ML20115H763

Forwards Response to NRC Concern,Delineated in Re Design of Tornado Barriers & Plant Doors.Results of Review Contained in Encl Confirmed That Plant Design Capable of Withstanding DBT as Defined in Updated FSAR

ML20115H763
Person / Time
Site:	Seabrook
Issue date:	10/23/1992
From:	Feigenbaum T NORTH ATLANTIC ENERGY SERVICE CORP. (NAESCO)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NYN-92146, NUDOCS 9210270320
Download: ML20115H763 (22)

Text

__ _ _ _ _ - - _ _ _

g

't .

+

P.0, Box 300

@f Seabrook NH 05874-

=- Telephone (603)474 9521 Energy. Service Corporation

( Facsimile (603)474 2987 Tod C. Felgenbaum Senior Vice President and Chief Nuclear Officer NYN-92146 Octob:r 23, 1992 United States Nuclear Regulatory Commission Washington, D.C. 20555 Attention: Document Control Desk l

References:

(a) Facility Operating License No. NPF-86, Docket No. 50-443 (b) USNRC 1 e:ter dated August 24, 1992, Mr. C.W. licht t a M r. T.C.

Feigenbaum (c) Telephoae Conference between Mr, E.M. Kelly, USNRC and Mr. T.L.

Harpster, North Atlantic, September 3,1992

Subject:

Tornado Design of Plant Doors Gentlemen:

North Atlantic Energy Service Corporation (North Atlantic) was notified by the NRC of a concern reg.. ding the design of tornado barriers at Seabrook Station [ Reference (b)].

The NRC requested that North Atlantic review this concern and report the 'results of the North Atlantic review.and disposition of this matter within 30 days of receipt 'of the letter.

North Atlantic requested and was granted an extension on September 3,1992 [ Reference (c)].

North Atlantic has completed a thorough evaluation of this matter and has initiated the appropriate corrective measures. The results of this evaluation confirmed that the plant ,

design was capable of withstanding a postulated worst case site specific tornado, liowever,  !

North Atlantic did nnt initially report the fact that six tornado doors were not-designed to withstand 'he desqn basis tornado as defined in the- Updated Final Safety Analysis Report (UFSAR) nor did it update the UFSAR with site specific tornado data. Details concerning this review are provided in Enclo;ure (1). This seview includes an evaluation that defines Scabrook site specific tornado data using the methodology presented in-NUREG/CR-3058.

Details concerning the development of the site specific tornado analysis are provided in ' i Enclosure (2).

During the most recent review, this condition was identified as being reportabits pursuant to 10 CFR 50.72(b)(ii)(B) and 10 CFR 50.73(a)(2)(ii)(B) on August 27, 1992. It -

was reported to the NRC via a verbal one hour notification on August 27,1992 and via a Licensee Event Report -(LER 92 013 00) on September 25, 1992.

270030 1

9210270320 921023 .a member of the Northeast Utilities system y PDR P

-ADOCK'05000443 PDR g{' j

United States Nuclear Regulatory Commission October 23, 1992 Attention: Document Control Desk Page two North Atlantic has evaluated the reasons why the doors were not included in the original design specification, why this condition was not reported when it was fi~ cst identified, and why the UFSAR was not updated to reflect the site specific tornado data. The results of this evaluation and corrective actions taken are discussed in Enclosure (3).

If , ,u have any questions concerning this response please contact hir. Neal A.

Pil'sbury, Director of Quality Programs, at (603) 474-9521, extension 3341.

Very truly yours,

" ~

ffc./6 2 Ted C. Feigenbaum T CF: MJM / ac t Enclosuie cc: hir. Thomas h1artin Regional Administrator U.S. Nuclear Regulatory Commission Region I 475 Allendale Road King of Prussia, PA 19406 hit. Gordon E. Edison, Sr. Project Nianager Project Directorate I-3 -

Division of Reactor Projects U.S. Nuclear Regulatory Commission Washington, DC 20555 hit. Noel Dudley NRC Senior Resident inspector P.O. Ilox 1149 Seabrook, Nil 03874

Norih ! Atlantic--

October 23, 1992 ENCLOSURE I TO NYN.92146 RESULTS OF Tile NORTil ATLANTIC REVIEW 3.

4 i

W G

II AC KG R OU N D r

North' Atlantic originally identified a concern with the design of tornado barriers al Seabrook Station in late 1990 while developing a consolidated design basis-document for plant barriers and while consolidating associated plant drawings. Plant bairiers covered by this design basis -

document included tho=c which blocked or limited the passage of air, water,: smoke, fire, radiation, or combinati E thereof. During this proce:,s seven safety related doors.were

~

identified .as tornado bou idary doors that had not been included in the United Ent,iacers and Constructors -(UE

• C) Doc? Specification as doors which are required to meet tornado design criteria. The seun doors are identified as follows:

- P-604; Primary Auxiliary Building Roof

- P-900; Residual llent Removal Vault Roof q- - EF-400 and EF 404; Emergency Feedwater Pumphouse

- EM-?r, & EM 210; Main Steam and Feedwater Pipe Chase ,_

F-104; Fue'. Storage Building Based on this information North Atlantic initiated a comprehensive review of tornado loading on Seabrook Station structures and components. This two phase review included walkdowns of plant structures and systems, and review of plant drawings and industry practice. The first phase was an evaluation of the site specific tornado characteristics. The second phase was a detailed evaluation of the ability of affected components to .. withstand the-depressurization effects of the design basis and/or site specific tornado. This preliminary review was completed in June 1991 and verified that the structures and components could ~

withstand either the design basis or site specific tornado, North Atlantic also initiated a design document to upgrade the locksets of five doors so that they would conform to the design basis tornado criteria in the UFSAR. The intent was to implement this modification during the fiist refueling outage.

At the time of the first refueling outage North Atlantic personnel were confident, based on the preliminary review, t'.iat the plant could withstand the worst case site specific tornado.

T'aus, implementation of the design change and UFS AR update (with the site specific tornado criteria) were not given a high priority. The UFSAR is schedul,:d to be updated by October _

31,1992. While the design change is not required to withstand the site specific tornado, the design change is scheduled to be implemented as a plant enhancement for two doors by May 31, 1993.

A confirmatory tornado design study was initiated in April 1992 to consider. tornado load protective design --features. This study- demonstrated that safety related structures and-

. components were capable of withstanding the worst case tornado postulated by site specific tornado data.

North Atlantic reported this condition per 10 CFR 50.72(b)(ii)(B) on August 27,1992 and per 10.CFR 50.73(a)(2)(ii)(B) on September 25, 1992 via LER 92-013-00. North Atlantic initially (1991) failed to recognize the reportability related to the discrepancy between -the actual-design of the discrepant plant doors and the design basis information in the Updated Final Safety Analysis Report. Timely action was taken to determine that the doors would in fact withstand the-loading of the postulated site-specific tornado and that the components in a vented area would not be adversely impacted by the site specific low atmospheric pressur:

transient.

1

4 DESIGN CRITERI A -

10 CFR 50 Appendix A, General Design Luteria 2, " Design Bases for protection Against Natural Phenomena *, requires, in part, that structures, systems, and components be designed

- to withstand the cIIects of natural phenomenn such as tornados without losing the capability to perform their safety function. The tornado design criteria used in the plant design was that specified in Regulatory Guide 1.76,

• Design llasis Tornado for Nuclear Power Plants *,

Regulatory Guide 1.117, " Tornado Design Classification", and the guidance in utandard Review Plan Sections 3.3.2 " Tornado Loadings

• and 3.5.1.4, ' Missiles Gencated by NMural Phenomena".

The general regulatory gaidance for ttructures, systems, and components which require tornado protection is contained in Regulatory Guide 1.117. In this Regulatory Guide the NRC provides un acceptable method for id:ntifying those structures, systems and components that should be protected from the effects of the design basis tornado. The Regulatory Guide states that structures, systems, and components so designated should be designed against building collapse, and provide an adequate missile barrier, if protective barriers are not-installed, the structure and components themselves should be designed t0 withstand the effects of the design basis tornado.

Regulatory Guide 1.76 describes the acceptable design basis tornados for each of three regions within the contiguous United States. These three regions conservatively envelope the expected hazards within each region. ' The Seabrook site is located in Tornado intensity Region I,- which encompasses all of the United States cast of = Colorado, and Lincludes midwestern and southwestern states which are prone to violent tornados. This categorization results in a specified tornado load criteria which is considerably more conservative than-that developed when considering site specific data. The Regulatory Guide 1.76 ' criteria for n design basis tornado in Region 'I are as follows:

A maximum wind speed of 360 miles per hour

- A rotational speed of 290 miles per hour

- A maximum translational speed of 70 miles per hour A minimum translational speed of 5 miles per hour

- A radius of maximum rotational speed of 150 feet

- A. pressure drop of 3.0 pounds per square inch A rate of pressure drop of 2.0 pounds per square inch per second T:.ese criteria are listed in section 2.3.1.2.b.2 of the Seabrook Station Updated Final Safety':

- Analysis Report:(UFSAR) as the criteria that Seismic Category structures are designed to withstand, with the ' cxceptton of the refueling water storage tank, spray additive tank enclosure and the cooling tower.

2

. ~ . . - .-. . , . - . - . . . .. .. , .-

o c ,

v .

LNorthfAtlantic has subsequently' developed a site specific _ tornado loading, in accordance _with :

--applicable NRC criteria, to__ provide a more realistic basis for performing the plant' door _ _

evaluation. This evaluation, which is' described in Enclosure (2), determined that _ the ' site-

. specific tornado has the following characteristics:

-- A maximum wind speed of 260 miles per hout

- A rotational speed of 203 miles per hour-

- .A maximum translational speed of 57 miles per hout

- A radius of maximum rotational speed of 453 feet

- A pressure drop of 1.46 pounds per squarc inch

- A rate of pressure drop of 0.27 pounds per square inch per second The option to utilize site-specific tornado - criteria existed when Seabrook- Station ~ design criteria was under development, but the . Architect-C' er (UE&C) conservatively chose to e 1.76. .Thus, the. existing design' apply the general ~ criteria specified in Regulatory >

criteria and stru1 ural analysis for the exterier e.. m and roofs of seismic _ category it -

structures is unaffected by this evaluation. The design of these structures considers the. full 4 3.0 psid- pressure drop.

Jp' "lAla EVAL,U ATION (January - June 1991)

For the seven doors that were not included in the UE&C door specification, the' controlling tornado load was determined to be the differential pressure drop. The force onlcach . door -

results from a drop in the outside atmospheric pressure as the. tornado passes; iThis drop--

in pressure produces a force inside the unvented structure which acts outward on the door, and is app'M to the door hinges and locksets. This effect produces a force which is directly prop,rtional to the magnitude of the design pressure drop. While, loss of door, integrity-in itself has no direct effect on safety related equipment, a previously.- unvented building could become vented and expose system. and components to a negative pressure -

which would require an evnluation lo ensure functionality.

- As a result of- the above evaluations, . North- Atlantic determined that one of the doors,

~

EM 204, had been designed to withstand the differential pressure of the design basis tornado; as: defined in the UFSAR (3.0 psid). The four single doors, P-604,- EF-404,: EM-210,1 and .

F-104, were shown- to have a differential pressure capacity of 1.5 psid.- It was determined that_ two other doors, EF-400 and P-900, could only withstand a differential pressure of 0.2 --

psid.

All seven doors were verified to not be seguired for protection against ' tornado-missiles.

. Missile entrance probabilit'y was shown to'bc extremely low for doors EF-400 and P 900.

The remaining five doors are not exterior doors and are shielded- by missile resistant concrete walls and slabs.

The effects of the-differential pressure _ loading on interior building areas was evaluated;for tornado loads acting on. doors P 900_ (Primary Auxiliary Building vault entrance) and EF-400 3

r- W4 M e- 9 -q- 4 e. um - *- # y

w- - - ~ _ - , --- . -- ~ - ~ . , . .- - . -,

-w l

. 4

-(Emergency Feedwater Pumphouse) whose capacities are less than the site specific tornado: n load. The identification of-the affected interior building areas is based on- the pressure =

capacities of the interior partitions _(e.g. walls, slabs, doors, metal partition walls, etc.) and existing openings that L allow one interior' area' to communicate with another. lA tornado depressurization evaluation and a tornado differential pressure droo venting evaluation were performed to determine the bases for-defining the affected interior building areas and the corresponding differential pressure -value. These evaluations which verified ;that 'the site specific . tornado' would not adversely affect plant equipment, are available for review at  ?

Seabrook Station.

Safety related components located in the above intcrior building areas were identified by either walkdowns and/or the Class 1E Equipment List. Identified interior structures _ and safety related components are designed to withstand the effects of_ the site specifie tornado differential pressure load. None of these coraponents are affected by tornado missiles or the differential pressure.

Furth Atlantic decided to modify the locksets on the remaining five doors such that the d sors could withstand the 3.0 psid differential pressure. A design change document was prepared and the work was scheduled to be performed during the first refueling outage.

In summary, as of June 1991, North Atlantic had established that tornado barriers were designed to withstand the design basis tornado defined in the UFSAR with the exception of six doors. All of the doors could meet the design basis tornado criteria with the exception of the 3.0 psid differential pressure. Of these, five of the doors could meet the design basis tornado differential pressure by replacing the existing locksels with : locksets of- higher.

capacity. The remaining door was evaluated to ensure that it woul.d withstand- the site specific tornado. Based upon the preliminary evaluation North Atlantic was confident at that time that the _ plant was capable of withstanding the worst case site specific tornado - Thus, implementation of the design change and the UFSAR update (with the site specific tornado criteria) were not giveu a high priority. However, North Atlantic failcd to identify the need to report this condition and to update the UFSAR to include the site specific tornado design criteria. The UFSAR is presently scheduled to be updated by October 31,' 1992. The design -

change to upgrade the locksets of two doors which are not designed m withstand the site-specific tornado is scheduled to be implemented as a plant enhancement by May 31, 1993.

1992 TORNADO DESIGN STUDY (April October 1092)

The issue of tornado barriers was reinitiated in- April 1992 in conjunction with the Individual Flant Evaluation for- External Events. This evaluation included other Seabrook' Station - .

, ' tornado load protective. design features in ' addition to tornado' doors. The objective of this evaluation was to reconfirm compliance with Regulatory Guide 1.117.in accordance with' UFSAR commitments. The evaluation of tornado birriers was completed in October 1992, The .UFSAR update will be completed by October 31,-1992.

An extensive re review of tornado protecteu structures was performed to ensure that the systems identified in Regulatory Guide 1.117 are protected from tornado loading. Portions of systems not ' completely tornado protected were verified to be capable of withstanding tornado loads or it was verified that alternate recovery methods are in place As previously stated, tornado loading includes the effects from-wind pressure loads, differential pressurc' load, and- tornado missiles.

4-i l, , , , . , . , , - . .

m _. . . . _

q.

R E. 4 Tornado prote'eted structures are cator,orized as either closed (barrier envelope openi_ngs are -

protected by tornado dampers), or opec (barrier openings exist orL barrier _ is not designed

for all tornado loads). For closed structures, interior structures and- components-_are noto-subjected to any differential pressure loading - The following is a listing of closed 'and open'-

structures.

Closed Structures: Electrical Tunnels Service Water-Pumphouse (Switchgear) -

Containment Building Primary Auxiliary Building (including the Service Water pipe slot)

Control Building Containment Enclosure Bui Sng Open Structures: Diesel Generator Building Cooling Tower Service Water Pumphouse (pump bays)

Fuel Storage Building Main Steam /Feedwater East and West Pipe Chases Personnel Hatch Ares Equipment Hatch Area Emergency Feedwater Pumphouse*

Equipment- Vaults *

• These structures are currently open structuies since two doors associated with. them -

(EF 400 and P-900) are not capable of withstanding the site specific tornado; differential' pressure. Safety related systems in these structures have been verified to be capable .of-withstanding tornado loads. These structures will be considered closed upon implementation of.a design change to replace the existing locksets with higher capacity ones.

Two other structures which require tornado missile protection only are the Condensate Storage Tank and the Intake and Discharge Transition Structure (Service Water ' valves only).

This review confirmed that safety related components located in closed tornado protected structures are completely shielded from tornado' loads, and that safety.related ; components in open structures can withstand the depressurization,cffects of a' Regulatory Guide L76 or:

site specific tornado as appropriate. i The confirmatory study consisted of a comprehensive review and evaluation of the following-attributes:

The tornado barrier envelope for tornado protected structures The tornado barrier envelope designs for tornado' loads

- _ Thci design of tornado barrie, doors

- Exterior penetration seals, exterior ductwork. design, and the shielding _ of exterior openings

.- ;ComponentsLin opa structures for differential pressure loads-m 5

3 ,im a , - - . . Jl _ , .....L-- - .i ~ ~.

. 1. ,Tpfnado liarrier Unvelone for' Tornado Protected Structures

'The envelope review was accomplished by a thorough inspection of the existing barrier

~

drawings with- subsequent confirmation by plant walkdowns. The walls,- floors and '

ceilings that comprised the ' barrier envelope were identified a' nd assigned an:

identification number on the barrier drawings. Some of these barriers are interior -

structures separating tornado protected areas from non protected areas. The-identified envelope barriers for both closed and open areas were inspected during plant walkdowns. hlarked up barrier drawings which show the identified envelope barricts for each structure are evallable for review at.Seabrook- Station.

The reason some of the seven doors were not oris/nally designed for differential pressure loading could be partially attributed to the fact that these doors are located on walls that are not outside perimeter walls. These walls are part of the tornado pressure envelope for a particular structure that is protected from differential pressure _

loads-but they are not outside walls, These walls are adjacent to structures that do -

not require protection from tornado ' differential pressure loads (eg. Emergen :y Feedwater Pumphouse adjacent to the East Pipe Chase). Therefore a plant wide a review of the tornado barrier envelope for terrado protected . structures was performed.

2. Tornado Envelone Barrier De<iens Walls, floors, and ceilings which comprise the tornado envelope for each tornado protected structure were reviewed to ensure they were properly designed for the tornado loads. This review included interior structures which form the tornado envelope, in addition, this review confirmed that the design of the structures, that comprise the tornado barrier envelope, did not ut.'ize venting to reduce the differential pressure load on the barrier envelope walls and slabs.

The evaluations performed for the review of the tornado envelope barrier design are documented in North Atlantic Engineering Evaluatian No. 92-29, " Tornado Barrier Evalut. tion", which is available for review at Scabrcok Station. .

4

3. Tornado Harrier Door Desiens The doors which are installed in the tornado envelope for each tornado protected structure were evaluated by reviewing design bases specifications. and by performing plant walkdowns. Doors on closed structures were verified ; to be designed to

~

withstand tornado loads. Some doors were exempted from requiring protection against

< tornado missile strikes. The b. mis for exempting these doors is documented in the report "Scabrook Nuclear Power Plant Tornado hiissile Analysis", which was performed by Applied Research Associates, Inc. in December 1984. This report is available for

'eview at Seabrook Station.

4. Exterior Penetration Seals. Ductwork Desien. and Onenines in Tornado Barriers Heating, Ventilation, and Air Conditioning (l!VAC) openings in the envelope barrier ,

for tornado protected structures were evaluated for the appropriate tornado loads in North Atlantic Engineering Evaluation.No. 92 29, " Tornado Barrier Evaluation", which 6

sui

w is available' for review at Seabrook Station! Openings were identified by .HVAC drawing reviews ard plant walkdowns. Openings were found to have missile shields -

unless the probabilistic analysis demoastrated that missile entrance was a very low -

probs.bility as defined in the report "Schbroo_k Nuclear Power Plant Tornado Missile-Analysis" (Applied Research Associates, December 198_4). Tornado dampers were

.. confirmed to be installed in openings for closed structures. Ductwork and-backdraft dampers were evaluated for positive tornado wind-pressure loads. Tornado ' dampers' were considered open for this evaluation. Small penetra '^ns such as piping' located in closed structures were confirmed to be. properly seated Ith ' pressure resistant material. 'rhese penetrations were identified by reviewing the seal list and plant walkdowns.

5. Components in Onen Structures Equipmrnt and interior structures located within the open structures were evaluated for the Ufects of the design basis tornado differential pressure load (3.0 psid),-

Affected interior structures and ductwork were identified _ by drawing reviews and plant walkdowns. Some interior ductwork and partition walls were evalueted for a site specific tornado with a maximum pressure drop of 1.u psid. Safety related equipment located in each open strue'ure was identified by Class 1E Equipment List. No impact on ~ safety related equipmert was ide ntified. . Natth Atlantic Engineering Evaluatioa No. 92-29, " Tornado llatrier Evaluation", which is available for review at Seabrook Station.

In summary, at the completion of the confirmatory study, North Atlantic has confirmed that the Seabrook Station tornado barriers, including the seven doors not originally _-specified as tornado boundary doors, are capable of withstanding the worst case site specific tornado.

Nane of the seven doors not originally specified as tornado boundary doors are required for tornado missile proteaion since other means of protection were provided in the original I design. The structural integrity of the seven doors is maintained under the effects of

• positive pressure resulting from direct conversion of wind yr 'ocity to static _ pressure. Four-of the doors satisfy the loading criteria for the negativt pressure effects of a site specific tornado (1.46 psid), and another door satisfies the criteria for the design basis tornado (3,0 psid). Safety related equipment in areas exposed to depressurization by the atier two doors is not affected and will continue :o perform its safety related functMns.

A UFSAR change is currently being developed to include'the site specific tornado design data as part of the current licensing basis. The UFSAR ~ Change is scheduled to = be_ ,

completed by _ October 31, 1992.

l1 l~  ;

l l -.

I.

7 -l i

,m-- - _ . , , _ _ _ - -

. _ __ . = _ . _ ._ .

North atlantic  ;

Oct ot,e r 23, 1992 l l

1

?

i ENCLOSURE 2 TO NYN.9214f.

TOHNADO DESIGN llASIS 1.OADINGS AT SEAllROOK STATION

'es--g .w~,-

Tornado Design Dasis Loadings At Boabrook Utation

,1. O B a_q&cr roufL4 liRC tornado design criteria is defined in Regulatory Guide 1.76 (1) and Standard Review Plan (SRp) Sections 3.3.2 (2) and 3.5.1.4 ( *., ) . .

These criteria were established to prov annual probability of exceedance of 1x10 . To jde meetdesign loads with an this objective, the IIRC has defined in Regulatory Guide 1.76, three Tornado Intensity Regions which conservatively envelope the expected design loads within the respective regions. Seabrook is in Tornado Intensity Region 1 which has the most severe tornadofollowing loading requirements. Regulatory Guide 1.76 specifies the tornado wind and dif ferential pressure loads for Tornado Intensity Region 1:

- loads resulting from a rotational wind velocity of 290 mph and a translational wind velocity of 70 mph for a caximum horizontal velocity of 360 mph.

- a differential pressure load resulting from an external pressure drop of 3.0 psi at a rate of 2.0 psi /sec.

- a radius of maximum rotational speed of 150 feet.

Basis for R'equlatory Guido 1.76 The basis for Regulatory Guide 1.76 is found in Wash-1300 (4) . The process emploved in Wash-1300 to establish design tornado wind speeds for each Tornado Intensity Region is outlined below:

- Tornado strike probabilities were determined for each five degree square of latitude and longitude in the contiguous United States based on a mean tornado damage path area of 2.82 square miles per tornado established by Thom (5) for tornadoes occurring in a ten-year period in Iowa. The mean number of tornadoes occurring within each five-degree square was determined from a data base compiled by paut: (6) for the years 1955 to 1967 (13 years of tornado data).

- Based on data developed by tornado climatologists (7,8) for tornadoes which occurred within the contiguous United States during 1971 and 1972, a tornado wind speed /F-scale intensity relationship was established independent of geographic location.

- For each five-degree square of latitude and longitude, a tornado wind speca was determined consistent. With an annual

probability of 1x10'I. Based on these wind speeds, design tornado wind speeds were established for each of the NRC's Tornado -

Intensity nogions.

Conservatisms in Wash-1300 The Wash-1300 tornado criteria is recogni:cd as being conservative, and as so stated in Wash-1300, it was considered at the time of its development as an interim criteria until morc realistic criteria could be developed. Major sources of conservatism of the Wash-1300 criteria as they relate to the Seabrook area are as follows:

- The nuan path area for tornadoes occurring within the 5 degree by 5 degree box contuining SeabrookAs is considerably noted above, this less .

than the 2.82 miles assumed in Wash-1300.

path area was based on tornadoes occurring in the mid-west for the (Iowa).

region Current estimates (9) of the mean path area surrounding Seabrook are 0.25 square miles for the 5 by 5 degree box surrounding the site and 0.19 square miles for the state of New Hampshire. These values are more than a factor of 10 lower than what was used in the Wash-1309 study. Application of the Wash-1300 -

methodology with a more complete dat9 base and a mean path area of 0.25 square miles results in a 1x10~ maximum wind speed of 260 mph (9) for the region containing the Seabrook site.

- The Wash-1300 methodology assumed that the maximum vind speed of a tornado occurs uniformly over the entiro damage area.

However, wind speeds vary along and across the path of a tornado (10). Mcdonald (11) has shown that on average, tornadoes rated have an F3 intensity for about 35% of their track. For the F3 remaining G5% of the track, the intensity is less than F3.

Similarly, tornadoes rated F4 have an F4 intensity for about 24% of the tracks, and F5 tornadoes have an F5 intensity for only about 19% of th2ir tracks.

2.0 current Tornado Hazard Modelg Substantial refinements have been made by Abbey and Fujita (12) and by Mcdonald (11) in the methodology for predicting tornado hazards ,

since the time the Wash-1300 study was performed (1974). Some of these refinements are:

- a larger database of reported tornadoes is available for statistical analyses than was available for the Wash-1300 study.

- estimates of tne number of unreported tornadoes are made and included in the hazard model..

- development of relationships between path length, path

width, and path crea as a functior. of tornado intensity.

- the variation of the tornado wind speed across the tornado path are explicitly accounted for in current models.

EUREG/CR-3058 (11) presents a methodology that incorporates all of the above refinements and also presents a detailed sample calculation. This methodology has been used for more than 30 site-specific studies, and is the basis for the Soabrook site-specific st.dy documented in SBC-44.#. (13).

3.0 Summary of Beabrpok Site Soecific Tornado study (BBC-4411_

SBC-441 follows the methodology presented in ITUREG/CR-3058. The

• data base used in SBC-441 was obtained from the National Severe Storms Forecast Center (NSSFC,14) and consists cf all reported tornadoes between 1950 and 1990 (41 years of data).

The tornado hazard probability analysis is based upon statistical analysis of tornadoes that have occurred in the region surrounding the Seabrook site. Four steps are involved in this analysis. They are:

i

1. Determination of an area-intensity relationship in a global region surrounding the site, l

2. Determination of an occurrenco-intensity relationship in a  !

local region surrounding the site,

3. Calculation of the probability of a point in the local region experiencing wind speeds in various wind speed inte rvals ,

4. Determination of the probability of wind speeds in the local region exceeding the interval value.

The global region (see Figure 1) is defined by the following latitudes and longitudes:

- Latitude 40 degrees to 45 degrees north

- Longitude 69 degrees to 75 degrees west Factors considered in selecting the glebal study region are (12):

- The region should genarally surround the site. 1

- The region should generally contain the same type cf terrain.

- The region should have common meteorological conditions on a synoptic scale, as they relate to.the. formation of tornadoes.

l

- The region should be of suf ficient si:o to givo an adoquato sampic size for determining the area-intensity relationship, i

Area-Intensity Relationship The area-intensity relationship is a continuous function of mean tornado damage path area versus wind speed. Within the global study region the NSSFC data base was sorted to determino all tornadoes that occurred between 1971 and 1990 and were completely defined in terms of F-scale, path length, and path width. The mean tornado damage path for each F-scale classification was then determined. A functional relationship between mean tornado damage path (area) and wind speed (intensity) was then developed.

Occurrence-Intensity Relationship The local region within the global region (see occurrouce-intensity Figure 1) is defined to determine a tornado frequency of relationship. For the Seabrook Station sita-specific study, the local region is defined as:

- Latitude 41 degrees to 44 degrees north

- Longitude 70 degrees to 73 degrees west For this local region the NSSFC data bare was sorted to determine all tornadoes that occurred between 1950 and 1990 (41 years) . This longer time fcame is used such that stable estimates of the rate of occurrence of each F-scale tornado can be determined. Tornadoes that occurred within the local region between 1950 and 1970 are not plotted on Figure 1.

For the local region, an occurrence-intensity relationship was

- developed which included an accounting for unreported tornadoes.

The expected number of tornadoes per year per F-scale classification is obtained by dividing the number of tornadoes in each F-scale classification by the number of years of record.

Probability of Exceedance for Tornado Wind Speeds A mean site-specific probability of exceedance curve for tornado wind speeds was determined by combining the global area-intensity relationship with the local occurrence-intensity relationship. The variation of wind speed across the width and along the path of the tornado is also accounted for in this analysis. Also, the area of the local region is adjusted to reficct only land area. The mean value wind speed at a probabili"r of exceedance of 1x10~7is a. bout 260 mph. Other tornado characteristics associated with a 260 mph maximum wind speed tornado can be found from ANSI /ANS-2,3 (15).

Figure 2 presents the results of this analysis relative to published results (15) . As can be seen the agreement is excellent.

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

/e

. I t

(

a u g.b-specific wind and Pressure Loadincs f

!M.ts from site-specific analyses (13) and ANSI /ANS-2. 3 (15) i support the following tornado wind and differential pressure loads for the Seabrook Sitet ,

- loads resulting from a rotational wind velocity of 203 mph-and a translational wind velocity of 57 mph for a maximum  :

horizontal velocity of 260 mph. j

- a differential pressure load resulting from an external pressure drop of 1.46 psi at a rate of 0.27 psi /sec (16). l

- a radius of maximum rotational speed of 453 feet.

These design loads ,,are consistent with an annual _ probability of .

exceedance of 1x10~

}.0 Confirmation of Site-Bogolfic Results i As part of the Systematic Evaluation Program-(SEP) the !TRC funded-their estimates consultant for the to develop site-specifiThe 1x10'9 SEP plants. tornado tornado and wind straight speed forwind.the Yankee Rowe site was estimated as 248 mph (17). The methodology used in (17) is consistent with that used in SBC-441. Also, the -f geographic location of both sites suggest that results for the Yankee Rowe study should-be'similar to the Seabrook site.

As stated earlier, the mean path area for tornadoes occurring within the 5 degree by 5 degree ~ box containing - Seabrook is. '

1 considerably less than the 2.82 miles assumed in' Wash-1300.

current estimates (9) of the mean path area for -_ the region -

murrounding Seabrook are 0.25 square miles for the 5 by 5. degree box surrounding the site and 0.19 square miles for the state of New Hampshire. .These values are more than a factor of 10 lower than whct was used in the Wash-1300 study. Application of the Wash-1300 .

methodology with a more complete dat9 b ase and:a mean path area of O.25 square miles results in a 1x10 maximum vind speed of 260 mph t (9) for the region containing the_Seabrook site.  !

Lastly, ANSI /ANS 2,3 recognizes that meteorological and topographic _

conditio.ts vary significantly within the continental United States,  ;

and that this variation influences- the frequency of occurrence and i intensity of t Figure 3 is from ANSI /ANS 2.3. .As can be '

.seenthe1x10-9rnadoes. wind speed for the Seabrook site would be. estimated:

at 250 mph. Furthermore,.the; tornado wind speeds in-the central United States, where most violent tornadoes occur, are higher than other regions. , Additional documentation.,that most violent

'h--,,me.., -~.,-,.,+~.ee,.w--,.,-

-+,.-ecy-e.,v...--m,,ww-3,,*,rew-*,- e ,m,r -vy-v+ rE , -,, r- t r=wr wm y--.-eee-, _ c-e- y g r- rwer- -,.er,,.=--w-wr- T w-v vy g*

and greater) occur between 80 to 105 degrees

. tornadoes (F4 longitude can be found in (9).

Based upon the above information it is concluded that results of the site-specific tornado study and the site-specific tornado characteristics for Seabrook given in 4.0 satisfy the 1x10'7tornado load requirements of Regulatory Guide 1.76.

.e t

6.0 HqfeIAAtu

' 1. U.S. Nuclear Regulatory Commission Regulatory Guide 1.76,

' Design Basis Tornado for Nuclear Power Plants,' Washington D.C. April 1974.

2. U.S. Nuclear Regulatory Commission Standard Review Plan Section D.C.,

3.3.2, Tornado Loadings,' NUREG-0800, Rev 2, Washington July 1981.

3. U.S. Nuclear Regulatory Commission Standard Review Plan Section

3. 5.1.4, ' Missiles Generated by Natural Phenomena, ' NUREG-0800, Rev 2, Washington D.C., July 1981.

and Sanders, K.E., ' Technical

4. Markee, E.H., Beckerley, J.G.,

Basis for Interim Regional Tornado Criteria,' Wash-1300, U.S.

Atomic Energy Commission, Office of Regulation, Washington, D.C., 1974.

5. Thom, H.C.S. , ' Tornado Probabilities, ' Monthly Weather Review, Vol. 91, pp 730-736, 1963.

6. Pautz, M.E., ed., ' Severe Local Storm occurrences 1955-1967,'

ESSA Technical Memo, WBTM FCST 12, Office of Meteorological Operations, Silver Springs, MD.

7. Fuj ita, T.T. , 'F-Scale Classification of 1971 Tornadoes,' SMRP Research Paper No. 100, 1972.

8. Fujita, T.T., Pearson, A.D., 'Results of FPP Classification of 1971 and 1972 Tornadoes,' Eighth Conference on Severe Local Storms, Denver, CO, 1973',

Ramsdell, J.V., Andrews, G.L., ' Tornado Climatology of the 9.

Contiguous United States,' Pacific Northwest Laboratory Operated by Batelle Memorial Institute, Prepared for U.S.

Nuclear Regulatory Commission, IMREG/CR-1461, 1986,

10. Twisdale, L.A., ' Tornado Data Characterization and Wind Speed Risk, ' Journal of the Structural Division, ASCE, Vol.104, No.

ST10, Proc. Paper 14096, pp. 1611-1630, 1978.

11. Mcdonald, J.R., 'A Methodology for Tornado Hazard Probability Assessment,' Institute for Disaster Research, Texas Tech University, Prepared for the U.S. Nuclear regulatory Commission, NUREG/CR-3058, 1983.

12. Abbey, R.F., Fujita, T.T.,

'The DAPPLE Method for Computing <

Tornado Hazard Probabilities: Refinements and Theoretical Considerations,' Preprints for the Eleventh Conference on Severe Local Storms, American Meteorological Society, Boston, l

MA,- 1979.

13. SBC-441, Tornado Hazard Probability Analysis,' July, 1991.

L t

L L

l L

, i

14. liational Savoro Stormo Forecast Center, liational Weathor  ;

Service, Kansas City, Missouri. ~

15. ' Standard for Estimating Tornado and Extreme Wind i Characteristics at !!uclear Power Sites, ' A11SI/ANS-2.3-1983, American !!uclear Society, La Grange Park, Illinois.

10. SBC-530, ' Tornado Ventir.g-Diesel Generator Building,8 '

September, 1992.

17. Letter, USliRC to Yankee Atomic Electric Company, ' Yankee Rowe -

SEP Topic II-2.A,' December 29, 1980. ,

i 4

6 9-E e

I l'

)

-r North Atlantic f October 23, 1992

-t s

i r

r ENCLOStJRE 3 TO NYN 92146 t ROOT CAtlSE AND CORRECTIVE ACTIONS ,

r

-I

.e y

s l

f b

.,,N-, , ~. , .- - , , . ,,. ,. -wg -.eS ,e- , , -,r - , , , , e

_ _ . _ . - _ _ __ , .- ._. _ _. ._- ._ __ ___ m _ _.

ROOT CAUSE AND CORRECTIVE ACTIONS There are two issues surrounding this condition which must be addressed. The first is the reason the seven doors were not included on the UEAC door specification as tornado barrier doors. The second is why, when it was discovered that the doors were not designed to meet the differential pressure value stated in the UFSAR, the condition was not reported to the j NRC and why the UFSAR data was not updated to reflect the site specific tornado data. ]

i ROOT CAUSE l i

1. Doors not included in the UE&C Snecificatin_p_

The responsibility for plant barricts during and immediately following the construction of the plant was divided among several engineering groups. This included providing tornado barrier information on numerous drawings. It was only when North Atlantic-was consolidating this information into a single design basis document that the seven doors were identified as tornado barriers which had not been included in the UE&C -

door specification. l A contributing factor is the fact that five of the seven doors were not on exterior perimeter walls, making them unilkely tornado barriers. Ilowever, when plant walkdowns were performed and plant barrier drawings consolidated as part of developing the Design Basis Document for Plant Barriers, the f act that these door. -

would be exposed to the differential pressure conditians imposed by a tornado became.

evident. The comprehensive review of plant barriers indicates that this condition is  ;

isolated to these seven doors.

2. Condition not reoorted and UFSAR not updated The North Atlantic personnel that identified the nonconform.ng condition properly

• cvaluated the condition and determined, based on a preliminary review, that the plant was able to withstand a site specific tornado. However, these personnel did not recognize that this condition was reportable and did not recognize the need to update the UFSAR. In the final analysis, thu failure to report the nonconforming condition and to update the UFSAR resulted from a failure to implement the North Atlantic corrective action process, CORRECTIVE ACTIONS Immediate Actions *

1. North Atlantic reported this condition via e one hour verbal notification pursuant to 10 CFR 50.72(b)(ii)(II) on August 27, 1992.

t Short Term Actions 1 North Atlantic followed up the one hour notification with a Licensee Event Report (LER 92-013 00) on September 25, 1992 pursuant to 10 CFR 50,72(a)(2)(ii)(B)< '

1 i

i. 4 ._m..e-,._-m . m ,,. , a ,_ . m , _ , , , , . . . , . _ , , ~ , , ,.s ....._m. , . , _ - , , - - , _ - , . _ _ . . , _ . . , . -*

2. Plant flarriers have been consolidated into a single design basis document, included in.this document is a list of the barrier drawings which clearly identify the barricts which must maintain their integrity against air. water, pressure, weather, fire, or a combination thereof. Itesponsibility for the these barriers has been clearly assigned.

3. North Atlantic completed a comprehensive reevaluation of plant design feet ures relative to tornado design criteria. This reevaluation verified that tornado design criteria are met by the existing plant design with the exception of the six doors.

These doors have been verified to be capable of withstanding the worst case site specific tornado or the affected plant' areas are capable of withstanding the corresponding depressurization.

E

4. North Atlantic confirmed the tornado barrier envelope by a thorough inspection of the existing plant barrier drawings with subsequent confirmation by plant walkdowns.

5. North Atlantic is preparing a change to the UFSAR to include the site specific .

tornado data. This change is scheduled to be included in the UFSAR by October >

31, 1992.

• I one Term Action,

1. In order to standardire design requirements, North Atlantic will implement a plant enhancement design change for the two doors (EF 400 and P-900)- to withstand a _

t differential pressure of at least 1.5 psid. This design change is expected to be impicmented by May 31, 1993.

2. Independent of the above events, North Atlantic has been developing 'a new procedurc~ .

which will enhance the corrective action process at Seabrook Station with the-intent of conso'idating deficiency reporting methods and standardizing the problem evaluation and resolution process. This procedure, which will address sequences of events such as those described in Enclosure (1), is expected to be implernented by June 30, 1993.

I

'2 D

y. g,--e- ,,, .b h [ p ... .- a- - . . y,_ y.- m- s , , , . ,,,E.

,,,m,%wv ,,,..y,,,,,.es,,,,_e___,# mm-.ew, rmw,c-w.---e~h----c--e*,.r..--- --=e-..-i,-t