ML20024E769

SEP Topic IX-5, `Ventilation Sys, Addl Study to Support CPL 830623 Pra

ML20024E769
Person / Time
Site:	Big Rock Point File:Consumers Energy icon.png
Issue date:	08/31/1983
From:	CAROLINA POWER & LIGHT CO.
To:
Shared Package
ML20024E765	List: ML20024E764 ML20024E769
References
TASK-09-05, TASK-9-5, TASK-RR NUDOCS 8309070035
Download: ML20024E769 (15)

Text

,

W ATTACHMENT Consumers Power Company big Rock Point Plant Docket 50-155 i

SEP Topic IX-5 " Ventilation Systems" -

Additional Study to Support the June 23, 1983 Consumers Power Company PRA August 31, 1983 8309070035 830831 85 e pages PDR ADOCK 05000155 p

PDR ic0883-0242bl42

Safety Evaluation Report Review of SEP Topic IX-5, Ventilation Systems Big Rock Point Nuclear Plant Docket Number 50-155 Revision 2

1.0 INTRODUCTION

As a result of its contractor's assessment, the NRC staff has concluded that the ventilation systems at Big Rock Point are in gonformance with current criteria for this topic, with three exceptions. These excep-tions are as follows:

1) Loss of the shop area ventilation system due to a loss of offsite power initiating event may result in unacceptable concentrations of hydrogen in the reactor depressurization system (RDS) uninter ruptable Power Supply (UPS) rooms.

2) Loss of ventilation cooling in the Electrical Equipment Room could cause overheating in this room, leading to failures of vital electrical equipment. Also, hydrogen generated during battery charging may reach inflammable levels.

3) The passive ventilation system in the emergency diesel generator room may be inadequate during extended operation of this generator.

These three areas will be discussed in the following sections.

l I

h i

l l

)

MIO81683-NL01 L

2.0 REVIEW CRITERIA The current licensing criteria are identified in Section 2 of Franklin

~

Research Center report TER-C5257-414. " Review of the Design and Opera-tion of Ventilation Systems for SEP Plants - Big Rock Point."

In determining which systems to evaluate under this topic, the staff used the definition of " systems important to safety" provided in Regulatory Guide 1.105.

The definition states that systems important to safety are those necessary to ensure (1) the integrity of the reactor coolant pressure boundary, (2) the capability to prevent or mitigate the consequences of accidents that could result in potential offsite exposures comparable to the guidelines of 10CFR Part 100,

" Reactor Site Criteria." This definition was used to determine which systems or portions of systems were " essential." Systems or portions of systems which performed functions important to safety were considered to be essential.

MIO81683-NL01

3.0 RESPONSE TO STAFF EVALUATION The following subsections' address the'three exceptions noted in the SER released for this topic. The probabiligtic risk assessment (PRA) performed for the Big Rock Point Plant was us,ed extensively in this analysis.

3.1 UNINTERRUPTABLE POWER SUPPLY ROOMS - SHOP AREA VCNTILATION The ventilation system for the UPS rooms is not powered from the emergency bus. Therefore, following a loss of offsite power initiating event there will be no circulation of air in these rooms. Each of these four rooms contains a power supply, in the form of batteries, to power one train of the RDS. Operating Procedures call for placing the batteries on the emergency bus, and once connected, continue to charge the batteries regardless of the status of the offsite power supply.

Fully charged batteries will generate hydrogen during charging, due to the electrolysis of water. Without ventilation, the concentration of hydrogen could build to 4 volume percent, where combustion is possible.

If this was to occur the power supplies for the RDS may be damaged and the RDS would be unavailable if required at some later time in the transient.

To determine the effect of loss of ventilation in this area following a loss of offsite power, it is necessary to estimate the amount of time required for the hydrogen concentration to build to inflammable levels.

i For this calculation it will be assumed that all batteries are fully charged at time T=0, and that the battery chargers are in operation.

For conservatism, it will be assumed that the battery chargers supply 6 amperes continuously to a fully charged battery, although the charging current tapers off as the batteries reach full potential. Hydrogen generation is greatest with a fully charged battery, since the charging current goes into the electrolysis of water (H 0-+ H2+ O ), rather 2

y than into charging the battery. This extreme conservatism will give gross overestimates for the amount of hydrogen gas generated. Actually, a fully charged battery receives only a trickle charge (fraction of a milliampere), whereas for a depleted battery, the charging current goes into charging the battery, rather than into generating hydrogen.

Also, it is assumed that hydrogen burning occurs immediately when the volumetric concentration of this gas reaches 4%.

This assumes that a source of ignition is present in the rooms which, from investigation, appears very unlikely. Finally, no credit for natural air addition or leakage through doorway seals is assumed. This is an added conver-satism since revised operating procedures provide for the opening of doors to the UPS Rooms and Elegtrical Equipment Room within two hours after a loss of station power.

The calculation proceeds as follows:

t i

M1081683-NL01

Calculation for time until hydrogen burning occurs in UPS Room

' Room size = 8'x12'x10' = 960 ft3 (Reference 5) i Volume of Batteries = (12"x6"x8") x 21 = 7.04 ft3 (Reference 7)

Volume of room cabinets = (56" x 75" x 36") = 87.5 f t g (Reference 10)

The gassing rate is a function of charging current. For a fully charged battery, the volyme of hydrogen generated is approximately 1 ft per 63A-hr per cell.

With 60 cells and a 6 ampere charging current.

3 Total hydrogen generation rate = (6A) (60 cells) 1 f t H, -

- 5.71 ft H,

~

63 A-hrs: cell hr Combustible concentration of hydrogen = 4% of free volume in room.

Free volume = total 3olume-bagteryvolung-cabinetvolume 7.04 ft - 87.5 ft

= 960 ft 3

= 865.5 ft 3

4% free volume = 34.6 ft With the normal concentration of hydrogen in air being near 0%, the time to reach 4% concentration is then:

34.6 ft /5.71 ft /hr 6.1 hrs

=

Therefore, if it is assumed that burning occurs in all four UPS rooms 6.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after the loss of offsite power and ventilation, that no offsite power is restored before this time, and that the burning disables the RDS, then the probability of RDS failure is 1.0 for all loss of offsite power accident sequences in which the RDS is required more than about 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> into the transient.

The time at which RDS actuation is required for different accident scenarios is given in the Big Rock Point PRA on Pages I-93, I-158, XVI-3, and XVI-20.

These items are:

Failure mode: E time 1.9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> i

E, (no make-up) 5-6 hours d (1 tube bundle 15.25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> E

l removing heat) i d (2 tube bundles 13.55 hours6.365741e-4 days <br />0.0153 hours <br />9.093915e-5 weeks <br />2.09275e-5 months <br /> E

removing heat)

IY 1.67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br /> The first four failure modes are different failures of the emergency condenser, whereas the last one is a failure of the main steam isolation valve c

to close coupled with a failure to provide inventory makeup. All of these failures will result in eventual depletion of the steam drum inventory so that l

MIO81683-NL01

RDS and core spray operation are required to maintain core cooling. Sequences which contain failure modes E or lY require the RDS before 6.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, and therefore are unaffected by tee generation of hydrogen. For conservatism E sequences will be included. However, those loss of.offsite power sequences In which are failures of the makeup system to the condenser sEell) d (both of which the emergency condenser fails due to failure modes E or E are affected.*

An investigation of the PRA accident sequences reveals that two such sequences exist: PE F C and PE F C.

Additionally, if the probability 'f RDS failure o

d (event C)"il equal to Enity, then other sequences are eliminated, since they are conditional on successful RDS operation. These sequences are PE F L and ms PE,F F L (long-term cooling failure sequences).

y The increase in core damage frequency due to the loss of the RDS following hydrogen burn is calculated as:

ACDF = PE,F C (new) + PE F,C (new) s d

- (PE F C (old) + PE F C (old) + PE F L (old) + PE F F L (old))

ms ds ms ms1 Note that this accounts or.ly for possible RDS failure due to hydrogen burning.

The failure probability of the RDS becomes unity, with the conservative assumptions employed, only if offsite power is not restored within 6.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

From Table III-13 of the PRA, the probability of failure to restore power within 6.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> (event F ) is approximately 0.10.

Using system failure values from the PRA, the Sew sequences are quantified as:

-5

~

PE,F,C = (0.13) (1.8x 10 ) (0.10) (1.0) = 2.3 x 10 E, = Failure to transfer demineralized water pump and air gmpressor to 2B bus (0.25) and failure of VEC-1 to open (7.1x10 )

-3

= 1.8x10 PE

"(*

} (**

(*

~

ds

-5 E = 7.2x10 d

The values of E and E determined here take into account the fact that d

the emergency cEndenser make-up valve is now a DC solenoid operated valve.

• Loss of offsite power sequences with failure of the emergency diesel generator are not affected because of the loss of power to the battery chargers, and therefore the cessation of hydrogen generation.

MIO81683-NLO1

The "old" sequences, as quantified in the PRA Table XI.3 (with modification 1) and adjusted-to account only for possible RDS failure, are:

~

PE F C

= 3.4x10 ma

-0 PE F,C

= 2.4x10 d

~

PE,F,L

= 3.4x10

-8 PE,F,F L = 5.1x10 y

The increase in core damage frequency is then:

- 5.1x10-8)

-(3.4x10-7 + 2.4x10-8 + 3.4x10~7 ACDF = 2.3 x }0-5 + 9.4 x 10-7

~

= 2.3x10 This is a very conservative value due to the conservative assumptions that:

1. Very-high current values are assumed to be charging fully charged batteries

2. No dilution occuro, ie, the rooms are essentially leak tight

3. Complete combustion of hydrogen occurs immediately at 4% concentration by volume and

4. The RDS is completely disabled once hydrogen burning occurs.

This increase in CDF becomes significant contributor to the overall core damage frequency as now calculated in the PRA. However, it is very easily decreased back to the values assumed in the PRA. Three potential solutions are presented here, with one of them having very obvious advantages. These solutions are:

1. Remove the battery chargers from the emergency bus within 6.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> of a loss of offsite power (if power is not restored). Place on the bus after power is restored.

2. Open the UPS room doors within 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> of the LOSP initiating event.

Close the doors when power is restored.

3. Place the power supply for the ventilation system of the UPS rooms j

onto the emergency bus.

The disadvantages of option (1) are many. Primarily, this option requires that the operator keep track of the running time of the event in progress.

He must take care, when removing the chargers from the emergency bus, that other essential equipment is not inadvertently removed from service. Also potentiallyimportantisthefactthatUPSbatgeriescansustaintheir l

outputs for only one hour if charging is lost MIO81683-NL01 l

Two disadvantages are.also apparent with option (2),'but these are much easier to_ reconcile.- The first-disadvantage is as in option (1),

namely the operator must keep track of the running. time of the event.

However, if the operator is instructed to open the UPS doors immediately following a LOSP initiating event, this problem is overcome.

-The second problem is that' opening the_ doors exposes the batteries to the danger of fire. However, the probability of a fire occurring near this area coincident with the LOSP is very low,'and the sequence can be operators to open UPS doors within 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> of a LOSP.gow call for considered incredible. Revised operating procedures Option 3 also has its disadvantages. Primary among these is that placingtheshopheatingandventilationunitandoneggantexhaustfan on the emergency bus would draw approximately 55 amps.

In addition to the loads already supported by th

• ** *1*"*#'#8 loads are not to exceed 150 amps.T0****8'"#Y Potential loads recommended for consideration by S0P 28 (Reference-10) are a reactor cooling water pump (50 amps), a control rod drive pump, (40 amps), a core spray pump (50 amps), and the clean-up pump (15 amps), among other equipment.

Obviously, placing the ventilation system on the bus will severely limit the additional equipment.which may be necessary. For this reason option 3 is not recommended for implementation, but option 2 is.

3.2 ELECTRICAL EQUIPMENT ROOM VENTILATION 3.2.1 Equipment Failure due to Thermal Overload The electrical equipment room ventilation subsystem services essential 4

equipment which includes the main plant batteries, two motor-generator sets, air compressors for instrument air, 480V switchgear; and cable spreading. The ventilation system is mainly a service-water-cooled, recirculation room cooler which is neither redundant nor powered by the emergency diesel. Loss of the service water system, or' failure of the recirculating room cooler, could cause the temperature in this room to rise to the point where equipment damage due to thermal overload may occur.

4 The sequences of concern for this area are only those in which offsite power is still available. A loss of offsite power would cause the sources of heat in the area (the 480V switchgear and the 2 motor generator sets) to cease operation. Without this heat, the room will remain cool.

i The service water system is comprised of two pumps, one of which is i

running while the other is on standby. These pumps are alternated once every week. Big Rock Point did not experience a failure of these pumps during the 87.648 hours0.0075 days <br />0.18 hours <br />0.00107 weeks <br />2.46564e-4 months <br /> for which data was compiled in the PRA (see Table III-3, page 111-10). Therefore, for conservatism, generic failure rates for pumps will be used to estimate the probability of service water system failure. From Table III-4b of the PRA 4 alternating pumps fail to start with a probabilgty of 3 x 10/ demand, and fail to operate with a probability of lx10 / hour. Loss of service water requires that the running pump fail to operate and the standby MIO81683-NL01

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

1 pump fail to start. Using one year as the mission time, the -5 probabilityoflossofse_rgicewaterisestimatedtobe(g10 / hour) x (168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> / weeks) x (3x10 ) x 52 or approximately 2.6x10 Now, if it is assumed that the recirculating room cooler is of a type similar to a lube o Q cooler, then Table III-4c provides an estimate of 1.5x10 / hour for the failure of this cooler to operate. A one year mission time will be assumed for g is cooler, with the resulting 4 failure probability being (1.5x10 / hour) x (8640 hours0.1 days <br />2.4 hours <br />0.0143 weeks <br />0.00329 months <br />) = 1.3x10 Therefore the probability of total loss of g ntilation cooling in the electrical equipment room is roughly 1.6x10 This failure probability appears to be insignificant by itself.

However, for completeness assume that this loss of ventilation approximates a loss of offsite power initiator, except that power is lost due to electrical equipment failure following overheating, and no credit for power restoration is possible since machinery in the room would require extensive repair or replacement. If it is assumed that' the heatup of the room occurs at a rate of 15'F/ hour, that equipment-failure begins at 100*F (both very conservative assumptions), and that the beginning temperature of the room is 75*F, equipment failure will begin in approximately two hours ((100-75)/15 = 1.67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br />). Under the worst case assumptions employed here, if restoration of ventilation is not accomplished within two hours of its failure, a loss of offsite power type of sequence will be initiated. This would require that the emergency condenser be in service to remove decay heat.

The emergency condenser requires cakeup if it is used in the long term.

Makeup under these conditions would require opening of the fire water makeup valve, SV-4947, since it is conservatively assumed that the air compressor (required to open the air operated makeup valve to the emergency condenser) is unavailable due to the failures in the electrical equipment room.

It is also assumed that the demineralized water pumps are out of service due to a loss of AC power.

If makeup is not provided, the steam drum inventory will eventually decrease af ter reactor pressure rises and steam is released through the safety valves.

i The RDS would be necessary so that core spray could be provided to cool i

the core. The sequence of concern, using the terminology of the Big Rock Point PRA, is denoted as P'E F 'C. This sequence is quantified as ms follows:

i

• Note: A procedure does exist at Big Rocc Point which provides cooling water to the air compressors with well water. Therefore, a loss of service water /

ventilation cooling would not by itself result in overheating of the air j

compressors. However, for this analysis it is assumed that the air compressors overheat. See Section 3.2.2 for sensitivity analyses concern-ing this assumption MIO81683-NL01

where P'E,F,'C = (1.6x10~4) (.01) (1.0) (3.7x10~ ) = 5.9x10~

P' = loss of power due to loss of ventilation cooling in electrical

~

equipment room

= 1.6x10 E,= Failure to provide makeup to the emergency condenser

= Failure of solenoid operated valve to open or failure of diesel fire pump to start (7.1x10~3 + 3.1x10~3)

F '= Failure to restore power in electrical equipment room before RDS s

operation

= 1.0 C = Failure of RDS/ core spray given that no AC power is available

-2

= 3.7x10 This sequence is below the 10~ cut-off point below which sequences are considered insignificant, and furthermore is very conversative in that no credit for operator action has been employed in this calculation.

Operating procedures now call for the cpening of doors and the possible use ofportablefansintheelecgricalequipmentroomifstationpoweris lost for more than two hours.

i 3.2.2 Sensitivity Analysis for Failure of Electrical Equipment Room Components due to Thermal Overload If it is assumed that the air compressors are available and are being cooled by well water, makeup could be provided with the demineralized water pump if power could be supplied for its operation. For this sensitivity analysis, assume that some source of power is available to the demineralized water pump (but not to the fire pumps). Then sequence P'E,F,'C is quantified as:

-8 P'E,F,C = (1.6x10-4) (2.53x10~ ) (1.0) (3.7x10~ ) = 1.5x10 where E = (Failure to transfer demineralized water pump to emergency power source or failure of demineralized water pump to run g failure of makeup valves to open) and (Failure of SV-4947 to open)

=(.25+lyx10

+ 2x10~3) x (.01)

~

= 2.53x10 Again, probability of this sequence is below the 10~ / year cut-off MIO81683-NL01

3.2.3 HYDROGEN GENERATION RATE FROM THE BATTERIES IN THE ELECTRICAL EQUIPMENT ROOM To determine the effect of a loss of ventilation cooling in the Electrical Equipment Room, the time required for the hydrogen concentration to build to inflammable levels must be estimated. It will be conservatively assumed that all batteries are fully charged at time t = 0 and all battery chargers are in operation. Note that the battery chargers for this room are not on the emergency bus.

If a loss of offsite power did contribute to ventilation cooling failure, operator action would be required to connect the chargers to the emergency bus.

The probabilistic risk assessment (PRA) evaluations of SEP Topic IX-5,

" Ventilation Systems," Rev. 1, incorporated an incorrect calculation for the rate of hydrogen generation during battery charging. The gassing rate is a function of current, not voltage. Since a battery cannot absorb all of the energy from the charging current toward the end of charge, the excess energy dissociates the H O by electrolysis 2

The hydrogen gassing rate f r a fully into the gasses H and 0,,,

Per 63 ampere-hours.9 2

This is charged cell is 1 cubic foot of H 2 equivalent to 7.49 cubic centimeters of hydrogen per ampere-minute per cell.

The batteries in this room, when fully charged, receive only a fraction of a milliampere of current. For depleted batteries, the charging current will jump up to a high level and decay rapidly. Note also that when less than fully charged, the charging current does not produce hydrogen at the above rate.

Assuming the batteries receive 15 amperes continuously (different battery chargers charge the station batteries and the UPS batteries) and that all current produces hydrogen, the generation rate is:

(15) (7.49) = 112.4 cc H / minute / cell 2

Since the batteries are comprised of 60 cells (125V total), a hydrogen generation rate of:

(60) (112.4) = 6740 cc H / minute 2

=.238 ft / minute is expected.

The free volume of the room is estimated to be 34,417 cubic feet; 4% of this is 1377 cubic feet.

l To produce a volume of 4% hydrogen would require

)

l 1377 = 5780 minutes = 4.0 days l

.238 l

The PRA attached to Consumers Power Company's March 31, 1983 submittal assumed rapid diffusion of hydrogen in the air. Also, it was assumed that no air flow occurred with the outside air, an unrealistic, but very conservative i

l MIO80583A-NL01 l

assumption. In fact, the door between the Electrical Equipment Room and the Machine Shop is a grid-like structure that should allow a free flow of air.

Furthermore, as documented in our 3/31/83 submittal, procedural revisions have already been incorporated to instruct operators to facilitate air flow by opening various vents and doors in the event that normal ventilation systems are disabled.

The possibility of hydrogen " pocketing" has arisen.12 However, hydrogen has a high diffusion coefficient and would tend to disperse into the ambient atmosphere, rather than rise due to its buoyancy. " Hazardous dust, fumes, vapors and gases are truly airborne, following air currents and are not subject tg3 ppreciable motion either upward or downward because of their own a

density."

Since the hydrogen will mix with the air, specific gravity effects are insignificant.

The rate of diffusion above thgbatteries can be estimated by starting from first principles of diffusion; Ficks law gives:

N

=-D H

A y

Ps 2

=

ux gm es/cm sec]

Where NA 3

C = concentration [g moles /cm ]

s = distance (cm]

2 D = diffusivity [cm /sec) y Using the ideal gas law, the above equation integrates to N

=D (C-C ) = D P (y-y )

A y

g g

BP RTB f

f D =.76 cm /sec (ref 15)

R=

82.06 cc atm/ degree-mole P = 1 atm T = 298K y = mole fraction y = mole fraction at interface -(y-y ) = 1.00

.04 (max) = 0.96 g

B = diffusion layer thickness. From Fick's law, we see that this distance f

corresponds to thg concentration change of 0.96.

Assume it is roughly 1 cm.,

which is about 10 mean free paths.

-5 2

N = 2.98 x 10 moles

=.73 cm /sec-cm A

2 sec-cm 2

Assumingaventareaontheorgerof100cm for the battery (63 vents, each with an area of approx. 1.5 cm ):

N x Area =

cm /sec A

MIO80583A-NL01

The generation rate for all of the cells is.118 cm /sec The diffusion and generation of hydrogen gas is occurring within several centimeters above the battery. Since these rates are of the same magnitude, and since the ceiling is on the order of meters above the batteries, buoyancy does not dominate, and hence " pocketing" of

-hydrogen does not occur.

3.3 EMERGENCY DIESEL GENERATOR ROOM VENTILATION The main emergency diesel generator room, located adjacent to the screenhouse, has a passive ventilation system. The passive ventilation system consists of ventilation louvers which allow outside air to enter the emergency diesel generator room. Exhaust from the diesel engine is through the roof of the building.

The heat from the engine will_ produce a temperature rise in the diesel generator room. The effect of the rise in room temperature could be to cause failures on the electrical control panel for the diesel during sustained operation.

The emergency diesel generator is run once each year for a period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for test purposes. During this time the temperature in the room does rise to the point where the operators open the door to provide additional air circulation. This additional ventilation is more than adequate for cooling purposes. However, these tests have' demonstrated the need for additional ventilation. As a result, a modification is currently in progress in which a damper will be powered directly from the diesel, and will open upon sensing a high temperature in the room.

l It is felt that the installation of this damper will adequately address the concerns of the staff regarding this room, and therefore no further action is necessary.

3.4 COST ANALYSIS A cost-to-cost risk reduction ratio (dollars / man-rem) can also be used i

as an aid in determining the maximum amount which should be spent for i

any modifications which reduce the risk investigated in this topic. For this purpose it will be assumed that if a proposed modification has a cost-to-benefit ratio of less than $100/ man-rem it should be considered for implementation, while if it has a ratio greater than $1000/ man-rem, it should not. A ratio in between these values should be considered more carefully. These values are consistent with generally accepted

(

views on societal risk and the criteria which were used to evaluate design modifications in the PRA.

The decrease in man-rem dose is calculated as follows:

M-R reduction = ( 6CRF) x (L.F.) x M-R x T i

L.F.

l l

l MIO80583A-NL01 l

i where.ACRF = change'in containment release frequency (per reactor. year) as a result of the modification (equals change in core damage frequency times probability of containment failure)

L.F._= latent fatalities expected'as a result of a core melt accident at Big Rock Point = 59.4 (see PRA, Reference 3)

M-R = man-rems required to result in one latent fatality = 10,000

~

L.F.

(assumed in PRA)

.If it is assumed that proposed modifications are perfect in reducing the riskduetoinadequateventilgion,thechangeincoredamagefrequency can be estimated to be 2.3x10 / year. The probability of containment failure as calculated in-the PRA is 0.064.

T, the remaining years of operation for Big Rock Point, is 18. Therefore, the maximum M-R reduction is:

-5 M-R reduction = 2.3x10 / year x.064x59.4 L.F. x (10,000 M-R) x 18 yrs

= 15.7 man-rems L.F.

Therefore, using $1000/ man-rem as an allowable expense, the maximum amount which can be spent is $1000x15.7, or $15,700.

j The modifications recommended here (ie, opening of the UPS room doors if

'l a LOSP initiating event occurs and opening of the roof fire damper and shop doors in the event electrical equipment room cooling is lost) involved revising currently existing plant procedures. The cost of incorporating such a revision is roughly $2000. This results in a cost / benefit ratio $2000/M-R or $127/M-R. These procedural revisions are therefore-cost effective.

f

4.0 CONCLUSION

S No further action appears necessary to address the concerns for the electrical equipment or emergency diesel generator rooms. Sequences of concern due to loss of ventilation in the electrical equipment room are probabilistically insignificant, and in addition can be further reduced by very simple operator actions. The currently on-going modification to the emergency diesel generator room will provide adequate cooling there.

This modification will be complete by the end of 1983.

Loss of A potential problem has been identified in the UPS rooms.

ventilation could, under extreme circumstances, disable the RDS through the process of hydrogen burning. A procedure has been implemented wherein the operators will open the doors to the four UPS rooms if offsite power is not restoud within two hours of its failure.

Failures of electrical equipment room components due to possible j

hydrogen burning or to thermal overload have been addressed. Under worst case conditions for both events, operator action, as outlined in the new operating procedures, should eliminate equipment failure.

MIO80583A-NLO1

e..

5.0 REFERENCES

1. Franklin Research Center, " Review of the Design and Operation of Ventilation Systems for SEP Plants - Big Rock Point," TER-C5257-44, August 13, 1982.

2. Letter LS05-82-10-030, Dennis Crutchfield to David VandeWalle, October 12, 1982, Docket No. 50-155.

3. Consumers Power Company, Probabilistic Risk Assessment, Big Rock Point Plant, March 1981, Docket No. 50-155.

4. Big Rock Point Operating Procedure, SOP 25. " Loss of Station Power",

and ONP 2.36, " Loss of Station Power".

5. Big Rock Point Piping and Instrumentation Drawing, " Equipment Location, Plan Above Grade," M-100.

6. Telephone Conversation, W. Brinsfield with P Donnelly, November 5, 1982. Fink and Carroll, Standard Handbook for Electrical Engineers, McGraw Mill Brook Company, 1969, Ch 24.

7. Fink and Carroll Standard Handbook for Electrical Engineers, 10th ed., McGraw-Hill Book,1969, Ch 24.

8. Consumers Power Company, Big Rock Point Plant Reactor Depressurization System Description. Operation and Performance Analysis, Revised 2/76, Docket 50-155, License DPR-6.

9. Big Rock Point Plant Manual, Vol. 3. Operating Procedures, SOP 25 Heating and Ventilation System, 3/10/83.

10. Big Rock Point Plant Manual, Volume 3B, Operating Procedures, S0P 28, Rev 9 " Station Power", 1/30/81.

l

11. Telephone Conversation, W. Brinsfield with H. Bazydlo, January 27, i

1983.

12. Telecon, Ray Scholl, USNRC SEPB to J Daiza, CPCo NLD.

13. American Conf. of Governmental Industraial Hygienists, Industrial Ventilation, A Manual of Recommended Practice, 15th ed.,

1979, pp 1-3, 1-4.

14. R.M. Perry and C.M. Chilton, editors Chemical Engineers Handbook, 5th ed., McGraw-Hill, 1973, P.14-4.

l

15. Scott and Cox, Canadian Journal of Chem. Eng., 38:201, 1960.

l

16. Big Rock Point Piping and Instrumentation Drawing, " Single Line Meter and Relay Diagram, Sheet 2", E-102.

l I

MIO80583A-NL01

!