average core bundle spray flow rates assuming the minimum required : 0 gpm flow rate to each channel. Moreover, since most of the bundles receia the lowest flows at 75 psig and because reactor pressure would be significantly l

less than the 75 pound test condition for the DBA, additional margin exists in the reported spray flows.

j Tables 1 and 3 summarizes the margin associated with the MABS and the test flow data.

{

i i

)

l P.

' NUCLEAR REGULATORY. COMMISSION 5

l BIG ROCK POINT PLANT l

ECCS DESIGN MARGIN l

April 15, 1993 Table 1 l

AVERAGE BUNDLE SPRAY FLOW, Nozzle Spray Sparger Spray MABS Minimum Flow Test Flow Test Flow Used in Tech Specs (gpm)

(gpm)

(gpm)

(gpm) 2.06 3.98 1.73 1.72 Note:

l 1 1.0 gpm per channel is sufficient flow to provide adequate core cooling [7] and [15]. Note that the Staff never published a Safety Evaluation accepting the adequacy of the 1.0 gpm channel flow rate.

Table 2 i

THE 13 CHANNEL FLOWS LESS THAN MABS Channel #

Nozzle Sparger Design Minimum Flow Flow Flow

- Fl ow Used in Tech MABS Specs (gpm)

(gpm)

(gpm)

(gpm)

)

I 17 1.88 3.19 2.0 1.88 18 1.88 2.16 2.0 1.88 35 1.88 3.03 2.0 1.88 37 1.929 3.46 2.0 1.929 38 1.929 4.22 2.0 1.929 40 1.88 1.71 2.0 1.71 45 1.88 3.14 2.0 1.88 47 1.929 3.59 2.0 1.929 48-1.929 2.92 2.0-1.929 i

50 1.88 3.16 2.0.

1.88 67 1.88 2.66 2.0 1.88 i

68 1.88 3.89 2.0 1.88 78 2.07 1.58 1.6 1.58 4

e w

-e---c, m,m

.3. - - -r

.t-

' NUCLEAR REGULATORY COMMISSION 6

BIG ROCK POINT PLANT ECCS DESIGN MARGIN April 15, 1993 j

Evaporization/Soray Flow To justify the 10 CFR 50 Appendix K spray cooling heat transfer coefficients, CPCo had to ensure that each bundle received sufficient flow.

Prior to Cycle 15, in the review of the performance adequacy of the new redundant spray system (nozzle spray system), CPCo and the staff examined numerous reports regarding the Full Length Emergency Cooling Heat Transfer Experiments (FLECHT) and the spray flows predicted for other BWRs of various vintages.

It was determined that " vaporization" or " evaporation" flow could be 'Jefined for each fuel bundle such that vaporization of that amount of spray flow i

would remove the total amount of heat being produced by the bundle. There-j fore, as long as the bundle spray flow is greater than the vaporization flow, then the Appendix K to 10 CFR 50 spray cooling heat transfer coefficients are conservatively justified. No credit was taken for the spray water subcooling effects.

Using the maximum allowed radial peaking factors with uncertainty, the ratio of spray cooling to the evaporation flow rate, in gpm, for the limiting 0.375 ft' break with the ANS recommended decay heat were calculated for the j

different set of spray flow data. The average channel evaporation ratios were I

calculated to be 1.3,1.5, and 2.9 for the design, nozzle, and sparger spray l

flows, respectively.

I 1

The Reference [11] spray cooling tests conducted for a 12-foot long, 49 rod, 4 MWt bundle with spray rates between 2.0 and 0.5 gpm had shown virtually no change in the maximum clad temperature transient for spray flow rates between 2.0 and 1.0 gpm. The 2.0 and 1.0 gpm values correspond to spray flow / evaporation rate ratios for the test initial conditions of 1.13 and 0.57, respectively. Furthermore for the 0.5 gpm test only slightly higher peak clad temperatures were recorded.

Because the FLECHT tests used a 7x7 fuel geometry compared to Big Rock's design of Ilxil and the test chamber was not insulated, the Staff never published a Safety Evaluation supporting use of the 1.0 gpm flow data. However, References [7] and [15] maintain that a minimum flow rate of 1.0 gpm will provide adequate core cooling in the event of a LOCA. Table 3 summarizes the above calculated margin.

The Tech Spec maximum bundle power calculations are based on the MABS values and the radial peaking factor of 1.4 used in the Exxon LOCA analysis [9]

which is equivalent to 4.0 MWt (i.e., 240 MWt/84 bundles x 1.4).

Each cycle is determined to be bounded by the LOCA analysis by ensuring that the power is restricted to a value where the Maximum Average Planer Linear Heat Generation Rates (MAPLHGR's) are met. The 4.0 MWt Tech Spec criterion is satisfied by ensuring that the bundle power never exceeds the limit and further more an additional 1% margin and a numerical uncertainty is applied to this restriction. Therefore bundle core power is limited by the following:

Core Power /84 bundles x 1.4 + 1% Margin + numerical uncertainty < 4.0 MWt.

The evaporation rates calculated in the Big Rock physics analysis incorpora-ting rating, the 1% administrative limit, and the numerical uncertainty predicts evaporation flows about 7% less than those calculated for the 1.3, 1.55, and 2.97 design, nozzle, and sparger average channel ratios respective-ly. Hence the spray flow-to-evaporation margin is understated.

' NUCLEAR REGULATORY COMMISSION ~

7 BIG ROCK POINT PLANT l

ECCS DESIGN MARGIN i

April 15, 1993 l

Time-to-Rated Sorav - tm The ECCS analyses, References 9 and 10, assumed no credit for spray cooling.

until the " time of rated spray" - to had been reached. This is the time interval at which the ECCS spray system motor operated valves (M0Vs) have received the comand to open and have traveled to the full open position.

Once the rated spray was reached the analyses employed the spray cooling heat transfer coefficients allowed by Appendix K of 10 CFR 50.

l 3

The tm value used in the Tech Spec maximum bundle power values is 26.9 seconds. The BRP LOCA ECCS MAPLHGR's are based on the limiting 0.375 ft2 split break. This particular break size requires rated core spray in 61.5 l

seconds. The Peak Clad Temperature (PCT) for the 0.375 ft2 is at a maximum due to the extended low flow in the core.

Flow stagnation occurs very early in the core which leads to early dryout followed by a long period of high-quality low-flow cooling. For larger break sizes the time interval of poor heat transfer is shortened and slightly.better blowdown heat transfer were calculated.

For smaller breaks significantly improved early blowdown heat transfer were calculated thereby resulting in lower PCTs.

Hence the time at which spray cooling must exceed the evaporation rate is notably _

later than the presently assumed 26.9 second value. Consequently by using the earliest to the highest bundle power which must be removed by spray i

flow was calculated. Using the ANS decay heat with 20% uncertainty, 16%

less spray flow is required to remove decay heat for the limiting PCT 0.375 ft2 LOCA. Table 3 summarizes the margin associated with the time-to-rated spray flow assumptions.

t The BRP LOCA analyses conservatively assumes no credit for any. spray flow.

prior to the " time of rated spray", though considerable spray would be 1

available.

Furthermore the tm value of 26.9 seconds assumes a 15 second i

delay for full valve stroke; however, design flow would be-available in I

less than the valve full stroke time. The effective flow area necessary to satisfy design flow is approximatly 60% of the valve stroke time, thus significant cooling would be available prior to the time of rated spray.

ANS Decay Heat l

The decay heat was calculated based on the correlations of the ANS standard I

" Decay Energy Release Rates Following Shutdown of Uranium-Field Thermal Reactors" supplemented by the calculated decay heat contribution of principle L

radioactive actinides Uranium-239 and Neptunium-239. The ANS, +20% for uncertainty, decay-heat fraction for infinite exposures was used.in the maximum bundle power calculations. These data were compared to the best--

estimate 'recomended' decay-heat for infinite irradiation time data listed in Appendix, page D-6, of Reference.[10]. Comparing the recomended decay -

heat to the ANS, +20% for uncertainty, and assuming the. limiting te value of 26.9 seconds results in a 25% margin in the bundle cooling flow.

Comparing the recomended decay heat and using the te value of 61.5 seconds associated with the limiting LOCA analysc:4 to the Tech Spec maximum bundle power data results in a 48% margin in bundle cooling flow. Table 3 summarizes the decay heat assumptions with respect to flow margin.

\\

~ NUCLEAR REGULATORY COMMISSION 8

i BIG ROCK POINT PLANT

• ECCS DESIGN MARGIN April 15, 1993 Summary ECCS Inherent Desian Marain i

Table 3 summarizes the specific design margins used in the maximum bundle power Technical Specifications.

i Table 3 EXAMPLES OF MAXIMUM BUNDLE _ POWER INHERENT _ MARGIN I

_ Assumption Descriotion Estimated Marain

[1]

The minimum bundle spray flow was Ratio of Actual determined by comparing the to minimum flow-i following:

used in the maximum bundle power calculations:

i Spray flow from NSS test 1.26 Spray flow from RSS test 2.41 MABS 1.01

[2]

Decay heat with a 1.2 reduction factor

_ Ratio of ANS,+20%

compared to the ANS recommended decay uncertainty, to heat for the follow ' time-to-rated ANS recommended spray values':

decay heat:

Tech Spec Value - 26.9 sec 1.25 Limiting 0.375 ft2 LOCA - 61.5 sec 1.27 i

[3]

Combined margin in decay heat and time to 1.48 rated flow for the limiting PCT LOCA:

l

[4]

Ratio of the average Nozzle, Sparger, and MABS flows-to the evaporation rate assuming the RETRAN based maximum radial peaking factors with uncertainty for the limiting 0.375 ft2 peak clad temperature LOCA; assuming reactor pressure at 75 psig, decay heat without uncertainty, and a tas of 61.5 seconds:

Nozzle 1.5 Sparger 2.9 Note that each cycle's maximum bundle power is further restricted by that additional 1%

administrative limit.

i t

I i

I i

l

\\

~ NUCLEAR REGULATORY COMMISSION 9

BIG. ROCK POINT PLANT ECCS DESIGN MARGIN April 15,1993 l

PALISADES FLOWNET MODEL VALIDATION / BENCHMARK The Palisades Service Water System FLOWNET Model Validation / Benchmark Report, dated October 1988 stated that "the summation of the measured component flows was 3 to 6 percent below the implied pump flow." The NRC staff assumed interpretation of this statement is that FLOWNET overpredicted flow by 3 to 6 percent.

Su'osequent to the Staff's concerns CPCo reviewed the benchmark and test and test reports for the three Service Water System tests.

Test Historv During the 1988 test, temporary flow elements were added to some parallel flow paths in the Palisades Service Water System. The pipe sections containing the Shutdown Cooling Heat Exchangers (E-54A and E-548) did not include flow elements. Therefore, the flow rates for those components were determined by measuring the pressure drop across the heat exchangers.

Since the measured pressure drop was about 5 psi, a water gauge was used to obtain the measurement test data. After 1988, flow elements were added to both heat exchangers. The new test data significantly lowered the flow measurement uncertainty for the heat exchangers.

Re-Reviewino the Bench Mark Test Data For each test, the benchmark report presents three system flow rate values.

The first flow rate is simply a sum of the measured and assumed flow rates for each branch in the service water system. The second flow rate is the value calculated by the FLOWNET computer program. Using the measured pump i

head, the data calculated the third flow rate using' the manufacture's pump head-flow curve.

For base test configuration (step 5.4.10 of the benchmark report), the flow rate predicted by FLOWNET'is one percent lower than the third flow rate value. This bias in the flow rates is consistent for all tests when one service water pump is running. This data suggests that the FLOWNET computer code can successfully predict the general performance of the service water system. The 3 to 6 percent difference occurs between the first and second flow rate values. The measured pressure increase across the pump provides the most accurate flow rate value. Since this flow rate is consistently higher than the sum of the measured flow rates, possible flow losses not accounted for in the FLOWNET model were investigated.

In the benchmark analysis for base test configuration, the flow rates through the isolated Containment Air Coolers and the Nontritical Service Water header are both assumed to be zero. The containment air coolers are isolated using a sixteen inch butterfly valve. This flow branch does contain two independent flow elements. For a design flow rate of 7000 gpm, the pressure drop across the flat plate orifice would be about 200 inches of water.

When the flow rate decreases to 250 gpm, the pressure drop will be about 0.3 inches of water. The pressure drop would only be 0.04 inches of water for a flow rate of 100 gpm. Therefore, the-flow rate through the isolated containment air coolers could exceed 100 gpm with no indication from the flow element. This

i NUCLEAR REGULATORY CONNISSION 10-BIG ROCK POINT PLANT

- ECCS DESIGN NARGIN April 15, 1993 leak path does not exist when the control valve is full open or throttled partially open. A second leak path is present for all test configurations.

The non-critical service water header is isolated using a twenty-four: inch l

diameter butterfly valve. There are no flow measuring devices in that flow path. Considering the size of the butterfly valve, the existence of a large pressure drop of about 40 psi and twenty years of service, a large leakage flow rate through.this flow path is very possible. Therefore, the flow rate differences are not a result of errors in the FLOWNET program, but due to leak i

paths not considered during the test data analysis.

i CONPARISON OF RETRAN AW FLOWNET j

The computational scheme in FLOWNET was checked using the RETRAN computer code. Since RETRAN uses the pressure calculated at the midpoint of the volume, some small volumes near pipe junctions to achieve the same modeling scheme used in the FLOWNET computer code were added. RETRAN uses smooth tube wall friction. To ensure a one-to-one comparison, the FLOWNET code was executed using a pipe roughness for a smooth tube. Using the same volume and junction input data for both codes, the results were nearly identical. 'This comparison demonstrated that the FLOWNET computer code models and solution method are acceptable and the code can accurately predict the ECCS l

l performance.

l BRP FLCWNET SENSITIVITY ANALYSES l

i A sensitivity analysis, Reference [13], of the BRP ECCS flow model, Reference

[12], was performed. Piping surface roughness and nozzle spray coefficients l

were analyzed.

i Two cases were analyzed for the surface roughness analysis. The first analysis investigated the roughness factor for a 50% cast iron and 50%

stainless steel system, the configuration of the BRP ECCS system. The second calculation examined a 100% cast iron system. The results indicated a 1 percent increase in flow for the case 1 analysis and a 2% reduction in flow for the case 2 assessment. Hence the baseline BRP ECCS FLOWNET model used a more conservative roughness factor than the actual configuration of 50% cast iron and 50% stainless steel geometry.

The sensitivity of the experimentally derived loss coefficients for the ring spray and nozzle spray tests with respect to the ECCS baseline requirements were analyzed. The largest pressure drop in the ECCS geometry i

is across the nozzles. The results indicated that a 5.7% increase in the:

ring spray nozzle friction loss coefficient would result in spray flow to the core slightly in excess of 292 gpm. A 49% increase in the nozzle spray friction loss coefficient would similarly result in a flow rate slightly in excess of the 296 gpm requiremev. Hence margin exists in:the key system experimentally derived friction loss coefficients.

CONCLUSIONS Considerable margin exists in the BRP LOCA analysis assumptions, the MABS specification, and the nozzle and sparger design flow requirements of 292

i i

' NUCLEAR REGULATORY CONNISSION 11 BIG ROCK POINT PLANT

• ECCS DESIGN MARGIN j

April 15, 1993 I

and 296 gpm.

Furthermore, over 20% margin exists in the minimum flows used in the Tech Spec maximum bundle power calculations when compared to the nozzle and sparger spray limiting test data respectively. Over 45%

margin exists when using the limiting 0.375 ft2 LOCA analysis tns value and assumed decay heat. Over 5% margin exists in the experimentally derived ring spray nozzle loss coefficient and over 45% margin exists in i

the experimentally developed nozzle spray loss coefficient.

Slight margin exists in the ECCS pipe roughness factor. Added margin exists in the form of the administrative limits imposed on the maximum allowed bundle power, that is, the 1% margin and the numerical uncertainty applied to the 4.0 MWt specification.

Because of these significant conservatisms, the comparison results of the l

FLOWNET code to the RETRAN code, and the additional information from the l

l Palisades FLOWNET benchmark data, it can be concluded that considerable margin i

exists in the BRP ECCS design.

L Similarly the NRC Staff concluded in the following excerpts:

{

"CPCo developed a maximum bundle power technical specification for each bundle such that the performance adequacy of both the RSS l

and NSS can conservatively be justified",

i in the Reference [6] License Amendment 26 report, and:

l-

"The redundant NSS is a fully-tested system that provides more than l

adequate flow to each fuel assembly in the anticipated LOCA steam environment. The calculations used to substantiate the system adequacy are quite conservative. Also, there is significant spray flow before l

the time that the ECCS model assumes spray cooling heat transfer",

i as stated in Reference [14].

Therefore, using FLOWNET to validate the fire pump monthly test results, i

thus ensuring that the Big Rock ECCS model satisfies the licensing ~ basis; performance requirements of 292 and 296 gpm, is appropriate.

REFERENCES

1. Memorandum and Order, by the Commissioners, NRC In the Matter of CPCo, Big Rock Point, dated May 26, 1976.

2. Letter to David A. Bixel CPCo, from Don K. Davis NRC, dated October 17, 1977 (

Subject:

Amendment #15 (i.e., Cycle 15 startup))

3.- Effects of Steam Environment on BWR Core SDray Distribution, Amendment 3 to NED0-20566, April 1977.

4. Ma Rock Point Core SoraY Test Report. Sinole Nozzle Test and Development Procram, NUS-3005, NUS Corporation, August 1977 (Included as attachment to the letter from W. S. Skibitsky, CPCo to Samuel J. Chilk, Secretary to the Commission, NRC, dated August 9, 1977).-

~

NUCLEAR REGULATORY CONNISSION 12 BIG ROCK POINT PLANT ECCS DESIGN MARGIN i

April 15,1993 REFERENCES (continued)

5. Letter from David A. Bixel CPCo, to Director of NRR NRC, dated September i

15, 1977 (

Subject:

Request for exemption).

6. Letter to David A. Bixel CPCo, from Dennis L. Ziemann NRC, dated April 10, 1979 (

Subject:

Amendment #26 (i.e., Cycle 16 start.4p))

' 7. Letter from David A. Bixel CPCo, to Director of NRR NRC, dated September

[

19, 1977 (

Subject:

Adequacy of the Redundant Core Spray System).

8. The Bio Rock Point Soarcer Rino Test Proaram,'NUS Corporation, NUS 3234, (GE Report NEDC-21974), dated September, 1978.

9. Exxon Nuclear Company, Bio Rock Point LOCA Analysis Usino The Exxon Nuclear Company WREM NJP-BWR ECCS Evaluation Model. MAPLHGR Analysis, XN-NF-79-21, Revision 1, April 1979.

10. General Electric Company. Power Generation in a BWR Followino Normal Shutdown of Loss-Of-Coolant Accident Conditions, NED0-10625, March 1973.

{

11. " Core Sorav and Core Floodina-heat Transfer Effectiveness in a Full-l Scale Boilina Water Reactor Bundle," APED-5529, June 1968.

I l

i

12. " Hydraulic Analysis of Emeraency Core Coolina System", EA-BRP-

~

ECCS-990916-GFP, Sepetember 21, 1988.

l

13. "FLOWNET Sensitivity Analysis on Pipe Rouchness," EA-NAP 92-13, l

October 27, 1992.

14. Amendment No. 15 to License DPR-6 dated October 17, 1977.

15. Letter from M. A. Costandi's GE, to S. S. Chan r v.0, dated August 12, l

1977.

. /qf sna )'/J L

Patrick M Donnelly Plant Manager l

l CC: Administrator, Region III, USNRC NRC Resident Inspector - Big Rock Point I

w r

w, e

~, + -

e --,