ML20091H875

Final Technical Evaluation Rept,Seabrook Station Unit 1 Station Blackout Evaluation

ML20091H875
Person / Time
Site:	Seabrook
Issue date:	12/17/1991
From:	SCIENCE APPLICATIONS INTERNATIONAL CORP. (FORMERLY
To:	NRC
Shared Package
ML20091H879	List: ML20091H875
References
CON-FIN-D-1311, CON-NRC-03-87-029, CON-NRC-3-87-29 SAIC-91-1801, TAC-M68601, NUDOCS 9112240173
Download: ML20091H875 (32)

Text

. _ _ _ _ _ _ _ _ _ - _ _ _ _ _

F Atu:nment i SAIC.91/1801 i

TECHNICAL EVALUATION REFORT SEABROOK STATION UNIT 1 STATION BLACKOUT EVALUATION ,

TAC No. 68601 SAIC Final December 17, 1991 Prepared fon U.S. Nuclear Regulatory Comadseion Washington, D.C. 20555 Contract NRC4387429 Task Order No. 38 mA'0

\P Poet OMice Bon 1.129,1710 Gooctnoge Orw. Mcleert, Vaprun 221W. Otul G14300

TABLE OF CONTENTS Section fagt 1.0 B A C K G R O U N D ... ....... ..._. .. .... ... ... ...... _.. 1 2.0 REVIEW PROCESS _ _. - -

3 3.0 EVALUATION . 5 3.1 Proposed Station Blackout Duration . 5 3.2. Station Blackout Coping Capability ............... ... 9 3.3 Proposed Procedures and Training ............... 22 3.4 Proposed Modifications . .. .............. .. .. . . 23 3.5 -

Quality Assurance and Technical Specifications . 23 t

- 4.0 - C01 CLUSIONS ... .. . . ... -_ .. 24 REFERENCES.

5.0 . . . _...... 28 1

e 1

P e

s il

)

. - ,y - , , - . ., . . + . . , 3,:n ,--., ,

yw.,,wy,,,.y,%,-y4,nry.,ymm-.-,,, ,,-w,,,,-y- ,-4 y,,- v. ,vv., , --m.-r._ +. ,y..-w ,,,-,,,.w .---my. ..,-- , , . . ,, .

TECHNICAL EVALUATION REPORT SEABROOK STATION UNIT I STATION BLACKOUT EVALUATION

1.0 BACKGROUND

On July 21,1988, the Nuclear Regulatory Commission (NRC) amended its regulations in 10 CFR Part 50 by adding a new section. 50.63, "Ioss of All Alternating Current Power" (1). The objective of this requirement is to assure that all nuclear power plants are capable of withstandmg a station blackout (SBO) and maintaining adequate reaciar core cooling and appropriate containment integrity for a required duration. This requirement is based on information developed under the commission study of Unresolved Safety Issue A-44," Station Blackout" (2-6).

The staff issued Regulatory Guide (RG) 1.155, " Station Blackout," to provide guidance for meeting the requirements of 10 CFR 50.63 (7). Concurrent with the development of this regulatory guide, the Nuclear Utility Management and Resource Council (NUMARC) developed a document entitled. " Guidelines and Technical Basis for NUM.GC Initiatives Addressing Station Blackout at Ught Water Reactors," NUMARC 87-00 (8). This document provides detailed guidelines and procedures on how to assess each plant's capabilities to comply with the SBO rule. The NRC staff reviewed the guidelines and analysis methodology in NUhMRC 87 00 and concluded that the NUMARC document provides ar2 acceptable guidance for addressing the ',0 CFR 50.63 requirements. The application of this method results in selecting a minimum acceptable SBO duration capability from two to sixteen hours depending on the plant's characteristics and vulnerabilities to the risk from station blackout. The plant's characteristics affecting the required coping capability are:

the redundancy of the onsite emergency AC power sources, the reliability of onsite emergency power sources, the frequency of loss of offsite power (LOOP), and the probable

time to restore offsite power.

1 1

l

In order to achieve a consistent systematic response from licensees to the SBO rule and to expedite the staff review process, NUMARC developed two generic response documents.

These documents were reviewed and endorsed (9) by the NRC staff for the purposes of plant specific submittals. The documents are titled:

1. " Generic Response to Station Blackout Rule for Plants Using Alteruate AC Power,"

and

2. " Gene.ric Response to Station Blackout Rule for Plants Using AC Independent Station Blackout Response Power."

A plant-specific submittal, t ing au , W ave generic formats, provides only a summary of results of the analp el e plm's station blackout coping capability.

Licensees are expected to ensure that the bueline assumptions used in NUMARC 87-00 are applicable to their plants and to verify the Le uracy of the stated results. Compliance with the SBO rule requirements is verified by review and evaluation of the licensee's submittal and audit review of the supporting documents as necessary. Follow up NRC inspections assure that the licensee has implemented the necessary changes as required to meet the SBO rule.

In 1989, a joint NRC/SAIC team headed by an NRC staff member performed audit reviews of the methodology and documentation that support the licensees'submittals for several plants. These audits revealed several deficiencies which were not apparent from the review of the licensees'submittals using the agreed upon generic response format. These deficiencies raised a generic question regarding the degree oflicensees' conformance to the requirements of the SBO rule. To resolve this question, on January 4,1990, NUMARC issued additional guidince as NUMARC 87-00 Supplemental Questions / Answers (10) addressing the NRC's concerns regarding the deficiencies. NUMARC requested that the licensees send their supplemental responses to the NRC addressing these concerns by March M,1990.

2

_ _. _ _ _ _ _ _ __ _ _ _ m.. _ _ _ _ _ . _ _ _ _ _ _ _. . _

. e 2.0 REVIEW PROCE.SS The review of the licensee's submittal is focused on the following areas consistent with i the positions of RG 1.155:

A. Minimum acceptable SBO duration (Section 3.1),

B. SBO copi- spability (Section 3.2),

C. Procedures and training for SBO (Section 3.4),

D. Proposed modiGeations (Section 3.3), and

- E Quality assurance and technical specifications for SBO equipment (Section 3.5).

For the determination of the proposed minimum acceptable SBO duration, the following -

factors in the licensee's submittal are reviewed: a) offsite power design characteristics, b) emergency AC power system configuration, c) determination of the emergency diesel generator (EDG) reliability consistent with NSAC-108 criteria (11), and d) determination of the accepted EDG target reliability. Once these factors are snown, Table 3 8 of NUMARC 87 00 or Table 2 of RG 1.155 provides a matrix for determining the required ,

coping duration.

For the SBO coping capability, the licensee's submittal is reviewed to assess .the availability, adequacy and capability of the plant systems and components needed to achieve.

and maintain a safe shutdown condition and recover from an SBO of acceptable duration which is determined a$ve. The review process follows the guidelines given in RG 1.155, Section 3.2, to assure:-

a. availability of sufficient coWmte inventory for decay-beat removal, 3

b adequacy of the class ;E battery capacity to support safe shutdown.

c. availability of adequate compressed air for air-operated valves necessary for safe shutdown,

d. adequacy of the ventilation systems in the vital and/or dominant areas that include equipment necessary for safe shutdown of the plant,

e. ability to provide appropriate containment integrity, and

f. ability of the plant to maintain adequate reactor coolant system inventory to ensure core cooling for the required coping duration.

The licensee's submittal is reviewed to verify that required procedures (i.e., revised existing and new) for coping with SBO are identified and that appropriate operator training will be provided.

The licensee's submittal for any proposed modifications to emergency AC tources, battery capacity, condensate capacity, compressed air capacity, ventilation systems, containment isolation valves, and primary coolant make-up capability is resiewed. Technical specifications and quality assurance set forth by the licensee to ensure high reliability of the equipment, specifically added or assigned to meet the requirements of the SBO rule, are assessed for their adequacy.

This SBO evaluation is based upon the review of the licensee's submittals dated April 17,1989 (12), March 30,1990 (13), and September 6, 1991 (15), and the information available in the Seabro6k Updated Final Safety Analysis Report (UFSAR) (14). An audit may be warranted as an additional confirmatory action. This determination would be made and the audit would be scheduled and performed by the NRC staff at some later date.

4 l

l- .

\ ..

J.0 EVALUATION 3.1 Pmposed Station Blackout Duration Licensee's Submittal ne licensee. New Hampshire Yankee (N1W), calculated (12 and 13) a minimum acceptable station blackout duration of four hours for the Seabrook Station Unit I site.

The ifcensee stated (12) that no equipment modifications are required to attain tbc proposed coping dumtion.

The plant factors used to estimate the proposed SBO duration are:

1. Offsite Power Design Characteristics The plant AC power design characteristic group is "P2" based on:

a. Independence of the plant offsite power system characteristics of '11/2,"

b. Expected frequency of grid-related LOOPS of less than one per 20 years,

c. Estimated frequency of LOOPS due to extremely severe weather (ESW) which places the plant in ESW Group "3." and

d. Estimated frequency of LOOPS due to severe weather (SW) which places the plant in SW Group "3."

5 4

c *

2. Esmerpacy AC (EAC) Power Configuration Group t

The EAC power configuration of the plant is "C." Seabrook is equipped with two ,

emergency diesel generators. One EAC power supply is necessary to operate safe-shutdown equipment following a loss of offsite power.

3. Tarpt Esserpecy Diesel Generator (EDG) RonaldHty The licensee has selected a target EDG reliability of 0.975. The selection of this target reliabilhy is based on having an average EDG reliability greater than 0.90, O.94, and 0.95 for the last 20,50, and 100 demands, respectively, consistent with ,

NUMARC 87 00, Section 3.2.4 Review of Ucessee's Sabinittal Factors which affect the estimation of the SBO coping duration are: the independence of the offsite power system g,ouping, the expected frequency of grid-related LOOPS, the estimated frequency of LOOPS due to ESW and SW conditions, the classification of EAC, and the selection of EDG target reliability. The bcensee stated that the '

independence of the plant offsite power system grouping is 11/2." A review of the Seabrook UFSAR shows that:

1. There is one switchyard for the site;

2. During normal operation, power is provided to the safety busses from the main .

generator through the unit auxiliary transformers (UATs);

3. Upon main generator trip, the generator breaker automatically _ opens and offsite i._

power is provided through the UATs; l

l 6

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

0 4 Upon loss of the UATs, per can be prmided to the safety busses through the reserve auxiliary transfonners (RATS).

Based on these and the criteria stated in Table 5 of RG 1.155, the plant independence of offsite power system group is "12."

With regard to the expected frequency of grid-related LOOPS at the site, we can not confirm the stated results. The available information in NUREG/CR 3092 (3), which gives a compendium of infonnation on the loss of offsite power at nuclear power plants in the U.S., only covers these incidems through the calendar year 1984. Seabrook did not enter commercial operation until 1990. In the absence of any contradictory information, we agree with the licensee's statemem.

The licensee stated that it used regional data to obtain an ESW grouping of "3." The licensee provided (15) information on the regional weather data which was used to determine that the site is in ESW grouping "3." The weather data provided by the licensee is not consistent with the weather data given in the plant UFSAR. Table 2 3-6, which gives the expected return frequency for selected fastest-mile wind speeds. The UFSAR data, if extrapolated, indicates that the site is in ESW Group "4," which is consistent with the data given in Table 3 2 of NUMARC 87-00. Since both the UFSAR and NUMARC data are consistent, we consider the Seabrook site to be in ESW Group "4."

With regard to the SW grouping, the licensee cated (15) that the Seabrook site has three transmission lines on two rights-of way. With all three transmission lines (Scobie Pond, Newington, and Tewksbury) in operation, it is assumed that the mhtimum number

~

of lines required for operation per technical specifications (Section 3.8.1.1) is two out of three lines. This could possibly be represented by the Scobie Pond and Tewksbury lines. These two lines sha.re a right-of way for five miles, and, therefore, it is possible for the plant to be opera:ional with one right-of.way. Based upon this the licensee 7

1

assumed a single right of way for its SW-grouping calculation. With a single right of-way, the site is in SW Group "3 " whereas with multiple rights-of way, the site is SW Group "2 " With an ESW Group of ~4." an SW Group of "3 " and an independence of offsite power system grouping of *12." the offsite power design characteristic is either "P3" (NUMARC Table 3 Sa) requiring an eight hour coping duration, or "P3'"

(NUMARC Table 3 5b) requiring a coping duration of four hours provided that pre-hurricane shutdown procedures are implemented. However, with ESW and SW groupings of "4" and "2" respectively, and an independence of offsite power system grouping of "12." the offsite power design characteristic is "I%" requiring a four hour coping durauon.

The licensee correctly categorized the EAC classification of Seabrook as "C." Each unit has two dedicated 7000 kW EDGs, one of which is necessary to safely shut down the reactor.

The licensee selected the EDG target reliability of 0.975 based upon the EDG reliability data for the last 20,50, and 100 demands. The information in NSAC-108 (11) gives the EDG reliability data at U.S. nuclear reactors for calendar years 1983 to 1985. Since Seabrook Station Unit I was not in commercial operation during this period, we do not have any information on the EDG reliability at Seabrook. However, the licensee can choose any EDG target reliability consistent with the minimum required SBO coping duration, provided that it is maintained. He licensee has provided this commitment in its submittal dated March 30,1990 (13). De licensee stated (13) that the se!ceted EDG reliability of 0.975 will be maintained by implementation of a diesel generator reliability program meeting the guidelines of RG 1.155.

In order for Seabrook to have a required coping duration of four hours, the licensee must either implernent pre-hurricane shutdown procedures or it must ensure that at least one line on each right-of-way is available during plant operations. If neither of these two conditions are met, then the licensee must resubmit its SBO coping analysis for an 3

eight. hour duration. Our analysis is based upon the licensee's meeting one of the two conditions and thus remaining a four. hour coping plant.

3.2 Station Blackout Coping Capability T e plant coping capability with an SBO event for the required duration of four hours

's assessed with on the following results:

1. Condensate leventory for Decay-Ilent Reinovat Licenwe's Submittal De licensee initially stated (12) that 107,745 gallons of water are required for decay-heat removal during the four-hour coping period. De licensee modified (15) this value to 131,137 gallons, which includes the condensate necessary for cooldown and steam generator level shrinkage. The individual elements are as follows:

Element Water Needed (gal)

Decay Heat Removal 76,955 Water Required to Remove Sensible Heat 30,790 Steam Generator Level Shrinkage 23302 Total 131,137 The minimum permissible condensate storage tank (CST) level per technical specifications corresponds to 212,000 gallons of water. The licensee concluded (15)

I that adequate supplies of condensate are available to cope with a four hour SBO event.

9

~ - . .

r i

Redew of UcensWs Sebaktal I' sing the expression provided in NUMARC 87-00, we have esthnated that the water required for removing decay heat during the four hour SBO would be ~77,000 gallons. This estimate is based on 102% of the licensed core thermal rating of 3411 M W t. Based on the information from similarly sized plants, we estimate that

~55.000 gallons of water are required for cooldown and ~38,000 gallons are needed for steam-generator level shrinkage. Based on our estimate,170,000 gallons of water are required to remove decay beat and to cooldown to a steam-generator pressure of 250 psig. Since the licensee stated that the minimum CST level corresponds to 212.000 gallons of water, we concur with the licensee's conclusion that there is an adequate condensate supply available to cope with a four hour 500 event.

2. Class-1E Battery Capacity -

Ucessee's Submittal ,

The licensee stated (12) that a battery-capacity calculation verifies that the class-1E batteries have sufficient capacity to meet SBO loads for four hours assuming loads not needed to cope with an SBO are removed from the DC busses. The licensee added that these loads are identified in plant procedure ECA 0.0 (Loss of all AC Power).

In response to questions regarding the difference between the two-hour battery capacity indicated in the UFSAR (Table 8.3-5) and the four hour capacity indicated in the licensee's submittal (12), the licensee stated (15) that the differences reflect the use of actual versus rated load for some loads and load shedding. The licensee added that in its battery capacity calculation, it followed IEEE Std-485, including a temperature correction factor to account for the batteries operating at the minimum

. temperature anticipated during an SBO event and an aging factor to ensure that the 10

batteries will have suf5cient capacity at the end of their design life. The licensee added that it did not include a specific design margin since this margin is included only in the initial battery sizing calculations to allow for future load growth. The load current used for the indisidual pieces of equipment was taken from the sizing calculation for the UFSAR load profiles which were based upon review of the devices in each circuit.

The licensee stated (15) that Seabrook has four safety related batteries and four DC busses with two batteries / busses per train. The normal con 5guration is to have each battery feed its respective bus (one battery /one bus). However, per technical specifications, it is permissible to operate for up to 30 days with the crosstie closed

, between the two busses within a train. i.e., one battery feeding two busses (one battery /two busses). De battery sizing calculation covers both the one battery /one bus and the one battery /two busses configumtions, even though there is a low probability of an SBO occurring at the same time as being in the one battery /two busses configuration.

Review of Licensee's Submittal We did not receive the licensee's battery-capacity calculation. De licensee provided (15) the load profile used to verify that the batteries have sufficient capacity to support the needed loads for four hours. The licensee stated that the load profiles represent the effects of load shedding and the use of actual loads instead of rated loads. Comparing the 15-120 minute UFSAR load profile (Table 83-5) with the 40-240 minute segment of the loads provided by the licensee, we found that the combination of the load shedding and the use of the actual equipment loads resulted in battery loads which are 173.4 ampere (A) lower for the division A battery (a combination of busses A and C) and 117.5 A lower for the division B battery (a combination of busses B and D) than the loads listed in Table 8.3-5. Based upon 11

the information available in the plant l'FSAR and that provided by the licensee, we have the following concerns:

l 1

1. In the load profile- provided by the licensee (15), the loads for the one battery /two busses arrangement is not a direct sum of the loads for the two one battery /one bus arrangements; for the 40 240 minute segment, the combined load profile of busses 11B and 11D (a total of 236.51 A) is not the sum of the l

loads on busses 11B (286.01 A) and 11D (31.9 A), in addition, the load during the 40-240 minute period on the combination of busses !!A and 11C (276.57 A) is less than the load on bus 11A alone (303.07 A); bus 11C should add an i additional 55.1 A to the combined load, which would bring the total load to 358.17 A. The licensee needs to provide justification for the discrepancy between the loads in the one battery /two busses arrangement and sum of the two one battery /one bus arrangements.

2. If we use the corrected battery loads (i.e., the sum of the two individual bus L loads) in conjunction with the battery performance characteristics for NCX.2250 ,

(the Seabrook batteries), we find that the 30< fay technical specification for plant operation with the two batteries crosstied will be in jeopardy (i.e., cannot be justified).

3. Contrary to the guidance of IEEE Std.485, the licensee used a zero design ,

margin (i.e., a factor of 1.0) in its calculation. The IEEE Std recommended design margin is 1.10-1.15. His is necessary to provide a capacity margin to allow for unforeseen additions to the DC system and less-than. optimum

, operating conditions of the battery due to improper maintenance, recent

~

discharge, or ambient temperature lower than anticipated.

4. The licensee used a temperature factor which corresponds to the minimum expected battery-room temperature. The licensee needs to verify that the 12 m.___. ____.m____-_ _ . _ _ _ .._. _. _ _ . _ . _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ . . _

l minimum temperature used is that of the electrolyte and ensure that under no circumstances will the electrolyte temperature drop below the assumed temperature.

5. From the load profiles provided (15) by the licensee, it appean that the load shedding will occur within the first 15 and 40 minutes of the SBO event. He guidance provided in NUMARC 8700 idemifies that loads can be shed commencing 30 minutes into the SBO event unless the loads are automatically shed. According to the plant UFSAR, the plant computer (600 A) is automatically shed from bus 11C at 15 minutes into the SBO event. From the licensce's load profile, the cornputer is the only load shed from the batteries within the fint 30 minutes. Therefore, the timing of the load shedding is consistent with the guidance provided in NUMARC 87-00. However, we did not receive any information on the loads which will be shed. He licensee needs to list the loads that will be shed and state why this load shedding will not adversely affect the ability to safely shut the plant down or maintain the plant in a safe shutdown condition.

6. The licensee used actual equipment loads instead of the rated loads for some equipinent. This approach is reasonable if the assumed loads are the maximum values taken from several tests, in addition, for the constant-power loads (i.e.,

uninterruptab!c power supplies) which are voltage-dependent, the licensee needs to consider the effect of a lower battery terminal voltage (i.e.,105 V) and the change in efficiency due to the reduced load in the actual current requirement for these loads. De licensee cannot use a one time test to justify the use of the actual loads in its calculation.

s 13

t l

3. Compressed Air i Ucensee's Submittal The licensee stated that air-operated valves relied upon to cope with a station ,

blackout for four hours can either be operated manually or have sufficient back up sources independent of the preferred and class-1E AC power supply. The licensee also stated that valves requiring manual operation or valves that require back up sources for operation are identified in plant procedure ECA 0.0 (l.oss of all AC Power).

~

Review of Ucensee's Submittal .

Upon review of the decay-heat removal systems (turbine-driven AFW system and atmospheric heat release system) we found that the steam generator atmospheric steam dump valves (ASDVs) would require compressed air for their operation.

Should cooldown following ECA 0.0 be required, the ASDVs will need to be .

operated. According to the plant UFSAR (Section 9.3.1.6.1), these valves are equipped with back-up nitrogen supplies which are capable of providing 10 complete operation cycles per valve. Therefore Seabrook has sufficient compressed air to cope with a four-hour SBO evem.

4. Effects of less of Ventilatloa Ucensee's Sebenhtal L

l

~

De licensee stated that the calculated steady-state ambient-air temperature for the steam-driven emergency feed water (EFW) pump room during an SBO-induced loss ,

of ventilation would be 128'F. The licensee also stated that the control-room l 14 1

I l

temperature will not exceed 120"F, and is therefore not a dominant area of concern (DAC).

The licensee stated (12) that reasonable assurance of the operability of SBO response equipment in the EFW pump room has been assessed using Appendix F to NUMARC 87 00. The licensee added that. no modification or procedure change is required to provide reasonable assurance for equipment operability.

In response to questions concerning its heat-up calculations, the licensee prtwided (15) a summary of its calculations. For the most part, the licensee used the NUMARC methodology. For some areas, the licensee stated (15) that it modified the method to account for external thermalinfluences. In the control room and the electrical tunnels, the licensee used existing plant-specific steady-state calculations to evaluate the area. The following table is a compilation of the information given by the licensee:

b.12 Temocrature m Methodolorv lailial flaAl f.Q EFW Pump House 104 128 165 5thMRC Switchgear Room A 104 114 130 NUMARC Switchgear Room B 104 112 130 NVMARC Containment:

Annular Compartment 120 1&& 255 MAAP 3.08

. Upper Companment 120 188 N/A MAAP 3.08 Lower Compartment 120 204 230 MAAP 3.0B Cavity 120 227 300 MAAP 3.0B MS/FW Pipe Chase (East & West) 130 206 225 steady state heat balance MS/FW Pipe Chase Electncal Room 118.5 132 13 0 modified NUMARC MS/FW Pipe Chase Stairwell (West) L% 133 13 9 NUMARC Merbmeal Penetration Area (MPA 1) 116 141 250 modified NUMARC Mechamcal Penetration Area (MPA 2) 111 142 250 modified ATMARC Mechanical Penetration Area (MPA.3) 118 135 250 modified NUMARC Mechanical Penetration Area (MPA 4) 115 136 250 modified NUMARC Mechanical Penetration Area (MPA 5) 105 123 250 modifmi NUhMRC Electrical Tunnels A & B t 111.3 130 steady state beat balance Control Room 75 <120 130 steady state heat balance N/A Not applicable. No SBO equipment is located in the upper compartment of the containment.

t No initial temperature prcmded.

15

In addition to the areas listed above, the licensee provided a statement concerning l 1

the battery room. The licensee assumed that the final temperature in the battery rooms was the same as the temperature of the surrounding space (the switchgear rooms) since the battery rooms have no significant heat :oads.

I The licensee stated that the heat loads used for the containment areas (annular compartment, upper contammem, lower containment, and cavity) are based on plant shutdown fro n full power. De major contributor to the containmem heat up are the reactor coolant system, the main steam system, and the assumed primary system leakage into the containmem (366.5 gpm).

The licensee stated (15) that the temperature in the MSA

• Pipe Chase Electrical Room pertains to the main steam isolation valve (MSIV) cabinets. MSIV closure will be perfonned in accordance with either Step 2 or Step 10 of SBO procedure ECA 0.0. In either case, MSIV closure will occur prior to the start of load shedding, which has been determined to begin within 30 minutes of the onset of an SBO event.

The licensee added that it would be reasonable to conclude that the MSIV closure would necessarily occur within the fust 30 minutes following the reactor trip. Once established, main steam isolation would be maintained for the duration of the SBO event. The four-hour temperature in the area is 13rF and the emironmental qualification (EO) temperature is 1307. It is expected that the temperature at 30 minutes into the event would be less than 1307. The licensee concluded that it is therefore reasonable to expect the MSIV cabinets to be capable of performing the intended function during an SBO event.

Redew of Ucensee's Sabadtt11 In response to questiorts, the licensee was asked to provide a summary of its heat up calculations. For each of the areas listed in the above table, the licensee provided the nmnwi initial temperature (except for the electrical tunnels), the room surface 16 N

^

area, the assumed heat load, the method used to determine the room temperature, final calculated temperature, and the EO temperature. The licensee's modified NUMARC method c,nsisted of taking the weighted average of the wall temperatures as the ini. room temperature. The change in temperature (.iT) was calculated using the NUMARC method.

Based on the information prmided, we concur with the licensee's conclusion for the EFW pump house, the pipe chases, and the mechanical penetration areas. With regard to the remaining rooms, we have the following comments:

Convol Room and Switchcear Room From the information provided (15) by the licensee, the heat loads assumed for the control room and switchgear rooms appear to be low. Most of the loads in these areas are due to equipment and instrumentation powered by the batteries. Since the battery loads, for the most part, are resistive loads, we estimated that all of the energy provided by the battery is lost as heat either in the control room or the switchgear rooms. The total heat load used by the licensee for the control room and switchgear rooms A and B is ~33 kW. The total DC loads are estimated to be ~62 kW, based upon the battery loads provided (15) by the licensee and an average battery voltage of 110 VDC. Since the heat loads directly affect the calculated temperature, the licensee needs to verify that its heat loads accurately reflect the loads expected during an SBO event. In addition, the licensee assumed an initial temperature of 75'F, which is non<omervative. If the licensee wishes to use a 75'F initial temperature, then it must place an administrative control which ensures that the control-room tempemture will not exceed the assumed temperature under any Circumstances.

Containment In its heat-up calculation for the containment, the licensee assumed a leak rate of 366.5 gpm. This leak rate would result in the entire primary system inventory 17

leaking to containment in four hours. T11e licensee's leak rate is considerably higher than the leak rate of 110 rpm (25 gpm per RCP and an estimated technical specifications leak rate of 10 gpm) postulated by NUMARC. Based upon the licensee's assumed leak rate, we concur with its conclusions for the cavity and the l annular, upper, and lower compartments. l MS/FW Pfoe Chase Electrical Room

^

The licensee's calculated final temperature (132'F) exceeds the EO temperature for this area (130*F). The licensec stated that the temperature pertains to the MSIV cabinets. The NUMARC methodology is for calculating the bulk room air temperature, and the temperature inside the cabinets would be ~15'F higher. The licensee needs to verify that the MSIVs will close before the temperature inside the MSIV cabinets exceeds the operability temperature. If the operability temperature for the MSIVs is exceeded prior to the closure of the valves, then the licensee needs to assess the consequences of the failure of the MSIVs to perform their function and to find a remedy for the situation.

Electrical Tunnels A&B Using the NUMARC methodology and the licensee's values for the room areas and heat loads, we calculated that the temperature increase for electrical tunnels A and B would be 4'F and 6*F, respectively. If an initial temperature of 1WF were assumed for these areas, the final temperature for both areas would be at least 20*F below their EQ temperatures of 130"F. Therefore we concur with the licensee's conclusion for these areas.

4 18

1

5. Contalament Isolation Licensee's Subarittal ne licensee stated (12) that the list of Seabmok containment isolation valves (CIW) has been reviewed to verify that valves which must be capable of being closed or that must be operated (cycled) under SBO conditions can be positioned with indication independent of the class-lE power supplies. The licensee added that no modifications or associated procedure changes were determined to be required to ensure that appropriate contamment integrity cr.n be prmided under SBO conditions at Seabrook.

In response to questions, the licensee provided a list of the CIVs which could not be excluded using the five criteria given in RG 1.155. The licensee used UFSAR Table 6.2-83 as the source for initialidentification of CIVs. The licensee noted (15) that this table contains several valves which are not considered essential for maintaining containment integrity during design-basis-accident conditions. The scope of valves considered essential for maintaining containment integrity is encompassed by:

1. Valves that automatically close on Phase A or B containment isolation signal and,

2. Valves that are included in the containment integrity monthly and cold shutdown surveillance procedure (OX 1456.76), which lists valves that are not automaticallyclosed on either a Phase A or Phase B isolation signal but are considered essential for maintaining containment integrity.

The licensee stated that specific consideration was given to the containment sump isolation valves. These valves would be in the closed position at all times except 19

during surveillance testing or in the evem of an accident, such as LOCA. a main steam line break, or a feedwater line break inside containment. In the event of such an accident, the containment sump valves would be opened to allow sump recirculation when containment isolation would be most likely initiated. The

!icensee added that, according to the American National Standard for Containment Isolation, these valves are technically not CIVs. Based upon the above, the licensee stated that it does not consider the containment sump valves as CIVs.

De licensee concluded that no valves that have been identified as CIVs of concern for station blackout are requaed to be operable during an SBO evem. Once the valves are closed or verified closed, they will remain in that position for the duration of the event.

Review of Licensee's Submittal ne licensee provided (15) a list of the CIVs which cannot be excluded by the five criteria given in RG 1.155. The licemee prmided justification for the exclusion of these valves. Upon review of the list of containment isolation valves (UFSAR Table 6.2 83), we concur with the licensee's conclusion with the exception of the containment sump valves. De licensee stated that these valves would be closed under all circumstances with the exception of surveillance testing and accident conditions. In order to be able to exclude this valve, the licensee needs to verify that the containment sump valves are closed before entering Mode 3, remain closed dt < , normal plant operations, and the surveillance testing of these valves is performed during cold shutdown or during a refueling outage.

20

6. Resctor Coolant Inventory Ucensee's Sutnnktal The licensee stated (12) that the ability to maintain adequate RCS inventory to ensure that the core is adequately cooled for four hours has been assessed. He licensee used tbc generic analyses listed in Section 2.5.2 of NUMARC 37-00 and stated that these analyses are applicable to the specific design of Seabrook Station Unit L 'l e expected rates of reactor coolant inventory loss under SBO conditions do not result in core uncovery in a four-hour SBO event. The licensee concluded that make-up systems under SBO conditions are not required to maintain core cooling under natural circulation (including reaux boiling).

Review of Ucensee's Submittal The licensee's use of a generic analysis without specific justiEcations for its applicability to the plant is not acceptable. We performed an independent evaluation of the RCS inventory using the available information in the plant UFSAR. Using a postulated leak rate of 110 gpm (25 gpm per pump per NUMARC 87-00 guideline and an estimated technical specifications maximum allowable leakage of 10 gpm), the total leakage from the RCS during the 4-hour SBO event is 26,400 gallons or -3500 ft3 . Upon review of the UFSAR (Table 5.1-1), we found that the total RCS volume to be 11.524 ft 3, leaving an RCS volume of

~8000 ft3 without any cooldowrt. If the primary system is cooled down following ECA 0.0, the RCS volume will be -5000 tr 3at the end of the SBO event, which is sufficient to keep the core covered. Therefore we concur with the licensee that suf5cient RCS inventory exists to keep the core covered, and natural circulation, through reflux boiling, will keep the core cooled.

21

NOTE:

The 25-com RCP sealleak rate was agreed to between NUMARC and the NRC staff pending resolution of Generic issue (GI) 23. If the final resolution of GI-23 defines higher RCP seal leak rates than assumed for the RCS inventory evaluation. the licensee needs to be aware of the potential impact of this resolution on its analyses and actions addressing conformance to the 5B0 rule.

3J Proposed Procedure and Training UcemmWs Sehedstal The licensee stated that the following procedures have been reviewed and modified to meet NUMARC 87-00 guidelines:

1. Station response,

2. AC power restoration, and

3. Severe weather guidelines.

Review of Licensee's Substittal We neither received not reviewed the affected procedures. We consider these procedures to be plant specific actions concerning the required actisities to cope with an SBO. It is the licensee's responsibility to revise and implement these procedures. as needed. to mitigate an SBO event and to assure that these procedures are complete and correct, and that the associated training needs are carried out accordingly..

l l .. _.

1

J.4 Proposed Medincation Uceasee's Subadstal The licensee stated that no modifications are required to are required to attain a 4-hour coping duration.

Review of Licensee's Sabanttaf We did not Snd the need for any modifications in order for Seabrook to be able to cope with an SBO event for four houn. However, our review has identiSed several concerns which may require modifications for their resolution.

3.5 Quality Assurance and Technical Specifications The licensee did not provide any information on how the plant complies with the requirement of RG 1.155, Appendices A and B.

i .

4.0 CONCLUSION

S i Based on our review of the licensee's submittals and the information available in the  !

UFSAR for Seabrook Station Unit 1. we find that the submittal conforms with the requirements of the SBO rule and the guidance of RG 1.155 with the following exceptions:

1. Offstee Power Design Characteristne/ Coping Duratnes The licensee used regional weather data to determine an ESW grouping of "3" and.

based upon a single right-of-way, an SW grouping of ~3." The licensee's estimate of the site ESW grouping is inconsistent with that obtained from both the NUMARC weather data and the data provided in the plant UFSAR: both th~ e NUMARC and the UFSAR data place the site in ESW group "4." With an ESW grouping of "4" and an SW grouping of "3," the offsite power design characteristic is "P3*." In order to be a four-hour coping plant the licensee needs to do one of two things (also see Section 3.1):

1. The licensee needs to implement pre-burricane shutdown procedures.

2. For the plant to be classified as "P2," se licensee needs to ensure that both transmission line rights-of-way will have an offsite power supply available to the plant.

2. Class-lE Battery Capacity Based upon the information available in the plant UFSAR and that provided by the licetisee, we have the following concerns:

1. In the load profile provided by the licensee (15), the loads for the one battery /two busses arrangement is not a direct sum of the loads for the two one 24

~

battery /one bus armrrgements. The licensee needs to provide justification for the discrepancy between the loads in the one battery /two busses arrangement and sum of the two one battery /one bus arrangements.

2. If we use the corrected battery loads (i.e the sum of the two indisidual bus loads) in conjunction with the battery performance characteristics for NCX 2250 (the Seabrook batteries), we find that the 30-day technical specification for plant operation with the two batteries crosstied will be in jeopardy (i.e., cannot be justified).

3. Contrary to the guidance of IEEE Std-485 which recommends a design margin of 1.10-1.15, the licensee used a design factor of 1.0 in its calculation.

4 De licensee used a temperature factor which corresponds to the minimum expected battery-room temperature. The licensee needs to verify that the minimum temperature used is that of the electrolyte and ensure that under no circumstances will the electrolyte temperature drop below the assumed temperature.

5. We did not receive any information on the loads which will be shed. The licensee needs to list the loads that will be shed and state why this load shedding will not adversely affect the ability to safely shut the plant down or maintain the plant in a safe shutdown condition.

6. He licensee used actual equipment loads instead of the rated loads for some equipment. ,This approach is reasonable if the assumed loads are the maximurn values taken from several tests. In addition, for the constant-power loads (i.e.,

uninterruptable power supplies) which are voltage-dependent, the licensee needs to consider the effect of a lower battery terminni voltage (i.e,105 V) and the change in efficiency due to the reduced load in the actual current requirement 25

for these loads. The licensee cannot use a one-time test to justify the use af actual loads in its calculation.

3. Imss of Ventilation Control Room and Switchzear R_oom From the iniw m. tion provided (15) by the licensee, the heat loads assumed for the control room and switchgear rooms appear to be low. Most of the loads in these areas are due to equipment and instrumentation powered by the batteries. Since the battery loads mostly power resistive loads, for comervatism, we esumated that all of the enery provided by the battery is lost as heat either in the control room or the switchgear rocms. The total heat load used for the control room and switchgear rooms A and B is ~33 kW whereas the total DC loads are estimated to be ~62 kW. ,

The licensee needs to verify that its heat loads accurately reflect the loads expected during an SBO event. In addition, the licensee assumed an initial temperature of 757, which is non conservative. However, if the licensee wishes to use a 757 irritial temperature, ien it must place an administrative control which errsures that the control-room temperature will not exceed the assumed temperature under any circumstances. -

MS/N Pioe Cha<c Electrical Room The licensee's calculated final temperature (1327) exceeds the EO temperature for this area (1307). The licensee stated that the temperature pertains to the MSIV cabinets. The licensee needs to verify that the MSIVs will' close before the temperature inside the MSIV cabinets exceeds the operability temperature. If the operabdity temperature for the MSIVs is exceeded prior to the closure of the valves, then the licensee needs to assess the consequences of the failure of the MSIVs to perform their function and to find a remedy for the situation.

26 H '

f..

A Costainment isolation The licensee needs to verify that the containment sump valves are closed before entering Mode 3, remain closed during normal plant operations, and the surveillance testing of these valves is performed during cold shutdown or during a refueling outage.

5. Proposed Modificaticas We did not 6nd the need for any modi 6 cations in order for Seabrook to be able to cope with an SBO event for four hours. However, our review has identified several concerns which may require modifications for their resolution.

6. Quality Assurance and Technical Specincations

, The licensee's submittal does not document the conformance of the plant's SBO equipment with the guidance of RG 1.155, Appendices A and B.

W

l s.o RertRzNces l l

1. The Office of Federal Register, " Code of Federal Regulations Title 10 Part 50.63," 10 CFR 50.63, January 1,1989.

2. U.S. Nuclear Regulatory Commission Evaluation of Station Blackout Accidents at Nuclear Power Plants - Technical Findmgs Related to Unresolved Safety Issue A-44,"

NUREG-1032. Baranowsky, P. W., June 1988.

3. U.S. Nuclear Regulatory Commission, " Collection and Evaluation of Complete and Partial Losses of Offsite Power at Nuclear Power Plants," NTREG/CR-3992, February 1985.

4 U.S. Nuclear Regulatory Commission, " Reliability of Emergency AC Power System at Nuclear Power Plants," NUREG/CR-2989, July 1983.

5. U.S. Nuclear Regulatory Commission, "Emeq;ency Diesel Generator Operating Experience, 1981-1983," NUREG/CR-4347, December 1985.

6. U.S. Nuclear Regulatory Commission, " Station Blackout Accident Analyses (Part of NRC Task Action Plan A-44)," NUREG/CR-3226, May 1983.

7. U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research.

" Regulatory Guide 1.155 Station Blackout," August 1988.

8. Nuclear Management and Resources Council Inc.," Guidelines and Technical Bases for NUMARC initiatives Addressing Station Blackout at Light Water Reactors," NTMARC 87-00, November 1987.

28

_-y. in w

.4 9 Thadani, A.C., Letter .to W,- Hi Rasin of NUMARC, " Approval of- NUMARC Documents on Station Blackout (TAC-40577)," dated October 7,1988.

10. Thadani, A.C., letter to A. Marion of NUMARC," Publicly-Noticed Meeting December

- 27. - 1989," ' dated January 3, 1990 (confirming "NUMARC 874X) Supplemental Questions /Aruwers," December 27,1989).

11. Nuclear Safety Analysis Center, 'The Reliability of Emergency Diesel Generators at U.S. Nuclear Power Plants," NSAC-108. Wyckoff - H., September 1966.

12. Thomas, G.S., letter to U. S. Nuclear Regulatory Commission Document Control Desk, "Information Submittal Required by 10 CFR 50.63," dated Apyd 17,1989.

13. Feigenbaum, T.C., letter to U. S. Regulatory Commission Document Control Desk,

" Supplemental Information Submittal on Station Blackout Rule," dated March 30,1990.

14. Seabrook Station Updated Final Safety Analysis Report, Revision 0, dated May 26,1991.

15. Feigenbaum T.C., letter to U. S.~ Nuclear Regulatory Commission Document Control f Desk, " Response to Request for Informatiort on Station Blackout Rule," dated September 6,1991.

W 29 l

- W + ,

-i.p-u. ,+igpy,--- ep -w.

e e - e l--w'-. -

us---r-r- + -- w