ML20206S285

Fort St Vrain Nuclear Generating Station Site-Specific Offsite Radiological Emergency Preparedness Alert & Notification Sys QA Verification

ML20206S285
Person / Time
Site:	Fort Saint Vrain
Issue date:	02/20/1987
From:	INTERNATIONAL ENERGY ASSOCIATES, LTD.
To:
Shared Package
ML20206S209	List: ML20206S203 ML20206S285
References
NUDOCS 8704220368
Download: ML20206S285 (45)

Text

mmm m WU W MW lNTERNATIONAL Energy and Environment Group FORT ST. VRAIN NUCLEAR GENERATING STATION SITE-SPECIFIC OFFSITE RADIOLOGICAL EMERGENCY PREPARED!!ESS ALERT AND NOTIFICATION S'ISTEM OUALITY ASSURANCE VERIFICATION Fregared for F-2deral Er..ergency Management Agency Nashir.gton, D. C.

20472 Under Contract No.

E.W-3 3 -C-1217 February 20, 198'i International Energy Associates Limited 8704220368 870402 3211 Jermantown Road PDR ADOCK 05000267 fair *ax, Wg nta 22030 F

PDR (703) 246-0200

FORT ST. VRAIN NUCLEAR GENERATING STATION SITE-SPECIFIC OFFSITE RADIOLOGICAL EMERGENCY PREPAREDNESS ALERT AND NOTIFICATION SYSTEM OUALITY ASSURANCE VERIFICATION Prepared for Federal Emergency Management Agency Washington, D.C.

20472 Under Contract No. EMW-83-C-1217 i

February 20, 1987 i

i

FORT ST. VRAIN NUCLEAR GENERATING STATION SITE-SPECIFIC OFFSITE RADIOLOGICAL EMERGENCY PREPAREDNESS ALERT AND NOTIFICATION SYSTEM OUALITY ASSURANCE VERIFICATION Prepared for Federal Emergency Management Agency Washington, D.C.

20472 Under Contract No. EMW-83-C-1217 February 20, 1987

,--n.-..r-.m---,,

,-,----,,,,n.-r--

er-n-

TABLE OF CONTENTS I.

INTRODUCTION 1

A.

Identification 1

1. Site Information 1

2. Governments Within The 5-Mile Emergency Planning Zone 2

B.

Scope Of Review 2

1. Emergency Plans For Offsite Response Organizations 2

2. Alert And Notification System Design Report 2

3. FEMA Evaluation Findings 3

II.

FINDINGS FOR EVALUATION CRITERION E.6 4

A. Administrative Means Of Alerting (E.6.1, FEMA-43) 5 B.

Physical Means Of Alerting (E.6.2, FEMA-43) 6

1. Tone Alert Radios (E.6.2.3, FEMA-43) 8

2. Special Alerting (E.6.2.4, FEMA-43) 11 III. FINDINGS FOR EVALUATION CRITERION N.1 12 A.

Findings For August 15, 1984 Demonstration 12 B.

Findings For June 18, 1985 Demonstration 22 C.

Findings For August 6, 1986 Demonstration 26 IV.

FINDINGS FOR EVALUATION CRITERIA E. 5, F.1, N. 2, N. 3, AND N. 5 37 REFERENCE LIST 38 APPENDIX A:

Sample Size Determination i

Fort St. Vrain Nuclear Generating Station Site-Specific Offsite Radiological Emergency Preparedness Alert And Notification System Quality Assurance Verification State Of Colorado Weld County I.

INTRODUCTION A.

Identification

1. Site Information The Fort St. Vrain Nuclear Generating Station is located about three and one-half miles northwest of the town of Platteville, Colorado, and about 37 miles north of Denver in Weld County.

It is owned and operated by the Public Service Company of Colorado (located in Denver, Colorado) and began commercial operation in 197 9.

The area within a few miles of the reactor site is characterized by irrigated farm and pasture land with gently rolling hills.1 In general, the majority of the land within 30 miles of the site is agricultural.

The population density within the 5-mile emergency planning zone (EPZ)

• surrounding the site is rela-tively low.

The population within the 5-mile EPZ is approximately 2,077 persons.3

• The Federal Emergency Management Agency / Nuclear Regulatory Commission (FEMA /NRC) Steering Committee concluded that small water-cooled reactors and the Fort St. Vrain Nuclear Genera-ting Station may use a 5-mile EPZ rather than the standard 10-mile EPZ. 2 1

2. Governments Within The 5-Mile Emergency Plannina Zone The Fort St. Vrain Nuclear Generating Station EPZ consists of a 5-mile-radius circle with the Fort St.

Vrain Nuclear Generating Station as the center point.

This EPZ lies entirely in Weld County and contains only one town, Platteville.

In case of an emergency at the Fort St. Vrain Nuclear Generating Station, the State of Colorado makes the decision to activate the early warning system.

Weld County is then responsible for physical activation.

B.

Scope Of Peview

1. Emergency Plans For Offsite Response Organizations All appropriate offsite radiological emergency plans and preparedness site-specific to the Fort St. Vrain Nuclear Generating Station have been reviewed by the Federal Emergency Management Agency (FEMA) Region VIII and the Regional Assistance Committee.

2. Alert And Notification System Design Report The physical means established for alerting the public within the Fort St. Vrain Nuclear Generating Station EPZ are documented in the following:

Federal Emergency Management Agency, letter from Alton D.

Cook, Regional Director, to Richard Krimm, Assistant Associate Director.

Subj ect :

Documents to be supportive of the public alerting and notification system evaluation, dated April 2, 1984;3 supplemented by 2

Public Service Company of Colorado, letter f rom O. R.

Lee, Vice President, to Alton D.

Cook, Federal Emergency Management Agency.

Subj ect :

Radiological alarm in Platteville, Colorado, dated August 16, 1985.7 l

These documents are hereinafter collectively referred to as the Design Report.

3. FEMA Evaluation Findings The offsite radiological emergency plans and pre-paredness site-specific to the Fort St. Vrain Nuclear Generating Station received approval under Title 44 of the Code of Fede ral Regulations, Part 350 (44 CFR 350), conditioned upon verification of the adequacy of the public alert and notification system, as documented in the:

Letter to the Honorable Richard D.

Lamm, Gov-ernor of Colorado, signed by Lee M. Thomas, Associate Director, State and Local Programs and Support, FEMA, dated January 29, 1982;4 and the Letter to William J.

Dircks, Executive Director for Operations, U.S.

Nuclear Regulatory Commis-sion, signed by Lee M.

Thomas, Associate Dir-ector, State and Local Programs and Support, FEMA, dated January 29, 1982.9 l

3

II.

FINDINGS FOR EVALUATION CRITERION E.6 The Design Report describing the alert and notification system for the Fort St. Vrain Nuclear Generating Station was reviewed against evaluation criterion E.6 and Appendix 3 of NUREG-0654/ FEMA-REP-1, Revision 1, " Criteria for Pre-paration and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of Nuclear Power Plants" (hereinaf ter ref erred to as NUREG-0654/ FEMA-REP-1, Rev.

1).

This evaluation criterion states:

Each organization shall establish administrative and physical means, and the time required for notifying and providing prompt instructions to the public within the plume exposure pathway Emergency Planning Zone.

(See Appendix 3.)

It shall be the licensee's responsibility to demon-strate that such means exist, regardless of who implements this requirement.

It shall be the responsibility of the State and mentstoactivatesuchasystem.gocalgovern-The bases for review against this evaluation criterion were the corresponding acceptance criteria of FEMA-43, " Standard Guide for the Evaluation of Alert and Notification Systems for Nuclear Power Plants" (hereinafter referred to as FEMA-43).5 Based upon this review, International Energy Associates Limited concluded that the design and implement-ation of the alert and notification system at the Port St.

Vrain Nuclear Generating Station and its supporting proce-dures conformed sufficiently to the acceptance criteria, as stated in FEMA-43, for evaluation criterion E.6 of NUREG-0654/ FEMA-REP-1, Rev.1, to support a FEMA finding that the alert and notification system is adequate.

Because the August 15, 1984 demonstration and subsequent public survey resulted in an approximate 40% affirmation alerting ratio for the tone alert system, a second demon-i stration was conducted on June 18, 1985.

This second i

4

demonstration and public survey resulted in an approximate 70% affirmation alerting ratio, which is considered to be only marginally acceptable.

Review of the first two surveys conducted on August 15, 1984 and June 18, 1985 identified Platteville, Colorado as an area where the alerting system could be enhanced.

Platteville, which is 37 miles northeast of Denver, is the only community within the 5-mile Port St. Vrain Emergency Planning Zone (EPZ).

Most of the approximately 2,000 persons within the EPZ live in Platteville.

The Public Service Company of Colorado committed, by letter to FEMA, to supplementing the existing early warning alert system by installing an audible alarm in the community of Platteville, Colorado.7 Consequently, the Fort St. Vrain tone / weather alert radio system was enhanced by installa-tion of a siren at the southwest corner of the Platteville City Hall.

This portion of the quality assurance verification evalu-ates the Fort St. Vrain Nuclear Generating Station's alert and notification system against FEMA-43 acceptance criteria in the following areas:

administrative means of alerting, physical means of alerting, and the special alerting methods.

A. Administrative Means Of Alerting (E.6.1, FEMA-43)

The information that is specifically cited in the Public Service Company of Colorado's Design Report addresses those individuals within that organization who are re-sponsible for recommending alert and notification system activation to the local governments.

The Design Report also specifies those individuals within the state and local governments who are responsible for alert and 5

notification system activation.

After reviewing the aforementioned documentation dealing with emergency procedures for the Fort St. Vrain Nuclear Generating Station, Weld County, and the State of Colorado, the decision logic as shown in Figure 1 was developed.

As Figure 1 indicates, the procedures satisfy the accep-tance criteria of FEMA-43.

These emergency procedures also specify the means by which the request to activate the alert and notification system at a specified time is conveyed f rom the Colorado Department of Health to the Weld County Sheriff- (who is responsible for alerting the af fected population).

The Fort St. Vrain Nuclear Generating Station's early warning system employa Weatheralert tone / weather alert radios to notify all residences and businesses within the 5-mile EPZ.

These radios are activated by the National Weather Service of the National Oceanic and Atmospheric Administration (NOAA).

Telephone hookups are in place to relay the message to NOAA from the Weld County Sheriff.

The emergency broadcast system (EBS) stations receive up-to-date reports on the status of an emergency from the Weld County Civil Def ense Director via a telephone communication system.

Additionally, the means are in place for the notification of local school authorities by commercial telephone to inform them of the situation.

B.

Physical Means Of Alerting (E.6.2, FEMA-43)

The physical means of alerting for the Fort St. Vrain Nuclear Generating Station consist of tone / weather alert radios, which have been offered to all residences within the 5-mile EPZ and are supplemented by one fixed siren.

6

i ~

~

i s

FIGl FORT ST. VRAIN NUCLEJ ALERT AND NOTIFICATION SYSTEM ACTIt Colorado Dept.

ggg Of Health is Of Thi Notified Of Decisi it.e Situation g33 I

I I

i fort St. Vrain l c]

Weld Co. Civil deld County Supervisor Oe.

l ;;

Defense Direc.

Emergency Op-termines Level

~

tor !$ Nott, erations Cen.

Of Emergency l*

And Begins The fled Of The ter (EOC) Is hottfication lC Situation Activated Process

%!k lE l'

Weld Co. Sher-iff is flo41

~

I fled Of The Situation l

g State Emergency Operations Cen-

)

ter (EOC) is Activated Weld Co. Com.

Colorado Divi-munications slon Of Disas.

Center (CC) is ter Emergency hotified Of Services (000ES)

The $ltuation is Notified Of forward Css The Situation Fort Lupton mand Post /t Police Depart-Fort Luptrn ment is Nati-is Activa;ed

~

fled Of The Colorado State Situation Patrol (Greeley Division) Is Notified Of The Situation

RE 1 LR GENERATIf1G STATION N gg ATION DECISI0fi/ ACTION SEQUEllCE DIAGRAM kNCAN Aho AvanSD Aperture Car sed On 4 s.ne ' t Comercial Broad-1:cident.

cast Stations (TV 4 Is Made T' And Radio) Are o Warn-Ix9 Notified To Begin The Warning Broad-cast

\\

W;ld Co. Civil EBS Stations Are

,(g',

Notified To Begin

+

Through CC Of The The Warning Broad-Decision To In-C

plement Warning W;1d Co Sheriff Weld County fational Weather is Notified Sheriff Segins Service is Nott-Tone Alert Thrnugh CC Of The To Disseminate fled Of Decision Radios Are Lecision lo In-Activated Warning To The 1o Begin The plement W rnin9 Public harning Broad-cast Civil Defense Sirens Are 000CS is Notified All EBS & Com-Sounded in Platte-Of the Cec t s t on mercial Broadcast ville If Available

+

Ts Implen.ent Stations Are Noti

• For Use Warning fled Of The Appro-priate Warning To Be Broadcast Loudspeaker.

Equipped Vehicles (Plane And Patrol Cars) Are Deployed School Authorities And Other Densely Populated Busi-nesses Are Nott-fled By Telephone Door-To-Door Notification Begins 87o yt zoM

-o/

l.

The siren, which is installed on a 47-foot tower, is a SiraTone Model EOWS-612, manuf actured by the Federal Signal Corporation.

The siren is radio-controlled and is capable of producing six standard signals through an array of six speakers.

According to the manufacturer's specifications, the sound level at 100 ft f rom the siren is 115 dB.

This siren was found to be adequate to enhance alerting coverage in the geographical area of concern in the vicinity of Platteville.8 Routine siren testing procedures and operability for the Fort St. Vrain Nuclear Generating Station siren warning system have been reviewed and determined to satisfy FEMA-43 operability requirements.

1. Tone Alert Radios (E.6.2.3, FEMA-43)

The early warning system for the Fort St. Vrain Nu-clear Generating Station incorporates a tone alert system utilizing Weatheralert Model TA-45 weather radios that operate on the National Weather Service communications system.

Access to the National Weather Service communications system to broadcast emergency messages concerning the Fort St. Vrain Nuclear Generating Station is described in an agree-ment between the Division of Disaster Emergency Ser-vices and the National Weather Service, both located in the State of Colorado.

FEMA has developed guidelines that should be followed to maintain an effective and continual alert and noti-fication system at the Fort St. Vrain Nuclear Genera-ting Station.

These guidelines are as follows:

. The program should offer the tone / weather alert radios to the public in geographical areas where needed and must make a "best-effort" attempt to place the radios.

This program should include a 8

record system (register) that contains an accur-ate list of addresses (names are optional) in geographical areas where tone / weather alert radios are needed.

Addresses where radius are offered to residents and refused by the residents should be noted.

A maintenance program of fering operating checks should be available at least annually to all res-idences in areas where tone / weather alert radios are needed.

The maintenance program and the register program mentioned above may be integrat-ed.

Tests of the tone alerting feature are desired at least monthly.

The final determination of test-ing f requency will rest with appropriate local government officials.

The results of these tests do not have to be monitored.

The purpose of these tests is to offer the public a means to self-test their receivers.

Written guidance should accompany the radio.

It should address (1) its general use, (2) self-testing f requency and method, (3) suggested placement to facilitate ef ficient monitoring, (4) the maintenance program, and (5) telephone num-bers for repair or replacements.

This informa-tion should be provided as a reminder to each tone alert radio holder annually.

This public information program may also be integrated with the register and maintenance programs mentioned a bov e.

Determination should be made that the broadcast medium for initiating the tone alert signal has adequate availability (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day, 7 days a week), signal strength, and signal quality.3 In early 1982, the Public Service Company of Colorado identified 1,077 residences and businesses that were located within the 5-mile EPZ.

Eleven tone / weather alert radios were not delivered.

Of these 11, eight residences could not be contacted (even af ter repeat-ed attempts), and three residents refused to accept them.3 The Public Service Company of Colorado is also dev-eloping a system for flagging electric meters, gas 9

meters, or both within the EPZ and will use this system to identify residences that may be vacated, sold, or rented to new persons.

This same system will provide information on any new building that is planned for the area.

The Public Service Company of Colorado has turned the tone / weather alert radio system over to the State of Colorado for its use but has agreed to maintain the system.

As indicated in the utility's information brochure and on the radio decal, persons within the EPZ have been given instructions to call the Public Service Company of Colorado in the event of any tone alert problems.

Additionally, batteries (the backup power) are mailed to each residence annually or upon r equest. The National Weather Service tests the alert system every Wednesday morning between 11:00 a.m. and 12:00 noon (Mountain Time).

The tone / weather alert radios were personally deli-vered by Public Service Company of Colorado repre-sentatives.

Operation of the radios was demonstrated and the alert system explained.

Each residence was left with a booklet of instructions, as well as a question and answer booklet.

In addition to the instruction booklet, a decal was placed on the radio to ensure ready access to emergency instructions.3 The National Weather Service operates two broadcast stations, providing adequate coverage for all of the residences involved within the S-mile EPZ.

The early warning system developed by the Public Ser-vice Company of Colorado for the Fort St. Vrain Nu-clear Generating Station, as described in the Design Report, meets FEMA guidelines addressing tone alert 10

radio systems, thus satisfying the criteria of NUREG-0654/ FEMA-REP-1, Rev. 1.

2. Special Alerting (E.6.2.4, FEMA-43)

The Fort St. Vrain Nuclear Generating Station's early warning system incorporates special alerting, as de-fined in FEMA-43.

The Weld County Sheriff is respon-sible for requesting the National Weather Service to broadcast warning messages, thereby activating the tone / weather alert radios.

He is also responsible for:

Sounding the Civil Defense warning sirens in the affected area (in coordination with the Weld County Civil Def ense Coordinator);

Deploying loudspeaker-equipped vehicles (an air-plane and patrol cars) in the affected area; Notifying school authorities, other densely popu-lated f acilities or institutions, and isolated f arm families via telephone and a citizen band radio system (in conjunction with the Weld County Civil Def ense Coordinator) ; and Dispatching personnel for door-to-door notifi-cation of known handicapped or infirm persons.

Finally, all businesses within the 5-mile EPZ have been given tone / weather alert radios.3 11

III. FINDINGS FOR EVALUATION CRITERION N.1 A.

Findings For The August 15, 1984 Demonstration On August 15, 1984, the physical means (tone / weather alert radios) used to alert the population within the EPZ for the Fort St. Vrain Nuclear Generating Station were demonstrated to satisfy the alert and notification aspects of 44 CFR 350.9(a).

This demonstration was conducted using the methods specified in Section N.l. (a, b).2 of FEMA-43.5 The resnits indicated that this portion of the alert and notification system evaluation was not acceptable and failed to conform to FEMA-43 and NUREG-0654/ FEMA-REP-1, Rev.1.2 The August 15, 1984, demonstration of the Fort St. Vrain Nuclear Generating Station early warning system con-sisted of the transmission of an alerting signal to the tone / weather alert radios within the EPZ, and a subse-quent telephone survey to estimate the proportion of EPZ households actually alerted and to identif; areas where enhancement of the alerting system might be needed.

The signal was transmitted at 10:00 a.m. Mountain Daylight Time (MDT).

The telephont survey of EPZ residences was begun at approximately 10:05 a.m.

(MDT) and was completed within one hour and 10 minutes.

This survey was conducted by approximately 40 telephone interviewers, each with a separate WATS line and computer terminal.

The universe of households to be surveyed was determined by establishing a 7.5-mile-radius circle around the latitude and longitude of the power staticn.

A sorted master list (addresses and telephone numbers) was 12 j

obtained of 2,390 households within the 7.5-mile-radius circle.

The address of each household was then checked to determine whether the household was within the 5-mile EPZ.

This review produced 826 households that appeared to be within the EPZ.

These 826 comprise nearly the entire universe of households within the Fort St. Vrain Nuclear Generating Station EPZ.

Replicated samples were developed t rom this list of households.

A sufficient number of these subsamples were developed to ensure that the required number of telephone calls could be made, i.e., to establish the proportion of households alerted to within a 5% precision at a 95% confidence level.

The method for sizing the sample to achieve this result is described in Appendix A of this report.

To ensure that the Spanish-speaking population was accurately surveyed, some interviewers also conducted the survey in Spanish.

The English and Spanish questionnaires used for the telephone survey are ' included as Figures 2 and 3 of this report.

As part of the telephone survey, 294 households within the Fort St. Vrain Nuclear Generating Station EPZ were contacted and their responses were collected in an automated data base.

However, before running the final tabulations, each of these addresses was checked (based on collected crossroads information) on a street map to validate its location.

Of these 294 addresses, 33 were outside the EPZ.

Therefore, the data were tabulated on the 261 respondent households located within the EPZ.

Respondents at 36 of these households had been away f sn home at the time of the demonstration of the early warning system and therefore were not included in the 13

1. 2106Q FIGURE 2 Chilton Research Services Study #8521 Radnor, Pennsylvania August,1984 OMB #3067-0103 (FEFA 9/83)

FEMA NUCLEAR POWER PLANT ALERTING AND N0liFICAfiON ST5icli: PUBLIC TELEPHONE SURVET FORT ST. VRAI!!

Time Began AM PM Interview #

(i-5)

Time Ended All PM Zip Code (6-io)

Sample Type lii)

RECORD BEFORE DIALING -Telephone #

(Area Gooe)

( Excnange)

(Numoer)

(i 2-21 )

RESPONDENT: Itale or Female head of household.

(ASK, DEPEHOING Ot! 'EX: Are you the (man of the house / lady of the house)?

INTRODUCTION:

Hello, my name is We're calling households long distance from Chilton Research Services as part of a survey. This survey is sponsored by The Federal Emergency Management Agency (FEMA) of the United States Government.

Your answers are voluntary and will be kept strictly confidential.

1.

First of all, is this (REPEAT # DIALED)?

Yes 1

TERMD! ATE AND DIAL AGAIll Ho 2

2.

As ycu may or may not know, there was a test of the public warning / alert notification system for THE FORT ST. VRAIN flVCLEAR GENERATING STATION. Did ycu, or any other member of. your nousencio, near a signai from a weatner alert radio around 10 A.M.

this morning?

22-SKIP TO 0, 1 Yes 1

SK!P TO 0. 4A Ho 2

nea rn f rcn SKIP TO 0. 4 another 3

scurce A5K 1F ANf GThEP. n0v5Eh0LD Don,,. Knew 8

MEMBER IS MORE K!!OWLEDGEABLE 14

FIGURE 2 (CONTINUED)

Sp. (23-36) 3.

THERE IS NO OUESTION #3.

4 (IF " HEARD D ERGENCY SIGNAL" ASK 0. 4 BELOW; OTHER!lISE SKIP TO 0. 4A)

Were you at this location when you heard the signal from the weather alert radio?

37-Yes 1

SKIP TO Q. 5 No 2

4A.

(IF "DID HnT HEAR EMERGENCY SIGNAL")

'dere you at this 1ocation at around 10:00 A.M. this morninc?

38-Yes 1

Ho 2

ASK Q. 48 Don't Know Y

48.

Ha:: this household ever been issued a Weather Aler: P.adio?

39-ASK 0. AC Yes 1

Ho 2

SKIP TO 0. 5 Don't Know Y

15

FIGURE 2 (CONTINUED)

AC. Was the red light on this household's weather alert radio lit today?

a0-fes 1

No 2

Don't Know Y

5.

Has this household ever received instructions which tell you what to do in the event of a "real" emergency at Fort St. Vrain? These brown brochures in both English and Scanish were sent out by the Public Service Company of Colorado and was entitled, "Infomation about the Ft. St. Vrain Radiological Emergency Response Plan." Do you remember receiving this infomation.?

al -

(es i

Ho 2

Don' t Know Y

6..

Because we need to detemine whether or not this household is within the 5 mile Emergency Planning Zone of Fort St. Vrain, would you please give me the address for this location? (PAUSE FOR AHSWER)

ADDRESS:

and the nearest intersection (or cross st eet) to this lccation.

On behalf of Chilton Research Services and the Federal Energency Management Agency, I woulo like to thank you for your time and for giving us this valuable infomation.

16

FIGURE 3 C2270Q Chilton Research Services Study #8521 Radnor, Pennsylvania August, 1984 FEMA NUCLEAR POWER PLANT ALERTING AND NOTIFICATION SYSTEM:

PUBLIC TELEPHONE SURVEY FORT ST. VRAIN - Spanish Version Horo cemenzado AM PM Entrevistator #

Hora Ter=inado AM

?M Zip Code Sa=ple Type ESCRIEA ANTES DE LLAMAR POR TELEFCNO - #

(Area Code)

(Exchange)

(Nc=ber)

RES?CNDIENTE: El senor o senora cabeza de familia.

(PREOUNTE, DEFENDIENDO DEL SEIO: Es Ud. (la cabeza) de familia?

INTRCrUCCION:

Buenas (tardes/ dias), =i nc=bre es Esta=os llamando de larga distancia desde Chilton Recearch Servi ~ces, como parte de u=a encuesta, patrocinada por la Agencia Federal del Manejo de E=ergencias (FEDERAL EMERGENCY MANAGEMENT AGENCY) del Gobierno de los Estados Unidos.

Sus respuestas son voluntarias y se =antendran en for=a confidencial, 1.

Pri= era = ente, es este al nt=ero (RE?ITA EL # DE TELEFONO)?

Si 1

f No TERMINE Y VUELVA A LLAMAR 2

2.

Puede que Ud. este enterado que se llevo a cabo una prueba de el alerta del siste=a publico de e=ergencias para la planta generadora nuclear Fort St. ? rain. Escucho Ud.,

o algun otro =ie= bro de su familia, una senal de alar =a en la radio que.avisa las condiciones at=cafericas, a las (HORA) de hey?

SKI? TO Q. 4 Si 1

SKI? TO 0. 4A No 2

SKI? TO Q. 4 Otro =edio 3

ASK IF CTHER HH MIM3ER MORE KNOWLEDGEABLE No Se Y

17

FIGURE 3 (CONTINUED) 4.

(SI CONTESTAN "ESCUCHE LA SENAL DE EMERGENCIA", PRECUNTE LA P.4 ABAJO, SI NO, PASE A LA P.4A).

Estaba Ud. en ese local cuando escucho la senal de alar =a en la radio que avisa las condiciones atnosfericas?

Si 1

PASE A P. 5 No 2

4A.

(SI CCNTESTA "NO ESCUCHE LA SENAL DE EMERGENCIA"):

Estaba Ud. hoy 7n ese local a las (HORA DE LA SENAL DE ALERTA)?

Si 1

No 2

No se Y

4E.

Le lan entregado and un radio para alertad de eltie=po?

ASK Q. 4C Si 1

SK!P TO Q. 5 No 2

No se Y

4C.

Estaba la lu: roja encendida en el radio de alerta en el dia de hoy?

Si 1

No 2

No se Y

18

FIGURE 3 (CONTINUED) 3.

Han recibido en su hogar instrucciones que dicen lo que hacer en caso de una "verdadera" e=ergencia en la planta generadora nuclear Fort St. Vrain? Estos folletos de color cafe en ingles y espanol fueron enviados por la Co=pania de Servicios Publicos de Colorado y se titula "INFORMACION SOERE EL PLAN PARA EMERGENCIA RADIOLOGICA EE FORT ST.VPAIN".

Ud. recuerda haber recibido esta infomacion?

Si 1

No 2

No se Y

6.

Debido a que debe=os saber si Ud. vive o no dentro de la zona de 5 millas del Plan de Energencia de Fort St. Vrain, podria dar=e su direccion?

DIRECCION:

y la calle principal o cruce principal cerca de su hogar es:

En no=bre de Chilton Research Services y de la Agencia Federal del Manejo de Energencias, deseo agradecerle su tie po y la atencion que =ostro al dar=e esta valiosa info macion.

19

alerting analysis.

Of the remaining 225 households, 41.3% (93) indicated that they had been alerted during the demonstration.

If one uses the estimated number of households within the EPZ (which is 793 according to the analyses of sample addresses) in the confidence interval expression in Appendix A of this report, it yields an estimated 95% confidence interval for the proportion of the total EPZ population alerted that ranges f rom 36.0%

to 46.9%.

In other words, at a 95% confidence level, between 36.0% and 46.9% of the households within the EPZ were alerted by the early warning system.

The sample of 261 households was also used to estimate the proportion of households within the EPZ that stated they received information about what to do in the event of a real emergency at the Fort St. Vrain Nuclear Generating Station.

Of these 261 households, 80.5%

(210) responded that they had received this information, 13.0% (34) responded that they had not received this information, and 6.5% (17) did not know or refused to state whether they had received this information.

Using the approach discussed previously, the following estimates for the entire EPZ population resulted (at the 95% confidence level) :

Between 76.2% and 84,1% of the households would have reported receiving the information; Between 10.0% and 16.7% of the households would have responded that they had not received the information; and Between 4.5% and 9.4% of the households would not have known or refused to state whether they had received the information.

20

A difficulty, related to drawing valid statistical inf er-ences f rom this survey, arose because the Public Service Company of Colorado conducted a similar concurrent tele-phone survey.

As would be expected concerning an EPZ containing such a small number of households, many house-holds were contacted by both surveys.

This may have af f ected the accuracy of the survey in two ways.

First, when our interviewers, unaware of the concurrent survey, contacted a number of households, they were told that the household had already been interviewed.

Once we became aware of the other survey, many households had to be recontacted, which added extra time to the interview-ing period.

The major reason that a relatively large number of interviewers are employed for these surveys is the concern that the information gathered immediately af ter the alert and notification system demonstration is more accurate than that gathered later.

Consequently, anything that delays the survey has the potential to adversely af fect its accuracy.

The concurrent Public Service Company of Colorado survey may also have introduced a subtle bias in our survey.

It is likely that the Public Service Company of Colorado survey sample was developed f rom the Fort St. Vrain tone alert radio register.

A similar survey has been shown to select a larger fraction of the alerted population than surveys drawn at random f rom EPZ households.6 ye is reasonable to expect that individuals contacted pre-viously by the Public Service company of Colorado survey would be more i tkely to refuse to respond to our sur-vey.

Therefore, it is possible that our survey under-estimated the actual proportion of the population that was alerted due to this sampling problem.

However, since the extent of this possible problem cannot be quantified, it is difficult, if not impo s sible, to cor-rect for it.

21

As this discussion indicates, the performance of con-current surveys by utilities or other organizations during an alert and notification system demonstration has the potential to cast doubt upon the overall accur-acy of the data gathered and should be avoided in the interest of ensuring a true measure of the alert and notification system performance.

The survey data were reviewed to identify areas where the alerting system could be enhanced.

The primary area identified was public instruction on the operation of tone / weather alert radios.

The survey revealed that 33.8% (76) of those individuals contacted, who were at home at the time of the demonstration, indicated that they were not operating their tone / weather alert radios in a manner that permitted the radio to annunciate upon receipt of an alerting signal.

This leads to the con-clusion that, at a 95% confidence level, between 28.7%

and 39.2% of the households within the EPZ would have reported that they were not operating their tone / weather alert radios so that the radio would annunciate upon receipt of an alerting signal.

B.

Findings For The June 18, 1985 Demonstration On June 18, 1985, the physical means (tone / weather alert radios) used to alert the population within the EPZ for the Fort St. Vrain Nuclear Generating Station were demonstrated for a second time to satisfy the alert and notification aspects of 44 CFR 350.9(a).

This demon-stration was conducted using the methods specified in Section N.1. (a,b,).2 of FEMA-435 and was to determine whether enhancements made to the alert and notification system subsequent to the August 15, 1984 demonstration were adequate to bring the system into conformance with the requirements of FEMA-43 and NUREG-0654/ FEMA-REP-1, 22

Rev. 1.2,5 The results of this second demonstration indicated that this portion of the alert and notifi-cation system was in marginal conformance with FEMA-43 and NUREG-0654/ FEMA-REP-1, Rev. 1.

The June 18, 1985, demonstration of the Fort St. Vrain Nuclear Generating Station alert and notification system consisted of the transmission of three separate alerting signals to the tone / weather alert radios within the EPZ and a concurrent telephone survey to estimate the pro-portion of EPZ households actually alerted.

The tone alert activation was originally scheduled for 12:15 p.m.

(MDT).

However, the first activation actually occurred at 10:32 a.m.

(MDT).

Upon notification of this activa-tion by a FEMA Region VIII representative, 32 inter-viewers were quickly assembled and brief ed, and the survey began at 11:08 a.m.

(MDT).

The second alerting signal was transmitted to the tone / weather alert radios while the survey was in progress, at approximately 11:25 a.m.

(MDT).

The third activation of the tone / weather alert radios occurred subsequent to the survey at 3:32

p. m.

(MDT).

The telephone su.rvey of EPZ residences was completed at 12:37 PM (MDT) (1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 29 minutes after the start of interviewing and 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> and 5 minutes after the first tone alert activation).

The universe of households to be surveyed was again determined by establishing a 7.5-mile-radius circle around the latitude and longitude of the power station.

A second sorted master list (addresses and telephone numbers) was obtained of 2,201 households within the 7.5-mile-radius circle.

The address of each household was then checked to determine whether the household was within the 5-mile EPZ.

This review produced 875 house-holds that appeared to be within the EPZ and which 23

comprise nearly the entire universe of households within the Fort St. Vrain Nuclear Generating Station EPZ.

Rep-licated samples were developed f rom this list of house-holds.

A sufficient number of these subsamples were developed to ensure that the required number of tele-phone calls could be made, i.e., to establish the propor-tion of households alerted to within a 5% precision at a 95% confidence level.

The method for sizing the sample to achieve this result is described in Appendix A of this report.

To ensure that the Spanish-speaking population was accurately surveyed, some interviewers were prepared to conduct the survey in Spanish.

The English and Spanish questionnaires used for this telephone survey were the same as those used in the August 15, 1984 survey (see Figures 2 and 3 of this report).

As part of the telephone survey, 378 households believed to be within the Fort St. Vrain Nuclear Generating Sta-tion EPZ were contacted, and the responses were collect-ed in an automated data base.

However, before running the final tabulations, each of these addresses was checked (based on collected address and crossroads information) on a street map to validate its location.

Of these 378 addresses, 69 were outside the EPZ.

There-fore, data were tabulated on the 309 respondent house-holds that were located within the EPZ.

Respondents at 51 of these households had been away from home at the time of the demonstration of the alert and notification system and, therefore, were not included in the alerting analysis.

Of the remaining 258 households, 69.8% (180) indicated that they had been alerted during the demon-stration.

If one uses the estimated number of house-j holds within the EPZ (which is 793 according to the 24

analysis of sample addresses) in the confidence interval expression in Appendix A of this report, it yields an estimated 95% confidence interval for the proportion of the total EPZ population alerted that ranges f rom 65.0%

to 74.2%.

In other words, at a 95% confidence level, between 65.0% and 74.2% of the households within the EPZ would have stated that they were alerted by the early warning system.

Respondents at the households that reported that they were not alerted were asked whether their tone / weather alert radios were turned on and set such that an alert-ing signal could be received.

Respondents at 35 house-holds indicated that their tone / weather alert radios were not being operated in a manner that permitted receipt of an alerting signal.

Thus, 80.7% of the households whose residents were home at the time of the demonstration and that had tone / weather alert radios operating in a manner that permitted receipt of an alerting signal reported that they were alerted.

Using the estimated number of households within the EPZ and the confidence interval expression in Appendix A, a

95% confidence interval ranging f rom 76.0% to 84.7% is obtained for the proportion of the total EPZ population with operating tone / weather alert radios that would have stated that they were alerted.

The sample of 309 households was also used to estimate the proportion of households within the EPZ that stated they received information about what to do in a real emergency at the Fort St. Vrain Nuclear Generating Sta-l tion.

Of these 309 households, 82.8% (256) responded that they had received this information,12.3% (38) responded that they had not received this information, l

25

and 4.9% (15) did not know whether they had received this information.

Using the approach discussed pre-viously, the following estimates for the entire EPZ population resulted (at the 95% confidence level):

Between 79.3% and 85.9% of the households would have reported receiving the information; Between 9.7% and 15.4% of the households would have responded that they had not received the informa-tion; and Between 3.3% and 7.1% of the households would not have known whether they had received the infor-mation.

Review of the survey data to identify areas where the alerting system could be enhanced indicated a need for additional public instruction on the operation of tone /

weather alert radios and the importance of operating these radios in a manner that permits receipt of an alerting signal.

Alternatively, the alerting system could be enhanced by installation of a physical means of alerting (such as an audible alarm) for areas of high population density that would not require operation by members of households within the EPZ to ensure that an alerting signal would be received.

C.

Findings For The August 6, 1986 Demonstration Prior to conducting the third demonstration on August 6, 1986, the Public Service Company of Colorado installed a SiraTone Model EOWS-612 siren at the southwest corner of the Platteville City Hall in Platteville, Colorado.

l 26 t-

On August 6,1986, the physical means (one siren and tone / weather alert radios) used to alert the population within the EPZ for the Fort St. Vrain Nuclear Generating Station were demonstrated for a third time to satisfy the alert and notification aspects of 44 CFR 350.9(a).

This demonstration was conducted using the methods specified in Section N.1. (a,b,).2 of FEMA-435 and was to determine whether enhancements made to the alert and notification system subsequent to the June 18, 1985 demonstration were adequate to bring the system into conformance with the requirements of FEMA-43 and NUREG-0654/ FEMA-REP-1, Rev. 3.2 The results of this third demonstration indicated that this portion of the alert and notification system was in conformance with FEMA-43 and NUREG-0654/ FEMA-REP-1, Rev. 1.

The August 6,1986, demonstration of the Fort St. Vrain Nuclear Generating Station alert and notification system began at approximately 5:28 p.m.

(Mr") with activation of the tone alert radios, followed by the siren sounding at approximately 5:29 p.m.

(MDT) and a concurrent tele-phone survey to estimate the proportion of EPZ house-holds actually alerted.

Upon notification of this activation by a FEMA Region VIII representative, 34 interviewers began dialing telephone numbers of the sampled households within the EPZ.

The telephone survey of EPZ residences was completed at approximately 7:03 p.m. (MDT) (1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 35 minutes after the start of interviewing).

i The universe of households to be surveyed was determined by establishing a 10-mile-radius circle around the lati-tude and longitude of the power station.

A second sort-ed master list (addresses and telephone numbers) was obtained of 2,700 households within a 10-mile-radius 27

ci rcl e.

The address of each household was then checked to determine whether the household was within the 5-mile EPZ.

This review produced 788 households that appeared to be within the EPZ and which comprise nearly the en-tire universe of households within the Fort St. Vrain Nuclear Generating Station EPZ.

Replicated samples were developed f rom this list of households.

A sufficient number of these subsamples were developed to ensure that the required number of telephone calls could be made, i.e., to establish the proportion of households alerted to within a 5% precision at a 95% confidence level.

The method for sizing the sample to achieve this result is described in Appendix A of this report.

To ensure that the Spanish-speaking population was accurately surveyed, some interviewers were prepared to conduct the survey in Spanish.

The English and Spanish questionnaires used for this telephone survey are included as Figures 4 and 5 of this report.

As part of the telephone survey, 371 households believed to be within the Fort St. Vrain Nuclear Generating Sta-tion EPZ were contacted, and the responses were collect-ed in an automated data base.

Powever, before running the final tabulations, each of these addresses was checked (based on collected address and crossroads infor-mation) on a street map to validate its location.

Of these 371 addresses, 7 were outside the EPZ.

Therefore, data were tabulated on the 364 respondent households that were located within the EPZ.

Respondents at 77 of these households had been away from home at the time of the demonstration of the alert and notification system and, therefore, were not included in the alerting analy-sis.

Of the remaining 287 households, 77.3% (222) in-dicated that they had been alerted during the demonstra-tion.

If one uses the estimated number of households 28 I

FIGURE 4 3255Q hiltra Research Services Study #6576 sdatr, Pennsylvania August 6, 1986 OMB #3067-0103 (FEMA 9/86)

FEMA NUCLEAR POWER PIAhT ALERTING AND NOTIFICATION SYSTEM:

PUBLIC TELEPHONE SURVEY Fort St. Vrain Tina Began AM PM Interview #

(1-5)

Tina Ended AM PM Zip Code (e-10)

Sample Type (11)

RECORD BEFORE DIALING -Telephone #

(Area Code)

(Exchange)

(Number)

(12-21)

TRODUCTION:

Hallo, my name is We're calling households long distance m Chilton Research Services as part of a survey. This survey is sponsored by The d::n1 Energency Management Agency (FEMA) of the United States Gcvernment.

Your answers are voluntary and will be kept strictly confidential.

First of all, is this (REPEAT # DIALED)?

Yes 1

TERMINATE AND DIAL AGAIN No 2

As you =ay or =ay not know, there was a test of the public warning / alert notification systs: for the Fort St. Vrain Nuclear Power Oenerating Station.

Did you, or any oths nember of this household, hear any type of emergency warning / alert signal from this test today?

22-CONTINUE Yes 1

SKIP TO Q. 4A No 2

Hearn fro:

CONTINUE another 3

source ASK IF ANY OTHER HCUSEh0LD Don't Know 8

MEMBER IS MCRE KNOWLEDGEABLE l

29 l

FIGURE 4 (CONTINUED)

ITnte type of emergency test signal did you hear?

(D0 NOT READ.

CIRCI.E ALL THAT APPLY)

(23-25)

Siren 1

SKIP TO Tone alert radio 2

Q. 4 Neighbor told me 3

Other family member told me 4

Other (SPECIFY) 9 CONTINUE Don't Know Y

Did you hear.

. -(RFAD LIST.

CIRCLE AL' THA* APPLY)

(30-32)

A siren i

A cone alert :sdio 2

From a neighbo:

3 From another fa=11y member 4

Or by means of so=ething else 9

(SPECI?Y)

DO NOT READ Don't Know Y

(I? " HEARD D'.ERGENCY SIGNAL" ASK Q. 4 3ELOW; O!FERWISE' SKIP TO Q. 4A)

Wcro you at home or away from home when you hen:d or were made aware of this c=argency test signal?

37-l Home 1

l SKIP To Q. 5 Away fro:

home 30

FIGURE 4 (CONTINUED)

(!? "DID NOT EE.AR EMERGENCY SIGNAL") Were you at home around 5:30 this evening?

38-Yes 1

l2 No Don'tKnowlY Eas your household ever been issued a Tone Alert Radio?

y 39-IF RESPONDENT WAS NOT HOME SKIP To Q. 5, OTHERVISE ASK Q. 4C Yes 1

No 2

SKIP TO Q. 5 Don't Know Y

> Uns it turned on at approximately 5:30 this evening?

40-Yes l1 No 2

Don't Know LY Has your household ever received infor=ation which tells you what to do in the event of a "real" emergency at Fort St. Vrain? These brown brochures in both English and Sp:nish were sent. out by the Public Service Company of Colorado in February and were titled, "Infor:ation About the it. S t. Vrain Radiological Emergency Response Plan".

Do you re=e:ber receiving this information?

41-Yes 1

No 2

Don't Know Y

32cause we seed to determine whether or not you live within the 5 cile Emergency Planning Zone of the Fort St. Vrain Nuclear Generating Station, would you please give

=s this address?

(FAUSE FOR ANSWER)

ADDRESS:

end the nearest intersection (or cross street) to this house.

Also, what co=sunity is this?

i l

, on bahalf of Chilton Research Setvices and the Federal Emergency Manage =ent Agency, I i

would like to thank you for your time and for giving us this valuable infor ation.

31

FIGURE 5

1. 3258Q Chilton Researen Services Study d6576 Radnor, Pennsylvania Augus: 6, 1986 OM3 #3067-0103 (FEMA 9/86)

FEMA NUCLEAR POWER PIANT ALERTING AND NOTIFICATION SYSTEM:

PUBLIC TELEPH0!TE SURVEY FORT ST. VRAIN - Spanish Version Hora al co=enza AM PM

1. de entrevista Hora el termina:

AM PM Zip Code Sample Type ESCRIBA ANTES DE LLAMAR - # de telefono, (Area Code)

(Ezenange)

(Nunber) ga INTRODUCCION:

Buenas (tardes/ dias), si nonbre es Esta=os lla=ando de larga distancia desde Chilton Research Services, como par:e de una encuesta, patrocinada por la Agencia Federal del Manejo de E=ergencias (FEDERAL EMERGENCY MANAGEMENT AGENCY) del Gobieno de los Estados Unidos.

Sus respuestas son voluntarias y se =antendran en forna confidencial.

1.

Pri=ero que todo, es este el utbero (REPITA EL # DE TELEFONO)?

Si 1

TE?.MINE Y VUELVA A LLAMAR No 2

Puede que Ud. este enterado que se llevo a cabo una prueba de el aler a del sistema puolico de emergencias para la planta generadora nuclear Fort St. Vrain.

Escucho' hoy Ud.,

o algun o::o =ie= bro de su familia, una senal de alar =a en la radio que avisa las condiciones a:nosfe'ricas?

CONTINUE Si 1

SKIP TO Q. 4A No 2

CONTINUE 0:ro =edio 3

ASK IF OTHE H'd MEM3 a MORE KNOWLEDGF_'diLE No Se Y

32

FIGURE 5 (CONTINUED)

. Que clase de senal de emerseccia oyo?

(NO LEA.

CIRCULE TODAS LAS RESPUESTAS QUE CO.4TESTEN)

Sirena 1

Tono alereta en radio 2

p,,,

Vecinos avisaron 3

Mie= bro de la facilia =e aviso 4 I p, 4 Otra respuesta 9 i CONTINUE No sabe Y

> Oyo usted.

(LEA IA LISTA.

CIRCULE TODAS LAS RESFUESTAS)

Una sirena 1 i Tono aler:a en radi6 2

De un veciso 3

De algun =ie= bro de fa=111a 4

Por o::o medio (ESPECIFIQUE) 0 NO LEER No sabe Y

33

FIGURE 5 (CONTINUED)

(SI CONTESTM "ESCUCHE LA SE'fAL DE EMERGENCIA", PREGi1NTE M P.4 A3AJO, SI NO, ?ASE A LA ?.4A).

,, Es:aba Ud. en casa o no cuando escucho' la setal de alama o cuando le avisaren?

Si 1i PASE A ?. 5 No 2 I i

4. (SI CONTESTA "_NO ESCUCHE LA SENAL DE Dd.ERGENCIA"):

Es:aba Ud. noy en casa a las 5:30 esta ta:de?

Si 1

No 2

No se Y

ii

l. Le han entregado a ud. un radio de cono de ala=a?

SI NO ESTA3A EN CASA PASE A P. 3, g *,

~

DE OTRA MANERA PREGUNTE ?. 4C SKI? TO Q. 5 No 2

No se Y

s

. Es:aba encendida la ala=a a las 5:30 esta tarde?

Si 2

No 2

No se Y

i i

34

within the EPZ (which is 793 according to the analysis of sample addresses) in the confidence interval expres-sion in Appendix A, it yields an estimated 95% confi-dence interval for the proportion of the total EPZ population alerted that ranges f rom 73.2% to 80.9%.

In other words, at a 95% confidence level, between 73.2%

and 80.9% of the households within the EPZ would have stated that they were alerted by the early warning system.

Respondents at the households that reported that they were not alerted were asked whether their tone / weather alert radios were turned on and set such that an alert-ing signal could be received.

Respondents at 29 house-holds indicated that their tone / weather alert radios were not being operated in a manner that permitted re-ceipt of an alerting signal.

Thus, 86.0% of the house-holds whose residents were home at the time of the demonstration and that had tone / weather alert radios operating in a manner that permitted receipt of an alerting signal reported that they were alerted.

Using the estimated number of households within the EPZ and the confidence interval expression in Appendix A, a

95% confidence interval ranging from 82.6% to 88.9% is obtained for the proportion of the total EPZ population with operating tone / weather alert radios that would have stated that they were alerted.

The sample of 364 households was also used to estimata the proportien of households within the EPZ that stated they received information about what to do in a real emergency at the Fort St. Vrain Nuclear Generating Sta-tion.

Ci these 364 households, 79.9% (291) responded that they had received this inf ormation.16.7% (61) 35

responded that they had not received this information, and 3.3% (12) did not know whether they had received this information, tising the approach discussed pre-viously, the following estimates for the entire EPZ population resulted (at the 95% confidence level) :

. Between 77.2% and 82.3% of the households would have reported receiving the information; Petween 14.6% and 19.2% of the households would have responded that they had not received the information; and Between 2.3% and 4.6% of the households would not have known whether they had received the infor-mation.

In conclusion, no areas of the Fort St. Vrain siren and tone / weather alert radio system were identified as need-ing enhancements.

36

IV.

FINDINGS FOR EVALUATION CRITERIA E.5, F.1, N.2, N.3, AND N.5 Those aspects of the alert and notification system address-ing evaluation criteria E.5, F.1, N.2, N.3, and N.5 of NUREG-0654/ FEMA-REP-1, Rev.1, have been reviewed by FEMA and found to be adequate to provide reasonable assurance that appropriate protective measures can be taken off site in the event of a radiological emergency.

This conclusion is documented in letters to the Honorable Richard D.

Lamm, Governor of Colorado, signed by Lee M.

Thomas, Associate Director, State and Local Programs and Support, FEMA, dated January 29, 1982;4 and William J.

Dircks, Executive Dir-ector f or Operations, U.S.

Nuclear Regulatory Commission, signed by Lee M.

Thomas, Associate Director, State and Local Programs and Support, FEMA, dated January 29, 1982.9 In these letters, the Fort St. Vrain Nuclear Generating Station received FEMA approval under 44 CFR 350, conditioned on an ultimate approval and verification of the public alert and notification system as called for in NUREG-0654/ FEMA-REP-1, Rev. 1.

In additon, a review was completed by FEMA Region VIII regarding those elements applicable to alert and notification, as called for in FEMA-43.

37

REFERENCE LIST 1.

Public Service Company of Colorado.

"Public Service Comp-any of Colorado, Fort St. Vrain final saf ety analysis report."

(No date).

2.

Nuclear Regulatory Commission and Federal Emergency Man-agement Agency.

1980.

" Criteria for preparation and evaluation of radiological emergency response plans and preparedness in support of nuclear power plants. "

NUREG-0 6 54/ FEMA-REP-1.

Fevision 1.

November 1980.

3.

Federal Emergency Management Agency.

1984.

Letter from Alton D.

Cook, Regional Director, to Richard Krimm, Assist-ant Associate Director.

Subj ect :

Documents to be suppor-tive of the public alerting and notification system evalua-tion.

April 2, 1984.

4.

Federal Emergency Management Agency.

1982.

Letter to the Honorable Richard D.

Lamm, Governor of Colorado, signed by Lee M.

Thomas, Associate Director, State and Local Programs and Support.

January 29, 1982.

5.

Federal Emergency Management Agency.

1983.

" Standard guide for the evaluation of alert and notification systems f or nuclear power plants. "

FEMA-43.

September 1983.

6.

International Energy Associates Limited.

1983.

" Analysis of tone alert pilot test."

IEAL-321.

September 27, 1983.

7.

Public Service Company of Colorado.

1985.

Letter from O.

R.

Lee, Vice President, to Alton D.

Cook, Federal Emergency Management Agency.

Subject:

Radiological alarm in Platteville, Colorado.

August 16, 1985, 8.

Federal Emergency Management Agency.

1986.

Memorandum from Alton D.

Cook, Regional Director, to Robert S.

Wilker. con, Chief, Technological Hazards Division.

Subject:

Siren System in Platteville, Colorado.

May 2, 1986.

9.

Federal Emergency Management Agency.

1982.

Letter to William J.

Dircks, Executive Director for Operations, U. S.

Nuclear Regulatory Commission, signed by Lee M.

Thomas, Associate Director, State and Local Programs and Support.

January 29, 1982.

38

APPENDIX A Sample Size Determination

APPENDIX A SAMPLE SIZE DETERMINATION The number of hobseholds that need to be surveyed is determined based upon the need to obtain a sample size sufficient to obtain a 95% confidence interval with precision (half-width) of 0.05.

for the estimate of the proportion alerted.

The exact number of households to be surveyed can be derived f rom the following statistical considerations.

For relatively large sample sizes (n 2 30), taken without replacement from a population (N), the 4

sampling distribution for proportions (e.g., the proportion of the population alerted) is nearly a normal distribution, the mean of which is the proportion (p) of the population alerted

[

and the variance of which is

"~"

p(1 p)/n If P fs the observed sample proportion, then for a particular confidence level with confidence coefficient Zcr (P - p)2 52 p(1 _ p)fn j-_

2 Thus, for this confidence level, the actual prop 6rtion of the population alerted satisfies the following inequalities:

,2 2

N - n\\

,"O N-n\\~,

p, 'c P(1 - P)

N-n 2n N - 1/

N - 1/

2 N - If c

n

$ p Md

~~

N-n\\

~c N - 1/

n 1

2 2

N-n\\

c N-n P(1 - P) IN-n c

p 2n N-1 c

n 2

(N - lj N - if PS 2

Z

/N-n e

n N-1 Thus, the precision (W) is simply given by N-n\\

P(1 - P) c N-n N - 1/

2 N - If Z

n w-2 IN-n\\

c (N - If n

E This equation can be solved to determine the sample size (n) required to yield a given precision (W) with a given observed sample proportion (P) as follows:

2 P(1 - ?) - 2W + W l - 4P (1 - P)

+P (1 - P)

~

~

~

n=

~

Z 2(1 - P)~

e 1+

P(1 - P) - 2W~

l+

}

+

W~

l - 4P(1 - P)

+P 2W'N Z

C Although this expression for n can be used directly, it is customary to make several approximations.

First, since the term in N in the denominator (the finite population term) is positive definite for all reasonable values of W (0 < W < 0.5), omitting this term will result in an approximation to n that is slightly larger than its true value.

This is an acceptable practice in sizing the sample since a larger sample gives greater precision.

2

A second approximation that can be made is to neglect the terms in W2 within the bracket in the numerator.

Analysis demonstrates that this underestimates n when P < 1/2 - 1/4 $[2 + 8W 2

or P > 1/2 + 1/4 T[2 + 8W2 and overestimates n for P between those two values.

For the case of interest (a 95% confidence interval with precision of 0.05), this approximation provides an overestimation of n when a sample size greater than 191 is required.

Since the sampling plan calls for a minimum sample size of 250, regardless of the value of P, this approximation is acceptable because it also yields an estimate of n larger than the true value.

Therefore, for the purposes of the pilot test and subsequent surveys, the following approximate equation can be used to determine whether a sample size larger than 250 is required:

2 z

n=

P(1 - P) w~

or using 1.96 for Zc and 0.05 for W, n = 1536.64 P(1 - P)

Data from the pilot test can be used to illustrate the effects of these approximations.

In the pilot test, the population of tone alert households from which the sample was to be drawn (N) was approximately 4500 and the observed proportion alerted (P) was 0.675.

This yields 311 as the exact result for n.

Neglecting the finite population term yields an estimate of 334 for n, and the simplified final approximation estimates n as 338.

Thus, the final simplified approxin.ation overestimates the required sample size by 27 in this case.

SOURCE: International Energy Associates Limited.

" Analysis of Tone Alert Pilot Test."

IEAL-321.

September 27, 1983.

3

_