ML20211G326

Forwards Revised Wind & Tornado Risk Analysis & Response to 860718 Request for Addl Info.Analysis Revised to Include Auxiliaru Shutdown Bldg Installed Per 10CFR50,App R.Encls Submitted to Complete SEP Topics III-2 & III-4.A

ML20211G326
Person / Time
Site:	Big Rock Point File:Consumers Energy icon.png
Issue date:	02/13/1987
From:	Frisch R CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
TASK-03-02, TASK-03-04.A, TASK-3-2, TASK-3-4.A, TASK-RR NUDOCS 8702250346
Download: ML20211G326 (62)

Text

,, ;

Consumers

@moeraRuus AEKM4Ar5 PRBERE55 Power General offices: 1946 West Parnell Road, Jackson, MI 49201 . (517) 788-0550 February 13, 1987 Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 DOCKET 50-155 - LICENSE DPR BIG ROCK POINT PLANT -

INTEGRATED PLAN ISSUE BN-025, WIND AND TORNADO LOADING, REVISED RISK EVALUATION AND RESPONSE TO REQUEST FOR ADDITIONAL-INFORMATION Consumers Power Company letter dated August 29, 1986, provided the NRC with

' Integrated Plan Semi-Annual Update No 5. In this Integrated Plan update Consumers Power Company committed to revise the risk analysis associated with Issue BN-025, Wind and Tornado Loading, to include the new auxiliary shutdown building and to respond to the NRC Staff's request for additional information letter dated July 18, 1986. This submittal provides the requested information.

The alternate shutdown building was installed as a part of 10CFR50 Appendix R modifications and was designed to be resistant to the effects of winds and tornadoes up to 250 mph. Extensive descriptions of the building and the equipment which is housed within the building has been provided as part of our Appendix R fire protection submittals. A brief description of the building, particularly those aspects important to wind and tornado and tornado missile resistance is included in the attachment to this submittal. The design basis for this structure was chosen as 250 mph because it is equal to the analyzed resistance of the containment building presented in our August 3, 1982, submittal and the NRC Staff's December 9, 1982, Safety Evaluation Report.

The attached risk analysis provides the following insights with respect to the capacity of the Big Rock Point Plant to withstand wind and tornado events.

With the installation of the alternate shutdown building, the reactor can be brought to hot shutdown and maintained in that condition for approximately four hours following wind and tornado and tornado missile events associated with wind speeds up to 250 mph. Long term cooling, however, depends on continued makeup to the emergency condenser. This is accomplished by use of fire system makeup to the emergency condenser. Any makeup to the primary coolant system which may be required due to losses to the main condenser, shrinkage and normal leakage is also accomplished with the fire system. The use of the fire system is dependent on operation of a fire pump located in the screenhouse. According to Consumers Power Company's analysis of August 3, 1982, the screenhouse is resistant to wind and tornado loadings up to 150 mph.

No tornado missile resistance is assumed for this structure because its roof is made up of a steel deck.

8702250346 870213 OC0287-0018A-NLO2 PR ADOCK 05000155 PDR _ ,

j

r i Nuclear Regulatory Commission 2 Document Control Desk Integrated Plan Issue BN-025, Wind and Tornado Loading, Revised Risk Evaluation and Response to Request for Additional Information February 13, 1987 It should be noted that there is one potential vulnerability of alternate shutdown system equipment to tornado missiles generated by events less than 250 mph. This is a small area in the vicinity of the equipment lock where power and control cabling for main steam isolation and the emergency condenser are located. This area is potentially vulnerable to a missile strike in that the top of the area is protected by a steel plate as opposed to reinforced concrete and the containment shell is on its remaining sides. The risk analysis takes this vulnerability into account. The importance of this area to wind and tornado and tornado missile risks is minimal for several reasons.

First, there are no active components in the area which will fail if exposed to the elements associated with the storm. Further, all conduit and junction boxes provided are NEMA weatherproof rated. Additionally, the area is small (less than 150 square feet) and the surrounding structures into which the circuitry is recessed provide a significant degree of shielding. For these reasons and for reasons similar to the NRC Staff's analysis of the core spray pump room in its November 29, 1982, missile evaluation, the risks associated with this limited area to the effects of winds and tornadoes is considered to be small.

Given these particular plant design features, the risk analysis indicates that the potential for core damage as a result of winds and tornadoes and tornado missiles is approximately 3.0 E-5/yr for the plant as it is currently de-signed. This is similar to or less than the risks associated with other rare events such as loss of coolant accidents or ATWS.

It should be noted that the attached risk analysis is considered to be conservative. That is, the derived core damage frequency for winds and tornadoes is higher than would be generated by a more realistic analysis and the perceived benefit of any backfits directed at wind and tornado risks is artificially enhanced as a result. The major conservatisms inherent in the risk analysis include:

- Hazards curves which include wind and tornado statistics from areas located in the southern part of Michigan.

- Assumptions that the functioning of all active components within a structure are lost any time the internal areas of the structure are i

exposed to the wind and tornado environment and that all active and passive components within a structure fail on the entry of a single missile.

- The use of deterministic capacities rather than fragility analysis to determine the capabilities of the various structures around the site.

- The lack of credit taken for missile resistance provided by reinforced concrete structures less than 12 inches in thickness.

Consumers Power Company also recognizes, however, that there are large uncer-tainties associated with external event risk analyses such as the one in the attachment to this submittal. Principally:

OC0287-0018A-NLO2

e 1 Nuclear Regulatory Commission 3 Document Control Desk Integrated Plan Issue BN-025, Wind and Tornado Loading, Revised Risk Evaluation and Response to Request for Additional Information February 13, 1987

- The initiating frequency of high winds or tornadoes can vary by one to two orders or magnitude from best estimates.

- Treatment of missiles is dependent on estimations of missile type and number and extrapolation of generic probabilistic missile evaluations.

With these uncertainties, Consumers Power Company proposed, in addition to a risk analysis, to determine the maximum wind speed at which safe plant shut-down could be achieved and to establish the recurrence interval associated with that event. The NRC Staff, in Sections 4.5 and 4.8 of the Integrated Plant Safety Assessment Report (IPSAR) additionally requested that alternative corrective actions ba identified to assure that the plant was capable of withstanding an NRC derived 1.0 E-4/yr to 1.0 E-5/yr range event at the upper 95% confidence limit. Cost-benefit analyses were to be provided to support any proposed modifications.

Using the NRC derived hazards curve for winds and tornadoes (Figure 6 of the NRC Staff's letter dated December 17, 1980, Safety Evaluation Report for SEP Topic.II.2.A Severe Weather Phenomena), the NRC Staff suggests that the Big Rock Point Plant be capable of withstanding wind and tornado and tornado missile loadings associated with wind speeds of 150 mph to 210 mph.

As noted from the results of the risk analysis, the Big Rock Point Plant as currently designed exceeds the NRC Staff's criteria for short term cooling conditions (the first four hours of the event) in that it is capable of withstanding events up to 250 mph. For long term cooling (on the order of a day or more) the plant just meets the suggested criteria for wind and tornado loads in that the screenhouse capacity falls into the lower end of the required range at 150 mph. The NRC Staff's suggested criteria is not met, however, for tornado missiles in that the screenhouse roof is assumed to provide minimal protection from missiles.

Among the modifications which could potentially reduce the vulnerability of the Big Rock Point Plant to tornado missiles is the installation of a portable pump capable of bypassing failures in the fire system which may occur in the screenhouse. The pump would be connected to the fire system through any of the hydrants around the site and would draw water from the intake bay or the lake to provide makeup to the emergency condenser and/or reactor, if neces-sary. The pump would be stored in the alternate shutdown building in order to provide the necessary protection from the wind or tornado event. The portable pump is a preferable backfit to providing missile protection to the screen-house because it is believed to be a less extensive modification and the availability of long term cooling systems is raised to well beyond the analyzed capacity of the screenhouse. On installation of this backfit, long term cooling capabilities would be provided for wind and tornado events up to the capacity of the alternate shutdown building (250 mph).

The risk analysis suggests that the portable pump modification reduces the risks associated with wind and tornado events by approximately a factor of two, to near 1.5 E-5/yr. At an estimated $100,000 for procurement, installa-tion, engineering and future testing and maintenance, the risk assessment suggests that the modification falls short of being cost-beneficial by an OC0287-0018A-NLO2

o e Nuclear Regulatory Commission 4 Document Control Desk Integrated Plan Issue BN-025. Wind and Tornado Loading, Revised Risk Evaluation and Response to Request for Additional Information February 13, 1987 order of magnitude beyond the traditionally used $1000/ person-rem measure of value impact.

However, Consumers Power Company, at this time, commits to the procurement of a portable pumping capability for the purpose of reducing risks associated with winds and tornadoes. This commitment is made to reduce the uncertainties associated with risk analysis of external hazards auch as winds and tornadoes, and because there is potential benefit to such a pumping capability in other accident scenarios which go beyond the design basis of the plant (such as flocding). Also, because the risk analysis assumes the auxiliary shutdown building is manned, Consumers Power Company further commits to revising procedures to provide guidance to the operator as to the weather conditions under which the alternate shutdown building should be manned.

This proposal was presented to the Technical Review Group at their last meeting and was ranked at that time. A schedule for the implementation of this modification will be provided as a part of the next Integrated Plan update.

It is recognized that this backfit also reduces the importance of some of the differences between the NRC Staff and Consumers Power Company as to the capacity of various structures at the Big Rock Point site. On installation of a portable pump, the capability of the plant to provide safe shutdown follow-ing a wind or tornado will depend primarily on the capacities of the contain-ment (with which the NRC Staff and Consumers Power Company are in agreement) and the alternate shutdown building. Consumers Power Company used the capaci-ties derived for the various buildings in our December 3, 1982, submittal in performing the attached risk analysis and believes they are defensible. In light of our commitment to provide portable pumping capabilities, if the NRC Staff still feels a need to resolve the differences between the NRC and Consumers Power Company's calculations of the capacities of structures other than the containment or alternate shutdown building, Consumers Power Company will provide additional information.

Other than those questions directed at the capacities of the various structures at the site, Appendix F in the attachment to this letter contains responses to the NRC Staff's request for additional information which was included as Enclosure 2 to their July 17, 1986, letter. Unless the NRC Staff requests additional information with respect to the capacities of the structures at the Big Rock Point site, this submittal completes the evaluation of SEP Topics III-2 and III-4.A concerning winds and tornadoes and tornado missiles.

% Ralph R Frisch Senior Licensing Analyst CC Administrator, Region III, NRC NRC Resident Inspector - Big Rock Point Plant Attachment OC0287-0018A-NLO2

_ . - - - - ~

r .t ATTACHMENT Consumers Power Company Big Rock Point Plant Docket 50-155 REVISED WIND AND TORNADO RISK ANALYSIS AND c RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION February 13, 1987 i

I i

l i

61 Pages j

OC0287-0018A-NLO2

e r 1

INTRODUCTION AND OVERVIEW To evaluate the Winds and Tornado SEP topics for the Big Rock event tree methodology will be employed. Taken from the original Big Rock Point Risk Assessment, the Loss of Offsite Power (LOSP) Event Tree with some simplifi-cations will be used as a basis for determining the relative risk of winds, tornadoes, and missiles at the Big Rock Point site.

The LOSP event tree was chosen as the most conservative tree due to the i assumed failure of the AC powered equipment. For this analysis, it will be assumed that all normal AC powered equipment is unavailable for the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of the transient. It should be noted that should station power not be lost there may exist equipment that would be lost only upon damage to the building in which it is housed, for instance, the main condenser circulating water pumps and the electric fire pumps. The assumption of losing offsite power greatly limits the equipment with which the operator has available to shut the plant down. Therefore, the LOSP was simplified to just the follow-ing headings:

i 1) Main Steam Isolation Valve (Primary System Isolation).

I

2) Turbine / Condenser Valves (Primary System Isolation).

3) Emergency Condenser Activation (decay heat removal).

4) Emergency Condenser Make-up (decay heat removal).

i

5) Primary System Inventory Make-up.

The event tree and a more detailed description of the tree and the headings is in Appendix B. The tree was composed with the following assumptions:

l

1) Offsite power is lost.

2) An operator has manned the Alternate Shutdown Building.

3) A 24-hour mission time is used.

l The assumption that an operator has manned the Alternate Shutdown Building was made knowing that currently there are no criteria available for the operators to use in judging when to man the building in the case of high winds or tornadoes. This assumption was made to insure control of the MSIV and emergency condenser outlet valves is available through the Alternate Shutdown Building thus reducing the importance of the outer electrical cable penetration room.

The mission time of the event tree, 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, was chosen for several reasons. Some of the sequences involve random failure of the main steam isolation valve. In these sequences if the primary system is isolated by the turbine / condenser valves, make-up water is necessary within the first

! 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to prevent the low reactor water level from being reached (2'9"

, above the core). Note that no make-up is required within the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />

) if the MSIV is closed early in the transient. Also, repair and recovery i actions are assumed available 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the winds or tornadoes strike the area, in which case offsite power may be restored or equipment damaged or assumed inoperable may be in service.

OC0287-0018A-NLO2

o o 2

Further, below a wind speed of 80 mph, no damage to the Big Rock Point Plant buildings or equipment occurs. The original design criteria for the buildings was an 80 mph wind. The plant has also experienced wind gusts up to 80 mph in the past. Between wind speeds of 80 mph and 110 mph the original LOSP event tree quantification will be used. A change will be made to the initiating event frequency from the total LOSP frequency to the frequency of the plant site experiencing winds in the interval 80 to 110 mph. The 110 mph wind speed is that which fails portions of the turbine building and thus the first pieces of equipment begin to fail due to winds.

The event tree was quantified at various wind speed intervals. These intervals were chosen based upon the wind speeds at which the various buildings fail. The failure of a building was assumed to lead to the direct failure of the equipment housed within. The intervals are:

1) Below 80 mph No damage is assumed to occur, original LOSP event tree quantification governs.

2) 80 to 110 mph Still no damage to plant buildings or equipment, LOSP j initiating frequency changed to frequency of winds in l this interval with missiles included as additional failure mode of the equipment (NOTE: Tornado missile frequencies below 80 mph were applied to this interval and were conservatively assumed to occur at straight wind frequencies).

3) 110 to 150 mph Damage to the turbine building is assumed to take out the demineralized water pump (which removes one source of emergency condenser make-up water) and emergency AC and DC power sources.

4) 150 to 190 mph Damage to the screenhouse occurs at 152 mph, the diesel fire pump is no longer available, and two sources of emergency condenser make-up have been lost as well as one PCS make-up source.

5) 190 to 250 mph Damage to the emergency diesel generator room occurs at about 193 mph and thus, a second source of PCS make-up is lost (control rod drive powered by the EDG through the alternate shutdown area).

6) Above 250 mph The containment building is assumed to fail at about 250 mph, therefore no credit for any available miti-gating systems is taken above this wind speed.

There are two quantifications performed in Appendix A, one using the current plant set up and the second taking credit for the proposed portable pump modification. The portable pumps can provide make-up water to both the primary system and the emergency condenser. The pumps will be housed within the Alternate Shutdown Building and thus are protected from winds and tornadoes as long as the building stands (a more detailed description of the Alternate Shutdown Building is in Appendix E). Since the Alternate Shutdown Building is capable of withstanding winds, tornadoes, and missiles above 250 mph the portable pumps will provide make-up through all wind speed intervals. The as-is quantification stops taking credit for mitigating systems above a wind speed of 150 mph due to no emergency condenser make-up OC0287-0018A-NLO2

3 water being available for long-term cooling. For the quantification above 150 mph, the core damage probability is equal to the frequency of experiencing winds in the respective interval. With the portable pumps avail-able, water is now available up to 250 mph and mitigating systems are used in the quantification.

For equipment relied upon during a particular wind speed interval, the random equipment failure rates coupled with the appropriate opetator error in using the equipment were used in the quantification. In addition, the effects of tornado missiles were added as an additional failure mode of the equipment.

These failure modes were determined using the EPRI target impact results from their Report NP-768. The impact values were then scaled appropriately for the differences between Big Rock and the generic EPRI plant. Thus, the effects of missiles were effectively included as a failure mode in the system (fault tree) logic models.

To do the quantifications, the probability for experiencing winds in any interval was determined by taking the probability of exceeding the lower wind speed of the interval and subtracting the probability of exceeding the upper wind speed of the interval. As an example, the interval from 80 to 110 mph is calculated below:

Interval ,

Probability of ,

Probability of Probability exceeding 80 mph exceeding 110 mph

= 0.01/ year -

7.0 E-4/ year

= 0.0093/ year The wind speed probability values are taken from Figure 6 located in Appendix C, and are the best estimate values.

The missile impact probabilities taken from Table 3 of Appendix C were

! applied to the Big Rock buildings to determine the additionel risk due to tornado missiles. The values in Table 3 were determined in EPRI Report 768.

These data will be used to determine the failure probability of the buildings i at Big Rock excluding the containment building and the alternate shutdown l building.

The impact probability for any particular wind interval is first multiplied by the ratio of Big Rock Point to EPRI generic plant wind frequencies for the interval. This is to account for the plants being in different tornado frequency regions. Next, this value is divided by the total surface area of the Big Rock buildings open for missile strike. Finally, the wind frequency for Big Rock is divided into this value to obtain a missile strike frequency for the Big Rock site. By multiplying this by a specific building surface area and the postulated number of missiles, a damage frequency is obtained for any piece of equipment housed within the particular building being analyzed. It should be noted that a conservative estimate is being used

! since only the impact frequency is being used from the EPRI report. Because l some of the buildings are made of concrete walls, a missile impact does not necessarily mean structural damage and therefore, the damage frequencies reported in the EPRI study may be used.

Missile damage was looked at for the buildings housing the equipment which is required in the event tree headings.

OC0287-0018A-NLO2

, e 4

A summary of results for the various wind speed intervals is shown below.

These results were obtained using the best estimate wind speed probabilities from Appendix C, Figure 6.

Wind Speed No Portable Pucp3 Portable Pumps Interval Available _ Available 80 to 110 mph 9.53 E-7 9.53, E-7 110 to 150 mph 1.58 E-5 8.92 E-6 150 to 190 mph 7.0 E-6 9.85 E-8 190 to 250 mph 2.6 E-6 6.46 E-8

>250 mph 4.0 E-7 4.0 E-7 2.67 E-5 1.04 E-5 i

i l

OC0287-0018A-NLO2

9 t APPENDIX A Event Tree Quantification t

I t

i OC0287-0018A-NLO2

a e l

1 QUANTIFICATION

SUMMARY

NO PORTABLE PUMPS Best Estimate 95% Limit 80 to 110 mph 9.53 E-7/yr 5.53 E-5/yr 110 to 150 mph 1.58 E-5/yr 1.35 E-4/yr 150 to 190 mph 7.0 E-6/yr

• 8.0 E-5/yr 190 to 250 mph 2.6 E-6/yr

• 1.3 E-5/yr

> 250 mph 4.0 E-7/yr

• 7.0 E-6/yr 2.67 E-5/yr 2.90 E-4/yr

• For these wind speed intervals, the core damage frequency is set equal to the frequency of experiencing winds in the respective interval. This is because above a wind speed of 150 mph the screenhouse is assumed failed and with it the diesel fire pump has failed removing the only available cooling water makeup to the emergency condenser, without a heat sink the core is assumed damaged with eventual release from containment.

PORTABLE PUMPS AVAILABLE Best Estimate 95% Limit 80 to 110 mph 9.53 E-7/yr 5.53 E-6/yr 110 to 150 mph 8.92 E-6/yr 7.65 E-5/yr 150 to 190 mph 9.85 E-8/yr 1.13 E-6/yr 190 to 250 mph 6.46 E-8/yr 3.23 E-7/yr

> 250 mph 4.0 E-7/yr 7.0 E-6/yr 1.04 E-5/yr 9.6 E-5/yr 1

OC0287-0018A-NLO2

2 Loss of Station Power Baseline Quantification Adjustments For Initiating Frequency And Missiles The LOSP event tree has been quantified for other reports and is only summarized here. The core damage frequency is determined primarily from ten sequences. These sequences are listed below with their respective core damage frequencies. Column one shows the sequence value with the current LOSP initiating frequency of 0.13/yr and column two shows the value using as the initiating frequency the frequency of experiencing winds between 80 and 110 mph which is 0.0093/yr.

1) P Z Fs Ycrd Lp 5.91 E-7 4.23 E-8

2) P Z Fs Ycrd Cs 1.39 E-6 9.94 E-8

3) P Z Fs Ycrd Cr 5.93 E-7 4.24 E-8

4) P Z Fs Lp 1.07 E-7 7.65 E-9

5) P Z W Fs Cs 1.07 E-6 7.65 E-8

6) P Z Q Fs Cr 3.27 E-7 2.34 E-8

7) P Z Q Fs Em F1 Lp 1.10 E-7 7.87 E-9

8) P Z Q Fs Em Cs 1.48 E-7 1.06 E-8

9) P Z Q Fs Em Cr 1.41 E-7 1.00 E-8

10) P Z Ev Fs th 3.72 E-7 2.66 E-8 4.85 E-6 3.47 E-7 One method of accounting for the increase in core damage probability due to 4 missiles is to include a factor for this missile damage into the system overall failure rates. For the headings listed in the above sequences, those vulnerable to missiles are: Cs, Cr, Em, and Yced. These headings all require equipment that is assumed inoperable should a missile damage the I building housing the required equipment.

To adjust the above sequence quantifications to include missile strikes, the first value needed is the probability that a missile will strike a building at Big Rock. This can be determined from the target impact results included in EPRI Report NP-768 and shown in Table 3 of Appendix B. In order to adjust this value for Big Rock several factors are needed, the ratio of Big Rock Point wind frequency to the EPRI plant frequency for the interval in

! question, the total Big Rock building surface area open for missile strike, I and the surface area of the specific building.

i The calculations will be as follows: 1) multiply the wind interval frequency by the ratio defined above, 2) dividing this by the Big Rock building area

! and then by the Big Rock specific wind frequency, 3) then multiply by the building specific area and the number of missiles. This yields a unitiess number that will be used to represent the probability of equipment damage due to missiles.

There are four wind intervals which are considered. These intervals were l chosen based upon wind speeds at which the buildings failed and the breakdown

( of the tornado characteristics. The intervals are shown below with the calcu-

! lations. The value shown need only be multiplied by the specific building l surface area.

l

! OC0287-0018A-NLO2

3 WIND SPEED RANGE 1.62 E-8/yr 2.09 E-5/yr 1 72-112 mph ----------- * ----------- * -----------

• 3000 126,600sqft 8.17 E-4/yr 2.09 E-5/yr

= 4.72 E-7/sqft 3.44 E-8/yr 9.77 E-6/yr 1 112-157 mph - * --- ---- * -----

• 3000 126,600sqft 1.19 E-3/yr 9.77 E-6/yr

= 6.65 E-7/sqft 157-206 mph 1.14 E-8/yr 3.03 E-6/yr 1

• 3000 j 126,600sqft 3.86 E-5/yr 3.03 E-6/yr

= 7.00 E-6/sqft 3.99 E-9/yr 6.85 E-7/yr 1 206-260 mph - - * * --

• 3000 126,600sqft 8.78 E-6/yr 6.85 E-7/yr

= 1.08 E-5/sqft l

The following table shows the additional failure probability that will be l used for equipment housed within a particular building. The buildings are

! shown with the surface areas used in the calculations.

l l

l OC0287-0018A-NLO2

4 BUILDING WIND SPEED RANGE (mph) 72-112 112-157 157-20f. 206-260 Screenhouse 3.63 E-3 5.27 E-3 building failed '

7,700 sq ft EDG Room 7.07 E-4 1.03 E-3 building failed 1,500 sq ft Turbine / Service 2.39 E-2 3.47 E-2 building failed 50,700 sq ft With the values above for the failure due to missiles the dominate sequences can now be recalculated. The value for missile failure has been incorporated in the respective heading failure probability. The value shown below is for l the initiating frequency based on the wind speed interval of 80 to 110 mph.

l i

1) P Z Fs Ycrd Lp 4.23 E-8

> 2) P Z Fs Ycrd Cs 1.49 E-7

3) P Z Fs Ycrd Cr 3.69 E-7

4) P Z Q Fs Lp 7.65 E-9

5) P Z Q Fs Cs 8.55 E-8

6) P Z Q Fs Cr 8.23 E-8

7) P Z Q Fs Em F1 Lp 1.03 E-7

8) P Z Q Fs Em Cs 1.95 E-8 l 9) P Z Q Fs Em Cr 6.82 E-8

10) P Z Ev Fs th 2.66 E-8 9.53 E-7 l

l l

i OC0287-0018A-NLO2

5 LOSP WITH WINDS BETWEEN 110 and 150 mph 50 PORTABLE PUMPS Mission Time of the Event Tree - 24 Hours The Event Tree for LOSP Transients Due to Winds is shown in Appendix B. The quantificction of the sequences follows. Values used for the quantification are included in Appendix C.

'e QUANTIFICATION

1) P Em = (6.9 E-4) (1.69 E-2) 1.17 E-5

2) P Ev = (6.9 E-4) (2.8 E-3) 1.93 E-6

3) P Z PCS = (6.9 E-4) (9.8 E-2)(2.59 E-3) 1.75 E-7

4) P Z Em = (6.9 E-4) (9.8 E-2)(1.69 E-2) 1.14 E-6

5) P Z Ev = (6.9 E-4) (9.8 E-2)(2.8 E-3) 1.89 E-7

6) PZI = (6.9 E-4) (9.8 E-2)(1.0 E-2) 6.76 E-7 1.58 E-5 Probability of experiencing winds between 110 and 150 mph is determined from the wind speed data in Appendix B, Best Estimate.

P = P(110) - P(150) = (7.0 E-4) - (9.0 E-6) = 6.9 E-4 l

l I

OC0287-0018A-NLO2

6 REQUANTIFICATION ASSUMING THE PORTABLE PUMPS ARE AVAILABLE TO THE OPERATORS Portable pumps allow the operators a diverse means of providing make-up water to both the emergency condenser and to the primary system if needed.

The pumps will have an assumed failure rate of 1.0 E-3 (includes both failing to start and failing to run).

With the portable pumps available, the event tree headings Em and PCS have new quantification values. The PCS make-up now includes the portable pumps and the core spray valves in addition to the CRD pumps. The same is true of the emergency condenser make-up heading Em. (See Appendix B for further discussion).

OC0287-0018A-NLO2

7 QUANTIFICATION - WINDS BETWEEN 110 and 150 mph In this interval, damage is assumed to occur to the turbine building and the cable penetration room. The domin water pump is unavailable, the MSIV and emergency condenser valves receive signals to close/open either from the LOSP event or from the alternate shutdown building.

There are now two sources of rake-up water for the emergency condenser (the portable pumps and the diesel fire pump) and three for Primary System Make-up (the diesel fire pump or portable pumps and the CRD pump coupled with the EDG).

QUANTIFICATION:

1) Pc Em = (6.9 E-4)(8.0 E-3) 5.5 E-6

2) P Ev = (6.9 E-4)(2.8 E-3) 1.93 E-6

3) P Z PCS = (6.9 E-4)(9.8 E-2)(1.19 E-3) 8.05 E-8

4) P Z Em = (6.9 E-4) (9.8 E-2)(8.0 E-3) 5.41 E-7

5) P Z Ev = (6.9 E-4)(9.8 E-2) (2.8 E-3) 1.89 E-7

6) PZI = (6.9 E-4)(9.8 E-2)(1.0 E-2) 6.76 E-7 8.92 E-6 The failure probability of the diesel fire pump or emergency diesel generator due to missiles impacts is not effected by the addition of the portable pumps and remain in the 1.0 E-7 to 1.0 E-8 range and therefore, will have little impact on the above quantification, as the random failure rates of the equip-ment is on the order of 1.0 E-2 to 1.0 E-3 range.

i l

l 1

I l

OC0287-0018A-NLO2

8 LOSP WITH WINDS BETWEEN 150 and 190 mph In this interval, the screenhouse is now assumed failed in addition to the cable penetration room and the domin water pump. Failure of the screenhouse results in the loss of both fire pumps. Therefore, make-up is only available via the portable pumps. The EDG room fails at 193 mph, therefore, the EDG is still available. The assumption will again be made that the MSIV and emer-gency condenser valves receive their signals prior to damage to the cables.

The probability of experiencing winds between 150 and 190 mph is:

P = P(150) - P(190) = (1.0 E-5) - (3.0 E-6) = 7.0 E-6 QUANTIFICATION

1) P Em =

(7.0 E-6)(9.0 E-3) 6.3 E-8

2) P Ev =

(7.0 E-6)(2.8 E-3) 1.96 E-8

3) P PCS = (7.0 E-6)(9.8 E-2)(1.35 E-3) 9.3 E-10

4) P Z Em = (7.0 E-6)(9.8 E-2)(9 E-3) 6.2 E-9

5) P Z Ev = (7.0 E-6)(9.8 E-2)(2.8 E-3) 1.9 E-9

6) PZI = (7.0 E-6)(9.8 E-2)(1 E-2) 6.9 E-9 9.85 E-8 I

I 1

OC0287-0018A-NLO2

. O 9

LOSP WITH WINDS BETWEEN 190 and 250 mph In this wind interval, the only buildings remaining undamaged are the contain-ment vessel and the alternate shutdown building. The portable pumps are now the sole source of make-up water to both the emergency condenser and the primary system.

Probability of winds between 190 and 250 mph:

(See Appendix C for wind speed data).

P = P(190) - P(250) = (3.0 E-6) - (4.0 E-7) = 2.6 E-6 QUANTIFICATION:

1) P Em = (2.6 E-6)(9.0 E-3) 2.34 E-8

2) P Ev = (2.6 E-6)(2.8 E-3) 7.3 E-9

3) P Z PCS = (2.6 E-6)(9.8 E-2)(8.6 E-2) 2.19 E-8

4) P Z Em = (2.6 E-6)(9.8 E-2)(9E-3) 2.29 E-9

5) P Z Ev = (2.6 E-6)(9.8 E-2)(2.8 E-2) 7.13 E-9

6) PZI = (2.6 E-6)(9.8 E-2) (IE-2) 2.55 E-9 6.46 E-8 From the March 16, 1982 letter from Consumers Power Company to the NRC, it was concluded that tornado missiles generated by a combined tornado wind speed of 250 mph, that only local buckling would occur to the containment vessel.

Therefore, missiles generated within the wind speed range of 190 and 250 mph will not affect the ability to cool down the plant assuming the portable pumps are available.

l l

OC0287-0018A-NLO2

10 WIND SPEEDS GREATER THAN 250 mph Since the containment veesel fails above 250 mph, the core damage probability for winds in excess of 250 mph is equal to the frequency of occurence of these wind speeds at Big Rock Poinc, which from the best estimate curve from Figure 6 in Appendix C is:

P(250) = 4.0 E-7 To determine the core damage estimates at the 95% confidence levels, the appropriate wind speed probabilities from the upper 95% curves are used and the sequences requantified. These valves are reported in the Quantification Summary Sheet.

OC0287-0018A-NLO2 l

APPENDIX B Event Tree and Discussion l

l f

l l

l l

i f

l OCO287-0018A-NLO2

c 1

LOSP EVENT TREE FOR WINDS AND TORNADOES LOSP RPS Z I Ev Em PCS

• • • • • • • • • • • • • • • • • • • • • 1

• • • • • • • • • • • • • • • • • • • • • ********************* 2

• ******************************* 3

• *********** 4

• * *********** *********** 5

• * *********** ********************* 6 k*khAh*khQn n k k

• *********** ******************************* 7

• ***************************************** 8

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 9 HEADINGS:

RPS: This heading is for the Reactor Protection System. Proper functioning of the RPS is successful SCRAM of the reactor.

For Winds and Tornadoes, the possibility of a failure to SCRAM in conjunction with the wind / tornado damage is not considered.

Z: This heading is for the Main Steam Isolation Valve (MSIV).

Closing the MSIV isolates the primary system from the secondary side, and c.liminates the need for primary system make-up during the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of the event (this assumes normal PCS leakage of about 2 gpm).

I: This heading is isolating the primary system with the turbine /

condenser valves. When isolating in this manner, PCS make-up is required within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> in order to prevent the Low Reactor Water setpoint of 2'9" from being reached.

Ev: The heat sink available is the Emergency Condenser. This heading is the operation of the emergency condenser, for the first four hours shell side make-up water is not required, after four hours make-up is needed to assure heat removal through the emergency condenser.

OC0287-0018A-NLO2

2 Em: Given successful operation of the emergency condenser shell side make-up water is needed. The make-up water source is dependent upon the wind speed interval under consideration. For wind speeds up to 80 mph, all water sources are available. At wind speeds above 110 mph, the domin water system is unavailable, at wind speeds of 150 mph and above the diesel fire pump is unavailable.

With the proposed portable pumps water is available up to 250 mph.

PCS: With the primary system isolated by the turbine / condenser valves there exists some small steam lines that allow primary system inventory loss and a make-up source is needed. The make-up is necessary in order to prevent the Low Reactor Water level from being reached. As with emergency condenser make-up, PCS make-up is dependent upon wind speed. Below wind speeds of 150 mph the make-up sources available are the No 1 control rod drive pump which can be powered from the emergency diesel generator if necessary (this is assumed for wind speeds above 80 mph). Also available in this range is the diesel fire pump providing water through manually opened core spray valves M0-7070 and MO-7071.

The CRD/EDO make-up source is availabla up to a wind speed of about 190 mph at which point the generator building has failed.

With the portable pumps available, make-up water is available up to a wind speed of 250 mph.

SEQUENCE DISCRIPTION:

1) In this sequence no adverse equipment failures have occurred and the plant can be successfully shutdown by the operators. The Reactor Protection System has SCRAMed the plant and the MSIV has closed isolating the primary system. The emergency condenser has been activated and within the four-hour limit make-up water has been established.

2) This sequence it similar to sequence number 1 except that the make-up water d urce to the emergency condenser has not provided sufficient water to cool down the primary system. Without a heat sink the core is assumed to eventually fail by sequence P:Em.

3) With this sequence the plant has SCRAMed and the MSIV closed.

However the emergency condenser has not activated to provide a heat sink for the decay heat. In this case the core damage sequence is P:Ev.

4) Sequences four through eight involve failure of the MSIV to close. In this sequence, the turbine / condenser valves have closed and minimize the steam flow outside of containment. The emergency condenser has activated and is removing decay heat.

Within the four hour limit, a make-up source to the condenser has been established providing for continued decay heat removal. With the primary side isolated by the turbine / condenser valves, a make-up water source to the primary was needed for this sequence to be successful in preventing core damage.

OC0287-0018A-NLO2

i 3

5) In this sequence, the primary system has been isolated by the turbine / condenser valves, and a PCS make-up water source was needed. However, attempts to provide make-up failed and PCS level drops too far and the core is assumed damaged by sequence P:Z:PCS.

6) The plant has SCRAMed and the MSIV failed to close, primary system isolation has been accomplished using the turbine /

condenser valves. The emergency condenser has activated and is removing decay heat. When a make-up source is needed, all attempts fail and decay heat cannot be removed. Therefore, even-

. tual core damage is assumed to occur with sequence P:Z:Em.

7) Successful isolation by the turbine / condenser valves has been performed. Attempts to place the emergency condenser in service have failed and no heat sink is available. The core damage for this sequence occurs by P:Z:Ev.

8) In this sequence, the plant has been successfully SCRAMed. The operator's attempts to isolate the primary system have failed.

With the primary system in this condition, core damage is assumed to occur by sequence P:Z:I.

9) This sequence is included in the tree for completeness. The addition of a failure to SCRAM on top of damage due to winds and tornado damage is not analyzed. The analysis of a failure to SCRAM is contained in other submittals.

i

(

1 l

OC0287-0018A-NLO2

APPENDIX C Quantification Data s

OC0287-0018A-NLO2

1 Event Tree Ouantification Data MSIV failing to close 9.80 E-2/ demand Turbine / Condenser valves 1.00 E-2/ demand fail to isolate Emergency condenser fails 2.80 E-2/ demand to activate Emergency Condenser Make-up 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> (4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after activation)

Diesel Fire Pump fails to start 3.06 E-3/ demand fails to run 2.0 E-5/hr

• 20 hrs out for maintenance 1.33 E-4/ year 3.60 E-3/ demand 110-150 mph Missile Damage 5.27 E-3 SV-4947 fire water to 7.00 E-3/ demand Emergency Condenser Operator fails to start 1.00 E-3/ demand pump /open valve Portable pumps failing to 1.00 E-3/ demand run/ start /out for maint.

EC make-up with both pumps available (portable pumps

• diesel pump) + SV-4947 FTO + Oper Error (1.00 E-3

• 8.87 E-3) + 7.00 E-3 + 1.00 E-3 = 8.00 E-3 I

Make-up with only portable pumps available 1.00 E-3 + 7.00 E-3 + 1.00 E-3 = 9.00 E-3 j

Make-up with diesel fire pump only 8.87 E-3 + 7.00 E-3 + 1.00 E-3 - 1,69 E-2 i

OC0287-0018A-NLO2

2 Primary System Make-up (5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />)

Emergency Diesel Generator FTS, 1.79 E-2/ demand FTR, 1.97 E-2/ hour

• 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> 0FM 3.00 E-3/ year

.1194/ demand 110-150 mph Missile Damage 1.03 E-3 150-200 mph Missile Damage 1.05 E-2

• ~

Control Rod Drive Pump FTS, 3.00 E-4/ demand FTR, 1.91 E-4/ hour

• 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> OFM 3.17 E-3/ year 4.43 E-3/ demand CV-4090 FTO 1.00 E-3/ demand Operator error loading .0323/ demand CRD pump to EDG Operator error in manually .0075/ demand opening core spray valves Valves fail to open 1.00 E-4/ demand PCS from DFP or Portable Pumps through core spray valves (DFP failure

• Portable Pumps fail) + operator error + valve FTO (8.87 E-3

• 1.00 E-3) + .0075 + 1.00 E-4 = .0076/ demand l

l PCS from DFP through core spray I 8.87 E-3 + . 0075 + 1.00 E-4 = .0165/ demand PCS from portable pumps through core spray 1.00 E-3 + . 0075 + 1.00 E-4 = .0086/ demand FTS - fails to start, FTR - fails to run, OFM - out for maintenance and FTO - fails to operate.

OC0287-0018A-NLO2

-r -- .-. -

I 3

PCS from CRD loaded to EDG CRD pumps fail + EDG fails + oper error + CV-4090 fails (See Note 1).

4.43 E-3 + .1194 + .0323 + 1.00 E-3 =

.157/ demand PCS from either'DFP or CRD/EDG

(.0165) * (.157) = 2.59 E-3 l

PCS from either DFP/PP and core spray or CRD/EDG  !

(.0076) * (.157) = 1.19 E-3 PCS from either PP or CRD/EDG l

(.0086) * (.157) = 1.35 E-3 l

NOTE: (1) Missile damage has minimal impact on the diesel generator failure  ;

probability and does not affect final results.

OC0287-0018A-NLO2

-I

\

\  % \ STRAIGHT ,

o , , in-2 \ \V'wNos ,

fg \

\

\

\ ,

\

E \

\

>- i , g3 s

\

hE \ \

W W '*# 4 \ s Wg \ \

t g i

> I x 10-5

\ \

HQ ,

i \ s v

i, gs \ \

E 95 PERCENT TORNA00ES'

\% \

CONFIDENCE LIMITS 3 1 : 10~7 50 10 0 LSO 200 250 300 350

, WINOSPEED MPH l

FIGURE 6. WINOSPEED HAZARD PROBASILITY MODEL FOR BIG ROCK POINT I

l

5 Table 1 Maximum

Maximum Wind Structure Pressure Velocity Element (psi) (mph)

Reactor Building Steel Spherical 5he11

. 1.35 250 Screen House / ~ Roof Decking Discharge struc- 0.41 182 Concrete Walls 1.35 ture 152 Emergency Diesel Roof Decking Generator Room 0.46 193 Concrete Walls 1.35 212

' 240-foot Stack Concrete Stack NA 200 Foundation NA 200 Condensate Water- Tank Storage Tank 1.35 250 Domineralized Tank i Water Storage 1.35 250 Tank

.i Solid Radwaste Superstructure Storage Vaults 0.17 100 Original " low level" Vault 1.04 250 Original "high level" Vault 1.35 250 New Vault 1.35 250 Turbine Building South Wall Intermediate Columns O'.17 110 Crane Columns and Roof Truss 0.21 121 North and South Wall Bracing NA 121 Wall Intermediate Colunns 0.22 125 Metal Siding D.24 138 Turbine Building Roof Bracing NA 140 East and West Wall Bracing NA 148 Roof Decking 0.49 198 Roof Purlins 0.82 >250 Service Building Safety-Related Block Walls 0.03 .16 NA Wall Bracing Column J NA 123 Exterior Co'umn 0.23 126 Metal Siding 0.24 138 Girts 0.28 140 Control Room South Wall 0.57 NA Roof Decking 0.81 233 Boiler Stack NA >250

~ _ . -_ . _ _ _ ____. _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _

6

, Table 1(Continued)

Maximum Maximum Wind Structury Pressure Velocity Element (psi) (mph)

Turtine Building Metal Siding 0.24 138 Passageway East and West Wall Column 0.36 159

" Blowout" Panel 0.50 NA Fuel Cask Loading Superstructure NA >250 Dc:k/ Core Spray Block Wall 0.03 NA Equipment Room I

l l

l 1

l

Table 2 WINO SPEED RANGE PROBA8ILITIES Wind Speed Rahge III Wind Speed Range (2)

F-Scale (Mcdonald) Probability (Twisdale) Probability 1 40-72 4.87 x 10-5 40-73 9.32 x 10-4 2 72-112 2.09 x 10-5 73-103 8.17 x 10-4 3 112-157 9.77 x 10-6 103-135 3.77 x 10-4 4 157-206 3.03 x 10 4 135-168 1.18 x 10-4 5 206-260 6.85 x 10~7 168-209 3.86 x 10-5 6 260-318 1.08 x 10'I 209-277 8.78 x 10-6 (1) From NRC wind-speed study for Big Rock Point (2) Used in EPRI tornado study I

3 l

Table 3 TARGET IMPACT PROBABILITIES F-Scale Expected Val.ue 95% Limit 2 1.62 x 10-8 2.84 x 10-8 3 2,53 x 10 4 4.23 x 10-8 4 9.12 x 10'I 1.37 x 10-8 5 1.14 x 10-8 1.93 x 10-8 6 3.99 x 10'I 7.11 x 10'I All 6.60 x 10-8 8.72 x 10-8 l

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

9 Table 4 TARGtT DAMAGE PROBABILITIES (ASSUMING 6-INCH THICK CONCRETE WALLS)

F-Scale Exper.ted Value 951 Limit 2 4.64 x 10-11 9.79 x 10'11 3 2.92 x 10-8 8.00 x 10-8 4 8.29 x 10-10 1.38 x 10-9 5 2.50 x 107' 4.38 x 10-9 6 1.11 x 10~8 1.89 x 10'I All 7.41 x 10-8 1.22 x 10-8 l

l l

l l __

i l

l

, DETERMINATION OF BIG ROCK BUILDING SURFACE AREAS i

1. 1 Containment radius =

65 ft (actuai) I radius + L/8 - 65.000 / 8.0 + 65.0 = 73.125 ft surface area of a sphere = 4*pi*r**2 l

SA actual = 4

• 3.14

• 65

• 65 = 53,066.00 sqft SA modified = 4

• 3.14

• 73.125

• 73.125 = 67,161.656 sqft

1. 2 Turbine and Service Building i

l S

e Tl**/

+

v o

15 %

^# #

p [ #+ n 61 4 i

I with G  : 30f+ +

l Dimensions actual + L/8 120 135.00 61 68.62 23 25.88 113 127.12 30 33.75 l

l l

j Surface areas Actual e/w walls 2 x 120 x 61 = 14,640 top turb b1dg 113 x 120 = 13,560 s wall turb b1dg 61 x 113 = 6,893 n/s serv b1dg 2 x 30 x 23 = 1,380 top serv b1dg 30 x 120 = 3,600 CC0287-0018A-:iLO2

n 40,073 sqft Actual + L/5, ,

e/w walls 2 x 135 x 68.62 =

18.527.40 top turb b1dg 127 x 135 = 17,145 s wall turb b1dg 68.62 x 127.12 = 8,722.97 n/s serv b1dg 2 x 33.75 x 25.88 = 1,746.90 top serv b1dg 33.75 x 135.0 = 4,556.25 50,698.52 sqft

1. 3 Alternate Shutdown Auilding x

+h lo kI+

V

• 2.5 & ~

Dimensions actual + L/8 i

25 28.12 21 23.62 10.5 11.81 Surface areas Actual e/w walls 2 x 21 x 10.5 = 441.00 n/s walls 2 x 25 x 10.5 = 525.00 top 21 x 25 = 525 1,491.00 sqft CC0287-0018A-NLO2

12 Actual + L/8 e/w walls 2 x 23.65 x 11.81 = 558.61 n/s walls 2 x 28.12 x 11.81 = 664.19 top 23.65 x 28.12 = 663.04 1,887.84 sqft

1. 4 Screenhouse and Emergency Diesel Generator Buildings C V l- J /

/

/ *

• Yf

' /

14 54 o #

7

• l'b ff  ; 13 4 +

Dimensions actual + L/8 i

14 15.75 61 68.62 33 37.12 13 14.62 l 30 33.75 Surface areas Actual

e wall screen house 61 x 14 = 854 v wall screen house 91 x 14 = 1,274 n/a wall screen house 33 x 14 x 2 = 924 cop screen house 91 x 33 = 3,003 6,055 sqft I OC0287-0016A-

ILO2

13 n/s wall edg room 13 x 14 x 2 = 364 e wall edg room 30 x 14 = 420 top edg room 30 x 13 = 390 1,174 sqft Actual + L/8 -

e wall screen house 68.625 x 15.75 = 1,080.84 w wall screen house 102.27 x 15.75 = 1,610.75 n/s wall screen house 37.12 x 15.75 x 2 = 1,169.28 top screen house 102.27 x 37.12 = 3,796.26 7,657.13 sqft n/s wall edg room 14.62 x 15.75 x 2 = 460.53 e wall eds room 33.75 x 15.75 = 531.56 top edg room 33.75 x 14.62 = 493.42 1,485.51 sqft l

! The total building surface area is therefore the sum of the individual

buildings which is

Actual 53066 + 40073 + 1491 + 6055 + 1174 = 101,859

Actual + L/8 67162 + 50700 + 1890 + 7654 + 1486 = 128,892 The original submittal reported a total Big Rock surface area, not including l

the offset impact factor, of 100,000 sqft. As can be seen from the above calculations the surface area including the offset factors varies considerably from the previous reported value. Therefore a calculation is performed below to see if the results originally submitted are still valid. A similar calculation as performed in the original submittal is shown below except that the areas including the offset impact are used. The calculation shown is for the Emergency t

Diesel Generator Room, which has total wall surface of 992 sqft and a top f surface area of 494 sqft (including L/8). The impact probability is for any l

missile hit (a missile hit is assumed to fail the roof) and the damage probability is for missile hits which damage the six inch concrete walls in the generic plant (the EDG room has ten inch concrete walls).

I l

l OC0287-0018A-NLO2 l

l

14 Probability (1/yr) 1/yr*sqft Impact 5.06 E-08 4.00 E-13 Damage 3.79 E-09 2.99 E-14 PQ = (4.00 E-14 * .023

• 494

• 1.0

• 3000)

+ (2.99 E-14 * .023

• 992

• 1.0

• 3000) l.567 E-08/yr

-From the original submittal the missile damage probability to the Emergency diesel generator was calculated to be 1.58 E-8 per year, it can be seen the the inclusion of the offset impact factor does not alter the results, and therefore the conclusions of the submittal will not change due to this factor.

i i

OCO287-0018A-NLO2

o .

APPENDIX D Justification for Applying EPRI Generic Data to Big Rock Point (Response to NRC Question 7)

OC0287-0018A-NLO2

• O 1

INTRODUCTION The EPRI reports (NP-768 and NP-769) contains an analysis of the impact and damage probabilities that may be expected for a tornado strike at a postulated generic plant. These results were used to determine the missile damage probabilities at Big Rock Point. In order to fully justify the use of these probabilities some assumptions made in the original Big Rock Point Winds and Tornado submittal require a more detailed explination.

The areas in question that need further review are:

1) How the Big Rock Plant site geometry compares to the geometry of the generic plant.

2) How the Big Re k Point Plant buildings compare in surface area to to the generic plant.

3) How the missiles which may surround the Big Rock Point plant compare to the missile set used for the generic analysis.

PLANT GEOMETRY REVIEW The damage and impact probabilities given in Tables 3 and 4 were determined using the Plant A safety envelope. This safety envelope is a polygon which encloses all of the buildings housing safety related equipment important in shutting down the reactor. A successful tornado strike is defined as any intersection between the tornado path and the plant safety envelope. A drawing of the generic plant with its safety envelope is shown in Figure 1, also included in the drawing are the Big Rock buildings. The Big Rock buildings are shown with their respective N-S-E-W orientation and with their approximate dimensions as dictated by the drawing scale of the generic plant.

The Big Rock buildings include the containment, the turbine building with offices, the screen-house / emergency diesel generator building, and the alternate shutdown building. As can be seen, the Big Rock buildings are all enclosed by the generic plants safety envelope. In addition, if the safety envelope for the Big Rock buildings is modeled as a circle, most of the circle is also enclosed by the generic safety envelope. If, on the other hand, the Big Rock safety envelope is modeled as a polygon, the entire safety envelope would be enclosed by the generic plants. Therefore, since the generic plants safety envelope is larger than the safety envelope expected for Big Rock, and

, keeping in mind that a successful tornado strike is the intersection of the j tornado path and the safety envelope, the results from the generic plant i contain more successful tornado strikes than would be expected for Big Rock.

This means that the potential for missile generation, which is determined only for successful tornado strikes, is greater than can be expected for Big Rock, and thus the generic plant results bound the results that would be l

expected for Big Rock.

I t OCO287-0018A-NLO2 I

i FIGURE 1.

2 s

lk 3

7 I

aman simm evunium U b.

e SCREsWl40W5E /

^ - EmAEWu DIEia.6EN.

__',( '

, evitoius 3 -

\ ,L  %

g , 1 CONTPitMMENT 2 Assil 11 ding #

3 Puel land 11ag Building SUILDIN(s

& Diesel Generator Building 3 waste Treatment building I 6.4 Service Water Istaka Structurae 7 solding Tanks (Iaclosed)

, Turbiae building (Neo-safety-related)

L i W.

I u 2100' 2600' 3100' s (a) Plas View aL 230' i

I 5 . .

h'

n. .... m0 .

(b) Elevation view I

l P

3 BUILDING REVIEW: SURFACE AREA COMPARISON The generic plant contains seven buildings which house the required safety related equipment needed to shutdown the plant. At Big Rock there are only four buildings, the containment vessel, the turbine / service building, the screen-house / emergency diesel generator building and the alternate shutdown building.

Since the original submittal the alternate shutdown building has been added onc' must now be incorporated into the total plant surface area calculation.

Froma the original submittal the total building surface area reported was 100,000 square feet, this value however did not include the additions to the buildings to account for the offset impacts as discussed in the EPRI report.

The factor used to determine the added dimension was L/8, or one eighth of the actual dimension was added before calculating the surface area. This was done for each of the four buildings and a new total of 129,000 square feet was determined. Since the two values are relatively close, the impact probabili-ties were checked to see if they changed at all. For the emergency diesel generator room the calculations are shown be' low:

PROBABILITY 1/yr 1/yr*sqft 1/vr*sqft (actual) (+ L/8)

IMPACT 5.06 E-8 5.06 E-13 4.00 E-13 DAMAGE 3.79 E-9 3.79 E-14 2.99 E-14 Surface area of walls = 992.25 sqft (840 sqft w/o L/8)

Surface area cr ro = 494 sqft (390 sqft w/o L/8)

PQ = (4.00 E-13)*(.023)*(494)*(1.0)*(3000) 1.36 E-8

+(2.99 E-14)*(.023)*(992.25)*(1.0)*(3000) 2.05 E-9 1.567 E-8 The value reported for the probability that a missile strikes the emergency diesel generator room and damages the unit, PQ, from the original report is 1.58 E-8, the PQ value calculated above is 1.57 E-8 and does not significantly differ from the original value. Therefore no change will result from the addition of the offset impact correction factor of L/8 and from the addition of the alternate shutdown building into the building surface area calcula-tions. It must-be kept in mind that the damage probabilities are only used for the concrete walls and that the results from the EPRI report are for six inch walls, the concrete walls at Big Rock are 10 inches thick.

OC0287-0018A-NLO2

4 SITE SPECIFIC MISSILE REVIEW A walkdown of the grounds surrounding Big Rock was conducted to identify the possible missiles that are present. The ground area considered was taken as a circle with a 2000 foot radius centered on the containment vessel. This circle was chosen as the missile generation zone for Big Rock. The generic plants missile generation zone was determined by enclosing all the ground which was within 2000 _eet from the nearest building which housed equipment important for safety. Since the buildings at Big Rock are grouped so closely together and the surrounding grounds are so uniform in composition, the circle was chosen as a good approximation of the missile generation zone. Drawings containing the plant site and the missile generation zone are shown in

~

Figures 2A and 2B. Within the 2000 foot circle approximately one quarter of the area is the surface of Lake Michigan and no missiles were postulated to originate from this area. Of the remaining three quarters of the area, roughly 75% to 80% is covered with trees (Big Rock is surrounded by a swampy type terrain with an abundance of cedar trees). The remaining 20% to 25% of the area is cleared land that contains the plant buildings, driveways, park-ing lot, office trailers and other small clearings. Also included in this area is cleared land for the transmission lines. The following table, Table 1, shows the various items found during the walkdown which were considered as possible missiles.

In reviewing this list, it can be seen that there are relatively few poten-tial missiles in the vicinity of the plant buildings. The most significant source of missiles is Area J, located to the wsst of the containment vessel.

This area contains most of the various construction type materials which are stored outside (ie, conduit, wooden blocks, metal barrels, etc).

TABLE 1 Possible Missile Sources - cuestde Stee Boundary Fence Area A: Walkway to well house

1) wooden shed used to cover sand pile, posts imbedded in ground

2) light pole

3) area within the fence surrounding the well house a) domestic water tank b) standby diesel generator trailer I c) small building covering the well pump I d) standby diesel generator transformer l

e) light pole Area B: Gravel parking area

1) two wooden sheds on cinder blocks

2) loose wooden skids and metal barrel

3) small water tank used for fire training

4) square metal tubes for waste disposal

5) two small steel tanks, cylinders

6) one metal rack, small

7) semi-trailer and spare axle

8) four transmission line poles

9) five concrete slabs (4'x8': 8'x12': 3-3'x4')

OC0287-0018A-NLO2

5 Area C: Radwaste building area

1) double wide trailer (currently being moved)

2) Radwaste processing building

3) LP gas storage tank

4) metal shipping containers (24)

5) three metal barrels

6) five tran: mission line poles

7) railroad tracks Area D: Employee parking lot

1) cars (7am to 4pm about 200: rest of the time about 20)

2) three metal light poles

3) one trash dumpster

4) LP gas storage tank Possible Missile Sources - Inside Site Boundary Fence Area E: Security building

1) metal siding of building

2) LP gas storage tank

3) metal walkway ramp

4) air conditioner

5) fire hose real container Area F: Front door of service building

1) flag pole

2) metal light pole

3) office building siding Area G: Screenhouse

1) three metal light poles

2) 2 metal camera poles

3) fire hose reel container

4) underground tank lids

5) junction boxes mounted on conduit, five feet from ground

6) hand rail around intake bay,

7) small metal hatch covers (,l'x 2')

Area H: North side of containment vessel

1) fire hose reel container

2) three metal light poles

3) security "Ident" station

4) emergency escape lock cover and two security cameras

5) equipment lock metal siding

6) cask handling tools

7) security "Ident" station OC0287-0018A-NLO2 1

. o 6

Area I: Alternate shutdown building area

1) LP gas storage tank Area J: Area west of containment vessel

1) miscellaneous construction material (bricks, wooden blocks, rack of pipes tied together, 36 empty barrels, two trailers, wooden shed, kerosene, small pipe rack)

2) yellov fiberglass construction building

3) conduit racks

4) fire hose real container

5) calibration shed, Chem /HP storage, gasoline storage

6) old QA/QC office building / warehouse

7) barrel rack with six barrels

8) paint storage shed

9) wooden shed

10) wheelbarrows

11) three light poles

12) various loose material and tools

13) three electrical panels Area K: Area surrounding the switchyard

1) gasoline fuel storage tank and stand

2) one metal light pole

3) two garbage dumpsters

4) fire hose reel container

5) two transformers

6) bottled gas storage area (connected to turbine building)

When the generic plant was analyzed for missile damage the missile set contained 1000 of each of the six basic missile types as defined in Table 3-2 of the EPRI report. These 6000 missiles were assumed to be equally distributed around the site missile generation zone. In comparing the Big Rock site missile list it can be seen that 6000 missiles seems an inordinately high number of missiles. The original submittal assumed a total of 3000 missiles to be dirtributed around the plant and that the same missile impact results would be generated as wich 6000 missiles. Since the site generation zones are of different area, with Big Rock's being smaller, the 3000 missiles distributed in this area are assumed to produce the same missile density as the 6000 missiles distributed among the generic plants generation zone.

SUMMARY

In summary, the results reported in the EPRI report for the generic Plant 'A' are assumed to encompass the results expected for the Big Rock Point Plant. The missile impact and damage probabilities for the generic plant were derived using some basic assumptions such as plant building surface area, site missile generation zone, and the number and types of possible missiles in and around the plant site. The review of Big Rock shows that the total surface area of the buildings is much less than the generic plant, the site missile generation zone is smaller for Big Rock than the generic plant as well as the number of possible missiles, the missile density however was assumed to be the same.

Therefore, the results obtained for the missile impact and damage probabilities for the generic plant will be used as close approximations for Big Rock.

OCO287-0018A-NLO2

FIGURE ZA .

/-

/

MICHIG/ N .

I p % l

/ N'3,a,;exn

. gf i '-

Q .c ..

N, ,

)

v

) ~, . . . . . .

~.. --...-..

ss~- -= * ' ~ * ' ' " ~ ' "'"d l

, s -

,0

..Y DLTERMATE 4 SCREEN Hot 15E./LDb BLDb Sg 311HTDOWN . .

y

., BlD6 '

,, CONTAlklM RORD t

l e

/ TURolNE / SERVICE BLM

? }.  ?

-a

i. _ _

f

i. ,, , < s,'s'

,t P* p ? .

l l

l l

'?* * ,s'ku R .

O N------------------ --

- l' N

i. .

Il M ISSIE GB MERATION 7_DNE '

1

! ' / -

ll C

/ i 1 ,0 t /*

l lLej *e t .'d ,

s I

/ //

l o

/ ,

~.

Table 3 TARGET IMPACT PROBADI!.ITIES F-Scale Expected Valme 95% Limit l 2 1.62 x 10 4 2.84 x 10-8 l

l 3 2,53 x 10 4 4.23 x 10-8 4 9.12 x 10*I 1.37 x 10-8 5 1.14 x 10-8 1.93 x 10-8 6 3.99 x 10-8 7.11 x 10-8 All 6.60 x 10 4 8.72 x 10-8

i .

l Table 4 TARGET DAMAGE PROBABILITIES (ASSUMING 6-INCHTHICKCONCRETEWALLS) l F-Scale Expected Value 955 Limit 2 4.64 x 10-I1 9.79 x 10'11 3 2.92 x 10?I 8.00 x 10'8 4 8.29 x 10-10 1.38 x 10-8

'S 2.50 x 107 9 4.38 x 10-8

. 6 1.11 x 10~I 1.89 x 10'I All 7.41 x 10 1.22 x 10-8

Table 3-2. Missile Set Definition Bseie Weight A- Postuisted Velocities By Missile Set Total Per Unit Length L d b L min Region (ft/sec) j Number Number Description (Ibs) (1bs/ft) (ft) (in) (in) d (in 2) 1 11 III 1 9 Wood Beam 114.6 9.55 12.0 12.0 4.0 12.0 48.0 272.6 229.9 190.5 2 3 6" Pipe 286.6 18.97 15.0 6.63 - 27.2 5.6 170.8 137.9 32.8 3 1 1" Steel Rod 8.8 2.67 3.0 1.0 - 36.0 0.8 167.5 131.4 26.3 4 2 Utility Pole 1124.3 32.00 35.0 13.5 - 31.1 143.1 180. 6, ., 157.6 85.4 5 3 12" Pipe 749.6 49.56 15.0 12.8 - 14.1 14.6 154.3 91.9 23.0

, 6 25 Automobile 4000.0 243.90 16.4 79.2 51.6 2.5 -

193.8 170.8 134.6 i

i O

i

APPENDIX E Alternate Shutdown Building Description OC0287-0018A-NLO2

I DESCRIPTION OF THE ALTERNATE SHUTDOWN BUILDING Originally proposed as a modification to reduce the consequences of fires at Big Rock Point, the Alternate Shutdown Building (ASDB) has been incorporated with several design features aimed specifically at several other SEP Topics.

Housed within the building are banks of batteries which supply the necessary power to operate the hain Steam Isolation Valve M0-7050 and the condensate return valves from the emergency condenser M0-7053 and MO-7063. These valves are normally controlled from the control room, however, control can be trans-ferred to the ASDB should conditions exist which would or have jeopardized normal control. Switches located inside the ASDB allow the operator to transfer control. The MSIV can only be closed from this location, whereas, the emergency condenser valves can be throttled to control the primary system cooldown rate, Instrumentation installed in the building provides for monitoring primary system pressure and steam drum level. An alarm is also present for low emergency condenser shell side water. Indicating lights are available for firewater flow to the emergency condenser and for the emergency condenser firewater make-up valve.

The ASDB is connected to the containment via an underground bank of conduits which carry the power and control cables for the equipment. There is a pene-tration area located underneath the equipment lock which is on the western side of the containment sphere.

The ASDB, structurally, is composed of 18-inch reinforced concrete walls covered by a 12-inch reinforced concrete slab. The entrance to the building is designed to provide missile protection to the doorway. The attached figures show the ASDB including the dimensions.

The equipment controlled from the ASDB including instrumentation was selected to provide for a diverse means of monitoring / controlling a plant shutdown l should the normal control room methods be unavailable (for example, damage to l the outer cable penetration room or the station power room).

As far as winds and tornadoes are concerned, the building can withstand winds in excess of 250 mph. For missiles, the building can withstand the impact of the two postulated missiles (small steel rod and a telephone pole) as described in SEP Topics III-2 and III-4.a for tornado winds at 250 mph.

Therefore, the building is as resistant to winds and tornadoes as the contain-ment vessel.

The underground bank of conduits are protected from winds and tornadoes from the ASDB to the-point where they enter the loading dock structure. At this point the conduits are run through the open area directly under the equipment lock. This area has very limited access. From above, the area is covered l mostly by the equipment lock itself. There is a small area, roughly 100 square feet or so, that is covered by metal plate sections welded together. The north and south walls are concrete approximately six inches thick. To the west is an open area that allows a liftable track bridge to be lowered allowing easy access for taking heavy equipment into containment. In this open area a small section below and to either side of the equipment lock is covered by heavy metal grating. The area of this opening is about 20 to l 30 square feet.

OCO287-0018A-NLO2

2 Overall, it will be assumed that 150 square feet of open area is available for a missile strike capable of damaging the ASDB conduits / cables. From the missile strike analysis performed for the screenhouse, the probability that a missile will impact the roof is about 1.0 E-7. The screenhouse has a much larger surface area open for impact and would, therefore, have a much higher frequency of damage due to a missile strike than would be expected for the ASDB cable penetration area. It is therefore concluded that the frequency of a missile impacting the area underneath the equipment lock is several orders of magnitude below the random equipment failure rates, since missile impacts are considered in the quantification they have no impact on the final result.

The same type of argument can be used when the control rod drive pump transfer switches are considered for missile impacts. The transfer switches are located on the end of a four-foot section of conduit which rises up from the loading dock working level. Assuming that the transfer switch box and conduit can be modeled as a 1x1x4 foot rectangle a total of 17 square feet is available for missile impact. This area is much smaller than the area of the screenhouse roof and therefore, missiles impacting the transfer switches are very remote occurrences.

OC0287-0013A-NLO2

w. s u .ss< a o a n ~<.

=

MISC STEEL. CE7 Ant.5

• 9 /LN " g ,1 ymmauCTURA4.(A ..l ROOF FRAMING ,' ~

n ar o- g 3

s P

r- g.. q- -

g' .qy -

~

006t3 71 M F/fl0 etterto orrmi 1-

@ $f T/ Ar/9M$D*Tr f, .m l e M R M/CE 8.4RA!/ R n'au.

a

.M JNN/N6S- /

'l 6 i *ttMl o KiiMFact/ MENT/ f 1/N6 .IN L A .?

• g , LIGH T f!170RE NITH N. ~ ~* .. '

- 4 *-On _ .I e e

• ' 4l

^^

I

' Mentum BAT [ HPS LA W S 34pg g pggoj pg$ 1 * <

It APPLCTCN MA LL MA S T[ R ~==

• M G WM

• 3 + L - 12 R - PC ,

R$ A M ND s/stl l ' Y CR EQUAL - fifLD LCC A ff' y:;j, e,gg gy rigig g,go,pcp

.STIDH$ !*DM80$r/0N s I t/PL Y offMIN6-2T'56L. I 3 elL.i ,tfCT. C04ttHT Stua$-Stt UN NGTn DMC67033 6 A

, n 3

' , _._-...J g.w, , W ((

e, A, i --

,* r g

4 i

(@ . J ..EX ?/_. \e y , . .

R i n n

I l+.r-- L-----fl *Qggj ,,[--- *ft M--)ta;-O-e m ,

.zumms s i -- --- ,

c o. <

• *f ra , _ _ _ upin6_ C A dj (f

• t.

)'_ _ _

4 .'

~

u77C i3' 2- 'oit d

4 k ,

1. eMQ-r- , If i if

/ * .T - ,,'==

g m.

. ~

co- T i- ,

w

, o c ii -

"g.3~~-(2h t

'&lWn!'/t*A%'fsifQ.&.

RM [bGE CDllMG g (

1 >-

y 4 >-- ~ x}\ -

l k

r

? fhI' hh k Y '

], Q

,a APPL &TCM hl

• ' / &'  ; 8 =

VnPCRTIGHTQ-2co w e ,

9*U 3+

4 , y ,

• el

• t

. ~

p ,-  :

= 81/PW 10756 TO e m ~

~

ve ginvir 44',,,,

y rump -h g* l . Qg._

, M-11

,1 i

u noom , , , n n -.; 3 '

rearo m mi

.g -

' 7:e rrp eces

~

i 'g L ' emis m einr ,

V

. ,],a 4 ~ [.-U -

- -p- - -

j

-yg,,,- _ - - >

9. - - - _ _-

pp Aaam restw-tr t

-Q !

, si,sias v

.)_ _e i a-v  !

, i

v. r  ;, .y ,o. .

s.. ,

-- n. o.' .

\ x l ;

e .

.: -. arrteisotra a RAN@REVATION 603'-5" non ucs.

Hvaw ,,,.,, o. w.ua step"oama

i * -

(G. ~ax en

,s.e / i.,c '

,' t, hk s ee M. - -

M q.= i  : -

N e~

SN !'

i $N55( ME/ - $, ' ' , h"gbe d' v

.w; q) s -

p- .. i f, .*ru ,~ j, -

2

g. ,  ; .__.

._.c , ,

- .- - _ - , ;s. . _ _ _

=E a 4 s}

kU" -'

lt # e. . . . -

u_ h T g q \

, ;_ - - - g - - - - -- -% a a g .Nl g _y',t%

Y- .

cf ,l -1 \[%,(( *N

, - - ^-e T *

. ., m I;) -

s! % :. l Gr~

i ga f ,* 'n '

w b

i b

.h1[l l h $, i

• i I'1

1. I 4,Q r

I d 4 I

9, N w

% s Y ,

t

5h

N N;;f>#'k L

- x_ jt..ee 3,-5 .,

hh es .i -~

,6 -b V -ts lR ' , ' , $qg @

________,L s .

7 . ,._ _ . _ _ _ _ _ _

l *

• l &

E. O g

gy 6' l i A > ,e w

f,y

• Qs i

pN l%

'i s

" .O 1 q . Ol

{t~

4:; '

\

m .

y , L= !s l r .- - -

n, i n E '

s.

u s

, ~. .. .

c

.I a i ti = ' a

~

s ,  ;

0

., ,! 3 l l,

'k $ -

Chk 2

x g ,' .

$ ,' i_ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ eN _ _ _ _

'I L t E h

g * '

=%

s l p_________.____

.' 'd 4 g 4 'I I '

i

-\ 8 4

y 9 S

I I

..w ,, y i

e C

s ,;N. a o

7 ~

g%  ::e  ;

i 4 '-

,. # 0 'd< '

/ P

[  % l' V '

t.: 7 T2

• k

) . t i' k. I

• I g j#

3'8

~ ,

'q ! x i ,,

L-

,N* -

a ,,, -

~_..__. - -

4 q

l

, , a* ca-- * + .d .f e . 4 *

% ~

l

[4 h 418 l

. i

• N

.5 c j s. *

.l 4

d i 9

-f u_______ g r q

a i p. u, p,,

h W 4 .s w I .. '

, t

! , o  ; '

ltl i

, tw d' e%r,I A ~

- N 1.

.y- - %

\

1

.. ? . 8 u e, i

I

'a% =

N ,a I q

m l 19NU/J 10103 . 9 , G o

l ,

-) ~?

&k) 4 *

~

,,c-l .. e .. .> ,

1 I

I i

1

APPENDIX F Response to NRC Questions from the July 18, 1986 Letter to Consumers Power Company l

OC0287-0018A-NLO2

1 Question 1:

Can the failure of one structure at a lower wind speed lead to damage or failure of other structures that were calculated to have a higher wind-speed resistance due to either a) structural interaction or b) generation of missiles?

Response

It is correct that interaction between st:uctures was not included as a viable failure mode for the structures considered important to shutdown of the Big Rock Point Plant during high winds or tornado conditions in our risk analysis of July 5, 1983. A review of the site layout (Figure 1) suggests that this assumption is appropriate. The important structures include the turbine / service building, cable penetration area, containment and screenhouse (which existed at the time of the 1983 analysis) and the alternate shutdown building (which was completed in the fall of 1985).

Two of the structures, the screenhouse and alternate shutdown building are not physically attached to other buildings and are not vulnerable to structural interaction.

The interface between the cable penetration area and the turbine building is made up of reinforced concrete walls of varying thicknesses of one foot or more. These walls do not make up the principle failure modes of the turbine building or cable penetration area, that being steel columns, trusses and the metal siding. As a result, even though portions of the turbine building are assumed to be vulnerable to lower wind

! speeds, their failures are not associated with structural members important to other structures such as the cable penetration area or containment.

The north wall of the cable penetration area is the containment shell itself. I.ike the turbine building, however, the principle failure modes for the ca' ale penetration area are the siding and steel columns which are not a physical part of the containment shell. Attachments to the containment shell do occur in the form of steel and aluminum flashing, but no physical attachment to important structural members of the cable

penetration area exist. The assumption that interaction between i structures plays little or no role in establishing the vulnerability of the Big Rock Point Plant to wind and tornado loads is considered appropriate.

Missile generation was considered to be inherent in the assumptions made in the missile analysis. Missile sources were considered for an area extending-2000 feet from the containment building which resulted in the assumption that thousands of missiles were possible during a tornado.

l The area includes all of the structures on the plant site. The

, additional missiles that would be expected at the failure of a structure l located at the site is expected to be only one of a number of contributors to the number of missiles already assumed in the analysis.

OC0287-0018A-NLO2 1

- e 2

The only situation in which this assumption might not be valid is in the event vulnerable structures are in the immediate vicinity of a building undergoing failure and.the resulting missile density is momentarily higher than is occurring at the rest of the site. Because of a lack of missile shielding, the structures most vulnerable to missiles include the turbine / service building, the cable penetration area and the screenhouse.

Again examining the site layout, the cable penetration area is in close proximity to the turbine teilding and may be more vulnerable to missiles generated by turbine building failure than other structures than was assumed in the missile analysis. Several features of the cable penetra-tion area and the missile analysis suggest that this additional vulnera-bility may be minimal, however:

the non-protected side of the cable penetration building is not oriented toward the turbine building and has as shielding between it and the turbine building a three-story high concrete wall several feet thick.

The missile analysis assumes that a single missile entering the room will fail all functions having circuitry in the room (which offsets any unconservatism that may occur as a result of assuming a uniform missile density).

Because of these features, it is not considered necessary to adjust the missile analysis.

Question 2:

The PRA analysis did not consider the effect of tornado and wind loadings below 150 mph. Structural capacities presented in Table 1 of the PRA analysis as well as capacities calculated by the staff indicate that the capacities of certain structures lie below 150 mph. How does this affect the conclusions reached in the PRA?

Response

j In response to this question, Consumsers Power Company elected to i generate the new risk analysis presented in this submittal which includes the risks associated with winds and tornado loadings below 150 mph. The conclusion of this analysis is that providing protection against winds and tornadoes for equipment associated with the alternate shutdown system plays a relatively significant role in reducing the risks of severe weather. This reinforces the conclusions reached in the original j evaluation.

Question 3:

The PRA analysis only considered structural capacity in terms of velocity pressure (wind flow around a structure) and did not consider capacity in terms of differential pressure (pressure drop). The capacity to resist differential pressure should be converted to an t associated windspeed and the lesser used to determine probability of exceeding that windspeed. Staff calculations do not always support

(

l OC0287-0018A-NLO2

e g 3

capacities presented by the licensee regarding resistance to differential pressure. Note that in Section 4.5.2 of the IPSAR, the staff concluded that the containment is adequate to resist any loads induced by differential pressure and thus this aspect of containment failure need not be considered in the PRA analysis.

Response

It is correct that the wind speed resistances were used as a basis for the risk analysis rather than the most conservative of wind or delta p resistances. A sensitivity study was performed to estimate the effects of including the delta p values in the risk analysis.

From the Texas Tech report (see NRC to CPC letter dated December 17, 1980) the effects of atmospheric pressure change must be taken into consideration above a windspeed of 140 mph as at this windspeed the tornado model begins to dominate over the straight wind model. The windspeed range at which this differential pressure would have the most effect is between 110 and 150 mph. In this windspeed range, the screen-house and the emergency diesel generator room have not failed due to straight winds. However when the differential pressure values from Table 1 of Appendix C are converted to wind speeds (a calculation was performed using information supplied in ANSI /ANS-2.3 1983) the screen-house would fail at 138 mph and the diesel generator room at 150 mph, as opposed to 152 and 193 respectively. If the windspeed range were to be lowered to 139 mph from 150 mph to keep the screenhouse intact, the change in initiating frequency for this interval would be insignificant (less than one percent). The original initiating frequency was deter-mined by subtracting the frequency of exceeding 150 mph from that of exceeding 110 mph which is (7.0 E-4) - (9.0 E-6) = 6.91 E-4, changing to 140 mph the initiating frequency would be (7.0 E-4) - (1.0 E-5) - 6.9 E-4.

For the emergency diesel generator room, a larger difference of 40 mph is obtained. The effect of losing the diesel generator room with a probability of one at 150 mph will be examined at in the'150 mph to l 190 mph range as there are fewer makeup systems available (in this range l

the screenhouse has failed and there is only the portable pumps and control rod drive pumps available for primary system makeup). From the baseline quantification (no portable pumps) no effect is made in the core damage estimate since the limiting building is in the screenhouse.

With the portable pumps installed lowering of the EDG room windspeed failure has minor impact on the core damage estimate for the 150 mph to 190 mph range only. With the EDG failed, the only PCS makeup system is the portable pumps through the core spray valves. Changing the system I failure value and requantifying the sequence results in changing the windspeed- interval core damage probability f rom 9.85 E-8 to 1.03 E-7.

The overall core damage estimate for winds and tornadoes and tornado missiles was calculated to be 1.04 E-5 assuming portable pumps. There-fore, when comparing the change in EDG windspeed failure of about 4.5 E-9 to the overall core damage esticate no effect would be realized and the effects of using differential pressure failure would not change the final results.

OC0287-0018A-NLO2

< e 4

Question 4:

Other than containment, it does not appear _that the effects of the loads imposed by missile impact on other structures has been considered.

Response

The effect of missile impact on structures was treated as follows in the risk analysis.

In both the original and the revised risk analysis it was assumed that penetration of a structure by a missile results in the failure of all equipment within that structure.

Penetration of concrete structures was handled as follows. In its November 29, 1982 safety evaluation on this topic, the NRC staff found equipment protected by eight inches of reinforced concrete to be protected from missiles the size of a one inch reinforcing rod and 12 inches of reinforced concrete to provide adequate protection from missiles as large as a telephone pole. In this regard, the staff found the reactor coolant pressure boundary, the reactor core, shutdown cooling, reactor cooling water, core spray and the spent fuel pool to be adequately protected from t'he effects of tornado missiles. EPRI report NP-768 found walls of l lesser thickness to provide some protection from missiles and assigned a probability of failure to these structures. Concrete walls less than 12 inches were not treated in a probabilistic manner in Consumers Power Company's analysis, as EPRI suggests, but were conservatively assumed to fail with a probability of one upon missile impact.

Structures providing no protection by way of reinforced concrete (such as masonry walls, metal siding or decking) were assumed to fail with a probability of unity if struck by a missile.

Question 5:

Provide your basis for the assumption that missile damage to the cable penetration room and station power room occurs prior to or simultaneous with a loss of offsite power 50 percent of the time.

Response

It was not necessarily conservative to assume that the loss of of fsite power always accompanied damage resulting from winds, tornados or tornado missiles. This is because the undervoltage condition brought about by the event is the actuating signal for certain safety features such as i

main steam isolation and emergency condenser actuation. By performing the risk analysis with and without AC power the risk associated with sequences with and without the primary system isolated could be assessed.

The 50 percent assumption was in effect a sensitivity study on the progression of the two types of transients associated with wind, tornado and tornado missile damage.

OC0287-0018A-NLO2 l

e .. .-

5 With the installation of the alternate shutdown system, missile damage to these systems is not as likely and the risk analysis was performed assuming no offsite power is available.

Question 6:

, Has consideration been given to the effect that dirt and debris generated by the tornade will have on components formerly protect'ei by siding.

Response

Yes, the risk analysis assumes that the functioning of any active components within a structure is lost at the time the structure fails and exposes the components to the environment associated with the storm. A small exception occurs in the case of the cable penetration area.

Circuitry within the room is assumed to remain functional for wind speeds slightly above the speed at which the siding is assumed to fail. This assumption is made for the purposes of simplifying the accident sequence quantification, and is considered justifiable in that the room contains only power and control cables (it contains no cetive components important to shutdown following a tornado) and it remains protected by reinforced concrete and the containment shell on all but one side after the siding is lost.

Question 7:

Tornado missile analyses performed per EPRI NP-768, 769 and 2005 are generic. For plant-specific reviews, the evaluation must take into account such items as plant geometry, potential missiles as a result of site review, and missile location relative to plant structures. It l appears that such an analysis has not been performed for Big Rock Point l but rather, extrapolations from other analyses were made. Justify your conclusions regarding tornado missiles in light of differences between the EPRI reports and the extrapolations performed for Big Rock Point.

I Response:

1 A detailed response to this question justifying the use of the generic EPRI data for the Big Rock Point Plant can be found in Appendix D of this report.

Question 8:

l Below a windspeed of approximately 165 mph (using the Mcdonald's upper j 95th percentile tornado estimate), the probability of exceeding l threshold windspeed for straight winds dominates over the probability of I

exceeding threshold windspeeds for tornadoes; however, the PRA presented by the licensee only considered plant risks from tornadoes. Therefore, plant risk from straight winds must also be considered, t

OC0287-0018A-NL02

r sn

• 6

Response

The updated analysis considers the effect of straight winds and the effects of tornadoes. When the windspeed interval probabilities were determined the frequency for any particular windspeed were taken directly from Figure 6 in Appendix C (Tornado and Straight Wind Hazard Probability Model for Big Rock Point). From this graph the effects of straight winds below about 140 mph dominate the model, and above '40 mph the tornado model, and above 140 mph the tornado model dominates with the respective contributions of pressure change and missiles. Therefore, the straight wind model was used for winds below 140 mph and the tornado model was used above 140 mph.

Question 9:

It was stated that 80 mph corresponds to the original design criteria, below which no damage from tornado missiles is postulated to occur.

Justify that no unacceptable missile damage will occur below 80 mph given that the original design, even through 80 mph, did not consider missiles.

Response.

The effects of missile damage has been incorporated into the event tree quantification by adding to the system logic models an additional factor which takes into account the probability of missiles impacting buildings. The data for the missile impacts was obtained from the EPRI study (refer to report NP-768) and contain impacts from missiles generated from winds down to 73 mph (see Tables 2, 3, and 4 in Appendix C). It is noted here that the Texas Tech report (see NRC to CPC letter dated December 17, 1980) states "For windspeeds above 139 mph the tornado model governs. In the case of a tornado, the effects of atmospheric pressure change and missiles must be taken into account in 1 addition to the wind effects." Therefore, given the cutoff of data from the EPRI study at 73 mph and the information supplied in the Texas Tech report, the effects of missiles has been adequately included in the event tree quantification.

l l

l l

l I

OCO287-0018A-NLO2

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